article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the quest to explore gravity theory beyond einstein s general theory relativity ( gtr ) on large cosmological scales , in recent years , has been aroused by the prospect of explaining the late time acceleration of the universe .
it has now been established beyond any doubt that our universe is currently going through an accelerated expanding phase which has been started in recent past @xcite . the theoretical approach to explain such an acceleration
can be broadly divided into two categories . in one approach , one has to add some exotic component with negative pressure ( known as dark energy ) in the energy budget of the universe as normal matter or radiation component can not initiate accelerated expansion .
simplest of such component , known as cosmological constant ( with equation of state @xmath5 ) can explain all the available cosmological data but at the same time is plagued by the embarrassing problem of fine tuning and cosmic coincidence @xcite .
one can also consider more exotic scalar fields @xcite to mimic the required negative pressure which may solve at least the cosmic coincidence problem and can have far more striking cosmological consequences . unfortunately the presently available cosmological data can not distinguish these two models decisively .
the other approach is to give up the idea of adding extra component in energy budget of the universe but to modify the gravity theory on large cosmological scales .
the idea is to look for departure from the gtr that can effectively mimic a cosmological constant and result the necessary acceleration .
@xmath0 gravity theory @xcite is one such example where instead of linear dependence of the einstein - hilbert ( eh ) action on the ricci scalar , one includes nonlinear dependence .
it is particularly interesting and has been studied extensively . recently
a generalization of such theory has been proposed which involves a non - minimal coupling between curvature and matter @xcite . in this model , one extends the presence of non - linear functions of the scalar curvature in the eh action by incorporating an additional term involving the coupling between the ricci scalar and the matter .
astrophysical and cosmological signatures of such non - minimally coupled @xmath0 gravity models have been studied by various authors @xcite ( see also @xcite for similar investigation ) .
however describing background cosmological evolution for any particular model is not sufficient to remove the observed degeneracies between different modified gravity models and @xmath1cdm .
growth of matter perturbation is extremely important in this regard as the evolution equations for growth of matter perturbation are completely different in two scenarios .
this can result characteristic signatures in cmb as well as in matter power spectrum from galaxy clusters which in turn can help to remove the degeneracies between the two models . in a recent paper @xcite the evolution of cosmological perturbations in the presence of non -
minimal coupling between curvature and matter has been studied . in this work ,
we extend the work done in @xcite to the observational front .
we study how far the present cluster data can distinguish the model with non - minimal coupling between curvature and matter from the standard @xmath1cdm model .
for this we proceed in a two stage manner .
we first look for the possible form for the @xmath0 that can give rise to the same background evolution as in @xmath1cdm universe . after reconstructing the suitable form for the @xmath0
, we calculate the matter power spectrum for such a model and see whether the currently available observational data can distinguish this model from the @xmath1cdm .
the action for non minimally coupled @xmath0 gravity models is given by @xmath6 where @xmath7 , @xmath0 is a dimensionless function of the scalar curvature @xmath8 and @xmath9 is the matter lagrangian . due to the presence of coupling between the curvature and the matter
, there will be an energy transfer between them which can be seen in the corresponding conservation equation given by @xmath10 the non vanishing term in the r.h.s of the above equation shows the energy transfer between curvature and matter .
this exchange of energy between the curvature and matter is a special feature in non minimally coupled @xmath0 gravity models .
but in a homogenous and isotropic universe , the energy momentum tensor for the matter should have a perfect fluid form and once we assume the form of the matter lagrangian as @xmath11 , @xcite(@xmath12 being the energy density for the matter ) , it is easy to show that @xmath13 is satisfied for the background evolution . varying the action ( [ action ] ) with respect to the metric tensor @xmath14 , one obtains the einstein s equation as @xmath15 where @xmath16 . assuming a spatially flat friedmann - robertson - walker spacetime with scale factor a(t ) @xmath17 the time component of above equation is @xmath18 our first goal is to look for those models that mimic the background evolution for the @xmath1cdm universe .
we know the form of the hubble parameter for the @xmath1cdm universe which is given by @xmath19 where @xmath20 , @xmath21 and @xmath22 are the hubble parameter , matter density parameter and density parameter for cosmological constant at present respectively . for a spatially flat universe , @xmath23 .
with such a background evolution , the equation ( [ zero ] ) takes the form @xmath24 here @xmath25 represents the first derivative of @xmath26 with respect to the scale factor .
one has to solve this equation to find the form for @xmath26 which can mimic the similar background evolution as in @xmath1cdm universe .
equation ( [ recons ] ) is a second order differential equation and needs two initial conditions , @xmath27 and @xmath28 , to solve .
we solve the equation from the era of decoupling ( @xmath29 ) till present day ( @xmath30 ) .
we observe that the behavior does not vary with @xmath28 but is very sensitive to @xmath27 . in figure 1
, we show the different behaviors for @xmath31 and @xmath32 with varying @xmath27 that give rise to the same background evolution as in @xmath1cdm model .
given these forms for @xmath0 that produce identical background evolution as in @xmath1cdm universe , the next question is how different is the growth history of the universe for these non - minimally coupled @xmath0 gravity models from the concordance @xmath1cdm model ? in the next section , we try to address this issue .
perturbed frw metric in longitudinal gauge takes the form : @xmath33 where @xmath34 and @xmath35 are the two gravitational potentials . using equations ( [ eins ] ) and ( [ perturb ] )
, one can obtain the equations for growth in the linear regime : @xmath36 @xmath37\delta \chi=\nonumber \\
\dot{\chi}(3h\phi+3\dot{\psi}+\dot{\phi})+(2\ddot{\chi}+3h\dot{\chi})\phi+\frac{\kappa^2}{3}(1+f)\delta\rho_m , \label{tracecomp}\end{aligned}\ ] ] @xmath38 here @xmath39 and @xmath40 .
next we assume the velocity perturbation as @xmath41 in the frw universe , @xmath42 . also with @xmath43 ,
one can write @xmath44 . the spatial part of the @xmath45 is the peculiar velocity .
writing the spatial part @xmath46 as gradient of a scalar : @xmath47 one can now get the following equations on perturbing equation ( 2 ) : @xmath48 @xmath49 where @xmath50 . using these two equations ,
one finally arrives at @xmath51 equation ( 16 ) governs the cosmological evolution of the growth of matter over density in a non minimally coupled @xmath0 gravity model which has identical background evolution as in the @xmath1cdm model .
we are interested in the perturbations which are deep inside the hubble radius for which @xmath52 .
we apply the following approximations to calculate the perturbations deep inside the hubble radius @xcite : @xmath53 this actually implies @xmath54 , where @xmath55 + under this approximation , equations ( 9 ) , ( 11 ) and ( 10 ) take the form @xmath56 from the corresponding @xmath1cdm value for different @xmath27 .
the solid , dashed , dot - dashed and dotted lines are for @xmath57 respectively .
@xmath58.,width=302 ] @xmath59 @xmath60\delta \chi=\frac{\kappa^2}{3}(1+f)\delta \rho_m \label{nwtrace}.\ ] ] using equations ( 18 ) , ( 19 ) and ( 20 ) , and under the approximation ( 17 ) , equation ( 16 ) finally becomes @xmath61 where @xmath62}{(\frac{\beta f_{,x}-1}{\beta f_{,x , x}})(1+\frac{3\omega_{0m}}{\alpha a^3})f_{,x}+x(\beta f_{,x}-1)-\frac{3}{\alpha a^2}\frac{k^2}{a^2}(\beta f_{,x}-1)}\ ] ] where @xmath63 and @xmath64 . to solve the equation ( 21 )
, we use the initial conditions @xmath65 and @xmath66 where we choose @xmath67 , which is the decoupling era .
this is to ensure that the growth evolution reproduces the matter - like behavior during the early time .
growth of matter overdensity is an important independent probe for different cosmological models that explain the late time acceleration of the universe .
the growth rate for matter over density is defined as @xmath68 in figure 2 , we show the percentage deviation of the growth factor @xmath69 for the non - minimally coupled @xmath0 gravity model from the corresponding @xmath1cdm model .
one can see that there can be four to six percent deviation one may expect for different values of @xmath27 .
the question is whether such deviation can be probed by the current observational data .
another important quantity related to the growth is the rms fluctuation of the linear density field at the scale @xmath70mpc , @xmath71 .
this defined as @xmath72 where @xmath73 is the window function defined as @xmath74 the quantity @xmath75 as defined in @xcite @xmath76 where @xmath77 is the normalized density contrast defined as @xmath78 , @xmath79 is the amplitude of the primordial curvature perturbation produced during inflation , @xmath80 = 0.05 @xmath81 is the pivot scale at which the primordial fluctuation is calculated .
@xmath82 is the transfer function which incorporates the effects of evolution of perturbations through horizon crossing and matter / radiation transition .
we use analytical form for @xmath82 as proposed in @xcite : @xmath83^{-0.25}\end{aligned}\ ] ] where @xmath84 and @xmath85 .
galaxy surveys are directly sensitive to the combination @xmath86 .
this combination is almost a model independent estimator for the observed growth history of the universe and that is why most of the surveys e.g. the 2df , vvds , sdss , 6df , boss , as well as the wiggle - z galaxy survey provide the measurement for this estimator . in a recent paper @xcite compilation of the current measurements for @xmath86 has been given and we use those measurements for our purpose . using these measurements , we calculate the posterior for our case , which is given by @xmath87 where @xmath88 corresponds to parameters in our theory .
@xmath89 is the prior probability distribution for different parameters . in our calculations
, we fix the value of the spectral index @xmath90 which is the best fit value obtained by the planck @xcite .
this is because , posterior is not very sensitive to the value of @xmath91 .
we use gaussian prior for the parameters @xmath92 and @xmath93 , with their valuse as @xmath94 and @xmath95 from recent measurement by planck @xcite . for the model parameter @xmath96
, we use a flat prior between @xmath97 to @xmath98 .
the likelihood function is defined as @xmath99 with @xmath100 covariance matrices for @xmath0 with non minimal coupling and for @xmath101 are constructed from the posterior . for non minimally coupled f(r ) with parameters as @xmath102 , @xmath103 and @xmath79 it becomes + @xmath104 covariance matrix for @xmath101 with parameters as @xmath102 and @xmath79 becomes @xmath105 using these covariance matrices of non minimally coupled @xmath0 as well as for @xmath1cdm , we have reconstructed evolution of the growth factor @xmath2 and @xmath71 at @xmath106 and @xmath107 confidence level . this has been shown in figure 3 .
one can see clearly that it is possible to distinguish the @xmath1cdm and the non - minimal @xmath0 model through the @xmath2 evolution both at @xmath106 and @xmath107 confidence level , whereas for the @xmath71 evolution , one can distinguish them at @xmath107 confidence level .
we should stress the fact that two models have identical background evolution and can not be distinguished by any observation related to background cosmology .
explaining late time acceleration is one of the most significant challenges for cosmologists today . modifying the einstein gravity at large cosmological scales like @xmath0 gravity models , is one interesting approach to explain such late time acceleration .
coupling curvature with matter in @xmath0 gravity has been recently proposed which has the advantage , that the singularity that occurs in the standard @xmath0 gravity models may not occur in non minimally coupled @xmath0 .
the background cosmology in such non - minimally coupled @xmath0 gravity has been studied .
recently its implication on inhomogeneous universe has also been explored in @xcite .
our goal is to extend further the work done in @xcite to investigate the possibility to distinguish such a non minimally coupled @xmath0 gravity model from the concordance @xmath1cdm model using the current measurements by different galaxy surveys . for this we do not assume any particular form for the function @xmath0 .
we take a different approach in this investigation .
the growth of matter overdensity in any modified gravity model depends both on the background expansion as well as the extra effects that arise solely due to the modification in the gravity action ( in this case it is the non minimal coupling between the curvature and matter ) . to study the effect of this non minimal coupling on the growth history and the subsequent deviation from the @xmath1cdm behavior , we fix the background evolution same as the that of the @xmath1cdm model thereby eliminating the effect due to the background evolution .
now departure from the @xmath1cdm evolution is solely due to the modified equation for the growth evolution in the non - minimally coupled case .
this is one crucial difference in approach from the investigation done in @xcite where a specific form for @xmath0 was assumed .
subsequently we calculate the allowed behaviors for @xmath0 modification , that can give rise to the same background cosmology as in @xmath1cdm universe . reconstructing this form
, we subsequently study the growth history and the @xmath108 normalization for such @xmath0 forms and compare them with @xmath1cdm universe .
we use the measurements for @xmath3 ( which is mostly model independent ) from different galaxy surveys to reconstruct the form for growth factor @xmath2 and @xmath4 for both the non minimally coupled @xmath0 model as well as for @xmath1cdm universe .
the results show that it is possible to distinguish these two models using the growth of matter over density , even if they have same background evolutions .
this result is the important extension of the work done in @xcite to observationally distinguish the non minimally coupled @xmath0 gravity model from the concordance @xmath1cdm model . we should mention that we have considered the simple form for the action as given in equation ( 1 )
one can consider more general form like @xmath109 . in that case , if one considers the same background evolution as in @xmath1cdm , there will be effects on the growth history from the modified part of the pure gravity action ( @xmath110 ) as well as the effect due to the non minimal coupling between the matter and the curvature .
one can then address two questions .
firstly whether such model can be distinguished from the concordance @xmath1cdm using the current observational data . and also given a specific form for the pure gravity lagrangian @xmath110 , whether the non minimal @xmath0 model can be distinguished from the standard minimal @xmath0 model using the observational data .
we shall address these issues in future .
the authors thank t.r . seshadri for useful discussions and comments .
st acknowledges the computational facilities provided by the ctp , jmi , new delhi , india .
st thank c.s.i.r , govt . of india for financial support through senior research fellowship .
aas . acknowledges the partial financial support provided by c.s.i.r .
govt . of india through the research grant ( grant no:03(1187)/11/emr - ii ) .
99 s.perlmutter _ et al .
_ , astrophys .
j.*517 * ( 1999 )
565 [ arxiv : astro - ph/9812133 ] .a . g. riess _
_ , astron.j.*116 * ( 1998 ) 1009 [ arxiv : astro - ph/9805201 ] d.n.spergel _ et al_.[wmap collaboration ] , astrophys.j.suppl . * 170 * ( 2007 ) 377 [ arxiv : astro - ph/0603449 ] ; u.seljak _ et al _ [ sdss collboration],phys.rev .
d * 71 * ( 2005 ) 103515 [ arxiv : astro - ph/0407372 ] ; d.j.eisenstein _ et al_.[sdss collboration ] , astrophys.j.*633 * ( 2005 ) 560 [ arxiv : astro - ph/0501171 ] ; b.jain and a.taylor , phys.rev lett .
* 91 * ( 2003 ) 141302 [ arxiv : astro - ph/0306046 ] e.j.copeland , m.sami and s.tsujikawa,int.j.mod.phys.d * 15 * , 1753 ( 2006 ) : m.sami , arxiv:0904.3445 ; v.sahni and a.a.starobinsky , int.j.mod.phys.d * 9 * , 373 ( 2000 ) ; t.padmanabhan,phys.rep . *
380 * , 235 ( 2003);e.v.linder , asrto - ph/0704.2064 ; j.frieman , m.turner and huterer , arxiv:0803.0982 ; r.caldwell and m.kamionkowski , arxiv:0903.0866 ; a.silvestri and m.trodden , arxiv:0904.0024 b. ratra and p.j.e .
peebles , phys .
d * 37 * 3406 ( 1988 ) ; r.r .
caldwell , r. dave and p.j .
steinhardt , phys .
lett . * 80 * 1582 ( 1998 ) ; a.r .
liddle and r.j .
scherrer , phys .
d * 59 * ( 1999 ) 023509 ; p.j .
steinhardt , l. wang and i. zlatev , phys .
d * 59 * ( 1999 ) 123504 c.armendariz-picon , t.damour , and v.mukhanov , phys.lett.b * 458 * , 209 ( 1999 ) j.garriga and v.f.mukhanov , phys.lett.b * 458 * , 219 ( 1999 ) t.chiba , t.okabe , m.yamaguchi , phys.rev.d * 62*,023511 c.armendariz-picon , v.mukhnov , and p.j.steinhardt , phys.rev.lett * 85 * , 4438(2000 ) c.armendariz-picon , v.mukhnov , and p.j.steinhardt , phys.rev.d * 63*,103510 ( 2001 ) t.chiba , phys.rev.d * 66 * , 063514 ( 2002 ) l.p.chimento and a.feinstein , mod.phys.lett.a * 19 * , 761 ( 2004 ) l.p.chimento , phys.rev.d * 69*,123517 ( 2004 ) r.j.scherrer , phys.rev.lett . *
93 * , 011301 ( 2004 ) a.y.kamenshchik , u.moschella , and v.pasquier , phys.lett.b * 511 * , 265 ( 2001 ) n.bilic , g.b.tupper , and r.d.viollier , phys.lett.b * 535 * , 17 ( 2002 ) m.c.bento , o.bertolami , and a.a.sen , phys.rev.d * 66 * , 043507 ( 2002 ) a.dev , j.s.alcaniz , and d.jain , phys.rev.d * 67 * , 023515(2003 ) v.gorini , a.kamenshchik and u.moschella , phys.rev.d * 67 * , 063509 ( 2003 ) r.bean and o.dore , phys.rev.d * 68*,23515 ( 2003 ) t.multamaki , m.manera and e.gaztanaga , phys.rev.d bf 69 , 023004 ( 2004 ) a.a , sen and r.j.scherrer , phys.rev.d * 72 * , 063511 ( 2005 ) t.p.sotiriou and v.faraoni , arxiv:0805.1726[gr - qc ] ; s.nojiri and s.d.odinstov , arxiv:0807.0685 ; l. amendola , r. gannouji , d. polarski and s. tsujikawa , phys .
d * 75 * 083504 , ( 2007 ) ; w. hu and i. sawicki , phys .
d * 76 * 064004 , ( 2007 ) ; a. a. starobinsky , jetp lett . * 86 * 157 , ( 2007 ) ; s. tsujikawa , phys .
d * 77 * 023507 , ( 2008 ) ; g. j. olmo , phys .
d * 72 * 083505,(2005 ) ; a. l. erickcek , t. l. smith and m. kamionkowski , phys .
d * 74 * 121501,(2006 ) ; v. faraoni , phys .
d * 74 * 023529 , ( 2006 ) ; t. chiba , t. l. smith and a. l. erickcek , phys .
d * 75 * 124014,(2007 ) ; p. brax , c. van de bruck , a. c. davis and d. j. shaw , phys .
d * 78 * 104021,(2008 ) ; i. thongkool , m. sami , r. gannouji and s. jhingan , phys .
d * 80 * 043523 , ( 2009 ) ; r. gannouji and d. polarski , jcap * 0805 * 018 , ( 2008 ) ; s nojiri , s d odintsov and d saez - gomez , phys .
lett . * b 681 * , 74,(2009 ) ; g cognola , e elizalde , s d odintsov , p.tretyakov and s zerbini , phys.rev.d * 79 * , 044001 , ( 2009 ) ; s nojiri and s d odintsov , arxiv:0706.1378 ; f. g. alvarenga et al . , arxiv:1302.1866 [ astro-ph.co ] ; a. abebe , a. de la cruz - dombriz and p. k. s. dunsby , arxiv:1304.3462 [ astro-ph.co ] ; a. abebe , m. abdelwahab , a. de la cruz - dombriz and p. k. s. dunsby , class . quant . grav . , * 29 * , 135011 ( 2012 ) ; a. de la cruz - dombriz , a. dobado and a. l.maroto , phys . rev .
d , * 77 * , 123515 ( 2008 ) .
o.bertolami and j.paramos arxiv:0805.1241v2[gr - qc ] ; o.bertolami , f. lobo and j. paramos , phys .
d , * 78 * , 064036 , ( 2008 ) ; o. bertolami , j. paramos , t. harko and f. lobo , arxiv:0811.2876 ; o. bertoami and m.c.sequeira , phys .
rev.d , * 79 * , 104010 , ( 2009 ) ; o. bertoami , c.g.boehmer , t.harko and f.lobo phys .
d , * 75 * , 104016 , ( 2007 ) ; v. faraoni , phys.rev . d. * 80 * , 124040 ( 2009 ) ; s. nojiri and s.d .
odintsov , phys.lett.b * 599 * , 137 , ( 2004 ) ; g. allemandi , a.borowiec , m.francaviglia , s.d.odintsov , phys.rev.d * 72 * 063505,(2005 ) ; s. nojiri , s.d . odintsov and p.tretyakov , arxiv:0710.5232 ; o. bertolami and j. paramos , arxiv:0906.4757 ; o. bertolami , p.frazao and j. paramos , arxiv:1003.0850 t. koivisto , class .
grav . , * 23 * , 4289 ( 2006 ) .
j.d.brown , class quantum grav .
* 10 * 1579 ( 1993 ) o.bertolami , pedro frazao and jorge paramos , jcap , * 1305 * , 029 ( 2003 ) .
s. tsujikawa , phys . rev .d * 76 * , 023514 , ( 2007 ) ; s. tsujikawa , r. gannouji , b. moraes , d. polarski , phys.rev .
d * 80 * , 084044 , ( 2009 ) d. huterer , arxiv:1010.1162 [ astro-ph.co ] .
m. bardeen , j.r .
bond , n. kaiser , k.s .
szalay , astrophysical journal * 304 * , 15(1986 ) .
s. basilakos , s. nesseris and l. perivolaropoulos , arxiv:1302.6051 [ astro-ph.co ] .
planck collaboration(p .
a. r. ade et al . ) , arxiv:1303.5076 [ astro-ph.co ] | non - minimally coupled @xmath0 gravity model is an interesting approach to explain the late time acceleration of the universe without introducing any exotic matter component in the energy budget of the universe . but
distinguishing such model with the concordance @xmath1cdm model using present observational data is a serious challenge . in this paper
, we address this issue using the observations related to the growth of matter over density @xmath2 as measured by different galaxy surveys . as background cosmology is not sufficient to distinguish different dark energy models ,
we first find out the functional form for @xmath0 ( which is non - minimally coupled to the matter lagrangian ) that produces the similar background cosmology as in @xmath1cdm .
subsequently we calculate the growth for the matter over density @xmath2 for such non minimally coupled @xmath0 models and compare them with the @xmath1cdm universe .
we also use the measurements of @xmath3 by different galaxy surveys to reconstruct the behavior for @xmath2 and @xmath4 for both the non - minimally coupled @xmath0 gravity models as well as for the @xmath1cdm .
our results show that there is a small but finite window where one can distinguish the non - minimally coupled @xmath0 models with the concordance @xmath1cdm . |
when a fermion multiplicity ( flavor ) exceeds a critical value @xmath7 in non - abelian gauge theories , a quantum phase transition from a chiral broken phase to the conformal window is anticipated to take place . in the conformal window ,
strong coupling gauge theories are asymptotically - free in ultra violet ( uv ) region , while dominated by a non - trivial infrared fixed point ( irfp ) in infrared ( ir ) region .
the irfp - conformal dynamics can be probed via a response to a small fermion mass ( @xmath8 ) perturbation , which gives rise to a bound state with a mass gap ( @xmath9 ) satisfying the conformal hyperscaling relation with a mass anomalous dimension @xmath2 . in the chiral broken phase near to the conformal window @xmath10 , it is advocated that the system shows an approximate conformality ( walking ) with a `` would - be '' @xmath2 , which is of great interest in physics beyond the standard model @xcite . in recent lattice studies ,
the precise determination of @xmath2 in many flavor qcd is reported , while there exist tensions among the values @xcite .
the discrepancy presumably originates to scale violations by uv / ir cutoffs ( lattice spacing and finite box size ) as well as a probe fermion mass . in this proceedings ,
we focus on the ir cutoff effects by using schwinger - dyson ( sd ) equation .
we consider the sd equation for the fermion propagator @xmath11^{-1}$ ] within improved ladder approximation in the landau gauge . after integrating over angular degrees of freedom in a momentum space ,
the sd equation is expressed as @xmath12\alpha(p^{2}+q^{2})}{4\pi } \left [ \frac{q^{2}}{p^{2}}\theta_{p^{2 } - q^{2}}+\theta_{q^{2}-p^{2}}\right ] \frac{\sigma(q^{2})}{q^{2}+\sigma^{2}(q^{2 } ) } , \label{sdeq}\ ] ] where @xmath8 , @xmath13 , @xmath14 , and @xmath15 $ ] represent a bare fermion mass , step function , running coupling constant in the two - loop perturbation theory , and quadratic casimir operator , respectively .
the uv / ir cutoffs @xmath16 are introduced in the momentum space integral . for a given fermion mass @xmath8 and in the presence of @xmath16 , we numerically solve the sd equation ( [ sdeq ] ) and evaluate the physical pole mass @xmath0 and pion decay constant @xmath1 as @xmath17 we take account of full momentum dependences of the two - loop running coupling without recourse to the usual step - function type approximation for the coupling .
we then perform the hyperscaling fit analyses for the obtained numerical data similarly to analyses in lattice gauge theories .
moreover , we derive the sd - based finite - size hyperscaling formula ( fss ) , which allows us to handle the ir cutoff artifacts , even in the case of @xmath18 .
these are the advantages to the previous work @xcite . in the color su(@xmath19 )
gauge theory .
middle and right : pole masses @xmath0 as a function of a bare fermion mass @xmath20 for various ir cutoffs ( @xmath21 $ ] , blue - thin - solid lines ) with the uv cutoff @xmath22 . in the chiral broken phase : ( middle panel ) @xmath23 , ( right ) @xmath24.,title="fig:",width=192 ] in the color su(@xmath19 ) gauge theory .
middle and right : pole masses @xmath0 as a function of a bare fermion mass @xmath20 for various ir cutoffs ( @xmath21 $ ] , blue - thin - solid lines ) with the uv cutoff @xmath22 . in the chiral broken phase : ( middle panel ) @xmath23 , ( right ) @xmath24.,title="fig:",width=192 ] in the color su(@xmath19 ) gauge theory .
middle and right : pole masses @xmath0 as a function of a bare fermion mass @xmath20 for various ir cutoffs ( @xmath21 $ ] , blue - thin - solid lines ) with the uv cutoff @xmath22 . in the chiral broken phase : ( middle panel ) @xmath23 , ( right ) @xmath24.,title="fig:",width=192 ]
first , we investigate the chiral broken phase in the presence of ir cutoffs .
the middle and right panels of fig . [ fig : sdeq - sol ] show the pole mass @xmath0 against @xmath8 for @xmath25 and @xmath26 , respectively .
each blue - thin - solid line represents a result with a different ir cutoff @xmath27 . denotes the intrinsic scale which is renormalization invariant and associated with the onset of non - perturbative dynamics @xcite .
] for a small @xmath27 , the pole masses remain finite at small @xmath8 , indicating the spontaneous chiral symmetry breaking .
we find a critical value of the ir cutoff above which the dynamical fermion mass @xmath28 is strongly suppressed as shown in the figures : @xmath29 thus , a _ fake conformality _ results from a large @xmath3 . as a system becomes closer to the conformal window ( @xmath30 ) , the fake conformality appears for a smaller @xmath3 .
this gives a caveat for lattice works studying the border region of chiral broken and conformal phases .
next , we investigate a conformal phase with ir cutoffs by using the sd results for @xmath31 .
the mass anomalous dimension @xmath2 is obtained by the scaling property of @xmath0 ( @xmath1 ) against @xmath8 .
the left panel of fig .
[ fig : fss ] shows a logarithm plot of the pole mass @xmath0 as a function of fermion bare masses @xmath8 for various ir cutoffs @xmath5 . for the smallest ir cutoff @xmath32 ( red - squares ) ,
the data points align on a straight line , therefore , the hyperscaling relation @xmath33 is satisfied . for the fit range of @xmath34 $ ] ( non - shaded region in the figure ) , we obtain @xmath35 . while , for larger ir cutoffs , the data points are on a polygonal line with a bend around @xmath36 . as seen in the figure , the fit range @xmath37 $ ] starts affected by the bend down with increasing @xmath3 . as a result ,
the fit with the ansatz @xmath33 results in a @xmath2 strongly suppressed , as shown in the middle panel of fig .
[ fig : fss ] . when the suppression sets in ,
the @xmath2 evaluated by @xmath0 ( red - squares ) and @xmath1 ( blue - circles ) get impaired , and thus the universality of @xmath2 does not hold any more .
we shall now derive a sd - based fss which allows us to handle the ir cutoff artifacts explained above . to this end , we further approximate the sd equation ( [ sdeq ] ) with a linearized fermion propagator and @xmath38 where @xmath39 denotes the running coupling at irfp , and obtain the analytic expression of the renormalization factor @xmath40 .
the @xmath41 is a function of dimensionless variables , @xmath42 , @xmath43 , @xmath44 , and @xmath45 , and the analytic expression allows us to derive the sd - based fss formula @xcite : @xmath46 the equation ( [ eq : fss0 ] ) is same as what dictated by the renormalization group equation @xcite .
remarkably , the sd predicts another fss formula ( [ eq : fssir ] ) which is characterized by the scaling variable @xmath47 rather than @xmath48 for larger @xmath3 .
we apply the sd - based fss ( [ eq : fss0 ] ) and ( [ eq : fssir ] ) for the data obtained by solving eq .
( [ sdeq ] ) numerically .
right panel of fig .
[ fig : fss ] shows the results .
the various symbols ( colors ) show data with different ir cutoffs . for the case of @xmath49
, we adopt the ansatz ( [ eq : fss0 ] ) .
the fit works well ( blue - solid line ) and we obtain @xmath50 which is fairly consistent to @xmath35 obtained in the previous section for the smallest ir cutoff @xmath51 .
we find the alignment of the data points , indicating the universal nature of @xmath52 .
for the case of @xmath53 , we adopt the ansatz ( [ eq : fssir ] ) .
the fit works well ( black - solid line ) and gives @xmath54 , which is somewhat smaller than @xmath55 but the strong suppression has disappeared .
thus , the right fss formula in the right place recovers the approximate universality @xmath56 , or equivalently , the approximate data alignments in the whole mass region including both @xmath49 and @xmath53 .
then , two scaling variables , @xmath57 and @xmath58 , are responsible for two slopes of the alignments . in lattice studies , indeed , the formula ( [ eq : fss0 ] ) is widely used in lattice study , but the formula ( [ eq : fssir ] ) is rarely used @xcite . however , these result are of the case without paying attention to the scope of application , and opposed to each other @xcite .
vs the fermion mass @xmath59 for various ir cutoffs @xmath60 in @xmath31 .
middle : the mass anomalous dimension @xmath2 obtained by fitting the non - shaded region data of the left panel with the conformal ansatz .
right : the pole mass @xmath61 vs @xmath62 in the conformal phase @xmath31 .
the data upper and lower than the shaded region are used for the fits , giving the blue- and black - solid lines . , title="fig:",width=192 ] vs the fermion mass @xmath59 for various ir cutoffs @xmath60 in @xmath31 .
middle : the mass anomalous dimension @xmath2 obtained by fitting the non - shaded region data of the left panel with the conformal ansatz .
right : the pole mass @xmath61 vs @xmath62 in the conformal phase @xmath31 .
the data upper and lower than the shaded region are used for the fits , giving the blue- and black - solid lines .
, title="fig:",width=192 ] vs the fermion mass @xmath59 for various ir cutoffs @xmath60 in @xmath31 .
middle : the mass anomalous dimension @xmath2 obtained by fitting the non - shaded region data of the left panel with the conformal ansatz .
right : the pole mass @xmath61 vs @xmath62 in the conformal phase @xmath31 .
the data upper and lower than the shaded region are used for the fits , giving the blue- and black - solid lines .
, title="fig:",width=192 ] | motivated by recent progress on many flavor qcd on a lattice , we investigate conformal / walking dynamics by using schwinger - dyson ( sd ) equation within an improved ladder approximation for two - loop running coupling . by numerically solving the sd equation , we obtain a pole mass @xmath0 , pion decay constant @xmath1 , and investigate the chiral symmetry breaking and mass anomalous dimension @xmath2 in the presence of ir cutoffs @xmath3 .
we find that the chiral symmetry breaking is suppressed if ir cutoff @xmath3 becomes larger than the critical value near the dynamical mass ( @xmath3 @xmath4 ) in the conformal phase the @xmath2 is strongly suppressed by ir cutoffs for @xmath5 @xmath6 .
we , then , obtain finite size hyperscaling ( fss ) relation by adapting a linearized approximation for the sd equation , and examine the @xmath2 the results offer valuable insight and suggestion for analyses in lattice gauge theories . |
in this section we give the details of the computation of the properties of the studied system which is equivalent to the discrete oscillator @xmath117 we begin with the ideal system described by ( [ eq : h_0 ] ) .
two lowest energy states of this problem correspond to wave functions that are non - zero on even or odd charges as explained in the main text . for brevity
, we shall refer to these states as odd and even states . in the limit @xmath15
the splitting of the even and odd states can be found analytically . in this
limit the wave functions in the phase representation @xmath118 are mostly localized near the minima of the first term in ( [ eq : h_0 ] ) .
quantum transitions between two minima result in the coherent mixture of these two states with the bonding state corresponding to the even and antibonding state to the odd state in charge representation . to prove the correspondence between bonding and even states we notice that the bonding state does not change under transformation @xmath119 , while the antibonding state changes sign under this transformation . in the charge basis
this implies that bonding state has only even components while antibonding - only odd ones . in the quasiclassical approximation
the amplitude of this transition in which the phase slips by @xmath0 is @xmath120 here @xmath121 , @xmath122 is the plasma frequency , the dimensionless parameter @xmath19 has a physical meaning of the extent of the wave function in the charge representation : @xmath123 . in the asymptotic limit of very large @xmath124 the prefactor in ( [ eq : t ] ) , @xmath125 but for moderately large @xmath126 @xmath127 is significantly different from its asymptotic value as shown in fig .
s[fig : sprefactor ] .
the phase slip transitions by @xmath128 lead to the same states . adding these amplitudes we get @xmath129 that gives energy splitting between odd and even states @xmath130 in eq .
( [ eq : t ] ) on @xmath19 . for realistic ( not - too - large ) values of @xmath126
, @xmath131 significantly deviates from its asymptotic value @xmath132 . ]
[ fig : sprefactor ] we now discuss the effect of perturbations .
small but non - zero @xmath133 term in the hamiltonian leads to the transition between odd and even states . in the limit of large @xmath134 the amplitude of this transition is @xmath135 so in the basis of odd and even states the hamiltonian becomes @xmath136 that gives energy splitting between ground and the first excited state @xmath137 we now discuss the effect of external noises .
the largest source of noise is produced by fluctuating potentials that are coupled to the charge operator .
another smaller but potentially dangerous source of noise are fluctuations of @xmath23 cause by flux fluctuations in each rhombus : @xmath138 where @xmath139 is the josephson energy of the small josephson junction .
these noises are described by the hamiltonian @xmath140 in the charge basis the operator @xmath141 is diagonal , so it remains diagonal in the basis of even and odd states . at @xmath27 the wave functions of these states are symmetric in @xmath10 , so the potential fluctuations are decoupled in the linear order at @xmath27 . for @xmath142
the diagonal matix elements of @xmath143 can be found by differetiation of @xmath144 : @xmath145 it contains the same exponential factor as the energy splitting . in the absence of @xmath23
the external potential leads only to the dephasing proportional to @xmath146 .
for instance , the johnson noise associated with the effective impedance @xmath147 of the central island leads to dephasing rate @xmath148 here @xmath149 is the quantum of resistance and @xmath150 is the typical frequency responsible for dephasing process , for realistic temperatures @xmath151 so @xmath152 this equation assumes slow frequency dependence of the @xmath153 in a relevant frequency range @xmath154 .
small @xmath23 results also in a decay : @xmath155 the effect of the flux fluctuations is more direct because they affect the energy difference between the levels ( [ eq : e_01 ] ) . assuming that flux noise is characterized by the @xmath156 spectrum with @xmath157 infrared cutoff that translates into the @xmath156 spectrum of @xmath23 fluctuations @xmath158 @xmath159 here @xmath160 is the typical scale of @xmath23 fluctuations caused by low frequency noise , it is related to the typical flux variations by @xmath161 where @xmath139 is the josephson energy of the small contact in individual rhombus .
the numerical coefficient @xmath162 in this equation is of the order of unity in the regime @xmath163 ( @xmath164 , similarly to @xmath8 it decreases at @xmath165 .
the main elements of the mw set - up are shown in fig .
the microwave response of the josephson rhombi chain coupled to the read - out @xmath32 resonator ( see fig .
2 ) was probed by measuring both the phase and amplitude of the microwaves traveling along a microstrip feedline coupled to the resonators .
this setup enabled the testing of several devices fabricated on the same chip in a single cooldown .
the microwaves at the probe frequency @xmath65 , generated by a microwave synthesizer , were coupled to the cryostat input line through a 16 db coupler and transmitted through the microstrip line coupled to the @xmath32 resonators .
the cold attenuators and low - pass filters in the input microwave lines prevent leakage of thermal radiation into the resonator . on the output line ,
two cryogenic pamtech isolators ( @xmath166 db isolation between 3 and 10 ghz ) anchored to the mixing chamber attenuate the 5 k noise from the cryogenic hemt amplifier ( caltech citcryo 1 - 12 , 35 db gain between 1 and 12 ghz ) .
the amplified signal is mixed by mixer m1 with the local oscillator signal at frequency @xmath167 , generated by another synthesizer .
the intermediate - frequency signal @xmath168 at @xmath169 is digitized by a 1 gs / s digitizing card ( alazartech ats9870 ) .
the signal is digitally multiplied by @xmath170 and @xmath171 , averaged over an integer number of periods , and its amplitude @xmath131 ( proportional to the microwave amplitude @xmath172 ) and phase @xmath31 is extracted as @xmath173 and @xmath174 $ ] , respectively . the reference phase @xmath175 ( which randomly changes when both @xmath65 and @xmath167 are varied in measurements ) is found using similar processing of the low - noise signal provided by mixer m2 and digitized by the second channel of the adc .
this setup enables accurate measurements of small changes @xmath176 unaffected by the phase jitter between the two synthesizers .
the low noise of this setup allowed us to perform measurements at a microwave excitation level of -133 dbm which corresponded to a sub - single - photon population of the tank circuit . in the second - tone measurements ,
the tested rhombi chain was excited by the microwaves at frequency @xmath66 coupled to the transmission line via a 16 db coupler . in the presence of charge fluctuation on the central island of the chain ,
we have applied a sweep - by - sweep averaging method .
since these fluctuations are slow ( on the time scale of seconds ) , we perform 10 - 100 full - range sweeps of the second tone frequency ( at a rate of 10ms per one value of @xmath66 ) and then perform averaging of the completed sweeps rather than a point - by - point averaging of the second tone sweep .
the sample was mounted inside an rf - tight copper box that provided the ground plane for the microstrip line and @xmath32 resonator .
this box was placed inside another rf - tight copper box in order to eliminate any stray infrared photons .
this nested - box construction was housed inside a cryogenic magnetic shield ( a4k ) , followed by a superconducting aluminum magnetic shield .
all the connectors inside the magnetic shields were non - magnetic ( ezform ) ; this reduces the stray magnetic field at the sample location down to 0.1 mgauss .
the entire sample and shielding construction was anchored to the mixing chamber of a cryogen - free dilution refrigerator with a base temperature of 20 mk . in order to control the charge on the central island of the chain , a dc gate voltage was applied to the microstrip line via a bias - t .
external noises at the dc port of the bias - t were attenuated with a voltage divider and a series of low - pass filters and stainless steel powder filters .
are transmitted through the microstrip line inductively coupled to the rhombi chain and the @xmath32 resonator located inside the rf sample box .
the microwaves at @xmath167 , after mixing with the microwaves at @xmath65 , provide the reference phase @xmath175 .
the signal at @xmath65 is amplified , mixed down to an intermediate frequency @xmath177 by mixer m2 , and digitized by a fast digitizer ( adc ) .
the microwaves at the second - tone frequency @xmath66 were used to excite the @xmath178 transitions in the qubit . ]
the oscillations in the @xmath32 resonance frequency versus the magnetic flux are shown in fig .
s3 near zero magnetic field ( a ) and full frustration @xmath28 ( b ) .
the oscillations in the @xmath32 resonance frequency reflect the dependence of the inductance of the chain in its ground state as a function of the phase across the chain , @xmath179 . near full frustration
, the period of oscillations corresponds to @xmath70 .
the range of fluxes where the @xmath0-periodicity was observed agrees well with the theory .
the dependence @xmath180 also demonstrates avoided crossings observed when the resonance frequency @xmath50 coincides with the resonance in the read - out @xmath32 resonator ( the spikes in fig .
s3b ) , but the second - tone measurements discussed in the main text provide much more detailed information on the chain spectrum . versus the probe - tone microwave frequency and the phase across the chain near zero field ( a ) and near full frustration @xmath28 ( b ) .
the period of the inductance oscillations changes from @xmath181 ( @xmath182 to @xmath183@xmath184 near full frustration .
, title="fig : " ] versus the probe - tone microwave frequency and the phase across the chain near zero field ( a ) and near full frustration @xmath28 ( b ) .
the period of the inductance oscillations changes from @xmath181 ( @xmath182 to @xmath183@xmath184 near full frustration .
, title="fig : " ] the energy @xmath8 can be estimated from the measurements of the probe - tone microwave amplitude @xmath172 versus the magnetic flux ( fig .
the energy @xmath8 is inversely proportional to the josephson inductance @xmath185 of the chain : @xmath186 .
both quantities depend on the phase across the chain - @xmath187,@xmath188 - so @xmath189 and @xmath190 when @xmath191 . using the equivalent circuit in fig .
2c , one can estimate @xmath192 from the swing of the resonance frequency of the @xmath32 resonator coupled to the chain ( @xmath193mhz in fig .
s2b ) . using @xmath194 and @xmath60 , we estimate @xmath195 , which translates into @xmath196 ( or @xmath197 ghz ) .
this estimate agrees very well with the fitting parameters used to simulate the chain spectrum .
on the phase @xmath69 across the chain near full frustration at @xmath27 ( blue dots ) and 0.5 ( red squares ) . for comparison ,
@xmath198 is shown on the same plot ( black triangles ) . ] | we have studied the low - energy excitations in a minimalistic protected josephson circuit which contains two basic elements ( rhombi ) characterized by the @xmath0 periodicity of the josephson energy .
novel design of these elements , which reduces their sensitivity to the offset charge fluctuations , has been employed .
we have observed that the life time @xmath1 of the first excited state of this quantum circuit in the protected regime is increased up to 70@xmath2 , a factor of @xmath3 longer than that in the unprotected state .
the quality factor @xmath4 of this qubit exceeds @xmath5 .
our results are in agreement with theoretical expectations ; they demonstrate the feasibility of symmetry protection in the rhombi - based qubits fabricated with existing technology .
in this letter we make an essential step towards building a protected josephson qubit by fabricating the simplest protected circuit and demonstrating that the first excited state of the circuit is protected from energy relaxation .
the idea of protection is illustrated in fig.1 .
let us consider the simplest chain of two @xmath6 elements .
they share the central superconducting island whose charge is controlled by the gate .
the hamiltonian of this quantum circuit can be written as @xmath7 where the energy @xmath8 describes the josephson coupling of the central superconducting island to the current leads , @xmath9 is the effective charging energy of the island , @xmath10 is the number of cooper pairs on the island , @xmath11 is the charge induced on the island by the gate .
the parity of @xmath10 is preserved if the transfer of single cooper pairs is blocked ( @xmath12 ) . in this case
the states of the system can be characterized by the quantum number @xmath13 .
the low energy states corresponding to number @xmath14 are shown in fig.1 .
the energy @xmath8 plays the role of the kinetic term that controls the spread of the wave functions along the @xmath10 axis .
provided @xmath15 , the number of components with different @xmath10 in these discrete gaussian wavefunctions is large : @xmath16 , and the energy difference between the two states , @xmath17 , is exponentially small ( see @xcite and supplementary material 1 ) : @xmath18 here @xmath19 , @xmath20 is the plasma frequency , @xmath21 ( supplementary material 1 ) .
furthermore , these states can not be distinguished by the noise operators , and , thus , the decay and dephasing rates are both reduced by the same large factor @xmath22 . in real circuits , the 2e processes are not completely suppressed , and a non - zero amplitude @xmath23 mixes the odd and even components ( fig.1 ) and increases @xmath24 . for a small amplitude @xmath23
the decay of the first excited state is due to the mixture of @xmath14 states and is suppressed by the factor @xmath25 in addition to the suppression by the factor @xmath22 that is common to decay and dephasing .
thus , for the coherence protection two conditions are required : @xmath26 ( i.e. slow energy relaxation ) and @xmath15 ( i.e. small dephasing rate ) .
element ; the charge of the central ( common to both rhombi ) island is controlled by the gate voltage .
bottom panel : two lowest - energy wavefunctions in the discrete harmonic potential shown for @xmath27 . ] the simplest @xmath6 josephson elements is represented by the josephson rhombus : a superconducting loop interrupted by four identical josephson junctions @xcite ( fig.2a ) .
when the rhombus is threaded by the magnetic flux @xmath28 ( @xmath29 is the flux quantum ) , its effective josephson energy @xmath30 becomes @xmath0-periodic in the phase difference across the rhombus , @xmath31 . in line with theoretical predictions , recent experiments @xcite have demonstrated that the properly designed small rhombi arrays can support a non - zero 4e supercurrent in the regime when the 2e supercurrent vanishes
. josephson element ( the josephson rhombus ) and its improved version which is less sensitive to the random offset charges on the upper and lower islands @xcite .
the chain contains two rhombi ; the charge of the central ( common to both rhombi ) island is controlled by the gate voltage applied between the central conductor of the microstrip line and the ground .
panel ( b ) : the on - chip circuit layout of the device inductively coupled to the microstrip transmission line .
panel ( c ) : the micrograph of the two - rhombi chain coupled to the read - out @xmath32 resonator via the kinetic inductance @xmath33 of a narrow superconducting wire .
the magnetic flux trough each rhombus is @xmath34 ; the flux in the phase loop is @xmath35 . ] in the current work , we have implemented the two - rhombi chain with an improved design of individual rhombi proposed in ref .
@xcite .
the key requirement for the protection is the cancellation of @xmath23 due to the destructive interference between the cooper pair transfer amplitudes along the upper and lower branches of a rhombus .
this cancellation is difficult to achieve in the quantum regime where the amplitudes depend on the uncontrolled offset charges on the side islands .
this dependence is due to the aharonov - casher effect ( see e.g. @xcite ) and the high probability of phase slips across the two small junctions in each arm of a rhombus . in the improved rhombus design @xcite one junction in each branch
is replaced by a short chain of larger junctions ( fig.2a ) .
the phase slips across the larger junctions are suppressed due to a large @xmath36 ratio ( @xmath37 for the studied devices ) . as a result
, an improved rhombus becomes insensitive to the offset charges on all islands except for the central island shared by both rhombi .
below we refer to the characteristic energies of the smaller and larger junctions as @xmath38,@xmath39 and @xmath40,@xmath41 respectively .
an optimal operation of the rhombus is realized for @xmath42 and @xmath43 , where @xmath44 is the number of larger junctions in the chain ( @xmath45 for the studied devices ) .
the chains , the readout circuits , and the microwave ( mw ) transmission line were fabricated using multi - angle electron - beam deposition of al films through a lift - off mask ( for fabrication details see refs .
@xcite ) .
the in - plane dimensions of the small and large junctions were @xmath46 and @xmath47 , respectively .
the ratio @xmath48 was chosen to realize the resonance frequency of the @xmath49 transition , @xmath50 , within the ( @xmath51 ) ghz range .
below we show the data for one representative device with @xmath52 , @xmath53 , @xmath54 , and @xmath55 ( throughout the paper all energies are given in the frequency units , @xmath56 ) .
the readout lumped - element @xmath32 resonator was formed by the meandered 2-@xmath57m - wide al wire with the kinetic inductance @xmath58 nh and an interdigitated capacitor @xmath59 ff .
the resonator was coupled to the chain via a narrow superconducting wire with a kinetic inductance of @xmath60 .
the chain and the wire formed a superconducting loop ; the flux @xmath35 of the external magnetic field in this phase loop controlled the phase difference across the chain @xmath61 ( fig.2c ) .
because the phase loop area ( @xmath62 ) was much greater than the rhombus area ( @xmath63 ) , the phase across the chain could be varied at an approximately constant value of @xmath34 .
several devices with systematically varied values of @xmath38 and @xmath39 were fabricated on the same chip and inductively coupled to the same microstrip line ( fig.2b ) .
the devices could be individually addressed due to different resonance frequencies of the @xmath32 resonators .
all measurements have been performed at @xmath64=20 mk . in the experiment
, the microwaves traveled along the microstrip line inductively coupled to the device .
the microwaves at the frequency @xmath65 probed the @xmath32 resonance .
the microwaves at the second - tone frequency @xmath66 excited the transitions between the quantum states of the chain , which resulted in a change of the impedance of this non - linear system .
the chain excitations were detected as a change in the amplitude @xmath67 and the phase of the microwaves at the probe - tone frequency @xmath65 ( for measurement details , see supplementary material 2 ) .
below we focus on the most interesting range of magnetic fields close to full frustration ( @xmath68 ) , where each rhombus represents a @xmath6 josephson element .
the measurements at the probe - tone frequency ( no microwaves at @xmath66 ) show that in this regime the response of the chain to the phase difference @xmath69 is indeed periodic with the period @xmath70 ( see supplementary material 3 ) .
figure 3 summarizes the spectroscopic data obtained in the two - tone measurements .
the inset in fig .
3 shows the resonance frequency @xmath50 measured as a function of @xmath11 at a fixed magnetic flux @xmath71 , @xmath72 . note that during the data accuisition time for the inset in fig.3 ( @xmath73 hours ) , no long - term shifts in the offset charge were observed .
also , quasiparticle poisoning was strongly suppressed due to ( a ) a larger superconducting gap of the central island ( in comparison with the nearest - neighbor islands ) and a relatively large @xmath9 @xcite , and ( b ) shielding of the device from stray infra - red photons by the double - wall light - tight sample holder ( see , e.g. @xcite ) . for perfectly symmetric rhombi , the two states corresponding to @xmath14 ( odd and even number of cooper pairs on the central island ) should become degenerate ( @xmath74 ) at @xmath75 . slight asymmetry of the studied rhombi results in a non - zero @xmath76 ( see also supplementary material 4 ) . at @xmath27 and @xmath77
, @xmath24 is equal to the effective charging energy @xmath78 .
the non - zero energy @xmath8 suppresses @xmath24 ( [ eq : e01 ] ) to 6.7 ghz .
the energies @xmath23,@xmath8 , and @xmath9 were obtained from the experimental dependences @xmath79 measured at different values of @xmath69 by fitting them with the computations of the spectra based on the full hamiltonian diagonalization . as a function of the phase difference across the chain
, @xmath8 oscillates with the period @xmath70 ( fig.3 ) .
the dependence @xmath80 agrees very well with the one expected theoretically : @xmath81 ( solid red line ) where @xmath82ghz is the energy of an individual rhombus , @xmath83 is the difference between these energies for two rhombi .
the fit shows that @xmath83 does not exceed @xmath84 for the studied chain .
the oscillations of @xmath8 result in a periodic dependence of the measured energy @xmath85 .
the estimated values of @xmath23 do not exceed 0.75 ghz around the optimal values @xmath86 corresponding to a maximum @xmath8 .
this small asymmetry in rhombi branches is consistent with the reproducibility of submicron junctions fabricated with the manhattan - pattern technique @xcite .
( blue dots ) and energies @xmath23 ( black triangles ) and @xmath8 ( red squares ) versus the phase @xmath69 across the chain near full frustration @xmath87 ) at @xmath27 .
the theoretical dependences @xmath88 ( solid blue curve ) and @xmath80 ( solid red curve ) were calculated for @xmath89 and @xmath78 .
the inset : the dependence @xmath79 at @xmath90 @xmath91 ) .
the energies @xmath23 and @xmath8 were extracted from fitting the dependences @xmath79 measured at different @xmath69 with the calculations based on hamiltonian diagonalization ( solid black curve ) .
the expected dependence @xmath79 for the case of perfectly symmetric rhombi is shown by the red dashed curve . ] because the bandwidth of the second - tone microwave line in our set - up was limited to @xmath92 , we could not access higher energy states directly . however , by increasing the second - tone power , we were able to observe the resonances corresponding to the multiphoton ( n=2,3,4 ) excitations of the @xmath93 transition around the optimal values @xmath94 the energy @xmath95 extracted from these measurements exceeded 50 ghz which is in good agreement with our expectations , and confirms the validity of the theoretical model ( eq .
1 ) .
the energy relaxation of the state @xmath96 was measured by exciting the rhombi chain with a @xmath0-pulse and then measuring its state after a variable time @xmath97 ( see fig.4a inset for pulse sequence ) .
the results of the energy relaxation measurements at the optimal working point ( @xmath28 , @xmath90 ) are presented in fig.4a and table 1 .
we also included in table 1 the @xmath1 data for the device with a large degree of asymmetry ( device 2 ) .
this asymmetry was caused by different values of the flux @xmath34 in the nominally identical rhombi : variations of the magnetic field across the chip were caused by conventional ( slightly magnetic ) microwave connectors on the sample holder .
after replacing these connectors with the non - magnetic ones , this source of asymmetry was eliminated , and all the measured devices consistently demonstrated @xmath98 .
the values of @xmath1 for the rhombi chains with a high degree of symmetry are 1 - 2 orders of magnitude greater than that for less symmetric rhombi circuits and unprotected josephson qubits coupled to the same read - out circuit @xcite .
the comparison shows that the symmetry protection supresses the decay rate by almost two orders of magnitude , in agreement with the large value of the decay suppression factor @xmath99 ( table 1 ) .
note that the value of @xmath1 at @xmath27 might be limited by the purcell decay into the @xmath32 resonator @xcite .
the quality factor @xmath4 for the protected rhombi chains exceeds @xmath100 ( see table 1 ) , and is comparable with that for the state - of - the - art transmons coupled to 3d cavities @xcite and tin coplanar resonators @xcite .
however , the reasons for such a large @xmath4 in the transmon and in the rhombi chain are different . in the former case
the decay is suppressed by the carefully designed microwave environment whereas in the latter case the rhombi symmetry prohibits the decay .
table 1 . [ cols="^,^,^,^,^,^,^",options="header " , ] @xmath101 in contrast to the long decay time , the decoherence time in the studied devices was relatively short ( @xmath102 ) .
the time @xmath103 was determined in ramsey measurements by applying an @xmath104 pulse followed by another @xmath104 pulse after a time @xmath105 t ( see fig.4b inset for pulse sequence ) . in the spin echo measurements
, a refocusing @xmath106 pulse was applied between the two @xmath104 pulses . from the ramsey and spin echo measurements we found @xmath107 and @xmath108 .
the dephasing time is expected to be long if @xmath109 . in the studied chain , these energies were of the same order of magnitude , which resulted in significant dephasing ( fig.4b ) . and @xmath71 .
panel ( a ) : energy relaxation of the first excited state after application of a @xmath0-pulse : @xmath110 ( blue circles ) , @xmath27 ( red squares ) .
inset : pulse sequence used in the measurement .
panel ( b ) : ramsey ( @xmath103 ) and spin echo ( @xmath111 ) measurements .
inset : pulse sequence used in the measurements . for the spin echo measurement ,
a refocusing @xmath0-pulse ( red dashed line ) was applied at @xmath112 . ] to conclude , we have demonstrated that a josephson circuit can be symmetry - protected from the energy decay .
we have studied the minimalistic protected circuit which contains two josephson rhombi .
the symmetry between the rhombus branches translates into the halving the periodicity of its josephson energy @xmath113 , and allows only the simultaneous transfer of pairs of cooper pairs on the central island mutual to both rhombi .
the logical states of the protected qubit correspond to the even and odd number of cooper pairs on this island .
our data indicate that the improved design of the josephson rhombi @xcite reduces the sensitivity of the chain spectrum to the offset charge asymmetry .
the measured phase and charge dependences of the energy of the @xmath49 transition are in good agreement with our numerical simulations .
symmetry protection results in a long energy relaxation time @xmath1 ( up to 70@xmath2 ) and a large quality factor @xmath114 of this qubit .
the experiments provide a solid foundation for the next stage the implementation of a qubit with much improved coherence due to a larger ratio @xmath115 which can be realized , e.g. , by parallel connection of a few rhombi chains .
we would like to thank b. doucot for helpful discussions .
the work was supported in part by grants from the templeton foundation ( 40381 ) , the nsf ( dmr-1006265 ) , and aro ( w911nf-13 - 1 - 0431 )
. 10 emanuel knill .
quantum computing with realistically noisy devices .
, 434:3944 , 2005 .
a.yu .
kitaev .
fault - tolerant quantum computation by anyons .
, 303:2 30 , 2003 .
l.b .
ioffe , m.v .
feigelman , a. ioselevich , d. ivanov , m. troyer , and g. blatter .
topologically protected quantum bits using josephson junction arrays .
, 415:503506 , 2002 .
l. b. ioffe and m. v. feigelman . possible realization of an ideal quantum computer in josephson junction array .
, 66:224503 , 2002 .
b. doucot and j. vidal .
pairing of cooper pairs in a fully frustrated josephson - junction chain .
, 88:227005 , 2002 .
b. doucot , m.v .
feigelman , and l.b .
ioffe .
topological order in the insulating josephson junction array .
, 90:107003 , 2003 .
b. doucot , m.v .
feigelman , l.b .
ioffe , and a.s .
ioselevich . protected qubits and chern - simon theories in josephson junction arrays .
, 71:024505 , 2005 .
i.v .
protopopov and m.v .
feigelman .
anomalous periodicity of supercurrent in long frustrated josephson - junction rhombi chains .
, 70:184519 , 2004 .
i.v .
protopopov and m.v .
. coherent transport in josephson - junction rhombi chain with quenched disorder .
, 74:064516 , 2006 .
p. brooks , a. kitaev , and j. preskill . protected gates for superconducting qubits .
, 87:052306 , 2013 .
b. doucot and l.b .
. physical implementation of protected qubits .
, 75:120 , 2012 .
s. gladchenko , d. olaya , e. dupont - ferrier , b. doucot , l.b .
ioffe , and m.e .
gershenson .
superconducting nanocircuits for topologically protected qubits . , 5:4853 , 2009 .
i. m. pop , k. hasselbach , o. buisson , w. guichard , b. pannetier , and i. protopopov .
measurement of the current - phase relation in josephson junction rhombi chains .
, 78:104504 , 2008 .
i. m. pop , b. dou ot , l. ioffe , i. protopopov , f. lecocq , i. matei , o. buisson , and w. guichard . experimental demonstration of aharonov - casher interference in a josephson junction circuit .
, 85:094503 , 2012 .
matthew t. bell , lev b. ioffe , and michael e. gershenson .
microwave spectroscopy of a cooper - pair transistor coupled to a lumped - element resonator .
, 86:144512 , 2012 .
m.t .
bell , i.a .
sadovskyy , l.b .
ioffe , a.yu .
kitaev , and m.e .
gershenson .
quantum superinductor with tunable non - linearity . , 109:137003 , 2012 .
j. aumentado , mark w. keller , john m. martinis , and m. h. devoret .
nonequilibrium quasiparticles and @xmath116 periodicity in single - cooper - pair transistors . , 92:066802 , 2004 .
p.j .
de visser , j.j.a .
baselmans , p. diener , s.j.c .
yates , a. endo , and t.m .
klapwijk .
number fluctuations of sparse quasiparticles in a superconductor .
, 106:167004 , 2011 .
a. a. houck , j. a. schreier , b. r. johnson , j. m. chow , jens koch , j. m. gambetta , d. i. schuster , l. frunzio , m. h. devoret , s. m. girvin , and r. j. schoelkopf . controlling the spontaneous emission of a superconducting transmon qubit .
, 101:080502 , 2008 .
hanhee paik , d.i .
schuster , l.s .
bishop , g. kirchmair , g. catelani , a.p .
sears , b.r .
johnson , m.j .
reagor , l. frunzio , l.i .
glazman , s.m .
girvin , m.h .
devoret , and r.j .
schoelkopf .
observation of high coherence in josephson junction qubits measured in a three - dimensional circuit qed architecture .
, 107:240501 , 2011 .
j.b .
chang , m.r .
vissers , a.d .
corcoles , m. sandberg , j. gao , d.w .
abraham , j.m .
chow , j.m .
gambetta , m.b .
rothwell , g.a .
keefe , m. steffen , and d.p .
pappas . improved superconducting qubit coherence using titanium nitride .
, 103:012602 , 2013 . |
it is well established that nuclear magnetic resonance ( nmr ) measurements of diffusing spins can be used to probe the geometry of porous media .
the nmr response is sensitive to the surface to volume ratio of the confining pore space @xcite . for smooth ,
ideal shapes , this relationship provides a means of estimating the pore size distribution @xcite . naturally occurring materials
may however have a more complicated pore structure than these ideal shapes , leading us to pose the question : can there be a direct link to pore size when the pore geometry is not ideal ?
the standard approach in a nmr relaxometry experiment is to measure the transverse ( spin - spin ) relaxation time of water protons in a porous medium and estimate the relaxation time @xmath0 @xcite .
in addition to the relaxation processes in the bulk fluid the protons interact with paramagnetic sites at the pore surface , which increases the relaxation rate @xcite . in this way , the relaxation experiment is sensitive to the surface - to - volume ( s / v ) ratio of the confining porous medium .
brownstein and tarr realized that the complex nature of the spin - surface interaction is well described by a robin boundary condition @xcite .
the magnetization @xmath1 in a pore is then described by the following bloch - torrey equation @xmath2 subject to a robin condition at the pore boundary @xmath3 where the scalar @xmath4 denotes the surface relaxivity and @xmath5 the diffusion coefficient of the spins . in an nmr relaxation experiment the relaxation rate(s )
are sought . omitting the bulk fluid relaxation properties we restate the classical result @xmath6 which has been derived on multiple occasions and is true in the limit of the time @xmath7 approaching zero and in the so - called fast - diffusion limit depicted by @xmath8 where r denotes the pore size @xcite .
equation [ eq : fast - diffusion - limit ] allows an estimate of the pore size distribution from nmr measurements in cases where the pores can be approximated by ideal shapes ( i.e. slabs , spheres or cylinders ) where a simple relationship between @xmath9 and the pore radius can be established .
experiments @xcite and numerical studies @xcite suggest that not only will a deviation from ideal shapes disrupt this relationship but that surface roughness may also have an impact ; no rigorous investigation of this latter effect has been conducted . in this paper
we show that when the surface is rough ( as is true for most naturally occurring materials ) the above expression is no longer directly translatable to the pore radius and that the derived pore size can differ substantially from the actual size . we demonstrate that in the presence of surface roughness the spins still behave as being in a smooth pore , but subject to a different , effective , surface relaxivity .
we provide a means of calculating this quantity , by introducing a magnetization rate coefficient describing the magnetic dissipation over the rough surface . with some a geometrical perturbation @xmath10 separated by a fictitious boundary @xmath11.[fig:1 ] ]
we begin our analysis by considering a smooth pore @xmath12 with surface @xmath13 and with a small geometrical perturbation @xmath10 , with surface @xmath14 ( see figure [ fig:1 ] ) representing surface roughenss .
`` small '' here means that the volume ( area in this 2d representation ) of the perturbation is much smaller than the volume ( area in this 2d representation ) of the original pore @xmath15 .
`` rough '' means that the perturbation is local with respect to the pore surface i.e. if we separate the two domains with a fictitious boundary @xmath11 then the surface ( length ) of @xmath11 is much smaller than the smooth pore surface ( length ) @xmath13 .
hence the geometrical perturbation @xmath10 can be seen as a local roughening of the surface .
we will utilize this fictitious boundary below . separating the diffusion equation depicted in equation [ eq : bloch - torrey ] yields the problem at hand : to find relaxation modes @xmath16 and relaxation times @xmath17 satisfying @xmath18 where @xmath19 denotes the outward pointing normal of the total pore surface @xmath20 and where @xmath21 denotes the total pore @xmath22 .
we now proceed by considering the right - hand side of equation [ eq : bt ] as a charge distribution @xmath23 .
the equations for the two sub - domains then become two coupled electrostatic problems @xmath24 for @xmath25 and some unknown charge distribution @xmath26 at the boundary @xmath11 .
these equations are satisfied when the charge distribution @xmath23 equals the original relaxation modes @xmath27 in equation [ eq : bt ] . a general solution to the coupled electrostatic problem in equation [ eq : poissons_equation ]
is given by @xmath28 where @xmath29 satisfies the laplace equation @xmath30 with the inhomogeneous boundary condition at @xmath11 : @xmath31 and where @xmath32 satisfies the poisson s equation @xmath33 with all homogeneous boundary conditions i.e. @xmath34 . by defining @xmath35
one may note that the solution @xmath36 satisfies the following inhomogeneous robin condition @xmath37 at @xmath11 where we call the parameter @xmath38 the magnetization transfer coefficient @xmath39 this parameter describes the magnetic dissipation over the fictious boundary @xmath11 .
the parameter @xmath40 may be estimated by noting that the relaxation times are proportional to the pore volume , i.e. @xmath41 hence , by forming a gaussian surface one may determine that @xmath42 with @xmath40 approaching zero as @xmath43 .
therefore the inhomogeneous boundary condition described by eq .
[ eq : inhomogeneous_rbc ] approaches a homogeneous b.c .
when the perturbation @xmath10 is much smaller than the original pore . by continuity the same boundary condition must hold in the larger pore @xmath12 as well .
the magnetic transfer coefficient @xmath38 can be estimated by assuming that it is constant over @xmath11 @xmath44 ( a good approximation when @xmath45 due to the form of the relaxation modes ) .
we then have @xmath46 induced by a rectangular box with base dimensions @xmath47 ( @xmath48 m ) , @xmath49 ( @xmath48 m ) and height @xmath50 ( @xmath48 m ) as a function of @xmath51 ; the base @xmath52 corresponds to the fictitious boundary ( @xmath11 ) , the opening to the larger pore .
three different values of the surface relaxivity were used : @xmath53 ( @xmath48 m / s ) ( dashed ) @xmath54 ( @xmath48 m / s ) ( dotted ) and @xmath55 ( @xmath48 m / s ) ( dashed - dotted ) .
the solid lines depict the initial slope and the asymptotes .
[ fig:2 ] ] in the case when the domain @xmath10 is a separable geometry and where @xmath56 denotes the normal to @xmath11 the laplace equation becomes @xmath57 where @xmath58 denotes the constant of separation .
the solution to equation [ eq : h_eq ] is then @xmath59 where @xmath60 denotes the eigenfunctions perpendicular to @xmath61 . when @xmath62 ( i.e. the spins inside the small domain can be considered to be in the fast - diffusion regime )
the dominant contribution will be from the lowest relaxation mode @xmath63 .
therefore one can consider the approximate expression @xmath64 where the subindex @xmath65 of @xmath38 denotes the average magnetic transfer coefficient and the first coefficient @xmath66 is @xmath67 a series expansion around @xmath68 gives @xmath69 and for @xmath70 we reach the asymptotic value @xmath71 we now represent the small geometrical perturbation as a rectangular box , and show in figure [ fig:2 ] the average magnetic transfer coefficient @xmath72 plotted as a function of the height @xmath51 of the box where the base area corresponds to the fictitious boundary @xmath11 .
as the height increases the increased surface area has less impact on @xmath72 , which reaches a plateau value .
the major impact comes however from the area of the base , revealed in the separation constant @xmath73 where @xmath74 and @xmath75 denote the lengths of the sides of the base of the rectangular box .
while there are sharp corners between the spherical surface and the rectangular box , we are working with the robin kernel , which has a smoothing effect due to the dirichlet nature that regularizes this type of sharp features @xcite .
as mentioned above , the identified robin condition in the electrostatic presentation tends towards a homogeneous boundary condition when @xmath76 . as a consequence
, the original eigenequation describing the spin relaxation in the total domain @xmath21 may be well approximated by the altered ( homogeneous ) boundary condition of the smooth domain @xmath12 in the following way @xmath77 by adding many small perturbations we reach a rough pore surface with a significantly increased surface area .
this problem may then be simplified further by defining an effective surface relaxivity @xmath78 in the following way @xmath79 and utilizing the well - known analytical solutions for a smooth pore @xcite .
as an example of the usefulness of these results consider a spherical pore with a radius of @xmath80 ( @xmath48 m ) where the sphere surface is roughened by modulating the surface using small rectangular boxes ( see inset in figure 2 ) .
we let the surface relaxivity of the pore surface ( including the surface of the rectangular boxes ) be @xmath53 ( @xmath48 m / s ) and let the rectangular boxes have the dimensions @xmath47 ( @xmath48 m ) , @xmath49 ( @xmath48 m ) where again the base area @xmath52 denotes the fictitious boundary @xmath11 .
setting the height of the boxes to @xmath50 ( @xmath48 m ) and using a diffusion constant @xmath81 ( @xmath82 @xmath83/s ) ( corresponding to water at @xmath84 c ) we get the average magnetic transfer coefficient @xmath85 a considerably larger value than the original assigned surface relaxivity @xmath86 .
this value describes the average relaxivity over the fictitious domain @xmath11 , as induced by the inclusion of the rectangular box to roughen the surface of the large pore .
if we assume that @xmath80% of the spherical surface is covered by such boxes we get the effective relaxivity @xmath87 the deviance between the original assigned @xmath86 and the effective @xmath78 is purely geometrical and due to the roughness of the pore surface . given the effective relaxivity it is straightforward to solve for the lowest relaxation time for this spherical pore @xcite : @xmath88 given the pore size , the original surface relaxivity and the diffusion coefficient , we would conclude that we would be justified in evaluating the pore size using the fast - diffusion limit ( @xmath89 ) .
this would however give a pore size of @xmath90 ( @xmath48 m ) , a value which deviates quite substantially from the assigned @xmath91 ( @xmath48 m ) of the pore we began with .
using instead the effective surface relaxivity ( obtained by calculating the average magnetic transfer coefficient ) , we find that the fast - diffusion limit no longer applies ( @xmath92 ) . in order to obtain the correct pore radius in this regime , one must instead solve for @xmath93 in the following non - linear problem ( derived by the solution to the robin problem for a sphere @xcite ) , @xmath94 this yields @xmath95 @xmath96 , which is in close agreement with the actual pore radius . as a function of surface - to - volume ( @xmath9 ) for a rough sphere of radius @xmath91 @xmath97 where the @xmath9 is increased by covering the @xmath98 of the surface of the sphere by small rectangular boxes of width @xmath99 @xmath97 and varying the length ( squares ) and alternatively covering @xmath98 of the surface by cylinders of radius @xmath100 @xmath97 and varying the length ( triangles ) .
the dashed line show the fast - diffusion approximation and the dotted line denote the limit of the slow - diffusion regime .
the diffusion coefficient was set to @xmath81 @xmath101 and two values of the surface relaxivity was used , @xmath102 @xmath103 and @xmath104 @xmath103 .
both cases are expected to follow the fast - diffusion limit depicted by the dashed lines and tend towards the slow - diffusion limit ( @xmath105 ) as @xmath9 increases .
[ fig:3 ] ] the magnetic transfer coefficient is straightforward to obtain analytically where the surface roughness can be modeled by analytically solvable geometries . for example
a cylinder of radius @xmath106 and height @xmath51 ( where the base area @xmath107 corresponds to @xmath11 ) has the separation constant @xmath108 which coincides with the rectangular box when the @xmath109 , i.e. the magnetization transfer coefficient becomes equal .
the cylinder gives the initial slope @xmath110 and the asymptotic value @xmath111 as @xmath51 grow large .
figure [ fig:3 ] depicts the inverse of the relaxation rate of a spherical pore of radius @xmath91 ( @xmath48 m ) where @xmath80% of the spherical surface is covered by such cylinders as a function of surface - to - volume where the increased surface - to - volume is modulated by varying the height of the cylinders .
finally , the developed homogenization procedure has been successfully validated in numerous numerical examples ( not included ) , with different type of surface roughness .
the presented theory allows us to investigate the effect of rough surfaces on nmr relaxation experiments by introducing a magnetic transfer coefficient describing the dissipation of magnetization over the rough surface .
this allows a convenient way of homogenizing rough pores to smooth equivalents with an effective surface relaxivity .
we find that the effective relaxivity increases dramatically in presence of rough surfaces , in particular where the surface roughness is narrow and deep .
the explicit expressions for the effective relaxivity demonstrates the possibility of determining a pore length scale for rough pores when the effective surface relaxivity is known .
our results agrees well with recent experimental findings by keating @xcite where surface roughness was induced by etching the surface of glass beads and with numerical simulations by mller - petke et al . | most theoretical treatments of nuclear magnetic resonance ( nmr ) assume ideal smooth geometries ( i.e. slabs , spheres or cylinders ) with well - defined surface - to - volume ratios ( s / v ) .
this same assumption is commonly adopted for naturally occurring materials , where the pore geometry can differ substantially from these ideal shapes . in this paper the effect of surface roughness on the @xmath0 relaxation spectrum is studied . by homogenization of the problem using an electrostatic approach it is found that the effective surface relaxivity can increase dramatically in the presence of rough surfaces .
this leads to a situation where the system responds as a smooth pore , but with significantly increased surface relaxivity . as a result : the standard approach of assuming an idealized geometry with known surface - to - volume and inverting the @xmath0 relaxation spectrum to a pore size distribution is no longer valid .
the effective relaxivity is found to be fairly insensitive to the shape of roughness but strongly dependent on the width and depth of the surface topology . |
the geometrical reciprocal spinors determine the rotational symmetries of the spinors . associated with each such symmetry
is a phase factor which we will calculate here .
suppose the rotation through an angle @xmath54 about the axis @xmath114 is a symmetry of the reciprocal spinor of @xmath15 .
then @xmath115 where @xmath52 is given by eq . (
[ eq : lambda ] ) . to prove this
, it is convenient to choose ( without loss of generality ) the axis of rotation to be parallel to the @xmath73-axis : @xmath116 . then by comparing the components of the spinors on the left- and righthand side of eq .
( [ eq : transformation ] ) we see that @xmath117 this equation implies that the non - zero components of the spinor must satisfy @xmath118 let us compare this result to the multiplicity of the root at the north pole , which corresponds to @xmath119 .
the characteristic equation is @xmath120 if the root at zero has multiplicity @xmath55 , then the coefficients of the first @xmath55 terms @xmath121 are all equal to zero , and the @xmath122 coefficient is nonzero .
this corresponds to the @xmath122 component of the spinor being nonzero : @xmath123 substituting @xmath124 into the condition for the nonzero components eq .
( [ eq : periodic ] ) gives the result eq .
( [ eq : lambda ] ) and the proof is complete . note that our discussion has been for a rotational axis aligned with the @xmath73-axis , but the result eq .
( [ eq : lambda ] ) has a conceptual formulation which does not depend on the coordinate system .
10 [ 2]#2 [ 1]#1 [ 1]#1 , , , , , and , , ( ) . , , , , , , , , and , , ( ) .
, , , , , , , and , , ( ) .
, , , , and , , ( ) .
, , , , and , , ( ) . and , , ( ) . and , , ( )
. and , cond - mat/0610071 . , , and , cond - mat/0612044 .
, , , , , , , and , , ( ) .
, , ( ) . and , , ( ) . , , and , cond - mat/0512602
, , ( ) . , , and , , ( ) . and , cond - mat/0610485
, , ( ) . and , , ( ) . , , and , , ( )
, , ( ) . , ( , ) . , cond - mat/0611171 | motivated by the recent realization of a @xmath0cr bose - einstein condensate , we consider the phase diagram of a general spin - three condensate as a function of its scattering lengths .
we classify each phase according to its reciprocal spinor , using a method developed in a previous work .
we show that such a classification can be naturally extended to describe the vortices for a spinor condensate by using the topological theory of defects .
to illustrate , we systematically describe the types of vortex excitations for each phase of the spin - three condensate . over the past few years
, there has been remarkable experimental progress in the study of spinor bose - einstein condensates @xcite .
one of the most intriguing examples is the recent realization of a @xmath0cr @xcite condensate which has since received substantial theoretical attention @xcite .
atomic chromium has six electrons in its outermost shell which combine into a state with maximal spin @xmath1 according to hund s rules .
these electrons also give rise to a magnetic dipole moment which is much larger than that of the alkalis where dipolar effects are typically unimportant .
in fact , such dipolar effects have been observed in the expansion profiles of @xmath0cr @xcite .
dipolar effects are important for magnetic traps where the spin degrees are frozen out . however , when purely optical traps are used it has been argued that the spin exchange interaction will overwhelm the dipolar interaction @xcite . due to such spin interactions , a dilute gas of spin
three bosons will interact according to the pair potential @xmath2 where @xmath3 projects into the state with total spin @xmath4 and @xmath5 where @xmath6 is the scattering length corresponding to spin @xmath4 . for the spin - three problem ,
this potential gives rise to a phase diagram ( as a function of the possible values of the scattering lengths ) of great richness as was revealed in @xcite .
in addition to having such rich phase diagrams , spinor condenstates in general and the @xmath0cr condensate in particular will have intriguing topological excitations .
topological excitations for spin - one , spin three - halves , and spin - two condenstates have received considerable attention in the literature @xcite since the pioneering work of ho @xcite and ohmi and machida @xcite . also , recently topological spin textures have been observed in @xmath7rb atoms in the spin - one hyperfine state @xcite . in a previous publication @xcite , we have developed a way of classifying spinor bec and mott insulating states and illustrated the method with an application to the spin - two boson problem .
more specifically , we showed that it is natural to identify a spin @xmath4 state @xmath8 with @xmath9 points on the unit sphere @xmath10 .
in fact , there is a one - to - one correspondence between a spinor and a `` reciprocal spinor '' up to phase and normalization @xmath11 where @xmath12 is the stereographic image on the complex plane of the point @xmath10 on the unit sphere .
the set of such complex numbers @xmath13 are explicitly constructed by finding the coherent states @xmath14 which are orthogonal to the spinor states @xmath15 .
the characteristic equation @xmath16 will then be a polynomial in @xmath17 of order @xmath9 and its @xmath9 roots projected on the sphere will immediately give the symmetries of the spinor .
spin rotations which leave this set of points invariant will take a spinor @xmath18 into itself modulo an overall phase which can be computed through geometrical considerations .
thus this method gives a complete description of the spinor state . in this work , we show that such a method is also useful for classifying the spinor vortices . to illustrate , we focus on the spin - three problem , motivated by the recent experimental development @xcite .
we first reproduce the phase diagrams for spin - three condensates obtained previously in @xcite .
then we find the symmetries of each phase ( see fig . [
fig : shapes ] ) by applying the method developed in @xcite .
we then systematically describe the types of topological excitations for such a system using the topological theory of defects @xcite .
such a scheme enumerates all possible types of vortex excitations , though it does not determine which are energetically stable against dividing into multiple vortices .
finally , we will briefly discuss the resulting microscopic wave functions for the vortex excitations .
such understanding of vortices will be relevant for rotating @xmath0cr condensates in optical traps .
in general , the topological vortex excitations ( line defects ) of an ordered medium are given by the first homotopy group @xmath19 ( the group of closed loops ) of the order parameter space @xmath20 .
more specifically , the conjugacy classes of @xmath21 correespond to the vortex types .
moreover , the way such vortices combine is determined by the multiplication table of the conjugacy classes . for a pedagogical review of the topological theory of defects ,
see @xcite ( whose notation we will adopt ) . a theorem from algebraic topology ( see e.g. @xcite ) which is used as a reliable tool for computing the first homotopy group
is the following .
let @xmath22 be a simply connected group and @xmath23 be a subgroup of @xmath22 .
then @xmath24 where @xmath25 are the elements of @xmath23 which are connected to the identity by continuous transformations . to apply this theorem to our system , we need to take @xmath22 to be a group of operators that act transitively on the spin groundstates and @xmath23 to be the subgroup of @xmath22 which leave the particular spinor state we are considering invariant
. then @xmath20 has the same topology as @xmath26 , so the theorem applies .
the first guess would be to use elements like @xmath27 ( where @xmath28 is the spin - three matrix and @xmath29 is a unit vector ) which act directly on the spinor states and superfluid phase .
however , these elements form a representation of @xmath30 , which is not simply connected , so eq . ( [ theorem ] ) will not apply .
thus we need instead to take the simply connected group @xmath31 where @xmath32 is the group of translations in 1d with elements satisfying the property @xmath33 ( which will be simply connected ) .
this is the covering group of @xmath34 . in this work
we will use the representation of @xmath35 by pauli matrices , thus we describe its elements in the form @xmath36 .
we then use the homomorphism @xmath37 given by @xmath38 which `` lifts '' our group acting directly on the spinor states into the larger simply connected group , @xmath22 . the subgroup @xmath39 will be the collection of elements @xmath40 which leave the particular spinor under consideration invariant .
once @xmath23 is constructed by such a method , eq .
( [ theorem ] ) can then be directly applied to construct the first homotopy group . as we will see , the majority of the phases in the spin - three phase diagram have only discrete symmetries .
for such states , @xmath25 is the trivial group containing one element .
for this situation , our computational theorem reduces to @xmath41 therefore , constructing @xmath42 for such states becomes straightforward : one first finds the rotations in @xmath43 which leave invariant the points on the unit sphere @xmath44 given by our classification scheme @xcite . for this purpose , it is helpful to note the possible finite subgroups of @xmath43 .
they are the cyclic groups @xmath45 , the dihedral groups @xmath46 , the tetrahedral group @xmath32 , the octahedral group @xmath47 , and the icosahedral group @xmath48 ( see table [ table ] ) .
one then lifts this subgroup of @xmath43 into the twice - as - large group @xmath35 .
on top of this `` spinor scaffolding '' one can impose a superfluid winding @xmath49 , but the allowed values of @xmath49 are related to the spin rotation chosen .
the winding is given in the form @xmath50 where the integer @xmath51 is arbitrary .
the offset @xmath52 is fixed and can be determined geometrically through the relation @xmath53 where @xmath54 is the angle of rotation , @xmath55 is the multiplicity of the spin - root on the axis of rotation , and @xmath4 is the total spin ( see appendix for proof ) .
this information on the superfluid phase determines the translation group elements @xmath56 .
this provides a direct way of constructing @xmath23 and therefore classifying the possible vortices which we illustrate below for the spin - three problem .
now we move on to discuss the possible phases of the spin - three condensate in the absence of an external magnetic field .
the interaction hamiltonian determined by the pair potential eq .
( [ pair - potential ] ) is given by @xmath57 where @xmath58 creates a zero - momentum boson in the spin @xmath59 state , @xmath60 are clebsch - gordan coefficients , and the sums over the greek indices are implicit . note that this interaction energy alone is sufficient to determine the spinor ground state since the rest of the hamiltonian will not have a spinor preference . using the variational state @xmath61 , where the coefficients are normalized @xmath62 the interaction energy to leading order in @xmath63
can be found by replacing @xmath64 by the scalar @xmath65 giving the interaction energy of @xmath66 where @xmath67 we numerically minimize this energy as a function of the seven complex variational parameters @xmath18 under the constraint of normalization @xmath62 for all possible values of the scattering lengths , and reproduce the phase diagram given in @xcite .
we give explicit forms of the wave functions up to @xmath43 rotation which are summarized in table [ table ] where , for comparison , we use the same notation as was used in @xcite .
a subtle difference is that we find the additional state hh which is degenerate with the phase h reported in @xcite .
it is important to note that the unspecified angles @xmath68 and @xmath69 will take unique values for particular regions in the phase diagram .
that is , there is no mean - field degeneracy as was reported for the spin - two case @xcite . [
cols="^,^,^,^",options="header " , ] .
the small white spheres are put at the origin as a visual aid .
points which correspond to degenerate roots are marked.,width=336 ] we now apply the classification scheme developed in @xcite to give the symmetries of the possible spinor condensates , and the results are given in fig .
[ fig : shapes ] .
shown are reciprocal spin - three spinors which are six points on the unit sphere which make the symmetries explicit .
all of the phases except for f and ff have no continuous symmetries . as was explained in @xcite , all of the scattering lengths for @xmath0cr
are known except @xmath70 .
this gives a line in parameter space of possible ground states for such a condensate where the phases a , b , and c are the candidate ground states of the @xmath0cr condensate .
we will now discuss the topological excitations of each phase .
for each case we will take @xmath71 which will act on the spinor states via the homomorphism @xmath72 .
the goal is to find the isotropy group @xmath23 which leaves the particular spinor state under consideration invariant .
we start with the two states having continuous symmetries , namely f and ff . for state ff , taking the state to be oriented along the @xmath73-axis , we find that the only operators leaving the spin state invariant are @xmath74 with @xmath1 .
note that the phase factor @xmath75 is necessary to cancel the phase from rotation . immediately
, we find that this gives the isotropy group @xmath76 the first homotopy group is then computed to be @xmath77 consistent with that in @xcite .
carrying through the similar analysis for state f , we find @xmath78 .
note that since these groups are abelian , the conjugacy classes are the original group elements giving .
the number of conjugacy classes for these cases are @xmath79 ( ff ) and @xmath80 ( f ) .
the number of possible types of vortex excitations is smaller by one here since the identity corresponds to the absence of a vortex . as noted
before , this is quite a nontrivial result .
this is in contrast to the spinless case where there are infinitely many types of vortices corresponding to different windings in the superfluid phase as the vortex is circled ( with @xmath81 ) .
the underlying reason for this is that rotating such a spinor along its symmetry axis is identical to changing the superfluid phase .
we now move on to consider the remaining phases which have only discrete symmetries .
for these situations , as mentioned before , @xmath25 is the trivial group , so we can apply eq .
( [ theorem_discrete ] ) directly .
we will start with the phases having the smallest isotropy groups , namely phases c and g. for phase g , with the spinor wave function written as in table [ table ] , the only symmetry is a rotation by @xmath82 about the @xmath73-axis . such an operation will not change the phase of the spinor , i.e. @xmath83 .
we then find for the isotropy group @xmath84 where @xmath85 is the @xmath86 identity matrix and we note that @xmath87 .
this group is abelian , so the conjugacy classes are just the group elements themselves .
note that the translational part of @xmath23 merely tells us that an arbitrary winding in the superfluid phase can be imposed on any vortex state .
_ in the following , we will suppress the translational part of @xmath23 _ , keeping this in mind .
the phase c will have a similar isotropy group , but we note here that rotation by @xmath82 about the symmetry axis will give a phase of @xmath88 .
phases e , h , and hh have @xmath89-fold rotational symmetries about a particular axis .
phases e and hh have a three - fold rotational symmetry about their symmetry axes , and the first homotopy group is found to be @xmath90 on the other hand , phase h has a five - fold rotational symmetry , and its first homotopy group is similarly @xmath91 as before , these groups are abelian
. therefore , the types of possible vortices will correspond to rotating the polyhedra about the axes of symmetries @xmath89 times as the vortex is circled .
in addition , the phase winds by @xmath92 where @xmath51 is an arbitrary integer and @xmath93 for the phases e , h , and hh respectively .
phase b is the first phase we encounter that has a nonabelian isotropy group .
this phase has a rotational symmetry of @xmath82 about the @xmath73-axis .
in addition , it also has the symmetry of a rotation about two axes in the @xmath94 plane , namely @xmath95 .
the isotropy group for this is @xmath96 since this group is nonabelian , the conjugacy classes are not just the original group .
this group is isomorphic to the quaternion group having eight elements .
the conjugacy classes corresponding to each possible type of vortex excitation are : @xmath97 phases a and d have larger isotropy groups ; we will not write out explicit representations here , but will merely state the results . for phase a which has the symmetry of the hexagon as seen in fig .
[ fig : shapes ] there is a six - fold rotational symmetry about the @xmath73-axis .
in addition to this there are six axes in the @xmath94 plane for which there will be a two - fold rotational symmetry .
this will make @xmath98 have @xmath99 elements , which will have @xmath100 conjugacy classes .
phase d has the symmetry of the octahedron . for this
, @xmath101 will have @xmath102 elements and @xmath103 conjugacy classes .
all of these results are summarized in table [ table ] .
now that the classification scheme for vortices based on homotopy theory has been completed , we will briefly consider the resulting microscopic wave functions .
we can express the order parameter explicitly in terms of the spinor as @xmath104 and write @xmath105 .
note that the amplitude of @xmath106 is taken to have constant magnitude which is realistic provided we are sufficiently far from the vortex core .
for @xmath107 it will prove useful to use the spinor - ket notation and write @xmath108 where @xmath109 is the reference state which can be taken as one of the spinors given in table [ table ] .
in this equation , @xmath110 is the axis about which the spinor is rotated as the vortex is circled and @xmath111 give the spatial dependence of the superfluid phase and spinor rotation around the vortex .
these functions must satisfy the requirement that the original wave function must be recovered after the vortex is completely circled .
for instance , suppose the spinor state under consideration is invarient under a rotation by angle @xmath54 .
then a vortex at the origin corresponding to this symmetry is given by @xmath112 and @xmath113 where @xmath72 is the azimuthal angle , @xmath51 is the winding number , and @xmath52 is given by eq .
( [ eq : lambda ] ) .
this work was supported by the nsf grant dmr-0132874 and the harvard - mit cua .
rb was supported by the sherman fairchild foundation .
we thank r. diener , t .- l . ho , and g. refael for useful discussions .
note added : after this work was completed , we became aware of a work by s .- k .
yip @xcite which considers a similar problem of vortices in spinor becs , but uses a different approach . |
we would like to express our gratitude to t. yanagida for making a chance to start this collaboration .
we thank k. kanaya for his help in preparing the manuscript .
the calculations were performed with ha - pacs computer at ccs , university of tsukuba and sr16000 at kek .
we would like to thank members of ccs and kek for their strong support for this work .
m. hayakawa , et al .
, pos lattice * 2010 * ( 2010 ) 325 .
m. a. clark and a. d. kennedy , phys .
* 98 * ( 2007 ) 051601 .
miransky , phys .
* d59 * ( 1999 ) 105003 ; l. del debbio and r. zwicky , phys . rev . *
d82 * 014502 .
ishikawa , y. iwasaki , yu nakayama and t. yoshie , to appear . | we give a new perspective on the dynamics of conformal theories realized in the su(@xmath0 ) gauge theory , when the number of flavors @xmath1 is within the conformal window .
motivated by the rg argument on conformal theories with a finite ir cutoff @xmath2 , we conjecture that the propagator of a meson @xmath3 on a lattice behaves at large @xmath4 as a power - law corrected yukawa - type decaying form @xmath5 instead of the exponentially decaying form @xmath6 , in the small quark mass region where @xmath7 : @xmath8 is the mass of the ground state hadron in the channel @xmath9 and @xmath10 is a constant of order 1 .
the transition between the `` conformal region '' and the confining region " is a first order transition .
our numerical results verify the predictions for the @xmath11 case and the @xmath12 case in the su(@xmath13 ) gauge theory with the fundamental representation .
conformal field theories are ubiquitous in nature and play important roles not only in particle physics beyond the standard model but also in condensed matter physics .
nonperturbative understanding of their dynamics in four - dimensional space - time is ardently desired . in this article
, we give a new perspective on the dynamics of conformal theories realized in the su(@xmath0 ) gauge theory when the number of flavors @xmath1 is within the conformal window @xcite by studying meson propagators with a finite infrared ( ir ) cutoff .
some preliminary results have been presented in @xcite .
our general argument that follows can be applied to any gauge theories with arbitrary representations as long as they are in the conformal window , but to be specific , we focus on su(3 ) gauge theories with @xmath1 fundamental fermions ( quarks " ) .
we define the conformal field theory in a constructive way @xcite : we employ the wilson quark action and the standard one - plaquette gauge action on the euclidean lattice of the size @xmath14 and @xmath15 with aspect ratio @xmath16 .
we impose periodic boundary conditions except for an anti - periodic boundary condition in the time direction for fermion fields .
we eventually take the continuum limit by sending the lattice space @xmath17 with @xmath18 keeping @xmath19 fixed .
when @xmath20 finite , the continuum limit defines a theory with an ir cutoff .
the theory is defined by two parameters ; the bare coupling constant @xmath21 and the bare degenerate quark mass @xmath22 at ultraviolet ( uv ) cutoff .
we also use , instead of @xmath21 and @xmath22 , @xmath23 and @xmath24 . for a later purpose
, we define the quark mass @xmath25 through ward - takahashi identities with renormalization constants being suppressed @xcite .
let us quickly remind ourselves of the renormalization group ( rg ) flow when @xmath1 is in the conformal window .
one important fact is finite size lattices in computer simulation always introduce an ir cutoff @xmath26 .
if the ir cutoff were zero , when quarks have tiny masses , the rg trajectory would stay close to the critical line , approaching the ir fixed point and finally would pass away from the ir fixed point to infinity .
therefore the ir behavior is governed by the `` confining region '' . only on the massless quark line the scale invariance
is realized at the ir fixed point .
see the left panel of fig . 1 .
when the cutoff @xmath2 is finite , the rg flow from uv to ir does stop evolving at the scale @xmath2 . when the typical mass scale ( e.g. that of a meson ) @xmath8 is smaller than @xmath27 , it is in the `` conformal region '' . on the other hand ,
when @xmath8 is larger than @xmath2 , the flow passes away from the ir fixed point to infinity with relevant variables integrated out , thus being in the `` confining region '' . see the right panel of fig . 1 .
this scenario implies that when physical quantities at ir ( e.g. hadron masses ) are mapped into a diagram in terms of physical parameters at uv ( e.g. the bare coupling constant and the bare quark mass ) , there will be gaps in the physical quantities along the boundary between the two phases .
there the phase transition will be a first order transition . to make the above rg argument more concrete
, we will study the propagator of the local meson operator @xmath28 when the theory is in the confining region " , it decays exponentially at large @xmath4 as @xmath29 due to the physical one - particle pole , where @xmath8 is the mass of the ground state hadron in the channel @xmath9 .
in contrast , we claim that in the conformal region " defined by @xmath30 where @xmath10 is a constant of order 1 which we will determine , the propagator @xmath31 behaves at large @xmath4 as @xmath32 which is a power - law corrected yukawa - type decaying form instead of the exponential decaying form ( eq.([exp ] ) ) observed in the confining region " .
we further claim that the boundary between the `` conformal region '' and the `` confining region '' is a first order transition .
we note that the behavior eq.([yukawa type ] ) is proposed based on the ads / cft correspondence with a softwall cutoff in the literature @xcite .
the meson propagator in the momentum space has a cut instead of a pole : @xmath33 .
the propagator in the position space ( after space integration ) takes the form eq.([yukawa type ] ) in the limit @xmath34 .
we distinguish @xmath35 in eq.([yukawa type ] ) from the pole mass @xmath8 in eq.([exp ] ) .
eq.([critical mass ] ) thus means the lower limit of @xmath8 is @xmath36 in the continuum limit with @xmath37 ( i.e. @xmath38 ) , the propagator on the massless quark line takes the form @xmath39 consistent with @xmath40 limit of eq .
( [ yukawa type ] ) .
if we take the coupling constant @xmath41 at the uv cutoff , @xmath42 takes a constant value , and the rg equation demands @xmath43 for the pseudo - scalar ( ps ) channel with @xmath44 being the anomalous mass dimension @xmath45 at @xmath46 .
the theory is scale invariant ( and shown to be conformal invariant within perturbation theory @xcite .
see also e.g. @xcite and references therein from ads / cft approach ) . when @xmath47 , @xmath42 depends slowly on @xmath4 as a solution of the rg equation . in the ir limit @xmath48
, we must retain @xmath49 .
the continuum limit with @xmath20 finite defines a continuum theory on @xmath50 which corresponds to a compact three - torus at finite temperature .
the ir cutoff @xmath51 is finite .
the propagator @xmath3 behaves at large @xmath4 as a power - law corrected yukawa type decaying form eq.([yukawa type ] ) .
the exponent @xmath42 in @xmath48 with @xmath52 takes the universal formula @xmath53 while with @xmath54 it takes a value depending on the dynamics , from which we can estimate the mass anomalous dimensions ( see @xcite ) .
[ exp_vs_yukawa ] [ expnf7nf16 ] now let us discuss the results of our numerical calculations .
we have performed simulations in the @xmath11 and @xmath12 cases , which we conjecture @xcite , are the boundaries of the conformal window .
the algorithms we employ are the blocked hmc algorithm @xcite for @xmath55 and the rhmc algorithm @xcite for @xmath56 in the case @xmath57 .
we specify the coupling constant @xmath58 for @xmath59 and @xmath60 for @xmath61 taking account of the fact that the ir fixed point for @xmath12 is @xmath62 in two - loop approximation .
we use the lattices of size @xmath63 and @xmath64 and vary the hopping parameter @xmath65 so that the quark mass takes the value from 0.40 to 0.0 : we simulate with 15 hopping parameters on the @xmath63 lattices , and 5 hopping parameters on the @xmath66 lattices .
we choose the run - parameters in such a way that the acceptance of the global metropolis test is about @xmath67 the statistics are 1,000 md trajectories for thermalization and 1,000 md trajectories or 500 md trajectories for the measurement .
we estimate the errors by the jack - knife method with a bin size corresponding to 100 hmc trajectories .
let us first discuss the results for the @xmath11 case on the @xmath63 lattices .
we define the effective mass @xmath68 by @xmath69 .
fig . 2 [ exp_vs_yukawa ] shows the @xmath4 dependence of the effective mass for the ps channel with three types of sources . on the left panel ,
we see the clear plateau of the effective mass at @xmath70 when quark mass is relatively large ; @xmath71 ( @xmath72 ) . on the other hand , on the right panel , we see the effective mass is slowly decreasing without no plateau up to @xmath73 when the quark mass is small ; @xmath74 ( @xmath75 ) , suggesting the power - law correction .
the effective masses for all cases with @xmath76 ( @xmath77 ) exhibit a similar behavior to the @xmath75 case .
we show the power - law corrected fit for the local - local data in the @xmath75 case with the fitting range @xmath78 $ ] on the left panel in fig .
3 .
the fit with @xmath79 reproduces the date very well .
in the case of @xmath12 , the effective masses on the @xmath63 lattices exhibit the power - law corrected yukawa - type decaying form for all cases when @xmath80 ( @xmath81 ) .
we show the power - law corrected fit for the local - local data in the @xmath82 case with the fitting range @xmath78 $ ] on the right panel in fig .
3 .
the fit with @xmath83 reproduces the data well .
we also report that the data on @xmath84 lattices for both cases of @xmath11 with @xmath85 and @xmath12 with @xmath81 clearly exhibit the power - law corrected yukawa type decaying form .
we obtain @xmath86 for @xmath75 in @xmath59 and @xmath87 for @xmath88 in @xmath89 , respectively .
thus we have confirmed that the propagators behaves as the power - law corrected yukawa type decaying form , when the quark mass is small enough .
we note in all cases the rho meson mass is degenerate with the pion mass within the one standard deviation .
this implies the phase is a chiral symmetric phase . in the remaining part of the article
, we try to identify the boundary between the `` conformal region '' and the `` confining region '' more precisely by investigating the transition region carefully with a small step of the value of @xmath65 for both of the @xmath11 and @xmath12 cases .
we present the results for @xmath25 and @xmath90 ( or @xmath91 ) in the @xmath11 case in fig . 4 .
we first note that the quark mass @xmath25 denoted by black points and line on the left panel is excellently proportional to @xmath92 in the whole region from @xmath93 to @xmath94 . for the propagators of the ps meson
, we observe the clear transition from the exponentially decaying form to the power - law corrected yukawa - type decaying form at @xmath95 . for @xmath96 , we present @xmath90 obtained from the fit to the exponentially decaying form with the fitting region [ 28:31 ] by green points and line .
for @xmath77 , we present @xmath91 obtained from the fit to the power - law corrected yukawa type decaying form with fitting range [ 15:31 ] by blue points and line .
the transition region is enlarged on the right panel in fig .
4 .
the values of @xmath90 and @xmath91 are different in the limit @xmath97 from smaller @xmath65 and from larger @xmath65 .
it should be noted that even if we had ignored the power - law corrections in yukawa type decay and had estimated the @xmath90 ( plotted in fig . 4 by black points ) by assuming exponential decay with the fitting range [ 28:31 ] , although no plateau is seen as shown on the right panel of fig 2 and in fig .
3 , there would have been a gap at the transition point .
it is worthwhile to stress that at @xmath98 and @xmath99 we observe two states , depending on the initial sate .
the existence of two states persists at least with 500 md trajectories .
we also note that the exponent @xmath42 does not vanish toward the transition point ; it rather increases up to @xmath100 thus the transition is a first order transition at @xmath101 these facts imply that there indeed exists a `` conformal region '' which is separated by a gap of physical quantities from the confining region " at finite quark mass .
the @xmath25 dependence of @xmath91 is rather complicated .
once @xmath91 becomes small at the transition point , it slightly increases as @xmath25 decreases , and then decreases at @xmath102 apparently the small @xmath91 region suffers from finite size effects . to verify the scaling relation @xcite for @xmath91 in terms of @xmath25 we have to control finite size effects . taking @xmath103 with @xmath104 ( @xmath105 ) , as the critical mass in eq.(3 )
, we estimate @xmath106 with our working definition of @xmath107 . in the @xmath12 case
we have observed similar results .
the @xmath25 dependence of @xmath91 is similarly complicated .
the transition occurs at @xmath108 with @xmath109 ( @xmath110 ) from which we estimate @xmath111 in the cases of @xmath84 lattices for both @xmath11 and @xmath12 cases , we have also observed the transition : in @xmath11 , we find exponential decay for @xmath112 ( @xmath113 ) and power - law corrected yukawa type decay for @xmath114 ( @xmath85 ) . in @xmath12 , we find power - law corrected yukawa type decay for @xmath80 ( @xmath81 ) .
the critical mass decreases compared with the @xmath63 lattices as expected since the ir cutoff decreases .
we conclude the paper with two main goals achieved in our study . we have given evidence that ( 1 ) : the meson propagator shows a transition from an exponentially decaying form to a power - law corrected yukawa - type decaying form at the critical hadron mass whose value is specified by the ir cutoff .
( 2 ) : the transition is first order . the existence of the conformal region we proposed gives a necessary ( but not sufficient ) condition for the conformal window , and the result for @xmath11 is consistent with our conjecture that @xmath11 is within the conformal window .
however we understand that we need more evidence to conclude that . in an accompanying publication @xcite
, we will discuss the physical interpretation of the power - law corrections in terms of the dynamics of the underlying conformal field theory by taking the continuum limit .
we also hope to report the physical applications of our findings in the near future . |
let @xmath0 be a fixed probability distribution on the real line such that @xmath1 and for any @xmath2 let @xmath3 denote the associated hermitian wigner matrix of size @xmath4 .
this means that @xmath5 where @xmath6 is a collection of real random variables with distribution @xmath0 .
the second - order correlation function of the characteristic polynomial of the random matrix @xmath7 is defined by @xmath8 where @xmath9 .
we are interested in the asymptotic behaviour of @xmath10 as @xmath11 , for certain sequences @xmath12 , @xmath13 which will be specified below .
furthermore , the correlation coefficient of the characteristic polynomial of the random matrix @xmath7 is defined by @xmath14 in the special case where @xmath0 is the gaussian distribution with mean @xmath15 and variance @xmath16 , the distribution of the random matrix @xmath7 is the so - called gaussian unitary ensemble ( gue ) ( see forrester @xcite or mehta @xcite but note that we work with a different variance ) . in this case
, it is well - known that @xmath17 where @xmath18 and the @xmath19 are the monic orthogonal polynomials with respect to the weight function @xmath20 ( see chapter 4.1 in forrester @xcite ) .
thus , up to scaling , the @xmath19 coincide with the hermite polynomials ( as defined in szeg @xcite ) , and it is possible to derive the asymptotics of the second - order correlation function @xmath21 from the well - known asymptotics of the hermite polynomials ( see theorem 8.22.9 in szeg @xcite ) .
more precisely , one obtains the following ( well - known ) results ( see also chapter 4.2 in forrester @xcite ) : for @xmath22 and any @xmath23 , @xmath24 where @xmath25 , @xmath26 , and @xmath27 for @xmath28 and any @xmath23 , @xmath29 where @xmath30 , @xmath31 and @xmath32 denotes the airy function ( see abramowitz and stegun @xcite ) . by symmetry ,
a similar result holds for @xmath33 . the functions in ( [ sine - kernel ] ) and ( [ airy - kernel ] ) are also called the sine kernel and the airy kernel , respectively .
furthermore , it is well - known that the eigenvalues of a random matrix @xmath7 from the gue are distributed roughly over the interval @xmath34 $ ] . that is why the results ( [ bulk - gue ] ) and ( [ edge - gue ] ) are also said to refer to the bulk and the edge of the spectrum , respectively .
recently , gtze and ksters @xcite have shown that the result ( [ bulk - gue ] ) for the bulk is ( almost ) `` universal '' in the sense that it holds not only for the gue , but also ( with minor modifications ) for more general hermitian wigner matrices as introduced at the beginning of this section . more precisely , under the assumption ( [ moments - hermitian ] )
, we have @xmath35 for all @xmath22 and all @xmath23 , where @xmath36 , @xmath37 and @xmath38 are the same as in ( [ bulk - gue ] ) .
therefore , it seems natural to ask whether the result ( [ edge - gue ] ) for the edge can also be generalized to more general hermitian wigner matrices .
the main purpose of this paper is to answer this question in the affirmative .
more precisely , our first result is as follows : [ theorem - hermitian ] under ( [ moments - hermitian ] ) , we have @xmath39 for all @xmath23 , where @xmath40 and @xmath41 are the same as in ( [ edge - gue ] ) . [ corollary - hermitian ] under ( [ moments - hermitian ] )
, we have @xmath42 for all @xmath23 .
moreover , it turns out that similar results hold for real - symmetric wigner matrices .
let @xmath43 be a fixed probability distribution on the real line such that @xmath44 and for any @xmath2 let @xmath45 denote the associated real - symmetric wigner matrix of size @xmath4 .
this means that @xmath46 where @xmath47 is a collection of real random variables with distribution @xmath43 .
then , similarly as above , the second - order correlation function and the correlation coefficient of the characteristic polynomial of the random matrix @xmath48 are defined by @xmath49 and @xmath50 respectively , where now @xmath51 .
following the approach by gtze and ksters @xcite , ksters @xcite recently showed that under the assumption ( [ moments - symmetric ] ) , we have @xmath52 for all @xmath22 and all @xmath23 , where @xmath37 is the same as in ( [ bulk - gue ] ) , @xmath53 , and @xmath54 thus , we also have universality ( in the same sense as above ) in the bulk of real - symmetric wigner matrices . in the special case where @xmath43 is the gaussian distribution with mean @xmath15 and variance @xmath55 ,
the distribution of the random matrix @xmath48 is the so - called gaussian orthogonal ensemble ( goe ) ( see forrester @xcite or mehta @xcite but , again , note that we work with a different variance ) . in this case ,
( [ bulk - symmetric ] ) had been obtained previously by brzin and hikami @xcite .
our second result shows that the result ( [ bulk - symmetric ] ) admits an analogue for the edge of the spectrum , too : [ theorem - symmetric ] under ( [ moments - symmetric ] ) , we have @xmath56 for all @xmath23 , where @xmath57 , and @xmath58 [ corollary - symmetric ] under ( [ moments - symmetric ] ) , we have @xmath59 for all @xmath23 . in the special case of the goe , ( [ edge - symmetric ] )
can already be found in brzin and hikami @xcite .
it seems interesting to note that the functions @xmath60 and @xmath61 arising for the bulk of the spectrum are related by the identity @xmath62 and that the functions @xmath63 and @xmath64 arising for the edge of the spectrum are related by the analogous identity @xmath65 see also brzin and hikami @xcite .
( to check the latter identity , use the fact that the airy function @xmath66 satisfies the differential equation @xmath67 . ) also , observe that in all the cases previously mentioned , the precise choice of the underlying distribution @xmath0 or @xmath43 enters into the asymptotic behaviour of the second - order correlation function of the characteristic polynomial only as a multiplicative factor depending on the fourth cumulant .
thus , the results are essentially `` universal '' .
let us mention some related results from the literature .
it is well - known that at the edge of the spectrum of wigner matrices , we have universality also for the correlation function of the eigenvalues themselves ( see soshnikov @xcite ) . in contrast to that , for the bulk of the spectrum of wigner matrices , only partial results are available in this direction ( see johansson @xcite ) .
furthermore , in the special cases of the gue and the goe , several authors have investigated the averages of more general products and even ratios of characteristic polynomials ( see brzin and hikami @xcite , fyodorov and strahov @xcite , baik , deift and strahov @xcite , strahov and fyodorov @xcite , akemann and fyodorov @xcite , vanlessen @xcite , borodin and strahov @xcite ) .
even more , at least in the hermitian setting , some of these results have been shown to be `` universal '' in that they continue to hold ( with some modifications ) for the class of unitary - invariant ensembles .
for wigner matrices , however , less seems to be known in this respect .
* acknowledgement : * i would like to thank alexander soshnikov for the suggestion to study the problem at the edge of the spectrum .
to prove theorems [ theorem - hermitian ] and [ theorem - symmetric ] , we will start from the fact that for fixed @xmath23 , the exponential generating functions of the sequences @xmath68 and @xmath69 are given explicitly by @xmath70 ( see lemma 2.3 in gtze and ksters @xcite ) and @xmath71 ( see lemma 2.3 in ksters @xcite ) , respectively .
this fact opens up the possibility to study the asymptotic behaviour of the second - order correlation functions by evaluating appropriate contour integrals of their exponential generating functions .
in fact , this strategy was already used by gtze and ksters @xcite and ksters @xcite to obtain the above - mentioned results for the bulk of the spectrum . here
we carry out a similar analysis for the edge of the spectrum .
since it does not require any additional efforts , it seems convenient to evaluate the values @xmath72 for arbitrary @xmath73 and @xmath74 , where @xmath75 denotes a contour around the origin . by the foregoing , the case @xmath76 corresponds to the hermitian case and the case @xmath77 corresponds to the real - symmetric case .
we remark in passing that for a general parameter @xmath73 , it is not hard to see that the exponential generating function under consideration can be interpreted as that of the second - order correlation function of the characteristic polynomial of a random matrix from ( a rescaled version of ) the tridiagonal beta ensemble introduced by dumitriu and edelman @xcite with @xmath78 . for this interpretation
, one should set @xmath79 . for @xmath73 and @xmath23 ,
put @xmath80 in the next section , we will prove the following results : [ theproposition2 ] for any @xmath73 , @xmath74 and @xmath23 , we have @xmath81 [ theproposition3 ] for any @xmath82 and @xmath23 , we have @xmath83 where @xmath84 denotes the @xmath85-fold application of the given differential operator .
( for @xmath86 , consider the continuous extension of the right - hand side . )
it is straightforward to check that @xmath87 and @xmath88 ( for the second identity , use the fact that the airy function @xmath66 satisfies the differential equation @xmath67 . )
thus , in view of the preceding comments , it is immediate that theorems [ theorem - hermitian ] and [ theorem - symmetric ] , which correspond to the special cases @xmath76 and @xmath77 , follow from propositions [ theproposition2 ] and [ theproposition3 ] . to prove corollaries [ corollary - hermitian ] and [ corollary - symmetric ] , observe that @xmath89 in the hermitian case and @xmath90 in the real - symmetric case .
moreover , it is not difficult to see that @xmath91 where @xmath92 is the @xmath4th hermite polynomial ( see section 5.5 in szeg @xcite ) .
thus , in both cases , the expectation of the characteristic polynomial is given by the _ same _ function @xmath93 .
therefore , to deduce corollaries [ corollary - hermitian ] and [ corollary - symmetric ] from theorems [ theorem - hermitian ] and [ theorem - symmetric ] , respectively , it will be sufficient to show that @xmath94 is asymptotically negligible in comparison to @xmath68 and @xmath69 .
slightly more generally , we will consider the case of an arbitrary parameter @xmath73 and investigate the asymptotic behaviour of @xmath95 for any @xmath23 . in the next section
, we will show that proposition [ theproposition2 ] entails the following result : [ theproposition5 ] for any @xmath73 , @xmath74 and @xmath23 such that @xmath96 , @xmath97 , we have @xmath98 furthermore , we will prove the following : [ theproposition6 ] for any @xmath82 , we have @xmath99 for all @xmath100 . in view of the preceding comments , it is obvious that corollaries [ corollary - hermitian ] and [ corollary - symmetric ] , which correspond to the special cases @xmath76 and @xmath77 , follow from propositions [ theproposition5 ] and [ theproposition6 ] .
we remark in passing that for a general parameter @xmath73 , @xmath101 can be interpreted as the correlation coefficient of the characteristic polynomial of a random matrix from ( the rescaled version of ) the tridiagonal beta ensemble ( with @xmath78 ) , since in this setting , the average of the characteristic polynomial is also given by the hermite polynomial ( see theorem 4.1 in dumitriu and edelman @xcite ) .
fix @xmath23 . we have to evaluate @xmath102 where @xmath103 , @xmath104 , and @xmath75 denotes a contour around the origin ( which will be chosen further below ) . setting @xmath105 and @xmath106 , ( [ intrep1 ] )
may be written as @xmath107 with @xmath108 in particular , the leading exponential factor in ( [ intrep2 ] ) is asymptotically the same as that in proposition [ theproposition2 ] .
similarly as in the proof of the main theorem in @xcite , the basic idea is that the main contribution to the integral in ( [ intrep3 ] ) comes from a small neighborhood of the point @xmath111 .
let @xmath112 ( where @xmath113 denotes the principal branch of the logarithm ) and @xmath114 then the left - hand side in ( [ intrep3 ] ) can be rewritten as @xmath115 put @xmath116 and @xmath117 , where @xmath118 .
we shall prove the following : thus , for @xmath118 sufficiently large , we have not only the conclusion of claim 2 , but also the inequality @xmath126 hence , in combination with claim 1 , it follows that @xmath127 for all @xmath124 sufficiently large .
since @xmath120 is arbitrary , this proves ( [ intrep3 ] ) .
_ proof of claim 1 : _ first of all , substituting @xmath128 , we have @xmath129 we will determine the asymptotics of @xmath130 as @xmath11 , for .
the @xmath131-bounds occurring in the sequel hold uniformly in this region . to begin with , note that @xmath132 therefore , for @xmath4 sufficiently large , we have the following approximations : @xmath133 @xmath134 @xmath135 @xmath136 @xmath137 @xmath138 putting these approximations together , the terms of highest order cancel out , and we end up with @xmath139 hence , by an application of the dominated convergence theorem , it follows that @xmath140 and claim 1 is proved .
_ proof of claim 2 : _ by symmetry , it suffices to consider the interval @xmath141 .
write the integral as @xmath142 since @xmath143 and @xmath144 are real and @xmath145 , this is clearly bounded by @xmath146 in the following , let @xmath147 denote constants which depend only on @xmath148 ( which are regarded as fixed ) but which may change from occurrence to occurrence
. then we have @xmath149 and @xmath150 ( as follows from a straightforward calculation ) and therefore @xmath151 since @xmath152 it follows that @xmath153 for some constant @xmath154 , say .
now let @xmath155 denote a constant such that @xmath156 for all @xmath157 $ ] . then
, for @xmath158 , @xmath159 and @xmath160 $ ] , we have @xmath161 and therefore @xmath162 thus , since @xmath163 , the integral under consideration is bounded by @xmath164 obviously , this upper bound can be made arbitrarily small by picking @xmath165 and @xmath4 large enough .
this proves claim 2 .
we start from the following well - known integral representation of the airy function : @xmath172 a standard application of cauchy s theorem shows that the contour @xmath168 can be deformed into the contour @xmath173 , @xmath174 .
thus , we obtain @xmath175 observe that the resulting integral exists in the lebesgue sense , since we have @xmath176 for any @xmath174 .
it follows from ( [ ai - int ] ) that @xmath177 substituting @xmath178 and doing a small calculation , we find that @xmath179 using the well - known relation @xmath180 ( where @xmath181 ) , it follows that @xmath182 or @xmath183 by another application of cauchy s theorem , the contour @xmath184 , @xmath185 , may be deformed back into the contour @xmath168 . replacing the contour @xmath168 in lemma [ airyproduct ] by the contour @xmath173 and substituting @xmath186
, we have @xmath187 by means of abbreviation , write @xmath188 for the numerator inside the integral . then , for any @xmath73 , we have @xmath189 @xmath190 and therefore @xmath191 the assertion of proposition [ theproposition3 ] now follows by induction .
fix @xmath23 , and put @xmath103 , @xmath104 . using well - known results about the asymptotic properties of the hermite polynomials ( see theorem 8.22.9(c ) in szeg @xcite ) , we find that the function @xmath93 given by @xmath192 ( @xmath193 ) satisfies @xmath194 for any @xmath73 .
setting @xmath195 , we therefore obtain @xmath196 where we have used proposition [ theproposition2 ] as well as the assumptions @xmath96 , @xmath197 .
this completes the proof of proposition [ theproposition5 ] .
first of all , note that the definition ( [ idefinition ] ) may be extended to the case @xmath198 and that @xmath199 for any @xmath100 by lemma [ airyproduct ] .
thus , @xmath200 for any @xmath100 , with strict inequality for @xmath201 ( since it is well - known that the airy function does not have any zeroes on the positive half - axis ) .
moreover , note that for any @xmath73 , @xmath202 for any @xmath100 .
since @xmath203 , this implies that for any @xmath73 , @xmath204 for any @xmath100 .
proposition [ theproposition6 ] now follows by a straightforward induction on @xmath85 . | we investigate the asymptotic behaviour of the second - order correlation function of the characteristic polynomial of a hermitian wigner matrix at the edge of the spectrum .
we show that the suitably rescaled second - order correlation function is asymptotically given by the airy kernel , thereby generalizing the well - known result for the gaussian unitary ensemble ( gue ) .
moreover , we obtain similar results for real - symmetric wigner matrices . |
extended red emission ( ere ) is a widely - observed interstellar photoluminescence phenomenon in the 500 nm to 900 nm spectral range and is seen in many environments where both dust and ultraviolet ( uv ) photons are present ( see witt & vijh ( 2004 ) for a recent review ) . in reflection nebulae ( rn )
( witt et al . 1984 ; witt & boroson 1990 ) , hii regions
( perrin & sivan 1992 ) and planetary nebulae ( pn ) ( furton & witt 1990 , 1992 ) the sky background is faint compared to the relatively high overall intensities of the ere and can be readily subtracted .
however , given the broadband nature of the ere profile , a persistent problem in observations of the ere is the separation of the ere from the normally much more dominant continuum radiation resulting from scattering by dust in the same spectral range in rn ( e.g. witt & boroson 1990 ) or from atomic recombination emission in pn ( e.g. furton & witt 1990 , 1992 ) , or both , as in the case of hii regions ( e.g. perrin & sivan 1992 ) .
observations have shown that the ratio of ere to scattered light intensities is highly variable from object to object as well as within individual objects ( witt & boroson 1990 ; witt & malin 1989 ) .
witt & boroson ( 1990 ) showed for the case of reflection nebulae that this large variation is predominantly due to the variations in the intensity of the scattered light , while ere , when present , exhibits more nearly constant intensities by comparison .
this is most readily explained , if the ere results from an isotropic emission process , while the scattered light intensity can vary greatly for a constant density of the illuminating radiation simply as a result of differences in the illumination geometry , combined with the strongly asymmetric , forward - directed scattering phase function of interstellar dust .
therefore , a great advantage arises for ere detection , when illumination geometries can be found where the dust - scattered light is largely directed away from the earth by virtue of the forward - directed phase function of scattering of interstellar dust .
such conditions exist in the red rectangle nebula ( cohen et al .
2004 ; vijh et al .
2006 ) , where the presence of an optically thick dust torus seen almost edge - on directs most of the dust - scattered light in directions perpendicular to the line of sight to earth . in retrospect , it is easy to see why ere was first detected in this object ( cohen et al . 1975 ; schmidt , margon , & cohen 1980 ) .
similarly favorable geometric conditions exist in high - latitude interstellar clouds , which are illuminated by the integrated radiation field of the milky way galaxy .
the intensity distribution of the galactic radiation field exhibits a pronounced peak in directions close to the galactic equator . a high - latitude cloud at galactic latitude @xmath3 receives its most intense illumination under an angle equal to @xmath3 or greater with respect to the line of sight .
consequently , the forward - directed phase function sends most of the scattered light into directions other than the line of sight to earth ( see figure 5 ) .
therefore , when present in such clouds , the ere can appear enhanced compared to the underlying scattered light continuum , because the isotropic emission pattern of the ere does not discriminate against any particular line of sight and the ere excitation process responds to incident radiation from all direction equally .
the first tentative identification of ere in the spectrum of a high - latitude dark nebula was made by chlewicki & laureijs ( 1987 ) , based on the previously published surface brightness spectrophotometry of ldn 1780 by mattila ( 1979 ) .
subsequently , guhathakurta & tyson ( 1989 ) determined the optical surface brightness and colors of small portions of the high - latitude counterparts to the galactic 100 cirrus detected by iras ( low et al .
they found these to exhibit colors significantly redder than expected for dust clouds scattering the integrated galactic radiation field .
they attributed this red color excess to ere .
similar results were reported by guhathakurta & cutri ( 1994 ) , based on the analysis of extensive optical imaging data of several high - latitude clouds .
gordon et al .
( 1998 ) conducted the first large - scale survey ( @xmath4 ) designed to detect ere at high galactic latitudes , using photometric all - sky data from pioneer 10 and pioneer 11 obtained outside the zodiacal dust belt of the solar system .
while providing data for large sections of the high - galactic latitude sky on the one hand , this survey suffered from the relatively large instantaneous field of view of the pioneer photometers of @xmath5 .
as a consequence , faint background stars contributed substantially to the measured signal , requiring deep star counts to correct for this stellar contribution . also , the large field of view of the pioneer photometers made it impossible to image individual high - latitude clouds , which typically have angular scales comparable to or smaller than the pioneer diaphragm apertures .
the observations yielded integrated , star - corrected surface brightness data that were well - correlated with hi column densities from large - scale hi surveys and equally well - correlated with the iras 100 maps of the galactic cirrus .
the investigation of the optical galactic background radiation corrected for stellar light contribution resulted in the detection in a strong red excess over color predictions from a radiative transfer model for the high galactic latitude diffuse galactic light .
this interpretation was confirmed spectroscopically by szomoru & guhathakurta ( 1998 ) , who obtained spectra of several high - latitude clouds , which revealed the spectral signature of the ere and confirmed the ere intensities deduced by gordon et al .
( 1998 ) .
recently , zagury ( 2006 ) published a re - analysis of the measurements and analysis of gordon et al .
( 1998 ) and concluded that ere was not a likely component of the high - latitude diffuse galactic light .
the resolution of this controversy is of importance , because so far only the high - latitude observations of the ere have provided the basis for an estimate of the photon conversion efficiency with which uv photons are being converted into ere photons as well as an estimate of the relative abundance of the ere carrier . both are crucial constraints in the ongoing efforts to identify the carrier of the ere . here
we report first results from an ongoing imaging survey of high galactic latitude interstellar clouds ( hglic ) , which has been designed to overcome most of the shortcomings of earlier efforts to identify the ere in high - galactic latitude dust environments . with this work
we follow in the footsteps of struve & elvey ( 1936 ) and of mattila ( 1970 ) , who showed that the surface brightness of individual high - latitude dark nebulae , which mainly results from scattering of the integrated galactic star light , can indeed be measured , and of sandage ( 1976 ) , who showed that the full complexity of the high - latitude galactic cirrus can be captured through large - field optical imaging with fast ground - based telescopes .
our survey objectives included the following desirable characteristics : ( 1 ) a large field of view to permit the simultaneous imaging of entire clouds and the surrounding sky , making possible reliable background subtraction ; ( 2 ) sufficient angular resolution to allow nebular surface brightness measurements without contamination from faint stars ; ( 3 ) use of four well - placed filter bands to define the approximate shape of the spectral energy distribution ( sed ) of the clouds ; ( 4 ) focus on optically thin clouds to assure a simplified radiative transfer analysis ; ( 5 ) a sufficiently large sample of clouds for representative results ; ( 6 ) fast telescope optics to assure high diffuse - source sensitivity ; and ( 7 ) reliable absolute intensity calibration . the successful implementation of these essential factors into a pilot observing program will be described in 2 of this paper .
we describe the data reduction and the resulting seds of our objects in 3 of this paper .
this is followed in 4 with an analysis that includes radiative transfer considerations leading to predictions for the expected cloud seds on the assumption that scattering by dust of the integrated galactic radiation field is the sole source of the clouds optical surface brightness .
we will derive red excesses which we interpret as ere . in 5
we will discuss the larger significance of our result for the interstellar medium in general , followed by our conclusions in 6 .
the imaging observations for this paper were carried out with a takahashi fsq 106 f/5 refractor of 106 mm aperture , operated remotely at new mexico skies observatories ( nms ) near mayhill , nm , at an altitude of 7300 feet .
the telescope was equipped with a santa barbara imaging group stl-6303e ccd camera , covering a field of view of @xmath6 .
the angular resolution determined by the 9 pixels was 3.60 pixel@xmath7 .
the system was equipped with both internal and external automatic guiders and was capable of automatic self - focusing between sub - exposures , a procedure followed routinely .
the 50 - mm filters , consisting of four astro - don broadband and one custom scientific narrow - band h - alpha filters , were housed in a 5-position filter wheel .
the band pass characteristics of the four broad filters are summarized in table 1 , and the transmission curves are displayed in figure 1 .
two of the band passes ( g , r ) were sensitive to the presence of ere , while the remaining broadband filters ( b , i ) were selected to define the scattered - light continuum at either end of the ere band .
the narrow - band h - alpha filter was used to assess the contribution of large - scale h - alpha emission to the surface brightness of hglics detected in the r - band .
the entire system was mounted on a software bisque paramount me equatorial mounting and was operated remotely through the internet .
five hglic were selected for this pilot study .
the clouds are identified by their number in the mbm catalog ( magnani et al . 1985 , 1996 ) ; their galactic coordinates and lsr radial velocities are listed in table 2 .
the membership in the mbm catalog implies that these clouds have been detected by their co line emission , and the lsr radial velocities are based on co observations by magnani et al .
( 1985 ) . in today
s terminology these clouds would be classified as diffuse molecular clouds ( snow & mccall 2006 ) or translucent clouds .
our five clouds were selected from the large mbm sample to be optically thin in the b - band .
we will show that the clouds met this requirement in 3.1.1 below .
individual distances of hglic are generally quite uncertain , but clouds with v@xmath8 10 km s@xmath7 typically are part of the thin layer of galactic interstellar medium which defines the galactic plane . as a result ,
such clouds have average z - distances of @xmath975 pc ; at latitudes of @xmath10 the distances from the sun are in the range from 80 pc to 180 pc ( magnani et al . 1985 ) .
individual distances can be estimated by several techniques , among them reddening observations of stars as a function of distance in the line of sight , or observations of interstellar absorption lines .
penprase ( 1992 , 1993 ) derived a distance to mbm 30 of 110 pc@xmath1110pc while magnani & de vries ( 1986 ) estimated the distance to mbm 32 as @xmath12 275 pc .
consequently , these objects are among the closest individual molecular clouds in the milky way galaxy .
mbm 41a and mbm 41d are sections of the draco nebula .
we follow the designations introduced by herbstmeier et al .
( 1993 ) in defining the sub - regions a and d. the draco nebula with its components mbm 41a and mbm 41d is a larger complex of intermediate - velocity gas which appears to be falling towards the milky way plane . despite several intensive efforts to determine the distance to the draco cloud ( e.g. goerigk & mebold 1986 ; gladders et al . 1998 ; penprase et al .
2000 ) , its value remains uncertain .
it is likely that the draco cloud distance is larger than those of the other clouds in our sample and might be found in the range derived in the latest attempt by penprase et al .
( 2000 ) of 800 pc to 1300 pc .
even at these values , the distance remains small in comparison to the diameter of the galactic disk , which implies that the density of the incident galactic radiation field should still be similar to that experienced by the other clouds in our sample , which in turn should be close to that in the solar vicinity .
however , the lack of accurate distances makes the interpretation of star counts ( e.g. goerigk et al .
1983 for the mbm 41 clouds ) highly problematical , because they are seen against a star field in which the volume density declines approximately exponentially with distance , as shown by mebold et al .
this results in a strong degeneracy involving distances and optical depths for these clouds .
the imaging observations were conducted during photometric nights during the months of january through april , 2006 .
the data were accumulated in the form of individual sub - exposures of 900 second duration .
the dates of observations and the total integration times for the images used in this study are listed in table 3 .
in addition to the nebular observations , photometric observations of the absolute flux standard sao 90153 were obtained through the same system to permit absolute calibration of the resulting data .
similarly , extinction stars were observed through a wide range of air masses to permit the derivation of coefficients of atmospheric extinction in our four broad - filter band passes for the nms site .
figures 2 a - e display the r - band images of our five objects . in each object
nebular regions labeled 1s through 10s are identified where the nebular surface brightnesses were measured ; similarly , regions labeled 1 through 3 indicate the locations where measurements of the sky background were obtained .
the cloud and sky regions were selected to be free of stars as faint as m @xmath9 21 mag and their integrated counts were measured with @xmath13 pixels ) apertures .
the sky - subtracted nebular counts were corrected for atmospheric extinction and converted into absolute intensities , using the calibration data obtained from the photometry of the absolute flux standard star sao 90153 .
we also obtained far - infrared iras 100 maps of our five hglics from the nasa / ipac infrared archive ( http://irsa.ipac.caltech.edu/applications/dust/ ) .
we identified our nebula and sky positions on these maps and determined the 100 intensity in units of mjy sr@xmath7 .
as has been noted before ( de vries & le poole 1985 ; guhathakurta & tyson 1989 ; paley et al .
1991 , guhathakurta & cutri 1994 ; zagury et al . 1999 ) , the detailed morphology of the hglics as seen at far - ir wavelengths are closely matched by the observable morphology at optical wavelengths , in particular in the r - band .
the reason for this is that most hglics are optically thin at optical wavelengths as they are at @xmath14 , and the intensities in both bands result from dust - related scattering or emission processes .
we confirmed the close similarity of appearance of our five clouds in the far - ir and optical bands ; as an example , we show a superposition of the iras 100 isophotes on the g - band image of mbm 32 in figure 3 .
the close similarity between the far - ir and optical morphologies is further supported by the linear relationship between the sky - subtracted nebular intensities at optical and far - ir wavelengths , as discussed in 3.1.1 .
we also established that the fields identified as _ sky _ are by no means dust - free but exhibit an easily detected residual surface brightness in the far - infrared .
the iras @xmath14 map clearly reveals that what appears as isolated clouds in the mbm co survey are merely density enhancement in an environment characterized by density fluctuations , with a typical enhancement factor of the order of three for a region defined as a _ cloud _ over the surrounding _ sky _ field accessible within the @xmath6 fields of view of our instrumentation .
the co morphologies plus their associated small - scale structure ( e.g. sakamoto 2002 ) mainly reflect localized volumes where the density and total shielding in the 912 - 1118 wavelength range , which contains co - destroying photons , are sufficiently high to permit co to survive ( van dishoeck & black 1988 ) .
our photometric measurements are therefore the result of differential photometry , in which we subtract not only a common foreground intensity stemming from atmospheric air glow and zodiacal light but also a substantial component of interstellar dust - scattered light and possible ere arising in the non - zero dust column density associated with our sky positions .
the averaged ( over 10 nebular positions ) , sky - subtracted optical intensities of our five hglics and the corresponding average infrared 100 intensities for cloud and sky regions are listed in table 4 . in the last two columns of table 4 we add the net iras 100 intensities of our clouds and the ratio of the total 100 intensities at the cloud and sky positions , providing a measure of the column density enhancement for the cloud over its nearby surroundings in the plane of the sky .
a basic assumption essential for our subsequent analysis is that our five hglics are optically thin in the b - band , and by inference , in the other optical bands employed in this study as well .
the averaging of the surface brightness measurements over ten nebular areas ( table 4 ) , which is necessary in order to overcome the noise from the dominant diffuse foreground sources ( air glow and zodiacal light ) , can only be justified , if the clouds are optically thin .
the same precondition applies to the use of a single - scattering approximation for the radiative transfer ( 3.1.2 ) .
finally , if ere is present in our hglics , it is excited most likely in the far - uv spectral region ( witt et .
al . 2006 ) .
consequently , the ere would arise predominantly near the cloud surfaces , where most of the interstellar far - uv radiation is absorbed .
only clouds that are optically thin in the visual spectral range would permit the ere from the far side to be fully transmitted , thus significantly enhancing the detection probability .
the determination of optical depths of hglics is difficult . given the uncertain distances of hglics , star counts against clouds embedded in a star field with exponentially declining volume density are practically useless ( mebold et al . 1985 ) .
hi column densities through hglics , coupled with a standard gas - to - dust ratio ( bohlin , savage , & drake 1978 ) , at best can yield a lower limit to the optical depth , because an unknown , possibly substantial , fraction of hydrogen is likely in molecular form in clouds where co is beginning to be stable ( burgh et al .
the far - ir intensity is generally the best direct measure of the dust column density through hglics , provided it can be calibrated reliably ( schlegel et al .
1998 ; schnee et al .
2005 , 2006 ) .
the relationship between the 100 intensities and the optical depths of hglics has been studied extensively .
schlegel et al . (
1998 ) have suggested a linear relation between the iras 100 intensities and the corresponding e(b - v ) , which is being used widely to estimate dust opacities at high galactic latitudes .
this relationship is based on the assumption of a constant dust temperature along sight lines . however , as shown by cambresy et al .
( 2001 ) , the assumption of a single linear scaling factor is not supported by data for the polaris molecular cirrus , where they find a@xmath15 values based on star counts that are smaller by factors of two to three compared to the values predicted by schlegel et al . ( 1998 ) .
it is not clear whether this disagreement is a consequence of the known problems with star counts for hglics .
even if all hglics are exposed to the same interstellar radiation field , one would not expect the relationship between iras 100 intensities and optical opacities to be linear , because the far - infrared emissivity depends sensitively upon the grain temperature , which is expected to be lower in portions of clouds which exhibit sufficient column densities of dust to produce some degree of self - shielding against more energetic uv irradiation ( bethell et al .
also , denser , more opaque clouds may experience grain growth , leading to a slightly lower grain emissivity at 100 with resulting lower grain temperature in contrast to grains in less dense , more optically thin clouds .
this is indeed borne out by more recent observations ( e.g. del burgo et al .
2003 ; lethinen et al .
2007 ) .
boulanger & perault ( 1988 ) correlated the iras 100 intensities and the integrated hi column densities over the galactic polar caps and found an average i(100 ) /a@xmath15 ratio of 15.9 mjy sr@xmath7 mag@xmath7 for a standard a@xmath15/n@xmath16 ratio of @xmath17 mag @xmath18 ( bohlin , savage , & drake 1978 ) . a similarly high ratio of i(100 ) /a@xmath15
ratio @xmath9 18.7 mjy sr@xmath7 mag@xmath7 was found by moritz et al .
( 1998 ) , who analyzed the x - ray shadow of the draco cloud , of which our targets mbm 41a and mbm 41d are parts .
the absorption of x - rays by the gas in foreground clouds has the advantage that both atomic as well as molecular gas is being sampled , as was done by bohlin et al .
the investigations of boulanger & perault ( 1988 ) and by moritz et al .
( 1998 ) are subject to the same uncertainty , stemming from the assumption of a constant gas - to - dust ratio throughout the galactic interstellar medium .
in fact , moritz et al . (
1998 ) suggest that the gas - to - dust ratio of bohlin et al .
( 1978 ) is probably too large by a factor of two for the draco cloud .
this is supported by optical observations of stars seen through the draco cloud ( e.g. penprase et al .
2000 ) , which lead to higher visual extinctions than implied by the total hydrogen column density .
other studies , which rely on a@xmath15 values derived from star counts or reddening observations , also find lower ratios of i(100 ) /a@xmath15 than deduced by boulanger & perault ( 1988 ) and by moritz et al .
. however , for a@xmath15 values based on star counts to be reliable , the optical depths of the associated clouds must be relatively high , the distances of the clouds need to be known , and the i(100 ) values are also much higher than those measured in our five target hglics .
for example , boulanger et al .
( 1998 ) studied the chamaeleon molecular clouds , where a@xmath15 values up to 2.5 mag are typical , associated with infrared intensities of up to 17 mjy sr@xmath7 .
depending on the optical depths of individual cloud portions , boulanger et al .
( 1998 ) found ratios of i(100 ) /a@xmath15 from @xmath95 mjy
sr@xmath7 mag@xmath7 for the densest clumps to @xmath910.5 mjy sr@xmath7 mag@xmath7 in more transparent regions .
similar values are found by de vries & de poole ( 1985 ) and by zagury et al .
( 1999 ) from the analysis of individual high - latitude clouds .
the lack of agreement between these results fully supports recent conclusions of schnee et al .
( 2005 , 2006 ) , who warn about the inherent uncertainties in deriving dust column densities from iras 100 surface brightnesses .
we conclude from these studies that for optically thin clouds as represented by our hglic sample a ratio of i(100 ) /a@xmath19 10 mjy sr@xmath7 mag@xmath7 is a fair , albeit uncertain estimate .
this conclusion is also supported by the recent investigation of ldn 1642 by lethinen et al .
( 2007 ) . fortunately , for the purpose of ere detection , _ the precise values of the optical depths of our hglics need not be known , provided they are entirely in the optically thin regime .
_ we are able to establish this empirically by plotting measured b - band net surface brightnesses of hglics against the corresponding iras 100 intensities at identical positions .
we are presenting such a plot in figure 4 for mbm30 , mbm32 , mbm41a , and mbm41d , where the iras 100 data are measured relative to a sky reference level of 0.65 mjy sr@xmath7 .
the relationship is linear and these four clouds obey the same relationship .
since these clouds are optically thin at 100 , this shows that they are also optically thin in the b - band , and that these clouds are exposed to the same effective radiation field .
de vries and poole ( 1985 ) have shown that the b - surface brightness of hglics begins to saturate for a@xmath20 mag .
this implies that the optically thin regime extends to about a@xmath19 0.7 mag , or @xmath21 0.83 .
figure 4 shows that our clouds are well within this linear , optically - thin regime .
this is reflected as well by the average optical depths determined for the individual clouds in table 5 . after converting the optical intensity units shown in figure 4 into mjy sr@xmath7 ( leinert et al .
1998 ) , we find @xmath22 = ( 2.13\pm0.10)\times 10^{-3 } ( i_{100 } - 0.65 ) \left[mjy\ sr^{-1}\right ] .\ ] ] the slope in eq .
1 agrees closely with a similar value found by zagury e al .
( 1999 ) , who derived a slope of @xmath23 for the cloud mcld123 + 24.9 , although those authors attributed the illumination of that cloud to the star polaris .
mbm 25 does not follow this relationship ( eq .
1 ) but rather a linear relation with a slope approximately twice as steep .
this indicates that mbm 25 is also optically thin but is illuminated by a radiation field which is either more intense or is scattered more effectively toward earth . to avoid confusion ,
however , the data for mbm 25 are not included in figure 4 .
we estimate average values of the net extinction of a@xmath24 mag deduced from the ( cloud - sky ) 100 intensities found for our five hglics in table 4 , with a spread of a factor two .
this is supported in general by the study of larson & whittet ( 2005 ) who determined the extinction to over 100 stars located behind high - latitude translucent clouds .
the peak in their distribution of extinctions is found at a@xmath25 0.5 mag .
thus , we conclude that the radiative transfer in our objects occurs under optically thin conditions and that multiple scattering and significant self - absorption are not important in our objects . with ere detected in other environments at wavelengths longer than 500 nm , only our measured intensities in the b - band can be assumed to be dominated by dust - scattered light , generated through external illumination by the galactic interstellar radiation field .
the subsequent section will therefore be restricted to estimating the expected surface brightness of our hglics in the b - band via a number of approximations with increasing complexity
. a critical component of any model prediction of hglic intensities is the density and directional distribution of the galactic interstellar radiation field ( irf ) in the solar vicinity .
we used the recent determination of the irf by porter and strong ( porter & strong 2005 ; moskalenko et al .
we are grateful to troy porter for providing us with data regarding the directional distribution of the intensity of the irf at the galactocentric radius of 8.5 kpc , both for a location in the plane ( z = 0 pc ) and for locations appropriate for many hglics , i.e. z = 100 pc . in figure 5
we show an aithoff projection of the radiation field for z = 100 pc for the wavelength of our b - filter , centered on the direction to the galactic center .
the positions of the hglics included in this study are indicated on this map , with mbm 41a representing both mbm41 regions .
porter & strong find a radiation density of u@xmath26 , corresponding to an average intensity of j@xmath27 for the b band .
these values are in near perfect agreement with the earlier determination of the irf density in the solar vicinity by witt & johnson ( 1973 ) and are only about 16% higher than the predictions for the average intensity of the irf by mathis et al .
the latter irs intensities were known to be too low to fully account for the observed far - infrared emission from the milky way galaxy ( mathis 1998 ; gordon et al .
1998 ) , a deficiency corrected in the recent self - consistent calculation of the irf by porter and strong .
early studies of the surface brightness of hglics ( sandage 1976 ; guhathakurta & tyson 1989 ) employed a single - scattering approximation in the optically - thin limit , assuming isotropically scattering grains with an albedo @xmath28 = 0.6 , to estimate the expected surface brightness @xmath29 of hglics .
thus , @xmath30 where @xmath31 is the line - of - sight optical depth through a cloud at the effective wavelength of our b - filter .
assuming @xmath31 = 0.5 and the radiation density @xmath32 from porter and strong , this approximation yields @xmath33 .
a comparison with our results in table 4 shows that this estimate exceeds the intensities actually observed in the b - band by factors of two to three .
while some of our clouds have values of @xmath34 , we attribute the higher - than - observed intensity primarily to the failure of the isotropic scattering assumption , given that it is well established that the phase function of interstellar grains is strongly forward directed ( gordon 2004 ) . in a high - latitude cloud
, an isotropic phase function is unable to respond correctly to the highly non - isotropic irf in which the illumination comes primarily from a ring - shaped source following approximately the galactic equator .
the flux from this principal source is incident on a hglic at galactic latitude @xmath3 under scattering angles of @xmath35 , and is thus scattered mostly not toward the earth .
a second contributing factor results from the fact that @xmath31 = 0.5 is already approaching the limits of the optically thin regime , which ignores all effects of self - shielding .
the latter limitation can be overcome by employing the exact radiative transfer solution for isotropic scattering for an externally illuminated cloud provided by van de hulst ( 1987 ) , which takes self - shielding and multiple scattering into account .
3 presents the bond albedo a for spherical clouds over a wide range of optical depths and grain albedos @xmath28 . for an optical diameter = 0.5 and a dust albedo @xmath28 = 0.6 , we interpolate to find @xmath36 = 0.18 .
the definition of the bond albedo employed by van de hulst then leads to @xmath37 where @xmath38 is the average intensity of the irf , averaged over the entire sky . with @xmath38
taken from porter and strong , we find @xmath39 .
this is much closer to the observed intensities but still somewhat higher than the average of the observed values .
the actual , forward - directed scattering phase function of interstellar dust grains responds most directly to the line - of - sight intensity of the irf , which is less than the average @xmath40 for hglics observed from a location near the galactic plane .
this may explain the remaining difference between observed and predicted intensities .
witt ( 1985 ) published a simple single - scattering approximation for the scattered intensity including self - shielding , which should work equally well , provided we remain near the optically thin limit .
for an externally illuminated cloud immersed in a radiation field with average intensity @xmath38 we find from witt ( 1985 ) @xmath41 where @xmath42 is the average absorption ( in contrast to extinction ) optical depth experienced by incident photons before their first scattering .
for isotropically incident radiation we approximate @xmath43 . with the same parameters as used in the previous examples we find @xmath44 , very similar to the result obtained from the exact solution by van de hulst ( 1987 ) for the same case , as well as relatively close to the observed intensities .
( 4 ) , as was the case with the other approximations , fails to take into account the forward scattering nature of interstellar grains .
this weakness was overcome in the approximation presented by stark ( 1995 ) , who treated the transfer of radiation through an externally illuminated hglic of @xmath45 with grains characterized by @xmath28 = 0.6 and a phase function asymmetry parameter @xmath46 = 0.75 , representing a strongly forward - directed phase function .
the most severely limiting approximation in this case was stark s assumption that the illumination source containing the irf was confined to a geometric sheet located below the cloud under investigation .
incident radiation is thus limited in origin to a hemisphere , with an average intensity of @xmath47 .
as expected , stark finds a strong dependence on galactic latitude in the amount of light back - scattered to the observer . interpolating his results for @xmath48 for @xmath49 ,
the reflected fraction of the average incident intensity is 0.033 .
thus , @xmath50 .
this value represents only about 50 % of the average b - intensities observed in our hglics .
this under - prediction of the intensity most likely is caused by the fact that the model ignores the non - zero illumination originating with the irf produced by sources in the upper hemisphere above the cloud .
a comparison of the simple models based on single scattering with an isotropic phase function with the model of stark , which includes forward scattering but assumes an unrealistic source distribution , suggests that the former produce better results , as long as the hglics considered are located at intermediate galactic latitudes , as is the case with our objects , and as long as self - shielding is included , as in eq .
( 4 ) .
the observed optical seds of hglics are expected to consist of two components , the scattered light continuum produced by scattering of the interstellar radiation field by the dust concentrated in the hglics , and the ere which is expected to be observable at wavelengths longer than 500 nm , if ere is present .
the presence of ere can be established by the comparison of the observed seds with an expected scattered light sed , yielding an excess in the red wavelength region with the usual spectral characteristics of the ere .
based on the considerations of sect .
3.1.2 , we have adopted the approach of witt ( 1985 ) , as expressed in eq .
( 4 ) , to predict the expected scattered light seds of hglics .
two specific adjustments were made , however .
first , to reflect the fact that our surface brightness measurements represent differential photometry measurements between nebular fields and background `` sky '' fields , which themselves have dust column densities roughly half of the corresponding nebular positions , we calculated the anticipated scattered light intensities for both sky and nebular fields separately and used the differences for comparison with our observed intensities .
second , we normalized the predicted scattered light seds at 454 nm to bring models and observations in agreement at a wavelength where the observed intensity is expected to consist of scattered light alone .
this normalization corrects for the fact that the scattering model does not include a forward - directed scattering phase function and also does not include the true directional distribution of the illuminating interstellar radiation field but uses the average intensity instead .
this normalization is justified by the facts that both the phase function asymmetry of the scattering dust grains ( draine 2003 ) and the geometry of the interstellar radiation field ( porter & strong 2005 ) are varying only slightly over the wavelength range covered by our observations . on the other hand
, our normalized models fully preserve the much more pronounced wavelength dependencies of the dust optical depth , the grain albedo , and the average intensity of the illuminating radiation field ( see table 6 ) .
accordingly , scattered light seds were calculated from the expression @xmath51 where @xmath52 is the albedo for interstellar dust with r@xmath15 = 3.1 from draine ( 2003 ) and @xmath53 is the average intensity of the galactic interstellar radiation field in the solar vicinity ( r = 8.5 kpc , z = 100 pc ) obtained from porter & strong ( 2005 ) . both quantities are listed for 15 wavelengths covering the spectral range of our filters in table 6 . the optical depths @xmath54 and @xmath55 are the line - of - sight optical depths through the hglics and their reference sky fields , respectively ; they are estimated from the corresponding 100 surface brightness found in the iras all - sky survey by using the relation @xmath56^{-1}$ ] for optically thin clouds , according to our discussion in sect .
the optical depths @xmath31 and @xmath57 are scaled as a function of wavelength according to the law of interstellar extinction for r@xmath15 = 3.1 , as given in table 6 . as in eq .
( 4 ) , we approximate the self - shielding absorption optical depth @xmath58 and @xmath59 .
this approximation reflects the fact that self - shielding only involves the non - scattered fraction , i.e. ( 1 - @xmath28 ) , of the incident light and that in externally illuminated clouds the average optical path for reaching a point of interaction is half the total optical depth through the cloud , assuming perfect symmetry .
the self - shielding absorption optical depths are scaled with wavelength by using the optical depth scale and the wavelength - dependent albedo values listed in table 5 .
the comparison of the observed surface brightness values and the predicted cloud seds based on scattering alone ( figures 6a - e ) shows that all five hglics exhibit a significant excess in the r filter , which is consistent with the presence of ere .
the ere spectra of the different clouds in our sample , which are obtained by subtracting the scattered - light sed from the observed surface brightnesses , show interesting differences , however . in three cases ( mbm25 , mbm41a and mbm41d ) , the ere intensity in the g- and r - bands
are nearly the same , suggesting a peak in the ere spectrum near 600 nm . in one cloud ,
mbm 32 , the excess is limited to the r - band alone , while in mbm 30 the highest ere intensity occurs in the g - band , followed by a lower intensity in the r - band .
these differences appear to be real , and they are consistent with the well - established fact that the peak of the ere band occurs at different wavelengths in different objects ( witt & boroson 1990 ; smith & witt 2002 ) .
further support for this conclusion can be found in the near - identity of the ere spectra for mbm 41a and mbm 41d .
the two data sets of these two separate parts of the draco cloud were obtained at different times and were treated as independent data throughout the analysis ; yet , they suggest that the draco cloud has an ere spectrum that does not vary within the cloud .
the observed intensities in the i - band essentially reverts to the predicted level for scattering alone for four clouds .
only mbm 25 exhibits a significant excess in the i - band as well .
this is not inconsistent with ere seen in other sources ; on the contrary , i - excesses in ere - bright regions are very common in reflection nebulae ( witt & schild 1985 ) . the bgri filter system employed in the present study provides insufficient resolution for a detailed investigation of differences in the ere spectra of hglics .
the scattered - light subtracted ere intensities in our five hglics are remarkably constant , typically @xmath60 .
the width of the ere band is difficult to estimate with only four filters , but the data in figures 6a - e suggest fwhm @xmath61 .
in contrast to the near - constancy of the ere intensity , the scattered light intensities within our sample of five hglics vary by almost a factor of two .
this range of scattered - light intensities matches almost exactly the range of the quantity ( cloud - sky ) in table 4 , the excess 100 intensity of the five hglics over that of the respective surrounding sky , which is a direct measure of the cloud optical depths .
however , the correlation between the optical depths of the individual clouds and the scattered light intensity is not perfect , e.g. the cloud with the brightest scattered light intensity , mbm 25 , has the second lowest optical depth .
thus , we conclude that the observed range of scattered light intensities must also be influenced by differences in the directional distribution of the illuminating radiation field . this should be expected , given the wide range of galactic longitudes of the hglcis in our sample ( see table 2 ) .
we note , therefore , that the results of our decomposition of the observed intensities into a relatively variable scattered scattered light component and a comparatively constant ere component is consistent with our original expectations .
the near - constancy of the ere is primarily caused by an isotropic emission process , which is responsive to the average intensity of the incident radiation field while remaining insensitive to its directional distribution .
the scattered light intensity , on the other hand , is strongly coupled to the directional anisotropy of the incident radiation through the forward - directed scattering phase function of the interstellar grains .
a second factor is likely to contribute to the different degrees of variability of the intensities of ere and scattered light in the case of clouds with a range of optical depths .
it is known that ere requires far - uv radiation for its initiation ( witt & schild 1985 ; darbon et al .
1999 ; witt et al .
the optical depth of a given cloud is between two and three times larger in the wavelength region of ere excitation compared to the optical depth at the wavelength of our b - filter ( 457 nm ) . while our clouds are optically thin at b ( sect .
3.1.1 ) , they will approach the optically thick condition in the far - uv , where ere is excited .
thus , nearly all uv photons capable of initiating ere are absorbed and the resulting ere intensity saturates to a constant level , arising mostly in the surface layers , independent of the optical depth .
by contrast , scattering at optical wavelengths , where the clouds are optically thin , occurs throughout the cloud and the resulting scattered light intensities vary with the differences in optical depth .
finally , we must recognize that the ere and scattered light intensities obtained from our differential photometry data are always lower limits , because the column density of interstellar dust associated with the `` sky '' positions generates both forms of radiation , with intensity levels comparable to the values shown in figures 6a - e. these sky - related ere and scattered light intensities are subtracted in the data reduction process .
therefore , to correct for this effect , we estimate that the absolute ere intensity in hglics is probably of the order of @xmath62 .
this estimate takes into account that the ere intensity is expected to increase linearly at very small optical depth before beginning to saturate at the optical depths found in our `` cloud '' positions .
the ratio of ere to scattered light intensity is expected to decrease with optical depth .
our hglic of smallest optical depth , mbm 41d , has a value of @xmath9 1/2 for this ratio , which implies that the ere intensity is approximately 1/3 of the total interstellar radiation in the r - band coming from the low - density diffuse cirrus medium at high latitudes .
this agrees remarkably well with the findings of gordon et al .
( 1998 , fig .
23 ) and szomoru & guhathakurta ( 1998 ) . also , our estimated total intensity of the ere , when integrated over the fwhm of the r filter of pioneer 10/11 ( 968 a ) results in a band - integrated ere intensity of @xmath63 , in excellent agreement with the results obtained with pioneer 10/11 by gordon et al .
from the outset , the clouds in our hglic sample have been selected to be optically thin . every cloud in our sample , albeit limited , exhibits ere with very similar intensities , apparently unrelated to the optical depths of the clouds , which vary over a range of about a factor two . as a result , the highest ratio of ere to
scattered light intensity occurs in the cloud with the smallest optical depth , mbm 41d .
the data fully support the suggestion that the ere results from an isotropic emission process that is initiated by the far - uv radiation component of the galactic interstellar radiation field ( witt & schild 1985 ; witt et al .
2006 ) . in the case at hand of isolated clouds immersed into the interstellar radiation field
, the maximum ere intensity will be reached quickly when the physical penetration depth for the exciting uv photons begins to equal or becomes less than the average cloud diameter , while the cloud remains still quite optically thin for the resulting ere photons .
this effect is expected to be particularly pronounced at high galactic latitudes where the wavelength dependence of interstellar extinction is characterized by r@xmath15 values smaller than the typical value of r@xmath15 = ( 3.05 @xmath11 0.15 ) found for the diffuse ism in the galactic plane ( larson & whittet 2005 ) .
a lower value of r@xmath15 generally indicates an extinction curve of increased steepness in the far - uv ( cardelli et al . 1989 ) , which implies a larger than typical ratio of optical depths for specific uv wavelengths to those at optical wavelengths . as a consequence , clouds with a range of optical thicknesses as represented in our sample , while more or less optically thin in the visual range , are close to being optically thick in the far - uv , with a resultant saturation in the ere excitation .
the near - constancy of the observed ere intensities has further important implications .
given the proximity to earth of the clouds in our hglic sample , we can assume that the average intensity and spectral energy distribution of the illuminating interstellar radiation field is essentially the same for all clouds and similar to the interstellar radiation field in the solar vicinity .
then , our results imply that the abundance of the ere carrier compared to the abundance of dust responsible for the uv / optical interstellar extinction is relatively constant
. the ere carrier must , therefore , be a common constituent of the interstellar medium in diffuse interstellar clouds , which makes a significant contribution to interstellar extinction through absorption , as we will show below .
gordon et al .
( 1998 ) and szomoru & guhathakurta ( 1998 ) compared the number of ere photons emitted to the number of photons absorbed from the interstellar radiation field by the diffuse ism over the 91.2 nm to 550 nm wavelength range , finding a ratio of 0.10 @xmath11 0.03 . assuming that at most one ere photon can be generated by photoluminescence for every uv / optical photon absorbed , they interpreted this ratio as a lower limit to the photon conversion efficiency of the ere process . inversely ,
if we assume that the photon conversion efficiency is at most 100% , this ratio can also be interpreted as the lower limit to the absorption cross section provided by the ere carrier as a fraction of the total continuous uv / optical interstellar absorption cross section exhibited by all interstellar dust components . a more realistic estimate for
the ere photon conversion efficiency may be 30% to 50% instead of the theoretical upper limit of 100% . in that case
, the ere carrier contributes from 20% to 30% of the total continuous uv / optical absorption cross section of the diffuse galactic ism .
depending on precisely where in the uv the absorption by the ere carrier peaks , only a small fraction of the energy of a photon absorbed by the ere carrier is re - emitted in the form of optical ere photons , which have typical energies near 1.8 ev .
the remainder of the energy , which could be as much as @xmath925% of the total absorbed interstellar radiation field energy , will have to be re - emitted at longer wavelengths , most likely in the near- and mid - ir .
this amount is remarkably close to the observed fraction of the total stellar radiation of the galaxy reprocessed by the ism and emitted in the near / mid - ir ( boulanger & perault 1988 ) .
we suggest , therefore , that the ere carrier is a major contributor to the near / mid - ir emission from the galactic cirrus , or inversely , that one characteristic of a large fraction of the particles giving rise to the near / mid - ir galactic emission is the ability to re - emit a small fraction of the energy absorbed in the uv / optical range in the form of optical photoluminescence photons appearing as ere .
it appears likely , therefore , that the ere is a significant component of the processing of galactic stellar radiation by interstellar dust . in the controversy concerning the existence of ere in the diffuse ism that has arisen between guhathakurta & tyson ( 1989 ) , gordon et al .
( 1998 ) , and szomoru & guhathakurta ( 1998 ) on the one hand and zagury ( 2006 ) on the other , our findings clearly support the former group of investigations .
we have conducted differential surface brightness photometry in four broad optical bands ( b , g , r , i ) of five optically thin hglics , employing the method of wide - field ccd imaging .
we compared the resulting intensities with spectral energy distributions expected for scattered light only , assuming that the source of illumination is the galactic interstellar radiation field of the solar vicinity .
we reached the following conclusions : 1 .
we detected excess intensities in the r - band in all cases , in the g - band in all but one case , and in the i - band in one case .
these excess intensities and their spectral energy distributions are consistent with the presence of ere in all hglics examined in this study .
the ere intensities are remarkably constant for clouds with a range of optical depths of about a factor two .
the constancy of the ere intensities is consistent with excitation of the ere through far - uv photons , which saturates well before clouds become optically thick in the red region where ere photons are emitted , and with a near - constant abundance ratio of ere carrier particles to dust grain causing the interstellar reddening .
3 . the absolute ere intensities in hglics , determined by correcting the results of our differential photometry for the ere contribution expected to be present in our reference sky positions are in excellent agreement with measurements of the ere in the high - latitude diffuse galactic light using pioneer 10/11 photometry by gordon et al . ( 1998 ) and the results of differential spectro - photometry of individual hglics by szomoru & guhathakurta ( 1998 ) .
we find that the ratio of the number of ere photons and the number of absorbed uv / optical photons in the interstellar radiation field , which leads to an estimate for a lower limit for the ere photon conversion efficiency of @xmath64% ( gordon et al .
1998 ; szomoru & guhathakurta 1998 ) , can also be interpreted to suggest that the ere carrier particles contribute about 20% to 30% of the continuous uv / optical absorption cross section in the diffuse ism and very likely are responsible for a major fraction of the near / mid - ir radiation emitted by the galactic cirrus .
this research has been supported through grants ast0307307 and ast0606756 from the national science foundation to the university of toledo .
our observing program at new mexico skies received additional support from corporate sponsors astrodon , rc optical systems , santa barbara instrument group , and software bisque , for which we are very grateful . , with north up and east to the left .
reference sky positions are indicated by labels 1 , 2 , 3 while the cloud positions at which surface brightness measurements were conducted are labeled 1s trough 10s in each case.,width=624 ] lccccc mbm 25 & 1/6/2006 & 1/6/2006 & 1/7/2006 & 1/9/2006 & 1/7/2006 + @xmath66 $ ] & 7200 & 5400 & 7200 & 7200 & 7200 + + mbm 30 & 2/28/2006 & 3/5/2006 & 3/5/2006 & 2/7/2006 & 3/25/2006 + @xmath66 $ ] & 7200 & 7200 & 7200 & 10800 & 15300 + + mbm 32 & 1/28/2006 & 1/28/2006 & 1/28/2006 & 1/29/2006 & 1/31/2006 + @xmath66 $ ] & 7200 & 5400 & 5400 & 7200 & 8100 + + mbm 41-a & 3/29/2006 & 3/29/2006 & 3/30/2006 & 4/9/2006 & 4/21/2006 + @xmath66 $ ] & 7200 & 6300 & 7200 & 14400 & 10800 + + mbm 41-d & 4/28/2006 & 4/28/2006 & 4/28/2006 & 4/30/2006 & 4/30/2006 + @xmath66 $ ] & 9000 & 7200 & 8100 & 7200 & 10800
+ ccccc mbm 25 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 + mbm 30 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 + mbm 32 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 + mbm 41a & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 + mbm 41d & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 & [email protected] @xmath67 + ccccccc mbm 25 & 3.67 & 1.94 & 1.73 & 1.89 & 0.44 & 0.23 + mbm 30 & 6.29 & 2.25 & 4.04 & 2.80 & 0.75 & 0.27 + mbm 32 & 5.99 & 2.47 & 3.52 & 2.43 & 0.71 & 0.29 + mbm 41a & 4.10 & 0.92 & 3.18 & 4.46 & 0.49 & 0.11 + mbm 41d & 2.75 & 1.30 & 1.45 & 2.12 & 0.33 & 0.15 + cccc 337 & 1.32 ( -7 ) & 1.29 & 0.670 + 390 & 1.44 ( -7 ) & 1.14 & 0.670 + 420 & 1.50 ( -7 ) & 1.08 & 0.670 + 454 & 1.42 ( -7 ) & 1.00 & 0.670 + 493 & 1.45 ( -7 ) & 0.90 & 0.657 + 527 & 1.50 ( -7 ) & 0.82 & 0.643 + 579 & 1.53 ( -7 ) & 0.73 & 0.623 + 607 & 1.51 ( -7 ) & 0.69 & 0.590 + 666 & 1.50 ( -7 ) & 0.62 & 0.563 + 705 & 1.48 ( -7 ) & 0.57 & 0.549 + 755 & 1.40 ( -7 ) & 0.53 & 0.516 + 802 & 1.34 ( -7 ) & 0.47 & 0.496 + 848 & 1.28 ( -7 ) & 0.42 & 0.469 + 918 & 1.22 ( -7 ) & 0.36 & 0.449 + 973 & 1.15 ( -7 ) & 0.33 & 0.422 + | nearby interstellar clouds at high galactic latitudes are ideal objects in which the interaction of interstellar dust with photons from the well - characterized interstellar radiation field can be studied .
scattering and uv - excited photoluminescence at optical wavelengths as well as thermal emission at mid- and far - infrared wavelengths are observable manifestations of such interactions . here
we report initial results from an optical imaging survey of optically thin high - galactic - latitude clouds , which is designed to study the surface brightness , structure , and spectral energy distribution of these objects .
the primary aim of this paper is to study the extended red emission ( ere ) that has been reported at high galactic latitudes in earlier investigations and which is attributed to ultraviolet - excited photoluminescence of an as yet unidentified component of interstellar dust .
we conduct this ongoing survey with remotely operated , fast , short - focal - length ( 0.5 m ) telescopes equipped with absolutely - calibrated ccd cameras yielding a field of view of @xmath0 .
the telescopes are located at new mexico skies at 7300 ft altitude near mayhill , nm . the optical surface brightness of our objects is typically a few percent of the brightness of the dark night sky , implying that the cloud seds must be deduced from differential surface brightness photometry in different filter bands .
we find strong evidence for dust emission in the form of a broad ( @xmath1 fwhm ) ere band with peak emission near 600 nm wavelength and peak intensity of @xmath2 in optically - thin clouds .
this amounts to about 30 % of the total optical surface brightness of these clouds , the remainder being consistent with expectations for dust - scattered light .
this supports claims for the ubiquitous presence of the ere carrier throughout the diffuse interstellar medium of the milky way galaxy .
we suggest that the ere carrier is involved in the radiative processing of about 20% to 30% of the dust - absorbed uv / optical luminosity of the milky way galaxy , with the bulk of this energy being emitted in the near- to mid - infrared spectral regions . |
measurement and theoretical predictions of kinetic isotope effects belong among the main tools of chemical kinetics . kinetic isotope effect ( kie )
is defined as the ratio @xmath2 of rate constants for two isotopomers @xmath3 and @xmath4 .
isotopomers @xmath3 and @xmath4 are two chemical species differing only by replacing a group of atoms in chemical species @xmath3 by their isotopes in species @xmath4 .
recently , observation of anomalously large kies has helped prove importance of quantum effects in enzymatic reactions at physiological ( i.e. surprisingly high ) temperatures @xcite .
this and similar results have changed our understanding of enzymatic catalysis and spurred an active experimental and theoretical research in the last several years . since the early days of chemical kinetics ,
kies have been predominantly described from the perspective of the transition - state theory ( tst ) @xcite .
this theory is intrinsically classical , although various quantum corrections have been incorporated in it over time .
these include corrections due to the zero - point - energy effects , high - temperature wigner tunneling correction @xcite , and various semiclassical approximations for treating the tunneling at low temperatures @xcite . on the other end of the spectrum
are exact quantum - mechanical methods for computing rate constants and kies @xcite , but in general these are not feasible for systems with many degrees of freedom .
one therefore resorts to various approximations that make a computation practicable but are less severe than the tst . among these
belongs a variety of quantum transition - state theories @xcite , the most recent of which is the quantum instanton ( qi ) approximation @xcite , motivated by an earlier semiclassical instanton model @xcite . in this contribution , we describe a method @xcite , based on the qi approximation , for computing kies directly , rather than via computing the rate constants for the two isotopomers first . because of the ultimate goal of applying a similar method to enzymatic reactions
, the method is implemented using a general path - integral approach that scales favorably with the number of degrees of freedom .
several alternative estimators for relevant quantities have been developed @xcite and their relative merits are compared in this contribution on the benchmark hydrogen atom
diatom exchange reaction .
the quantum - instanton approximation for the rate constant was introduced in ref .
@xcite . a simpler alternative derivation @xcite described in detail in ref .
@xcite , starts with the miller
tromp formula @xcite for the thermal rate constant @xmath5 , @xmath6 here @xmath7 is the reactant partition function ( per unit volume for bimolecular reactions ) and @xmath8 is the symmetrized flux flux correlation function , @xmath9 with hamiltonian operator @xmath10 and flux operator @xmath11 denoting the flux through the dividing surface @xmath12 .
quantum instanton expression follows by multiplying and dividing the integrand of eq.([miller_schwartz_tromp ] ) by the delta delta " correlation function @xmath13 defined below in eq . ( [ cdd ] ) , assuming that @xmath14 varies slowly compared with @xmath13 , and applying the steepest descent approximation to the resulting integral .
assuming further that the stationary - phase point is at @xmath15 , we obtain the qi thermal rate constant , @xmath16 here @xmath17 is a specific type of energy variance @xcite , @xmath18^{1/2 } , \label{dh}\ ] ] and the delta
delta correlation function @xmath13 is defined @xcite as @xmath19 the generalized delta operator @xmath20 will be defined below in eq .
( [ gen_delta ] ) . in applying the qi approximation to the kies , it is useful to consider a continuous change of the isotope mass . if the two isotopomers are @xmath3 and @xmath4 , a real parameter
@xmath21 $ ] can be defined such that @xmath22 where @xmath23 and @xmath24 are the masses of the @xmath25th atom in the isotopomers @xmath3 and @xmath4 , respectively . within the qi approximation ( [ k_qi ] )
, the kie can be expressed as @xmath26 where the argument denotes the value of @xmath27 and for simplicity the time argument of the correlation functions has been omitted since it is always @xmath15 in the qi approximation . also , for convenience , both numerator and denominator have been divided by @xmath28 .
four types of quantities must be evaluated in order to compute the kies from eq .
( [ qi_kie_1 ] ) : the ratio of the partition functions @xmath29 , the ratio of the delta correlation functions @xmath30 , and the energy variance @xmath31 and the velocity " factor @xmath32 for @xmath33 and 1 .
the last two quantities are in the form of thermodynamic averages ( for a given @xmath27 ) and therefore can be directly computed by metropolis monte - carlo techniques ; the relevant estimators have been derived in refs . @xcite .
the most general forms are listed in ref .
the first two quantities can not be evaluated directly since they are ratios of quantities for two different values of @xmath27 .
an elegant solution exists , however .
here is where considering a continuous isotope change ( using a parameter @xmath27 ) becomes useful : instead of computing the ratios directly , we use the _ thermodynamic integration _ idea @xcite , applied to the parameter @xmath27 ( i.e. , to the masses of the isotopes instead to the usual inverse temperature @xmath34 ) .
we can express the two ratios as an exponential of the integrals of logarithmic derivatives , @xmath35,\label{charging_qr}\\ \frac{c_{\mathrm{dd}}(1)}{c_{\mathrm{dd}}(0)}&= & \exp\left[\int_0 ^ 1\d\lambda\,\frac{\d\log c_{\mathrm{dd}}(\lambda)}{\d\lambda}\right ] .
\label{charging_cdd}\end{aligned}\ ] ] since the logarithmic derivatives can be expanded as @xmath36 they are normalized quantities ( thermodynamic averages ) which can be directly computed by the metropolis algorithm .
we can compute ratios of both the reactant partition functions and the delta
delta correlation functions at @xmath33 and 1 , by computing the values of the corresponding logarithmic derivatives for enough values @xmath27 between 0 and 1 , and then by integrating over @xmath27 and exponentiating , according to eqs .
( [ charging_qr ] ) and ( [ charging_cdd ] ) .
in fact , in a cruder version of the qi method , called the simplest quantum instanton ( sqi ) approximation @xcite , the ratios of the partition and delta delta correlation functions are all we need , since within that approximation , the kinetic isotope effect is just @xmath37 the relevant estimators for the logarithmic derivatives have been derived in refs . @xcite . in ref .
@xcite , _ thermodynamic _ estimators have been derived that differentiate the kinetic part of the action ; in ref .
@xcite , _ virial _ estimators have been derived that differentiate the potential part of the action . in both cases ,
the derivations have been done for general systems with @xmath38 atoms in @xmath39 dimensions , even for cases with unbound degrees of freedom ( such as the center - of - mass motion ) . in the next section
, we present a simplified derivation of these estimators for a single particle in a one - dimensional external potential .
this choice significantly simplifies notation , but preserves the main ingredients of the many - dimensional derivation .
below we derive three types of estimators for the logarithmic derivatives of both @xmath7 and @xmath40 .
we refer to the three types of estimators as thermodynamic , virial , and generalized virial because of their resemblance to corresponding thermodynamic @xcite , virial @xcite , and generalized virial @xcite estimators for the kinetic energy .
let us consider a single particle of mass @xmath41 in a one - dimensional potential @xmath42 .
since we have only one mass @xmath41 , we do not need to define an additional parameter @xmath27 : we can just take @xmath41 itself to be the parameter for the thermodynamic integration .
the pi representation of the partition function is @xmath43,\\ \phi&=&\frac{pm}{2\pi\hbar^2\beta^2}\sum_{s=1}^p\left(r^{(s)}-r^{(s-1)}\right)^2 + \frac{1}{p}\sum_{s=1}^{p}v\left(r^{(s)}\right).\nonumber\end{aligned}\ ] ] here @xmath44 , denotes the beads of the discretized paths ( @xmath45 is identical to @xmath46 ) . in general
, we will obtain the estimators for the logarithmic derivative directly , by computing the logarithmic derivative @xmath47 of the particular form of the discretized pi . applying this approach to the pi ( [ qr_pi_cartesian ] )
, we obtain the _ thermodynamic _ estimator @xmath48 above , @xmath49 denotes the average over paths weighted with the weight @xmath50 , @xmath51 alternatively , we can define new , mass - scaled coordinates as @xmath52 in these new coordinates , the partition function becomes @xmath53 simplest _ virial _
estimator for the logarithmic derivative can again be derived by direct differentiation of pi ( [ qr_pi_rescaled ] ) , @xmath54 } { \partial m}\right>_{\rho_r}=\label{qr_est_virial}\\ & = & -\frac{\beta}p\left<\sum_{s=1}^p\frac{\partial v\left[(m+\delta m)^{-1/2}m^{1/2}r^{(s)}\right ] } { \partial\delta m}\biggl|_{\delta m=0}\right > _ { \rho_{r}}=\nonumber\\ & = & \frac{\beta}{2p}\left<\sum_{s=1}^{p}r^{(s)}\frac{\partial v\left(r^{(s)}\right)}{\partial r^{(s)}}\right>_{\rho_{r}}.\nonumber\end{aligned}\ ] ] above are three estimators for the logarithmic derivative : the first one suitable if the mc simulation is done in mass - scaled coordinates @xmath55 , the other two for original cartesian coordinates @xmath56 .
the first two suggest evaluation of the derivative numerically , by finite differences , which will be in fact , more efficient in many dimensional systems than the analytical third expression that requires the knowledge of the gradient of the potential . only for systems with few degrees of freedom and available
gradient of the potential , the third expression may be preferable .
the trick of using numerical derivatives with respect to a single parameter was originally used by predescu for computing heat capacities @xcite and higher temporal derivatives of the flux
flux correlation function @xcite where the parameters were the inverse temperature and the imaginary time , respectively .
the simplest virial estimators ( [ qr_est_virial ] ) have one shortcoming compared to the thermodynamic estimators , namely , they only work in bound systems .
this can be immediately seen by considering a free particle with @xmath57 .
this shortcoming can be remedied if the rescaling is done only after subtracting an arbitrarily chosen ( but fixed ) slice from the remaining @xmath58 slices . to be more explicit ,
let us define relative coordinates as @xmath59 since the jacobian of the transformation is unity , we have @xmath60+\nonumber\\ & & + \frac1p\biggl[\,\sum_{s=1}^{p-1}v(r^{(p)}+y^{(s)})+v(r^{(p)})\biggr].\nonumber\end{aligned}\ ] ] now we define mass - scaled coordinates as @xmath61 in these coordinates , the partition function becomes @xmath62+\nonumber\\ & & + \frac1p\biggl[\,\sum_{s=1}^{p-1}v(r^{(p)}+m^{-1/2}x^{(s)})+v(r^{(p)})\biggr].\nonumber\end{aligned}\ ] ] the _ generalized virial _
estimator for the logarithmic derivative follows by differentiating the pi expression ( [ qr_pi_relative_rescaled ] ) , @xmath63 } { \partial m}\biggr>_{\rho_{r}}=\\[12pt ] \disty\qquad = \frac{1}{2m}-\frac{\beta}{p}\biggl<\sum_{s=1}^p \frac{\partial v\left[r^{(p)}+(m+\delta m)^{-1/2}m^{1/2 } \left(r^{(s)}-r^{(p)}\right)\right]}{\partial\delta m}\biggl|_{\delta m=0}\biggr>_{\rho_r}=\\[12pt ] \disty\qquad = \frac1{2m}+\frac{\beta}{2p}\biggl < \sum_{s=1}^{p}\frac{\partial v\left(r^{(s)}\right ) } { \partial r^{(s)}}(r^{(s)}-r^{(p)})\biggr>_{\rho_r}\ , .
\end{array}\ ] ] since we have chosen the slice @xmath46 arbitrarily , we can do the same for any slice @xmath64 , derive a corresponding estimator , and then take an average of these estimators .
the result is @xmath65 and in general , we can replace @xmath66 in all three forms ( [ qr_est_virial_generalized ] ) of the estimator by @xmath67 .
since the number of slices @xmath68 appears explicitly only in the denominator of the generalized virial estimator ( [ qr_est_virial_generalized ] ) or ( [ qr_est_virial_centroid ] ) , the statistical error should be independent of @xmath68 for a fixed number of monte - carlo samples .
on the other hand , @xmath68 appears explicitly in the numerator of the thermodynamic estimator ( [ qr_est_thermodynamic ] ) , so the error is expected to grow with @xmath68 .
this will be confirmed in the numerical example in section 4 .
the derivation for @xmath40 is similar .
however , due to the constraints to the two dividing surfaces , a new term appears in the estimator .
the pi representation of @xmath40 is @xmath69 \delta\left[\xi_b\left(r^{(p/2)}\right)\right]\e^{-\beta\phi}.\nonumber\end{aligned}\ ] ] the generalized delta function @xmath70 is defined @xcite as @xmath71\equiv\left\vert\frac{m}{\partial_r\xi}\right\vert^{1/2 } \delta\bigl[\xi(r)\bigr]\ , .
\label{gen_delta}\ ] ] for numerical purposes , it is convenient to replace the strict delta function by a gaussian approximation @xcite , @xmath72&\approx & \tilde{\delta}\left[\xi\left ( \bar{r}^{(s)}\right)\right],\label{gauss_delta}\\ \bar{r}^{(s)}&\equiv&\frac{1}{2}\,\left(r^{(s)}+r^{(s+1)}\right),\nonumber\\ \tilde{\delta}\bigl[\xi(r)\bigr]&\equiv & \left(\frac{2p}{\pi\hbar^2\beta}\right)^{1/2 } \exp\left\{-\frac{2pm}{\hbar^2\beta}\left [ frac{\xi(r)}{\partial_r\xi(r)}\right]^2\right\}.\nonumber\end{aligned}\ ] ] we can define an effective action @xmath73 , which includes the constraint potential @xmath74 ^ 2 . \label{v_constr}\ ] ] the logarithmic derivative will have one extra term due to this constraint potential , @xmath75 where @xmath76 denotes that the corresponding estimator for @xmath7 given in eq .
( [ qr_est_thermodynamic ] ) , ( [ qr_est_virial ] ) , or ( [ qr_est_virial_generalized ] ) should be used except that the sampling is done according to weight @xmath77 instead of @xmath78 .
using the pi representation ( [ cdd_pi_cartesian ] ) of @xmath40 in cartesian coordinates , we find the additional term from eq .
( [ logderivative_cdd ] ) to the thermodynamic estimator ( [ qr_est_thermodynamic ] ) to be @xmath79 ^ 2 .
\label{cdd_est_thermodynamic}\ ] ] rescaling coordinates according to eq .
( [ rescaling ] ) gives a constraint potential @xmath80 ^ 2\,.\ ] ] the additional term from eq .
( [ logderivative_cdd ] ) to the simple virial estimator ( [ qr_est_virial ] ) becomes @xmath81 ^ 2\,,\label{cdd_est_virial}\\ r_{\text{resc}}&\equiv&\left(\frac{m}{m+\delta m}\right)^{1/2}r\,.\nonumber\end{aligned}\ ] ] finally , if we rescale coordinates according to eq .
( [ rescaling_subtract ] ) , or better , as @xmath82 we obtain the same estimator as ( [ cdd_est_virial ] ) , only the rescaled coordinate is defined as @xmath83 the generalization to more - dimensional systems is fairly straightforward . only in the case of the simple virial estimator ( [ qr_est_virial ] ) or ( [ cdd_est_virial ] ) , care must be taken to account for the unbound ( free ) degrees of freedom by appropriately rescaling the corresponding volume .
for instance , for bimolecular reactions , the potential in the reactant region is independent of the center - of - mass - coordinate and the relative coordinate of the two molecules . for details ,
see ref . @xcite .
@xcite , the qi procedure for evaluating kies was successfully tested on several problems of increasing complexity : the one - dimensional eckart barrier and the isotopic variants of both the collinear and the three - dimensional hydrogen - exchange reaction h+h@xmath84 h@xmath85+h .
the results for the kie = @xmath5(h+h@xmath85)/@xmath5(d+d@xmath85 ) as a function of the inverse temperature @xmath86 for both collinear and three - dimensional versions of the reaction are also shown here in fig . 1 .
the figure compares the exact quantum - mechanical result @xcite with the results of the qi , sqi , and tst approximations .
the three - dimensional version also shows the result of the canonical variational tst with semiclassical tunneling correction ( cvt ) @xcite . in general , the results of the qi approximation are very good : the error is smaller than @xmath87 for temperatures 250 to 600 k. for lower temperatures , the larger error is due to using a single dividing surface : results can be improved by considering two separate dividing surfaces . at high temperatures ,
the error is due to classical recrossing and can not be corrected within the qi approximation .
for further details of the calculation see ref .
@xcite .
-5 mm three types of estimators for the logarithmic derivative of @xmath7 , given in eqs .
( [ qr_est_thermodynamic ] ) , ( [ qr_est_virial ] ) , and ( [ qr_est_virial_centroid ] ) , are compared in fig .
this calculation is for the collinear version of the kie = @xmath5(h+h@xmath85)/@xmath5(d+d@xmath85 ) at 300 k. the calculation was done with 10 walkers and a fixed number @xmath88 monte - carlo moves for all @xmath68 .
the left part of the figure shows the convergence of the partition - function ratio @xmath89 as a function of the number of slices @xmath68 .
the ratio was computed via the thermodynamic integration ( [ charging_qr ] ) in which the three different estimators ( [ qr_est_thermodynamic ] ) , ( [ qr_est_virial ] ) , and ( [ qr_est_virial_centroid ] ) for @xmath90 were used .
the right panel shows the @xmath68-dependence of the relative error of @xmath89 . as expected , for large @xmath68 , the error of the generalized virial estimator ( [ qr_est_virial_centroid ] ) is almost independent of @xmath68 . on the other hand the error of the thermodynamic estimator ( [ qr_est_thermodynamic ] ) grows with @xmath68 . even for small @xmath68 ,
the generalized virial estimator is superior .
finally , we can see the importance of subtracting the centroid motion before rescaling in eq .
( [ rescaling_subtract ] ) by comparing errors of the simple ( [ qr_est_virial ] ) and generalized ( [ qr_est_virial_centroid ] ) virial estimators .
the difference is due to the fact that we have two free degrees of freedom in the reactant region of the collinear bimolecular reaction .
similar conclusions ( not shown here ) can be obtained for the ratio @xmath30 , except that the error of the generalized virial estimator has a weak dependence on @xmath68 arising from the additional term due to the constraint to the dividing surfaces .
judging from the numerical results in the previous section , the qi approach for computing kinetic isotope effects is very promising . the procedure is quite general : multi - dimensional estimators for all relevant quantities are presented in refs .
although the path - integral approach has been chosen because of its favorable scaling with the number of degrees of freedom , the computations for many - dimensional systems are still difficult .
one obstacle is the difficulty of efficient sampling of a many - dimensional configuration space . with proper estimators
, we may decrease statistical and systematic discretization errors , but it is difficult to avoid systematic errors due to long correlations .
for this reason , it may be efficient to use a different number of imaginary time slices @xcite for different degrees of freedom , which is a generalization of more crude mixed quantum - classical methods .
another obstacle to obtaining a good match between theory and experiment is the potential energy surface for the reaction .
while more accurate _ ab initio _ potentials are computationally very expensive , the much faster molecular - mechanics force fields are often too crude . in ref .
@xcite , in which the qi method is used to compute the kie for the isomerization of _
cis_-pentadiene , two approaches are taken : in one , an empirical valence bond ( evb ) potential is formed from the equilibrium potentials for reactants and products ; in the other , a more accurate but also a more - expensive semi - empirical potential is used . because of the computational expense already for this system with 39 degrees of freedom , it appear that due to their better accuracy such semi - empirical potentials will be the potentials of choice for intermediate - size systems , and the evb potentials based on molecular - mechanical force fields the potentials of choice for truly many - dimensional systems .
j. vanek would like to thank y. li and c. predescu for useful discussions .
this work was supported by the office of naval research grant no .
n00014 - 05 - 1 - 0457 and by the director , office of science , office of basic energy sciences , chemical sciences , geosciences , and biosciences division , u.s .
department of energy under contract no .
de - ac02 - 05ch11231 . 99 a. kohen , r. cannio , s. bartolucci and j.p .
klinman : nature * 399 * ( 1999 ) 496 . l.c.s .
melander : _ isotope effects on reaction rates_. ronald press , new york , 1960 .
collins and n.s .
bowman : eds . : _
isotope effects in chemical reactions_. van nostrand reinhold , new york , 1971 .
miller : j. chem .
phys . * 62 * ( 1975 ) 1899 .
marcus and m.e .
coltrin : j. chem . phys .
* 67 * ( 1977 ) 2609 .
truhlar and b.c .
garrett : annu .
phys . chem . *
35 * ( 1984 ) 159 .
miller : faraday discuss .
* 110 * ( 1998 ) 1 .
voth , d. chandler and w.h .
miller : j. chem .
* 91 * ( 1989 ) 7749 .
hansen and h.c .
andersen : j. phys .
* 100 * ( 1996 ) 1137 .
truong , d. lu , g.c .
_ : comput .
* 75 * ( 1993 ) 143
miller , y. zhao , m. ceotto and s. yang : j. chem .
* 119 * ( 2003 ) 1329 .
j. vanek , w.h .
miller , j.f .
castillo and f.j .
aoiz : j. chem . phys .
* 123 * ( 2005 ) 054108 .
j. vanek and w.h .
miller : in preparation .
m. ceotto and w.h .
miller : private communication . w.h .
miller , s.d .
schwartz and j.w .
tromp : j. chem .
* 79 * ( 1983 ) 4889 .
t. yamamoto and w.h .
miller : j. chem .
* 120 * ( 2004 ) 3086 .
y. zhao , t. yamamoto and w.h .
miller : j. chem .
* 120 * ( 2004 ) 3100 .
d. chandler : _ introduction to modern statistical mechanics_. oxford university press , new york , 1987 .
j. barker : j. chem .
phys . * 70 * ( 1979 ) 2914 .
herman , e.j .
bruskin and b.j .
berne : j. chem .
* 76 * ( 1982 ) 5150 .
m. parrinello and a. rahman : j. chem .
* 80 * ( 1984 ) 861 . c. predescu , d. sabo , j.d . doll and d.l .
freeman : j. chem . phys .
* 119 * ( 2003 ) 12119 . c. predescu : phys .
e * 70 * ( 2004 ) 066705 .
truhlar , a. kuppermann and j.t .
adams : j. chem .
* 59 * ( 1973 ) 395 .
fast , j.c . corchado and d.g .
truhlar : j. chem . phys .
* 109 * ( 1998 ) 6237 . y. li and w.h .
miller : mol .
* 103 * ( 2005 ) 203 . | a general method for computing kinetic isotope effects is described .
the method uses the quantum - instanton approximation and is based on the thermodynamic integration with respect to the mass of the isotopes and on the path - integral monte - carlo evaluation of relevant thermodynamic quantities .
the central ingredients of the method are the monte - carlo estimators for the logarithmic derivatives of the partition function and the delta delta correlation function .
several alternative estimators for these quantities are described here and their merits are compared on the benchmark hydrogen - exchange reaction , h+h@xmath0h@xmath1+h on the truhlar kuppermann potential energy surface .
finally , a qualitative discussion of issues arising in many - dimensional systems is provided . |
the use of chiral lagrangians in combination with unitary techniques in coupled channels of mesons and baryons has been a very fruitful scheme to study the nature of many hadron resonances .
the analysis of meson baryon scattering amplitudes shows poles in the second riemann sheet which are identified with existing baryon resonances . in this way
the interaction of the octet of pseudoscalar mesons with the octet of stable baryons has lead to @xmath3 resonances which fit quite well the spectrum of the known low lying resonances with these quantum numbers @xcite .
similarly , the interaction of the octet of pseudoscalar mesons with the decuplet of baryons also leads to many resonances that can be identified with existing ones of @xmath4 @xcite .
sometimes a new resonance is predicted , as in the case of the @xmath5 , where all the chiral approaches find two poles close by , rather than one , a fact that finds experimental support in the analyses of refs .
the nature of the resonances is admittedly more complex than just a molecule of pseudoscalar and baryon , but the success of this picture in reproducing many experimental data on decay and production of the resonances provides support to claim very large components of this character for the resonance wave function . in some cases
one can even reach the limits of the model and find observables that call for extra components , even if small , but essential to explain some data .
this can be seen for instance in the radiative decay of the @xmath6 @xcite , or in the helicity form factors of the @xmath7 @xcite .
another promising approach is the one followed in @xcite , based on the naturalness of the subtraction constants in the dispersion relations that hint in some cases at the existence of non meson baryon components , particularly in the case of the @xmath7 .
another step forward in this direction has been the interpretation of low lying @xmath8 as molecular states of two pseudoscalar mesons and one baryon @xcite .
much work has been done using pseudoscalar mesons as building blocks , but the consideration of vectors instead of pseudoscalars is only beginning to be exploited . in the baryon sector
the interaction of the @xmath9 @xmath10 has been recently addressed in @xcite , where three degenerate @xmath11 states around 1800 mev and three degenerate @xmath10 states around @xmath12 mev , with @xmath13 , are found .
this work has been recently extended to the su(3 ) space of vectors and baryons of the decuplet in @xcite .
the underlying theory for this study is the hidden gauge formalism @xcite , which deals with the interaction of vector mesons and pseudoscalars in a way respecting chiral dynamics , providing the interaction of pseudoscalars among themselves , with vector mesons , and vector mesons among themselves .
it also offers a perspective on the chiral lagrangians as limiting cases at low energies of vector exchange diagrams occurring in the theory . in a more recent work , looking for poles in the @xmath14 scattering amplitudes , the @xmath15 channel is also included @xcite and a resonance around 1700 mev is dynamically generated , having the strongest coupling to this later channel . in the meson sector , the interaction of @xmath16 within this formalism has been addressed in @xcite , where it has been shown to lead to the dynamical generation of the @xmath17 and @xmath18 meson resonances , with a branching ratio for the sensitive @xmath19 decay channel in good agreement with experiment @xcite .
this work has been extended to the interaction of the su(3 ) vector mesons in @xcite , where several known resonances are also dynamically generated . in the present work we study the interaction of the octet of vector mesons with the octet of stable baryons , using the unitary approach in coupled channels .
we shall see that the scattering amplitudes lead to poles in the complex plane which can be associated to some well known resonances . under the approximation of neglecting the three momentum of the particles versus their mass
, we obtain degenerate states of @xmath0 , a pattern which seems to be followed qualitatively by the experimental spectrum , although in some cases the spin partners have not been identified . a different approach to account for the vector - baryon interaction
is the one followed in @xcite where pseudoscalar mesons , vector mesons and baryons mix , advocating su(6 ) symmetry for the interaction .
the approach leads to the same pseudoscalar - baryon interaction than the hidden gauge approach , but to different results when it comes to the interaction of the vector mesons with baryons .
in particular , the spin degeneracy predicted by the hidden gauge approach does not show up in the matrix elements of the potential in the su(6 ) scheme .
we follow the formalism of the hidden gauge interaction for vector mesons of @xcite ( see also @xcite for a practical set of feynman rules ) . the lagrangian involving the interaction of vector mesons amongst themselves is given by @xmath21 where the symbol @xmath22 stands for the trace in the @xmath23 space and @xmath24 is given by @xmath25\ , \label{vmunu}\ ] ] where @xmath26 is @xmath27 with @xmath28 mev the pion decay constant .
with the value of @xmath26 of eq .
( [ g ] ) one fulfills the ksfr rule @xcite which is tied to vector meson dominance @xcite .
the magnitude @xmath29 is the @xmath23 matrix of the vectors of the octet of the @xmath9 @xmath30 the lagrangian @xmath31 gives rise to a contact term coming from @xmath32[v_\mu , v_\nu]$ ] @xmath33 as well as to a three vector vertex from @xmath34 it is convenient to rewrite the lagrangian of eq .
( [ l3v ] ) as @xmath35 in this case one finds an analogy to the coupling of vectors to pseudoscalars given in the same theory by @xmath36v^{\nu}\rangle \ , \label{lagrvpp}\ ] ] where @xmath37 is the su(3 ) matrix of the pseudoscalar fields . in a similar way
, one obtains the lagrangian for the coupling of vector mesons to the baryon octet given by @xcite @xmath38\rangle + \langle \bar{b}\gamma_{\mu}b \rangle \langle v^{\mu}\rangle \right ) \label{lagr82}\ ] ] where @xmath39 is now the su(3 ) matrix of the baryon octet @xmath40 with these ingredients we can construct the feynman diagrams that lead to the @xmath41 and @xmath42 interaction , by exchanging a vector meson between the pseudoscalar or the vector meson and the baryon , as depicted in fig .
[ fig : feyn ] . from the diagram of fig .
[ fig : feyn](a ) , and under the low energy approximation of neglecting @xmath43 in the propagator of the exchanged vector , where @xmath44 is the momentum transfer , one obtains the same amplitudes as obtained from the ordinary chiral lagrangian for pseudoscalar - baryon octet interaction @xcite , namely the weinberg - tomozawa terms .
one could anticipate some analogy between the vector - baryon amplitudes with the pseudoscalar - baryon ones , given the similarity of the lagrangians in the way we have written them in eqs .
( [ l3vsimp ] ) and ( [ lagrvpp ] ) . however , one also anticipates differences . indeed , in the case of the pseudoscalar , fig .
[ fig : feyn](a ) , there is only one vector meson in the @xmath45 coupling and this must necessarily be the one that is exchanged in the diagram . however , in the vector case , fig .
[ fig : feyn](b ) , the @xmath46 vertex involves three vector mesons and any of them can correspond in principle to the exchanged vector in the diagram .
nevertheless , one can proceed consequently with neglecting @xmath43 as implicit in the chiral lagrangians @xcite , by also neglecting the three momentum of the external vectors versus the vector mass . in this case
, the polarization vectors of the external vector mesons have only spatial components , since the zeroth component is either zero for the transverse polarizations , or negligible for the longitudinal one , ( @xmath47 ) .
then , by examining the lagrangian of eq .
( [ l3vsimp ] ) one realizes that the field @xmath48 can not correspond to an external vector meson .
indeed , if this were the case , the @xmath49 index would be spatial and then the partial derivative @xmath50 would lead to a three momentum of the vector mesons which are neglected in the approach .
we can then conclude that the field @xmath48 corresponds to the exchanged vector and the analogy with the pseudoscalar and vector interaction is then evident .
indeed , they are formally identical , by substituting the octet of pseudoscalar fields by the octet of the vector fields , with the additional factor @xmath51 in the case of the interaction of the vector mesons .
note that @xmath52 becomes @xmath53 and the signs of the lagrangians also agree .
a small amendment is in order in the case of vector mesons , which is due to the mixing of @xmath54 and the singlet of su(3 ) , @xmath55 , to give the physical states of the @xmath56 and the @xmath57 mesons : @xmath58 given the structure of eq .
( [ eq : omephi ] ) , the singlet state which is accounted for by the v matrix , @xmath59 , does not provide any contribution to eq .
( [ l3vsimp ] ) , in which case all one must do is to take the matrix elements known for the @xmath60 interaction and , wherever @xmath37 corresponds to the @xmath61 , the amplitude should be multiplied by the factor @xmath62 to get the corresponding @xmath63 contribution , and by @xmath64 to get the corresponding @xmath57 contribution . upon the approximation consistent with neglecting the three momentum versus the mass of the particles ( in this case the baryon ) , we can just take the @xmath65 component of eq .
( [ lagr82 ] ) and then the transition potential corresponding to the diagram of [ fig : feyn](b ) is given by @xmath66 where @xmath67 are the energies of the incoming and outgoing vector meson .
the @xmath68 coefficients of eq .
( [ kernel ] ) can be obtained directly from @xcite with the simple rules given above for the @xmath56 and the @xmath57 mesons , and substituting @xmath69 by @xmath9 and @xmath70 by @xmath71 in the matrix elements .
the coefficients are obtained both in the physical basis of states or in the isospin basis . here
we will directly study the interaction in isospin basis and in the appendix we collect the tables of the @xmath68 coefficients for different states of isospin , @xmath72 , and strangeness , @xmath73
. the tables immediately show the sectors where there is attraction and therefore give chances to find bound states or resonances .
we can see that the cases with @xmath74 , @xmath75 and @xmath76 , the last two corresponding to exotic channels , are repulsive and do not produce poles in the scattering matrices .
however , the sectors @xmath77 , @xmath78 , @xmath79 and @xmath80 are attractive and we expect to find bound states and resonances in these cases .
the next step to construct the scattering matrix implies solving the coupled channels bethe salpeter equation in the on shell factorization approach of @xcite @xmath81^{-1}\ , v \label{eq : bethe}\ ] ] with @xmath82 being the loop function of a vector meson and a baryon which we calculate in dimensional regularization using the formula of @xcite @xmath83 \right\ } \ , \label{eq : gpropdr}\end{aligned}\ ] ] with @xmath84 a regularization scale , which we take to be 630 mev , and with a natural value of the subtraction constant @xmath85 of @xmath86 , as determined in @xcite .
the iteration of diagrams implicit in the bethe salpeter equation in the case of the vector mesons has a subtlety with respect to the case of the pseudoscalars .
the @xmath51 term of the interaction forces the intermediate vector mesons in the loops to propagate with the spatial components in the loops . we need to sum over the polarizations of the internal vector mesons which , because they are tied to the external ones through the @xmath51 factor , provides @xmath87
as shown in @xcite , the on shell factorization leads to a correction coefficient in the @xmath82 function of @xmath88 versus unity , which is negligibly small , and which we also neglect here in consonance with the approximations done .
in this case the factor @xmath51 , appearing in the potential @xmath89 , factorizes also in the @xmath90 matrix for the external vector mesons .
the formalism described above would provide results obtained using fixed masses for the vector mesons and no width .
the mass distributions of the @xmath9 and @xmath71 mesons are sufficiently extended to advise a more accurate calculation that takes this large width into account .
we follow the traditional method of convoluting the @xmath82 function with the mass distributions of the @xmath9 or @xmath71 mesons , as is customarily made @xcite .
one can prove that this convolution is equivalent to calculating the loop function with the dressed vector meson propagator written in terms of its lehmann representation , as is done in calculations of medium effects in the scattering matrices @xcite .
the method amounts to replacing the @xmath82 function by @xmath91 obtained as @xmath92 with @xmath93 being the normalization factor , and @xmath94 the decay width of the meson ( @xmath95 ) , which we take to be 149.4 mev and 50.5 mev for the @xmath9 and @xmath71 meson , respectively .
the energy dependent width @xmath96 for the @xmath9 meson , obtained from its decay into two pions in @xmath97-wave , is given by @xmath98 a similar expression gives the energy dependent width of the @xmath71 meson from its decay into a @xmath70 meson and a pion .
we will see that , using fixed masses for the vector mesons , one finds bound states in the @xmath15 and @xmath99 amplitudes , i.e. states having zero width . however , when @xmath91 is used in eq .
( [ eq : bethe ] ) and , therefore , both the @xmath9 and @xmath71 vector mesons are taken with their corresponding mass distribution , there is phase space for the decay of each of these bound states into some of the mass components of the vector meson and the nucleon , thereby acquiring an appreciable width .
we search for poles in the scattering matrices in the second riemann sheet , as defined in previous works @xcite , basically changing @xmath100 by to @xmath101 in the analytical formula of the @xmath82 function , eq . ( [ eq : gpropdr ] ) , for channels where re@xmath102 is above the threshold of the corresponding channel .
when one has a mass distribution of the @xmath9 and @xmath71 mesons , and hence a fuzzy description of the threshold for some channels , one could take different prescriptions for going to the optimal riemann sheet that better reflects the behavior of the amplitude in the real axis , which is where the physical information is contained .
the results are very similar in all cases , expect when one has a resonance very close to threshold , where the convolution can distort the shape of the amplitude and make even the pole disappear . in view of that , for these cases the couplings are obtained from the amplitudes in the real axis as follows . assuming these amplitudes to behave as @xmath103 where @xmath104 is the position of the maximum of @xmath105 , with @xmath106 being the channel to which the resonance couples more strongly , and @xmath107 its width at half - maximum , one then finds @xmath108 up to a global phase , this expression allows one to determine the value of @xmath109 , which we take to be real .
the other couplings are then derived from @xmath110 this procedure to obtain the couplings from @xmath111 in the real axis was used in @xcite where it was found that changes in the input parameters that lead to moderate changes in the position and the width of the states affected the couplings more smoothly . having this in mind
, we also calculate , as a check , the couplings in the case of the particles without width from the residues of the amplitudes in the first riemann sheet of the complex plane .
the couplings obtained in this way are very similar to those obtained by means of eqs .
( [ couprealaxis ] ) and ( [ couprealaxis2 ] ) employing amplitudes in the real axis obtained with the convolution method .
the differences between the couplings obtained with the two methods are of the order of 10 - 20 % .
in this section we show our results obtained in the attractive sectors mentioned above . in fig .
[ fig : s0 ] we show the results of @xmath112 as a function of @xmath113 for the different channels in the @xmath77 sector .
we can see a neat bump for @xmath114 around 1700 mev , a few mev below the @xmath15 threshold .
the width of the peak is due to the convolution of the @xmath9 mass . for a @xmath9 of fixed mass and no width
one obtains a bound state with a small binding .
the bump around 1700 mev is also seen in the @xmath115 channel but is absent or barely visible in the @xmath116 , @xmath117 and @xmath118 channels . on the other hand , with the same strangeness and isospin ,
one finds a second peak around 1970 mev , which is clearly visible in the @xmath115 , @xmath116 , @xmath117 and @xmath118 channels but not visible in the @xmath15 one . this special behavior has to do with the coupling of the resonances to the different channels which will be studied below . in fig .
[ fig : sm1 ] we show @xmath112 for two cases , @xmath124 and @xmath79 .
for @xmath125 we see again a clear peak behavior of @xmath112 around 1800 mev in the @xmath126 channel .
the structure is also seen in all the other channels , except in the @xmath127 one . in the @xmath128 channel
one can see another structure around 1920 mev , below the @xmath128 threshold , which could be associated to a bound state of this system .
finally in the @xmath127 channel one finds another clear structure corresponding to a resonance around 2140 mev , practically invisible in the other channels , which could qualify as a @xmath127 bound state .
the @xmath129 case shows a clear peak around 1830 mev in the @xmath130 channel .
this structure is also seen in other channels , but it could be the cusp due to the opening of the @xmath126 channel .
we will come back to this when we make a search for the poles of the @xmath90 matrix . in the @xmath131 channel , and
also visible in the @xmath127 channel , we see another less pronounced structure around 1990 mev . in fig .
[ fig : sm2 ] we display the sector @xmath132 . in this case
we find a clear structure of resonant type around 2030 mev which is visible in the @xmath133 and @xmath134 channels .
another clear peak appears around 2080 mev , shortly below the @xmath118 threshold , which is more pronounced in the @xmath135 channel and appears as an interference minimum in the @xmath134 channel .
since the spin dependence only comes from the @xmath51 factor and there is no dependence on the spin of the baryons , the interaction for vector - baryon states with @xmath136 and @xmath137 is the same and then one should associate each peak with the existence of two degenerate states .
the spin degeneration also appears in some quarks models @xcite .
.pole position and coupling constants to various channels of the resonances found in the @xmath138 sector .
@xmath139 the properties of this state have been determined from the amplitudes in the real axis , as explained in the text . [
cols="^,^,^,^,^",options="header " , ] n. kaiser , p. b. siegel and w. weise , phys .
b * 362 * , 23 ( 1995 ) [ arxiv : nucl - th/9507036 ] .
n. kaiser , p. b. siegel and w. weise , nucl .
a * 594 * , 325 ( 1995 ) [ arxiv : nucl - th/9505043 ] .
n. kaiser , t. waas and w. weise , nucl .
a * 612 * , 297 ( 1997 ) [ arxiv : hep - ph/9607459 ] .
e. oset and a. ramos , nucl .
a * 635 * ( 1998 ) 99 .
j. a. oller and u. g. meissner , phys .
b * 500 * , 263 ( 2001 ) [ arxiv : hep - ph/0011146 ] . c. garcia - recio , m. f. m. lutz and j. nieves , phys .
b * 582 * ( 2004 ) 49 . c. garcia - recio , j. nieves , e. ruiz arriola and m. j. vicente vacas , phys .
d * 67 * , 076009 ( 2003 ) [ arxiv : hep - ph/0210311 ] .
t. hyodo , s. i. nam , d. jido and a. hosaka , phys .
c * 68 * , 018201 ( 2003 ) [ arxiv : nucl - th/0212026 ] .
t. hyodo , d. jido and a. hosaka , phys .
d * 75 * , 034002 ( 2007 ) [ arxiv : hep - ph/0611004 ] .
m. doring , e. oset and s. sarkar , phys .
c * 74 * , 065204 ( 2006 ) [ arxiv : nucl - th/0601027 ] .
v. k. magas , e. oset and a. ramos , phys .
lett . * 95 * , 052301 ( 2005 ) [ arxiv : hep - ph/0503043 ] .
d. jido , m. doering and e. oset , phys .
c * 77 * , 065207 ( 2008 ) [ arxiv:0712.0038 [ nucl - th ] ] .
t. hyodo , d. jido and a. hosaka , phys . rev .
c * 78 * , 025203 ( 2008 ) [ arxiv:0803.2550 [ nucl - th ] ] .
a. martinez torres , k. p. khemchandani and e. oset , phys .
c * 77 * , 042203 ( 2008 ) [ arxiv:0706.2330 [ nucl - th ] ] .
a. martinez torres , k. p. khemchandani and e. oset , eur .
phys . j. a * 35 * , 295 ( 2008 ) [ arxiv:0805.3641 [ nucl - th ] ] .
k. p. khemchandani , a. martinez torres and e. oset , eur .
j. a * 37 * , 233 ( 2008 ) [ arxiv:0804.4670 [ nucl - th ] ] .
d. jido and y. kanada - enyo , phys .
c * 78 * , 035203 ( 2008 ) .
y. kanada - enyo and d. jido , phys .
c * 78 * , 025212 ( 2008 ) [ arxiv:0804.3124 [ nucl - th ] ] .
m. bando , t. kugo and k. yamawaki , phys .
* 164 * , 217 ( 1988 ) .
m. harada and k. yamawaki , phys
. rept . * 381 * , 1 ( 2003 ) [ arxiv : hep - ph/0302103 ] .
m. doring , c. hanhart , f. huang , s. krewald and u. g. meissner , arxiv:0903.4337 [ nucl - th ] .
r. molina , d. nicmorus and e. oset , phys .
d * 78 * , 114018 ( 2008 ) [ arxiv:0809.2233 [ hep - ph ] ] .
h. nagahiro , j. yamagata - sekihara , e. oset and s. hirenzaki , arxiv:0809.3717 [ hep - ph ] .
l. s. geng and e. oset , phys .
d * 79 * , 074009 ( 2009 ) arxiv:0812.1199 [ hep - ph ] . c. garcia - recio , j. nieves and l. l. salcedo , phys .
d * 74 * , 034025 ( 2006 ) [ arxiv : hep - ph/0505233 ] .
f. klingl , n. kaiser and w. weise , nucl .
a * 624 * ( 1997 ) 527 [ arxiv : hep - ph/9704398 ] .
j. e. palomar and e. oset , nucl .
a * 716 * , 169 ( 2003 ) [ arxiv : nucl - th/0208013 ] .
g. ecker , prog .
35 ( 1995 ) 1 a. ramos , e. oset and c. bennhold , phys .
lett . *
89 * , 252001 ( 2002 ) [ arxiv : nucl - th/0204044 ] .
t. inoue , e. oset and m. j. vicente vacas , phys .
c * 65 * , 035204 ( 2002 ) [ arxiv : hep - ph/0110333 ] .
h. nagahiro , l. roca and e. oset , eur .
phys . j. a * 36 * , 73 ( 2008 ) [ arxiv:0802.0455 [ hep - ph ] ] .
l. roca , e. oset and j. singh , phys .
d * 72 * , 014002 ( 2005 ) [ arxiv : hep - ph/0503273 ] .
l. tolos , a. ramos and e. oset , phys .
c * 74 * , 015203 ( 2006 ) [ arxiv : nucl - th/0603033 ] .
m. kirchbach and c. b. compean , eur .
j. a 39 , 1 ( 2007 ) arxiv:0805.2404 [ hep - ph ] . c. amsler
_ et al . _
[ particle data group ] , phys .
b * 667 * , 1 ( 2008 ) . c. schutz , j. haidenbauer , j. speth and j. w. durso , phys .
c * 57 * , 1464 ( 1998 ) ; o. krehl , c. hanhart , s. krewald and j. speth , phys . rev .
c * 62 * , 025207 ( 2000 ) ; a. m. gasparyan , j. haidenbauer , c. hanhart and j. speth , phys .
c * 68 * , 045207 ( 2003 ) .
m. doring , c. hanhart , f. huang , s. krewald and u. g. meissner , arxiv:0903.1781 [ nucl - th ] .
b. m. k. nefkens , aip conf .
proc . * 870 * , 405 ( 2006 ) .
j. w. price _
[ clas collaboration ] , phys .
c * 71 * , 058201 ( 2005 ) [ arxiv : nucl - ex/0409030 ] . | we study the interaction of vector mesons with the octet of stable baryons in the framework of the local hidden gauge formalism using a coupled channels unitary approach .
we examine the scattering amplitudes and their poles , which can be associated to known @xmath0 baryon resonances , in some cases , or give predictions in other ones .
the formalism employed produces doublets of degenerate @xmath0 states , a pattern which is observed experimentally in several cases .
the findings of this work should also be useful to guide present experimental programs searching for new resonances , in particular in the strange sector where the current information is very poor .
@xmath1 departamento de fsica terica and ific , centro mixto universidad de valencia - csic , institutos de investigacin de paterna , aptdo .
22085 , 46071 valencia , spain + @xmath2 departament destructura i constituents de la matria and institut de cincies del cosmos , universitat de barcelona , 08028 barcelona , spain + |
fig.1 : the schematic representation of excitation - deexcitation processes taking place at the band edges .
fig.2 : the energy levels in units of @xmath69=@xmath70 for zinc blende inas qds as a function of r in the ( a ) absence and ( b ) presence of e - h coulomb interaction(without exchange ) .
the energy reference is taken as the top of the valence band without the spin orbit effects included .
eigenvalues in the absence(@xmath43 ) as well as presence(@xmath44 of dielectric mismatch effects are also shown in ( b ) for both the cases of infinite and finite barrier where @xmath71 .
( c ) eigenvalues including e - h coulomb and exchange interactions .
the states are labelled with @xmath61 with superscripts standing for upper and lower state .
the midpoint of the exciton energies with @xmath63=@xmath72 and @xmath63=@xmath73 is taken as the reference energy ( zero of energy ) .
fig.3 : the energy levels in units of @xmath74 for wurtzite cdse qds as a function of dot radius r in the ( a ) absence and ( b ) presence of e - h coulomb interaction and dielectric mismatch effects(without exchange ) for both the cases of infinite and finite barrier .
the energy reference is taken as the top of the valence band with the spin orbit effects included .
the states are labelled as @xmath75 etc , where capital letters represent the dominant @xmath21 present and the subscripts stand for x- , y - like or z - like states , the arrow indicating the spin state .
( c ) eigenvalues including the e - h coulomb and exchange interactions .
fig.4 : singlet triplet splitting in inas in the absence @xmath43 and presence @xmath44 of dielectric mismatch effects for both the cases of infinite and finite barrier .
the inset gives the energy differences of s and p states .
experimental data(empty circle and square ) are taken from ref.4 .
fig.5 : the singlet triplet splitting in cdse in the absence and presence of dielectric mismatch effects for both the cases of infinite and finite barrier .
the inset gives the energy difference of s and p states with @xmath34 and @xmath54 respectively .
experimental data(empty circle ) are taken from ref.3 . | the mechanism of stokes shift in semiconductor quantum dots is investigated by calculating the energy of the excitonic states .
we have taken into account all possible contributions to the total electronic energy in the dot , i.e , dielectric mismatch between dot and surrounding medium , the effects of finite barrier height and electron - hole exchange interaction .
the stokes shift is calculated as a function of radius of dot and compared with experimental data on two different semiconductor based quantum dots .
these results provide evidence for exchange splitting of excitonic states , as the mechanism of stokes shift in quantum dot the red shift of the emission spectra with respect to absorption spectra is known as stokes shift , which is commonly observed in semiconductor quantum dots(qds ) and the most important quantity that determines the optical properties of qds .
however the underlying mechanism of stokes shift is still controversial . in case of semiconductor qds
, this shift decreases with the increase in radius of the dot and disappears beyond a certain radius .
the red shift , principally , occurs if either the top of the valence band is an optically passive state ( say p state ) or if the electron and the hole are in a triplet state .
absorption of a photon from the top of the valence band in such cases is not allowed and is possible only from an optically active state lying deeper in the valence band .
the exciton , once formed after absorption , can not decay to the top of the valence band by direct dipole transition and hence is called `` dark exciton''@xcite .
deexcitation eventually takes place with the help of phonons thus giving rise to red shifted photons .
two different mechanisms@xcite@xmath0@xcite have been proposed to explain the origin of stokes shift in semiconductor qds , as shown in fig .
1 . absorption takes place from the s - state which lies deeper than the p - state in the valence band to form an exciton in the singlet state of the electron and hole . according to the first mechanism , ( 1 ) of fig .
1 , deexcitation takes place into the p state with the help of phonons giving rise to a stokes shift by an amount @xmath1 representing the difference in energy between the exciton states formed with s and p hole states(ref.2 ) .
according to the second mechanism the exciton in the singlet state first thermalizes into a triplet state from where it deexcites with the help of phonons either ( a ) to the deeper lying original s state , ( 2 ) of fig .
1 , giving rise to a stokes shift by an amount @xmath2 representing the singlet triplet splitting(ref .
3 ) , or ( b ) to the p state at the top of the valence band , ( 3 ) of fig .
1 , with the stokes shift given by @xmath3 .
in addition to this , the red shift is also observed if the final excited state has a different atomic configuration than the initial ground state .
such a red shift , known as frank condon shift , is commonly observed in molecules and point defects in solids and may be relevant in ultra small qds@xcite . in this letter
we investigate the origin of stokes shift in iii - v(inas ) and ii - vi(cdse ) semiconductor based qds by calculating the energies of excitonic states .
we have taken into account the dielectric mismatch effects , the effect of finite barrier height and the effect of the electron hole exchange interaction and then compared the results with experimental data .
inas and cdse are chosen because of their ( i ) different crystal structures , ( ii ) different crystal field effects which is zero in zinc blende structures(i.e inas ) but significantly large in wurtzite structure(cdse ) , ( iii ) different effective masses which is small in inas , but large for cdse and ( iv ) different nature of bonds .
it has been observed experimentally@xcite@xmath0@xcite@xmath0@xcite that the red shift decreases with radius and disappears beyond a certain radius .
essentially , the strong dependence of stokes shift on the radius of the qd , is used to investigate the origin of stokes shift in qds based on these semiconductors(inas , cdse ) .
optical transitions from the electron state @xmath4 , where @xmath5 are the bloch functions of the conduction band , are possible only to hole states in the valence band which have the s - state as a component because @xmath6 .
the optical transition probability from the bottom of the conduction band to a hole state in the valence band is given by @xcite @xmath7 , where c is the component of the s - state @xmath8 in the hole wavefunction , @xmath9 is the spin projection of the removed electron in the hole state and @xmath10 are the bloch wavefunctions of the hole states at the valence band top .
@xmath11 is zero if @xmath12 .
optical transitions from the bottom of the conduction band to the top of the valence band is not possible if either ( 1 ) the hole state does not have a component of the s - state ( i.e if c=0 ) , or ( 2 ) the spin projections of the removed electron in the hole state and the electron in the conduction band are not equal , @xmath12 . since the hole spin is obtained by flipping the spin of the removed electron , it follows that optical transitions are not possible if the spins of the electron and the hole are parallel i.e if they are in a triplet state .
thus , if the top of the valence band is a p - state , or the electron and the hole are in a triplet state , a _ dark exciton _ is formed .
the exciton states are obtained from the hamiltonian @xmath13 where @xmath14 are the electronic hamiltonian , the hole hamiltonian , the spin - orbit interaction and the e - h coulomb interaction respectively given in ref.2 and 7 .
the last two terms represent the surface polarization energies arising due to the difference in the dielectric constants between the semiconductor quantum dot and the surrounding medium@xcite .
@xmath15 is the self energy of the electron and hole due to their image charges and @xmath16 is the mutual interaction energy between the electron and hole via image charges .
the excitonic states obtained from eq.([h1 ] ) will be split further due to exchange interaction@xcite @xmath17 between the electron and hole .
since the nature of the state at the top of the valence band plays a pivotal role in the formation of dark and bright excitons , we first investigate the hole states for both the zinc - blende and wurtzite structures using the * k.p * hole hamiltonian @xmath18 given in ref.9 .
the energy eigenvalues of the hole states without inclusion of excitonic effects as a function of dot radius for inas(zinc - blende structures ) and cdse(wurtzite structures ) are shown in figs .
2(a ) and 3(a ) respectively .
these eigenvalues get modified when coulomb interaction between electron in the conduction band and hole in the valence band are taken into account .
the corresponding results including excitonic effects are shown in figs .
2(b ) and 3(b ) respectively . from fig .
2(a ) we observe that in inas the optically passive @xmath19 is the ground state which lies below the first optically active state @xmath20 state .
the states are labeled @xmath20 , @xmath19 etc where the capital letters correspond to the lowest @xmath21 present and the subscripts gives the total angular momentum @xmath22=@xmath23 , where @xmath21 is the orbital angular momentum , @xmath24(=1 ) is angular momentum of the bloch wavefunction at the valence band top and @xmath25 is the spin of the hole .
the difference in energy between the optically passive @xmath19 state and the optically active @xmath20 state decreases as the radius of the dot increases and eventually @xmath20 becomes the ground state at around @xmath26 .
hence dark exciton formation and associated stokes shift will be observed below this radius . for the wurtzite structure ,
the hamiltonian is axially symmetric and only the z - component @xmath27 of the total angular momentum is conserved . from fig.3(a )
it is observed that for cdse the optically passive @xmath28 with @xmath29 state is the ground state at @xmath30 .
the energy of @xmath28 in units of @xmath31 ( @xmath32 is a luttinger parameter ) has an upward slope as a function of radius and crosses the relatively flat @xmath33 state with @xmath34 at @xmath35 .
dark exciton formation and stokes shift is predicted for cdse below @xmath35 .
the trend of s - state and p state can be explained by observing that @xmath33 with @xmath34 is predominantly made from only one state with composition @xmath36 whose energy decreases as @xmath37 but in units of @xmath38 is independent of r and hence the state is relatively flat .
but the optically passive p state with @xmath29 can be made either from the state with composition @xmath39 or the state with the composition @xmath40 .
these two states are separated from each other by spin - orbit interaction . at very low radii spin orbit effects
are negligible relative to kinetic energy terms due to confinement .
hence the two states are almost degenerate at @xmath30 .
the coupling between these two states give rise to a splitting of the states with the lower p state lying below the s state . as the radius increases the energies of both
the s and p states decrease .
the energy of s state decreases as @xmath37 but the energy of the p state decreases more slowly because of the presence of the spin - orbit interaction energy which becomes important as the kinetic energy starts decreasing with the increase of radius
. eventually the two levels cross . in figs . 2 and 3 the energies have been shown in the units of @xmath38 . when the coulomb interaction between the electron in @xmath41 state and the hole is taken into account ,
the energies decrease . the decrease in energy of s - state is more than that of p state , the overlap between @xmath41 state and s state of the hole being greater .
this causes the crossing between s and p states to occur at lower values of r as can be seen from comparison of figs .
2(b ) and 3(b ) with the corresponding figs . 2(a ) and 3(a ) .
the differences between the excitonic s and p states @xmath42 as a function of r , are shown in the insets to figs . 4 and 5 .
the surface polarization charges arising due to the dielectric mismatch between the dot and the surroundings also affect the excitonic energies .
the contributions of self energy @xmath15 and attractive energy @xmath16 are opposite in sign and comparable in magnitude@xcite .
hence the dielectric mismatch effects are greatly reduced with the net effect being an increase in energy depending on the amount of dielectric mismatch between the dot and the surrounding .
the increase in energy for the p state is more than that for the s state resulting in a crossover to occur at a lower dot radius .
this can be seen from figs.2(b ) and 3(b ) which gives the comparison between zero and finite dielectric mismatch , i.e @xmath43 and @xmath44 . the self energy of a charge confined to a spherical dielectric is minimum@xcite at r=0 . since @xmath45 for the p - state is larger compared to the s - state , the self energy @xmath15 is higher for the p - state . but the attractive energy @xmath16 is higher for the s state compared to the p state .
this leads to a higher net increase of polarization contribution @xmath46 $ ] for the p - state as compared to the s - state and hence to a shifting of the s - p crossover to lower values of r as can be seen from figs .
2(b)and 3(b ) . another cause of dark exciton formation and the consequent stokes shift is the splitting of the active exciton states due to electron hole exchange interaction@xcite @xmath47 where @xmath48 is the exchange strength constant , @xmath49 the lattice constant , @xmath50 the pauli spin matrices representing the electron spin and @xmath51 is the hole spin matrix .
the exchange interaction gives rise to a splitting between triplet and singlet state of the electron and hole , with the triplet state lying lower in energy .
the triplet state being passive , excitation takes place to the higher singlet state .
deexcitation can take place from the triplet state reached by thermalization , thus giving rise to a red shift of the emmision spectra with respect to the absorption spectra . in inas the first optically active s state ( fig .
2(a ) ) has total anguar momentum @xmath52 @xmath53(@xmath21 = 0 , @xmath24=1 , @xmath25=@xmath54 ) and is fourfold degenerate .
the @xmath41 electron state in the conduction band being doubly degenerate , the exciton states are eightfold degenerate in the absence of exchange interaction .
the exchange interaction splits these degenerate levels into two : the lower optically passive triplet state with @xmath55=2(@xmath56 ) and the upper optically active singlet state with @xmath55=1 ( fig .
2(c ) ) .
the difference in energies @xmath2 between the singlet and triplet states as a function of dot radius is shown in fig .
4 . @xmath2 decreases with r and vanishes at @xmath57 .
the exchange interaction is enhanced with decrease in radius because of the increase in the overlap of electron and hole wavefunctions due to confinement . in cdse ( fig .
3(c ) ) the ground state with @xmath58 can be obtained by coupling a hole with @xmath59 with the electron state with @xmath60 .
since these combinations form triplet states , the ground state with @xmath58 is optically passive and forms a dark exciton .
the first excited state with @xmath61=@xmath62 can be obtained either by coupling @xmath63=@xmath64 , @xmath65=@xmath66 or alternatively by @xmath63=@xmath67 with @xmath65=@xmath67 .
the state @xmath68 will be a mixture of singlet and triplet states and will be optically active .
the difference in energies @xmath2 between the optically passive ground state and the first optically active excited state as function of r for cdse is shown in fig .
5 . as in the case of inas , @xmath2 for cdse also decreases with r and vanishes around @xmath26 . as seen from figs . 2 , 3 , 4 and 5
the curves corresponding to finite barrier case lie lower than infinite barrier case .
both the exchange interaction and coulomb interaction depend on the amount of overlap of electron and hole wavefunctions , which is reduced in the finite barrier case due to penetration of the substantial amount of electron wavefunction outside the dot resulting in the decrease of exchange energy and coulomb binding energy . having theoretically calculated @xmath1 and @xmath2 we are in a position to investigate the various mechanisms for stokes shift given in fig . 1 by comparison with experimental data . in figs .
4 and 5 are shown the plots of @xmath2 as a function of the dot radius together with the experimentally observed stokes shift for cdse and inas respectively .
@xmath1 as a function of r for these cases are shown in the insets of the corresponding figures .
it is observed that @xmath1 is much larger than the experimental red shift for inas . for cdse , as seen from the inset of fig .
5 at @xmath30 @xmath1 is almost the same as the experimental stokes shift but for larger values of r the experimental values are much higher .
it is seen from figs . 4 and 5 that @xmath2 fits the experimental data rather well .
we observed that the fit for cdse corresponds to an infinite barrier at the dot radius but for inas a good fit is obtained for a finite barrier height .
this can be explained by observing that the effective mass of electron in inas is very small and hence the lowest eigenvalue lies near the top of the barrier height and there is substantial probability of finding the electron outside the dot . for cdse ,
the effective mass of electron is large and the lowest eigenvalue lies near the bottom of the potential well , making the infinite barrier effects predominant . in fig .
5 we have also plotted the theoretical curve for stokes shift obtained by efros et al@xcite in which the coulomb interaction between the electron and hole was not taken into account . we observe that better agreement with experimental stokes shift is obtained by our theoretical curve which takes into account the coulomb interaction . from a comparision of the theoretical results with experimental data
we conclude that the stokes shift observed in qds is caused by the splitting of the exciton states by the electron hole exchange interaction through the deexcitation mechanism ( 2 ) in fig
. 1
. 99 m. nirmal , d. j. norris , m. kuno , m. g. bawendi , al .
l. efros , and m. rosen , phys .
rev .
lett . * 75 * , 3728 ( 1995 ) .
j. li and j. b. xia , phys .
rev .
b. * 61 * , 15880 ( 2000 ) .
. l. efros , m. rosen , m. kuno , m. nirmal , d. j. norris and m. g. bawendi , phys .
rev .
b. * 54 * , 4843 ( 1996 ) .
u. banin , j. c. lee , a. a. guzelian , a. v. kadavanich and a. p. alivisatos , superlattices microstruct .
* 22 * , 559 ( 1997 ) a. franceschetti and s. t. pantelides , phys .
rev .
b. * 68 * , 033313 ( 2003 ) .
al .
l. efros , phys .
rev .
b. * 46 * , 7448 ( 1992 ) .
a.bagga , p. k. chattopadhyay and s.ghosh , phys .
rev .
b. ( in press ) .
t. takagahara , phys .
rev .
b. * 47 * , 4569 ( 1993 ) .
a. bagga , p. k. chattopadhyay and s. ghosh , phys .
rev .
b. * 68 * , 155331 ( 2003 ) .
v. a. fonoberov , e. p. pokatilov and a. a. balandin , phys .
rev .
b. * 66 * , 085310 ( 2002 ) .
l. e. brus , j. chem .
phys * 79 * , 5566 ( 1983 ) . |
let @xmath14 denote the rational numbers and @xmath15 $ ] denote the ring of polynomials over @xmath14 .
many classical combinatorial sequences can be defined as connection coefficients between various basis of the polynomial ring @xmath15 $ ] .
there are three very natural bases for @xmath15 $ ] . the usual power basis @xmath16 , the falling factorial basis @xmath17 , and the rising factorial basis @xmath18 .
here we let @xmath19 and for @xmath7 , @xmath20 and @xmath21 .
then the stirling numbers of the first kind @xmath22 , the stirling numbers of the second kind @xmath23 and the lah numbers @xmath24 are defined by specifying that for all @xmath25 , @xmath26 the signless stirling numbers of the first kind are defined by setting @xmath27 .
then it is well known that @xmath28 , @xmath23 , and @xmath24 can also be defined by the recursions that @xmath29 , @xmath30 if either @xmath31 or @xmath32 , and @xmath33 for all @xmath34 .
there are well known combinatorial interpretations of these connection coefficients .
that is , @xmath23 is the number of set partitions of @xmath35 = \{1 , \ldots , n\}$ ] into @xmath36 parts , @xmath28 is the number of permutations in the symmetric group @xmath37 with @xmath36 cycles , and @xmath24 is the number of ways to place @xmath38 labeled balls into @xmath36 unlabeled tubes with at least one ball in each tube . in @xcite , we introduced fibonacci analogues of the number @xmath22 , @xmath23 , and @xmath24 .
we started with the tiling model of the @xmath0 of @xcite .
that is , let @xmath39 denote the set of tilings a column of height @xmath38 with tiles of height 1 or 2 such that bottom most tile is of height 1 .
for example , possible tiling configurations for @xmath40 for @xmath41 are shown in for each tiling @xmath42 , we let @xmath43 is the number of tiles of height 1 in @xmath44 and @xmath45 is the number of tiles of height 2 in @xmath44 and define @xmath46 it is easy to see that @xmath47 , @xmath48 and @xmath49 for @xmath50 so that @xmath51 .
we then defined the @xmath52-fibo - falling factorial basis @xmath53 and the @xmath52-fibo - rising factorial basis @xmath54 by setting @xmath55 and setting @xmath56 for @xmath7 .
our idea to define @xmath52-fibonacci analogues of the stirling numbers of the first kind , @xmath57 , the stirling numbers of the second kind , @xmath58 , and the lah numbers , @xmath59 , is to define them to be the connection coefficients between the usual power basis @xmath60 and the @xmath52-fibo - rising factorial and @xmath52-fibo - falling factorial bases .
that is , we define @xmath57 , @xmath58 , and @xmath59 by the equations @xmath61
@xmath63 for all @xmath25 .
it is easy to see that these equations imply simple recursions for the connection coefficients @xmath57s , @xmath58s , and @xmath59s .
that is , @xmath57s , @xmath58s , and @xmath59s can be defined by the following recursions @xmath64 plus the boundary conditions @xmath65 and @xmath66 if @xmath67 or @xmath32 .
if we define @xmath68 , then @xmath69s can be defined by the recursions @xmath70 plus the boundary conditions @xmath71 and @xmath72 if @xmath67 or @xmath32 .
it also follows that @xmath73 in @xcite , we developed a new rook theory model to give a combinatorial interpretation of the @xmath69s and the @xmath58s and to give combinatorial proofs of their basic properties .
this new rook theory model is a modification of the rook theory model for @xmath23 and @xmath28 except that we replace rooks by fibonacci tilings
. the main goal of this paper is to show how that model can be modified to give combinatorial interpretations to two new @xmath10-analogues of the @xmath74s and the @xmath75s .
let @xmath76_q = 1 $ ] and @xmath11_q = \frac{1-q^x}{1-q}$ ] .
when @xmath38 is a positive integer , then @xmath35_q = 1 + q+ \cdots + q^{n-1}$ ] is the usual @xmath10-analogue of @xmath38 .
then there are two natural analogues of the falling and rising fibo - factorial basis .
first we let @xmath11_{\downarrow_{q , f,0 } } = \overline{[x]}_{\downarrow_{q , f,0 } } = [ x]_{\uparrow_{q , f,0 } } = \overline{[x]}_{\uparrow_{q , f,0}}=1 $ ] . for @xmath12 , we let @xmath12 , @xmath77_{\downarrow_{q , f , k } } & = & [ x]_q [ x - f_1]_q \cdots [ x - f_{k-1}]_q , \\ \ \overline{[x]}_{\downarrow_{q ,
f , k}}&= & [ x]_q ( [ x]_q-[f_1]_q ) \cdots ( [ x]_q-[f_{k-1}]_q ) , \\ \ [ x]_{\uparrow_{q , f , k}}&= & [ x]_q [ x+f_1]_q \cdots [ x+f_{k-1}]_q , \ \mbox{and } \\ \
\overline{[x]}_{\uparrow_{q , f , k}}&= & [ x]_q ( [ x]_q+[f_1]_q ) \cdots ( [ x]_q+[f_{k-1}]_q).\end{aligned}\ ] ] then we define @xmath78 and @xmath79 by the equations @xmath80_{\uparrow_{q , f , n}}= \sum_{k=1}^n \mathbf{cf}_{n , k}(q ) [ x]_q^k\ ] ] and @xmath81}_{\uparrow_{q , f ,
n}}= \sum_{k=1}^n \overline{\mathbf{cf}}_{n , k}(q)[x]_q^k.\ ] ] similarly , we define @xmath82 and @xmath83 by the equations @xmath84_q^n=
\sum_{k=1}^n \mathbf{sf}_{n , k}(q ) [ x]_{\downarrow_{q , f , k}}\ ] ] and @xmath85_q^n= \sum_{k=1}^n \overline{\mathbf{sf}}_{n , k}(q ) \overline{[x]}_{\downarrow_{q , f , k}}\ ] ] one can easily find recursions for these polynomials .
for example , @xmath86_q^{n+1}&= & \sum_{k=1}^{n+1 } \mathbf{sf}_{n+1,k}(q ) [ x]_{\downarrow_{q , f , k } } = \sum_{k=1}^n \mathbf{sf}_{n , k}(q ) [ x]_{\downarrow_{q , f , k}}[x]_q \\ & = & \sum_{k=1}^n \mathbf{sf}_{n , k}(q ) [ x]_{\downarrow_{q , f ,
k}}([f_k]_q + q^{f_k}[x - f_k]_q)\\ & = & \sum_{k=1}^n [ f_k]_q \mathbf{sf}_{n , k}(q ) [ x]_{\downarrow_{q , f , k } } + \sum_{k=1}^n q^{f_k}\mathbf{sf}_{n , k}(q ) [ x]_{\downarrow_{q , f , k+1}}.\end{aligned}\ ] ] taking the coefficient of @xmath11_{\downarrow_{q , f , k}}[x]_q$ ] on both sides shows that @xmath87_q \mathbf{sf}_{n , k}(q)\ ] ] for @xmath88 .
it is then easy to check that the @xmath82s can be defined by the recursions ( [ sfrec ] ) with the initial conditions that @xmath89 and @xmath90 if @xmath32 or @xmath31 .
a similar argument will show that @xmath83 can be defined by the initial conditions that @xmath91 and @xmath92 if @xmath32 or @xmath93 and the recursion @xmath94_q \overline{\mathbf{sf}}_{n , k}(q).\ ] ] for @xmath88 .
similarly , @xmath78 can be defined by the initial conditions that @xmath95 and @xmath96 if @xmath32 or @xmath93 and the recursion @xmath97_q \mathbf{cf}_{n , k}(q),\ ] ] for @xmath88 , and @xmath79 can be defined by the initial conditions that @xmath98 and @xmath99 if @xmath32 or @xmath31 and the recursion @xmath100_q \overline{\mathbf{cf}}_{n , k}(q)\ ] ] for @xmath88 .
the main goal of this paper is to give a rook theory model for the polynomials @xmath78 , @xmath79 , @xmath82 , and @xmath83 .
our rook theory model will allow us to give combinatorial proofs of the defining equations ( [ cfdef ] ) , ( [ ovcfdef ] ) , ( [ sfdef ] ) , and ( [ ovsfdef ] ) as well as combinatorial proofs of the recursions ( [ sfrec ] ) , ( [ ovsfrec ] ) , ( [ cfrec ] ) , and ( [ ovcfrec ] ) .
we shall see that our rook theory model @xmath78 , @xmath79 , @xmath82 , and @xmath83 is essentially the same as the the rook theory model used in @xcite to interpret the @xmath58s and @xmath58s but with a different weighting scheme .
the outline of the paper is as follows . in section 2 , we describe a ranking and unranking theory for the set of fibonacci tilings which will a crucial element in our weighting scheme for our rook theory model that we shall use to give combinatorial interpretations of the polynomials @xmath78 , @xmath79 , @xmath82 , and @xmath83 . in section 3 , we shall review the rook theory model in @xcite and show how it can be modified for our purposes . in section 4 , we shall prove general product formulas for ferrers boards in our new model which will specialize ( [ cfdef ] ) , ( [ ovcfdef ] ) , ( [ sfdef ] ) , and ( [ ovsfdef ] ) in the case where the ferrers board is the staircase board whose column heights are @xmath101 reading from left to right . in section 5 , we shall prove various special properties of the polynomials @xmath78 , @xmath79 , @xmath82 , and @xmath83 .
there is a well developed theory for ranking and unranking combinatorial objects .
see for example , williamson s book @xcite . that is , give a collection of combinatorial objects @xmath102 of cardinality @xmath38 , one wants to define bijections @xmath103 and @xmath104 which are inverses of each other . in our case
, we let @xmath105 denote the set of fibonnaci tilings of height @xmath38
. then we construct a tree which we call the fibonacci tree for @xmath0 .
that is , we start from the top of a fibonacci tiling and branch left if we see a tile of height 1 and branch right if we see a tiling of height 2 .
for example , the fibonacci tree for @xmath106 is pictured in figure [ fig : ftree ] .
then for any tiling @xmath107 , we define the rank of @xmath44 for @xmath0 , @xmath108 , to be the number of paths to the left of the path for @xmath44 in the fibonacci tree for @xmath0 .
clearly @xmath109 so that @xmath110_q$ ] .
it is , in fact , quite easy to see compute the functions @xmath111 and @xmath112 in this situation .
that is , suppose that we represent the tiling @xmath44 as a sequence @xmath113 where reading the tiles starting at the bottom , @xmath114 if there is a tiling @xmath115 of height @xmath116 that ends at level @xmath117 in @xmath44 , @xmath118 if there is @xmath115 of height @xmath119 that ends at level @xmath117 in @xmath44 , and @xmath120 if there is no tile @xmath115 that ends at level @xmath117 in @xmath44 .
for example , the tiling of @xmath44 height 9 pictured in figure [ fig : f9 ] would be represented by the sequence @xmath121 . for any statement @xmath122 , we let @xmath123 is @xmath122 is true and @xmath124 if @xmath122 is false .
then we have the following lemma .
suppose that @xmath107 is a fibonacci tiling such that @xmath125
. then @xmath126 .
the theorem is easy to prove by induction .
it is clearly true for @xmath127 and @xmath128 .
now suppose @xmath129 .
then it is easy to see from the fibonacci tree for @xmath0 that if @xmath130 so that @xmath131 , then the tree that starts at level @xmath132 which represents taking the path to the left starting at level @xmath38 is just the fibonacci tree for @xmath133 and hence this tree will contain @xmath133 leaves which will all be to the left of path for the tiling @xmath44 .
then the tree that starting at level @xmath134 which represents taking the path to the right starting at level @xmath38 is just the fibonacci tree for @xmath135 and the number of paths in this tree which lie to the left of the path for @xmath44 is just that the number of paths to the left of the tiling @xmath136 such that @xmath137 in the fibonacci tree for @xmath135 .
thus in this case @xmath138 on the other hand if @xmath139 , then we branch left at level @xmath38 so that the number of paths to the left of the path for @xmath44 in the fibonacci tree for @xmath0 will just be the number of paths to the left of the tiling @xmath140 such that @xmath141 in the fibonacci tree for @xmath133 .
thus in this case @xmath142 for example , for the tiling @xmath44 in figure [ fig : f9 ] , @xmath143 . for the unrank function
, we must rely on zeckendorf s theorem @xcite which states that every positive integer @xmath38 is uniquely represented as sum @xmath144 where each @xmath145 and @xmath146 .
indeed , zeckendorf s theorem says that the greedy algorithm give us the proper representation .
that is , given @xmath38 , find @xmath36 such that @xmath147 , then the representation for @xmath38 is gotten by taking the representation for @xmath148 and adding @xmath149 .
for example , suppose that we want to find @xmath44 such that @xmath150
. then 1
. @xmath151 so that we need to find the fibonacci representation of @xmath152 .
2 . @xmath153 so that we need to find the fibonacci representation of @xmath154 .
3 . @xmath155 .
thus we can represent @xmath156 so that @xmath157
in this section , we shall give a rook theory model which will allow us to give combinatorial interpretations for the @xmath82s and the @xmath78s .
this rook theory model is based on the one which bach , paudyal , and remmel used in @xcite to give combinatorial interpretations to the @xmath58s and the @xmath69s .
thus , we shall briefly review the rook theory model in @xcite .
a ferrers board @xmath158 is a board whose column heights are @xmath159 , reading from left to right , such that @xmath160 .
we shall let @xmath161 denote the ferrers board @xmath162 .
for example , the ferrers board @xmath163 is pictured on the left of figure [ fig : ferrers ] and the ferrers board @xmath164 is pictured on the right of figure [ fig : ferrers ] classically , there are two type of rook placements that we consider on a ferrers board @xmath165 .
first we let @xmath166 be the set of all placements of @xmath36 rooks in @xmath165 such that no two rooks lie in the same row or column .
we shall call an element of @xmath166 a placement of @xmath36 non - attacking rooks in @xmath165 or just a rook placement for short .
we let @xmath167 be the set of all placements of @xmath36 rooks in @xmath165 such that no two rooks lie in the same column .
we shall call an element of @xmath167 a file placement of @xmath36 rooks in @xmath165 .
thus file placements differ from rook placements in that file placements allow two rooks to be in the same row . for example , we exhibit a placement of 3 non - attacking rooks in @xmath168 on the left in figure [ fig : placements ] and a file placement of 3 rooks on the right in figure [ fig : placements ] . given a ferrers board @xmath169 , we define the @xmath36-th rook number of @xmath165 to be @xmath170 and the @xmath36-th file number of @xmath165 to be @xmath171 .
then the rook theory interpretation of the classical stirling numbers is @xmath172 the idea of @xcite is to modify the sets @xmath166 and @xmath167 to replace rooks with fibonacci tilings .
the analogue of file placements is very straightforward .
that is , if @xmath158 , then we let @xmath173 denote the set of all configurations such that there are @xmath36 columns @xmath174 of @xmath165 where @xmath175 such that in each column @xmath176 , we have placed one of the tilings @xmath177 for the fibonacci number @xmath178
. we shall call such a configuration a fibonacci file placement and denote it by @xmath179 let @xmath180 denote the number of tiles of height 1 that appear in @xmath181 and @xmath182 denote the number of tiles of height 2 that appear in @xmath181 . then in @xcite , we defined the weight of @xmath181 , @xmath183 , to be @xmath184 .
for example , we have pictured an element @xmath181 of @xmath185 in figure [ fig : fibfile ] whose weight is @xmath186 .
then we defined the @xmath36-th @xmath52-fibonacci file polynomial of @xmath165 , @xmath187 , by setting @xmath188 if @xmath189 , then the only element of @xmath173 is the empty placement whose weight by definition is 1 .
then in @xcite , we proved the following theorem concerning fibonacci file placements in ferrers boards .
[ thm : ferrersfilerec ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
let @xmath193 .
then for all @xmath194 , @xmath195 to obtain the @xmath10-analogues that we desire for this paper , we define a new weight functions for elements of @xmath173 where @xmath158 is ferrers board .
that is given a fibonacci file placement @xmath196 in @xmath197 , let @xmath198 be the sequence of columns in @xmath165 which have no tilings , reading from left to right .
then we define @xmath199 note that the only difference between these two weight functions is that if @xmath200 is column that does not contain a tiling in @xmath181 , then it contributes a factor of @xmath201 to @xmath202 and a factor of 1 to @xmath203 .
we then define @xmath204 and @xmath205 , by setting @xmath206 if @xmath189 , then the only element of @xmath173 is the empty placement @xmath207 so that @xmath208 and @xmath209
. then we have the following analogue of theorem [ thm : ferrersfilerec ] . [ thm : qferrersfilerec ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
let @xmath193 .
then for all @xmath194 , @xmath210_q \mathbf{ft}_{k-1}(b^-,p , q)\ ] ] and @xmath211_q \mathbf{ft}_{k-1}(b^-,p , q).\ ] ] we claim ( [ qfibtilerec ] ) results by classifying the fibonacci file placements in @xmath173 according to whether there is a tiling in the last column .
if there is no tiling in the last column of @xmath181 , then removing the last column of @xmath181 produces an element of @xmath212 .
thus such placements contribute @xmath213 to @xmath214 since the fact that the last column has no tiling means that it contributes a factor of @xmath215 to @xmath202 .
if there is a tiling in the last column , then the fibonacci file placement that results by removing the last column is an element of @xmath216 and the sum of the weights of the possible fibonacci tilings of height @xmath217 for the last column is @xmath218_q$ ] .
hence such placements contribute @xmath219_q \mathbf{ft}_{k-1}(b^-,q)$ ] to @xmath214 .
thus @xmath220_q \mathbf{ft}_{k-1}(b^-,p , q).\ ] ] a similar argument will prove ( [ ovqfibtilerec ] ) .
if @xmath158 is a ferrers board , then we let @xmath221 denote the board that results by adding @xmath222 rows of length @xmath38 below @xmath165 .
we label these rows from top to bottom with the numbers @xmath223 .
we shall call the line that separates @xmath165 from these @xmath222 rows the _ bar_. a mixed file placement @xmath181 on the board @xmath221 consists of picking for each column @xmath200 either ( i ) a fibonacci tiling @xmath224 of height @xmath200 above the bar or ( ii ) picking a row @xmath225 below the bar to place a rook in the cell in row @xmath225 and column @xmath117 .
let @xmath226 denote set of all mixed rook placements on @xmath165 .
for any @xmath227 , we let @xmath180 denote the number of tiles of height 1 that appear in @xmath181 and @xmath182 denote the set tiles of height 2 that appear in @xmath181 . then in @xcite , we defined the weight of @xmath181 , @xmath183 , to be @xmath184 .
for example , figure [ fig : mixed ] pictures a mixed placement @xmath181 in @xmath221 where @xmath228 and @xmath222 is 9 such that @xmath229 . also in @xcite , we proved the following theorem by counting @xmath230 in two different ways .
[ thm : ferrersfileprod ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
@xmath231 to obtain the desired @xmath10-analogues for this paper , we must define new weight functions for mixed placements @xmath227 . that is
, suppose that @xmath232 is the fibonacci tile placement @xmath233 , and suppose that , for the rooks below the bar in columns @xmath234 , the rook in column @xmath235 is in row @xmath236 for @xmath237
. then we define @xmath238 that is , for each column @xmath117 the choice of a fibonacci tiling @xmath224 of height @xmath200 above the bar contributes a factor of @xmath239 to @xmath240 and the choice of picking a row @xmath225 below the bar to place a rook in the cell in row @xmath225 and column @xmath117 contributes a factor of @xmath241 to @xmath240 .
similarly , for each column @xmath200 the choice of a fibonacci tiling @xmath224 of height @xmath200 above the bar contributes a factor of @xmath239 to @xmath242 and the choice of picking a row @xmath225 below the bar to place a rook in the cell in row @xmath225 and column @xmath117 contributes a factor of @xmath243 to @xmath242 .
then we have the following analogue of theorem [ thm : ferrersfileprod ] . [ thm : qferrersfileprod ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
then for all positive integers @xmath222 , @xmath244_q [ x+f_{b_2}]_q \cdots [ x+f_{b_n}]_q = \sum_{k=0}^n \mathbf{ft}_k(b , q ) [ x]_q^{n - k}\ ] ] and @xmath245_q+[f_{b_1}]_q ) ( [ x]_q+[f_{b_2}]_q ) \cdots ( [ x]_q+[f_{b_n}]_q ) = \sum_{k=0}^n \overline{\mathbf{ft}}_k(b , q ) [ x]_q^{n - k}\ ] ] to prove ( [ qfibfileprod ] ) , fix @xmath222 to be a positive integer and consider the sums @xmath246 for @xmath247 , in a given column @xmath117 , our choice of the fibonacci tiling of height @xmath200 will contribute a factor of @xmath248_q$ ] to @xmath247 .
our choice of placing a rook below the bar in column @xmath117 contribute a factor of @xmath249_q\ ] ] to @xmath247 . as @xmath250_q + q^{f_{b_i}}[x]_q = [ x+f_{b_i}]_q$ ] , each column of @xmath200 of @xmath165 contributes a factor of @xmath251_q$ ] to @xmath247 so that @xmath252_q.\ ] ] for @xmath253 , in a given column @xmath117 , our choice of the fibonacci tiling of height @xmath200 will contribute a factor of @xmath248_q$ ] to @xmath247 .
our choice of placing a rook below the bar in column @xmath117 contribute a factor of @xmath254_q\ ] ] to @xmath253 .
thus each column @xmath200 contributes a factor of @xmath11_q+ [ f_{b_i}]_q$ ] to @xmath247 so that @xmath255_q + [ f_{b_i}]_q).\ ] ] on the other hand , suppose that we fix a fibonacci file placement @xmath256 .
then we want to compute @xmath257 which is the sum of @xmath258 over all mixed placements @xmath259 such that @xmath259 intersect @xmath165 equals @xmath181 .
it it easy to see that such a @xmath259 arises by choosing a rook to be placed below the bar for each column that does not contain a tiling .
each such column contributes a factor of @xmath260_q$ ] in addition to the weight @xmath202 .
thus it follows that @xmath261_q^{n - k}$ ] .
hence it follows that @xmath262_q^{n - k } \sum_{p \in \mathcal{ft}_k(b ) } \mathbf{w_{b , q}}(p ) \\ & = & \sum_{k=0}^n \mathbf{ft}_k(b , q ) \ [ x]_q^{n - k}.\end{aligned}\ ] ] the same argument will show that @xmath263_q^{n - k}.\ ] ] now consider the special case of the previous two theorems when @xmath264 .
then ( [ qfibtilerec ] ) implies that @xmath265_q \mathbf{ft}_{n - k}(b_n , q).\ ] ] it then easily follows that for all @xmath266 , @xmath267 note that @xmath268 for all @xmath269 since there are no fibonacci file placements in @xmath270 since there are only @xmath132 non - zero columns
. moreover such a situation , we see that ( [ cnkpqinterp ] ) implies that @xmath271_q [ x+f_1]_q [ x+f_2]_q \cdots [ x+f_{n-1}]_q = \sum_{k=1}^n \mathbf{cf}_{n , k}(q ) \ [ x]_q^k.\ ] ] thus we have given a combinatorial proof of ( [ cfdef ] ) .
similarly ( [ ovqfibtilerec ] ) implies that @xmath272_q \overline{\mathbf{ft}}_{n - k}(b_n , q).\ ] ] it then easily follows that for all @xmath266 , @xmath273 moreover such a situation , we see that ( [ snkpqinterp ] ) implies that @xmath271_q ( [ x]_q+[f_1]_q ) ( [ x]_q+[f_2]_q ) \cdots ( [ x]_q+[f_{n-1}]_q ) = \sum_{k=1}^n \overline{\mathbf{cf}}_{n , k}(q ) \ [ x]_q^k.\ ] ] thus we have given a combinatorial proof of ( [ ovcfdef ] ) .
the fibonacci analogue of rook placements defined in @xcite is a slight variation of fibonacci file placements .
the main difference is that each tiling will cancel some of the top most cells in each column to its right that has not been canceled by a tiling which is further to the left .
our goal is to ensure that if we start with a ferrers board @xmath190 , our cancellation scheme will ensure that the number of uncanceled cells in the empty columns are @xmath274 , reading from left to right .
that is , if @xmath158 , then we let @xmath275 denote the set of all configurations such that that there are @xmath36 columns @xmath174 of @xmath165 where @xmath175 such that the following conditions hold . * in column @xmath276 , we place a fibonacci tiling @xmath277 of height @xmath278 and for each @xmath279 , this tiling cancels the top @xmath280 cells at the top of column @xmath225 .
this cancellation has the effect of ensuring that the number of uncanceled cells in the columns without tilings at this point is @xmath281 , reading from left to right . * in column @xmath282 , our cancellation due to the tiling in column
@xmath283 ensures that there are @xmath284 uncanceled cells in column @xmath285 .
then we place a fibonacci tiling @xmath286 of height @xmath284 and for each @xmath287 , we cancel the top @xmath288 cells in column @xmath225 that has not been canceled by the tiling in column @xmath283 .
this cancellation has the effect of ensuring that the number of uncanceled cells in columns without tilings at this point is @xmath289 , reading from left to right . * in general ,
when we reach column @xmath290 , we assume that the cancellation due to the tilings in columns @xmath291 ensure that the number of uncanceled cells in the columns without tilings is @xmath292 , reading from left to right .
thus there will be @xmath293 uncanceled cells in column @xmath290 at this point .
then we place a fibonacci tiling @xmath294 of height @xmath295 and for each @xmath296 , this tiling will cancel the top @xmath297 cells in column @xmath225 that has not been canceled by the tilings in columns @xmath298 .
this cancellation has the effect of ensuring that the number of uncanceled cells in columns without tilings at this point is @xmath299 , reading from left to right .
we shall call such a configuration a fibonacci rook placement and denote it by @xmath179 let @xmath180 denote the number of tiles of height 1 that appear in @xmath181 and @xmath182 denote the number of tiles of height 2 that appear in @xmath181 . then in @xcite , we defined the weight of @xmath181 , @xmath183 , to be @xmath184 .
for example , on the left in figure [ fig : fibrook ] , we have pictured an element @xmath181 of @xmath300 whose weight is @xmath301 . in figure
[ fig : fibrook ] , we have indicated the canceled cells by the tiling in column @xmath117 by placing an @xmath117 in the cell .
we note in the special case where @xmath302 , then our cancellation scheme is quite simple .
that is , each tiling just cancels the top @xmath36 cells in each column to its right which has not been canceled by tilings to its left .
for example , on the right in figure [ fig : fibrook ] , we have pictured an element @xmath181 of @xmath303 whose weight is @xmath304 .
again , we have indicated the canceled cells by the tiling in column @xmath117 by placing an @xmath117 in the cell .
we define the @xmath36-th @xmath52-fibonacci rook polynomial of @xmath165 , @xmath305 , by setting @xmath306 if @xmath189 , then the only element of @xmath173 is the empty placement whose weight by definition is 1 . then in @xcite , we proved the following two theorems concerning fibonacci rook placements in ferrers boards .
[ thm : ferrersrookrec ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
let @xmath193 .
then for all @xmath194 , @xmath307 [ thm : ferrersrookprod ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 .
@xmath308 to obtain the @xmath10-analogues that we want for this paper , we need to define two new weight functions on fibonacci rook tilings .
that is , suppose that @xmath190 is a ferrers board and @xmath309 is an fibonacci rook tiling in @xmath275 .
then we know that the number of uncanceled cells in the @xmath310 columns which do not have tilings are @xmath311 reading from left to right .
suppose that the number of uncanceled cells in the columns with tilings are @xmath312 reading from left to right so that tiling @xmath177 is of height @xmath313 for @xmath314 .
the we define @xmath315 for example , if @xmath316 and @xmath317 is the rook tiling pictured in figure [ fig : fibrook ] , then @xmath318 , @xmath319 and @xmath320 and one can check that @xmath321 , @xmath322 , and @xmath323 .
thus @xmath324 and @xmath325 .
if @xmath189 , then the only element of @xmath173 is the empty placement @xmath207 which means that @xmath326 and @xmath327 .
then we define @xmath328 by setting @xmath329 and @xmath330 note that because of our cancellation scheme , there is a very simple relationship between @xmath328 and @xmath331 in the case where @xmath332 . that is , in any placement @xmath333
, the empty columns have @xmath274 uncanceled cells , reading from left to right , so that @xmath334 let @xmath158 be a ferrers board and @xmath222 be a positive integer .
then we let @xmath335 denote the board where we start with @xmath221 and add the flip of the board @xmath165 about its baseline below the board .
we shall call the the line that separates @xmath165 from these @xmath222 rows the _ upper bar _ and the line that separates the @xmath222 rows from the flip of @xmath165 added below the @xmath222 rows the _
lower bar_. we shall call the flipped version of @xmath165 added below @xmath221 the board @xmath336 .
for example , if @xmath337 , then the board @xmath338 is pictured in figure [ fig : aug ] .
the analogue of mixed placements in @xmath335 are more complex than the mixed placements for @xmath221 .
we process the columns from left to right . if we are in column 1 , then we can do one of the following three things .
* we can put a fibonacci tiling in cells in the first column in @xmath165 .
then we must cancel the top - most cells in each of the columns in @xmath165 to its right so that the number of uncanceled cells in the columns to its right are @xmath339 , respectively , as we read from left to right .
this means that we will cancel @xmath340 at the top of column @xmath117 in @xmath165 for @xmath341 .
we also cancel the same number of cells at the bottom of the corresponding columns of @xmath336 .
* we can place a rook in any row of column @xmath116 that lies between the upper bar and lower bar .
this rook will not cancel anything .
* we can put a flip of fibonacci tiling in column @xmath116 of @xmath336 .
this tiling will not cancel anything .
next assume that when we get to column @xmath225 , the number of uncanceled cells in the columns that have no tilings in @xmath165 and @xmath336 are @xmath342 for some @xmath36 as we read from left to right .
suppose there are @xmath200 uncanceled cells in @xmath165 in column @xmath225 .
then we can do one of three things .
* we can put a fibonacci tiling of height @xmath200 in the uncanceled cells in column @xmath225 in @xmath165 .
then we must cancel top - most cells of the columns in @xmath165 to its right so that the number of uncanceled cells in the columns which have no tilings up to this point are @xmath343 , we also cancel the same number of cells at the bottom of the corresponding columns of @xmath336 * we can place a rook in any row of column @xmath225 that lies between the upper bar and lower bar .
this rook will not cancel anything .
* we can put a flip of fibonacci tiling in the @xmath200 uncanceled cells in column @xmath225 of @xmath336 .
this tiling will not cancel anything we let @xmath344 denote set of all mixed rook placements on @xmath335 . for any placement @xmath345
, we define @xmath346 and @xmath347 as follows . for any column @xmath117 ,
suppose that the number of uncanceled cells in @xmath165 in column @xmath117 is @xmath115 .
then the factor @xmath348 that the placement in column @xmath117 contributes to @xmath346 is 1 .
@xmath349 if there is tiling @xmath224 in @xmath165 in column @xmath117 , 2 .
@xmath350 if there is a rook in row @xmath351 row from the top in the @xmath222 rows that lie between the upper bar and lower bar , and 3 .
@xmath352 if there is a flip of a tiling @xmath224 in column @xmath117 of @xmath336
. then we define @xmath353 similarly , the factor @xmath354 that the tile placement in column @xmath117 contributes to @xmath347 is 1 .
@xmath349 if there is tiling @xmath224 in @xmath165 in column @xmath117 , 2 .
@xmath355 if there is a rook in row @xmath351 row from the top in the @xmath222 rows that lie between the upper bar and lower bar , and 3 .
@xmath352 if there is a flip of a tiling @xmath224 in column @xmath117 of @xmath336
. then we define @xmath356 for example , figure [ fig : aug2 ] pictures a mixed placement @xmath181 in @xmath335 where @xmath228 and @xmath222 is 7 where @xmath321 , @xmath357 , and @xmath358 where @xmath224 is the tiling in column @xmath117 for @xmath359 .
the rooks columns 2 and 6 are in row 5 and the rook in column 3 is in row 3 so that @xmath360 and @xmath361 . thus @xmath362 our next theorem results from counting @xmath363 in two different ways .
[ thm : ovqferrersrookprod ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 and @xmath364
. then @xmath365_q^n = \sum_{k=0}^n \overline{\mathbf{rt}}_{n - k}(b , q ) ( [ x]_q-[f_{b_1}]_q ) ( [ x]_q-[f_{b_2}]_q)\cdots ( [ x]_q-[f_{b_k}]_q).\ ] ] fix @xmath222 to be a positive integer and consider the sum @xmath366 .
first we consider the contribution of each column as we proceed from left to right .
given our three choices in column 1 , the contribution of our choice of the tilings of height @xmath367 in column 1 of @xmath165 is @xmath368_q$ ] , the choice of placing a rook in between the upper bar and the lower is @xmath11_q$ ] , and the contribution of our choice of the tilings of height @xmath367 in column 1 of @xmath336 is @xmath369_q$ ] .
thus the contribution of our choices in column 1 to @xmath247 is @xmath368_q+[x]_q -[f_{b_1}]_q = [ x]_q$ ] . in general
, after we have processed our choices in the first @xmath225 columns , our cancellation scheme ensures that the number of uncanceled cells in @xmath165 and @xmath336 in the @xmath225-th column is @xmath200 for some @xmath370 .
thus given our three choices in column j , the contribution of our choice of the tilings of height @xmath200 in column @xmath225 of @xmath165 is @xmath250_q$ ] , the choice of placing a rook in between the upper bar and the lower is @xmath11_q$ ] , and the contribution of our choice of the tilings of height @xmath200 in column @xmath225 of @xmath336 is @xmath371_q$ ] .
thus the contribution of our choices in column @xmath225 to @xmath247 is @xmath250_q+[x]_q -[f_{b_i}]_q = [ x]_q$ ] .
it follows that @xmath372_q^n$ ] .
on the other hand , suppose that we fix a fibonacci rook placement @xmath373 .
then we want to compute the @xmath374 which is the sum of @xmath347 over all mixed placements @xmath259 such that @xmath259 intersect @xmath165 equals @xmath181 .
our cancellation scheme ensures that the number of uncanceled cells in @xmath165 and @xmath336 in the @xmath36 columns that do not contain tilings in @xmath181 is @xmath342 as we read from right to left . for each such @xmath375 , the factor that arises from either choosing a rook to be placed in between the upper bar and lower bar or a flipped fibonacci tiling of height @xmath200 in @xmath336
is @xmath11_q-[f_{b_i}]_q$ ] .
it follows that @xmath376_q-[f_{b_i}]_q.\ ] ] hence it follows that @xmath377_q-[f_{b_i}]_q\right ) \sum_{p \in \mathcal{nt}_k(b ) } \overline{\mathbf{w_{b , q}}}(p ) \\ & = & \sum_{k=0}^n \overline{\mathbf{rt}}_{n - k}(b , q ) \left ( \prod_{i=1}^k [ x]_q-[f_{b_i}]_q\right).\end{aligned}\ ] ] [ thm : qferrersrookprod ] let @xmath190 be a ferrers board where @xmath191 and @xmath192 and @xmath378
. then @xmath379_q^n = \sum_{k=0}^n \mathbf{rt}_{n - k}(b , q ) [ x - f_{b_1}]_q [ x - f_{b_2}]_q\cdots [ x - f_{b_k}]_q.\ ] ] it is easy to see from our cancellation scheme that @xmath380 thus it follows from ( [ ovqfibrookprod ] ) that @xmath271_q^n = \sum_{k=0}^n \mathbf{rt}_{n - k}(b , q)q^{-(f_{b_1}+ \cdots + f_{b_k } ) } ( [ x]_q-[f_{b_1}]_q ) ( [ x]_q-[f_{b_2}]_q)\cdots ( [ x]_q-[f_{b_k}]_q).\ ] ] however since @xmath381 for every @xmath117 , @xmath271_q-[f_{b_i}]_q = q^{f_{b_i}}[x - f_{b_i}]_q\ ] ] so that @xmath271_q^n = \sum_{k=0}^n \mathbf{rt}_{n - k}(b , q ) [ x - f_{b_1}]_q [ x - f_{b_2}]_q\cdots [ x - f_{b_k}]_q.\ ] ] now consider the special case of the previous three theorems when @xmath264 .
then ( [ qfibtilerec ] ) implies that @xmath382_q \mathbf{rt}_{n - k}(b_n , q).\ ] ] similarly ( [ ovqfibtilerec ] ) implies that @xmath383_q \overline{\mathbf{rt}}_{n - k}(b_n , q).\ ] ] it then easily follows that for all @xmath266 , @xmath384 and @xmath385 note that @xmath386 for all @xmath269 since there are no fibonacci rook placements in @xmath387 since there are only @xmath132 non - zero columns .
moreover such a situation , we see that ( [ qsfnkpqinterp ] ) implies that for @xmath388 , @xmath271_q^n = \sum_{k=1}^n \mathbf{sf}_{n , k}(q ) [ x]_q [ x - f_1]_q [ x - f_2]_q\cdots [ x - f_{k-1}]_q\ ] ] thus we have given a combinatorial proof of ( [ sfdef ] ) .
similarly , ( [ ovqsfnkpqinterp ] ) implies that for @xmath388 , @xmath271_q^n = \sum_{k=1}^n \overline{\mathbf{sf}}_{n , k}(q ) [ x]_q ( [ x]_q-[f_1]_q ) ( [ x]_q-[f_2]_q)\cdots ( [ x]_q-[f_{k-1}]_q)\ ] ] thus we have given a combinatorial proof of ( [ ovsfdef ] ) .
in this section , we shall derive various identities and special values for the fibonacci analogues of the stirling numbers @xmath82 , @xmath83 , @xmath78 , and @xmath79 .
1 . @xmath390 and @xmath391 .
2 . @xmath392_q$ ] and @xmath393_q$ ] .
3 . @xmath394_q(\sum_{j = i}^{n-2 } [ f_j]_q)$ ] and @xmath395_q(\sum_{j = i}^{n-2 } [ f_j]_q)$ ] .
4 . @xmath396 and @xmath397 .
5 . @xmath398 and @xmath399 . 6 .
@xmath400 and @xmath401 . for ( 2 )
, we can see that @xmath404_q$ ] because placements in @xmath405 have exactly one column which is filled with a fibonacci tiling .
if that column is column @xmath406 , then @xmath407 and the sum of the weights of the possible tilings in column @xmath117 is @xmath408_q$ ] .
the fact that @xmath393_q$ ] then follows from ( [ i1 ] ) . for ( 3 )
, we can classify the placements in @xmath409 by the left - most column which contains a tiling .
if that column is column @xmath406 , then @xmath407 and the sum of the weights of the possible tilings in column @xmath117 is @xmath410_q$ ] .
moreover , any tiling in column @xmath117 cancels one cell in the remaining columns so that number of uncanceled cells in the columns to the right of column @xmath406 will be @xmath411 , reading from right to left .
it then follows that @xmath412_q(\sum_{j = i}^{n-2 } [ f_j]_q).\ ] ] the fact that @xmath413_q(\sum_{j = i}^{n-2 } [ f_j]_q)\ ] ] then follows from ( [ i1 ] ) . for ( 4 ) , note that the elements in @xmath414 have a tiling in every column .
given our cancellation scheme , there is exactly one such configuration .
for example , the unique element of @xmath415 is pictured in figure [ fig : fullrooks ] where we have placed @xmath117s in the cells canceled by the tiling in column @xmath117 .
thus the unique element of @xmath414 is just the fibonacci rook placement where there is tiling of height one in each column .
thus @xmath416 since the rank of each tiling height 1 is 0 . for ( 5 ) , note that the elements in @xmath417 have exactly one column @xmath418 which does not have a tiling .
given our cancellation scheme , if the column with out a tiling is column @xmath418 , then any non - empty column to the left of column @xmath117 will be filled with a tiling of height @xmath116 and every column to the right of column @xmath117 will be filled with a tiling of height 2 . for example , the unique element of @xmath419 is pictured in figure [ fig:2fullrooks ] where we have placed @xmath117s in the cells canceled by the tiling in column @xmath117 . since the ranks of the tilings of heights 1 and 2 are 0 , it follows that @xmath420 .
the fact that @xmath421 then follows from ( [ i1 ] ) . for ( 6 )
, we proceed by induction .
note that we have proved @xmath422 now assume that @xmath129 and @xmath423 .
then @xmath424_q \mathbf{sf}_{n,3}(q)\\ & = & q\left(q(n-1 ) \right)+(1+q)\left((1+q)^{n-1}-((n-1)q+1)\right ) \\ & = & q^2(n-1 ) + ( 1-q)^n -(n-1)q-(n-1)q^2 -q -1\\ & = & ( 1-q)^n -(nq+1).\end{aligned}\ ] ] the fact that @xmath425 then follows from ( [ i1 ] ) .
next we define @xmath426 for @xmath7 it follows from theorem [ thm : id1 ] that @xmath427 then for @xmath428 , @xmath429_q \overline{\mathbf{sf}}_{n-1,k-}(p , q)\right ) t^{n-1 } \\ &
= & t^k + t\left(\sum_{n > k } \overline{\mathbf{sf}}_{n-1,k-1}(q ) t^{n-1}\right ) + [ f_k]_qt\left(\sum_{n > k } \overline{\mathbf{sf}}_{n-1,k}(q ) t^{n-1}\right ) \\ & = & t^k + t(\overline{\mathbb{sf}}_{k-1}(q , t ) -t^{k-1 } ) + [ f_k]_qt \overline{\mathbb{sf}}_k(q , t).\end{aligned}\ ] ] it follows that @xmath430_qt ) } \overline{\mathbb{sf}}_{k-1}(q , t).\ ] ] the following theorem easily follows from ( [ eq : id2 ] ) and ( [ eq : id3 ] ) . 1 . for all @xmath439 , @xmath440 .
2 . for all @xmath441 , @xmath442 .
3 . for all @xmath443 , @xmath444 .
4 . for all @xmath445 , @xmath446 . 5 . for all
@xmath447 , @xmath448 6 . for all @xmath447 , @xmath449 for ( 1 ) , note that a placement @xmath181 in @xmath450 must have @xmath451 empty columns among columns @xmath452 .
if @xmath453 , then it must be the case that all the tilings in the columns which contain tilings in @xmath181 must have rank 0 so that the tiling must contain only tiles of height 1 .
thus @xmath181 is completely determined by the choice of the @xmath451 empty columns among columns @xmath452 .
thus @xmath454 .
next fix @xmath462 and assume by induction that @xmath463 for all @xmath464
. then we shall prove by induction on @xmath38 that @xmath465 .
the base case @xmath466 holds since @xmath467 .
but then assuming that @xmath465 , we see that @xmath468 for example , by ( 3 ) , @xmath469 so that our formula holds for @xmath470 .
now suppose that @xmath462 and our formula holds for @xmath451 .
that is , @xmath471 next observe that @xmath472 since @xmath473 so that our formula holds for @xmath466 .
note also that for @xmath462 , @xmath474 .
but then for @xmath447 , @xmath475_q \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^2 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^2}+\left((1+q+q^2 ) \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^2 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^2}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^0}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q}+\overline{\mathbf{sf}}_{n , k}(q)|_{q^2}\\ & = & ( k-4 ) \binom{n-1}{k-1 } + \binom{k-2}{2}\binom{n-1}{k } + \binom{n-1}{k-1 } + ( k-2)\binom{n-1}{k } \\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^2 } \\ & = & ( k-3 ) \binom{n-1}{k-1 } + \binom{k-1}{2}\binom{n-1}{k } + \overline{\mathbf{sf}}_{n , k}(q)|_{q^2}.\end{aligned}\ ] ] this gives us a recursion for @xmath476 in terms of @xmath477 which we can iterate to prove that @xmath478 @xmath480_q \overline{\mathbf{sf}}_{n,4}(q)\right)|_{q^3 } \\ & = & \overline{\mathbf{sf}}_{n,3}(q)|_{q^3}+\left((1+q+q^2 ) \overline{\mathbf{sf}}_{n,4}(q)\right)|_{q^3 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^3}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^2}+\overline{\mathbf{sf}}_{n , k}(q)|_{q^3}\\ & = & \binom{n-1}{5 } + 2\binom{n-1}{4 } + \left(\binom{n-1}{4}+3\binom{n-1}{5}\right ) + \overline{\mathbf{sf}}_{n , k}(q)|_{q^3 } \\
& = & 3\binom{n-1}{4 } + 4\binom{n-1}{5 } + \overline{\mathbf{sf}}_{n,4}(q)|_{q^3}.\end{aligned}\ ] ] next assume that @xmath485 .
first we note that @xmath486 since @xmath473 so that our formula holds for @xmath466 .
note also that for @xmath485 , @xmath487 .
now suppose our formula holds for @xmath451 .
that is , @xmath488 next observe that @xmath489 since @xmath473 so that our formula holds for @xmath466 .
note also that for @xmath462 , @xmath474 .
but then for @xmath490 , @xmath491_q \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^3 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^2}+\left((1+q+q^2+q^3 ) \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^3 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^3}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^0}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^2}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^3}\\ & = & ( k-5 ) \binom{n-1}{k-1 } + \left(\binom{k-2}{2}+\binom{k-3}{2}-1\right ) \binom{n-1}{k } + \binom{k-1}{3}\binom{n-1}{k+1}\\ & & + \binom{n-1}{k-1}+(k-2)\binom{n-1}{k}+ ( k-3)\binom{n-1}{k } + \binom{k-1}{2}\binom{n-1}{k+1 } \\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^3}\\ & = & ( k-4 ) \binom{n-1}{k-1 } + \left(\binom{k-1}{2}+\binom{k-2}{2}-1\right ) \binom{n-1}{k } + \binom{k}{3}\binom{n-1}{k+1 } \\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^3}. \\\end{aligned}\ ] ] this gives us a recursion for @xmath492 in terms of @xmath493 which we can interate to prove that @xmath494 @xmath495_q \overline{\mathbf{sf}}_{n,4}(q)\right)|_{q^4 } \\ &
= & \overline{\mathbf{sf}}_{n,3}(q)|_{q^4}+\left((1+q+q^2 ) \overline{\mathbf{sf}}_{n,4}(q)\right)|_{q^4 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^4}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^2}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^3}+\overline{\mathbf{sf}}_{n , k}(q)|_{q^4}\\ & = & \binom{n-1}{6 } + \binom{n-1}{4 } + 3\binom{n-1}{5 } + 3\binom{n-1}{5}+4\binom{n-1}{6 } + \\ & & \overline{\mathbf{sf}}_{n , k}(q)|_{q^4 } \\ & = & \binom{n-1}{4 } + 6\binom{n-1}{5 } + 5\binom{n-1}{6}+\overline{\mathbf{sf}}_{n,4}(q)|_{q^3}.\end{aligned}\ ] ] next assume that @xmath485 .
first we note that @xmath499 since @xmath473 so that our formula holds for @xmath466 .
note also that for @xmath485 , @xmath487 .
now suppose our formula holds for @xmath451 .
that is , @xmath500 next observe that @xmath501 since @xmath473 so that our formula holds for @xmath466 .
note also that for @xmath485 , @xmath487 .
but then for @xmath490 , @xmath502_q \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^4 } \\ & = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^4}+\left((1+q+q^2+q^3+q^4 ) \overline{\mathbf{sf}}_{n , k}(q)\right)|_{q^4 } \\
& = & \overline{\mathbf{sf}}_{n , k-1}(q)|_{q^3}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^0}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^2}+ \overline{\mathbf{sf}}_{n , k}(q)|_{q^3}\\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^4}\\ & = & ( k-5 ) \binom{n-1}{k-1 } + \left(\binom{k-2}{2}+\binom{k-3}{2}+\binom{k-3}{2}-3\right ) \binom{n-1}{k } + \\ & & + \left(\binom{k-1}{3}+\binom{k-2}{3}-(k-1)+1\right)\binom{n-1}{k+1 } + \binom{k}{4}\binom{n-1}{k+2}\\ & & + \binom{n-1}{k-1}+(k-2)\binom{n-1}{k}+ ( k-3)\binom{n-1}{k } + \binom{k-1}{2 } \binom{n-1}{k+1 } \\ & & + ( k-4 ) \binom{n-1}{k } + \left(\binom{k-1}{2}+\binom{k-2}{2}-1\right ) \binom{n-1}{k+1 } + \binom{k}{3 } \binom{n-1}{k+2}\\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^4}\\ & = & ( k-4)\binom{n-1}{k-1 } + \left ( \binom{k-1}{2}+\binom{k-2}{2}+\binom{k-3}{2}-3 \right ) \binom{n-1}{k}\\ & & + \left(\binom{k}{3}+\binom{k-1}{3}-k+1\right ) \binom{n-1}{k+1 } + \binom{k+1}{4}\binom{n-1}{k+2 } \\ & & + \overline{\mathbf{sf}}_{n , k}(q)|_{q^4}. \end{aligned}\ ] ] this gives us a recursion for @xmath503 in terms of @xmath504 which we can iterate to prove that @xmath505 a sequence of real numbers @xmath506 is is said to be _ unimodal _ if there is a @xmath507 such that @xmath508 and is said to be _ log - concave _ if for @xmath509 , @xmath510 where we set @xmath511 .
if a sequence is log - concave , then it is unimodal .
a polynomial @xmath512 is said to be _ unimodal _ if @xmath506 is a unimodal sequence and is said to be log - concave if @xmath506 is log concave .
it is easy to see from theorem [ thm : id1 ] that @xmath83 is unimodal for all @xmath513 when @xmath514 .
computational evidence suggests that @xmath515 is unimodal for all @xmath516 and that @xmath517 is unimodal for all @xmath518 .
however , it is not the case that @xmath519 is unimodal for all @xmath520 .
for example , one can use part 3 of theorem [ thm : id1 ] to compute @xmath521 it is not difficult to see that for any ferrers board @xmath522 , the coefficients that appear in the polynomials @xmath214 and @xmath205 are essentially the same .
that is , we have the following theorem .
let @xmath158 be a skyline board
. then @xmath523_qz)\right)|_{z^k}\ ] ] and @xmath524_qz)\right)|_{z^k } = q^{\sum_{i=1}^n f_{b_i}}\left ( \prod_{i=1}^n ( 1+\frac{1}{q}[f_{b_i}]_{\frac{1}{q}}z)\right)|_{z^k}\ ] ] it is easy to see that if we are creating a fibonacci file tiling in @xmath173 , then in column @xmath117 , we have two choices , namely , we can leave the column empty or put a fibonacci tiling of height @xmath200 . for @xmath205 ,
the weight of an empty column is 1 and the sum of weights of the fibonacci tilings of height @xmath200 is @xmath250_q$ ] .
thus @xmath525_qz)\right)|_{z^k}$ ] is equal to the sum over all fibonacci file tilings where exactly @xmath36 columns have tiling which is equal to @xmath205 .
similarly , for @xmath204 , the weight of an empty column @xmath117 when it is empty is @xmath201 and the sum of weights of the fibonacci tilings of height @xmath200 is @xmath250_q$ ] .
thus @xmath526_qz)\right)|_{z^k}$ ] is equal to the sum over all fibonacci file tilings where exactly @xmath36 columns have tiling which is equal to @xmath204 .
it follow that for any @xmath38 , the coefficient of @xmath527 in @xmath525_qz)\right)|_{z^k}$ ] is equal to the coefficient of @xmath528 in @xmath529_{\frac{1}{q}}z)\right)|_{z^k}$ ] .
it follows that @xmath530 it is easy to see from ( [ ftone ] ) that @xmath531_q\ ] ] so that coefficient of @xmath532 in @xmath533 weakly decreases as @xmath36 goes from 0 to @xmath534 .
it follows that the coefficient of @xmath532 in @xmath535 weakly increase .
similarly , it is easy to see that @xmath536_q\ ] ] so that @xmath537 is just the rank generating function of a product of chains which is know to be symmetric and unimodal , see @xcite . from our computational evidence
, it seems that the polynomials @xmath538 are unimodal . however , it is not the case @xmath79 are unimodal for all @xmath38 and @xmath36 .
for example , @xmath539 starts out @xmath540 finally , our results show that the matrices @xmath541 and @xmath542 are inverses of each other
. one can give a combinatorial proof of this fact .
indeed , the combinatorial proof of @xcite which shows that matrices @xmath543 and @xmath544 are inverses of each other can also be applied to show that the matrices @xmath541 and @xmath542 are inverses of each other . | let @xmath0 denote the @xmath1 fibonacci number relative to the initial conditions @xmath2 and @xmath3 . in @xcite , we introduced fibonacci analogues of the stirling numbers called fibo - stirling numbers of the first and second kind .
these numbers serve as the connection coefficients between the fibo - falling factorial basis @xmath4 and the fibo - rising factorial basis @xmath5 which are defined by @xmath6 and for @xmath7 , @xmath8 and @xmath9 .
we gave a general rook theory model which allowed us to give combinatorial interpretations of the fibo - stirling numbers of the first and second kind .
there are two natural @xmath10-analogues of the falling and rising fibo - factorial basis .
that is , let @xmath11_q = \frac{q^x-1}{q-1}$ ] .
then we let @xmath11_{\downarrow_{q , f,0 } } = \overline{[x]}_{\downarrow_{q , f,0 } } = [ x]_{\uparrow_{q , f,0 } } = \overline{[x]}_{\uparrow_{q , f,0}}=1 $ ] and , for @xmath12 , we let + @xmath11_{\downarrow_{q , f , k } } = [ x]_q [ x - f_1]_q \cdots [ x - f_{k-1}]_q$ ] , @xmath13}_{\downarrow_{q , f , k}}= [ x]_q ( [ x]_q-[f_1]_q ) \cdots ( [ x]_q-[f_{k-1}]_q)$ ] , + @xmath11_{\uparrow_{q , f , k}}= [ x]_q [ x+f_1]_q \cdots [ x+f_{k-1}]_q$ ] , and @xmath13}_{\uparrow_{q , f , k}}= [ x]_q ( [ x]_q+[f_1]_q ) \cdots ( [ x]_q+[f_{k-1}]_q)$ ] . in this paper , we show we can modify the rook theory model of @xcite to give combinatorial interpretations for the two different types @xmath10-analogues of the fibo - stirling numbers which arise as the connection coefficients between the two different @xmath10-analogues of the fibonacci falling and rising factorial bases . |
lattice density functional theory is drawing increasing attention on account of its wide range of applicability to phenomena of strong interest in current research.@xcite its applications include ordering phenomena in metallic alloys , submonolayer adsorbate systems,@xcite colloid - polymers mixtures , @xcite fluids in porous media , @xcite and dna denaturation.@xcite in certain cases , exact density functionals in zero and one dimension can be written in a so - called fundamental measure form , which allows one to construct approximate functionals in higher dimensions that reduce to the exact ones upon dimensional crossover .
@xcite moreover , the theory can be extended to time - dependent phenomena .
@xcite particles with shapes that interact solely via hard - core repulsion on a lattice or in a continuum do produce interesting effects including phase transitions , e.g. for hard hexagons.@xcite however , the inclusion of attractive or repulsive forces on contact or at some distance is important for more realistic modelling.@xcite models with square - well interaction potentials including potentials of the zero - range , sticky - core type have proven to exhibit realistic physical features for many systems : colloidal suspensions,@xcite crystallization of polymers , @xcite micelles , @xcite protein solutions , @xcite dna coated colloids @xcite ionic fluids@xcite , and microemulsions.@xcite it is well established that many effects of generic short - range interactions can be reproduced by contact forces of a strength that yields matching second virial coefficients.@xcite in this paper we develop ideas found in the work of buschle _ et al . _ @xcite to derive an exact free - energy functional for hard rods with first - neighbor coupling of arbitrary shape and limited range . onto a one - dimensional lattice of @xmath0 sites we place non - overlapping rods of one size @xmath1 ( in unit of the lattice spacing ) as illustrated in fig .
[ fig : fig1 ] . to each lattice site
@xmath2 we assign occupation number @xmath3 if it is the location of the left end of a rod and @xmath4 otherwise .
the hard - core exclusion condition implies that @xmath5 for @xmath6 .
hard walls at @xmath7 imply that @xmath4 for @xmath8 and @xmath9 .
hard rods of size @xmath10 interacting with a potential of range @xmath11 .
the rods on the left are in contact .
the interaction between the two rods is nonzero for @xmath12.,scaledwidth=40.0% ] the two terms in the hamiltonian , @xmath13 represent an interaction @xmath14 between successive rods and an external potential @xmath15 .
we write @xmath16 for microstates
( for easy notation , we include the @xmath4 for @xmath17 ) .
the range @xmath18 of the interaction is assumed to be shorter than two rod lengths .
hence we have @xmath19 for @xmath20 and @xmath21 for @xmath22 , where @xmath23 .
we present explicit results for rods subject to contact forces @xmath24 , in which case our model is a lattice version of the sticky - core continuum model analyzed by baxter.@xcite . for the case
@xmath25 our model reduces to the familiar ising lattice gas .
the chemical potential , which controls the average number of rods in a grandcanonical ensemble , is conveniently absorbed as a constant in the external potential . in the following
we derive exact joint probability distributions for the positions of the rods on the lattice as functionals of their density profile , from which we infer exact expressions for the two - point functions and density functionals .
contact is made with lattice fundamental measure theory .
we then present exact and explicit results for the thermodynamics pertaining to models with attractive and repulsive contact potentials . in connection with the markov property
underlying the markov chain approach , it is interesting to note that it corresponds to kirkwood s `` superposition principle '' in the integral equation method for evaluating distribution functions in continuum fluids.@xcite for one - dimensional continuum fluids with nearest - neighbor interactions , analytical expressions for many - particle densities were derived by using laplace transform techniques and it was shown that the markov property ( or `` superposition principle '' ) then becomes exact.@xcite our approach for lattice systems here starts with the markov property and by this we are able to derive microstate distributions as functionals of the density profile .
for given interaction @xmath26 , the external potential @xmath27n_i$ ] , which yields the set @xmath28 of mean occupation numbers @xmath29 ( `` density profile '' ) in equilibrium , is a ( unique ) functional of @xmath30.@xcite our goal in this section is to derive the corresponding joint probability @xmath31 as functional of @xmath30 , where @xmath32n_i$ ] , @xmath33=u_i[\mathbf{p}]-\mu$ ] .
all energies here and in the following are given in units of @xmath34 .
our derivation , inspired by the method developed in ref . , proceeds in three steps . in the first step ,
we use a markov property to express @xmath35 in terms of marginal probabilities @xmath36 for sites within range of the interactions . in the second step the latter
are expressed in terms of the @xmath37 and the correlators @xmath38 by making use of the exclusion constraint associated with the hardcore repulsion . in the last step ,
the @xmath39 $ ] are given as functional of @xmath30 by comparing @xmath40 with the boltzmann expression for a few simple configurations @xmath41 .
the derivation of @xmath40 constitutes a rare case , where the `` mermin potential '' @xmath42-\langle v(\mathbf{n})\rangle-\ln\chi_\mathbf{p}(\mathbf{n})$ ] can be stated explicitly .
we begin by writing @xmath35 as a product of conditional probabilities , @xmath43 here @xmath44 is the probability of finding the occupation number @xmath45 at site @xmath46 under the condition that all occupation numbers left to site @xmath46 are given .
because of the finite interaction range and because we are dealing with a one - dimensional system , one can prove the following markov property , @xmath47 where @xmath48 denotes the joint probability of finding the set @xmath49 of occupation numbers .
it is useful to subdivide the range @xmath50 of sites into two compartments of labels , @xmath51 and @xmath52 .
the hardcore exclusion dictates that each compartment contains at most one label of a nonzero occupation number .
if both compartments are occupied the two labels must satisfy @xmath53 .
let us introduce the abbreviated notation , @xmath54 , @xmath55 , @xmath56 , and @xmath57 , for the remaining options of compartmental occupancy .
the first and second argument refer to the first range and second range , respectively .
a zero in one argument means that all occupation numbers in the corresponding range are zero and @xmath58 in an argument means that @xmath59 in the corresponding range .
for example , @xmath60 .
we thus arrive at the following representation for the marginal probabilities : @xmath61 where @xmath62 and @xmath63 refer to indices running over the first and second range , and the specification @xmath64 means that the two indices run over the set @xmath65 , @xmath66 .
equation ( [ eq : phi - decomp ] ) amounts to expressing the joint probability in terms of the @xmath67 , where case selections from all possible configurations are encoded in the exponents .
for example , for the configuration @xmath68 , the associated exponent @xmath69 is one for @xmath3 and all @xmath70 in the second range , while otherwise it is zero and thus giving an ( irrelevant ) factor one in eq .
( [ eq : phi - decomp ] ) .
note that , different from a case selection by products in the exponents , e.g.by @xmath71 for the configuration @xmath68 , we have written sums .
this is allowed because at most one occupation number can be one in the two ranges and @xmath72 .
the function @xmath73 in the denominator of eq .
( [ eq : markov ] ) has a representation identical in structure but now the range of indices indicated by @xmath62 refers to @xmath74 and the range of indices indicated by @xmath64 refers to @xmath75 , @xmath66 .
the range indicated by @xmath63 remains unchanged .
the corresponding shortened notation is indicated by a tilde , i.e. @xmath76 , @xmath77 etc .
for example , @xmath78 .
we continue by relating the functions @xmath79 to the occupancies @xmath29 and correlators @xmath80 . for this
we only need basic properties and the normalization condition : [ eq : pc ] @xmath81 where @xmath82 . from these relations
we infer [ eq : phi12 ] @xmath83 the corresponding expressions for the functionals @xmath84 have the @xmath46 as upper limit of sums replaced by @xmath85 .
hence we have reduced the joint probability @xmath35 to a functional of densities @xmath37 and correlators @xmath86 .
what remains to be accomplished is to reduce the correlators to functionals of the densities . to this end
we compare @xmath35 calculated from above with the boltzmann probability for configurations with zero , one and two rods , [ eq : chi-2 ] @xmath87 where @xmath88 , and @xmath89 in eq .
( [ eq : chi-2-c ] ) .
these probabilites thus satisfy the relation @xmath90 the task ahead is cumbersome but manageable : express all four joint probabilities of ( [ eq : chi - rel ] ) in terms of the marginal probabilities @xmath79 and @xmath84 by using eqs .
( [ eq : chi-1 ] ) , ( [ eq : markov ] ) , and ( [ eq : phi - decomp ] ) .
then substitute relations ( [ eq : phi12 ] ) in order to extract the desired functional dependence of the correlators @xmath86 on the densities @xmath37 .
for example , for a configuration with two rods at site @xmath2 and @xmath91 with @xmath92 we have @xmath93 \frac{\phi_j(1_j,0)}{\tilde\phi_j(0,0 ) } % & { } \times \left[\prod_{s = j+1}^{i-1}\frac{\phi_s(1_j,0)}{\tilde\phi_s(1_j,0)}\right ] \frac{\phi_i(1_i,1_j)}{\tilde\phi_j(0_i,1_j)}\nonumber\\ & \hspace{1em}{}\times \left[\prod_{s = i+1}^{j+\xi}\frac{\phi_s(1_i,1_j)}{\tilde\phi_s(1_i,1_j)}\right ] \left[\prod_{s = j+\xi+1}^{i+\sigma-1}\frac{\phi_s(1_i,0)}{\tilde\phi_s(1_i,0 ) } \right]\nonumber\\ & \hspace{1em}{}\times \left[\prod_{s = i+\sigma}^{i+\xi}\frac{\phi_s(0,1_i)}{\tilde\phi_s(0,1_i ) } \right ] \left[\prod_{s = i+l+1}^{l}\frac{\phi_s(0,0)}{\tilde\phi_s(0,0)}\right]\,.\end{aligned}\ ] ] the different terms in this equation are a consequence of the markov chain in eq .
( [ eq : markov ] ) with progressing @xmath46 index : the first four factors arise from successively capturing the rods located at sites @xmath91 and @xmath2 .
the next four terms are associated with the following specific @xmath46-values : if @xmath94 , the rod at site @xmath91 falls out of the interaction range , and if @xmath95 , the rod at site @xmath2 is no longer in the first range ( see above ) with respect to site @xmath46 .
if @xmath96 , the rod at site @xmath2 eventually falls out of the interaction range . decomposing the other joint probabilities in eq
( [ eq : chi-2 ] ) in an analogous way and inserting the @xmath97 , @xmath98 from eq .
( [ eq : phi12 ] ) , yields , after elementary algebra , @xmath99e^{-v_{i , j}}\ ] ] for @xmath100 . with the @xmath86 determined from eqs .
( [ eq : c - p - rel ] ) as functional of @xmath30 , the distribution of microstates becomes also a functional of @xmath30 using eqs .
( [ eq : phi12 ] ) , ( [ eq : phi - decomp ] ) , ( [ eq : markov ] ) , and ( [ eq : chi-1 ] ) , i.e.@xmath101 . for general interactions @xmath14 , an analytic solution eqs .
( [ eq : c - p - rel ] ) appears out of reach and we must resort to a numerical evaluation . in the remainder of this section we focus on a system with contact interactions @xmath102 . in this case
( [ eq : c - p - rel ] ) simplify into @xmath103[p_{i-\sigma } -c_{i-\sigma , i}]}{[1-\sum_{k = i-\sigma}^i p_k + c_{i-\sigma , i}]}e^{-v_c},\ ] ] with physically relevant solutions , @xmath104^{1/2 } } { 2(e^{-v_c}-1)}\,,\ ] ] where @xmath105 figure [ fig : prob - contact ] shows , as an example , the contact correlators @xmath106 for a spatially homogeneous bulk system ( @xmath107 ) with mean occupation numbers @xmath108 . for all graphs shown in the following
we use the coverage @xmath109 as independent variable , which can be interpreted as mass density of sorts .
we consider rods of sizes @xmath110 and @xmath111 and contact interactions of zero , finite , and infinite strength ( attractive and repulsive ) .
the solid curve in each panel represents the result for the non - interacting case , where @xmath112 for @xmath110 , because correlations are absent in the simple fermi lattice gas . for larger rod size
, we have @xmath113 for @xmath114 .
this is a consequence of the well - known entropy effect in systems with athermal exclusion interactions : bringing neighboring rods closer to each other gives the remaining rods more configurational freedom and in total the system more configurational space .
repulsive interactions @xmath115 cause dispersal of rods , which suppresses contact correlations .
attractive interactions @xmath116 , on the other hand , lead to clustering of rods , which enhances contact correlations .
infinitely strong attraction produces a single cluster that grows with @xmath117 , whereas infinitely strong repulsion makes the rods avoid all contacts if possible , i.e. for @xmath118 .
these attributes account for the ( piecewise ) linear dependence of @xmath119 on @xmath117 .
contact correlators @xmath120 for spatially homogeneous systems of hard rods of sizes @xmath110 and 5 , and different strength of repulsive and attractive contact interactions @xmath121.,title="fig : " ] contact correlators @xmath120 for spatially homogeneous systems of hard rods of sizes @xmath110 and 5 , and different strength of repulsive and attractive contact interactions @xmath121.,title="fig : " ] + contact correlators @xmath120 for spatially homogeneous systems of hard rods of sizes @xmath110 and 5 , and different strength of repulsive and attractive contact interactions @xmath121.,title="fig : " ] contact correlators @xmath120 for spatially homogeneous systems of hard rods of sizes @xmath110 and 5 , and different strength of repulsive and attractive contact interactions @xmath121.,title="fig : " ]
based on the gibbs - bogoliubov inequality the following functional is defined in density functional theory , @xmath122&=\sum_{\mathbf{n}}\chi_{\mathbf{p}}(\mathbf{n } ) \left[\ln \chi_{\mathbf{p}}(\mathbf{n})+v(\mathbf{n})+u(\mathbf{n})- \mu n\right]\nonumber\\ & = f[\mathbf{p}]+ \sum_{k=1}^l ( u_k-\mu)p_k\ , , \label{eq : omega}\end{aligned}\ ] ] where @xmath123=\sum_{\mathbf{n } } \chi_{\mathbf{p}}(\mathbf{n})[\ln \chi_{\mathbf{p}}(\mathbf{n})+v(\mathbf{n})]$ ] is the free energy functional and the @xmath124 is the external in eq .
( [ eq : hamiltonian ] ) . minimizing @xmath125 $ ] yields the equilibrium density profile @xmath126 . substituting the results of sec .
[ sec : microstates ] and using the abbreviated notation , @xmath127 , we can write the free - energy functional in the form @xmath128&= \sum_{i , j } v_{i , j}c_{i , j}\nonumber\\ & { } + \sum_{s=1}^{l}\biggl\ { \phi\bigl(p_s\!-\!\sum_{i = s-\xi}^{s-\sigma}c_{s , i}\bigr)\nonumber\\ & \phantom{{}+\sum_{s=1}^{l}\bigl\ { } { } + \phi\bigl(1\!-\!\sum_{i = s-\xi}^sp_i \!+\!\sum_{i = s-\xi+\sigma}^s\sum_{j = s-\xi}^{i-\sigma}c_{i , j}\bigr)\nonumber\\ & \phantom{{}+\sum_{s=1}^{l}\bigl\ { } { } -\phi\bigl(1\!-\!\sum_{i = s-\xi}^{s-1}p_i + \sum_{i = s-\xi+\sigma}^{s-1}\sum_{j = s-\xi}^{i-\sigma-1}c_{i , j}\bigr)\nonumber\\ & \phantom{{}+\sum_{s=1}^{l}\bigl\ { } { } + \sum_{i = s-\xi}^{s-\sigma}\bigl\ { \phi\bigl(c_{s , i}\bigr)+\phi\bigl(p_{i}-\sum_{j = i+\sigma}^{s}c_{j , i}\bigr)\nonumber\\ & \phantom{{}+\sum_{s=1}^{l}\bigl\{{}+\sum_{i = s-\xi}^{s-\sigma}\bigl\ { } { } -\phi\bigl(p_{i}-\sum_{j = i+\sigma}^{s-1}c_{j , i}\bigr)\bigr\}\biggr\}\ , , \label{eq : f}\end{aligned}\ ] ] with the @xmath39 $ ] extracted from eq .
( [ eq : c - p - rel ] ) . for contact interactions , @xmath102 ,
it can be rendered more compactly : @xmath129 & = \sum\limits_s \bigl\ { c_{s-\sigma , s}v_{s-\sigma , s } + \phi\bigl(p_s - c_{s-\sigma , s}\bigr)\\ & { } + \phi\bigl(1-\sum_{k = s-\sigma}^s p_k + c_{s-\sigma , s}\bigr ) -\phi\bigl(1-\sum_{k = s-\sigma}^{s-1 } p_k\bigr ) \nonumber\\ & { } + \phi\bigl(c_{s-\sigma , s}\bigr ) + \phi\bigl(p_{s-\sigma}-c_{s-\sigma , s}\bigr)- \phi\bigl(p_{s-\sigma}\bigr)\bigl\ } \nonumber\end{aligned}\ ] ] for the special case @xmath110 ( ising lattice gas ) , this functional agrees with a previous result of ref . , for which lafuente and cuesta @xcite derived a fundamental measure form also .
as it happens , the fundamental measure form can be extended to hard rods ( @xmath130 ) with contact interaction by eliminating @xmath131 in eq .
( [ eq : f - contact-1 ] ) in favor of correlators and densities via eq .
( [ eq : c - p - rel ] ) : @xmath132&=\sum_{s=1}^l\left(\mathcal{f}_{2}[p_{s-\sigma},\ldots , p_s ] -\mathcal{f}_1[p_{s-\sigma},\ldots , p_{s-1}]\right)\end{aligned}\ ] ] where @xmath133=\nonumber\\ & \hspace{2em}p_{s-\sigma}\ln p_{s-\sigma}+(1-\sum_{i = s-\sigma}^{s-1}p_i ) \ln(1 -\sum_{i = s-\sigma}^{s-1}p_i)\end{aligned}\ ] ] and @xmath134=\nonumber\\ & \hspace{3em } p_s\ln(p_s - c_{s-\sigma , s})+p_{s-\sigma}\ln(p_{s-\sigma}-c_{s-\sigma , s})\nonumber\\ & \hspace{3em } { } + ( 1-\sum_{i = s-\sigma}^sp_i)\ln(1-\sum_{i = s-\sigma}^sp_i + c_{s-\sigma , s})\end{aligned}\ ] ] are the free energy functionals of one - particle and two - particle cavities , respectively
. a one - particle cavity refers to a range of successive lattice sites , where at most one occupation number can be one , in analogy to the zero - dimensional cavity in rosenfeld s fundamental measure theory in continuous space . for discrete lattice gas systems , following lafuente and cuesta
, @xcite an @xmath135-particle cavity refers to a range of successive lattice sites , where at most @xmath135 occupation numbers can be one .
notice that the size of an @xmath135-particle cavity can vary between @xmath136 ( minimal @xmath135-particle cavity ) and @xmath137 ( maximal @xmath135-particle cavity ) . in this respect ,
@xmath138 in eq .
( [ eq : f1 ] ) refers to a maximal one - particle cavity and @xmath139 in eq .
( [ eq : f2 ] ) to a minimal two - particle cavity .
fundamental measure forms allow for a straighforward extension to approximate functionals in higher dimensions that become exact under dimensional reduction.@xcite
here the focus is entirely on bulk thermodynamic properties of homogeneous systems with attractive or repulsive contact interactions .
the internal energy per site @xmath140 is directly related to the contact correlator @xmath119 from eq .
( [ eq : cij - contact ] ) : @xmath141 its dependence on coverage @xmath117 for @xmath142 is represented by the curves in fig . [ fig : prob - contact ] , appropriately rescaled by @xmath143 ( positive or negative ) .
not surprisingly , the magnitude of @xmath144 increases with crowding .
we infer the free energy per site @xmath145 from eq .
( [ eq : f - contact-1 ] ) : @xmath146 the entropy per site @xmath147 follows directly : @xmath148 their dependences on @xmath117 are shown in figs .
[ fig : entropy ] and [ fig : free_energy ] .
entropy @xmath46 per lattice site as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] entropy @xmath46 per lattice site as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] + entropy @xmath46 per lattice site as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] entropy @xmath46 per lattice site as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] attractive and repulsive contact interactions both lead to an entropy reduction .
the underlying causes are different . for @xmath149
the rods have a tendency to form clusters .
this ordering tendency is largely independent of coverage , producing a relative entropy reduction that depends only weakly on @xmath117 . in the limit @xmath150 , the rods form a single cluster , implying @xmath151 for any @xmath117 . for @xmath152 ,
by contrast , the rods have a tendency to disperse , i.e.to avoid contact . the associated ordering tendency in the face of space constraints
strongly depends on coverage . above a critical strength @xmath153 of the contact interaction ,
the entropy develops a double - hump structure with a minimum at @xmath154 for @xmath110 , and close to @xmath155 for @xmath130 . at the coverage @xmath156 , perfect ordering , corresponding to a vanishing entropy ,
is obtained in the limit @xmath157 ( @xmath158 ) . for lower or higher coverages , residual entropy @xmath46 persists at significant levels .
entropy profiles similar to those in fig .
[ fig : entropy ] were recently found by a study of jammed granular matter in a narrow channel.@xcite a point of some interest is the location of the entropy maximum in dependence of @xmath1 and @xmath143 ( first maximum for @xmath159 in the case of repulsive interactions ) .
the curves show that for @xmath160 the coverage of maximum entropy shifts to right from @xmath154 as the size of the rods grows from @xmath110 and that , for given @xmath130 , it shift to left as @xmath161 increases . this shift can be understood intuitively by viewing rods and vacancies as two species with numbers @xmath162 and @xmath163 , respectively . neglecting entropy related correlation effects in the non - interacting case , the number @xmath164 of possible configurations
could then be estimated by @xmath165 . for @xmath110 ,
this is clearly maximal for equal number of particles and vacancies @xmath166 , while for @xmath167 , one has to take into account that @xmath168 decreases with increasing @xmath117 .
accordingly , the location of the entropy maximum occurs left to @xmath156 , where @xmath169 . in the interacting case , the coverage of maximum entropy moves to the left for increasing interaction strength , because the enhanced configurational restrictions for larger @xmath161 can be partly compensated by lowering @xmath117 .
analytical results for the entropy of hard rods with nearest - neighbor interactions in one - dimensional continuum have been earlier obtained by percus.@xcite its overall behavior as function of the density in the case of contact interaction agrees with the one displayed in fig .
[ fig : entropy ] .
free energy per site versus @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] free energy per site versus @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] + free energy per site versus @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] free energy per site versus @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] because of absence of phase transitions in one dimension ( if not considering `` exotic '' cases of interaction with particular long - range behavior @xcite ) the free energy shown in fig .
[ fig : free_energy ] does not show any peculiarities as a function of @xmath117 . in the case of attractive interaction , it approaches the line @xmath170 for large @xmath161 due to the aggregation of the rods into one cluster . for strong repulsive interactions , i.e. for @xmath143 significantly larger than @xmath171
, the free energy essentially follows the ( negative ) entropy for @xmath172 , increases linearly for @xmath173 due to the linear increase of non - avoidable contacts until reaching @xmath174 for @xmath175 . for repulsive interactions , the density @xmath156 should also show up as a particular value in the behavior of the chemical potential , because the free energy amount to add a rod to the system is expected to increase strongly around this point .
the chemical potential @xmath176 is given by @xmath177\right\}\ ] ] and plotted in fig .
[ fig : fugacity ] as function of @xmath117 .
it indeed shows a step - like change around @xmath156 for strong repulsive interactions , which could be utilized for determining the rod size from thermodynamic measurements . for comparison the behavior for attractive interaction is also displayed in fig .
[ fig : fugacity ] .
chemical potential @xmath178 as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] chemical potential @xmath178 as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] + chemical potential @xmath178 as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ] chemical potential @xmath178 as a function of the coverage @xmath117 for a homogeneous system of hard rods of sizes @xmath110 and 5 , and various contact interactions @xmath143.,title="fig : " ]
our solution for the distribution of microstates of hard - rod lattice gases with a general nearest - neighbor - range interaction potential provides a promising basis for future studies related to applications .
for example , the results can be utilized to describe formation of molecular nanowires on surfaces , where molecules interact via van der waals interactions and form hydrogen bonds when coming into contact . in the model , such situation could be accounted for by an interaction profile @xmath14 with strong attractive interaction @xmath179 at contact distance @xmath1 and a smoothly decreasing , weaker attractive part for @xmath180 . also , based on experiences in other contexts , @xcite the explicit exact expressions for
the density functionals should allow one to faithfully study the kinetics of wire formation by employing time - dependent density functional theory .
similarly , our results may be helpful in the future to treat diffusion of molecules through nanopores or membrane channels .
our derivation of a fundamental measure form of the hard - rod lattice gas with contact interaction enables a straightforward extension to higher dimensions
. this should be useful to account for transitions between different phases in corresponding systems , which generally resemble nematic ( and other ) phases of liquid crystals.@xcite it has been shown @xcite for athermal hard - rod lattice gases ( @xmath21 ) that extensions to higher dimensions are a powerful means to treat corresponding phase transitions . up to now , we did not succeed to identify fundamental measure forms for general nearest - neighbor - range interactions @xmath14 .
however , there seem to be other possibilities of extensions , which become exact under dimensional reduction
. these will be explored elsewhere .
considering the core in the derivation of the distribution of microstates , it is important to realize that the procedure can in principle be extended to interactions of longer range covering several rod lengths . for a given range @xmath18 , the exclusion constraint leads to a natural decomposition of the set @xmath82 into ranges covering the lattice sites @xmath181 ( first range ) , @xmath182 ( second range ) , and so on . in each of these ranges at
most one occupation number can be equal to one .
the total number of ranges that need to be taken into account is @xmath183 , where @xmath184 denotes the integer ceiling division , i.e. the smallest integer larger than @xmath185 .
accordingly , we would need to consider higher - order correlators @xmath186 , where @xmath187 , specifies the location of the occupied site in the @xmath188th range as in sec .
[ subsec : jointprob ] .
the distribution of microstates can then be expressed in term of these correlators @xmath186 and the densities @xmath37 . by equating with the boltzmann formula for simple configurations ,
relations between the @xmath186 and the @xmath37 eventually can be obtained , which need to be solved for expressing the @xmath186 in terms of the @xmath37 . by working out the long - range asymptotic behavior of correlation functions it should be possible also to study the occurrence and behavior of widom - fisher lines.@xcite these lines separate regions in the temperature - density and temperature - pressure planes , where in one region the pair correlation decays monotonically , while in the other region it oscillates with decaying amplitude .
the abrupt change in the behavior at the lines takes place without any singularities in the thermodynamics . in the original work
, @xcite transition lines were calculated for a one - dimensional continuum fluid with square - well interaction potential and a lattice fluid as considered in this work , corresponding to hard rods of size @xmath189 with first - neighbor coupling .
access to the correlation properties of systems with larger @xmath1 and interactions of longer range may allow one to gain deeper insight into general features of these lines .
let us finally note that the rod length defines a length scale independent of the lattice spacing , which implies that a continuum limit of the results should be accessible .
this , the possible extension to larger interaction ranges , and evaluations of widom - fisher lines open new and challenging possibilities for further investigations .
58ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase { 10.1103/physreve.47.718 } [ * * , ( ) ] link:\doibase { 10.1063/1.471150 } [ * * , ( ) ] link:\doibase 10.1103/physreve.61.422 [ * * , ( ) ] @noop * * , ( ) link:\doibase { 10.1103/physreve.72.031405 } [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.055701 [ *
* , ( ) ] link:\doibase 10.1103/physreva.20.1671 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.63.980 [ * * , ( ) ] link:\doibase { 10.1088/0953 - 8984/14/46/314 } [ * * , ( ) ] link:\doibase { 10.1088/0953 - 8984/15/27/304 } [ * * , ( ) ] link:\doibase 10.1063/1.1615511 [ * * , ( ) ] link:\doibase { 10.1103/physreve.68.066120 } [ * * , ( ) ] link:\doibase { 10.1103/physrevlett.93.130603 } [ * * , ( ) ] \doibase 10.1023/b : joss.0000012518.27400.2a [ * * , ( ) ] link:\doibase 10.1103/physreve.65.066112 [ * * , ( ) ] link:\doibase { 10.1209/0295 - 5075/93/50003 } [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.060603 [ * * , ( ) ] http://stacks.iop.org/0305-4470/13/i=3/a=007 [ * * , ( ) ] link:\doibase 10.1007/bf01011623 [ * * , ( ) ] link:\doibase 10.1063/1.471920 [ * * , ( ) ] link:\doibase 10.1063/1.473286 [ * * , ( ) ] link:\doibase 10.1063/1.1384419 [ * * , ( ) ] link:\doibase 10.1063/1.1645781 [ * * , ( ) ] link:\doibase 10.1063/1.1758693 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1063/1.2647150 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.99.098301 [ * * , ( ) ] link:\doibase 10.1103/physreve.77.051206 [ * * , ( ) ] link:\doibase 10.1063/1.3528226 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.047801 [ * * , ( ) ] link:\doibase 10.1063/1.477746 [ * * , ( ) ] link:\doibase 10.1063/1.1601605 [ * * , ( ) ] link:\doibase 10.1063/1.3081144 [ * * , ( ) ] link:\doibase 10.1063/1.3174449 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.105.068001 [ * * , ( ) ] link:\doibase 10.1063/1.473201 [ * * , ( ) ] link:\doibase 10.1063/1.470751 [ * * , ( ) ] link:\doibase 10.1063/1.1461358 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.106.228102 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.048301 [ * * , ( ) ] link:\doibase 10.1007/bf02183346 [ * * , ( ) ] link:\doibase 10.1063/1.481695 [ * * , ( ) ] link:\doibase 10.1063/1.456635 [ * * , ( ) ] link:\doibase 10.1103/physreve.74.051407 [ * * , ( ) ] @noop * * , ( ) \doibase 10.1023/a:1018652808652 [ * * , ( ) ] link:\doibase 10.1063/1.1670482 [ * * , ( ) ] link:\doibase 10.1143/ptp.21.421 [ * * , ( ) ] link:\doibase 10.1063/1.1699116 [ * * , ( ) ] link:\doibase 10.1103/physrev.137.a1441 [ * * , ( ) ] @noop * * , ( ) @noop ( ) http://stacks.iop.org/0953-8984/1/i=17/a=011 [ * * , ( ) ] link:\doibase 10.1103/physrev.86.821 [ * * , ( ) ] link:\doibase 10.1103/physreve.75.051708 [ * * , ( ) ] link:\doibase 10.1103/physreve.78.031705 [ * * , ( ) ] link:\doibase 10.1103/physreve.81.021706 [ * * , ( ) ] link:\doibase 10.1063/1.1671624 [ * * , ( ) ] | we derive exact density functionals for systems of hard rods with first - neighbor interactions of arbitrary shape but limited range on a one - dimensional lattice .
the size of all rods is the same integer unit of the lattice constant .
the derivation , constructed from conditional probabilities in a markov chain approach , yields the exact joint probability distribution for the positions of the rods as a functional of their density profile . for contact interaction ( `` sticky core model '' ) between rods
we give a lattice fundamental measure form of the density functional and present explicit results for contact correlators , entropy , free energy , and chemical potential .
our treatment includes inhomogeneous couplings and external potentials . |
an extended color sector is an integral part of many theories with new dynamics that attempt to address unresolved challenges with the ordinary standard model ( sm ) .
models including topcolor @xcite , the flavor - universal coloron @xcite , chiral color @xcite , chiral color with unequal gauge couplings @xcite and flavor non - universal chiral color @xcite belong to this class of theories beyond the sm . in such models , the minimal extension of the color sector
consists of enhancing the color gauge group to @xmath4 , which is spontaneously broken to the diagonal @xmath5 subgroup .
ordinary qcd is then identified with this diagonal unbroken gauge group , containing the usual massless gluon color - octet . the spontaneous symmetry breaking produces , furthermore , a set of massive color - octet gauge bosons , which we generically refer to as colorons .
spontaneous breaking of the enhanced color symmetry may be facilitated in a renormalizable manner in conjunction with electroweak symmetry breaking by means of an enlarged scalar sector @xcite . as a consequence , new ( colored and uncolored ) scalar degrees of freedom are present in this formalism ; most notably there is a color - singlet ( heavy ) higgs - like scalar in addition to the electroweak higgs boson .
moreover , cancelation of potential anomalies brought forth by the possible chiral couplings of the ordinary quarks to colorons may require the existence of additional spectator fermions .
hence , if this formalism corresponds to the correct description of nature , a rich spectrum of new ( colored ) scalar , fermionic , and vector particles with novel properties are predicted , the discovery of which may lie within the reach of the lhc .
a first complete study of hadron collider production of colorons at next - to - leading order was presented in @xcite .
the scalar sector of the renormalizable coloron model has also been constrained in a previous analysis @xcite , by imposing limits from the stability of the potential , unitarity , electroweak precision tests , and properties of the 125 gev higgs - like scalar , discovered at the lhc @xcite .
furthermore , the latter study presented a brief discussion of the lhc heavy higgs searches at @xmath6 tev @xcite , applied to the additional cp - even scalar of the renormalizable coloron model . in the present paper ,
we further explore the properties of the additional scalar boson , with a mass up to 1 tev , arising from the minimally extended color sector .
specifically , we discuss the implications of the expected reach of the @xmath7 tev lhc with an integrated luminosity of 300 and 3000 fb@xmath8 for the heavy higgs @xcite , and at 300 fb@xmath8 for the detailed properties of the discovered 125 gev higgs @xcite .
if no signal of a heavy scalar is found nor any deviations from the current measurements of the 125 gev higgs detected by the @xmath7 tev lhc with the quoted luminosities , these projections will impose complimentary constraints at 95% c.l . on the model s parameter space .
we present the plots depicting the current exclusion limits , as well as the projected 14 tev exclusion limits , for three selected scenarios with 0 , 1 , and 3 spectator fermion generations , varying the remaining free parameters within each model to demonstrate the sensitivity of the analyses to these variables .
we show thatthe large regions of the parameter space are sensitive to future probing by the lhc . in section [ review ]
, we briefly describe the formalism of the renormalizable coloron model @xcite , review the particle content , and introduce relevant notation .
section [ shiggs ] is devoted to setting up an effective lagrangian utilized for the subsequent phenomenological study of the heavy @xmath0 scalar , arising from this minimally extended color sector .
the phenomenology of the lhc heavy scalar searches and projections , along with other considerations such as the total width of the heavy @xmath0 boson in the renormalizable coloron model , are presented in section [ spheno ] and summarized in plots , covering the free parameter space for representative scenarios .
we further discuss the results in section [ disc ] for the different models and their ranges of free parameters , and we demonstrate that all scenarios may be fully explored by the lhc . if no heavy scalar is discovered at or below a 1 tev mass , the searches will exclude the renormalizable coloron model within the tev range .
furthermore , we elaborate on distinguishing a potentially discovered heavy scalar , originating from the renormalizable coloron model , from other color - singlet `` higgs portal '' scenarios @xcite . finally , we present our conclusions in section [ concl ] .
we start by briefly reviewing the formalism of the minimal renormalizable coloron model @xcite , and summarizing the gauge , scalar , and fermionic sectors of the theory . in the renormalizable coloron model ,
the color gauge group of the standard model ( sm ) is extended to @xmath4 , which is spontaneously broken to the diagonal @xmath5 subgroup , and the latter is identified with the ordinary qcd .
schematically , we have @xmath9 spontaneous symmetry breaking in this enlarged color sector , along with spontaneous breaking of the sm electroweak gauge group , proceeds via an extended scalar sector .
in addition to the usual sm higgs doublet , @xmath10 , the scalar sector contains a @xmath11 scalar , @xmath12 , that is bi - fundamental under the @xmath4 color group , but singlet under the electroweak interactions .
component - wise these two scalars may be written as @xcite @xmath13 where @xmath14 is the sm higgs boson with the corresponding vacuum expectation value ( vev ) @xmath15 gev , @xmath16 are the electroweak nambu - goldstone bosons , and @xmath17 are the gell - mann matrices . as explained in @xcite , the fields @xmath18 and @xmath19 are , respectively , cp - even and cp - odd gauge - singlet scalars , whereas @xmath20 denotes a set of scalars that transform as a color - octet under the @xmath5 color group . in addition , the @xmath12 field contains a set of massless colored nambu - goldstone bosons , @xmath21 .
the cp - even gauge - singlet , @xmath18 , develops a non - zero vev , @xmath22 , resulting in spontaneous breaking of the extended color symmetry .
it has been shown in @xcite that the most general renormalizable scalar potential is of the form @xmath23 } \right)}^2 + \frac{\kappa_s}{2 } { \rm tr}{\left [ { \left ( \phi^\dagger \phi \right)}^{2 } \right ] } -\frac{\lambda_s + \kappa_s}{\sqrt{6}}\,r_\delta v_{s } { \left ( { \rm det}\phi + { \rm h.c . }
\right ) } -\frac{\lambda_s + \kappa_s}{6 } \ , v_s^{2 } { \left ( 1 - r_{\delta } \right ) } { \rm tr}{\left [ \phi^\dagger \phi \right ] } \\ & + \frac{\lambda_h}{6}{\left ( \phi^\dagger \phi - \frac{v^2_h}{2 } \right)}^2 + \lambda_m{\left ( \phi^\dagger \phi - \frac{v^2_h}{2 } \right ) } { \left ( { \rm tr}{\left [ \phi^\dagger \phi \right ] } - \frac{v^2_s}{2 } \right ) } \ , \end{split}\ ] ] with @xmath24 , @xmath25 , @xmath26 , @xmath27 , and @xmath28 all dimensionless couplings .
moreover , it was demonstrated that this potential is bounded from below , with a global minimum coinciding with the vevs shown above , given the conditions @xmath29 with @xmath30 . in the mass eigenstate basis , the cp - even scalars @xmath14 and @xmath31 in and
are mixed due to the @xmath32 term in the potential and due to the fact that they both develop non - zero vevs .
all the other scalar masses remain diagonal in both mass and field eigenstate bases .
the corresponding mass eigenstates @xmath1 and @xmath0 composed of the original cp - even scalars may be defined using an orthogonal rotation @xmath33 with the mixing angle , @xmath34 , given by @xmath35 } \ .\ ] ] the diagonal scalar masses then read @xmath36}\sec 2\chi \right\ } } \ , \label{mhs } \\
m_{\cal a}^{2 } = & \ , \frac{v_{s}^{2 } } { 2 } \ , r_{\delta } \lambda_{s}^{\prime } \ , \qquad m_{g_{h}}^{2 } = \frac{1}{3 } { \left [ v_{s}^{2 } \ , \kappa_{s}+ 2 m_{\cal a}^{2 } \right ] } \ . \label{magh}\end{aligned}\ ] ] the @xmath1 scalar is assumed to be the lighter of these two cp - even degrees of freedom , and is identified with the discovered 125 gev higgs - like state at the lhc @xcite . for small mixing angle values motivated by the experimental constraints @xcite
the @xmath1 is more sm - like " , whereas the heavier @xmath0 is more singlet - like " ( c.f . ) .
they are both capable of interacting with the ordinary sm particles , albeit with suppressed tree - level couplings proportional to @xmath37 and @xmath38 , respectively , as compared with a pure sm higgs .
analyzing the phenomenology of the heavier scalar , @xmath0 , constitutes the main subject of the present study . examining and together with the conditions , one further notes the following mass relations @xcite @xmath39 a scalar color - octet mass , @xmath40 , ranging from 50 to 125 gev is already excluded by the tevatron searches @xcite ; hence , following @xcite , we treat this color - octet set as being heavier than the discovered @xmath1 scalar ( @xmath41 gev ) throughout our analyses .
the spontaneous breaking of the enhanced color symmetry to its diagonal subgroup produces , in addition to the usual massless @xmath5 gluons , a set of massive color - octet vector bosons , generically referred to as colorons .
colorons obtain their mass by eating " the colored nambu - goldstone bosons , @xmath42 in ; the mass is given by the expression @xcite @xmath43 here , @xmath44 represents the mixing angle in the orthogonal matrix rotating the @xmath45 and @xmath46 gauge eigenstates ( with the corresponding couplings @xmath47 and @xmath48 ) into the gluon and coloron mass eigenstates .
furthermore , @xmath49 denotes the qcd gauge coupling , which may be expressed as the combination @xcite @xmath50 note that tevatron and current lhc searches constrain the coloron mass , @xmath51 , to be at least in the tev region @xcite .
as described in @xcite , in the context of the renormalizable coloron model , the fermionic matter sector needs modifications as well , possibly including the addition of heavy spectator quarks @xcite .
these extra fermionic degrees of freedom can be necessary to cancel potential anomalies of the theory ( arising once the couplings of the quarks to the @xmath4 color gauge group are chosen to be chiral @xcite ) , and must have the opposite chirality charges as the ordinary quarks .
allowing for the spectator fermions to obtain their mass via yukawa interactions with the @xmath12 scalar , one deduces @xmath52 where , for simplicity , a flavor - universal yukawa coupling , @xmath53and hence spectator fermion mass scale is assumed .
furthermore , it is conjectured @xcite that the spectator flavors have the same electric charges as their corresponding quark partners , and that they are vectorial under the electroweak interactions .
flavor - changing couplings of the colorons result in strong constraints on potential mixing between these extra fermionic states and the ordinary quarks , rendering such a mixing negligible @xcite .
in addition to those assumptions regarding the properties of the heavy spectator fermions and their interactions , the chiral nature of a specific model determines the number of necessary spectator flavors to cancel potential anomalies @xcite .
no spectators are required when all quarks are vectorially charged under the extended color group , since no anomaly will be introduced .
if , however , the chiral couplings of the third quark generation are chosen to be opposite to those of the first two generations , then one spectator generation ( one up - like and one down - like spectator ) is necessary .
flavor - universal chiral interactions of the quarks under the enlarged color gauge group require , on the other hand , three generations of spectator fermions to cancel all anomalies .
following @xcite , we shall study the phenomenological results for three cases described above , where 0 , 1 , or 3 spectator quark generations may be present in the theory . in summary ,
the renormalizable coloron model extends the sm color gauge group minimally ( c.f . ) , while appropriately enlarging the scalar sector ( to also accommodate spontaneous breaking of the enhanced color symmetry ) , as well as the fermionic matter sector ( to cancel potentially introduced anomalies ) . setting the electroweak vev at @xmath54 gev and the mass of the @xmath1 higgs at @xmath55 gev
, it adds eight new free parameters to the usual sm , which may be taken as the set @xcite @xmath56 with @xmath57 the number of spectator fermion generations , which we keep as a free parameter throughout our analyses . for completeness
, we exhibit the explicit dependence of the lagrangian parameters on the set , which can be straightforwardly derived from , , , and , @xmath58 it is worth noting that the sign of @xmath32 ( i.e. an attractive or repulsive interaction between the scalar fields @xmath10 and @xmath12 in ) does not affect the stability of the potential ( c.f . ) , and is reflected only in the sign of the parameter @xmath38 in . as we shall demonstrate , however , this sign has notable effects on the phenomenology of the heavy higgs - like scalar , @xmath0 .
as explained in the previous section , the heavy @xmath0 scalar is an admixture of the sm @xmath14 higgs and the gauge - singlet @xmath31 boson ( c.f . ) , and is mostly singlet - like for the small mixings favored by the lhc 125 gev signal data and the electroweak precision tests @xcite .
this heavy scalar is , nevertheless , capable of interacting weakly with the sm particles , while its tree - level coupling is suppressed by the mixing angle factor , @xmath38 , with respect to that of the usual sm higgs boson , @xmath14 .
the lhc heavy higgs searches @xcite and projected 14 tev exclusion limits @xcite in the vector boson channels can , thus , be utilized to investigate and constrain the properties of the @xmath0 boson arising from the extended color sector . to this end
, we set up an effective lagrangian @xcite to parametrize all the relevant decay channels of the heavy @xmath0 higgs at the lhc @xmath60 in addition to the tree - level sm gauge boson and heavy fermion channels , this effective lagrangian describes the one - loop induced @xmath0 decay into a pair of gluons , as well as its tree - level decays into pairs of 125 gev @xmath1 higgses , @xmath61 pseudoscalars , and @xmath62 scalar color - octets .
decay into di - photons is a small contribution to the total decay width and is not relevant to our analysis .
] given the tev lower bounds on the coloron mass @xcite , we have neglected a potential decay of the @xmath0 boson into a pair of colorons in .
furthermore , we focus our attention on the case where the spectator fermions are heavy , with @xmath63 in the tev range @xcite , such that the @xmath0 decay into a pair of spectator fermions is also kinematically prohibited . in this regime ,
the @xmath0 boson production and decay amplitudes depend crucially on the number of spectator quarks , but are relatively insensitive to the precise mass of these fermions . since the coupling of the @xmath0 boson to a coloron or spectator fermions is proportional to the coloron s or spectator s mass ( and , respectively ) , the contributions of the spectator fermions and colorons to @xmath0 boson production via gluon fusion do not decouple " in the heavy mass limit .
the production cross section is , therefore , not significantly suppressed in this regime . in case of the @xmath0 decay into a sm final state , such as @xmath64 or @xmath65 ,
the dimensionless coefficient @xmath66 in represents the appropriate deviation of the corresponding coupling from its sm value ( i.e. , if @xmath0 were replaced by the usual sm higgs boson , @xmath67 ) .
the tree - level coefficients are , in this case , given by @xmath68 whereas , the coefficient associated with the one - loop effective gluon coupling ( with colorons , spectators , and scalar color - octets running in the loop , in addition to the ordinary sm heavy quarks ) may be parametrized by @xcite @xmath69\label{delcgs } \ .\end{aligned}\ ] ] here , @xmath70 , the subscript @xmath71 ( @xmath72 ) represents the coloron ( spectator ) , @xmath57 is the number of spectator fermion generations , and the vector , fermion , and scalar loop form factors are defined as @xmath73 } \
, \quad a_f(\tau)\equiv \frac{3}{2\tau^2}{\left [ \tau- ( 1-\tau)f(\tau ) \right ] } \ , \quad a_s(\tau)\equiv\frac{1}{8\tau^2}{\left [ \tau - f(\tau ) \right ] } \ , \\
f(\tau ) \equiv&\ , \begin{cases } \arcsin^{2}\sqrt{\tau } \qquad \qquad \qquad \qquad \ ; \,\tau\leq1\\ -\frac{1}{4}{\left [ \log \frac{1+\sqrt{1-\tau^{-1}}}{1-\sqrt{1-\tau^{-1}}}-i\pi \right]}^2 \qquad \tau>1 \end{cases } \ . \end{split}\ ] ] it is important to notice the sign in associated with @xmath38 : an attractive or repulsive interaction between the scalars @xmath10 and @xmath12 as parametrized by the sign of @xmath32 in the potential see also
. this may result in a constructive or destructive interference between the two terms .
in addition , a possible cancellation between the spectator and coloron contributions may occur in , depending on the chosen number of spectator generations , while the scalar term can be of either sign .
this complicated interplay between the various parameters lead to non - trivial phenomenological consequences for the @xmath0 boson that will be discussed in section [ spheno ] .
the remaining dimensionful scalar coupling coefficients were also derived in @xcite , and we list them here for completeness @xmath74 \ , \label{chs}\\ c_\mathcal{a}^s = & \,-\frac{m_\mathcal{a}^2+m_s^2}{2v_s}\cos\chi \ , \label{cas}\\ c_{g_h}^s=&\ , -\frac{m_s^2 + 2m_{g_h}^2-\frac{2}{3}m_\mathcal{a}^2}{2v_s}\cos\chi \label{cghs } \ .\end{aligned}\ ] ] armed with the effective lagrangian , in the following section we shall adapt the lhc heavy higgs searches and projections in the vector boson channels to study and constrain the properties of the @xmath0 boson .
the lhc heavy higgs searches @xcite and projected 14 tev exclusion limits @xcite generally assume that the narrow - width approximation is valid . before applying these studies to examine the @xmath0 boson s properties , it is , therefore , necessary to confirm the validity of this approximation in the context of the heavy @xmath0 scalar originating from an extended color sector . using the constructed effective lagrangian and the coefficients - , the total width of the @xmath0 boson as a function of its mass , @xmath75 ,
is given by @xcite @xmath76}+\frac{\gamma(s\to hh)+\gamma(s\to\mathcal{a}\mathcal{a})+8\ , \gamma(s\to g^a_hg^a_h)}{\gamma_{h_{0}}(m_s ) } \ , \ ] ] where @xmath77 denotes the total decay width of the sm @xmath67 higgs when @xmath78 ( c.f . ) .
the expression takes into account the decay of @xmath0 into the usual sm final states , as well as into the additional scalar degrees of freedom present in the theory , such as the pseudoscalar @xmath61 and the scalar color - octet @xmath20 . inserting the coefficients - , the corresponding widths of the @xmath0 decay into scalar pairs
are defined as @xmath79 the narrow - width approximation concerns , specifically , the ratio of the total width of the heavy @xmath0 boson to its mass ( @xmath80 ) , which must remain small over the entire relevant range of the mass .
this ratio has been plotted in fig .
[ nwa ] for three selected values of the singlet vev , @xmath22 , covering a heavy @xmath0 boson up to @xmath81 tev . in each panel , three representative values of the mixing angle have been displayed , as motivated by the experimental bounds @xcite . in all plots , the pseudoscalar and the scalar color - octet
have been chosen to be light enough to contribute to the @xmath0 width , making the total width as broad as possible for the mass range of interest , with @xmath82 gev and @xmath83 gev . into pairs of @xmath1 , @xmath84 and @xmath20 in computing its total width .
an off - shell decay of @xmath0 into the usual sm pairs ( e.g. @xmath85 ) is , however , taken into account .
] hence , the plots depict a scenario in which decay to a pair of @xmath61-bosons contributes to the @xmath0 width .
the heavy tev - range coloron and spectator masses render the ratio insensitive to the precise values of @xmath51 and @xmath86 parameters .
sensitivity to the number of spectator fermion generations , @xmath87 , is also negligible .
one can deduce from fig .
[ nwa ] that the @xmath0 boson s total width remains relatively narrow in the entire mass range of interest for all possible scenarios of the model ; in fact , the @xmath0 boson is much narrower than a corresponding heavy sm - like @xmath67 higgs .
the narrow - width approximation , therefore , remains valid and the aforementioned searches and projections apply throughout our heavy @xmath0 analyses .
boson s total width to its mass as a function of its mass for the range @xmath88 gev , exhibiting the validity of the narrow - width approximation .
the panels correspond to three benchmark values of the singlet vev , @xmath22 , each displaying this ratio for three different mixing angles , @xmath38 , as motivated by the experimental searches @xcite .
a `` light '' pseudoscalar ( @xmath82 gev ) and scalar color - octet ( @xmath83 gev ) have been chosen to make the width as broad as possible for the entire range of @xmath89 , demonstrating the `` worst - case '' scenarios .
the ratio is , furthermore , insensitive to the ( tev range ) coloron and spectator masses , and the number of spectator fermion generations.,title="fig : " ] boson s total width to its mass as a function of its mass for the range @xmath88 gev , exhibiting the validity of the narrow - width approximation .
the panels correspond to three benchmark values of the singlet vev , @xmath22 , each displaying this ratio for three different mixing angles , @xmath38 , as motivated by the experimental searches @xcite .
a `` light '' pseudoscalar ( @xmath82 gev ) and scalar color - octet ( @xmath83 gev ) have been chosen to make the width as broad as possible for the entire range of @xmath89 , demonstrating the `` worst - case '' scenarios .
the ratio is , furthermore , insensitive to the ( tev range ) coloron and spectator masses , and the number of spectator fermion generations.,title="fig : " ] boson s total width to its mass as a function of its mass for the range @xmath88 gev , exhibiting the validity of the narrow - width approximation .
the panels correspond to three benchmark values of the singlet vev , @xmath22 , each displaying this ratio for three different mixing angles , @xmath38 , as motivated by the experimental searches @xcite .
a `` light '' pseudoscalar ( @xmath82 gev ) and scalar color - octet ( @xmath83 gev ) have been chosen to make the width as broad as possible for the entire range of @xmath89 , demonstrating the `` worst - case '' scenarios .
the ratio is , furthermore , insensitive to the ( tev range ) coloron and spectator masses , and the number of spectator fermion generations.,title="fig : " ] at this stage , we have developed all the necessary tools in order to investigate the properties of the heavy @xmath0 boson by utilizing the lhc heavy higgs searches @xcite and projected 14 tev exclusion limits @xcite in the vector boson channels .
in particular , employing the narrow - width approximation , these measurements and simulations constrain the @xmath0 production cross section times its branching ratio , as compared with that of a sm higgs , @xmath67 , with the same mass . as discussed in @xcite , the amount of mixing between the sm @xmath67 higgs and the gauge - singlet @xmath18 ( c.f . )
is constrained to small values , @xmath90 , by the electroweak precision data and the properties of the discovered 125 gev lhc @xmath1 signal .
furthermore , it was shown that , in this small mixing region , the @xmath0 production via gluon - fusion dominates over the other mechanisms .
hence , in order to adapt the lhc searches and exclusion projections in vector boson channels to study the @xmath0 boson , we may construct the following parametrization @xmath91^{-1 } \label{ggfsvv } \ , \end{aligned}\ ] ] where we have inserted the appropriate coefficients from , and the ratio of the widths is given by . in the previous treatment @xcite ,
we have briefly discussed the constraints on the @xmath0 boson phenomenology , based on the lhc heavy higgs searches in the vector boson channels for @xmath6 tev @xcite . in this section ,
we extend that discussion , by further considering the expected @xmath0 production cross section times branching ratio for @xmath7 tev lhc , with an integrated luminosity of 300 and 3000 fb@xmath8 .
these simulations have been performed for the atlas detector assuming a sm - like heavy higgs @xcite , and for the cms detector assuming a heavy higgs in the context of the 2hdm @xcite .
if the upcoming lhc run at @xmath7 tev with the quoted luminosities does not detect a heavy higgs - like scalar at or below the mass of 1 tev , the projected exclusion limits will provide complementary constraints at 95% c.l . on the parameter space of the renormalizable coloron model . to be specific ,
we utilize the heavy higgs search and exclusion projections to understand the lhc sensitivity to the @xmath92 parameter ( left - hand side of ) in each case , and compare the latter with the theoretically derived expression ( right - hand side of ) .
if no heavy higgs signal is detected by the lhc , this may be translated into 95% c.l .
constraints on the model s variables .
the @xmath6 tev heavy higgs searches @xcite directly quote the @xmath92 parameter in various decay channels , and we use the strongest bounds determined by the @xmath0 decays into @xmath93 and @xmath94 final states .
the @xmath7 tev heavy higgs exclusion projections @xcite , however , quote the obtained production cross section times decay branching ratios , with the most stringent limits arising from the @xmath94 decay channel . to construct the @xmath92 parameter for these simulations , we divide their obtained production cross section times @xmath94 decay branching ratio by that of the sm @xmath67 higgs , the latter being provided in @xcite . fig .
[ allmu ] displays the most stringent values of the @xmath92 parameter as a function of the @xmath0 boson mass , incorporating the @xmath6 tev lhc heavy higgs search data ( @xmath93 and @xmath94 final states ) @xcite , as well as the exclusion projections for @xmath7 tev with an integrated luminosity of 300 and 3000 fb@xmath8 ( @xmath94 final states ) , based on search projections for a sm - like @xcite and a 2hdm - like @xcite heavy higgs .
clearly , future searches with higher center - of - mass energies and luminosities are anticipated to yield more stringent upper bounds on the @xmath92 parameter and the model s parameter space at 95% c.l .
, if no heavy higgs - like scalar is detected at the lhc in the displayed mass range .
nonetheless , one notes the similarity in the obtained results for the sm - like and the 2hdm - like heavy higgs projections .
the differences between the two may be attributed to the differences in widths of the two particles .
given this similarity , in the following , we shall apply only the sm - like higgs expected @xmath95 projections @xcite to our analyses of the @xmath0 boson phenomenology .
higgs production cross section times its vector boson decay branching ratio , divided by that of a corresponding sm @xmath67 higgs with the same mass , as parameterized by @xmath92 in , for @xmath96 gev .
the lhc @xmath6 tev heavy higgs exclusion limits at 95% c.l .
@xcite are displayed by the dot - dashed line , with the most stringent bounds arising from the @xmath93 and @xmath94 final states .
these data are complemented by the @xmath7 tev projected exclusion limits at 95% c.l . with an integrated luminosity of 300 and 3000 fb@xmath8 for atlas ( solid lines ) assuming a sm - like heavy higgs @xcite , and for cms ( dashed lines ) assuming a heavy higgs in the 2hdm @xcite . in both 14 tev projections ,
the strongest limits are determined by the @xmath94 final states . ] separately , we consider the projected ability of the atlas experiment at @xmath7 tev atlas with 300 fb@xmath8 of integrated luminosity to more precisely measure the properties of the existing 125 gev @xmath1 higgs @xcite .
similar to the case of the heavy @xmath0 scalar , these @xmath1 boson measurements present additional constraints on the parameter space of the model , assuming no large deviations from the current measurements of the @xmath1 properties are discovered at the @xmath7 tev lhc run .
the strongest bounds in this study arise from the gluon - fusion production mechanism and the @xmath94 decay mode ; hence , we take only these particular channels into account .
the analogous expression for the @xmath92 parameter of the @xmath1 higgs has been provided in @xcite ; it formally depends on all the input parameters , except for the mass of the @xmath0 boson , @xmath89 . according to @xcite , at 300 fb@xmath8
, the experimental relative uncertainty of the signal strength is @xmath97 .. the theoretical uncertainties , however , are expected to decrease and therefore , in this paper , we illustrate the sensitivity of the 14 tev lhc to this model using the experimental uncertainties alone . ] assuming @xmath98 , and requiring no exclusion by the previous @xmath99 tev best fit at 95% c.l .
@xcite , we impose the 95% c.l .
exclusion constraint from the @xmath100 tev @xmath1 projections at 300 fb@xmath8 to find the sensitivity of high - energy lhc data to this model . as discussed in section [ review ] , we present our plots for three scenarios containing various spectator fermion generations , @xmath101 , 1 , and 3 .
the results are exhibited , respectively , in figs .
[ resultnq0]-[resultnq3 ] .
the panels incorporate the constraints explored in @xcite arising from unitarity , electroweak precision tests , and lhc 125 gev @xmath1 higgs data , as well as the lhc heavy higgs searches with @xmath6 tev @xcite , and the @xmath0 and @xmath1 higgs atlas projections for @xmath7 tev with an integrated luminosity of 300 fb@xmath8 at 95% c.l .
@xcite . highlighting the importance of the sign associated with @xmath38 for the lhc analyses ( see the discussion below ) , the current plots are extended to cover the entire range @xmath102 .
higgs and the heavy @xmath0 higgs depend on it . ] for each scenario , three benchmark values of the singlet vev , @xmath103 , have been selected , within which the mass of the color - octet scalar boson , @xmath104 , is varied from light to heavy . a universal pseudoscalar mass , @xmath105 gev , is also assumed .
region only for @xmath106 gev , imposing a lower bound on @xmath89 ( c.f . ) .
hence , for the quoted pseudoscalar masses , this theoretical constraint is absent in our displayed figures . ]
we have , furthermore , set the spectator and coloron tev - range masses to @xmath107 tev and @xmath108 tev throughout , as reference values . for colorons and spectators this heavy , the explicit dependence of the analysis on the precise values of @xmath109 and @xmath63 is negligible .
the limits arising from more precise measurements of the 125 gev @xmath1 boson properties are weaker or at most comparable to those from the @xmath0 scalar projections in all depicted scenarios , except for @xmath110 with very high singlet vev and color - octet masses .
the @xmath1 higgs projection constraints are , therefore , omitted in the remaining scenarios . ,
represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
a universal pseudoscalar mass , @xmath112 gev , has been used for the purpose of illustration .
three selected values of the singlet vev , @xmath103 , are displayed in the three rows , within which the scalar color - octet mass , @xmath104 , is varied from light ( left plot ) to heavy ( right plot ) .
the exclusion limits arise from imposing unitarity ( dot - dashed purple ) , electroweak precision tests ( dotted orange ) , lhc direct measurements of the 125 gev @xmath1 higgs ( vertical dashed gray ) , and the @xmath6 tev lhc heavy @xmath0 boson searches ( solid yellow ) .
the inclined - shaded green region corresponds to the heavy @xmath0 boson exclusion projections for @xmath7 tev atlas with an integrated luminosity of 300 fb@xmath8 .
given their heavy nature , the dependence on the coloron and spectator masses , @xmath109 and @xmath63 , is negligible throughout.,title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ] , represented in the @xmath111 plane , covering the heavy @xmath0 scalar mass range @xmath96 gev and the full range of mixing angle values @xmath102 .
the additional horizontally - shaded gray region in the @xmath113 tev case accounts for the exclusion by more precise measurements of the 125 gev @xmath1 higgs .
( for details , see the caption of fig .
[ resultnq0]),title="fig : " ]
within each @xmath87 scenario , it is evident that a larger singlet vev , @xmath103 , enlarges the viable parameter space by alleviating the exclusion constraints , and , at the same time , accommodates heavier scalar color - octets .
moreover , for the different signs of the mixing angle , @xmath38 , one observes an asymmetry in all lhc @xmath1 and @xmath0 searches and projections , indicating the sensitivity of these analyses to this sign .
the origin of this sensitivity may be traced back to the constructive or destructive role played by the sign in the heavy @xmath0 ( and @xmath1 ) production via gluon fusion ( see the discussion below ) .
it is , furthermore , interesting to note that the pseudoscalar mass , @xmath114 , has the effect of determining the lower bound on the mass of the heavy @xmath0 scalar near the @xmath115 region .
this is due to the fact that the @xmath0 decay channel into a pair of pseudoscalars is proportional to @xmath116 and becomes accessible at a lower @xmath89 for a lighter @xmath84 , further suppressing its decay branching ratios into pairs of vector bosons .
as anticipated , the projections demonstrate that the @xmath7 tev lhc is far more sensitive to a potential heavy scalar signal than the @xmath6 tev searches already carried out .
in fact , the heavy @xmath0 boson projections with 300 ft@xmath8 luminosity already cover practically the entire remaining region of the parameter space in all possible scenarios , except for the very high singlet vev and color - octet masses ( deep in the tev region ) that are only available within the @xmath110 scenario .
this signals the interesting predictive power of the model , which can be thoroughly probed by the imminent lhc searches .
given this increased sensitivity , in case no heavy higgs signal is detected by the lhc at @xmath7 tev , the displayed projections at 300 ft@xmath8 tightly constrain at 95% c.l . the model s parameter space , while the 3000 ft@xmath8 luminosity data ( not shown ) completely rule out the model in the depicted region of interest ( c.f . fig .
[ allmu ] ) . a comparison of the three models with different spectator generation contents , as presented in figs
. [ resultnq0]-[resultnq3 ] , reveals that the model with three spectator fermion generation ( @xmath110 ) is the scenario that is least constrained by the current theoretical and experimental exclusions for a wide range of the free parameters , such as the scalar color - octet mass and the singlet vev .
this is attributable to the large cancellation occurring between the coloron and spectator contributions in the effective @xmath0 coupling to the gluon pair .
the projected bounds on the properties of the @xmath1 higgs , on the other hand , do not suffer from such cancellations in this region of parameter space and become relevant , providing additional potential constraints at 300 ft@xmath8 .
consequently , in all three cases , it is generally true that nearly the entire parameter space of the models lies within reach of the @xmath3 tev lhc .
finally , let us elaborate on how a potential discovery of a heavy scalar boson by the lhc may be identified with the @xmath0 boson of the minimally extended color sector . as discussed in section [ review ] , within the context of the renormalizable coloron model , the heavy @xmath0 boson
is necessarily accompanied by a multitude of additional scalar , vector , and fermionic degrees of freedom .
being color - charged many of these additional states may reveal their existence in a high center - of - mass energy hadron collider .
a heavy @xmath0 scalar in contrast with the scalar in many other proposed `` higgs portal '' models @xcite is likely to be discovered in association with other ( colored ) resonances of comparable mass .
moreover , the existence of these additional colored states may enhance the effective gluon coupling to the heavy @xmath0 boson , potentially compensating for the suppressed coefficient proportional to the small mixing angle ( @xmath38 ) , which is favored by the current experimental constraints .- boson , another potential distinguishing feature of this state .
] this may , in turn , enhance the heavy higgs production cross section an effect absent in the color - singlet higgs portal extensions and potentially visible in the global fitting analyses of the data .
we have studied the collider phenomenology of a heavy color - singlet scalar boson originating in models with a minimally extended color sector .
the properties of the heavy scalar
a color - singlet with some degree of weak interactions turn out to be closely related to the properties of additional ( colored ) states in the theory , which may include scalars , vectors , or fermionics . incorporating the projected @xmath117 tev 95% c.l .
exclusion limits , along with the previously obtained exclusion bounds @xcite imposed by unitarity , electroweak precision tests , lhc 125 gev @xmath1 higgs searches , and @xmath6 tev lhc heavy higgs searches , we have investigated various scenarios within the renormalizable coloron model , with zero , one , or three generations of the spectator fermions .
we have demonstrated the sensitivity of the future high center - of - mass energy and luminosity lhc searches to this model .
our results are summarized in figs .
[ resultnq0]-[resultnq3 ] , covering the free parameter space of the theory for the scenarios with different numbers of spectator fermion generations .
we find that the upcoming lhc searches should be sensitive to an @xmath0 scalar of mass less than 1 tev for essentially all of the model parameter space in which the @xmath1 state differs from the higgs boson of the sm . more precisely ,
unless the mixing angle , @xmath118 , is zero , the 14 tev lhc will be sensitive to the presence of the non - standard heavy @xmath0 scalar that is characteristic of the renormalizable coloron model .
r.s.c . and e.h.s .
are supported , in part , by the u.s .
national science foundation under grant no .
phy-0854889 . during the completion of this work , a.f . was supported in part by the tsinghua outstanding postdoctoral fellowship , and by the nsf of china ( under grants 11275101 , 11135003 ) .
is supported by the china scholarship council , and by the nsf of china ( under grants 11275101 , 10625522 , 11135003 ) .
p. h. frampton and s. l. glashow , phys .
b * 190 * , 157 ( 1987 ) .
m. v. martynov and a. d. smirnov , mod .
a * 24 * , 1897 ( 2009 ) [ arxiv:0906.4525 [ hep - ph ] ] .
p. h. frampton , j. shu and k. wang , phys .
b * 683 * , 294 ( 2010 ) [ arxiv:0911.2955 [ hep - ph ] ] . c. t. hill and s. j. parke , phys
. rev .
d * 49 * , 4454 ( 1994 ) [ hep - ph/9312324 ] .
d. a. dicus , b. dutta and s. nandi , phys .
d * 51 * , 6085 ( 1995 ) [ hep - ph/9412370 ]
. y. bai and b. a. dobrescu , jhep * 1107 * , 100 ( 2011 ) [ arxiv:1012.5814 [ hep - ph ] ] .
r. s. chivukula , a. farzinnia , j. ren and e. h. simmons , phys .
d * 88 * , 075020 ( 2013 ) [ arxiv:1307.1064 [ hep - ph ] ] .
r. s. chivukula , a. farzinnia , e. h. simmons and r. foadi , phys .
d * 85 * , 054005 ( 2012 ) [ arxiv:1111.7261 [ hep - ph ] ] .
r. s. chivukula , a. farzinnia , j. ren and e. h. simmons , arxiv:1303.1120 [ hep - ph ] . g. aad _ et al .
_ [ atlas collaboration ] , phys .
b * 716 * , 1 ( 2012 ) [ arxiv:1207.7214 [ hep - ex ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b * 716 * , 30 ( 2012 ) [ arxiv:1207.7235 [ hep - ex ] ] .
[ cms collaboration ] , cms - pas - hig-13 - 002 ; cms - pas - hig-13 - 003 .
t. aaltonen _ et al . _
[ cdf collaboration ] , [ arxiv:1303.2699 [ hep - ex ] ] .
e. h. simmons , phys .
d * 55 * , 1678 ( 1997 ) [ hep - ph/9608269 ] .
i. bertram and e. h. simmons , phys .
b * 443 * , 347 ( 1998 ) [ hep - ph/9809472 ] .
g. aad _ et al . _ [ atlas collaboration ] , jhep * 1301 * , 029 ( 2013 ) [ arxiv:1210.1718 [ hep - ex ] ] .
[ atlas collaboration ] , atlas - conf-2012 - 148 .
s. chatrchyan _ et al . _
[ cms collaboration ] , arxiv:1302.4794 [ hep - ex ] .
r. schabinger and j. d. wells , phys . rev .
d * 72 * , 093007 ( 2005 ) [ hep - ph/0509209 ] .
r. barbieri , t. gregoire and l. j. hall , hep - ph/0509242 .
b. patt and f. wilczek , hep - ph/0605188 .
cms collaboration [ cms collaboration ] , cms - pas - b2g-12 - 015 .
[ atlas collaboration ] , atlas - conf-2013 - 018 . for a general review of heavy electroweak vector fermions ,
see e.g. s. a. r. ellis , r. m. godbole , s. gopalakrishna and j. d. wells , arxiv:1404.4398 [ hep - ph ] .
[ atlas collaboration ] , atl - phys - pub-2013 - 014 | we discuss the properties and lhc phenomenology of a potentially discoverable heavy scalar boson ( @xmath0 ) that arises in the context of the renormalizable coloron model ; the model also contains a light scalar , @xmath1 , identifiable with the 125 gev state discovered by the lhc .
these two scalar mass eigenstates are admixtures of a weak doublet gauge eigenstate and a weak singlet gauge eigenstate .
a previous study set exclusion limits on the heavy @xmath0 scalar , using the stability of the scalar potential , unitarity , electroweak precision tests , lhc searches for the 125 gev higgs ; it also briefly discussed the @xmath2 tev lhc searches for a heavy higgs . in this work ,
we show how the projected lhc sensitivity at @xmath3 tev to the presence of a heavy higgs and to the detailed properties of the 125 gev higgs will further constrain the properties of the new heavy @xmath0 scalar .
since the renormalizable coloron model may contain spectator fermions to remove anomalies , we examine several representative scenarios with different numbers of spectator fermions .
our results are summarized in plots that overlay the current exclusion limits on the @xmath0 boson with the projected sensitivity of the @xmath3 tev lhc to the new state .
we find that the upcoming lhc searches should be sensitive to an @xmath0 scalar of mass less than 1 tev for essentially all of the model parameter space in which the @xmath1 state differs from the higgs boson of the sm . more precisely ,
unless the mixing between the weak doublet and weak singlet gauge - eigenstate scalars is zero , the 14 tev lhc will be sensitive to the presence of the non - standard heavy @xmath0 state that is characteristic of the renormalizable coloron model . |
the recent discovery of apparently massive and relaxed clusters of galaxies at redshifts @xmath12 offers a unique opportunity to study the evolution of gravitationally bound systems over an extended look - back time .
moreover , if such systems are proven to be massive , especially those at @xmath13 , their sheer existence can impose stringent constraints on viable cosmological models @xcite @xcite .
in addition , a large sample of confirmed clusters , spanning a broad redshift range , is of great interest for constraining models of formation and evolution of early - type galaxies and large - scale structure , and for the selection of targets in different redshift intervals for a variety of other applications . while the existence of high - redshift clusters is by
now generally accepted , current samples of spectroscopically confirmed clusters , predominantly @xmath14-ray selected , are small and unlikely to grow in the near future , at least not until deep observations with chandra and xmm are carried out and analyzed .
this is particularly true at the highest redshifts ( @xmath15 ) , where only a handful of systems have been confirmed so far @xcite @xcite @xcite .
an alternative approach is to use moderately deep @xmath16-band surveys to identify overdensities in the projected distribution of galaxies .
surveys suitable for this purpose include the palomar distant cluster survey @xcite , covering over 5 square degrees in the northern hemisphere , and the eso imaging survey ( eis ) @xcite @xcite @xcite , covering about 17 square degrees in the southern hemisphere .
the main advantage of optical searches is that large areas of the sky can be covered using wide - field imagers .
however , the yield of physical associations from such samples is largely undetermined , as projection , among other effects , may lead to spurious detections .
therefore , assessing the success rate of the identification of real systems at different redshifts is an essential step in determining the possible role , and the best design , of ground - based wide - angle surveys in building a large sample of confirmed clusters over a large redshift range suitable for evolutionary studies . in this contribution , we describe the ongoing effort to establish the yield of real clusters from the eis distant cluster candidate list . to this end
we combine the results obtained from follow - up observations by different groups .
this ongoing work includes optical / infrared imaging and spectroscopic observations , which are used to provide either direct or circumstantial evidence supporting the identification of real systems .
the confirmation of a cluster candidate requires one or more of the following conditions : the identification of concordant redshifts from spectroscopic observations ; the existence of a distinctive red - sequence in optical / infrared color - magnitude diagrams ; the detection of a statistically significant concentration of galaxies with similar colors or photometric redshifts .
the original eis catalog of cluster candidates was constructed by applying a matched - filter algorithm @xcite to galaxy catalogs extracted from the eis @xmath0band survey images @xcite @xcite @xcite .
the survey covers four patches of the sky spread over the right ascension range @xmath17 and the object catalogs reach a limiting magnitude of @xmath18 .
details about these data , the location of the patches and the eis cluster candidate catalog can be found in the original papers @xcite @xcite @xcite .
more recently , most of the area covered by the @xmath16-band survey has been observed in @xmath19 and @xmath20 as part of the public survey being conducted by eis using the wide - field camera now available at the mpg / eso 2.2 m telescope at la silla @xcite @xcite .
a summary of the data currently available , combining these new data with those in the @xmath19 and @xmath20 passbands of the original survey @xcite , is shown in table [ tab : coverage ] .
the table gives , for each eis patch , the total area covered in each passband .
the survey is still ongoing and it is expected to cover roughly 14 square degrees in @xmath21 . the available data provide color information for a significant fraction of the eis cluster candidates .
nearly all candidates already have @xmath22 and over half of them have @xmath21 data . 0.5 cm .*wide - angle survey ( square degrees ) * [ cols="<,>,>,>,>,>",options="header " , ] also available are data obtained by different groups from optical / infrared observations of eis cluster candidates with @xmath23 .
these data were obtained at the ntt @xcite ( infrared observations for 15 clusters ) and the danish 1.5 m telescope on la silla , and the nordic optical telescope @xcite ( @xmath20 and @xmath24 observations for 50 clusters ) .
these observations are still ongoing , which will enable doubling the sample within the next year . even though there is no substitute for direct spectroscopic confirmation of candidate cluster of galaxies ,
these observations are extremely time consuming .
in fact , for this type of work it is not only important to have a high - confidence that the selected candidates are real associations but also to be able to select individual galaxies likely to be cluster members .
one way of pre - selecting candidates is to use the available multi - color data to search for galaxies with red colors corresponding to the cluster early - type galaxy population .
such red - sequences have been detected over a broad redshift range using optical and optical / infrared colors for low and high redshift clusters , respectively @xcite @xcite .
as shown below , from the data we have , optical colors are useful for @xmath25 , while optical / infrared data are required to extend the redshift range beyond this limit . in order to find supporting evidence that the eis candidates are physical associations , those with available data in the @xmath20 and @xmath16 filters have been analyzed to search for the presence of a red - sequence .
this has been done either by direct inspection of the @xmath26 color - magnitude diagrams where , at least for clusters with @xmath27 , a red - sequence appears as a distinctive feature@xcite , or by evaluating the statistical significance of spatial concentrations of galaxies in different color slices .
the significance of a detection is determined using mock catalogs containing the same number of objects as the data , as described in more detail in forthcoming papers @xcite @xcite .
considering the 105 detections at the 95% confidence level obtained from the color - slice analysis we show , in the left panel of figure [ fig : vizmf ] , the relation between the @xmath1 color characterizing the detected concentrations and the redshift .
for the eis candidates we take the redshift to be @xmath28 as given by the matched - filter . for @xmath25
we show the median and the quartiles of the color distribution of the detected red galaxy concentrations . at higher redshifts ,
the error bars represent the full range of colors covered by the detections using deeper @xmath20 data .
also shown are curves corresponding to the expected variation of the @xmath29 ) color with redshift for a non - evolving elliptical galaxy and a passively evolving one , assuming a formation epoch at @xmath30 , a single burst of star formation , and a subsequent evolution in a low - density cosmological model .
the colors of early - type galaxies along the red - sequence of spectroscopically confirmed clusters are also shown @xcite @xcite . from the figure we find that is by and large a fair indicator of the cluster redshift , albeit the large scatter , predicting colors consistent with the models and empirical data .
the scatter in color is to a large extent due to the uncertainties in the detection procedure , errors in the galaxy colors and the uncertainties associated with .
the consistency between the medians of the eis candidate colors , the models and the empirical data up to @xmath25 is remarkable . for larger redshifts most cluster candidates are bluer than expected for their estimated redshift . from the visual inspection of these candidates ( in different passbands )
we find that the blue colors are either due to contamination by a foreground cluster or they simply reflect the fact that galaxies in high - redshift clusters are not detected in the @xmath20-band .
these results show that the effectiveness of the @xmath1 color in confirming cluster candidates is in practice limited to systems with @xmath25 , with deeper @xmath20 data providing little additional leverage .
we have also examined the cm - diagrams based on the available optical / infrared data for 15 clusters with @xmath31 . from the analysis of @xmath9 color slices ,
significant overdensities are detected near the nominal center of the candidates in 10 of the observed fields .
the same systems are also identified in @xmath32 , but the weak dependence of this color on the redshift yields no independent constraint on the redshift . for the @xmath395% confidence level detections ,
the colors range from about 2.4 to 3.6 for @xmath9 and 1.6 to 2.1 in @xmath32 .
the dependence of the @xmath9 color of cluster galaxies on redshift is shown in the right panel of figure [ fig : vizmf ] .
for comparison we also show the expected dependence of the @xmath9 color of ellipticals on redshift , for the same models presented earlier , and the color - redshift relation for clusters observed spectroscopically @xcite @xcite . given the small number of candidates considered ,
we show the color of each detected candidate .
the number in parenthesis indicates cases of overlap .
while for low redshifts this color is a poor redshift indicator , it does give some leverage for redshifts @xmath33 .
note , in particular , that the colors of the eis candidates are at least consistent with their estimated redshifts and the general trend of the theoretical models , with a few cases having colors consistent with @xmath15 .
an alternative approach for testing the reality of candidate clusters is to resort to photometric redshifts .
the main advantages are that the method makes optimal use of the multi - color data and that it enables one to consider not only early - type galaxies but the whole range of seds .
tests carried out with eis - deep data have demonstrated @xcite that the minimal set of filters required to obtain reasonable estimates of redshifts , taking into account the redshift range of interest and the limiting magnitudes in the various colors , is composed of @xmath34 .
currently , we have such a five - band coverage for 14 clusters .
photometric redshifts have been assigned@xcite@xcite , for all galaxies brighter than @xmath35 within a @xmath36 arcmin field of the sofi detector .
the redshift distribution in the field of one of the candidates considered is shown in figure [ fig : map ] , where it is compared to the background redshift distribution as measured from all the available fields . as can be seen the background distribution peaks at @xmath37 , while for the field considered the redshift distribution peaks at @xmath15 .
to evaluate the significance of these peaks we have resorted to the same method employed above in analyzing color slices , but considering redshift slices .
we find that in six out of the seven cases analyzed so far a concentration is detected .
figure [ fig : map ] illustrates the distribution of galaxies brighter than @xmath35 in photometric redshift slices 0.2 in width , for one of the cases considered .
an overdensity of galaxies is clearly seen in the redshift interval 0.9 - 1.1 , which coincides with the position obtained by employing the matched - filter algorithm which , by the way , also predicts @xmath38 .
more importantly , from these redshift estimates individual galaxies can be selected for spectroscopic observations over a much larger area than that covered by early - type galaxies detected using a single color criterion .
the spectroscopic data are from observations conducted at the 3.6 m telescope at la silla using the efosc2 spectrograph @xcite @xcite and from a single 3 hours test exposure using the 2df spectrograph at aat @xcite .
the efosc2 observations yielded 69 redshifts in the field of seven eis cluster candidates selected in the redshift interval @xmath39 .
the observations conducted at the 2df cover the vicinity of 58 cluster candidates in patch d , nearly all of the clusters over about two - thirds of the total area of the @xmath16-band survey , irrespective of their estimated redshift .
a total of 340 galaxies were observed yielding 197 redshifts .
a likelihood analysis of the redshift distribution of the galaxies observed with efosc2 was performed , based on a comparison with results obtained from random re - sampling of the cfrs redshift distribution .
this analysis has shown that four of the seven systems correspond to a real system in redshift space with a probability @xmath40% @xcite .
two of them have a mean redshift in agreement with @xmath28 , while the other two have significantly smaller redshifts . however , it is important to note that the redshift distribution in each cluster field is under - sampled and the identification of bound systems from the redshift distribution alone may not suffice for a conclusive determination about the nature of the system and its redshift . in this case
it is important to take into consideration other constraints such as the detection of overdensities in color slices , the spatial distribution of galaxies with measured redshifts and their brightness .
therefore , we supplement the above spectroscopically confirmed sample with systems for which an overdensity of red galaxies is detected and at least one galaxy along this red - sequence has a measured redshift . when these cases are considered three additional clusters are identified with redshifts much closer to those estimated by the matched - filter . in one case , even though only one redshift is available , it corresponds to the brightest galaxy in the field .
moreover , this galaxy lies on a clearly visible red - sequence , consistent with the measured redshift ( @xmath41 ) @xcite .
furthermore , in one case the low redshift assigned by the likelihood analysis is very likely due to the superposition of a nearby system , since one of the measured redshifts is consistent with that inferred from the color analysis and with that estimated by the matched - filter .
when combined , all the information available suggests that all seven clusters can be confirmed and , with exception of one , all observed candidates have @xmath12 . clearly , denser sampling of these fields is required to resolve some of these ambiguities . while no detailed analysis has been carried out of the accumulated 2df data , the completeness of this preliminary survey as a function of magnitude immediately indicates that only clusters with @xmath27 can be identified from the redshifts . from a preliminary inspection of the redshift distribution , identifying cases with two or more concordant redshifts and , as discussed above , by examining measured redshifts and color detections
, we find that for 22 candidates , corresponding to 60% of those with @xmath42 , it is possible to tentatively assign a spectroscopic redshift to the cluster .
combining all the available spectroscopic data we can assess how well the redshift estimated by the matched - filter reflects the true redshift . in figure
[ fig : zspeczmf ] , we compare the median of the measured redshift distribution to the original .
error bars again represent the full range of the distributions .
the results show that the matched - filter redshift is indeed a fair indicator of the cluster redshift at least out @xmath43 . on average
it overestimates the redshift by @xmath44 , comparable to the observed scatter . even though these preliminary results are admittedly based on a small sample they are encouraging since most of the candidate clusters observed
were drawn directly from the matched - filter list without any prior information .
in this paper we have investigated the reality of the @xmath16-band selected eis cluster candidates over an extended redshift baseline .
our main findings are : 1 . for most cluster candidates with @xmath45
, red - sequences can be clearly seen in the @xmath26 cm diagram leaving little doubt that they are real clusters .
the success rate of detecting concentrations of red galaxies is @xmath46 70% out to @xmath47 0.8 , dropping to about 40% at @xmath47 1 .
there is a clear correlation between colors and the redshift estimated by the matched - filter algorithm .
this correlation is seen up to @xmath48 when the @xmath1 color is used and up to @xmath49 when @xmath9 is considered .
the color - redshift relation is consistent with models of stellar population evolution and/or the color measured for elliptical galaxies in clusters with measured redshifts .
preliminary spectroscopic observations show that the measured redshifts are consistent with those estimated by the matched - filter algorithm with the estimated success rate from spectroscopy at @xmath50 and from imaging for larger redshifts we have demonstrated that large samples of clusters are within reach comprising low ( @xmath51 ) to very high - redshift clusters ( @xmath15 )
. such a sample would be ideal for evolutionary studies of galaxies and large - scale structures .
the eis cluster catalog being compiled will also be of great value for many other applications and in support to vlt observations as well as targets for follow - up observations with chandra and xmm .
there is currently no comparable sample in the southern hemisphere . | candidate clusters of galaxies drawn from the sample identified from the moderately deep @xmath0band data of the eso imaging survey ( eis ) , have been used for follow - up optical / infrared imaging and spectroscopic observations .
the observations were conducted to assess the nature of these candidates over a large range of redshifts .
currently , 163 eis candidates have @xmath1 colors , 15 have @xmath2 and 65 cluster fields have been observed spectroscopically . from a preliminary analysis of these data , we find that @xmath3 65% of the candidates studied show strong evidence of being real physical associations , over the redshift range @xmath4 .
the evidence in some cases comes directly from spectroscopic measurements , in others indirectly from the detection of overdensities of objects with either the same color or the same photometric redshift , or from a combination of color and spectroscopic information .
preliminary results also suggest that the redshift derived from the matched - filter algorithm is a reasonable measure of the cluster s redshift , possibly overestimating it by @xmath5 , at least for systems at @xmath6 .
overdensities of red objects have been detected in over 100 candidates , 38 of which with estimated redshifts @xmath7 , and six candidates in the interval @xmath8 have either been identified directly from measured redshifts or have been confirmed by the measurement of at least one redshift for galaxies located along a red - sequence typical of cluster early - type galaxies .
lastly , five candidates among those already observed in the infrared have @xmath9 colors consistent with them being in the redshift interval @xmath10 .
the sample of `` confirmed '' clusters , already the largest of its kind in the southern hemisphere , will be further enlarged by ongoing observations .
1@xmath11 astr .
. # 1 |
kh 15d , an extraordinary pre - main - sequence eclipsing binary system , is located in the young ( 24 myr ) open cluster ngc 2264 ( @xmath11 @xmath12 760 pc ) . in 1970 , a russian circular announced the discovery of eight new variable stars in ngc 2264 ( badalian & erastova 1970 ) .
kh 15d was among these , and was given the name svs 1723 .
it was found to vary by 1.1 mag and was classified as an irregular variable . in 1972 , the new variables discovered by badalian & erastova ( 1970 ) were included in the 58th name - list of variable stars ( kukarkin et al .
1972 ) , and kh 15d was given the name v582 mon .
prior to october 2000 , kh 15d had been observed in various surveys of the open cluster ngc 2264 .
flaccomio et al .
( 1999 ) and park et al .
( 2000 ) both observed kh 15d at least once while the system was bright and included it in their catalogs , each identifying it as # 391 and # 150 , respectively .
earlier surveys of ngc 2264 ( herbig 1954 ; walker 1956 ; vasilevskis , sanders , & balz 1965 ) did not identify kh 15d as a cluster member most likely because of its proximity to the b2 iii star , hd 47887 ( see fig .
its light was simply lost in the bright glow of the b2 iii star .
this object gained its notoriety as kh 15d when it was reported to have unusual properties in 1995 based on a project at the van vleck observatory ( vvo ) to photometrically monitor star - forming regions ( kearns & herbst 1998 ) .
this object was observed to undergo eclipses that were remarkably deep ( @xmath133.5 mag in @xmath14 ) and unusually long ( @xmath1316 days in 1999/2000 ) .
of central interest was the evolution of the shape and duration of the eclipses that had been observed over a 5-year period ( hamilton et al .
2001 ) . out of eclipse ,
the observable star in this system was determined to have a spectral type of k6/k7 ( hamilton et al .
2001 ; agol et al .
2004 ) , while during eclipse , the observed spectrum was simply an attenuated version of that out of eclipse .
an international campaign to monitor this intriguing system was begun during july 2001 .
the primary goal of this project was to obtain much information about the structure of the intervening material . continued photometric observations from around the world provided nearly complete coverage of three consecutive eclipses during the 2001/2002 observing season ( herbst et al .
these consecutive eclipses demonstrated that , in addition to the yearly changes observed in the phased data , differences in the shape of each individual eclipse could be observed over the course of a season .
the shapes of ingress and egress were successfully modeled by the steady advance or retreat of a knife edge " across a limb - darkened star . in order to match the duration of ingress and egress , however , herbst et al .
( 2002 ) determined that the occulting edge must be inclined by @xmath1315@xmath8 to the motion of the star .
these authors also reported that the object s color becomes slightly bluer during eclipse , which indicates that most ( if not all ) of the light received during eclipse is due to scattering .
polarization measurements made by agol et al .
( 2004 ) and gary schmidt ( private communication ) show that out of eclipse , the system exhibits low polarization consistent with zero . during eclipse , however , the polarization is observed to increase to roughly 2% across the optical spectrum .
these results support the conclusion that the star is likely completely eclipsed so that the flux during eclipse is due entirely to scattered radiation , and that the obscuring material is most likely made up of relatively large grains ( @xmath1310 @xmath15 m ; agol et al .
2004 ) .
high - resolution spectra obtained during the december 2001 eclipse ( hamilton et al .
2003 ) revealed that the system is still undergoing accretion and driving a bipolar outflow .
recently , extended molecular hydrogen emission near 2 @xmath15 m was observed spectroscopically by deming , charbonneau , & harrington ( 2004 ) .
given the spatial extent of the emission , and the observed h@xmath16 line profile , deming et al .
( 2004 ) suggest that the ambient h@xmath16 gas is being shocked by a bipolar outflow from the star and/or disk .
further evidence for an extended and well - collimated outflow was provided by tokunaga et al .
( 2004 ) , who obtained broad - band and narrow - band infrared images of kh 15d , showing a jet - like h@xmath16 emission filament that extends out to @xmath17 from the object .
the position angle ( @xmath1320 ) of the in - eclipse polarization is nearly parallel to this h@xmath16 filament ( agol et al .
2004 ) .
many hypotheses have been proposed since 2001 to explain the evolving eclipses of kh 15d .
most of these involve circumstellar material in some way .
hamilton et al .
( 2001 ) , herbst et al .
( 2002 ) , winn et al .
( 2003 ) , and agol et al .
( 2004 ) all proposed the idea of an edge - on circumstellar disk with a warp that periodically passes in front of the star .
barge & viton ( 2003 ) proposed that the eclipses were caused by an orbiting vortex of solid particles , and grinin & tambotseva ( 2002 ) suggested that an asymmetric common envelope in a binary system could accurately reproduce the eclipses .
historical studies by winn et al .
( 2003 ) found that observations made between 1913 and 1950 were consistent with no eclipses .
johnson & winn ( 2004 , hereafter jw04 ) discovered that between 1967 and 1982 , the system alternated between bright and faint states with the same period as observed today , but 180@xmath8 out of phase with the current eclipses .
additionally , the eclipse depth was much smaller ( only 0.67 @xmath0 0.07 mag in @xmath14 instead of 3.5 mag ) , and when out of eclipse , the system was brighter by 0.9 @xmath0 0.15 mag than it is today ( jw04 ) . while herbst et al .
( 2002 ) first suggested the possibility of a binary companion , the light curve between 1967 and 1982 led jw04 to believe that the light from a second brighter star once contributed to the flux coming from this system , and that the second star is now completely obscured .
these results have been confirmed and extended by data obtained at various observatories between 1954 and 1997 ( johnson et al .
2005 ) .
models of the kh 15d system that explained the historical and modern - day light curves were put forward almost simultaneously by winn et al .
( 2004 ) and chiang & murray - clay ( 2004 ) .
each model involves an eccentric pre - main - sequence binary whose motion periodically carries it behind a precessing circumbinary disk .
these models are supported by the results of a recent high - resolution spectroscopic monitoring program of kh 15d : johnson et al .
( 2004 ) have observed radial - velocity variations that are consistent with a binary companion with an orbital period in agreement with the 48-day photometric period . in summary ,
kh 15d is now known to be a pre - main - sequence binary system with a strongly eccentric orbit of 48.38 days .
the brighter component b is currently totally obscured by the circumbinary ring or disk , while only the fainter member a emerges from behind the obscuring material for just less than half the period . to explain the secular variations observed in the light curve over the past 9 years , it has been suggested that the eclipses occur whenever the motion of a star carries it behind the ring , with a period equal to that of the binary orbital period .
if this ring is also precessing , it would be possible to explain the changing length of the eclipse .
in this paper we present the analysis of photometric data on kh 15d obtained at vvo between 1995 and 2000 , color data obtained at the cerro tololo inter - american observatory ( ctio ) during the 2001/2002 season , and the results of the 20022004 observing campaigns as compared to those of 2001/2002 .
section 2 describes the principal sources of optical photometric observations , the data reduction , and the resulting photometry for each participating observatory in the 20022004 campaigns .
a comparison of individual data sets is shown in section 3 , while the results and analysis of these data are presented in section 4 . a brief discussion of our findings in light of the recent models
is also presented , while a more complete interpretation is deferred to a future paper ( winn et al .
2005 , in preparation ) .
the principal sources of optical photometric data analyzed here are telescopes of 0.62.2 m aperture .
these include the us naval observatory s ( usno ) flagstaff observing station in arizona , mount maidanak observatory ( mmo ) in uzbekistan , vvo in connecticut , the european southern observatory ( eso ) and ctio in chile , kitt peak national observatory ( kpno ) in arizona , wise observatory in israel , teide observatory in the canary islands , and konkoly observatory in hungary .
we also obtained data with automated telescopes at the tenagra observatory in nogales , arizona , and the katzman automatic imaging telescope ( kait ) , located at lick observatory atop mt .
hamilton in california .
while attempts were made to encourage uniformity in observing and reduction procedures within the initial group of participating observatories during the 2001/2002 season , the excitement that this object ignited inspired observers at telescopes around the world to contribute data to this project during the subsequent seasons . variations in seeing , image quality , sky brightness , flat - fielding procedures , and ccd characteristics made it impossible to enforce substantial uniformity .
although the photometry parameters varied widely between data sets for the 20022004 observing seasons , which are discussed in detail here , we compared data obtained at different observatories on the same nights to search for consistency , as described below .
it should also be noted that not all observatories used the same reference stars in their data reduction .
we begin by listing the contributing observatories , telescopes , ccd parameters , filters , and dates in table 1 .
( ut dates are used throughout this paper . )
most observations were made with a cousins @xmath14 filter to reduce the nebular contamination of ngc 2264 , as well as the contribution from the nearby b2 iii star , hd 47887 , which is @xmath18 away . here
we will discuss the 20022004 data in detail .
exposure times varied per observatory according to telescope and ccd pixel size .
observatories with similar observing procedures and reduction / photometry techniques are grouped together .
photometry parameters for each dataset from 20022004 are included in table 2 .
attempts were made to observe kh 15d nearly every clear night during the 2002/2003 observing season at the usno , flagstaff station .
@xmath19 , @xmath20 , @xmath21 , and @xmath14 filters were used while kh 15d was bright , and only @xmath20 , @xmath21 , and @xmath14 during eclipse .
in 2003/2004 , multiple observations were obtained during three consecutive egresses using only the @xmath20 , @xmath21 , and @xmath14 filters .
exposure times generally ranged from 2 to 3 minutes , depending on whether the object was bright or faint .
all frames were bias - subtracted and flat - fielded in real time by local code .
photometry was performed on individual frames with an aperture radius that depended on the full width at half - maximum ( fwhm ) of unblended stars in the image . to maximize the signal - to - noise ratio ( s / n ) , a radius of @xmath22 was employed .
the background parameters were set such that the scattered light from hd 47887 was minimized ( see table 2 ) .
differential photometry of kh 15d was performed using seven reference stars selected from all - sky photometry obtained with the 1.0 m telescope at the usno , flagstaff station .
a finding chart for the 7 comparison stars is shown in figure 1 , and the adopted standard johnson @xmath23 and kron - cousins @xmath24 magnitudes for each comparison star is given in table 3 . to determine the magnitude for each frame ( @xmath19 , @xmath20 , @xmath21 , and @xmath14 ) , the @xmath25 magnitude difference was calculated for each non - saturated standard star .
this difference was then applied to the @xmath26 magnitude of kh 15d .
no color terms were used since the usno filters are well matched to the standard system and the local standards span the colors of kh 15d .
the resultant uncertainty includes the spread in the @xmath27 due to not including a color term .
kh 15d was generally observed once every clear night at vvo , tenagra , and kait .
typically , five 1-minute exposures were obtained per night ; however , during ingress and egress , vvo increased the number of exposures .
all images obtained at vvo and tenagra were bias - subtracted , corrected for dark current , and flat - fielded using standard iraf tasks ( ccdproc ) .
images obtained at kait were bias - subtracted , corrected for dark current , and flat - fielded automatically using a local code ( li et al .
2003 ) . after each image was processed , sets of five were averaged to increase the s / n , and the result was aligned to a single reference image using an idl ( interactive data language ) code .
the fwhm for stars in each image was measured using the iraf task imexamine , and was used to assess the seeing on that night . any image that exhibited a very large fwhm ( @xmath13 4.5@xmath28 or larger )
was not used . for the analysis performed here , the average fwhm for the observing season was calculated and used to determine the size of the aperture radius utilized in the photometry process .
this method was chosen so that all the data from one season could be processed and analyzed most simply in batch mode .
we find that the standard deviation from this mean is typically less than 0.5 pix . to maximize the s / n , a radius of @xmath29 was chosen .
instrumental magnitudes and uncertainties were computed for all the local comparison stars ( see fig . 1 ) and kh 15d using the iraf task phot .
calibration of the instrumental magnitudes from one observation to the next was accomplished by means of a set of comparison stars , which were averaged together to produce a synthetic stable comparison . however , since even these stars are variable at some level , the final set of stars used were identified by an iterative process , and represent the most stable ones , with a typical standard deviation of 0.005 mag based on their night - to - night scatter .
stars a , c , e , and g ( see fig .
1 ) were found to be most stable and were used as the comparison stars for both the 2002/2003 and 2003/2004 seasons at vvo . due to saturation issues , stars a , c , e , and f were used for the tenagra images , while only stars c , e , and f could be used for the kait data due to the small size of the detector . a time series of differential magnitudes relative to the synthetic comparison star was then produced . to place kh 15d on the standard magnitude scale ,
the adopted @xmath14-band magnitudes ( converted to flux ) of each comparison star were averaged to produce an @xmath30 comparison magnitude .
this was then added to the differential magnitudes calculated for kh 15d during the 2002/2003 and 2003/2004 observing seasons .
kh 15d was observed multiple times on nearly every clear night at mmo ( uzbekistan ) during the 20022004 observing seasons .
exposure times typically ranged from 2 to 5 minutes , and all observations were made with a bessell @xmath14 filter .
raw frames were bias - subtracted and flat - fielded using the imred package within the iraf environment .
sky flat - field images were usually taken for each night , but in some cases only dome flats were available .
differential magnitudes for kh 15d were computed in reference to star 16d from kearns et al .
( 1997 ) . in order to place these differential magnitudes onto our standard scale
, we needed to know the true magnitude of 16d .
thus , 16d was photometered during the 20012004 seasons at vvo in addition to the 7 local standards and kh 15d ( see section 2.2.2 ) , and the weighted average flux of 16d for the season was converted to a magnitude .
this magnitude was then added to the differential magnitudes computed for 15d at mmo .
the weighted average @xmath14 magnitudes computed for 16d between 2001 and 2004 are given in table 4 .
since error bars were not produced when computing the differential magnitudes , we could not assign formal photometric uncertainties to these data .
kh 15d was observed multiple times on nearly every clear night at konkoly observatory during the 2002/2003 observing seasons .
exposure times ranged from 2 to 20 minutes , but were typically 5 minutes .
all observations were made with a cousins @xmath14 filter .
raw frames were bias - subtracted and flat - fielded using the imred package within the iraf environment .
sky flat - field images were usually taken for each night , but in some cases only dome flats were available .
the differential magnitudes of kh 15d were measured in reference to local secondary standard stars c , d , and f ( see fig .
1 ) . to determine the raw magnitudes of stars in the frames ,
aperture photometry was performed using the iraf / daophot package ( stetson 1990 ) .
the fwhm of the stellar profile was determined for each night and an aperture of @xmath22 was chosen .
the ring from which we estimated the background contribution to the measured fluxes started 2 pixels from our aperture and had a width of 5 pixels in every case .
the differential magnitudes were calculated by defining the variable s brightness as @xmath31 .
these differential magnitudes ( @xmath32 ) were transformed onto the standard photometric system by taking the adopted @xmath14 magnitudes ( converted to flux ) of each comparison star and averaging them together to produce an @xmath30 comparison magnitude .
this was then added to @xmath32 to get the true @xmath14 magnitude of kh 15d .
no color term was applied since the konkoly @xmath14-band filter is well matched to the standard system .
finally , three cousins @xmath14 magnitudes of kh 15d were calculated on each frame by adding the magnitudes of the comparison stars to the differential magnitudes .
we calculated the brightness by averaging the three different fluxes determined for kh 15d .
the uncertainty of the brightness of kh 15d was estimated from the standard deviation of the three individual magnitudes .
kh 15d was observed multiple times every clear night during the 2002/2003 observing season at the teide observatory .
exposure times varied from 5 to 20 minutes , and all observations were made with a cousins @xmath14 filter .
raw frames were bias - subtracted and flat - fielded using the imred package within the iraf environment .
sky flat - field images were usually taken for each night , but in some cases only dome flats were available .
differential photometry of kh 15d was performed using an aperture of @xmath33 .
the sky background was computed using a circular annulus with an inner radius of @xmath34 and a width of 8 pixels .
differential photometry and calibration from instrumental magnitudes to real ones was accomplished with respect to the reference stars 16d , 17d , 27d , 31d , and 39d from kearns et al .
( 1997 ) .
data obtained at vvo between the years of 1995 and 2000 were re - examined with parameters that were consistent with those of the 20012004 observing seasons .
this was done so that an absolute comparison between the earliest data and the most recent data could be made .
one difficulty in completing this task was the fact that from the fall of 1995 through the spring of 1998 , vvo used a chip that was only 512 @xmath35 512 pixels in size .
therefore , only one of the current `` standard '' stars ( d ) was available , and it also happened to be the most variable star .
therefore , we consulted the original photometric studies of ngc 2264 conducted at vvo ( kearns et al .
1997 ; kearns & herbst 1998 ) and examined 5 of their quoted comparison stars ( 16d , 17d , 21d , 27d , and 31d ) . since these stars did not have known magnitudes , each was photometered on the images taken at vvo during the 20012004 observing seasons , allowing us to determine their magnitude on our standard scale .
table 4 lists the weighted average @xmath14 magnitude and uncertainties for each star during 20012004 .
once a standard magnitude was known for each comparison star , all 5 stars and kh 15d were photometered on the images obtained at vvo between 1995 and 1998 .
table 5 lists the comparison stars used for each season .
the eso observations of kh 15d contributed here were the byproduct of a larger photometric monitoring program of ngc 2264 ( lamm et al .
full details describing the observing procedure , reduction methods , and photometry parameters can be found in lamm et al .
. in summary , monitoring of ngc 2264 , and hence of kh 15d , took place during a period of two months between 30 december 2000 and 01 march 2001 with the wide field imager ( wfi ) on the mpg / eso 2.2 m telescope on la silla ( chile ) .
all observations were carried out with a cousins @xmath14-band filter .
the wfi consists of a mosaic of 4 @xmath35 2 ccds with a total array size of 8k @xmath35 8k .
exposures of 5 s , 50 s , and 500 s were obtained , but the data used in this contribution come only from the 500 s images .
image processing was done separately for each of the individual wfi chips using standard iraf tasks .
each image was bias - subtracted using the overscan region in each frame , and flattened using an illumination - corrected dome flat .
the data presented here are based on differential photometry relative to a set of non - variable reference stars .
the daophot / apphot task was used to measure the brightness of kh 15d .
the aperture was chosen to be 8 pixels ( 1.9@xmath28 ) in diameter for all measurements in order to maximize the s / n .
the sky was calculated as the median of an annulus with an inner diameter of 30 pixels ( 7.1@xmath28 ) and a width of 8 pixels centered on kh 15d . because only differential magnitudes were calculated for the eso data , we computed the average out - of - eclipse magnitude for the 2000/2001 season based only on the data obtained at vvo .
an appropriate out - of - eclipse magnitude was then added to the eso differential magnitudes such that these data could be placed on our standard scale .
kh 15d was observed extensively during the 2001/2002 observing season with the ctio / yale 1 m telescope .
all observations were made with a single instrument ( andicam ; see depoy et al .
2003 for a description ) using johnson @xmath20 and @xmath14 filters .
kh 15d was observed on nearly every usable night between 30 august 2001 and 18 april 2002 .
typically , several images were obtained each night using 300 s exposures .
the data were reduced in the usual manner using standard iraf routines : an overscan ( bias ) was subtracted and a flat - field was applied to all the images .
kh 15d was measured relative to two nearby stars shown in figure 2 . labeled as 1 " and 2 , " they are also known as 5d and 25d from kearns et al .
( 1997 ) , with @xmath14 = 13.250 @xmath0 0.001 mag and @xmath14 = 16.464 @xmath0 0.005 mag , respectively .
we find that neither comparison star varied by more than 0.005 mag over the duration of the observations of kh 15d .
a 3.6@xmath28 diameter aperture was synthesized on each of the ccd images for both of the reference stars and kh 15d using standard iraf aperture photometry routines .
the sky was determined in a 4@xmath28 to 6@xmath28 annulus immediately adjacent to each of the sources .
we calibrated the measurements of kh 15d relative to the two stars using images obtained at the 1.3-m mcgraw - hill telescope of the michigan - dartmouth - mit ( mdm ) observatory .
the data were obtained on the photometric night of 19 october 2003 .
the scale for the 1024 @xmath35 1024 pixel ccd that we used is 0.5@xmath28 pixel@xmath4 . throughout the night we observed photometric standard stars taken from landolt ( 1992 ) at a range of airmasses . near the end of this night we also obtained 12 @xmath20-band and 7 @xmath14-band images of the kh 15d field .
we reduced the data using standard iraf techniques , including bias - subtraction and flat - fielding using twilight - sky flats obtained during the evening twilight of the same night .
we performed aperture photometry on the mcgraw - hill data , using a 20 pixel ( 10@xmath28 ) diameter aperture for both the standard stars and the target stars in the kh 15d field .
the seeing throughout the night was roughly 1.4@xmath28 , so aperture effects should be small .
a photometric solution was found using all 36 standard - star measurements in @xmath20 and @xmath14 .
we solved only for the zero point and extinction coefficient terms and held the color term constant at zero .
this solution was then applied to the kh 15d field stars , and the calibration was used to convert the relative photometry to the standard system .
we estimate that ignoring the color term in our calibration observations and between the mdm and the ctio / yale observations gives a systematic error in the kh 15d light curve of @xmath360.02 mag .
this is based on intercomparison of other standard - star measurements .
this paper combines a number of data sets obtained at various observatories , analyzed by different means , as described in the preceding sections . before beginning the process of analyzing the results of these data ,
random and systematic errors between data sets were assessed . to accomplish this , we selected all the photometry obtained at different places on the same julian date . if there were multiple observations made at a given observatory , a nightly mean was calculated .
this nightly mean was then used in the analysis described below . out - of - eclipse and
in - eclipse observations were analyzed separately since the photometric errors are typically much larger during eclipse owing to the faintness of the object .
additionally , times of ingress and egress were avoided since the magnitude of the system changes too rapidly for comparison of data , even as close as one day apart .
each season is addressed separately .
one of the main objectives of the observing campaign of 2001/2002 was to study , in detail , the eclipse of december 2001 .
the original contributing observatories included wise , vvo , mmo , usno , and kpno , and the results were reported by herbst et al .
we now have data from the ctio / yale 1 m telescope that has provided us with perhaps the highest sampling rate yet of the eclipses that occurred during the 2001/2002 season .
these are shown in figure 3 .
the central reversal is evident in these data , as well as some variation in the shape of ingress / egress from eclipse to eclipse .
additionally , substantial variability is seen outside of eclipse , as explored further in section 4.1.2 .
since these data were originally obtained through johnson @xmath20 and @xmath14 filters , we could not immediately combine them with the rest of our cousins @xmath14-band data .
the coverage of nearly 5 eclipses , including dense sampling throughout ingress and egress , provided excellent color information about this system .
we therefore examined the johnson colors separately from the other data sets and present the results in section 4 .
however , in order to add these johnson data to the full @xmath14-band light curve , we computed a transformation from johnson @xmath14 to cousins @xmath14 by comparing nearly simultaneous johnson @xmath20-band and cousins @xmath14-band data that were obtained at the usno ( see next section ) . in 2002/2003
, seven different observatories contributed to the light curve of kh 15d ( see table 1 ) . to evaluate the systematic differences between each contribution , we began by selecting all the data that were obtained on the same julian dates outside of eclipse , for the entire observing season .
we define out - of - eclipse " to correspond to phases between 0.3 and 0.7 for both the 2002/2003 and 2003/2004 seasons .
the out - of - eclipse 2002/2003 data are shown in figure 4 . for each
julian date , the mean brightness of the star was formed based on all measurements .
these contributions might include a single observation from one observatory and a mean observation ( due to multiple exposures throughout the night ) from another .
the difference between each observatory s measurement and the total nightly mean was then calculated and examined as a function of the total mean brightness . a histogram indicating the number of times an observatory s measurement differed from the mean by 0.01 mag , 0.02 mag , etc . is shown in figure 5 . in general , there appears to be good agreement between the data sets .
most observations differ from the mean by @xmath130.01 mag , which is typically the size of the estimated random errors associated with the photometry .
thus , we can say that the data sets are consistent with each other to within @xmath130.02 mag .
variations in the light curve on the order of @xmath130.08 mag are observed out - of - eclipse ( see fig .
4 ) , which we take to be real , and not an artifact of combining the data sets .
this will be explored further in section 4 .
figure 6 shows multiple data obtained on the same julian dates while in deep eclipse during 2002/2003 .
we defined deep eclipse " to correspond to phases between 0.83 and 0.17 for both 2002/2003 and 2003/2004 , with phase = 0.0 ( and 1.0 ) representing the approximate phase for mid - eclipse .
the same analysis described above was performed on these data .
figure 7 shows that the differences of each observatory s observation from the mean is on average @xmath130.15 mag , which is consistent with the photometric observational errors . in 2003/2004
, only four observatories contributed to the light curve of kh 15d ( see table 1 ) .
there were significantly fewer multiple observations out - of - eclipse during 2003/2004 as compared to 2002/2003 .
we compared these individual measurements in the same manner as discussed above .
the agreement between these data sets is not as good as was seen in the previous season . on average ,
the data from tenagra were @xmath130.04 mag brighter than those from vvo , and the data obtained at vvo were @xmath130.02 mag brighter than those obtained at mmo ( although that is not always the case ) .
there are generally only 2 observations contributing to the calculation of the mean on any given night .
therefore , the difference between each observatory s measurement and the mean gives us a sense of the systematic differences between each data set .
the average difference is @xmath130.02 mag , which is slightly larger than the typical photometric errors ( @xmath130.01 mag or smaller ) .
the in - eclipse data from 2003/2004 showed a great deal of scatter , but for the most part , appear to be correlated with each other . on average , the data depart from the calculated mean by @xmath130.15 mag , which is on the order of the quoted photometric errors associated with the in - eclipse measurements .
overall , the data sets from both 2002/2003 and 2003/2004 mesh fairly well , and do not appear to be grossly different from each other .
therefore , we feel confident that the shape of the light curve and the depth of the eclipses are accurate to within @xmath130.02 mag out - of - eclipse and @xmath130.15 mag in deep eclipse .
the full data set now contains 6694 measurements of cousins @xmath14 magnitude covering nine seasons of observation between 1995 october and 2004 march , as well as additional color measurements .
magnitudes in @xmath14 are given in table 6 and colors are given in table 7 .
both tables , in full , are available electronically . in this section
we describe the light and color variations and propose a rotation period for the star .
we also provide some qualitative guidance to the interpretation of the variations based on the models of winn et al .
( 2004 ) and chiang & murray - clay ( 2004 ) .
detailed modeling of the photometric ( and spectroscopic ) data will be done by winn et al .
we turn first to a discussion of the light variations in @xmath14 , where most of the data are available .
figure 8 shows the data from the last two seasons , which is the primary addition to the data set in this paper .
clearly , we have obtained good coverage of a total of nine eclipses over two seasons .
results indicate that the trends noticed by hamilton et al .
( 2001 ) ( reduction in amplitude of the central reversal and widening of the eclipse ) are continuing , and it would appear that the duration of the eclipse is now about one - half of the period ( compare with fig .
the shapes of ingress and egress are not identical and show some level of variation , not only from eclipse to eclipse , but also from year to year .
since 2001 , the duration of the eclipse has been growing by roughly 2 days year@xmath4 .
however , as will be seen in section 4.1.2 , this has not always been the case . in earlier years , such as from 1998/1999 to 1999/2000 , the change in eclipse length was close to 1 day . in figure 9 , the last two seasons are compared with the seven previous ones for which modern , ccd photometry is available . some of the features and a distinct secular variation of the light curve can be seen . here
, it appears that the out - of - eclipse brightness of the star has not changed by more than @xmath130.1 mag since 1995 , while the depth of the eclipse has grown by @xmath131 mag over that same interval .
in fact , it appears as though the eclipse depth has been increasing at a rate of @xmath130.2 mag year@xmath4 . to determine the best estimate of the mean @xmath14 magnitude out - of - eclipse , we looked solely at the data obtained at vvo since it was the only observatory from which we had data for every year .
we began by calculating a mean @xmath14 magnitude per year to see if there was any variation in the out - of - eclipse brightness with time .
the phases that corresponded to the out - of - eclipse state changed as the width of the eclipse evolved with time ( discussed below ) , and are listed in table 8 .
figure 10 shows the mean @xmath14 magnitudes versus observing season at vvo . the straight line represents a simple linear fit to the data .
it appears as if the currently visible star might also be fading a bit , but we can not be sure whether this is a significant trend . as shown below , the increasing depth of the eclipse is accompanied by an increased time duration and a diminishing height for the central brightness reversal .
the current data show that these trends , noticed previously ( see hamilton et al .
2001 ; herbst et al .
2002 ) , have continued in the last two seasons . during the first season of monitoring at vvo
there was one unusually bright data point , taken on only the sixth night that the system was observed .
repeated checks have shown that this is a valid datum and can not be discarded as an error . as will be shown below
, its phase also corresponds to a point close to mid - eclipse .
we now interpret that point as the only observation in our data set during which the relatively unobscured light from the currently invisible star ( b ) was seen directly .
this point occurred close to periastron when star b was peaking above the disk for just about the last time . on that night , the system s light was only due to star b , since star a was hidden behind the disk .
if this interpretation is correct , it means that the magnitude of star b is equal to or brighter than @xmath14 = 14.01 @xmath0 0.01 ( see table 6 ) . since star b
might also have been partly hidden itself , we can not say for sure based on one measurement whether star b is brighter than @xmath14 = 14.01 @xmath0 0.01 , but it certainly can not be fainter . this conclusion is consistent with the historical light curve .
johnson et al .
( 2005 ) show that prior to 1990 , the system was at times as much as 1 mag brighter than it appears today .
if we take the difference between the combined light ( star a + star b ) and the light of star a alone to be @xmath37 mag , then the magnitude difference between star a and b is 0.45 .
the mean @xmath14 magnitude out - of - eclipse based on the vvo data ( see fig . 10 ) shows that star a has a mean @xmath14 = 14.47 @xmath0 0.03 mag . assuming a minimum @xmath14 magnitude of 14.01 @xmath0 0.01 for star b , the mean brightness of star b is 0.46 @xmath0 0.03 mag brighter than star a. to make further progress we wish to phase the light curve with the appropriate period for the system , a task to which we now turn .
it is important to establish the period of this system to the highest accuracy possible so that the data can be properly phased and the evolution of the light curve with time correctly displayed .
it is now known that the 48-day period comes from the orbital motion within the binary . as such
, the most definitive determination of the period should come from radial - velocity measurements and the spectroscopic - binary solution .
this yields a period of 48.38 days with an uncertainty of about 0.01 days ( johnson et al .
the problem with this method is that only a portion of the full orbit of one star can currently be seen , and it is the part near apastron which provides the least leverage on the solution .
a second complication is that the visible star may suffer from a rossiter effect ( worek 1996 ) of unknown size as it enters or emerges from eclipse , depending on the inclination angle of the rotation axis to the line of sight .
therefore , the data that provide the greatest leverage on the radial - velocity solution ( i.e. , those taken during late egress or early ingress ) may also be the data which are compromised in this regard . hence , we believe it is important to ask independently of the radial velocities whether the period of the system can be determined .
the problem with the photometric method is , of course , that the light curve is evolving with time due to the progressive occultation by a foreground screen .
this puts an additional time - dependent behavior into the system , albeit on a much longer time scale than the orbital period . to proceed , we first did a fourier transform of all of the modern photometric data using the scargle periodogram technique ( scargle 1982 ; horne & baliunas 1986 ) .
this yielded a highly significant peak in the periodogram at 48.367 days ( see fig .
11 ) , which we round to 48.37 , as discussed below .
this is gratifyingly close to the spectroscopic period and to the previously reported photometric period of 48.36 days ( herbst et al . 2002 ) .
we attempted to refine the photometric period by focusing on a particular feature in the light curve whose phase we believe is stable and whose location we could determine from season to season .
the only such feature that exists is the peak of the central reversal ( see fig .
12 ) , which we take to be the point in the orbit when the currently unseen star is closest to the edge of the obscuring cloud .
assuming that this is a stable , fixed location in the orbit , if we could phase these central peaks we would have a good estimate of the period .
unfortunately , it turns out that this method lacks discrimination because the location of the central peak is not sufficiently well defined by the light curves , especially the early ones where the data density is too low .
a difference of 0.01 days in period is not significant and does not affect our results , as it corresponds to only 0.75 days over the 75 cycles covered by our monitoring and the early data do not constrain the time of the central peak to better than that .
therefore , we find it impossible to improve on the accuracy of the period determinations with just these data .
it might be possible to do this photometrically using the historical data , but any discussion of that is postponed to a later contribution ( winn et al .
2005 , in prep . ) .
here we simply adopt the result of the periodogram analysis , 48.37 days , noting that it is consistent with the spectroscopic determination ; the 0.01 day difference does not affect any of the analysis or conclusions reported here . during the 20012004 observing seasons , a great deal of effort was made to observe kh 15d out - of - eclipse as well as in eclipse .
significantly more data populated the bright phase of the light curve during these seasons than ever before , and small - scale variations ( @xmath130.08 mag over the course of 814 days in 2002/2003 ) are evident ( see fig . 4 and section 3.1 ) . since the visible star in the kh 15d system is a weak - lined t tauri star ( wtts ; hamilton et al .
2001 ) , it is most likely spotted in some fashion .
it has been determined that this k7 star is still actively accreting ( hamilton et al .
2003 ) , and could therefore exhibit accretion hot spots on its surface , in addition to any dark , magnetically induced star spots .
if this is the case , a closer look at the out - of - eclipse data from the 20012004 seasons may provide an estimate for the rotation period of the star .
the analysis above indicates that systematic errors between data sets are relatively small , being at the level of 0.02 mag or less .
however , when searching for the tiny amplitudes ( less than 0.1 mag ) that may characterize rotationally induced spotted - star variations , even errors of this size may make the search more difficult . therefore , we have searched each observatory s data set independently for periodicity during the out - of - eclipse phases .
it is also prudent to confine the search to a single season of observation since spots are well known to evolve on time scales of less than one year .
hence , one rarely finds phase coherence of spot variations extending over more than one observing season . detecting spots on wttss that are not undergoing
eclipses is difficult enough , and generally only 15% or less of the stars in a cluster field will exhibit coherent periodic variations over a season s observing . in this star
we have greater difficulty because during about half the observing season the star is in eclipse .
nonetheless , we have found that periodogram analyses of the available data sets reveals highly significant periods in three cases ( see fig .
13 ) , as listed in table 9 .
notably , two of these periods are identical to within the errors , averaging to 9.6 days .
the chance of finding two exactly equal , highly significant periods in these data is vanishingly small if the periodicity is not real ; we thus identify the detected period as the likely rotation period of the visible star .
it is interesting that the third significant period found in the periodograms ( 7.95 d ) is very close to the beat period between 9.6 and 48.37 days . supporting evidence for our interpretation comes from the measured @xmath2sin@xmath3 of the star . as discussed in appendix a
, we derive a value of @xmath2sin@xmath3 = 6.9 @xmath0 0.3 km s@xmath4 for the visible component based on high - resolution spectra obtained out of eclipse . since sin@xmath3 is likely to be very close to 1.0 for this star ( assuming that the spin and orbital angular momenta vectors are nearly aligned )
, one can easily calculate that for a radius of @xmath5r@xmath6 ( hamilton et al .
2001 ) , the expected rotation period is 9.6 @xmath0 0.1 days , consistent with our result . both the photometry and spectroscopy of this system out - of - eclipse concur in suggesting a rotation period of 9.6 days for the visible component .
to within the errors , this is a 5:1 resonance with the orbital period ( 48.37/5 = 9.67 ) , although it is not clear whether this is physically significant or just a numerical coincidence .
a rotation period of 9.6 days is somewhat long for a 0.6 m@xmath6 wtts in ngc 2264 , although not unprecedented .
the ( bi)modal values for stars more massive than 0.25 m@xmath6 are around 1 and 4 days .
however , about 12% of the 182 cluster stars in that mass range with known rotational periods have @xmath38 days ( lamm et al .
could the slightly slower than typical rotation of the visible star in the kh 15d system be caused by a tidal interaction between the components ?
the effect of tidal friction on rotation can be estimated quantitatively based on the seminal theory of zahn ( 1977 ) for stars with convective envelopes .
he shows that the synchronization time for a star s rotation to become tidally locked to its revolution is approximately given by @xmath39 where @xmath40 is the mass ratio , @xmath41 is the semimajor axis , @xmath21 is the stellar radius , and @xmath42 is given in years .
most of the torque in a highly eccentric system will naturally occur at periastron . with @xmath43 and @xmath44
, we find a synchronization time of @xmath135 my .
hence , it is not unreasonable to suggest that tidal friction is responsible for some degree of slowing of this wtts .
we further note that since the orbit is rather eccentric , one would not expect to reach synchronous rotation , but only pseudosynchronous rotation .
the pseudosynchronization timescale describes a near synchronization of revolution and rotation around periastron ( hut 1981 ) .
hut ( 1981 ) has shown in elegant fashion that for binaries with eccentricities exceeding @xmath130.3 , as must surely be the case for kh 15d , tidal interaction near periastron is expected to produce an equilibrium angular rotation velocity of about 0.8 times the orbital angular velocity at periastron .
one may write , therefore , that the pseudosynchronous rotation period ( @xmath45 ) predicted by this theory is related to the orbital period ( @xmath46 ) by @xmath47 where @xmath48 is near 0.8 for @xmath49 and is given precisely by hut ( 1981 ) .
the range of particular interest for kh 15d is @xmath50 as discussed by johnson et al .
( 2004 ) , because these solutions are consistent with the radial velocities and with the inferred masses of the components based on astrophysical constraints . over this limited range , and
somewhat beyond it , @xmath48 = 0.81 @xmath0 0.01 . adopting @xmath51 and @xmath46 = 48.37 @xmath0 0.01 yields a rather precise prediction for @xmath1 of 0.65 @xmath0 0.01 .
this is the required orbital eccentricity if the star is in pseudosynchronous rotation and if the theory of hut ( 1981 ) is correct .
remarkably , this eccentricity is very close to those inferred on the basis of two independent methods .
as already noted , the orbital solution based on radial velocities , together with astrophysical constraints on the masses of the components , leads to @xmath52 , just barely outside of the present result .
model fits to the light curve and its evolution over more than fifty years , by winn et al .
( 2004 ) and winn et al .
( 2005 , in prep . ) , also suggest an eccentricity of between 0.55 and 0.7 . since the rotation rate of the visible component is notably slower than that of most stars of comparable mass and age , and
since it agrees so well with the predicted pseudosynchronous period , we suggest that the star has , indeed , been tidally locked into its current configuration . that , in turn , puts a rather severe limit on the possible eccentricity of the orbit , assuming that the locking has reached equilibrium and that the theory of hut ( 1981 ) is accurate in its prediction of @xmath48 .
we note that the strong sensitivity of @xmath45 to @xmath1 means that there is only a very small range of measured rotation periods which would have given consistent results with the orbital solution and light - curve fitting techniques . for comparison , a normal rotation period of around 14 days for this star would have predicted a value of @xmath1 = 0.80.95 for this binary , rather extreme solutions that are inconsistent with the other results .
while most of the monitoring has been in the @xmath14 band due to the brightness of the star during eclipse , we have also obtained some data at other wavelengths which reveal interesting features of the system . here
we discuss the color results obtained at ctio in 2001/2002 and at the usno during several seasons .
the discovery reported by herbst et al .
( 2002 ) that the star is generally bluer when fainter is confirmed , but we now have much more detailed and intriguing information on the color variations and their phase dependence .
extensive @xmath20-band and @xmath10-band observations were made during the 2001/2002 season at ctio with the yale 1 m telescope ( see sections 2.8 and 3.1 for details ) and are shown in figures 14 and 15 .
the solid and dashed lines in figure 14 are provided to indicate the approximate beginning of ingress and ending of egress , respectively .
while the johnson magnitudes can be transformed to cousins for comparison with other data sets , there is no need to do so in this section , so we plot the results in the johnson system .
the high density of color data obtained at ctio during this season reveals an interesting new feature of the behavior of kh 15d : it becomes bluer by a small but significant amount in very steady fashion as it enters eclipse and shows an analogous reddening as it emerges from eclipse .
this is quite easily seen in both the time and phased versions of the light curves ( fig .
14 and 15 , respectively ) .
it is also confirmed by the less dense but more temporally extended color monitoring done at the usno as described in the next section . while the general trend of bluer when fainter had been recognized for this star previously ( herbst et al .
2002 ) and has been seen in other young stars , such as uxors ( e.g. , herbst & shevchenko 1998 ) , the smooth decrease ( increase ) in @xmath53 seen during ingress ( egress ) in figure 15 is remarkable and unprecedented . before discussing its interpretation , we complete a description of the color behavior during eclipse .
as both figures 14 and 15 show , the color change that occurs as the star goes into eclipse does not continue smoothly throughout the eclipse .
while the data are noisy due to the low brightness of the star at those phases , it is clear that the star does not maintain a steady color with phase during eclipse .
our impression of figure 15 is that there are actually three phases when the star has its bluest colors , peaking at about @xmath54 = 1.21.3 mag , or 0.30.4 mag bluer than the out - of - eclipse value of @xmath54 = 1.6 mag .
these phases occur around @xmath00.17 and close to phase 0 .
it is interesting that there is also a distinctive feature in the light curve at each of these three phases . at @xmath00.17 phase ( in 2001/2002 )
there is a distinct change of slope in the decline rate during ingress and the rise rate during egress .
while this distinctive change of slope can be seen in the 2001/2002 light curve , it is even more noticeable in the light curves of the last two years ( see fig . 8) , where it occurs at phases of about @xmath00.20 and @xmath00.24 , respectively .
the other feature in the light curve which seems to correspond to a blue extreme in the colors is the well - known central reversal near phase 0 . before discussing the interpretation of these interesting features
, we turn to a description of the other major set of color data , that obtained at the usno .
@xmath20 , @xmath55 , and @xmath56-band monitoring took place at the usno 1.3 m telescope throughout the 2001/2002 season , while @xmath19 , @xmath20 , @xmath55 , and @xmath56-band monitoring took place at the 1.0 m telescope during 2002/2003 ( see section 2.1 for details ) .
it was determined from the first observing season that the object was simply too faint in @xmath19 to provide any reliable photometry at those wavelengths during eclipse . over the second season of monitoring ,
@xmath19-band measurements were obtained out - of - eclipse , in addition to the standard @xmath20 , @xmath55 , and @xmath56 observations . during 2003/2004 , only @xmath20 ,
@xmath55 , and @xmath56 data were obtained during three consecutive egresses . in figure 16
, we show the behavior of all of the cousins - band colors monitored at the usno during the 2002/2003 season .
the trend is clearly quite similar to what was seen in the ctio data in 2001/2002 namely , there is a distinctive bluing of the system just as it enters eclipse or corresponding reddening as it emerges from eclipse .
in addition , there are clearly some significant color variations out - of - eclipse which we attribute to star - spot activity on the visible star . combining the color data obtained over many seasons at the usno
, we illustrate the overall trend of color with brightness in figure 17 .
this confirms our impression that , out of eclipse , the star is about as red as it gets and shows relatively little color variation beyond what may be explained through star spots . as it enters ( emerges from ) eclipse the star gets bluer ( redder ) by 0.10.2 mag .
near minimum light there are substantial color variations ; the star is sometimes at its bluest extreme , which is 0.30.4 mag bluer in @xmath53 than out - of - eclipse .
however , sometimes it shows very little , if any , color effect . a full explanation of the color data must await improved understanding of the details of this system , which will only come from an extensive phenomenological model currently being constructed ( winn et al .
2005 , in prep ) . however , qualitative explanations for the observed trends can be proposed based on a few simple arguments .
these lead to a set of questions that need to be addressed , and they suggest further observations that may clarify the situation . basically , blue colors indicate either the importance of wavelength - dependent scattering , such as from small grains or molecules , or the presence of a hotter component in the system , or both .
it is likely that there is a hotter component in the system , since the currently invisible companion to the k7 star is known to be more luminous , as discussed above .
all models of pre - main - sequence contraction indicate that the more massive star in a coeval pair is also more luminous and hotter , so it is reasonable to suppose that the currently invisible component of this system is also bluer than the k7 star . however , since the system was never substantially brighter in the past , even when both stars were visible , it is also true that this invisible component can not be much more massive , luminous , or hotter than the k7 star .
in fact , the lack of detectable change of spectral type during minimum means there is little other than k6 or k7 light in the system , as far as a stellar component . the original interpretation proposed for the color variation by herbst et al .
( 2002 ) is that near minimum light , we see the system primarily or only by reflected light and that some small grains are involved in the reflection . this is supported by the increased polarization detected near minimum light by agol et al .
the current data show , however , that there is almost certainly not a single color to which the star moves during eclipse , but rather that there is a great deal of real variability , some or all of which is phase dependent .
therefore , it can not simply be that all of the light of the system is coming from distant , scattered radiation .
in particular , we are impressed by the relative peak in blueness that occurs near phase zero in the color - phase plots ( fig .
15 and 18 ) .
this is the time when the hotter , currently invisible , star is closest to the edge of the occulting cloud .
we propose that the blue peak at this phase may be caused by this fact .
we are apparently seeing a very small amount of either transmitted light or increased importance in the reflected component due to the light of this star , or , perhaps , the extension of a hot , circumstellar nebula associated with it , above the obscuring wall . in any event , we tentatively attribute the central peak in blue color and the phase - dependent variations around it as due to the orbital motion of a slightly bluer component ( than the k7 star ) relative to the edge of the occulting disk and , perhaps , relative to the principal scatterers .
the blue peaks associated with phases @xmath00.17 in 2001/2002 are an entirely new feature of the light of this system revealed by the intensive color monitoring at ctio during that season and deserve some explanation .
we find it interesting , as noted above , that they seem to correlate with a distinct change in slope that occurs in the decline ( rise ) in magnitude during late ingress ( early egress ) .
one problem in interpreting these data is that we do not yet know how sharp the occulting edge of the cloud is .
let us suppose , for the purposes of this qualitative study , that the edge is very sharp . in that case
, we can identify the change of slope in decline ( rise ) rate as the point where the photosphere has just been completely covered ( has first appeared ) .
the fact that the color peaks blue at that point indicates that the blue light is closely associated with the photosphere , arising just above it .
we suggest , therefore , that the light in this system contains a hotter component associated with an extended chromosphere or corona of the k7 star ( and possibly its invisible companion as well ) .
we have no idea whether the physical nature of this is a scaled - up solar - like chromosphere , a magnetically channeled accretion column , or something else .
but the color evidence suggests an extended hotter , dense , optically thick zone located close to the photosphere of the k7 star .
the persistence of the bluing of the light curve shows that it must be extended on a length scale comparable to a stellar radius .
an alternative model is that one is seeing a peculiar scattering effect just as the star enters ( emerges ) from full eclipse .
one could imagine , for example , that strongly forward scattered light might glance off the top of an occulting disk providing a bluish glint " at these phases .
this would require , of course , wavelength - dependent scattering and , therefore , some component other than the obscuring cloud ( which produces no reddening of the transmitted light ) .
if wavelength - dependent reflection is heavily involved here , then there should be some dramatic increases in polarization during these phases
. it would definitely be interesting to extend the polarization study of agol et al .
( 2004 ) to a wider variety of phases .
it will also be important to continue to monitor the color during these phases in future eclipses , hopefully including the @xmath57 band with larger telescopes , which will give important information on the nature of the scattering or the temperature of the emission zone .
the light curve of the remarkable system kh 15d continues to evolve , evidently as a result of an obscuring screen moving across the orbit of a binary .
a periodogram analysis of our data indicates that the photometric period is in excellent agreement with the spectroscopic period .
analysis of the data collected outside of eclipse suggests that the visible star has a rotation period of @xmath139.6 days .
this conclusion is in accord with the predicted rotation period , which is derived from the measured @xmath2sin@xmath3 ( 6.9 km s@xmath4 ) and taking sin@xmath3 @xmath12 1 .
interpretation of the color information leads us to believe that some extended blue emission is revealed in the colors as the currently visible star goes into or comes out of eclipse .
the eclipse length is now greater than half the orbital period of the system and continues to evolve , expanding at a rate that is @xmath132 days year@xmath4
. continued photometric monitoring will help to further constrain the models .
this object will clearly provide us with the opportunity to learn about disk evolution , the circumstellar environment of young stars , and the possibility of planet formation .
thus , we strongly encourage continued monitoring before it disappears !
we would like to thank josh winn for helpful discussions regarding this system .
c. m. h. thanks christopher johns - krull for supporting her during the last year of this project and providing the k7 template used in the determination of @xmath2sin@xmath3 .
w. h. gratefully acknowledges support by the national aeronautics and space administration under grant nag5 - 12502 , issued through the origins of solar systems program .
p. a. , m. k. , a. m. , j. b. , and s. c. are supported by the hungarian national science fund under numbers t043739 and t043504 .
a. v. f. and w. l. are supported by nsf grant ast-0307894 .
a. v. f. is also grateful for a miller research professorship at uc berkeley , during which part of this work was completed . v.
j. s. b. would like to acknowledge prof .
rafael rebolo for informing him of the international campaign to monitor kh 15d and for inviting him to participate .
kait was made possible by generous donations from sun microsystems , inc .
, the hewlett - packard company , autoscope corporation , lick observatory , the national science foundation , the university of california , and the sylvia & jim katzman foundation .
in order to determine the @xmath2sin@xmath3 of the visible star in the kh 15d system , the uves / vlt spectra ( hamilton et al . 2003 ) were revisited .
first , a k7 v model spectrum was produced using a temperature of 4000 k , a log(@xmath58 ) = 3.5 ( @xmath58 in cgs units ) , and a macroturbulence value of 0 .
the template was then artificially broadened to resemble that of a higher velocity star by convolving it with a rotation profile that was generated from a theoretical model , and including a macroturbulence value of 2 km s@xmath4 .
this was accomplished by using an idl code and various values of @xmath2sin@xmath3 believed to span the range within which the @xmath2sin@xmath3 of the visible k7 star is expected to be found .
each broadened spectrum was then cross - correlated against the original narrow - lined template , and the fwhm of the peak of the cross - correlation function was measured . a plot of the fwhm values versus @xmath2sin@xmath3 a calibration curve was produced by fitting a second - order polynomial function to the data .
the 29 nov .
2001 spectrum was then cross - correlated against the narrow - lined template and the fwhm value of the cross - correlation peak was measured .
this value was determined to be 0.489 .
the polynomial fit to the calibration curve was evaluated at this fwhm value to obtain a @xmath2sin@xmath3 = 6.9 @xmath0 0.3 km s@xmath4 for the k7 star .
the uncertainty was determined by first broadening our template spectrum to the value of 6.9 km s@xmath4 .
we then performed a monte carlo test where noise , appropriate to that which is found in the uves / vlt spectrum , was added to the broadened template .
this broadened , noisy template was then cross - correlated against the narrow - lined model spectrum .
the fwhm of the peak of the cross - correlation function was measured for each test , and the standard deviation of these values provided us with our estimated uncertainty .
vvo & 0.6 m & 0.5k @xmath35 0.5k , 0.6@xmath28 pix@xmath4 & @xmath14 & 27 oct 1995 25 mar 1998 + vvo & 0.6 m & 1k @xmath35 1k , 0.6@xmath28 pix@xmath4 & @xmath14 & 10 dec 1998 16 mar 2004 + eso & 2.2 m & 8k @xmath35 8k , 0.238@xmath28 pix@xmath4 & @xmath14 & 30 dec 2000 1 mar 2001 + ctio & 1.0 m & 2k @xmath35 2k , 0.3@xmath28 pix@xmath4 & @xmath59 & 22 aug 2001 18 apr 2002 + mmo & 1.5 m & 2k @xmath35 0.8k , 0.26@xmath28 pix@xmath4 & @xmath14 & 12 sept 2001 18 mar 2004 + usno & 1.3 m & 2k @xmath35 4k , 0.6@xmath28 pix@xmath4 & @xmath60 & 28 nov 2001 3 apr 2002 + usno & 1.0 m & 1k @xmath35 1k , 0.68@xmath28 pix@xmath4 & @xmath61 & 11 nov 2002 30 apr 2003 + usno & 1.0 m & 1k @xmath35 1k , 0.68@xmath28 pix@xmath4 & @xmath60 & 23 dec 2003 12 mar 2004 + kpno & 0.9 m & 2k @xmath35 2k , 0.6@xmath28
pix@xmath4 & @xmath62 & 1 22 dec 2001 + wise & 1.0 m & 1k @xmath35 1k , 0.7@xmath28 pix@xmath4 & @xmath14 & 2 22 dec 2001 + tenagra & 0.81 m & 1k @xmath35 1k , 0.87@xmath28 pix@xmath4 & @xmath14 & 2 nov 2002 4 apr 2004 + kait & 0.76 m & 0.5k @xmath35 0.5k , 0.8@xmath28 pix@xmath4 & @xmath14 & 20 sept 2002 30 mar 2003 + teide & 0.82 m & 1k @xmath35 1k , 0.432@xmath28 pix@xmath4 & @xmath14 & 28 sept 2002 28 mar 2003 + konkoly & 1.0 m & 1k @xmath35 1k , 0.33@xmath28 pix@xmath4 & @xmath14 & 1 oct 2002 2 mar 2003 + vvo & 7 ( 4.2@xmath28 ) & 10 ( 6.0@xmath28 ) & 5 ( 3.0@xmath28 ) + mmo & 10 ( 2.6@xmath28 ) & 11 ( 2.86@xmath28 ) & 5 ( 1.3@xmath28 ) + usno & 4.5 ( 3.06@xmath28 ) & 6.5 ( 4.42@xmath28 ) & 2 ( 1.36@xmath28 ) + tenagra & 3 ( 2.61@xmath28 ) & 5 ( 4.35@xmath28 ) & 2 ( 1.74@xmath28 ) + kait & 3 ( 2.4@xmath28 ) & 5 ( 3.2@xmath28 ) & 2 ( 1.6@xmath28 ) + teide & 2.5 ( 1.1@xmath28 ) & 20 ( 8.6@xmath28 ) & 8 ( 3.5@xmath28 ) + konkoly & 12 ( 3.96@xmath28 ) & 15 ( 4.95@xmath28 ) & 5 ( 1.65@xmath28 ) + a & 14.401 @xmath0 0.022 & 14.115 @xmath0 0.021 & 13.368 @xmath0 0.020 & 12.952 @xmath0 0.020 & 12.548 @xmath0 0.021 + b & 13.514 @xmath0 0.022 & 13.341 @xmath0 0.021 & 12.624 @xmath0 0.020 & 12.196 @xmath0 0.020 & 11.734
@xmath0 0.021 + c & 13.563 @xmath0 0.022 & 13.566 @xmath0 0.021 & 12.969 @xmath0 0.020 & 12.617 @xmath0 0.020 & 12.240 @xmath0 0.021 + d & 15.808 @xmath0 0.028 & 15.921 @xmath0 0.022 & 15.013 @xmath0 0.020 & 14.328 @xmath0 0.021 & 13.645 @xmath0 0.021 + e & 15.803 @xmath0 0.028 & 15.004 @xmath0 0.022 & 13.885 @xmath0 0.020 & 13.211 @xmath0 0.021 & 12.614 @xmath0 0.021 + f & 15.319 @xmath0 0.024 & 14.745 @xmath0 0.021 & 13.869 @xmath0 0.020 & 13.375 @xmath0 0.021 & 12.902 @xmath0 0.021 + g & 16.842 @xmath0 0.041 & 15.642 @xmath0 0.022 & 14.230 @xmath0 0.020 & 13.350 @xmath0 0.021 & 12.479 @xmath0
0.021 + * 16d & 14.303 @xmath0 0.001 & 14.304 @xmath0 0.001 & 14.302 @xmath0 0.002 + * 17d & 15.085 @xmath0 0.001 & 15.098 @xmath0 0.002 & 15.158 @xmath0 0.003 + * 21d & 15.023 @xmath0 0.001 & 14.991 @xmath0 0.002 & 15.006 @xmath0 0.003 + * 27d & 13.783 @xmath0 0.001 & 13.785 @xmath0 0.001 & 13.794 @xmath0 0.002 + * 31d & 14.423 @xmath0 0.001 & 14.426 @xmath0 0.001 & 14.411 @xmath0 0.002 + * * * * * 2450017.8375 & 14.449 & 0.011 & vvo + 2450021.7443 & 14.521 & 0.024 & vvo + 2450026.7416 & 17.132 & 0.074 & vvo + 2450028.6983 & 14.687 & 0.014 & vvo + 2450028.8949 & 14.435 & 0.015 & vvo + 2450031.7696 & 14.014 & 0.007 & vvo + 2450039.7466 & 14.562 & 0.010 & vvo + 2450054.6473 & 14.466 & 0.009 & vvo + 2450056.7206 & 14.466 & 0.007 & vvo + 2450056.8106 & 14.457 & 0.009 & vvo + 1.619 & 0.037 & 14.475 & 0.024 & 2452237.7643 & usno + 1.582 & 0.024 & 14.525 & 0.018 & 2452241.8421 & usno + 1.416 & 0.064 & 18.061 & 0.037 & 2452261.8464 & kpno + 1.379 & 0.040 & 17.833 & 0.026 & 2452262.8831 & kpno + 1.516 & 0.047 & 17.601 & 0.031 & 2452263.8295 & kpno + 1.565 & 0.039 & 16.739 & 0.023 & 2452264.8739 & kpno + 1.584 & 0.013 & 14.695 & 0.009 & 2452266.7740 & usno + 1.621 & 0.014 & 14.519 & 0.008 & 2452269.8189 & usno + 1.680 & 0.025 & 14.435 & 0.021 & 2452279.6966 & usno + 1.716 & 0.091 & 17.427 & 0.073 & 2452306.8015 & usno + 1995/1996 & 0.200.80 + 1996/1997 & 0.200.80 + 1997/1998 & 0.200.80 + 1998/1999 & 0.250.75 + 1999/2000 & 0.250.75 + 2000/2001 & 0.250.75 + 2001/2002 & 0.300.70 + 2002/2003 & 0.300.70 + 2003/2004 & 0.350.65 + | we present results from the most recent ( 20022004 ) observing campaigns of the eclipsing system kh 15d , in addition to re - reduced data obtained at van vleck observatory ( vvo ) between 1995 and 2000 .
phasing nine years of photometric data shows substantial evolution in the width and depth of the eclipses .
the most recent data indicate that the eclipses are now approximately 24 days in length , or half the orbital period .
these results are interpreted and discussed in the context of the recent models for this system put forward by winn et al .
( 2004 ) and chiang & murray - clay ( 2004 ) .
a periodogram of the entire data set yields a highly significant peak at 48.37 @xmath0 0.01 days , which is in accord with the spectroscopic period of 48.38 @xmath0 0.01 days determined by johnson et al .
( 2004 ) .
another significant peak , at 9.6 days , was found in the periodogram of the out - of - eclipse data at two different epochs .
we interpret this as the rotation period of the visible star and argue that it may be tidally locked in pseudosynchronism with its orbital motion .
if so , application of hut s ( 1981 ) theory implies that the eccentricity of the orbit is @xmath1 = 0.65 @xmath0 0.01 .
analysis of the uves / vlt spectra obtained by hamilton et al .
( 2003 ) shows that the @xmath2sin@xmath3 of the visible star in this system is 6.9 @xmath0 0.3 km s@xmath4 . using this value of @xmath2sin@xmath3 and the measured rotation period of the star
, we calculate the lower limit on the radius to be @xmath5r@xmath6 , which concurs with the value obtained by hamilton et al .
( 2001 ) from its luminosity and effective temperature . here
we assume that @xmath7 = 90@xmath8 since it is likely that the spin and orbital angular momenta vectors are nearly aligned .
one unusually bright data point obtained in the 1995/1996 observing season at vvo is interpreted as the point in time when the currently hidden star ( b ) made its last appearance .
based on this datum , we show that star b is 0.46 @xmath0 0.03 mag brighter than the currently visible star a , which is entirely consistent with the historical light curve ( johnson et al .
2005 ) . finally , well - sampled @xmath9 and @xmath10 data obtained at the ctio / yale 1-m telescope during 2001/2002 show an entirely new feature : the system becomes bluer by a small but significant amount in very steady fashion as it enters eclipse and shows an analogous reddening as it emerges from eclipse .
this suggests an extended zone of hot gas located close to , but above , the photosphere of the currently visible star .
the persistance of the bluing of the light curve shows that its length scale is comparable to a stellar radius . |
radiatively induced mass models are one of the promising candidate to include a dark matter ( dm ) candidate naturally , which connect the standard model ( sm ) fermions and dm . along this line of idea , there exist a lot of papers , _ i.e. _ , @xcite at one - loop level , @xcite at two - loop level , @xcite at three - loop level , and
@xcite at four - loop level . however authors mainly focus on the neutrino sector , but not so many on the quark sector @xcite . in this paper
, we propose all the sm fermion masses are induced at one - loop level except for the third generation , and the inert type of boson dm couples to all these fermions .
the masses of third generation fermions are generated via the vacuum expectation value ( vev ) of sm higgs field to be consistent with sm higgs properties observed by lhc experiments such as gluon fusion cross section and @xmath3 branching fractions @xcite .
furthermore it would be natural to require first and second generation masses are generated at loop - level from the fermion mass hierarchy . then we add extra local u(1 ) symmetry to restrict the yukawa interaction associated with sm higgs field in anomaly free way .
the vector - like fermions are also introduced to write relevant one - loop diagrams for fermion mass generation . in our model , therefore , the light fermion masses are generated at the one - loop level induced by the yukawa interactions among sm fermions , inert scalar fields and vector - like fermions which are invariant under the new u(1 ) .
we note that all these yukawa couplings can not be so large to induce the relevant relic abundance of dm in our parameter choices , and the nature of dm is the same as the two higgs doublet model with one inert @xmath4 doublet boson . in order to reproduce the observed mixing matrices and masses for the lepton and quark sector
, we have to take into account the flavor changing neutral currents ( fcncs ) and lepton flavor violations ( lfvs ) where mediated boson masses ( including dm ) are restricted by the both sector . also positive contribution to the anomalous magnetic moment
are induced from the lepton sector via one - loop diagram in which the mediated boson are included .
therefore , an economical scenario including the quark sector might be achieved in a sense .
this paper is organized as follows . in sec .
ii , we show our model , and establish the quark and lepton sector , and derive the analytical forms of fcncs , lfvs , muon anomalous magnetic moment . in sec .
iii , we have a numerical analysis , and show some results .
we conclude and discuss in sec .
.field contents of fermions and their charge assignments under @xmath5 , where each of the flavor index is defined as @xmath6 and @xmath7 . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] _ lepton flavor violations _ : the lepton flavor ( lfvs ) violation processes give the constraints on our parameters .
the most known processes are @xmath8 , and its branching ratio is given by @xmath9 where @xmath10 is the fine - structure constant , @xmath11 for ( @xmath12 ) , @xmath13 gev@xmath14 is the fermi constant , @xmath15 is computed as @xmath16,\\ & f_{\ell_j\to\ell_\alpha\gamma}(m_{1},m_{2 } ) = \frac{2 m_1 ^ 4 - 4 m_1 ^ 2 m_2 ^ 2 + m_2 ^ 4 + 4 m_1 ^ 4 \ln\left[\frac{m_2}{m_1}\right]}{2(m_1 ^ 2-m_2 ^ 2)^3}.\end{aligned}\ ] ] _ muon anomalous magnetic moment @xmath17 _ : through the same process from the above lvfs , there exists the contribution to @xmath17 , and its form @xmath18 is simply given by @xmath19 this value can be tested by current experiments @xcite . _
@xmath1 conversion _
: the @xmath1 conversion process can be found in the same diagram as the process of @xmath8 with @xmath20 line being attached to nucleons , where additional contribution is taken into account by replacing @xmath20 with @xmath21 boson .
then the @xmath1 conversion rate @xmath22 is given by @xcite @xmath23,\\ & f_{\mu e}(m_1,m_2,m_3)= \int_0 ^ 1dx_3\int_0^{1-x_3}dx_4 \frac{x_4(1-x_4)}{x_3 m_1 ^ 2 + ( 1-x_3 ) m_2 ^ 2+x_4(x_3 + x_4 - 1 ) m_3 ^ 2 } , \ ] ] where @xmath24 , and @xmath25 , and @xmath26 gev , @xmath27 , @xmath28 , @xmath29 , @xmath30 .
the values of @xmath31 , @xmath21 , @xmath32 , @xmath33 , and @xmath34 depend on the kind of nuclei . here
we focus on titanium , because its sensitivity will be improved by several orders of magnitude @xcite in near future compared to the current bound @xcite , as can be seen in the table [ tab : cif ] . in this case , these values are determined by @xmath35 sec@xmath36 , @xmath37 , @xmath38 , @xmath39 , and @xmath40 @xcite .
now that we have all the formulae for the quark and lepton sector , and we carry out numerical analysis to find what kind of regions are allowed . here
we randomly select values of the fifteen parameters within the corresponding ranges @xmath41,\quad m_{h_1 } \in [ 600\ \text{gev } , 800\ \text{gev}],\quad m_{h_2 } \in [ 4\ \text{tev } , 6\ \text{tev}],{\nonumber}\\ & m_{q'_1 } \in [ 4.5\ \text{tev } , 5\ \text{tev}],\quad m_{q'_2 } \in [ 1.7\ \text{tev } , 2.2\ \text{tev}],\\ & m_{l'_1 } \in [ 7\ \text{tev } , 7.5\ \text{tev}],\quad m_{l'_2 } \in [ 9\ \text{tev } , 10\ \text{tev}],\\ & \ { ( y_u)_{12 } , ( y_u)_{22 } \ } \in [ -1,1],\quad ( y_u)_{11 } \in [ -0.1,0.1 ] , \quad ( y_u)_{21 } \in [ -0.002,0.002 ] , { \nonumber}\\ & \ { ( y_\nu)_{21 } , ( y_\nu)_{22 } \ } \in [ -7\times10^{-12},7\times10^{-12}],\quad ( y_\nu)_{11 } \in [ -3\times10^{-13},3\times10^{-13 } ] , { \nonumber}\\ & ( y_\nu)_{12 } \in [ -2\times10^{-13},2\times10^{-12 } ] , \label{range_scanning}\end{aligned}\ ] ] to reproduce quark masses , ckm mixings for the quark sector , and neutrino oscillation data and satisfy the constraints of lfvs for the lepton sector . in this analysis
, we are preparing 10 million sample points .
notice here that the other yukawa couplings such as @xmath42 , @xmath43 , @xmath44 , @xmath45 are numerically solved by using the best fit values of the measurements in ref .
@xcite for quark sector and ref .
@xcite for lepton sector
. then we obtain the sets of yukawa couplings where all we need to take care is not to exceed the perturbative limit that we take @xmath46 as upper limit .
the sets of yukawa couplings are applied to calculate @xmath47 , @xmath48 , @xmath18 , @xmath49 and @xmath1 conversion .
_ numerical results _ :
the numerical results are shown in fig .
[ fig : results ] .
the left figure is the scattering plot in terms of @xmath47 and @xmath48 normalized by [ mev]@xmath50 , and the black solid lines represents the experimental upper bound .
here we have found 734 allowed points , and the most stringent constraint comes from the process of @xmath51 .
it tells us that our @xmath48 values are the same order as current experimental constraint and some parameter sets are excluded .
similarly @xmath47 values can be comparable to the current experimental bound .
thus they can be tested in the near future .
the right figure is the scattering plot in terms of @xmath52 and @xmath53 .
it tells us that the maximal value for @xmath2 is around @xmath54 , which is lower than the current bound by three order of magnitude .
@xmath55 is also much lower than the current bound , however it will be test in the future experiment such as comet @xcite , which will reach @xmath56 as shown in the previous section . and @xmath48 normalized by [ mev]@xmath50 , and the black solid lines represents the experimental upper bound .
the right figure is the scattering plot in terms of @xmath52 and @xmath53 .
, title="fig : " ] and @xmath48 normalized by [ mev]@xmath50 , and the black solid lines represents the experimental upper bound .
the right figure is the scattering plot in terms of @xmath52 and @xmath53 .
, title="fig : " ]
we have proposed a radiatively induced quark and lepton mass model in the first and second generation , in which we have analyzed the allowed regions simultaneously to satisfy the fcncs for the quark sector and lfvs including @xmath1 conversion in addition to the quark mass and mixing and the lepton mass and mixing .
also we have estimated the typical value for the @xmath2 .
then we have found @xmath48 and @xmath47 can be the same order as the current experimental bound where some parameter sets are excluded .
thus our model can be further tested in the near future . as for the lepton sector
, we have found that the maximal value for @xmath2 is around @xmath54 , which is lower than the current bound by three order of magnitude .
@xmath55 is also much lower than the current bound , however it will be test in the future experiment such as comet , which will reach @xmath56 .
we note that the @xmath57 boson from @xmath58 gauge symmetry has flavor violating interaction due to our choice of charge assignment for sm fermions .
since @xmath57 couples to both sm quarks and leptons these interaction could be tested in future lhc experiments .
particularly lepton flavor violating signals @xmath59 would be interesting signatures of the model .
detailed simulation studies of the signal is beyond the scope of this paper and we left it as a future work . at the end of this paper
, we mention the dark matter candidate . in our case ( and our parametrization ) , @xmath60 can be a dark matte candidate , which is the imaginary component of the @xmath4 doublet inert boson .
the dominant annihilation processes are induced through the gauge interactions , since yukawa couplings related to @xmath61 , @xmath62 and @xmath45 , are expected to be tiny .
thus its nature is the same as the two higgs doublet model with one inert boson and serious analysis can be found in ref .
it can also be detected through the spin independent direct detection searches such as lux @xcite , because it has two higgs portal interactions with the nucleon .
this situation might relax the experimental constraint compared to the one higgs portal scenario , activating the cancellation mechanism between two cp - even bosons @xcite .
h. o. is sincerely grateful for all the kias members , korean cordial persons , foods , culture , weather , and all the other things .
a. zee , phys .
b * 93 * , 389 ( 1980 ) [ erratum - ibid .
b * 95 * , 461 ( 1980 ) ] , t. p. cheng and l. f. li , phys .
d * 22 * , 2860 ( 1980 ) , e. ma , phys . rev .
d * 73 * , 077301 ( 2006 ) [ hep - ph/0601225 ] , b. s. balakrishna , phys .
lett . * 60 * , 1602 ( 1988 ) , b. s. balakrishna and r. n. mohapatra , phys .
b * 216 * , 349 ( 1989 ) , x. g. he , r. r. volkas and d. d. wu , phys .
d * 41 * , 1630 ( 1990 ) , t. hambye , k. kannike , e. ma and m. raidal , phys .
d * 75 * , 095003 ( 2007 ) [ hep - ph/0609228 ] , p. -h .
gu and u. sarkar , phys .
d * 77 * , 105031 ( 2008 ) [ arxiv:0712.2933 [ hep - ph ] ] , n. sahu and u. sarkar , phys .
d * 78 * , 115013 ( 2008 ) [ arxiv:0804.2072 [ hep - ph ] ] , p. -h .
gu and u. sarkar , phys .
d * 78 * , 073012 ( 2008 ) [ arxiv:0807.0270 [ hep - ph ] ] , d. aristizabal sierra and d. restrepo , jhep * 0608 * , 036 ( 2006 ) [ hep - ph/0604012 ] , r. bouchand and a. merle , jhep * 1207 * , 084 ( 2012 ) [ arxiv:1205.0008 [ hep - ph ] ] , k. l. mcdonald , jhep * 1311 * , 131 ( 2013 ) [ arxiv:1310.0609 [ hep - ph ] ] , e. ma , phys .
b * 732 * , 167 ( 2014 ) [ arxiv:1401.3284 [ hep - ph ] ] , y. kajiyama , h. okada and k. yagyu , nucl .
b * 887 * , 358 ( 2014 ) [ arxiv:1309.6234 [ hep - ph ] ] , s. kanemura , o. seto and t. shimomura , phys .
d * 84 * , 016004 ( 2011 ) [ arxiv:1101.5713 [ hep - ph ] ] , s. kanemura , t. nabeshima and h. sugiyama , phys .
b * 703 * , 66 ( 2011 ) [ arxiv:1106.2480 [ hep - ph ] ] , s. kanemura , t. nabeshima and h. sugiyama , phys . rev .
d * 85 * , 033004 ( 2012 ) [ arxiv:1111.0599 [ hep - ph ] ] , d. schmidt , t. schwetz and t. toma , phys .
d * 85 * , 073009 ( 2012 ) [ arxiv:1201.0906 [ hep - ph ] ] , s. kanemura and h. sugiyama , phys .
d * 86 * , 073006 ( 2012 ) [ arxiv:1202.5231 [ hep - ph ] ] , y. farzan and e. ma , phys .
d * 86 * , 033007 ( 2012 ) [ arxiv:1204.4890 [ hep - ph ] ] , k. kumericki , i. picek and b. radovcic , jhep * 1207 * , 039 ( 2012 ) [ arxiv:1204.6597 [ hep - ph ] ] , k. kumericki , i. picek and b. radovcic , phys . rev .
d * 86 * , 013006 ( 2012 ) [ arxiv:1204.6599 [ hep - ph ] ] , e. ma , phys .
b * 717 * , 235 ( 2012 ) [ arxiv:1206.1812 [ hep - ph ] ] , g. gil , p. chankowski and m. krawczyk , phys .
b * 717 * , 396 ( 2012 ) [ arxiv:1207.0084 [ hep - ph ] ] , h. okada and t. toma , phys . rev .
d * 86 * , 033011 ( 2012 ) arxiv:1207.0864 [ hep - ph ] , d. hehn and a. ibarra , phys . lett .
b * 718 * , 988 ( 2013 ) [ arxiv:1208.3162 [ hep - ph ] ] , p. s. b. dev and a. pilaftsis , phys .
d * 86 * , 113001 ( 2012 ) [ arxiv:1209.4051 [ hep - ph ] ] , y. kajiyama , h. okada and t. toma , eur .
j. c * 73 * , 2381 ( 2013 ) [ arxiv:1210.2305 [ hep - ph ] ] , t. toma and a. vicente , jhep * 1401 * , 160 ( 2014 ) [ arxiv:1312.2840 , arxiv:1312.2840 [ hep - ph ] ] , s. kanemura , t. matsui and h. sugiyama , phys . lett .
b * 727 * , 151 ( 2013 ) [ arxiv:1305.4521 [ hep - ph ] ] , s. s. c. law and k. l. mcdonald , jhep * 1309 * , 092 ( 2013 ) [ arxiv:1305.6467 [ hep - ph ] ] , b. dasgupta , e. ma and k. tsumura , phys .
d * 89 * , 041702 ( 2014 ) [ arxiv:1308.4138 [ hep - ph ] ] , s. baek and h. okada , arxiv:1403.1710 [ hep - ph ] , s. kanemura , t. matsui and h. sugiyama , phys .
d * 90 * , 013001 ( 2014 ) [ arxiv:1405.1935 [ hep - ph ] ] , s. fraser , e. ma and o. popov , phys .
b * 737 * , 280 ( 2014 ) [ arxiv:1408.4785 [ hep - ph ] ] , a. vicente and c. e. yaguna , jhep * 1502 * , 144 ( 2015 ) [ arxiv:1412.2545 [ hep - ph ] ] , s. baek , h. okada and k. yagyu , jhep * 1504 * , 049 ( 2015 ) [ arxiv:1501.01530 [ hep - ph ] ] , a. merle and m. platscher , phys .
d * 92 * , no . 9 , 095002 ( 2015 ) [ arxiv:1502.03098 [ hep - ph ] ] , d. restrepo , a. rivera , m. snchez - pelez , o. zapata and w. tangarife , arxiv:1504.07892 [ hep - ph ] , a. merle and m. platscher , jhep * 1511 * , 148 ( 2015 ) [ arxiv:1507.06314 [ hep - ph ] ] , w. wang and z. l. han , phys . rev .
d * 92 * , 095001 ( 2015 ) [ arxiv:1508.00706 [ hep - ph ] ] , y. h. ahn and h. okada , phys .
d * 85 * , 073010 ( 2012 ) [ arxiv:1201.4436 [ hep - ph ] ] , e. ma , a. natale and a. rashed , int .
j. mod .
a * 27 * , 1250134 ( 2012 ) [ arxiv:1206.1570 [ hep - ph ] ] , a. e. carcamo hernandez , i. d. m. varzielas , s. g. kovalenko , h. ps and i. schmidt , phys .
d * 88 * , 076014 ( 2013 ) [ arxiv:1307.6499 [ hep - ph ] ] , e. ma and a. natale , phys .
b * 723 * , 403 ( 2014 ) [ arxiv:1403.6772 [ hep - ph ] ] , e. ma , phys .
b * 741 * , 202 ( 2015 ) [ arxiv:1411.6679 [ hep - ph ] ] , e. ma , arxiv:1504.02086 [ hep - ph ] , e. ma , phys .
lett . * 112 * , 091801 ( 2014 ) [ arxiv:1311.3213 [ hep - ph ] ] , h. okada and k. yagyu , phys .
d * 89 * , 053008 ( 2014 ) [ arxiv:1311.4360 [ hep - ph ] ] , h. okada and k. yagyu , phys . rev .
d * 90 * , no . 3 , 035019 ( 2014 ) [ arxiv:1405.2368 [ hep - ph ] ] , v. brdar , i. picek and b. radovcic , phys .
b * 728 * , 198 ( 2014 ) [ arxiv:1310.3183 [ hep - ph ] ] , h. okada , y. orikasa and t. toma , arxiv:1511.01018 [ hep - ph ] , f. bonnet , m. hirsch , t. ota and w. winter , jhep * 1207 * , 153 ( 2012 ) [ arxiv:1204.5862 [ hep - ph ] ] , f. r. joaquim and j. t. penedo , phys .
d * 90 * , no .
3 , 033011 ( 2014 ) [ arxiv:1403.4925 [ hep - ph ] ] , h. davoudiasl and i. m. lewis , phys .
d * 90 * , no . 3 ,
033003 ( 2014 ) [ arxiv:1404.6260 [ hep - ph ] ] , m. lindner , s. schmidt and j. smirnov , arxiv:1405.6204 [ hep - ph ] , h. okada and y. orikasa , arxiv:1412.3616 [ hep - ph ] , y. mambrini , s. profumo and f. s. queiroz , arxiv:1508.06635 [ hep - ph ] , s. m. boucenna , s. morisi and j. w. f. valle , adv .
high energy phys .
* 2014 * , 831598 ( 2014 ) [ arxiv:1404.3751 [ hep - ph ] ] , a. ahriche , s. m. boucenna and s. nasri , arxiv:1601.04336 [ hep - ph ] , s. fraser , c. kownacki , e. ma and o. popov , arxiv:1511.06375 [ hep - ph ] , s. fraser , e. ma and m. zakeri , arxiv:1511.07458 [ hep - ph ] , r. adhikari , d. borah and e. ma , arxiv:1512.05491 [ hep - ph ] , h. okada and y. orikasa , arxiv:1512.06687 [ hep - ph ] , a. ibarra , c. e. yaguna and o. zapata , phys .
d * 93 * , no . 3 , 035012 ( 2016 ) [ arxiv:1601.01163 [ hep - ph ] ] , c. arbelaez , a. e. c. hernandez , s. kovalenko and i. schmidt , arxiv:1602.03607 [ hep - ph ] , a. ahriche , k. l. mcdonald , s. nasri and i. picek , phys .
b * 757 * , 399 ( 2016 ) [ arxiv:1603.01247 [ hep - ph ] ] , w. b. lu and p. h. gu , arxiv:1603.05074 [ hep - ph ] , c. kownacki and e. ma , arxiv:1604.01148 [ hep - ph ] , a. ahriche , k. l. mcdonald and s. nasri , arxiv:1604.05569 [ hep - ph ] , a. ahriche , a. manning , k. l. mcdonald and s. nasri , arxiv:1604.05995 [ hep - ph ] , e. ma , n. pollard , o. popov and m. zakeri , arxiv:1605.00991 [ hep - ph ] , t. nomura , h. okada and y. orikasa , arxiv:1605.02601 [ hep - ph ] , c. hagedorn , t. ohlsson , s. riad and m. a. schmidt , arxiv:1605.03986 [ hep - ph ] , o. antipin , p. culjak , k. kumericki and i. picek , arxiv:1606.05163 [ hep - ph ] , t. nomura and h. okada , arxiv:1606.09055 [ hep - ph ] .
a. zee , nucl .
b * 264 * , 99 ( 1986 ) , k. s. babu , phys . lett .
b * 203 * , 132 ( 1988 ) , k. s. babu and c. macesanu , phys .
d * 67 * , 073010 ( 2003 ) [ hep - ph/0212058 ] , d. aristizabal sierra and m. hirsch , jhep * 0612 * , 052 ( 2006 ) [ hep - ph/0609307 ] , m. nebot , j. f. oliver , d. palao and a. santamaria , phys .
d * 77 * , 093013 ( 2008 ) [ arxiv:0711.0483 [ hep - ph ] ] , d. schmidt , t. schwetz and h. zhang , nucl .
b * 885 * , 524 ( 2014 ) [ arxiv:1402.2251 [ hep - ph ] ] , j. herrero - garcia , m. nebot , n. rius and a. santamaria , nucl .
b * 885 * , 542 ( 2014 ) [ arxiv:1402.4491 [ hep - ph ] ] , h. n. long and v. v. vien , int .
j. mod .
a * 29 * , no . 13 , 1450072 ( 2014 ) [ arxiv:1405.1622 [ hep - ph ] ] , v. van vien , h. n. long and p. n. thu , arxiv:1407.8286 [ hep - ph ] , m. aoki , s. kanemura , t. shindou and k. yagyu , jhep * 1007 * , 084 ( 2010 ) [ erratum - ibid . *
1011 * , 049 ( 2010 ) ] [ arxiv:1005.5159 [ hep - ph ] ] , m. lindner , d. schmidt and t. schwetz , phys .
b * 705 * , 324 ( 2011 ) [ arxiv:1105.4626 [ hep - ph ] ] , s. baek , p. ko , h. okada and e. senaha , jhep * 1409 * , 153 ( 2014 ) [ arxiv:1209.1685 [ hep - ph ] ] , m. kohda , h. sugiyama and k. tsumura , phys .
b * 718 * , 1436 ( 2013 ) [ arxiv:1210.5622 [ hep - ph ] ] , m. aoki , j. kubo and h. takano , phys .
d * 87 * , no .
11 , 116001 ( 2013 ) [ arxiv:1302.3936 [ hep - ph ] ] , y. kajiyama , h. okada and k. yagyu , nucl .
b * 874 * , 198 ( 2013 ) [ arxiv:1303.3463 [ hep - ph ] ] , y. kajiyama , h. okada and t. toma , phys .
d * 88 * , 015029 ( 2013 ) [ arxiv:1303.7356 ] , s. baek , h. okada and t. toma , jcap * 1406 * , 027 ( 2014 ) [ arxiv:1312.3761 [ hep - ph ] ] , h. okada , arxiv:1404.0280 [ hep - ph ] , h. okada , t. toma and k. yagyu , phys .
d * 90 * , no .
9 , 095005 ( 2014 ) [ arxiv:1408.0961 [ hep - ph ] ] , h. okada , arxiv:1503.04557 [ hep - ph ] , c. q. geng and l. h. tsai , arxiv:1503.06987 [ hep - ph ] , s. kashiwase , h. okada , y. orikasa and t. toma , arxiv:1505.04665 [ hep - ph ] , m. aoki and t. toma , jcap * 1409 * , 016 ( 2014 ) [ arxiv:1405.5870 [ hep - ph ] ] , s. baek , h. okada and t. toma , phys .
b * 732 * , 85 ( 2014 ) [ arxiv:1401.6921 [ hep - ph ] ] , h. okada and y. orikasa , arxiv:1509.04068 [ hep - ph ] , d. aristizabal sierra , a. degee , l. dorame and m. hirsch , jhep * 1503 * , 040 ( 2015 ) [ arxiv:1411.7038 [ hep - ph ] ] , t. nomura and h. okada , phys .
b * 756 * , 295 ( 2016 ) [ arxiv:1601.07339 [ hep - ph ] ] , t. nomura , h. okada and y. orikasa , arxiv:1602.08302 [ hep - ph ] .
l. m. krauss , s. nasri and m. trodden , phys .
d * 67 * , 085002 ( 2003 ) [ arxiv : hep - ph/0210389 ] , m. aoki , s. kanemura and o. seto , phys . rev . lett . * 102 * , 051805 ( 2009 ) [ arxiv:0807.0361 ] , m. gustafsson , j. m. no and m. a. rivera , phys .
lett . * 110 * , 211802 ( 2013 ) arxiv:1212.4806 [ hep - ph ] , a. ahriche , s. nasri and r. soualah , phys .
d * 89 * , 095010 ( 2014 ) [ arxiv:1403.5694 [ hep - ph ] ] , a. ahriche , c. s. chen , k. l. mcdonald and s. nasri , phys .
d * 90 * , no . 1 , 015024 ( 2014 ) [ arxiv:1404.2696 [ hep - ph ] ] , a. ahriche , k. l. mcdonald and s. nasri , jhep * 1410 * , 167 ( 2014 ) [ arxiv:1404.5917 [ hep - ph ] ] , c. -s . chen , k. l. mcdonald and s. nasri , phys .
b * 734 * , 388 ( 2014 ) [ arxiv:1404.6033 [ hep - ph ] ] , h. okada and y. orikasa , phys .
d * 90 * , no . 7 , 075023 ( 2014 ) [ arxiv:1407.2543 [ hep - ph ] ] , h. hatanaka , k. nishiwaki , h. okada and y. orikasa , nucl . phys .
b * 894 * , 268 ( 2015 ) [ arxiv:1412.8664 [ hep - ph ] ] , l. g. jin , r. tang and f. zhang , phys . lett .
b * 741 * , 163 ( 2015 ) [ arxiv:1501.02020 [ hep - ph ] ] , p. culjak , k. kumericki and i. picek , phys .
b * 744 * , 237 ( 2015 ) [ arxiv:1502.07887 [ hep - ph ] ] , h. okada , n. okada and y. orikasa , phys .
d * 93 * , no . 7 , 073006 ( 2016 ) [ arxiv:1504.01204 [ hep - ph ] ] , c. q. geng , d. huang and l. h. tsai , phys . lett .
b * 745 * , 56 ( 2015 ) [ arxiv:1504.05468 [ hep - ph ] ] , a. ahriche , k. l. mcdonald , s. nasri and t. toma , phys .
b * 746 * , 430 ( 2015 ) [ arxiv:1504.05755 [ hep - ph ] ] , k. nishiwaki , h. okada and y. orikasa , arxiv:1507.02412 [ hep - ph ] , h. okada and k. yagyu , arxiv:1508.01046 [ hep - ph ] , a. ahriche , k. l. mcdonald and s. nasri , arxiv:1508.02607 [ hep - ph ] , y. kajiyama , h. okada and k. yagyu , jhep * 10 * , 196 ( 2013 ) arxiv:1307.0480 [ hep - ph ] , s. f. king , a. merle and l. panizzi , arxiv:1406.4137 [ hep - ph ] , s. kanemura , k. nishiwaki , h. okada , y. orikasa , s. c. park and r. watanabe , arxiv:1512.09048 [ hep - ph ] , h. okada and k. yagyu , phys .
b * 756 * , 337 ( 2016 ) [ arxiv:1601.05038 [ hep - ph ] ] , p. ko , t. nomura , h. okada and y. orikasa , arxiv:1602.07214 [ hep - ph ] , t. nomura , h. okada and y. orikasa , arxiv:1603.04631 [ hep - ph ] , t. t. thuc , l. t. hue , h. n. long and t. p. nguyen , arxiv:1604.03285 [ hep - ph ] , d. cherigui , c. guella , a. ahriche and s. nasri , arxiv:1605.03640 [ hep - ph ] , b. s. balakrishna , phys .
lett . * 60 * , 1602 ( 1988 ) .
b. s. balakrishna and r. n. mohapatra , phys .
b * 216 * , 349 ( 1989 ) .
e. ma , phys .
lett . * 63 * , 1042 ( 1989 ) .
x. g. he , r. r. volkas and d. d. wu , phys .
d * 41 * , 1630 ( 1990 ) . c. arbelaez , a. e. c. hernandez , s. kovalenko and i. schmidt , arxiv:1602.03607 [ hep - ph ] . c. kownacki and e. ma , arxiv:1604.01148 [ hep - ph ] .
a. ibarra and a. solaguren - beascoa , phys .
b * 736 * , 16 ( 2014 ) [ arxiv:1403.2382 [ hep - ph ] ] .
m. baumgart , d. stolarski and t. zorawski , phys .
d * 90 * , no . 5 , 055001 ( 2014 ) [ arxiv:1403.6118 [ hep - ph ] ] .
p. ko , y. omura and c. yu , phys .
b * 717 * , 202 ( 2012 ) [ arxiv:1204.4588 [ hep - ph ] ] .
s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
b * 720 * , 63 ( 2013 ) [ arxiv:1212.6175 [ hep - ex ] ] .
g. aad _ et al .
_ [ atlas collaboration ] , phys .
d * 90 * , no . 5 , 052005 ( 2014 ) [ arxiv:1405.4123 [ hep - ex ] ] .
the atlas collaboration , atlas - conf-2015 - 070 .
f. jegerlehner and a. nyffeler , phys .
rept . * 477 * , 1 ( 2009 ) [ arxiv:0902.3360 [ hep - ph ] ] .
m. benayoun , p. david , l. delbuono and f. jegerlehner , eur .
j. c * 72 * , 1848 ( 2012 ) [ arxiv:1106.1315 [ hep - ph ] ] .
t. hambye , f .- s .
ling , l. lopez honorez and j. rocher , jhep * 0907 * , 090 ( 2009 ) erratum : [ jhep * 1005 * , 066 ( 2010 ) ] [ arxiv:0903.4010 [ hep - ph ] ] . l. lopez honorez , e. nezri , j. f. oliver and m. h. g. tytgat , jcap * 0702 * , 028 ( 2007 ) [ hep - ph/0612275 ] .
s. baek , p. ko , w. i. park and e. senaha , jhep * 1211 * , 116 ( 2012 ) [ arxiv:1209.4163 [ hep - ph ] ] . | we propose a radiatively induced quark and lepton mass model in the first and second generation with extra @xmath0 gauge symmetry and vector - like fermions .
then we analyze the allowed regions which simultaneously satisfy the fcncs for the quark sector , lfvs including @xmath1 conversion , the quark mass and mixing , and the lepton mass and mixing .
also we estimate the typical value for the @xmath2 in our model .
kias - p16046 |
at the end of the large electron positron ( lep ) operations , an exclusion limit was reached by combining the results of all the higgs search channels performed by its four experiments , establishing a lower limit on @xmath2 of 114.4 @xmath3 at 95% confidence level .
more recently , the direct higgs searches performed by the cdf and d0 experiments at the tevatron accelerator have reached better sensitivity , making possible the exclusion of a sm higgs boson with a mass between 158 and 174 @xmath3 .
the lhc is continuously collecting luminosity at a center of mass energy of 7 @xmath0 since march 2010 .
the projected luminosity to be delivered before the shut - down to go to higher energies is 10 @xmath4 by november 2010 and 1 @xmath1 by the end of 2011 . the main production modes for the standard model ( sm ) higgs at 7 @xmath0 at the lhc , presented in fig . [ production modes ] , are : gluon gluon fusion ( gg ) , dominant in the whole mass ranges ; vector boson fusion ( vbf , qqh ) , increasingly important for high masses and leading to a characteristic signature in the final state with two forward jets and associated production , with w and z boson or top quarks ( wz , zh , t@xmath5h ) , that is relevant in the low mass region and allows for easy triggering . concerning the decay modes , presented in fig .
[ decay modes ] , for low masses ( @xmath6 ) , h@xmath7b@xmath8 decay has the highest branching ratio , although this channel is experimentally challenging due to the huge qcd background present .
another important decay mode for low masses is h@xmath9 , accessible through vbf , and h@xmath10 , in which the higgs mass peak can be reconstructed with good resolution . for high masses ( @xmath11 ) ,
the h@xmath7t@xmath5 decay can be studied , however , it presents difficulties in the selection due to the small signal - to - noise ratio .
the h@xmath7ww@xmath12 and h@xmath7zz@xmath12 decay channels are powerful in the whole mass range .
the higgs decay into w bosons gives the earliest sensitivity and it is the dominant decay mode for a wide mass range .
higgs to zz@xmath12 has a very clean experimental signature with four leptons where a narrow mass peak is reconstructed above a smoothly varying background .
the higgs searches at cms during the lhc first period of data taking will overlap with the last runs of the tevatron experiments , cdf and d0 . by the end of 2011
, the tevatron is expected to have 10 @xmath1 at a center of mass energy of 1.96 @xmath0 compared with 1 @xmath1 at 7 @xmath0 for the lhc . to give a rough estimate of the expected performance of lhc with respect to tevatron the different parton luminosities of the two colliders need to be taken into account . in the high mass regime ( @xmath13 )
gluon gluon luminosity at lhc is more than 15 times higher than at tevatron .
being the dominant background to the main channels ( h@xmath7zz and h@xmath7ww ) mainly produced by q@xmath14 which rises relatively slowly it can be concluded that lhc will be competitive with 1 @xmath1 .
contrary in the low mass region the rise in the cross section is less pronounced and the q@xmath14 processes ( e.g. higgs - strahlung ) are not so favorable at lhc compared to at tevatron .
the signal to background ratio is also worsen by the dominant gluon fusion mechanism which enhances the main backgrounds channels ( e.g t@xmath5 , w / zb@xmath8 ) .
the previous considerations suggest to use the challenging @xmath15 mode .
the cms experiment utilizes a wide range of higgs decay channels and have public results available for different @xmath16 .
the higgs sensitivity at 7 @xmath0 with 1 @xmath1 , has been studied using projections made from the results obtained using monte carlo samples at different center of mass energies , namely 10 and 14 @xmath0 .
these projections are not new analysis done with 7 @xmath0 monte carlo samples and new detector simulation and reconstruction software .
they are done starting from public results at 10 and 14 @xmath0 .
the signal and background event counts have been re - scaled by the ratio of 7 @xmath0 to 14 @xmath0 cross - sections and then normalized to 1 @xmath1 .
no corrections for higher acceptance at smaller @xmath17 or for improvements in the reconstruction are applied .
systematic and statistical uncertainties have been also re - scaled conservatively .
the main higgs search channels relevant for the current running conditions of the lhc will be presented in the following .
the h@xmath7ww@xmath18ll@xmath19 decay channel is considered the discovery channel for a sm higgs boson in a wide mass range at the lhc .
the higgs to ww@xmath20 branching ratio is close to 1 in the 2@xmath21 mass range and the leptonic decay of the w bosons gives a clear experimental signature characterized by the two high p@xmath22 leptons with opposite charge and a small transverse opening angle .
missing transverse energy is also expected , due to the undetected neutrinos .
no central jet activity is characteristic of the gg@xmath7h process while two high rapidity jets are expected for the vbf process ( smaller cross - section ) . the backgrounds to consider in the analysis are all sources of real or fakes multi - lepton final states and missing e@xmath22 , like the irreducible continuum ww production ( plus other di - boson processes such as wz and zz ) , t@xmath5 process , drell - yan and w+jets , amongst others .
the cms analysis consider the three final states ee , @xmath23 and e@xmath24 . as no mass peak
can be reconstructed due to the presence of the neutrinos , the best knowledge of the backgrounds is mandatory .
this is achieved by using control regions and data - driven methods .
the systematic uncertainties are carefully addressed .
different approaches have been studied . a sequential cut - based analysis independently optimized for
the three considered final states in the 0-jet bin ( h+0j ) , a multivariate analysis considering all the three final states together and a dedicated study for the vbf channel .
for an integrated luminosity of 1 @xmath1 at 7 @xmath0 , as shown in fig .
[ ww exclusion ] and fig .
[ ww discovery ] , exclusion would be possible at 95% cl in this channel from around 140 to 180 @xmath3 , and discovery level sensitivity ( 5@xmath25 ) would be achieved for masses around 165 @xmath3 .
the h@xmath7zz@xmath18llll decay channel is known as the golden higgs decay , as it has the cleanest experimental signature for discovery with a narrow four - lepton invariant mass peak on top of a smooth background .
it is powerful in a wide high mass range , however it can be a challenge for higgs masses between 120 and 150 @xmath3 , where one of the z bosons is off - shell . in the analysis
two pairs of same flavor , opposite - sign leptons ( 4e , 4@xmath24 , 2e2@xmath24 ) , are selected and the higgs mass is reconstructed .
the main backgrounds are the irreducible zz@xmath12 , zb@xmath8 and t@xmath5 . due to the presence of two
non - isolated and displaced leptons the last two can be largely reduced with isolation and impact parameter cuts .
the rate of the zz is assessed from data z events . with an integrated luminosity of 1 @xmath1 at 7 @xmath0 , as shown in fig .
[ zz exclusion ] , the sm higgs boson can not be excluded anywhere in the entire mass range
. however , the higgs boson with a mass @xmath26 would be excluded , if a fourth generation of quarks exists .
indeed an extra doublet of quarks would make the gluon fusion production rate about 9 times larger , regardless of how massive the two extra 4@xmath27 generation quarks might be . in the low mass range , @xmath28
, h@xmath10 is a promising channel .
two high energy isolated photons in the final state allow for a mass peak reconstruction , but due to the small branching ratio , this channel is considered a high luminosity analysis .
the backgrounds are the irreducible @xmath15 , @xmath29+jets , jets and drell - yan .
the background can be assessed from the sidebands .
cms has proposed two complementary analyzes : a cut - based and an event - by - event kinematic likelihood ratio . for an integrated luminosity of 1 @xmath1 at 7 @xmath0 as shown in fig .
[ gammagamma exclusion ] the sm higgs boson can not be excluded anywhere in the entire mass range .
the case of a fermiophobic higgs would largely suppress the gluon fusion cross section enhancing at the same time the @xmath15 decay mode . under this condition and assuming a conservative selection it would be possible to improve the sm exclusion limit by a factor four or down to 110 @xmath3 .
[ combined exclusion ] shows the projected exclusion limits for a sm higgs by combining results of the h@xmath7ww@xmath12 , h@xmath7zz@xmath12 , and h@xmath10 channels .
the expected exclusion mass range is @xmath30 .
the higgs boson with a mass @xmath31 would be excluded , if a fourth generation of heavy quarks exists .
the main channel for minimal super symmetric standard model ( mssm ) higgs searches in the first period of running of the lhc will be heavy neutral mssm higgs @xmath32 produced via b quark association with subsequent decay to @xmath33 , b@xmath34 , @xmath35 .
the main backgrounds are z+b@xmath8/c@xmath36/jets and t@xmath5 . in the proposed analysis
final states with isolated pairs of @xmath33 decaying to hadrons and leptons are selected .
missing e@xmath22 is also required as well as one b - tagged jet and a veto on extra jets in the event .
the analysis is performed by counting events in the @xmath37 invariant mass window , reconstructed by collinear approximation where the @xmath33 decay products are assumed to be in the same direction as the @xmath33 itself .
[ mssm discovery ] shows the projected discovery and exclusion contours in the mssm ( @xmath38 , tan@xmath39-plane ) for the search for a neutral mssm higgs bosons in the pp@xmath7b@xmath34 , b@xmath40 channel with the cms experiment . at @xmath41 ,
discovery is possible for tan@xmath42 and the exclusion limit is expected to reach down to tan@xmath43 .
the performance of the cms detector in collision data has been very good since the beginning of the data taking allowing to produce the first results after few hours . at 7 @xmath0 , with enough luminosity ( 1 @xmath1 )
, the cms experiment will begin to explore a sizable range of higgs mass , reaching sm higgs discovery sensitivity for masses between 160 and 170 @xmath3 and exclusion between 140 and 200 @xmath3 , while low mass sm higgs searched will require higher center of mass energy and integrated luminosity .
barate , r. et al . , search for the standard model higgs boson at lep , phys .
b565 ( 2003 ) 61 - 75 .
the tevnph working group , combined cdf and d0 upper limits on standard model higgs boson production with up to 6.7@xmath1 of data , fermilab - conf-10 - 257-e , 2010 .
the cms collaboration , the cms experiment at the cern lhc , jinst 3:s08004 , 2008 .
the cms collaboration , the cms physics reach for searches at 7 tev , http://cdsweb.cern.ch/record/1264009 , cms note 2010/008 , 2010 .
n. berici et al . , the fourth standard model family and the competition in standard model higgs boson search at tevatron and lhc , http://arxiv.org/abs/0908.2653 , eur .
j. c66:119 - 126 , 2010 . | an overview on the prospects for the higgs boson searches with the cms detector is presented .
projections have been made to estimate the potential to a possible discovery or exclusion of the higgs boson during the run at a center of mass energy of 7 @xmath0 at lhc , with a recorded integrated luminosity of approximately 1 @xmath1 , conditions expected by the end of 2011 . |
one of the central issues in condensed matter physics is to understand the interplay between diverse competing phases : for example , antiferromagnetism and superconductivity in the high-@xmath7 cuprates @xcite and charge - density - wave ( cdw ) order and superconductivity in transition metal dichalcogenides @xcite .
very recently , the cdw compound tise@xmath3 has been successfully doped by a cu intercalation between adjacent layers @xcite .
this allows for tuning the electronic properties of cu@xmath8tise@xmath9 by controlling the cu concentration @xmath10 .
the cu doping drives the system to a superconducting state around @xmath11 . in this regard
, the cu@xmath8tise@xmath9 system provides a unique opportunity to study the competition between the cdw and superconducting phases in a controlled way .
the cu@xmath8tise@xmath9 compound has a layered structure with trigonal symmetry where the @xmath12 atoms are in octahedral coordination with se .
the cu atoms occupy the positions between the tise@xmath9 layers which leads to an expansion of the lattice parameters @xcite .
the parent compound tise@xmath3 is one of the first cdw compounds , where a commensurate cdw at @xmath13 k with a wave vector @xmath14 was reported @xcite .
the cdw transition is not driven by the nesting of a fermi surface .
instead it is either due to a transition from a small indirect gap into a larger gap state with a slightly different location in the brillouin zone or due to a particle - hole symmetry @xcite .
the cdw transition in tise@xmath3 can be suppressed continuously by a cu doping while superconductivity emerges with a maximum @xmath15 k at @xmath16 .
the electronic structure has been investigated by band structure calculations and arpes measurements @xcite .
the remarkable feature is that the cdw order parameter competes microscopically with superconductivity in the same band and that the parent compound tise@xmath9 is a correlated semiconductor and the cu doping enhances the density of states and raises the chemical potential .
various measurements were also performed to further investigate the role of the cu doping @xcite .
thermal conductivity suggests that this system exhibits conventional _ s_-wave superconductivity due to the absence of a residual linear term @xmath17 at very low temperatures @xcite .
the weak magnetic field dependence of thermal conductivity indicates a single gap that is uniform across the fermi surface .
furthermore , it is suggested that the @xmath18 se band is below the fermi level and superconductivity is induced because of the cu doping into the @xmath19 ti bands @xcite .
an optical spectroscopy investigation on cu@xmath20tise@xmath9 reveals that the compound has a low carrier density and has an anomalous metallic state because of the substantial shift of the screened plasma frequency @xcite .
this is corroborated by the temperature - independent hall coefficient @xmath21 observed in heavily cu - doped samples @xcite .
raman scattering measurements as a function of doping suggest that the @xmath10-dependent mode softening is associated with the reduction of electron - phonon couplings and the presence of a quantum critical point within the superconducting region @xcite . in this paper
we address the issue of electronic correlations in this system as a function of cu doping using @xmath0se and @xmath1cu nmr .
we studied two doping levels of cu@xmath8tise@xmath9 which exhibit cdw and/or superconductivity at zero field : @xmath4 ( @xmath22 k , @xmath23 k ) and @xmath24 ( @xmath25 k ) . at high magnetic fields
, superconductivity is suppressed and we find that the spin - lattice relaxation rate is described by a modified korringa relation with a doping - dependent korringa factor .
the large values of the extracted interaction parameters signify the importance of electron - electron interactions and/or the se _ p_- and ti _
d_-band contribution for these materials .
this is a key ingredient to the superconductivity for these compounds .
cu nmr spectrum showing the two peaks from the sample and the copper coil .
( b ) temperature dependence of @xmath1cu nmr knight shift at 8 t for x=0.05 and 0.07 .
the arrows indicate the transition temperatures at zero field .
( c ) @xmath1cu spin - lattice relaxation rate @xmath5 versus temperature in cu@xmath20tise@xmath3 at 8 t for * * b**@xmath26_ab_-plane .
the solid line is the temperature dependence of @xmath27 for @xmath1cu metal .
inset : representative longitudinal magnetization recovery curves at 10 k and 80 k fitted by the master equation described in the text.,width=3 ]
single crystals of cu@xmath8tise@xmath9 were grown by chemical iodine - vapor transport method and characterized by x - ray , transport , and susceptibility measurements @xcite .
the plate - like samples with average thickness of @xmath28 m along the @xmath29-axis were cut in rectangular shapes with approximate dimensions @xmath30 .
parallel stacks of rectangular samples were placed in a rectangular - shaped copper coil for a better filling factor with the magnetic field b applied along the _ ab _ plane ( parallel with the plates ) for nmr measurements .
two doping levels ( @xmath31 ) of cu@xmath8tise@xmath9 were studied under the same conditions . @xmath0se and
@xmath1cu nmr spectra , knight shift , and spin - lattice relaxation rates were obtained using a homebuilt nmr spectrometer equipped with a high homogeneity 15 t sweepable magnet .
@xmath32 was measured by the recovery of the spin echo intensity as a function of time after saturation .
se nmr spectra of cu@xmath2tise@xmath3 ( @xmath24 ) at 11.1 t for * * b**@xmath26_ab_-plane .
( b ) the corresponding @xmath0se nmr shift as a function of temperature.,width=3 ] figure [ fig:1](a ) shows the representative @xmath1cu ( spin-3/2 , @xmath33 mhz / t , 69.1 % natural abundance ) nmr spectrum .
the @xmath1cu nmr spectrum consists of two peaks , which are assigned to the signals from the sample and copper coil as indicated in the figure .
the symmetric , sharp spectral shape of the nmr signal indicates that the cu site is not affected by cdw modulation .
the insensitivity of @xmath1cu nuclei to an electronic state should be ascribed to the fact that the cu atoms are located in between the tise@xmath9 layers which are bounded by van der waals interaction .
quadrupolar effects are expected for cu nuclei which occupies a site of non - cubic symmetry but the satellites are not observable .
the quadrupole frequency @xmath34 for @xmath1cu is usually on the order of 20 - 30 mhz @xcite , thus the quadrupolar satellites are probably too broad and washed out due to disorder since cu is dilute in this system . as such
, the recovery of the longitudinal nuclear magnetization is appropriately fitted by the master equation for a spin-3/2 central transition using a single saturating pulse : @xmath35 $ ] @xcite ( see fig.[fig:1](c ) inset ) .
the temperature dependence of @xmath1cu nmr knight shift and spin - lattice relaxation rate @xmath5 at 8 t is summarized in figures [ fig:1](b ) and ( c ) .
the @xmath1cu knight shift data exhibit very weak temperature dependence with @xmath36 % for @xmath4 and @xmath37 % for @xmath24 , whereas in comparison @xmath1cu metal has @xmath38 % @xcite .
there is no hint that its temperature dependence scales to the magnetic susceptibility .
in addition , no anomalies are accompanied by the cdw transition .
this might be due to the fact that the external field of 11.1 t fully suppresses the residual cdw state by shifting the chemical potential and/or changing the electronic density of states of ti 3__d _ _ bands .
the @xmath1cu spin - lattice relaxation rate for @xmath24 varies linearly with temperature , obeying korringa behavior with a slope approximately 1/18 times that of a pure @xmath1cu metal . since the se atoms are located on the layers , they probe the fermi surface and provide more information on electronic correlations in this material .
figure [ fig:2 ] displays temperature dependence of @xmath0se ( spin-1/2 , @xmath39 mhz / t , 7.5 % natural abundance ) spectra and nmr shift of the @xmath24 compound at b=11.1 t. the @xmath0se nmr signal shows the symmetric , sharp spectral shape as well .
the nmr shift is constant from 280 k to 150 k but increases monotonically upon further cooling .
we note that the room temperature nmr shift @xmath40 for the cu@xmath2tise@xmath3 samples ( @xmath41 for @xmath42 and @xmath43 for @xmath44 ) are close to the reported nmr shift values of the selenide compounds tise@xmath3 ( @xmath45 ) @xcite and tase@xmath3 ( @xmath46 ) @xcite . in figure [ fig:3 ]
we compare temperature dependence of @xmath5 of cu@xmath2tise@xmath3 ( @xmath47 ) single crystals with that of a @xmath0se metal @xcite and powdered tise@xmath9 @xcite .
a single - exponential fitting equation @xmath48 $ ] was used to extract the @xmath49 values from the magnetization recovery curves .
it is apparent that an increase of the cu doping concentration leads to an enhancement of the relaxation rate as well as an increase of the slope of the korringa - like behavior .
the inset of fig .
[ fig:3 ] shows the non - linear @xmath5 vs t behavior of the parent compound tise@xmath9 and the subsequent linear behavior upon cu doping which suppresses the cdw state . in order to understand the underlying mechanism of the observed @xmath5 behaviors we invoke a model involving a korringa factor @xmath50 due to the effect of electron - electron interaction in metals and alloy compounds developed by moriya @xcite and later refined by narath and weaver @xcite . in simple metals ,
the dominant relaxation mechanism of nuclear spins is due to their interaction with conduction electron spins .
this is given by the korringa relation @xmath51 wherein only fermi contact interaction is taken into account for noninteracting electrons .
however electron - electron interactions can not be neglected especially for metal alloys .
the coulomb interaction has an effect on the magnetic susceptibility which is proportional to the knight shift k@xmath52 @xcite . the internal magnetic field experienced by a nucleus
is given by @xmath53 where @xmath54 is the average internal magnetic field and @xmath55 is the fluctuating magnetic field .
the nuclear spin lattice relaxation rate is a measure of the perpendicular fluctuating magnetic field @xmath56 given by @xmath57 . in metals with strictly non - interacting electron gas , the relaxation rate is proportional to the temperature and square of the knight shift @xcite , known as the korringa relation : se spin - lattice relaxation rates @xmath5 for @xmath0se metal ( ref.@xcite ) , cu@xmath2tise@xmath3 ( x=0.05 , 0.07 ) single crystals from this work , and the powdered parent compound tise@xmath3 from ref . @xcite .
inset : log - log plot of @xmath6 vs _ t_.,width=3 ] @xmath58 where @xmath59 , @xmath60 are the nuclear and electron gyromagnetic ratios , respectively .
moriya modified the korringa relation by incorporating the factor @xmath50 that includes the effect of a @xmath61-function - potential electron - electron interaction @xcite : @xmath62 where the korringa factor @xmath63^{2}}$ ] is a function of the interaction parameter @xmath64 , a measure of electron - electron interaction . here
@xmath65 and @xmath66 $ ] from the free electron approximation with @xmath67 where @xmath68 is a constant , _
n _ is the number of electrons , _ q _ is the wavevector , and @xmath69 , @xmath70 are the fermi energy and wave number , respectively @xcite .
the value of @xmath50 ranges from 0 to 1 . for @xmath71 , @xmath72 which gives us eqn.([eq : korr1 ] ) .
a corrected version was later formulated by narath appropriate for alkali and noble metals where the function potential has a non - zero interaction range less than the atomic radius @xcite . before delving into the korringa factor calculation , we note that the orbital shift contribution can not , in the present case , be accurately separated from the spin part . typically , the orbital contribution can be unambiguously determined in the superconducting state where the spin part of the shift vanishes at low temperature due to pairing .
however , access of nmr in the superconducting state , believed to be of bcs type @xcite , is hampered by the low upper critical field ( @xmath73 t ) @xcite of these compounds .
we can however make some estimates based on the nmr data from the parent compound tise@xmath3 @xcite .
the total nmr shift @xmath40 can be expressed as : @xmath74 where @xmath75 is the _
s_-electron contact term , @xmath76 is the _
p_-electron core polarization term , @xmath77 is the van vleck orbital contribution , and @xmath78 is the chemical shift .
experimental estimates from a previous @xmath0se nmr study on tise@xmath3 @xcite at room temperature yield @xmath79 , @xmath80 , @xmath81 thus the temperature - independent part @xmath82 is 0.071% .
since cu doping into the tise@xmath3 layers increases the density of states ( dos ) at the fermi level , this will increase @xmath75 and @xmath76 , but the terms @xmath77 and @xmath78 should be unaffected since they do not depend on dos .
we can then estimate the spin - part of the total shift for cu@xmath2tise@xmath3 system to be approximately ( @xmath83 ) .
this gives us the high temperature knight shift values of 0.07% for @xmath84 ( @xmath85 ) and 0.079% for @xmath86 ( @xmath87 ) doping level ( see table [ tab : kalpa ] ) .
.korringa factor @xmath50 and interaction parameter @xmath64 due to electron - electron interactions based on @xmath0se nmr data in cu@xmath8tise@xmath9 ( @xmath88 ) and tise@xmath9 @xcite at room temperature . [ cols="^,^,^,^,^,^",options="header " , ] [ tab : kalpa ] vs interaction parameter @xmath64 from models set by moriya and a modified version from ref .
the arrows indicate that the corresponding interaction parameter @xmath64 for cu@xmath8tise@xmath9 ( @xmath89 ) and tise@xmath9 ( @xmath90 ) @xcite based on the k(@xmath64 ) values summarized in table i.,width=3 ] the korringa factors @xmath50 for cu@xmath2tise@xmath3 ( @xmath91 ) system were calculated using eqn.([eq : korr ] ) . by comparing our data to the enhanced korringa factor @xmath50 versus interaction parameter @xmath64 calculated from ref.@xcite [ see fig .
[ fig:4 ] ] , we are able to determine the interaction parameter @xmath64 .
the calculated values of @xmath50 and @xmath64 are summarized in table [ tab : kalpa ] .
first we revisit the @xmath1cu data in fig.[fig:1 ] which yield @xmath92 , @xmath93 for @xmath94 % and @xmath95 , @xmath96 for a pure @xmath1cu metal .
the fact that @xmath97 indicates that the exchange enhancement in the cu metal is more dominant than in the cu@xmath20tise@xmath3 compound . for the @xmath0se data , @xmath98 , @xmath99 for @xmath100 % and @xmath101 , @xmath102 for @xmath94 % doping .
the large value of @xmath64 highlights the significance of electron - electron interactions . in comparison ,
data from the parent compound tise@xmath3 @xcite at room temperature ( see table [ tab : kalpa ] ) yield @xmath103 , @xmath104 .
the small value of @xmath50 for tise@xmath3 may suggest strong ferromagnetic correlations @xcite but this compound is viewed as a correlated semiconductor which may warrant its deviation from this theoretical interpretation @xcite . for cu@xmath8tise@xmath9 system , the result @xmath105 indicates that cu intercalation monotonically increases the exchange - enhancement in the se site .
in addition to correlation effects , the contribution of the density of states ( dos ) should also be considered in the observed relaxation behavior . here
cu@xmath2tise@xmath3 compounds evolve into better metals as evidenced by decreasing resistivity and increasing susceptibility as cu doping is increased @xcite .
this points to the fact that increasing cu content introduces carriers into the tise@xmath3 conduction band , increasing the dos at the fermi level @xcite thereby enhancing the relaxation rates in agreement with our experimental data .
we have presented @xmath0se and @xmath1cu nmr studies of cu@xmath2tise@xmath3 ( @xmath106 ) .
we find that @xmath1cu knight shift shows weak variations with temperature and the spin - lattice relaxation rate @xmath6 follows a modified korringa relation .
the observed korringa factor was discussed in terms of moriya s exchange - enhanced relaxation model and density of states contribution .
our results suggest that enhanced electronic correlations and se _ p_- and ti _ d_-band play a key role in inducing superconductivity in the cu@xmath2tise@xmath3 compounds .
this work is supported by the us national science foundation ( nsf ) dmr grant 0602859 , priority research center program ( 2009 - 0093817 ) and basic science research program ( 2009 - 0077376 ) of korea nrf , and the national basic research program for china 2006cb922005 .
the experiments were performed at the national high magnetic field laboratory which is supported by nsf dmr-0084183 , by the state of florida , and by the department of energy . | we report @xmath0se and @xmath1cu nuclear magnetic resonance ( nmr ) investigation on the charge - density - wave ( cdw ) superconductor cu@xmath2tise@xmath3 ( @xmath4 and 0.07 ) . at high magnetic fields where superconductivity is suppressed ,
the temperature dependence of @xmath0se and @xmath1cu spin - lattice relaxation rates @xmath5 follow a linear relation .
the slope of @xmath6 vs _ t _ increases with the cu doping .
this can be described by a modified korringa relation which suggests the significance of electronic correlations and the se 4__p__- and ti 3__d__-band contribution to the density of states at the fermi level in the studied compounds . |
the determination of the quadrupole excitation strength @xmath7 in the region of the @xmath8 resonance has been the aim of considerable experimental and theoretical activities . within the harmonic oscillator quark model ,
the @xmath9 and the nucleon are both members of the symmetrical 56-plet of @xmath10 with orbital momentum @xmath11 , positive parity and a gaussian wave function in space . in this approximation the @xmath9 may only be excited by a magnetic dipole transition @xmath12 @xcite . however , in analogy with the atomic hyperfine interaction or the forces between nucleons , also the interactions between the quarks contain a tensor component due to the exchange of gluons .
this hyperfine interaction admixes higher states to the nucleon and @xmath9 wave functions , in particular @xmath13-state components with @xmath14 , resulting in a small electric quadrupole transition @xmath7 between nucleon and @xmath9 @xcite .
in addition quadrupole transitions are possible by mesonic and gluonic exchange currents @xcite .
therefore an accurate measurement of @xmath7 is of great importance in testing the forces between the quarks and , quite generally , models of nucleons and isobars . the e2/m1 ratio , @xmath15 has been predicted to be in the range @xmath16 in the framework of constituent quark @xcite , relativized quark @xcite and chiral bag models @xcite .
considerably larger values have been obtained in skyrme models @xcite .
a first lattice qcd calculation resulted in a small value with large error bars @xmath17 @xcite .
however , the connection of the model calculations with the experimental data is not evident .
clearly , the @xmath9 resonance is coupled to the pion - nucleon continuum and final - state interactions will lead to strong background terms seen in the experimental data , particularly in case of the small @xmath18 amplitude .
the question of how to correct the experimental data to extract the properties of the resonance has been the topic of many theoretical investigations .
unfortunately it turns out that the analysis of the small @xmath18 amplitude is quite sensitive to details of the models , e.g. nonrelativistic vs. relativistic resonance denominators , constant or energy - dependent widths and masses of the resonance , sizes of the form factor included in the width etc . in other words , by changing these definitions the meaning of resonance vs. background changes , too . in order to study the @xmath9 deformation ,
pion photoproduction on the proton has been measured by the legs collaboration @xcite at brookhaven and by the a2 collaboration @xcite at mainz using transversely polarized photons , i.e. by measuring the polarized photon asymmetry @xmath19 .
in particular , the cross section @xmath20 for photon polarization in the reaction plane turns out to be very sensitive to the small @xmath18 amplitude .
assuming for simplicity that only the @xmath21-wave multipoles contribute , the differential cross section is @xmath22 where @xmath23 and @xmath24 are the pion and photon momenta and @xmath25 is the pion emission angle in the @xmath26 frame . neglecting the ( small ) contributions of the roper multipole @xmath27 , one obtains @xcite @xmath28 because the isospin @xmath29 amplitudes strongly dominate the cross section for @xmath30 production . in order to obtain the c2/m1 ratio and the form factors as functions of @xmath4
pion electroproduction has been studied . at mainz ,
bonn , bates and jlab different experiments have been performed , without polarization as well as single and double polarization .
while the experiments at mainz and bates measured at @xmath31 gev@xmath5 in order to get the c2/m1 ratio close to the photon point , the jlab experiment was motivated by the possibility of determining the range of momentum transfers where perturbative qcd ( pqcd ) would become applicable . in the limit of @xmath32 , pqcd predicts @xcite that only helicity - conserving amplitudes contribute , leading to @xmath33 and @xmath34 .
starting from fixed-@xmath0 dispersion relations for the invariant amplitudes of pion photoproduction , the projection of the multipole amplitudes leads to a well known system of integral equations , @xmath35 where @xmath36 stands for any of the multipoles @xmath37 and @xmath38 for the corresponding ( nucleon ) pole term .
the kernels @xmath39 are known , and the real and imaginary parts of the amplitudes are related by unitarity . in the energy region below two - pion threshold , unitarity is expressed by the final state theorem of watson , @xmath40 where @xmath41 is the corresponding @xmath42 phase shift and @xmath43 an integer .
we have essentially followed the method of schwela et al @xcite to solve eq .
( [ inteq ] ) with the constraint ( [ watson ] ) .
in addition we have taken into account the coupling to some higher states neglected in that earlier reference . at the energies above two - pion threshold up to @xmath44 gev , eq .
( [ watson ] ) has been replaced by an ansatz based on unitarity @xcite . finally , the contribution of the dispersive integrals from @xmath45 gev to infinity has been replaced by @xmath0-channel exchange , parametrized by certain fractions of @xmath46- and @xmath47-exchange .
furthermore , we have to allow for the addition of solutions of the homogeneous equations to the coupled system of eq .
( [ inteq ] ) .
the whole procedure introduces 9 free parameters , which have been determined by a fit to the data .
@xcite in figure 1 we show the @xmath48 multipoles @xmath49 and @xmath50 . our dispersion theoretical analysis ( solid line )
agrees very well with our single - energy fit and with the single - energy fit of beck et al . @xcite .
the only systematic deviation becomes visible in the electric multipole above the resonance position .
this can be due either to our truncation of partial waves or to systematics in the experiment at the highest energies . in a new experiment
a full angular coverage of the differential cross section for @xmath51 over a wide range of energy from threshold up to @xmath52 has been taken and in the near future a new and very precise multipole analysis should also clarify this small deviation . according to the watson theorem , at least up to the two - pion threshold ,
the ratio @xmath53 is a real quantity .
however , it is not a constant but even a rather strongly energy dependent function .
if we determine the resonance position as the point , where the phase @xmath54 , we can define the so - called `` full '' ratio @xmath55 we note that this ratio is identical to the ratio obtained with the @xmath56-matrix at the @xmath56-matrix pole @xmath57 .
this can be seen by using the relation between the @xmath58- and the @xmath56-matrix , @xmath59 therefore , at @xmath57 we find @xmath60 the recent , nearly model - independent value of the mainz group at @xmath61 mev is @xmath62 @xcite is in excellent agreement with our dispersion theoretical calculation that gives @xmath63 , see table 1 . .e2/m1 ratios for @xmath4=0 from different analyses . [
cols="<,<",options="header " , ] we have also re - analyzed older desy @xcite data measured at @xmath64 and found significantly different results for the ratios compared to a previous analysis @xcite . however , since these data also give a @xmath65 value @xmath66 below our fit curves in figure 2 , we did not include them in our fits of the ratios . in a similar way we also analyzed recent bates measurements @xcite for unpolarized differential cross sections , @xmath67 response function and @xmath68 asymmetry for @xmath69 at @xmath70 and obtained the following preliminary values for the e2/m1 and c2/m1 ratios @xmath71 finally , in a double polarization experiment at mainz @xcite , measuring the recoil polarization of the proton , @xmath72 at @xmath73 , a preliminary value of @xmath74
could be extracted in a rather model - independent way from the x - component of the recoil polarization @xmath75 , which is very sensitive to the resonant @xmath76 multipole .
our extracted values for @xmath77 and @xmath78 and a comparison with the results of ref .
@xcite are presented in table 2 and shown in figure 3 . the main difference between our results and those of ref .
@xcite is that our values of @xmath77 show a clear tendency to cross zero and change sign as @xmath4 increases .
this is in contrast with the results obtained in the original analysis @xcite of the data which concluded that @xmath77 would stay negative and tend toward more negative values with increasing @xmath4 .
in the framework of fixed - t dispersion relations with the new and very precise data obtained at mami in mainz we have obtained a new partial wave analysis for pion photoproduction . the uncertainties in most multipoles could be considerably improved compared to previous analyses . at the resonance position , where the phase passes @xmath79
, we obtain an e2/m1 ratio of @xmath80 . at the pole in the complex plane
we obtain the ratio of the resonant electric and magnetic multipoles as @xmath81 .
this is a further model - independent ratio that can be determined in any analysis or calculation of pion photoproduction . for pion electroproduction ,
we have re - analyzed the recent jlab data for electroproduction of the @xmath2 resonance via @xmath82 with two models for pion electroproduction , both of which give excellent descriptions of the existing data .
in contrast to previous findings , our models indicate that @xmath77 , starting from a small and negative value at the real photon point , actually exhibits a clear tendency to cross zero and change sign as @xmath4 increases . it will be most interesting to have data at yet higher momentum transfer in order to see whether such a trend continues , which would be a sign for a rather slow approach towards the pqcd region .
furthermore , the absolute value of @xmath78 is strongly increasing , which indicates that the pqcd prediction of @xmath83 is not yet reached .
we wish to thank r. beck , r. leukel , h. schmieden , r. gothe and c. papanicolas for their contribution on the experimental data .
this work was supported in part by nsc under grant no .
nsc89 - 2112-m002 - 038 , by deutsche forschungsgemeinschaft ( sfb443 ) and by a joint project nsc / dfg tai-113/10/0 .
99 c. m. becchi and g. morpurgo , phys
* 17 * ( 1965 ) 352 .
r. koniuk and n. isgur , phys .
d * 21 * ( 1980 ) 1868 .
s. s. gershteyn and g. v. dzhikiya , sov .
. phys . * 34 * ( 1981 ) 870 .
n. isgur , g. karl and r. koniuk , phys .
d * 25 * ( 1982 ) 2394 .
a. j. buchmann , e. hernandez and a. faessler , phys .
c * 55 * ( 1997 ) 448 .
d. drechsel and m. m. giannini , phys . lett .
* 143 * b ( 1984 ) 329 .
s. capstick and g. karl , phys .
d * 41 * ( 1990 ) 2767 .
s. capstick , phys . rev .
d * 46 * ( 1992 ) 2864 .
k. bermuth , d. drechsel , l. tiator and j. b. seaborn , phys .
d * 37 * ( 1988 ) 89 .
a. wirzba and w. weise , phys .
b * 188 * ( 1987 ) 6 .
d. b. leinweber , t. draper and r. woloshyn , contr .
baryons 92 , ( 1992 ) p. 29 .
g. blanpied _ et al .
* 79 * ( 1997 ) 4337 .
r. beck _ et al .
_ , phys . rev
* 78 * ( 1997 ) 606 .
s.j . brodsky and g.p .
lepage , phys .
d * 23 * , 1152 ( 1981);c.e .
carlson and j.l .
poor , phys .
d * 38 * ( 1988 ) 2758 .
d. schwela and r. weizel , z. physik * 221 * ( 1969 ) 71 . w. pfeil and d. schwela , nucl . phys . *
b45 * ( 1972 ) 379 .
o. hanstein , d. drechsel , and l. tiator , nucl . phys .
* a632 * ( 1998 ) 561 .
r. beck _ et al .
c * 61 * ( 2000 ) 035204 .
arndt , i.i .
strakovsky and r.l .
workman , phys .
c * 56 * ( 1997 ) 577 .
davidson and n.c .
mukhopadhyay , phys .
* 79 * ( 1997 ) 4509 .
. groom _ et al .
_ , eur . phys
c * 15 * ( 2000 ) 1 .
et al . _ , phys .
c * 59 * ( 1999 ) 1059 .
g. hhler and a. schulte , @xmath42 newsletter * 7 * ( 1992 ) 94 .
yang , j. phys .
g * 11 * ( 1985 ) l205 .
d. drechsel , o. hanstein , s.s .
kamalov and l. tiator , nucl . phys .
* a645 * ( 1999 ) 145 ; http://www.kph.uni-mainz.de/maid .
kamalov and s.n .
yang , phys .
lett . * 83 * ( 1999 ) 4494 .
arndt , i.i .
strakovsky and r.l .
workman , phys .
c * 53 * ( 1996 ) 430 .
et al . _ ,
. lett . * 82 * ( 1999 ) 45 .
j. m. laget , nucl . phys .
* a481 * ( 1988 ) 765 .
gothe , prog .
* 44 * ( 2000 ) 185 .
r. haidan , report no .
desy - f21 - 79 - 03 , 1979 ( unpublished ) . v. burkert and l. elouadrhiri , phys .
* 75 * ( 1995 ) 3614 .
s. kamalov , s.n .
yang , d. drechsel , o. hanstein , l. tiator , nucl - th/0006068 .
c. mertz _ et al .
_ , nucl - ex/9902012 . h. schmieden , nucl
* a663 * ( 2000 ) 21 . | a partial wave analysis of pion photoproduction has been obtained in the framework of fixed-@xmath0 dispersion relations valid from threshold up to 500 @xmath1 . in the resonance region
we have precisely determined the electromagnetic properties of the @xmath2 resonance , in particular the e2/m1 ratio @xmath3 . for pion electroproduction recent experimental data from mainz ,
bates and jlab for @xmath4 up to 4.0 ( gev / c)@xmath5 have been analyzed with two different models , an isobar model ( maid ) and a dynamical model .
the e2/m1 ratios extracted with these two models show , starting from a small and negative value at the real photon point , a clear tendency to cross zero , and become positive with increasing @xmath4 .
this is a possible indication of a very slow approach toward the pqcd region .
the c2/m1 ratio near the photon point is found as @xmath6 . at high @xmath4 the absolute value of the ratio is strongly increasing , a further indication that pqcd is not yet reached . |
_ dehn surgery _ can be regarded as an operation to make a ` new ' 3-manifold from a given one .
of course the trivial dehn surgery leaves the manifold unchanged , but ` most ' non - trivial ones would change the topological type .
in fact , gordon and luecke showed as the famous result in @xcite that any non - trivial dehn surgery on a non - trivial knot in the 3-sphere @xmath6 never yields @xmath6 .
as a natural generalization , the following conjecture was raised . * cosmetic surgery conjecture * ( ( * ? ? ?
* problem 1.81(a ) ) ) : two surgeries on inequivalent slopes are never purely cosmetic .
here we say that two slopes are _ equivalent _ if there exists a homeomorphism of the exterior of a knot @xmath7 taking one slope to the other , and two surgeries on @xmath7 along slopes @xmath8 and @xmath9 are _ purely cosmetic _ if there exists an orientation preserving homeomorphism between the pair of the surgered manifolds .
the cosmetic surgery conjecture for `` chirally cosmetic '' case is not true : there exist counter - examples given by mathieu @xcite .
in fact , for example , @xmath10- and @xmath11-surgeries on the right - hand trefoil knot in @xmath6 yield orientation - reversingly homeomorphic pairs of 3-manifolds for any non - negative integer @xmath12 , and it can be shown that such pairs of slopes are inequivalent . that is to say , the trefoil knot admits chirally cosmetic surgery pairs along inequivalent slopes . in this paper
, we consider cosmetic surgeries on a well - known class of knots in @xmath6 , the _ two - bridge knots_. first , by using known results , we have the following in section [ 2-bridge ] . [ atmost9 ] all the two - bridge knots of at most @xmath13 crossings other than @xmath0 $ ] admit no purely cosmetic surgery pairs . here
the knot @xmath3 in the rolfsen s knot table is the two - bridge knot of the schubert form @xmath14 and the conway form @xmath15 $ ] . in view of this , let us focus on the knot @xmath3 .
previously , for the same reason , the first author considered this knot in @xcite , and it was shown that some pairs of surgeries give distinct manifolds . in this paper , for a family of knots including the knot @xmath3 , we have the following . [ mainthm ]
let @xmath16 be a two - bridge knot of the conway form @xmath1 $ ] with @xmath2 .
then @xmath16 admits no cosmetic surgery pairs yielding homology 3-spheres , i.e. , any @xmath17- and @xmath18-surgeries on @xmath16 are not purely cosmetic for @xmath19 . in other words ,
all the homology 3-spheres obtained by dehn surgeries on @xmath16 are mutually distinct .
[ rmk2 ] this can not be achieved by using known invariants ; the ( original ) casson invariant and the @xmath20-invariant defined by ozsvth - szab in @xcite , and the correction term in heegaard floer homology .
see section [ alexpoly ] for details .
our advantage in this paper is to use the @xmath5 version of the casson invariant .
very roughly speaking , for a closed orientable 3-manifold @xmath21 , the @xmath5 casson invariant @xmath22 is defined by counting the ( signed ) equivalence classes of representations of the fundamental group @xmath23 in @xmath5 . based on the method to enumerate the boundary slopes for two - bridge knots developed in @xcite , we give calculations of the @xmath5 casson invariant for the knots @xmath16 s .
the calculations will be given in section [ calculation ] . before that the formulae and the method used in the calculations will be explained in section [ sl(2,c)casson ] .
practically our method can be applied further .
however it seems not enough to prove that all the @xmath16 s have no purely cosmetic surgery pairs .
here we recall basic definitions and terminology about dehn surgery . a _ dehn surgery _
is the following operation for a given knot @xmath7 ( i.e. , an embedded circle ) in a 3-manifold @xmath24 .
take the exterior @xmath25 of @xmath7 ( i.e. , the complement of an open tubular neighborhood of @xmath7 in @xmath24 ) , and then , glue a solid torus to @xmath25 .
let @xmath26 be the slope ( i.e. , an isotopy class of a non - trivial unoriented simple loop ) on the peripheral torus of @xmath7 in @xmath24 which is represented by the curve identified with the meridian of the attached solid torus via the surgery .
then , by @xmath27 , we denote the manifold which is obtained by the dehn surgery on @xmath7 , and call it the 3-manifold obtained by dehn surgery on @xmath7 along @xmath26 . in particular , the dehn surgery on @xmath7 along the meridional slope is called the _
trivial _ dehn surgery . when @xmath7 is a knot in @xmath6 , by using the standard meridian - longitude system , slopes on the peripheral torus are parametrized by rational numbers with @xmath28 .
thus , when a slope @xmath26 corresponds to a rational number @xmath29 , we call dehn surgery along @xmath26 _ @xmath29-surgery _ , and use @xmath30 in stead of @xmath27 .
for two - bridge knots , we use standard definitions based on @xcite .
see also @xcite . to show proposition [ atmost9 ] , we use the following two known results .
one ingredient is the casson invariant of 3-manifolds introduced by casson . by using the casson invariant ,
boyer and lines in @xcite proved that a knot @xmath7 in @xmath6 satisfying @xmath31 has no cosmetic surgeries . here
@xmath32 denotes the ( symmetrized ) alexander polynomial for @xmath7 .
that is , @xmath32 satisfies that @xmath33 and @xmath34 .
the other one is the following excellent result recently obtained by ni and wu in @xcite .
suppose that @xmath7 is a non - trivial knot in @xmath6 and @xmath35 are two distinct slopes such that the surgered manifolds @xmath36 are orientation - preservingly homeomorphic .
then @xmath37 satisfy that ( a ) @xmath38 , ( b ) @xmath39 for @xmath40 , ( c ) @xmath41 , where @xmath20 is the invariant defined by ozsvth - szab in @xcite .
this result is obtained by using heegaard floer homology .
we remark that , for alternating knots , @xmath42 holds ( * ? ? ?
* theorem 1.4 ) , where @xmath43 denotes the signature of @xmath7 .
now proposition [ atmost9 ] follows from table [ table ] . to fill the table
, we use the values given in _ knotinfo
also we can use the facts that the half of @xmath44 is equal to the second coefficient of the conway polynomial .
this well - known fact is due to casson , and , for details , see @xcite and ( * ? ? ?
* section 1 ) for example .
.two - bridge knots of at most 9 crossings with trivial @xmath20-invariant [ cols="^,^,^,^,^",options="header " , ]
we here recall briefly the definition of the @xmath5 casson invariant , denoted by @xmath45 , based on @xcite .
let @xmath24 be a closed , orientable 3-manifold with a heegaard splitting @xmath46 with handlebodies @xmath47 and a heegaard surface @xmath48 , that is , @xmath49 and @xmath50 .
then the inclusion maps @xmath51 and @xmath52 for @xmath53 induce surjections on the fundamental groups .
it then follows that @xmath54 , where @xmath55 , @xmath56 , @xmath57 and @xmath58 denote the @xmath5-character varieties for @xmath24 , @xmath59 , @xmath60 and @xmath48 respectively .
there are natural orientations on all the character varieties determined by their complex structures .
the invariant @xmath45 is ( roughly ) defined as an oriented intersection number of the subspaces of characters of irreducible representations in @xmath61 and @xmath62 , which counts only compact , zero - dimensional components of the intersection .
see @xcite and @xcite , also @xcite for detailed definition . for the 3-manifolds obtained by dehn surgeries on two - bridge knots , boden and curtis studied the @xmath5 casson invariant @xmath45 in detail in @xcite , and showed that @xmath45 can be calculated as follows ( ( * ?
* theorem 2.5 ) ) : let @xmath7 be a two - bridge knot with schubert form @xmath63 and @xmath64 the 3-manifold obtained by @xmath65-surgery on @xmath7 .
suppose that @xmath65 is not a strict boundary slope and no @xmath66-th root of unity is a root of @xmath32 , where @xmath67 if @xmath68 is odd and @xmath69 if @xmath68 is even .
then @xmath70 \dfrac{\| p / q \|_{t}}{2 } - \dfrac{\alpha - 1}{4 } & \text{if $ p$ is odd . }
\end{cases}\ ] ] here @xmath71 denotes the total culler - shalen seminorm of @xmath65 .
recall that a slope on the boundary of a knot exterior @xmath24 is called a _
boundary slope _
if there exists an essential surface @xmath48 embedded in @xmath24 with nonempty boundary representing the slope , and a boundary slope is called _ strict _ if it is the boundary slope of an essential surface that is not the fiber of any fibration over the circle . in this paper , we omit the detailed definition of the total culler - shalen norm ( see @xcite for example ) , while the calculation of the total culler - shalen seminorm of a slope for a two - bridge knot was essentially given in @xcite .
in fact , the following explicit formula is presented as ( * ? ? ?
* proposition 2.3 ) .
@xmath72 here @xmath73 denote the boundary slopes for a two - bridge knot @xmath7 . by the result given in @xcite , a boundary slope for a two - bridge knot @xmath74
is associated to a continued fraction expansion of @xmath75 .
then @xmath76 is set to be @xmath77 for the continued fraction expansion @xmath78 $ ] associated to @xmath79 .
combining these formulae , we see the following . @xmath80 in particular , we have the following when @xmath81 .
@xmath82 consequently , together with the result of ni and wu given in @xcite , a two - bridge knot has no purely cosmetic surgery pairs yielding homology 3-spheres if @xmath83 holds . on the other hand , in (
* theorem 2 ) , the following method to enumerate all the continued fractions associated to boundary slopes for a two - bridge knot was given .
the boundary slopes of a two - bridge knot with schubert form @xmath63 are associated to the continued fractions obtained by applying the following substitutions at non - adjacent positions in the _ simple continued fraction _
( i.e. , the unique one with all terms positive and the last term greater than 1 ) of @xmath84 .
the following exhibit the substitutions at position 2 .
substitution 1 : + @xmath85 \mapsto [ b_0 + 1 , ( -2 , 2)^{b_1 -1 } , -2 , b_2 + 1 , b_3 , \dots , b_n ] $ ] + substitution 2 : + @xmath86 \mapsto [ b_0 + 1 , ( -2 , 2)^{b_1 } , - b_2 - 1 , - b_3 , \dots , - b_n ] $ ] let us recall how to calculate the boundary slopes from a continued fraction . by the result given in @xcite
, a continued fraction expansion is associated to a boundary slope if it has partial quotients which are all at least two in absolute value .
we call such a continued fraction a _ boundary slope continued fraction_. given a two - bridge knot with schubert form @xmath63 , consider a boundary slope continued fraction expansion @xmath87 $ ] of @xmath88 with integer part @xmath89 and @xmath90 for @xmath91 . compare the signs of the terms @xmath92 to the pattern @xmath93 $ ] , and let @xmath94 ( resp .
@xmath95 ) be the number of terms matching ( resp .
not matching ) the pattern .
note that , among the boundary slope continued fractions , there is a unique one having all terms even ; that is associated to the longitude ( i.e. , the boundary slope of a seifert surface ) .
let @xmath96 and @xmath97 be the corresponding values for the continued fraction associated to the longitude .
then the boundary slope associated to the continued fraction is presented as @xmath98 .
in this section , we give a proof of theorem [ mainthm ] .
as explained in the previous section , to prove the theorem , it suffices to enumerate all the boundary slopes by using the substitution method , and calculate @xmath99 and @xmath100 for the obtained boundary slopes .
first we consider the case @xmath4 , that is , the case of @xmath3 .
we start with the simple continued fraction of @xmath101 , which is represented as the continued fraction @xmath102 $ ] .
we use @xmath103-tuples of the form @xmath104 with @xmath105 to show where substitutions are applied . as an example , @xmath106 means the substitution rule is applied at positions @xmath107 and @xmath103 .
then we have a boundary slope continued fraction @xmath108 $ ] which is the longitude continued fraction . hence we see that @xmath109 and @xmath110 . here recall that each term of boundary slope continued fractions must be at least two in absolute value .
hence @xmath111 does not fit in our case since the term of @xmath112 at position @xmath113 remains after substitutions .
similarly , we can eliminate the possibility of @xmath114 and @xmath115 .
we also note that no two terms of @xmath112 are adjacent in a @xmath103-tuple .
it is therefore enough to consider the following 10 cases to obtain all the boundary slope continued fractions .
. then we have @xmath108 $ ] .
hence @xmath116 , @xmath117 and @xmath118 .
. then we have @xmath120 $ ] .
hence @xmath121 , @xmath122 , @xmath123 and @xmath124 .
. then we have @xmath126 $ ] .
hence @xmath127 , @xmath128 and @xmath129 .
. then we have @xmath131 $ ] .
hence @xmath132 , @xmath133 , @xmath134 and @xmath135 .
@xmath136 .
then we have @xmath137 $ ] .
hence @xmath138 , @xmath139 , @xmath140 and @xmath141 .
. then we have @xmath143 $ ] .
hence @xmath144 , @xmath145 , @xmath146 and @xmath147 .
. then we have @xmath149 $ ] .
hence @xmath150 , @xmath151 , @xmath152 and @xmath153 .
. then we have @xmath155 $ ] .
hence @xmath156 , @xmath157 , @xmath158 and @xmath159 .
. then we have @xmath161 $ ] .
hence @xmath162 , @xmath163 , @xmath164 and @xmath165 .
. then we have @xmath167 $ ] .
hence @xmath168 , @xmath169 , @xmath170 and @xmath171 .
we therefore see that @xmath172 next we consider the general case , where @xmath173 .
we remark that the schubert form of the knot @xmath16 is described as @xmath174 .
thus its simple continued fraction is given as @xmath175 $ ] .
we in turn use @xmath13-tuples of the form @xmath176 with @xmath105 to show where substitutions are applied .
the longitude continued fraction is obtained from @xmath177 and is @xmath178 $ ] .
there is no possibility of @xmath179 , @xmath180 , @xmath181 , @xmath182 and @xmath183 since each term of boundary slope continued fractions is at least two in absolute value .
we again note that no two terms of @xmath112 are adjacent in a @xmath13-tuple .
it is therefore enough to consider the following 25 cases to obtain all the boundary slope continued fractions .
case 1 . @xmath184
. then we have @xmath185 $ ] .
hence @xmath186 , @xmath187 and @xmath118 .
. then we have @xmath189 $ ] .
hence @xmath190 , @xmath191 , @xmath192 and @xmath193 .
. then we have @xmath178 $ ] .
hence @xmath194 , @xmath195 and @xmath129 .
. then we have @xmath197 $ ] .
hence @xmath198 , @xmath199 , @xmath200 and @xmath201 .
. then we have @xmath203 $ ] .
hence @xmath204 , @xmath205 , @xmath206 and @xmath207 .
. then we have @xmath209 $ ]
. hence @xmath210 , @xmath211 , @xmath212 and @xmath213 .
. then we have @xmath215 $ ] .
hence @xmath216 , @xmath217 , @xmath218 and @xmath219 .
. then we have @xmath221 $ ] .
hence @xmath222 , @xmath223 , @xmath224 and @xmath225 .
. then we have @xmath227 $ ] .
hence @xmath228 , @xmath229 and @xmath230 .
. then we have @xmath232 $ ] .
hence @xmath233 , @xmath234 , @xmath235 and @xmath236 .
. then we have @xmath238 $ ] .
hence @xmath239 , @xmath240 , @xmath241 and @xmath242 .
. then we have @xmath244 $ ] .
hence @xmath245 , @xmath246 , @xmath247 and @xmath248 .
@xmath249 .
then we have @xmath250 $ ] .
hence @xmath251 , @xmath252 and @xmath253 .
. then we have @xmath255 $ ] .
hence @xmath256 , @xmath257 , @xmath258 and @xmath259 .
. then we have @xmath261 $ ] .
hence @xmath262 , @xmath263 , @xmath264 and @xmath265 .
. then we have @xmath267 $ ] .
hence @xmath268 , @xmath269 and @xmath270 .
. then we have @xmath272 $ ] .
hence @xmath273 , @xmath274 , @xmath275 and @xmath276 .
case 18 . @xmath277
. then we have @xmath278 $ ] .
hence @xmath279 , @xmath280 , @xmath281 and @xmath282 .
. then we have @xmath284 $ ] .
hence @xmath285 , @xmath286 , @xmath287 and @xmath288 .
@xmath289 .
then we have @xmath290 $ ] .
hence @xmath291 , @xmath292 , @xmath293 and @xmath294 .
@xmath295 .
then we have @xmath296 $ ] .
hence @xmath297 , @xmath298 , @xmath299 and @xmath300 .
@xmath301 .
then we have @xmath302 $ ] .
hence @xmath303 , @xmath304 and @xmath305 .
@xmath306 .
then we have @xmath307 $ ] .
hence @xmath308 , @xmath309 , @xmath310 and @xmath311 .
. then we have @xmath313 $ ] .
hence @xmath314 , @xmath315 , @xmath316 and @xmath317 .
. then we have @xmath319 $ ] .
hence @xmath320 , @xmath321 , @xmath322 and @xmath323 . since we are assuming @xmath324 , @xmath325
in this section , we justify remark [ rmk2 ] in section 1 as follows .
let @xmath16 be a two - bridge knot with conway form @xmath326 $ ] for @xmath2 .
then @xmath327 and @xmath328 hold . here
@xmath329 denotes the alexander polynomial of @xmath16 normalized to be symmetric and to satisfy @xmath330 .
let @xmath16 be a two - bridge knot @xmath7 with conway form @xmath1 $ ] for @xmath2 .
then @xmath16 is a slice knot , originally observed by casson and gordon , and see ( * ? ? ?
* lemma 8.2 ) for a proof . on the other hand ,
the invariant @xmath20 must vanish for slice knots as shown in ( * ? ? ?
* corollary 1.3 ) .
thus we have @xmath328 for our knot @xmath16 with @xmath2 .
now let us calculate the alexander polynomial for @xmath16 .
this is just a straightforward calculation , but we include it for readers convenience
. in general , a two - bridge knot with conway form @xmath331 $ ] is depicted as in figure [ fig1 ] .
note that such a knot is of genus three , and any two - bridge knot of genus three has such a conway form . in the figure , @xmath332 to @xmath48 denote the numbers of horizontal full - twists with signs of the twists .
( 10,2 ) ( 0,0 ) to @xmath48 denote the numbers of full - twists.,title="fig : " ] ( .8,.94)@xmath332 ( 3.9,.94)@xmath333 ( 6.95,.94)@xmath334 ( 2.3,.43)@xmath335 ( 5.4,.43)@xmath336 ( 8.45,.43)@xmath48 such a seifert surface of genus three can be deformed into the one as shown in figure [ fig2 ] . to calculate the seifert matrix , we set a basis @xmath337 of the first homology group of the surface , as illustrated in figure 2 .
( 10,3.3 ) ( 0,0 ) ( 1.15,3)@xmath338 ( 2.7,3)@xmath339 ( 4.15,3)@xmath340 ( 5.6,3)@xmath341 ( 7.1,3)@xmath342 ( 8.6,3)@xmath343 ( 1.2,.33)@xmath332 ( 2.68,.33)@xmath335 ( 4.18,.33)@xmath333 ( 5.64,.33)@xmath336 ( 7.13,.33)@xmath334 ( 8.63,.33)@xmath48 then we have the seifert matrix as follows . @xmath344 then @xmath345 is obtained as @xmath346 we then have the following polynomial of degree 6 ; @xmath347 now we consider the conway form @xmath1 $ ] , that is , @xmath348 this implies that @xmath349 .
after normalization , we have the following .
@xmath350 it follows that ; @xmath351 @xmath352 @xmath353
the authors would like to thank hitoshi murakami for letting them know the relationship of the casson invariant and the conway polynomial . during our study ,
computer - experiments were quite useful .
the experiments were performed by using the dunfield s program @xcite , which implements the hatcher - thurston s algorithm to enumerate boundary slopes for two - bridge knots .
s. a. bleiler , c. d. hodgson and j. r. weeks , cosmetic surgery on knots , in _ proceedings of the kirbyfest ( berkeley , ca , 1998 ) _ , 2334 ( electronic ) , geom .
monogr . , 2 , geom .
coventry . c. l. curtis , erratum
to : `` an intersection theory count of the @xmath355-representations of the fundamental group of a 3-manifold '' [ topology * 40 * ( 2001 ) , no .
4 , 773787 ] , topology * 42 * ( 2003 ) , no . 4 , 929 | we consider the cosmetic surgery problem for two - bridge knots in the 3-sphere .
it is seen that all the two - bridge knots at most 9 crossings other than @xmath0 $ ] admits no purely cosmetic surgery pairs .
then we show that any two - bridge knot of the conway form @xmath1 $ ] with @xmath2 admits no cosmetic surgery pairs yielding homology 3-spheres , where @xmath3 appears for @xmath4 .
our advantage to prove this is using the @xmath5 casson invariant . |
* experimental details * the experimental procedure has been described in previous publications @xcite .
a degenerate fermi gas of @xmath0li atoms was first prepared in an optical trap , using laser cooling and sympathetic cooling with @xmath132na atoms .
a variable spin mixture of the two lowest hyperfine states @xmath45 and @xmath46 ( corresponding to the @xmath133 and @xmath134 states at low magnetic field ) was created at a magnetic field @xmath135 g. the final evaporative cooling by lowering the trap depth and all measurements were performed at @xmath136 g. the temperature of the cloud was controlled by the lowest value of the trap depth in the evaporative cooling process .
the axial trap frequency was @xmath137 hz .
the two transverse trap frequencies are equal within less than @xmath138 .
the optical signal in the phase - contrast imaging is proportional to the net phase shift of the imaging beam passing through a fermi mixture , i.e. @xmath139 , where @xmath140 is the probe frequency of the imaging beam , and @xmath141 and @xmath142 are the resonance frequencies of the optical transition for the states @xmath45 and @xmath46 , respectively . when the probe beam is tuned to the middle of the two transitions , i.e. @xmath143 , the optical signal reflects the density difference @xmath144 with @xmath145 . in our experiment ,
two phase - contrast images of the same sample were taken consecutively with different probe frequencies , @xmath146 and @xmath147 .
the two images record the density difference @xmath148 and the weighted density difference @xmath149 .
@xmath146 was determined by zeroing the optical signal with an equal mixture and @xmath150 was determined by the signal ratio between the first and the second image for a highly imbalanced fermi mixture with @xmath151 ( an almost fully polarized gas ) .
finally , we obtained @xmath152 and @xmath153 .
the difference between @xmath146 and @xmath147 was chosen to lie between 8 and 13 mhz .
the time interval between the two images was @xmath154s , and the pulse duration of each probe beam was @xmath155s . because the probe beam was off - resonant
, no heating effect of the first pulse was observed in the second image .
low - noise profiles were obtained by averaging the column density distribution of phase - contrast images along elliptical equipotential lines ( @xmath156 ) . for the measurement of the critical spin polarization ,
the averaging region was restricted to @xmath157 m in order to preserve the sharp features at the phase boundary .
the diffraction limit for our imaging system was about @xmath158 m .
for the determination of local quantities in the profiles , we averaged over @xmath159 m around a given position . for temperature determination , the averaging region was restricted to an axial sector of @xmath160 to avoid corrections due to radial anharmonicities ( see appendix ) .
the relative temperature @xmath161 is determined as @xmath162 , where @xmath163 is the fugacity obtained from the fit ( @xmath164 is the poly - logarithmic function of order @xmath165 ) . in our previous work
@xcite , temperatures have been determined by fitting the spatial wings of the majority component after expansion .
however , we found that one can neglect collisions with the minority atoms in the core only for large population imbalances . in a simplified picture , one can regard collisions with the inner core as collisions with a moving wall , which moves outward radially and inward axially ( due to the magnetic trapping potential ) .
this results in different average kinetic energies ( transversely and axially ) of the free majority atoms in the outer region .
figure [ f : expansion ] shows the density distribution of the majority and minority components after expansion .
although the temperature has been overestimated by only @xmath167 for typical experimental conditions ( @xmath168 ) in refs
@xcite , we do not regard this technique as well - calibrated absolute thermometry .
one other concept for thermometry determines temperature as the derivative of entropy with energy .
so far , this concept could be implemented only for balanced fermion mixtures with certain approximations , and due to the need of determining a derivative , could only be used to obtain temperatures averaged over a certain range @xcite . and ( d , e ) transverse @xmath169 directions after expansion .
( c ) the azimuthally averaged column density profiles of the majority ( red ) and the minority ( blue ) cloud are obtained from ( a ) and ( b ) , respectively .
the excess majority atoms in the outer region interact with the core during expansion .
the contour lines of the outer part of the majority cloud ( color inset ) are not elliptical and have the shape of a horse - track .
this shows that the minority cloud pushes the outer majority atoms in the transverse direction , which is also indicated by the hump of the majority profile at the edge of the minority cloud .
the population imbalance was @xmath170 .
[ f : expansion],width=307 ] for the determination of temperatures from the spatial in situ profiles it was necessary to address the anharmonicity of the trapping potential .
our trap is generated by a weakly focused ( beam waist @xmath171 m ) infrared gaussian laser beam ( wavelength 1064 nm ) near the saddle point of a magnetic potential .
the total trapping potential is given as @xmath172 where @xmath173 .
note that gravity has been compensated by a magnetic field gradient .
the axial confinement comes mainly from the magnetic potential with oscillation frequency of @xmath137 hz .
the transverse magnetic potential is anti - trapping and limits the trap depth as @xmath174 where @xmath48 is the transverse oscillation frequency in the central harmonic region .
when the trap depth is comparable to the fermi energy of a sample , the transverse anharmonicity will affect the shape of the cloud .
although in our experiments , the inner core and the outer cloud had the same aspect ratio as the trapping potential , indicating the absence of anharmonic effects , anharmonicities were not negligible in the spatial wings used to determine the temperature . of the averaging sector .
for a large angle , the large-@xmath175 region is included in the averaged profile , resulting in a broadening of the spatial wings and consequently higher value of the fitted temperature .
the red line shows the results of a simulation using the same parameters as the experiment ( @xmath176 , @xmath177k and the trap depth @xmath178k).[f : anharmonicity],width=230 ] this issue was addressed by adjusting the angular averaging region ( fig .
[ f : anharmonicity ] ) .
since the trapping potential is only anharmonic for large @xmath179 , we could reduce the effect by decreasing the angle of the averaging sector around the @xmath180 direction . both the experimental data and
an exact simulation for an ideal fermi gas show that the fitted temperature remains almost constant up to a certain angle and then increases when the averaging sector includes more of the transversely outer region . in our temperature determination
, we chose the averaging sector to be @xmath160 which was large enough to create low - noise profiles , but kept the effect of the anharmonicities to below @xmath181 .
the 1d fit to angularly averaged profiles was computationally more efficient than a 2d fit to a selected region of the image . in a 2d fit
, one could also include anharmonic terms in the fitting function . when the two spin components have a chemical potential difference @xmath182 , the bcs - type superfluid has two branches of quasiparticles with excitation energies @xmath183 where @xmath184 . at finite temperature ,
the superfluid is polarized due to the large thermal population of the lower branch compared to the upper branch .
an interesting situation arises when @xmath185 becomes larger than @xmath105 , i.e. the lower branch has negative energy quasiparticles , implying that even at zero temperature the superfluid state would have a finite polarization .
our experiments show @xmath186 at very low temperature , suggesting that a polarized superfluid state exists only at finite temperature .
the breached - pair state with @xmath187 at zero temperature has been predicted in a stronger coupling region ( on the bec side of the feshbach resonance ) . since @xmath105 gradually decreases with higher temperature , it might be possible to have @xmath187 at finite temperature , at least in the weakly - interacting bcs limit where @xmath105 smoothly approaches zero at a second order phase transition point .
one interesting problem is identifying this gapless region of @xmath188 in the phase diagram for various coupling regimes . | the pairing of fermions is at the heart of superconductivity and superfluidity .
the recent experimental realization of strongly interacting atomic fermi gases has opened a new , controllable way to study novel forms of pairing and superfluidity .
a major controversial issue has been the stability of superfluidity against an imbalance between the two spin components when the fermions interact resonantly .
here we present the phase diagram of a spin - polarized fermi gas of @xmath0li atoms at unitarity , mapping out the superfluid phase versus temperature and density imbalance . using tomographic techniques ,
we reveal spatial discontinuities in the spin polarization , the signature of a first - order superfluid - to - normal phase transition , which disappears at a tricritical point where the nature of the phase transition changes from first - order to second - order . at zero temperature
, there is a quantum phase transition from a fully - paired superfluid to a partially - polarized normal gas .
these observations and the implementation of an in situ ideal gas thermometer provide quantitative tests of theoretical calculations on the stability of resonant superfluidity .
superfluidity and superconductivity of fermions are based on the formation of fermion pairs .
the stability of these pairs determines the robustness of the superfluid state , and the quest for superconductors with high critical temperature is a search for systems with strong pairing mechanisms .
ultracold atomic fermi gases present a highly controllable model system for studying strongly interacting fermions @xcite .
tunable interactions utilizing feshbach collisional resonances and control of population or mass imbalance among the spin components provide unique opportunities to investigate the stability of pairing @xcite , and possibly to search for new exotic forms of superfluidity @xcite .
the case of unitarity , when the two spin components resonantly interact and the behavior of the system becomes independent of the nature of the interactions , has become a benchmark for experimental and theoretical studies over the last few years .
however , there is an ongoing debate about the stability of resonant superfluidity , reflected in major discrepancies in predicted transition temperatures for the balanced spin mixture @xcite , and an even more dramatic discrepancy for the critical imbalance of the two spin components , called the chandrasekhar - clogston ( cc ) limit of superfluidity @xcite .
recent quantum monte - carlo ( qmc ) calculations predicted that superfluidity would be quenched by a density imbalance around 40% @xcite , whereas other studies predicted a critical imbalance above 90% @xcite .
our earlier work @xcite suggested the lower limit but other experiments @xcite were interpreted to be consistent with the absence of the cc limit .
this huge discrepancy reveals that even qualitative aspects , such as the role of interactions in the normal phase , are still controversial .
the lack of reliable thermometry for strongly interacting systems limits the full interpretations of experimental results . here
we resolve this long standing debate by presenting the phase diagram of a spin - polarized fermi gas at unitarity .
we observe that the normal - to - superfluid phase transition changes its nature . at low temperature , the phase transition occurs with a jump in the spin polarization as the imbalance increases , which we interpret as a first - order phase transition . the local spin polarization or local density imbalance is defined as @xmath1 , where @xmath2 and @xmath3 refer to the two spin components with densities @xmath4 . at high temperature ,
the phase transition is smooth and therefore of second - order .
the two regimes are connected by a tricritical point @xcite for which we estimate the position as @xmath5 , where @xmath6 is the fermi energy of the majority component of density @xmath7 ( @xmath8 is the boltzmann constant , @xmath9 is the planck constant divided by @xmath10 and @xmath11 is the atomic mass of @xmath0li ) .
our low - temperature results confirm a zero - temperature quantum phase transition at a critical polarization @xmath12 .
this work required the introduction of several novel techniques .
a tomographic reconstruction of local fermi temperatures and spin polarization allowed us to obtain the phase diagram for the homogeneous system , no longer affected by the inhomogeneous density of the trapped samples . furthermore
, absolute temperatures were obtained using in situ thermometry applied to the non - interacting fully - polarized fermi gas in the outer part of the trapped samples , an ideal thermometer with exactly - known thermal properties .
in contrast to previous works @xcite , this is a direct measurement without any approximations . and
@xmath3 ) fermi mixture is confined in an external potential @xmath13 with the chemical potentials of each spin components @xmath14 ( @xmath15 is the shift for the spin @xmath3 component due to interactions ) .
( b ) density distributions of the majority component @xmath16 ( red ) and the minority component @xmath17 ( blue ) . ( c )
spin polarization @xmath18 .
at zero temperature , the trapped fermi mixture has radially a three - layer structure : ( i ) the core region ( @xmath19 ) of a fully paired superfluid with equal densities of the two components , ( ii ) the intermediate region ( @xmath20 ) of a partially polarized normal gas , and ( iii ) the outer region ( @xmath21 ) of a fully polarized normal gas .
the critical polarization @xmath22 ( @xmath23 ) is defined as the minimum ( maximum ) spin polarization of the normal ( superfluid ) region .
the non - interacting case is shown in insets .
a.u .
, arbitrary units.[f : model],width=240 ] * spatial structure of a trapped fermi mixture * our experiments are carried out in a trapping potential @xmath24 .
the local chemical potential of each spin component is given as @xmath25 , where @xmath26 are the global chemical potentials .
when @xmath27 due to imbalanced populations , the chemical potential ratio @xmath28 spatially varies over the trapped sample and thus , under the local density approximation ( lda ) the trapped inhomogeneous sample is represented by a line in phase diagrams of the homogeneous system .
figure [ f : model ] illustrates the spatial structure of a strongly interacting fermi mixture in a harmonic trap . in the inner region , where @xmath29 is closer to unity , a superfluid with zero ( small ) spin polarization will form at zero ( low ) temperatures , having a sharp phase boundary against the partially - polarized normal gas in the outer region .
the spin polarization shows a discontinuity at the phase boundary @xmath30 , a signature of the phase separation of a superfluid and a normal gas @xcite .
the critical polarization @xmath31 ( @xmath32 ) represents the minimum ( maximum ) spin polarization for a stable normal ( superfluid ) gas . at higher temperature
, the discontinuity in the density imbalance disappears .
the main result of this paper is the observation and quantitative analysis of such density profiles .
since we have no experimental evidence , we are not discussing the exotic partially - polarized phases @xcite which could exist only in the transition layer between the superfluid core and the normal outer region . and ( b ) the weighted density difference @xmath33 , respectively .
the phase - contrast images show the 2d distribution of the column density difference , @xmath34 where the integral describes the line - of - sight integration .
the field of view for each image is @xmath35m@xmath36 m .
( c ) the distributions of the column density difference @xmath37 ( black ) and @xmath38 ( red ) along the central line ( the dashed lines in ( a ) and ( b ) ) .
the profiles of the integrated linear density difference , ( d ) @xmath39 and ( e ) @xmath40 show the identical flattop feature except scaling .
the aspect ratio of the trapping potential was @xmath41 , the majority atom number was @xmath42 , the population imbalance was @xmath43 , and the relative temperature was @xmath44 ( see text for definitions).[f : doublepc],width=268 ] we prepared a variable spin mixture of the two lowest hyperfine states of @xmath0li atoms , labeled as @xmath45 and @xmath46 , at a magnetic field of 833 g. a broad feshbach resonance is located at 834 g and the interactions between the two spin states are resonantly enhanced .
our sample was confined in a 3d harmonic trap with cylindrical symmetry .
the in situ density distributions of the majority ( spin @xmath2 ) and minority ( spin @xmath3 ) components were determined using a phase - contrast imaging technique . since the trapped sample was observed to have an elliptical shell structure of the same aspect ratio @xmath47 as the trapping potential over our entire temperature range , where @xmath48 ( @xmath49 ) is the oscillation frequency in the transverse ( axial ) direction ( fig .
[ f : doublepc ] ) , we obtained the low - noise profiles @xmath50 by averaging the column density distribution along the equipotenital line ( defined as @xmath51 for a given radial position @xmath52 ) .
the region for averaging was restricted depending on the type of analysis .
the density profiles @xmath53 were determined from the 3d reconstruction using the inverse abel transformation of the column densities @xmath50 @xcite .
deviations from the trap aspect ratio were only found for the outer thermal wings and will be discussed below . more detailed description of the experimental procedure and the image processing is provided in appendix .
most of our measurements were performed at a total population imbalance @xmath54 , where @xmath55 refers to the total atom numbers in the sample , @xmath56 and @xmath57 of the spin @xmath2 and @xmath3 components , respectively .
figure [ f : profiles ] displays the radial profiles of the densities @xmath16 and @xmath17 of each component and the corresponding spin polarization @xmath58 for various temperatures .
the discontinuity in the spin polarization , clearly shown at very low temperature , demonstrates the phase separation of the inner superfluid of low polarization and the outer normal gas of high polarization . at low temperature
the core radius @xmath59 is determined as the kink ( and/or peak ) position in the column density difference profile . at high temperature
( but still in the superfluid regime ) , the discontinuity in @xmath58 disappears . at our lowest temperature ,
the radii of the minority cloud and the core region were measured as @xmath60 and @xmath61 ( at @xmath43 ) , respectively , and agree with recent theoretical calculations @xcite within the experimental uncertainties due to the determination of @xmath62 . here , @xmath63 is the radius of the majority cloud , measured as @xmath64 ( @xmath65 is the thomas - fermi ( tf ) radius of a non - interacting fermi gas with the same atom number of @xmath56 ) , reflecting the attraction between the two components .
( see text for definition ) was determined from the outer region ( @xmath66 ) of the averaged column density difference profile ( black line ) fitted to a finite temperature fermi - dirac distribution ( red line ) .
the radius of the minority cloud @xmath67 was determined from a fit of the wing profile of the minority component ( black dashed line ) to a zero temperature tf distribution ( red dashed line ) .
( a ) @xmath68 and @xmath43 , and ( b ) @xmath69 and @xmath70.[f : insitut],width=192 ] * thermometry of a strongly interacting fermi gas * in most cold atom experiments , temperature is determined from the density distribution after ballistic expansion , reflecting the momentum distribution of thermal atoms in the cloud . in the case of a population - imbalanced fermi mixture ,
the outer part of the majority component , having no spatial overlap with the minority component , is a non - interacting fermi gas in thermal equilibrium with the inner part and fulfills the definition of an ideal thermometer , namely a substance with exactly - understood properties in contact with the sample to be characterized .
this concept was first introduced in ref .
@xcite , where temperature was determined from the majority wing profile after expansion .
however , we found that during expansion some of the outer majority atoms experience collisions with the minority atoms in the inner region , causing a modification of their kinetic energy ( see appendix ) .
we avoided this problem by analyzing the in situ profiles .
the outer part of the averaged column density difference profile ( @xmath66 ) was fit to a finite temperature fermi - dirac distribution in a harmonic trap ( fig [ f : insitut ] ) and the relative temperature @xmath71 was determined , where @xmath72 is the fermi energy of the non - interacting fermi gas which has the same density distribution in the outer region as the majority cloud ( @xmath73 is the central density of the non - interacting fermi gas at zero temperature ) .
we verified that anharmonicity of the trapping potential does not affect the fitted temperature ( see appendix ) . *
phase diagram for a homogeneous system * the critical lines of the phase diagram of a homogeneous spin - polarized fermi gas were obtained by determining the local temperature and spin polarization at the phase boundary .
the local relative temperature @xmath74 was derived from the local density @xmath75 as @xmath76 .
since we observe no jump in the majority density within our resolution , @xmath77 is well - defined at the phase boundary .
the critical spin polarizations @xmath22 and @xmath23 were measured as @xmath78 and @xmath79 was more robust than a fitting procedure , but precludes that @xmath23 will be equal to @xmath22 at high temperature .
therefore , @xmath23 should be regarded as a lower bound for the polarization of the superfluid at the phase boundary ] .
the discontinuity in the spin polarization profile implies that there is a thermodynamically unstable window , @xmath80 , leading to a first - order superfluid - to - normal phase transition . as the temperature increases
, the unstable region reduces with @xmath22 decreasing and @xmath23 increasing . for high temperature when the bimodal feature in the spin polarization profile disappears , we recorded the condensate fraction as an indicator of superfluidity , using the rapid field - ramp technique @xcite .
as the temperature decreases , the condensate fraction gradually increases with a finite central polarization @xcite .
such a smooth variation of the density profile and condensate fraction across the phase transition are characteristic of a second - order phase transition .
the phase diagram is characterized by the three distinguished points : the critical temperature @xmath81 for a balanced mixture , the critical spin polarization @xmath82 of a normal gas at zero temperature , and the tricritical point @xmath83 where the nature of the phase transition changes from second - order to first - order . due to the lack of a predicted functional form for the phase transition line in the @xmath84-@xmath85 plane
, we apply a linear fit to the measured critical points , suggesting @xmath86 , @xmath87 and @xmath88 .
the value for @xmath82 agrees well with the prediction of qmc calculation of 0.39 @xcite .
the extrapolation of the phase diagram to @xmath89 is only tentative , since the in situ thermometry could not be applied to small population imbalances due to the narrowness of the non - interacting outer region .
it is possible that the first- and second - order transition lines meet at the tricritical point with different slopes , similar to the case of liquid @xmath90he-@xmath91he mixtures @xcite .
-@xmath85 phase diagram for a homogeneous spin - polarized fermi gas with resonant interactions .
the critical polarizations @xmath22 ( black solid circles and square ) and @xmath23 ( gray solid circles ) are displayed along the local @xmath92 at the phase boundary .
the yellow area ( @xmath93 ) represents a thermodynamically unstable region , leading to the phase separation . above the tricritical point ,
the phase transition in the center of the cloud was observed by the onset of pair condensation . for this , a cloud was evaporatively cooled , until it crossed the phase transition on a trajectory almost perpendicular to the phase transition line ( see appendix ) .
the critical spin polarization and temperature were obtained by interpolating between points without and with small condensates ( black solid square ) .
the linear fit to the @xmath22 s is shown as a guide to the eye for the normal - to - superfluid phase transition line .
each data point consists of five independent measurements and error bars indicate standard deviation .
the blue open symbols show theoretical predictions for the critical temperature of a homogeneous equal mixture ( @xmath94 : bulgac _ et al . _
@xcite , @xmath95 : burovski _ et al . _
@xcite , @xmath96 : haussmann _ et al . _
@xcite ) and the critical polarization at zero temperature ( @xmath97 : lobo _ et al . _
@xcite ) .
the blue solid square is the measured critical temperature from luo _
et al . _
@xcite , multiplied by @xmath98 with @xmath99 @xcite to obtain local @xmath100 at the center .
finite temperature correction may increase the effective value of @xmath101.[f : phasediagram],width=249 ] the zero - temperature phase diagram and in particular the value for the cc limit are significantly affected by strong interactions in the normal phase .
the cc limit reflects the energetic competition between a superfluid state and a partially - polarized normal state . when the chemical potential difference @xmath102 is larger than a critical difference @xmath103 , the normal state is energetically favorable and the superfluid state breaks down .
in bcs theory , valid for weak interactions , the critical difference is @xmath104 @xcite ( @xmath105 is the pairing gap ) . at unitarity ,
qmc studies predict @xmath106 @xcite with the assumption of a non - interacting normal gas . here , @xmath107 .
the condition @xmath108 implies that @xmath109 in a non - interacting normal gas , and consequently @xmath110 , i.e. the absence of a partially - polarized normal phase .
mean - field approaches @xcite , which can not treat the interactions in the normal phase in an accurate way , also predict a high critical imbalance @xmath111 .
strong interactions between the atoms in the normal phase , however , have been observed through the compressed shape of the minority cloud @xcite and the shift in the rf excitation spectrum @xcite . the data in figure [ f : phasediagram ] clearly establish a zero - temperature cc limit for @xmath82 in the range of @xmath112 to @xmath113 .
the density profiles at our lowest temperature provide quantitative information on the zero - temperature thermodynamics @xcite . at zero temperature ,
the global chemical potential of a fully - paired superfluid in the core is given as @xmath114 where @xmath115 is the local fermi energy and @xmath116 is the majority ( or minority ) density at the center , whereas @xmath117 and @xmath118 . from the thermodynamic equilibrium condition @xmath119 , we obtain the chemical potential ratio as @xmath120 in our coldest sample ( @xmath121 ) , the normalized central density and the radii for the phase boundary and the minority cloud were measured to be @xmath122 , and @xmath123 , respectively , yielding @xmath124 and @xmath125 with @xmath99 @xcite .
furthermore , the critical difference is given as @xmath126 .
since theory clearly predicts @xmath127 @xcite , we have @xmath128 .
if @xmath129 were larger than @xmath105 , polarized quasi - particles would have negative energies and form already at zero temperature .
therefore , up to our observed value of @xmath129 , the fully - paired superfluid state is stable , and a polarized superfluid exists only at finite temperature .
the interface between two immiscible fluids involves a surface energy , leading to at least a small violation of the lda .
however , the observed sharp interface along the an equipotential line and the flattop structure of the linear density difference profiles ( fig .
[ f : doublepc]d and e ) imply that corrections to the lda are smaller than the resolution of our experiment .
these observations are inconsistent with the interpretations given for the experimental results reported in ref .
@xcite , where it has been shown that highly - elongated small samples are deformed by surface tension @xcite .
the scaling of those surface effects to our parameters predicted a deviation of the aspect ratio of the superfluid core of @xmath130 from the trap aspect ratio @xcite , whereas we observe this deviation to be smaller than 2% .
note that surface tension would add energy in the phase - separated superfluid regime and would shift the cc limit to smaller values .
ref .
@xcite concluded that the cc limit should be @xmath131 which is ruled out by our observations .
we are not aware of any suggested effect which can reconcile the data of ref .
@xcite with our phase diagram for a resonant superfluid . to indentify this finite size effect and to fully understand the nature of the normal state @xcite are still open questions for imbalanced fermi gases . * conclusions * we have established the phase diagram of a homogeneous spin - polarized fermi gas with resonant interactions in the @xmath84-@xmath85 plane .
this includes the identification of a tricritical point where the critical lines for first - order and second - order phase transitions meet , and the final confirmation of a zero - temperature quantum phase transition , the cc limit of superfluidity , for a gas at unitarity .
so far , predicted exotic superfluid states such as the breached - pair state in a stronger coupling regime ( bec side " ) @xcite and the fflo state in a weaker coupling regime ( bcs side " ) @xcite have not been observed , but the novel methods of tomography and thermometry will be important tools in the search for those states .
we thank m. w. zwierlein and a. keshet for a critical reading of the manuscript .
this work was supported by nsf and onr . |
boosted by the discovery of superconductivity in cecu@xmath0si@xmath0 over twenty years ago,@xcite the relationship between superconductivity and magnetism has been extensively investigated in various d and f heavy fermion ( hf ) compounds .
a consensus has developed that hf superconductivity is mediated by spin fluctuations,@xcite mainly because superconductivity was found close to a magnetic instability at @xmath5 , sometimes described as a quantum - critical point ( qcp ) , often attained by applying pressure .
a recent development is that an essentially gapless superconducting ( sc ) state has been identified by nmr / nqr measurements in the region where the sc state coexists with antiferromagnetism,@xcite consistent with a theoretical prediction.@xcite in this paper we further explore the possibility that at a pressure @xmath6gpa , a second qcp , associated with f - electron occupation number , has a major role to play in the superconductivity of cecu@xmath0si@xmath0 and related compounds.@xcite cecu@xmath0si@xmath0 has a superconducting ground state at ambient pressure with a critical temperature @xmath7 , around 0.7 k. it is firmly believed that the compound is close to an antiferromagnetic qcp at slight negative pressure , accessible for example by partial substitution of si with ge.@xcite when pressure is applied , @xmath7 initially remains close to its ambient pressure value , followed by a sudden increase to around @xmath8k at about @xmath9gpa .
further increase in pressure results in a slower suppression of @xmath7 to zero .
this non - monotonous behavior of @xmath10 was first explored by resistivity in the quasi - hydrostatic conditions of bridgman anvil cell.@xcite subsequent investigations by susceptibility@xcite and resistivity @xcite were carried out in various pressure media , and showed considerable variation in @xmath7 between samples , especially at high pressure . with increasing pressure cecu@xmath0si@xmath0 passes from a nearly trivalent @xmath11 behavior , with kondo coupling between conduction and f - electrons , to behavior at very high pressure characteristic of intermediate valence ( iv ) systems , whose valence fluctuates between the @xmath12 and @xmath13 $ ] electronic configurations . as a result ,
deep in this iv regime , the resistivity , for instance , resembles that of lacu@xmath0si@xmath0 , which lacks @xmath14 electrons
. similar @xmath10 dependence to that found in cecu@xmath0si@xmath0 is seen in the isoelectronic sister compound , cecu@xmath0ge@xmath0 , offset by about 10@xmath15gpa due to the larger atomic volume of ge.@xcite apart from this shift of the pressure scale , no qualitative difference occurs between the two compounds beyond unavoidable sample dependence . from a more theoretical point of view , there exist at least three reasons to believe that critical valence fluctuations are at the origin of the pressure - induced peak of the sc transition temperature @xmath7 .
first , the @xmath16 coefficient of the @xmath17 resistivity law decreases drastically by about two orders of magnitude around the pressure corresponding to the @xmath7 peak.@xcite since @xmath16 scales as @xmath18 in the so - called kondo regime , this implies that the effective mass @xmath19 of the quasiparticles also decreases sharply there .
this fall of @xmath19 is possible only if there is a sharp change of ce valence , deviating from ce@xmath20 , since the following approximate formula for the renormalization factor @xmath21 holds in the strongly correlated limit:@xcite @xmath22 where @xmath23 is the f - electron number per ce ion .
second , the so - called kadowaki - woods ( kw ) ratio,@xcite @xmath24 , where @xmath25 is the sommerfeld coefficient of the electronic specific heat , crosses over quickly from that of a strongly correlated class to a weakly correlated one.@xcite the inverse of the sommerfeld coefficient , @xmath26 , can be identified with the kondo temperature @xmath27 , which is experimentally accessible by resistivity measurements .
this indicates that the mass enhancement due to the dynamical electron correlation is quickly lost at around @xmath28 , in agreement with the previous point .
the phenomenon can be understood if we note the fact that @xmath25 consists essentially of two terms : @xmath29 where @xmath30 is due to the so - called band effect and @xmath31 is due to the many - body correlation effect , with @xmath32 being the self - energy of the correlated electrons .
@xmath33 and @xmath16 are related to each other through the kramers - krnig relation , leading to the large value of the kw ratio,@xcite and when @xmath34 , this is indeed seen . on the other hand ,
if @xmath35 , the ratio @xmath24 should be reduced from the kw value considerably because the effect of @xmath30 can not be neglected in its denominator .
third , there is a sharp peak in the residual resistivity @xmath36 at around @xmath37,@xcite which can be understood as a many - body effect enhancing the impurity potential ( in fact we define the pressure @xmath4 experimentally by the maximum of @xmath38 ) .
in the forward scattering limit , this enhancement is proportional to the valence susceptibility @xmath39 , where @xmath40 is the atomic f - level of the ce ion , and @xmath41 is the chemical potential.@xcite physically speaking , local valence change coupled to the impurity or disorder gives rise to a change of valence in a wide region around the impurity which then scatters the quasiparticles quite strongly , leading to the increase of @xmath36 .
the enhancement of @xmath36 can be thus directly related to the degree of sharpness of the valence change , because the variation of the atomic level @xmath40 is considered to be a smooth function of the pressure .
these circumstantial clues to the importance of critical valence fluctuations have been backed up by a microscopic calculation of @xmath7 for @xmath42-wave pairing as a function of @xmath43.@xcite this showed that sudden valence change occurs if a moderately sized coulomb repulsion @xmath44 is taken into account between the conduction c- , and localized f - electrons , with the peak structure of @xmath7 being qualitatively reproduced . [ cols="<,<,^,^",options="header " , ] table [ tab : properties ] summarizes the current experimental evidence of anomalies seen in cecu@xmath0(ge / si)@xmath0 around @xmath45 .
part * ( i ) * of table [ tab : properties ] refers to direct evidence for a valence transition of the ce ion : cell volume@xcite and @xmath46 x - ray absorption@xcite measurements show discontinuities as a function of pressure .
the drastic decrease of the @xmath16 coefficient of the @xmath47 resistivity law , along with the @xmath16 vs @xmath48 scaling relation , indicate that the system is leaving the strongly correlated regime characterized by a f - occupation number close to unity .
( @xmath48 is defined in fig .
[ fig : tmax ] and assumed to be proportional to @xmath27 . )
part * ( ii ) * refers to anomalies observed close to the maximum of @xmath7 predicted by critical valence fluctuation theory.@xcite these are the maximum of @xmath7 itself and the enhanced residual resistivity , @xmath38.@xcite part * ( iii ) * refers to properties following from the extended treatment of the critical valence fluctuations found in section [ sec : theory ] of this paper .
this includes the linear resistivity,@xcite and the maximum in @xmath25 , both found around @xmath4 . in part * ( iv ) * are listed the remaining features that are observed in cecu@xmath0si@xmath0 and cecu@xmath0ge@xmath0 around the maximum in @xmath7 but which are so far not fully explained .
for example the merging of @xmath48 and @xmath49 , where the latter ( also defined in fig .
[ fig : tmax ] ) is believed to reflect the effect of the excited crystalline electric field ( cef ) split f - levels .
many of the anomalies noted in table [ tab : properties ] have also been observed to coincide with the maximum of @xmath7 in other hf superconductors , from cepd@xmath0si@xmath0@xcite to cecu@xmath50au@xcite , the latter showing traces of superconductivity under pressure .
previous work on cecu@xmath0si@xmath0 has shown a lot of variation in low - temperature behavior between different samples .
the fact that not all reports have shown every anomaly is not entirely surprising , since large variations in the electronic properties of cecu@xmath0si@xmath0 are well known to result from extremely small differences in composition.@xcite the extension of these variations with pressure has not been systematically explored , but almost all samples so far studied have shown an enhancement of @xmath7 , along with effects such as the enhancement of the residual resistivity , to be discussed below .
this variability under pressure may be due to the samples themselves , or to pressure inhomogeneities caused by non - hydrostatic pressure media .
we were therefore motivated to use solid helium as a pressure medium , due to its near - ideal hydrostaticity at low temperature .
by simultaneously probing resistivity and specific heat in the same sample , we were able to explore both percolative transport and bulk evidence for superconductivity .
high pressure was induced using diamond anvils with a 1.5@xmath15 mm culet.@xcite a stainless steel gasket was specially prepared to absorb the large volume decrease of the helium pressure medium from ambient pressure , and to avoid severing the measurement wires .
these were insulated from the gasket using a mixture of al@xmath0o@xmath51 powder and epoxy resin .
the pressure was measured to within 0.02@xmath15gpa at various temperatures down to 4.2@xmath15k using the ruby fluorescence scale .
the cecu@xmath0si@xmath0 sample was prepared by reaction of its constituent elements with a slight excess of cu , with a nominal initial composition cecu@xmath52si@xmath0 . the product
was then melted in an induction furnace and slowly allowed to crystallize under 50 bars ar in a bazro@xmath51 crucible ( see ref . for more details ) .
the small monocrystal used in this work was cut and polished to @xmath53m@xmath54 , and six 5@xmath55 m @xmath56 wires ( four gold and two au + 0.07% at .
fe ) were spot welded to the sample .
the @xmath57-axis of the tetragonal structure was parallel to its smallest dimension .
the magnetic field , when applied , was parallel to the @xmath57-axis .
the six wires spot - welded to the sample allowed multiple redundant measurements to be performed . this improved reliability and enabled us to verify the calorimetry measurements using several different configurations .
the sample resistance could be measured by a four - point method ; a knowledge of the sample dimensions then enabled the absolute resistivity to be determined to within 10% .
the two thermocouple junctions were formed from an au / fe pair at either end of the sample .
an alternating resistive heating current was passed through one ( to avoid passing the current through the sample ) , while the signal from the other was measured using a lock - in amplifier .
the resulting temperature oscillations serve as a sensitive measure of the sample heat capacity.@xcite a simple model of the a.c .- calorimetry system predicts the amplitude and phase of the temperature oscillations ( @xmath58 ) induced by a.c .
heating : @xmath59 where @xmath60 is the heating power , @xmath61 the thermal conductance to the bath , @xmath62 the heat capacity , and @xmath63 the excitation frequency , assumed to be low enough that the thermometer can follow the temperature oscillations ( the factor of @xmath64 may be assumed implicitly from this point ) .
the signal therefore contains a contribution from the specific heat and from thermal coupling to the surroundings . for frequencies @xmath65 , where @xmath66 is the cut - off frequency @xmath67 , the sample contribution dominates the signal , and @xmath68 can be considered to be inversely proportional to the heat capacity ( which we assume to be dominated by the sample ) . for @xmath69 , the signal approaches the d.c .
limit and gives a measure of the mean elevation of the sample temperature over that of the bath . for intermediate measuring frequencies , information from the phase @xmath70 , can be used to extract the specific heat : @xmath71 alternatively , one can subtract a background signal taken at a different frequency , with @xmath72 where ideally @xmath73 .
the sample temperature must also be corrected for the constant d.c .
component of the oscillatory joule heating .
this was done by repeating the measurement well below the cut - off frequency , also providing the background signal in order to estimate @xmath74 using eq.@xmath15([eq : tdc ] ) .
the cut - off frequency @xmath66 turned out to be very temperature dependent , varying between 200@xmath15hz at 0.5@xmath15k and over 2@xmath15khz at 1.5@xmath15k , presumably due to the thermal properties of the surrounding material .
fortunately , while complicating the data analysis , the reduction in @xmath66 at the lowest temperatures allows the technique to be used down to @xmath75100@xmath15 mk .
the two estimates of @xmath74 using eqs .
( [ eq : phase ] ) and ( [ eq : tdc ] ) are in good agreement below @xmath752@xmath15k .
the working frequency @xmath76 was generally of the order of @xmath66 .
sources of systematic error in the result might come from : variation of the fe thermopower under pressure ; temperature and/or frequency dependent addenda to the measured specific heat due to the pressure medium , gasket and/or anvils ; or any irreversibility or first - order character in the transitions being observed .
these potential problems will be addressed in the discussion .
we present five principal results from the sample reported in this paper , and by drawing on previous work , we aim to place our work in a broader context .
we will try to highlight common features found in many samples of cecu@xmath0si@xmath0 , one of the defining characteristics of which is its variability .
* we present the superconducting phase diagram obtained using various criteria for @xmath7 , and compare it to the widely quoted phase diagram determined under hydrostatic conditions by susceptibility . *
we examine the details of the superconducting transition , which provides some insight into the nature of the sc state and into the sample itself .
* we estimate the variation of the sommerfeld coefficient @xmath25 , with pressure , and compare it to previous results obtained by analysis of the upper critical field . *
we report the pressure dependence of the residual resistivity @xmath38 , and exponent @xmath77 , determined by a fit to the normal state resistivity of @xmath78 ( @xmath79 denoting a free exponent as opposed to the quadratic coefficient @xmath16 ) .
a comparison of @xmath80 between different samples reveals a scaling relation which can be related to the theoretical enhancement of impurity scattering .
* we explore the deviation from the scaling relation @xmath81 , which indicates the sharp change in f - electron occupation number described in the introduction .
the enhancement of @xmath7 and the other results described above are shown to occur around the same pressure .
@xmath10 in cecu@xmath0si@xmath0 determined from resistivity and specific heat measurements .
the triangles show @xmath7 determined from the onset of the resistive transition ( @xmath82 ) , the squares show its completion ( @xmath83 ) , and the filled circles show the midpoint of the specific heat jump .
the numbers indicate the sequence of pressures .
the dotted line shows @xmath7 determined by susceptibility in a different sample , also in a helium pressure medium.@xcite ] figure [ fig : tc ] shows the superconducting phase diagram determined by both resistivity and specific heat , both on increasing and decreasing the pressure .
two qualitatively different types of behavior can be seen in the same sample , represented by the onset and completion of the resistive transition .
if we follow the transition onset @xmath84 , one sees the sharp kinks similar to those seen in ref .
( dashed line ) , along with a linear decrease of @xmath7 between 3.3 and 4.8@xmath15gpa at a rate of 0.14@xmath15k@xmath85gpa@xmath86 .
superconductivity is observed however over a much smaller pressure range in our sample than in ref . .
the temperature @xmath87 , at which the resistance vanishes , behaves differently from @xmath88 .
it has a narrower peak with a maximum at slightly higher pressure .
@xmath83 agrees closely however with the transition seen in the specific heat ( see below ) . when a magnetic field was applied , @xmath83 and the specific heat anomaly shifted in agreement .
the large resistive transition widths found in cecu@xmath0si@xmath0 at high pressure are often blamed on a lack of hydrostaticity due to the pressure medium . as helium was used in this case , we can rule out pressure inhomogeneities and concentrate on the sample itself .
further information about the sc state comes from the effect of measurement current on the transition width .
for example , at 1.78@xmath15gpa high current led to the upper part of the transition disappearing , and a resistive transition can even be recovered with a narrow width comparable to that close to ambient pressure .
this is presumably due to the presence of filamentary superconductivity , with a higher @xmath7 , whose critical current density is exceeded .
these broad resistive transitions appear to be a universal feature of cecu@xmath0si@xmath0 at high pressure .
let us recall that even for the highest @xmath89 measured in a single crystal , at 2.4@xmath15k , a tail of 1% of the normal state resistivity remained well below 2@xmath15k , vanishing only at 1.5@xmath15k.@xcite the status of the superconductivity of cecu@xmath0si@xmath0 between @xmath90 and @xmath83 remains mysterious .
superconducting transition at three pressures in ( a ) resistivity and ( b ) specific heat .
note the width of the resistive transitions , and the fact that the start of the jump in specific heat coincides with the completion of the resistive transition . ]
figure [ fig : cp ] compares the superconducting transition in resistivity and specific heat at three different pressures . at 2.38@xmath15gpa
the resistive transition is broad and the sharp specific heat jump at 0.73@xmath15k begins at the point where the resistance falls to zero . at 3.67@xmath15gpa the specific heat jump , at 1.35@xmath15k ,
is much larger , and remains sharp ( and did so at intervening pressures ) , while the corresponding resistive transition has narrowed considerably . at 4.07@xmath15gpa ( not shown ) where @xmath83 has a maximum around 1.6@xmath15k in both @xmath91 and @xmath74 , the specific heat peak has already started to broaden and collapse in amplitude , while at the same pressure the resistive transition is at its narrowest since ambient pressure . as @xmath7 is driven to zero at high pressure , the superconducting @xmath74 jump becomes smaller and broader ( as shown at 4.67@xmath15gpa ) until it is no longer visible .
when the pressure was reduced , the @xmath74 peak recovered its shape , indicating the reversibility of the bulk pressure - induced behavior .
the dramatic increase in the apparent size of the superconducting jump is intriguing , and might suggest the presence of strong coupling,@xcite or other qualitative change in the sc state .
although the apparent value of @xmath92 is clearly less than the bcs ratio of 1.43 , similar a.c .
measurements on cecoin@xmath50 in an argon pressure medium indicate that there is a substantial contribution to the measured heat capacity from addenda.@xcite in helium we would expect this to be even more significant the increase in the @xmath74 jump size might itself be an artefact of the uncalibrated a.c .- calorimetry method ; nevertheless @xmath93 does appear to show a maximum at a pressure coinciding with the increase in @xmath7 .
furthermore , the assumption of strong coupling provided the best fit to @xmath94 for measurements of the upper critical field in another sample.@xcite estimate @xmath95 of the sommerfeld coefficient from a.c .- calorimetry signal at 2@xmath15k ( triangles ) , scaled for comparison with that deduced from @xmath94 measurements ( squares).@xcite the lines are guides for the eye . ] the electronic specific heat coefficient @xmath25 , and hence the effective mass @xmath96 , can be estimated by following the calorimetric signal @xmath97 , at a fixed temperature and measurement frequency above the superconducting transition , though this includes constant or slowly varying addenda from the helium , diamonds etc .
figure [ fig : gamma ] shows the estimate @xmath95 , along with the value deduced from measurements of the upper critical field in ref . .
a single constant scale factor has been introduced , showing that the two curves can be superimposed .
there is a clear anomaly in @xmath98 at 4@xmath15gpa ( just below the pressure corresponding to @xmath99 ) , superimposed on a constant reduction with pressure .
the effective mass is also reflected in the initial slope of the upper critical field @xmath100 , which in our sample also had a maximum at the same pressure as the peak in @xmath98 .
enhancement of residual resistivity in several different cecu@xmath0si@xmath0 samples , scaled to a universal pressure dependence , with @xmath101 and @xmath102 being normalizing factors .
the maximum in @xmath38 is at a pressure slightly higher than that corresponding to the maximum in @xmath7 .
the inset shows @xmath77 for a fit of @xmath78 .
filled squares indicate results from this work , the rest are reported in ref .
( s and c refer to the original labels and a retained for continuity ) . ] the residual resistivity @xmath38 , has a huge peak at a pressure slightly higher than the maximum in @xmath7 .
the magnitude of this peak varies by a factor of more than ten between samples.@xcite however , it is possible to scale the residual resistivities from different samples onto the same lorentzian curve ( see fig .
[ fig : rho0 ] ) .
a constant value @xmath102 , different for each sample , is subtracted from @xmath38 at each pressure , and the result is multiplied by a scale factor , @xmath101 ( i.e. @xmath103 ) , so that all lie on the curve defined by sample s1 , which has the highest residual resistivity ( i.e. @xmath104 ) . according to the theoretical prediction , the residual resistivity @xmath36 is given as @xcite @xmath105 where the coefficient @xmath106 depends on the band structure of host metals , @xmath107 is the concentration of impurities with moderate scattering potential @xmath108 coming from disorder other than ce ions , @xmath109 is the density of states of quasiparticles around the fermi level , and the last term represents the residual resistivity due to unitary scattering mainly arising from any deficit or defect of the ce ions .
the scaling behavior of @xmath36 shown in fig . [ fig : rho0 ] would be possible if the universal form is given by @xmath110 .
it is an open question whether the observed lorentzian form is indeed reproduced by the theory of ref . .
there is a striking correlation between the scaling factor @xmath101 and the behavior of @xmath7 .
the sample measured in helium reported in this paper and sample c1 , pressurized in steatite , both have similar values of @xmath101 , and both @xmath111 and @xmath83 agree over almost the entire pressure range .
sample s1 , with the highest @xmath38 at @xmath4 , has a lower @xmath1 ( @xmath112k ) , and the superconductivity disappears at a lower pressure .
samples c2 and s2 have scaling factors @xmath101 around 14 , and show a higher maximum @xmath7 , with superconductivity extended over a greater pressure range than in the samples with larger residual resistivities .
these differences between samples , both in @xmath38 and @xmath7 , are vastly amplified from their appearance at ambient pressure .
the inset in fig .
[ fig : rho0 ] shows the result of a fit to @xmath78 above @xmath7 .
there are two important points to note here .
firstly , at the pressure slightly higher than the maximum @xmath7 , @xmath113 is linear in @xmath3 up to about 25@xmath15k .
secondly , the exponent appears surprisingly large ( @xmath114 ) at the slightly higher pressure corresponding to the maximum @xmath38 .
this is difficult to understand without taking into account the resistivity due to impurity scattering . in sample s1 ,
reported in ref . , the residual resistivity reaches @xmath115 cm at @xmath4 , compared to a maximum of @xmath116 cm for the sample reported here .
@xmath113 then showed a falloff with temperature very similar to that of a kondo impurity system .
in other samples , this behavior is hidden by the usual positive temperature dependence of the resistivity .
contrary to the usual situation , where the lowest @xmath38 possible is sought , this example shows how samples whose residual resistivities are large at ambient pressure can reveal interesting physics at high pressure .
even if a negative temperature dependence is not seen , the power - law fit to the resistivity is affected , deviating from the linear relationship predicted in section [ sec : theory ] and leading to anomalous values of @xmath77 . at lower pressure ,
the @xmath117 coefficient is an order of magnitude larger , so ( for example ) almost linear resistivity is observed at the pressure corresponding to @xmath99 .
note that a quadratic temperature dependence of @xmath91 was recovered at the lowest temperatures when superconductivity was suppressed by a magnetic field greater that @xmath94 .
the normal state resistivity of heavy fermions can usually be understood in terms of the kondo lattice model.@xcite at high temperature the f - electron moments are localized and disordered , the resistivity is large and dominated by the scattering from spin disorder , with a characteristic @xmath118 slope . as the temperature is reduced , kondo singlets form below a characteristic temperature @xmath27 , and coherence effects in a periodic lattice cause the resistivity to drop below a maximum , at @xmath48 , which can be considered as proportional to @xmath27 . for @xmath119 away from a critical point , fermi - liquid - like behavior is recovered , with @xmath120 , where @xmath121 and reflects the hugely enhanced effective mass caused by interactions between the f - electrons . in a real system where @xmath27 is not too large , a second peak in the resistivity occurs at @xmath122 , due to the crystalline electric field ( cef ) effect@xcite ( see inset of fig .
[ fig : tmax ] ) . the low temperature behavior
then reflects the characteristics of the lowest cef - split f - level .
when pressure is applied , @xmath49 remains fairly constant , while @xmath27 rapidly increases , seen via the rise in @xmath48 .
when @xmath123 ( @xmath124 is the cef splitting between the ground and excited states ) the full 6-fold degeneracy of the j=5/2 @xmath125 multiplet is recovered , even at the lowest temperatures . as a result the resistivity maxima at @xmath48 and @xmath49 merge into a single peak.@xcite similar behavior in the magnetic component of the resistivity is found in all ce compounds studied ( such as cecu@xmath50au,@xcite cepd@xmath0si@xmath0,@xcite cepd@xmath0ge@xmath0@xcite ) . plotted against @xmath48 ( defined in inset ) , a measure of the characteristic energy scale of the system , are ( a ) the bulk superconducting transition temperature , ( b ) the residual resistivity and estimate @xmath98 of the sommerfeld coefficient , and ( c ) the coefficient @xmath16 of the @xmath126 law of resistivity , including data from cecu@xmath0ge@xmath0 .
note the straight lines where the expected @xmath127 scaling is followed .
the maximum of @xmath7 coincides with the start of the region where the scaling relation is broken , while the maximum in residual resistivity is situated in the middle of the collapse in @xmath16 .
pressure increases towards the right - hand side of the scale ( high @xmath48 ) . ] in fig .
[ fig : tmax ] the @xmath16 vs @xmath48 scaling is explored in both cecu@xmath0si@xmath0 and cecu@xmath0ge@xmath0 .
the value of @xmath16 was determined from the slope of the normal state resistivity versus @xmath47 , despite the non - fermi liquid behavior shown in the inset of fig .
[ fig : rho0 ] .
however , if one allows the exponent @xmath77 to vary between 1 and 2 , the resulting coefficient will not vary more than a factor of two , which is within the scatter of the data .
there are two regions where the predicted @xmath128 relationship is followed , separated by an abrupt drop in @xmath16 of over an order of magnitude .
the collapse of @xmath16 seems closely connected with the enhancement of superconductivity , it is at the start of this drop that @xmath7 has maximum , and the superconductivity has disappeared by the point where the @xmath129 scaling is recovered .
the residual resistivity however , peaks at around the midpoint of the drop in @xmath16 , and this is the point where @xmath4 is defined .
various unconventional properties observed around @xmath28 have been explained , at least qualitatively , by a series of theoretical investigations on the basis of an extended anderson lattice model.@xcite however , the @xmath3-linear temperature dependence of the resistivity observed in a narrow region around @xmath28 remains as yet unexplained . in ref .
, microscopic calculations showed that the static limit of the effective interaction @xmath130 between quasiparticles is enhanced greatly around @xmath131 , and is almost independent of @xmath21 , the momentum transfer , up to @xmath753/2 of @xmath132 , reflecting the local nature of critical valence fluctuations .
this implies that the valence fluctuation response function @xmath133 , is also almost @xmath21-independent in the low frequency region .
based on this observation , we present here a phenomenological theory explaining the @xmath3-linear resistivity and the enhancement of the sommerfeld coefficient @xmath25 around @xmath28 .
we adopt an exponentially decaying phenomenological form for the valence - fluctuation propagator ( dynamical valence susceptibility ) @xmath134 : @xmath135\rangle \\ & = & { k\over \omega_{\rm v}-{\rm i}\omega } , \quad{\hbox{for $ q
< q_{\rm c}\sim p_{\rm f}$ } } \label{chiv1}\end{aligned}\ ] ] where @xmath136 is the fourier component of the number of f - electron per ce site , @xmath61 is a constant of @xmath137 , and @xmath138 parameterizes the closeness to criticality .
@xmath138 is inversely proportional to the valence susceptibility @xmath139 .
the real and imaginary parts of the retarded self - energy @xmath140 respectively give a measure of the quasiparticle effective mass and lifetime .
they can be calculated using a simple one - fluctuation mode exchange process ( see fig . [
fig : feyn ] ) and given as follows : @xmath141 @xmath142 where @xmath143 is the coupling between quasiparticles and the valence fluctuation modes , and @xmath144 is the dispersion of the quasiparticle . for simplicity
, @xmath143 is assumed to be constant without wavenumber or frequency dependence .
feynman diagram for the self - energy given by eqs .
( [ chiv2 ] ) and ( [ chiv3 ] ) .
the solid line represents the green function of the quasiparticles , the wavy line the propagator of the valence fluctuations , and the filled circle the coupling between valence - fluctuation modes and the quasiparticles .
@xmath145 and @xmath146 are the matsubara frequency of the quasiparticle and fluctuation propagators , respectively . ] in typical limiting cases , ( [ chiv3 ] ) can be straightforwardly calculated in the approximation @xmath147 , where @xmath70 is the angle between @xmath148 and @xmath149 , @xmath150 is the quasiparticle velocity , and @xmath151 is assumed to be on the fermi surface , i.e. , @xmath152 : + + @xmath5 , @xmath153 : @xmath154 where @xmath155 is the cutoff wavenumber of the order of @xmath156 .
+ + @xmath157 , @xmath158 : @xmath159 where @xmath160 . since @xmath161 holds in the dominant region of q - space , the integration with respect to @xmath162 can be performed , to a good accuracy , leading to @xmath163 where we have made approximation that the range of integration is restricted as @xmath164 in which the last factor in ( [ chiv5 ] ) is approximated as @xmath165 .
then , @xmath166 the latter result , @xmath167 for @xmath168 , implies that almost all the critical valence - fluctuation modes can be regarded as classical at @xmath169 , and @xmath3-linear dependence stems from the asymptotic form of @xmath170 , essentially the classical approximation of the bose distribution function .
the real part of the self - energy , ( [ chiv2 ] ) , can be calculated easily at @xmath5 and @xmath171 , leading to @xmath172^{2 } } \nonumber \\ & & \quad\qquad + { 2\omega_{\rm v}^{2}\over[\omega_{\rm v}^{2}+(vqt)^{2}]^{2}}\ln\biggl|{\omega_{\rm v}\over vqt}\biggr| \biggr ] , \label{chiv6}\end{aligned}\ ] ] where @xmath173 . in the limit @xmath174 ,
integration with respect to @xmath175 in ( [ chiv6 ] ) leads to @xmath176 where @xmath177 .
the @xmath3-linear dependence of @xmath178 , for @xmath169 , ( [ chiv5b ] ) , implies @xmath3-linear resistivity , as the quasiparticles are subject to the large angle scattering by the critical valence - fluctuation modes .
these are effective in a wide region in the brillouin zone due to their local nature and easily couple to the umklapp process of quasiparticle scattering .
this result is consistent with the experimental fact that @xmath3-linear resistivity is observed in a narrow pressure region around @xmath4 , which is considered to correspond to a nearly critical valence transition of the ce ion .
such a @xmath3-linear dependence has been discussed in the context of high-@xmath7 cuprates with a marginal fermi liquid ( mfl ) assumption,@xcite and charge transfer fluctuations were once considered as an origin for mfl,@xcite while further theoretical models have been put forth up to now.@xcite excepting the @xmath3-linear resistivity , the present result is different from mfl behavior .
the self - energy exhibits different energy dependence , while the idea for the origin of our singular behavior shares aspects similar to the first idea of a charge transfer mechanism for high-@xmath7 cuprates.@xcite @xmath32 in the mfl model is given as @xmath179 @xcite which is indeed different from the present case [ eqs.([chiv4 ] ) and ( [ chiv7a ] ) ] . in any case
, it is to be noted that @xmath3-linear resistivity is accompanied by the peak of @xmath7 in both systems , high-@xmath7 cuprates and cecu@xmath0si@xmath0 .
the result ( [ chiv7a ] ) implies that the mass enhancement @xmath180 is expected around @xmath131 .
namely , the effective mass is given by @xmath181 where @xmath182 is the effective mass renormalized by the conventional correlation effect , leading to heavy electrons , i.e. not including the effect of critical valence fluctuations .
this latter effective mass @xmath182 exhibits a drastic decrease around @xmath28 , while the second factor in ( [ chiv8 ] ) is enhanced .
both effects should be reflected in the sommerfeld coefficient @xmath25 , so that the peak of @xmath183 is shifted to the lower pressure ( larger @xmath184 ) side , and the anomaly of @xmath25 due to the valence fluctuations may be smeared to some extent . nevertheless , some trace should be observed .
( the shift of peak of @xmath25 can be understood as the superposition of the two trends using a model @xmath185-dependence of @xmath182 and @xmath138 . )
indeed , the present experimental result presented in figs .
[ fig : gamma ] and [ fig : tmax ] may be explained by this effect .
our calorimetric results in such extreme conditions deserve some discussion , in particular the considerable apparent increase in the specific heat jump at the superconducting transition when @xmath4 is approached
. a very large specific heat jump at @xmath7 would be strongly reminiscent of the huge value found in cecoin@xmath50.@xcite it is therefore a legitimate question to ask how much the results of the uncalibrated ac calorimetry technique under pressure can be relied on to give an accurate measurement of the specific heat .
the model used to extract the specific heat from the amplitude and phase of the temperature oscillations takes no account of the heat capacity of the solid helium , diamonds , or surrounding pressure apparatus , or the essentially three - dimensional nature of the situation .
secondly , the thermopower of the fe thermocouple has been assumed not to vary with pressure ( ref .
indicates that it varies by no more than 20% up to 12 gpa ) .
nevertheless , the superconducting transition observed corresponds to @xmath186 of the signal amplitude , indicating that the addenda are a minority contribution to the total signal . runs at several different frequencies agree to within 1020% after the amplitude and phase are combined , with the discrepancy possibly due to frequency - dependent addenda .
kapitza resistance between the sample and helium is likely to better decouple the sample from its surroundings at very low temperature . if the specific heat is calculated using the two - frequency method [ eq .
( [ eq : tdc ] ) ] , the result agrees ( @xmath187 ) with that calculated using the amplitude and phase up to at least @xmath188 . given these observations
, it seems reasonable to accept our results as a good first approximation to @xmath74 , to within a constant scaling factor , and with an unknown but relatively small component due to addenda .
furthermore , the apparent anomaly in the normal state specific heat shown in fig .
[ fig : gamma ] was measured at a fixed temperature and frequency above the superconducting transition , with pressure the only independent variable .
the small peak in @xmath25 is consistent with the maximum in the initial slope of the upper critical field observed at the same pressure , though the interpretation of the latter depends on whether the sample can be considered to be in the clean or dirty limit , or somewhere in between .
having addressed the experimental questions , let us discuss some other remaining points .
the merging of @xmath189 and @xmath190 seems to be be a general feature at @xmath4 in compounds where a critical valence transition is thought to exist .
it can be understood as follows : the so - called kondo temperature @xmath191 , related to @xmath192 ( @xmath193 ) , depends crucially on the degeneracy @xmath194 of the local f - state : @xmath195 $ ] , where @xmath196 is the bandwidth of conduction electrons , @xmath197 the density of states of conduction electrons at the fermi level , and @xmath198 the c - f exchange coupling constant.@xcite even though the 6-fold degeneracy of the 4f - state is lifted by the cef effect , leaving the kramers doublet ground state and excited cef levels with excitation energy @xmath124 , the kondo temperature @xmath191 is still enhanced considerably by the effect of the excited cef levels.@xcite the technical degeneracy relevant to the kondo effect is affected by the broadening @xmath199 of the lowest cef level . if @xmath200 , the degeneracy relevant to @xmath191 is 2-fold . on the other hand ,
if @xmath201 , it increases to 4- or 6-fold .
the level broadening is given by @xmath202 where @xmath203 is the strength of c - f hybridization , and @xmath204 is the renormalization factor which gives the inverse of mass enhancement in the case of a lattice system .
it is crucial that @xmath199 is very sensitive to the valence of ce ion because @xmath204 is essentially given by @xmath21 [ eq.([eq : m*nf ] ) ] .
in particular , the factor @xmath204 increases from a tiny value in the kondo regime , @xmath205 , and approaches unity in the so - called valence fluctuation regime . since the factor @xmath206 in general for ce - based heavy electron systems , the ratio @xmath207 , which is much smaller than 1 in the kondo regime , greatly exceeds 1 across the valence transformation around @xmath28 , leading to the increase of the technical degeneracy of f - state , _ irrespective _ of the sharpness of the valence transformation .
therefore , @xmath189 should merge with @xmath190 , which corresponds to 4- or 6-fold degeneracy of 4f - state due to the effect of finite temperature , i.e. , @xmath208 .
this may be the reason why @xmath189 increases and approaches @xmath190 at pressure where @xmath7 exhibits the maximum , and the kw ratio changes between strongly and weakly correlated classes .
while the experimental picture of cecu@xmath0si@xmath0 presented in this paper is more complete than the theoretical , a large number of the features found around @xmath4 follow directly from the valence fluctuation approach and the addition of a @xmath44 term to the hamiltonian .
the linear resistivity is explained in section [ sec : theory ] , as is the local maximum in the electronic specific heat , possibly due to the renormalisation of the effective mass due to valence fluctuations , superimposed on an overall decrease with pressure .
the enhancement of the residual resistivity at low temperature follows from the renormalisation of impurity potentials by valence fluctuations .
the relative positions of the peaks in @xmath7 , @xmath25 , and @xmath38 are consistent with the valence fluctuation scenario , but for a more precise comparison more detailed calculation would be needed .
other features yet to be fully addressed with the current model are observed to occur in the valence fluctuation region .
they are the apparent increase in the specific heat jump at @xmath7 , the temperature dependence of the impurity contribution to the resistivity , and the nature of the superconducting state between the onset and completion of the superconducting transition .
the presence , and indeed enhancement , of superconductivity so far from the disappearance of magnetic order calls into question whether magnetic mediation is really the sole mechanism of superconductivity in cecu@xmath0si@xmath0 .
the evidence presented here , along with other anomalous behavior seen at a pressure well separated from the disappearance of magnetism , strongly suggests the presence of a second quantum critical point in cecu@xmath0si@xmath0 , this time related to quantum fluctuations between electronic configurations rather than to collective spin instabilities .
while magnetic pairing may be responsible for superconductivity at the magnetic qcp , critical valence fluctuations are responsible for pairing at @xmath4 .
the recent result in cecu@xmath0(si@xmath209ge@xmath210)@xmath0,@xcite where two separate peaks of @xmath7 are observed , suggests the validity of the present point of view .
schematic @xmath185-@xmath3 phase diagram for cecu@xmath0(si / ge)@xmath0 showing the two critical pressures @xmath211 and @xmath4 . at @xmath211 , where the antiferromagnetic ordering temperature @xmath212 , superconductivity in region @xmath213 is mediated by antiferromagnetic spin fluctuations ; around @xmath4 , in the region @xmath214 , valence fluctuations provide the pairing mechanism and the resistivity is linear in temperature .
the temperatures @xmath48 , and @xmath48 , merge at a pressure coinciding with @xmath4 . ]
figure [ fig : schemphadiag ] shows a schematic phase diagram for the cecu@xmath0(si / ge)@xmath0 system .
the two critical pressures , @xmath211 and @xmath4 , are respectively defined by the disappearance of magnetic order as @xmath215 tends to zero , and by the region of linear resistivity where @xmath38 has a maximum and @xmath216 , accompanied by a maximum in @xmath7 . in cecu@xmath0si@xmath0 and cecu@xmath0ge@xmath0 the two critical pressures are widely separated . in compounds such as cepd@xmath0si@xmath0 on the other hand , superconductivity is found in a narrow pocket , seemingly directly connected to the disappearance of magnetism as @xmath212 .
however , many of the other anomalies listed in table [ tab : properties ] are still observed in this system , and are difficult to explain within a purely spin fluctuation picture .
if a valence instability is present in cepd@xmath0si@xmath0 , @xmath4 is superimposed on @xmath211 , as identified by the pressure at which @xmath216.@xcite the physics associated with valence change in cecu@xmath0si@xmath0 may thus also play an important role in other heavy fermion superconductors .
linear resistivity and an enhancement of @xmath38 have also been seen in the cetin@xmath50 compounds,@xcite where t is co , rh or ir . for this family
, superconductivity extends over a relatively broad pressure range , and it may be that valence fluctuations also play a role with a critical valence pressure separate from any magnetic instability . finally , it is worth addressing the physical interpretation of the valence fluctuation mediated pairing interaction .
we emphasize that this intuitive explanation is rather speculative , but we think that it is sufficiently useful to merit inclusion .
a clue comes from the likely nearest neighbor pairing , implied by the largely local nature of the interaction , and the prediction of @xmath42-wave pairing symmetry .
one can imagine an almost filled f - band , with each occupied @xmath217 site experiencing a coulomb repulsion @xmath44 from the respective conduction electrons .
as the pressure is increased and @xmath40 moves closer to the fermi level @xmath218 , there will come a point where @xmath219 and the f - band will start to empty . on an individual @xmath220
` hole ' site , the @xmath44 interaction will be absent , thus an increased density of conduction electrons would be energetically favorable at this position .
if this extra ` screening ' conduction electron density is not strictly localized onto the atom itself , but spills onto neighboring sites , the f - electrons on ce atoms around the original ` hole ' site will feel an increased repulsion .
the tendency to transfer electrons from the f to conduction bands will be locally reinforced , explaining intuitively the increasingly catastrophic drop in @xmath23 for larger @xmath221 , predicted in ref . .
for large enough @xmath44 , phase separation would be expected to occur for some values of @xmath40 .
the attractive pairing interaction can be understood as follows : consider an isolated pair of @xmath220 ` holes ' , accompanied by their cloud of conduction electrons .
if these are separated by two lattice positions , with an intervening filled @xmath125 site , the two clouds of conduction electrons will overlap at the intermediate site , further increasing the coulomb energy at that point .
it would therefore be energetically favorable for the two ` holes ' to be on neighboring atoms , thus the attractive interaction .
the attractive interaction between ` holes ' is equivalent to that between ` electrons ' , so that this argument would give an intuitive understanding of the origin of the valence - fluctuation mechanism of superconductivity .
the enhancement of superconductivity in cecu@xmath0si@xmath0 under pressure is found to coincide with a number of anomalies in the superconducting and normal state properties that are hard to explain in a purely spin fluctuation scenario .
many of these anomalies are directly related to an abrupt change in valence of the ce ion , while others can be indirectly connected to such a transition .
we propose a second critical pressure @xmath4 at around 4.5@xmath15gpa where critical valence fluctuations provide the superconducting pairing mechanism . an extended anderson lattice model with coulomb repulsion between the conduction and f - electrons
predicts an abrupt change in ce f - level occupation .
the associated fluctuations are sufficient to explain the observed enhancement of @xmath7 , the @xmath3-linear normal state resistivity , the enhancement of the residual resistivity , and the peak in the electronic specific heat coefficient @xmath25 . a.h . and d.j .
would like to thank a. demuer , f. bouquet and a. junod for careful reading of the manuscript and invaluable comments , and r. cartoni for technical assistance .
one of the authors ( k.m . ) acknowledges stimulating communications with h.q .
yuan and f. steglich .
k.m . was supported in part by the coe program ( no.10ce2004 ) by mext of japan . | simultaneous resistivity and a.c .- specific heat measurements have been performed under pressure on single crystalline cecu@xmath0si@xmath0 to over 6 gpa in a hydrostatic helium pressure medium .
a series of anomalies were observed around the pressure coinciding with a maximum in the superconducting critical temperature , @xmath1 .
these anomalies can be linked with an abrupt change of the ce valence , and suggest a second quantum critical point at a pressure @xmath2gpa , where critical valence fluctuations provide the superconducting pairing mechanism , as opposed to spin fluctuations at ambient pressure .
such a valence instability , and associated superconductivity , is predicted by an extended anderson lattice model with coulomb repulsion between the conduction and f - electrons .
we explain the @xmath3-linear resistivity found at @xmath4 in this picture , while other anomalies found around @xmath4 can be qualitatively understood using the same model . |
the likelihood function for a complex multivariate model may not be available or very difficult to evaluate , and a composite likelihood function constructed from low - dimensional marginal or conditional distributions has become a popular alternative ( varin , 2008 ; varin , reid & firth , 2011 ) .
suppose @xmath0 is a @xmath1-dimensional random vector with probability density function @xmath2 , with a @xmath3-dimensional parameter vector @xmath4 .
given a set of likelihood functions @xmath5
, @xmath6 , defined by the joint or conditional densities of some sub - vectors of @xmath0 , the composite likelihood function ( lindsay , 1988 ) is defined as @xmath7 where the @xmath8 s are nonnegative weights and the component likelihood @xmath5 might depend only on a sub - vector of @xmath9 .
the choice of @xmath5 and the weights @xmath10 is critical for improving the efficiency of the resulting statistical inference ( lindsay , 1988 ; joe & lee , 2009 ; lindsay , yi & sun , 2011 ) . in this paper
we focus on the two most commonly used composite likelihood functions in literature , independence likelihood and pairwise likelihood , which are defined as @xmath11 and @xmath12 , respectively . given a random sample @xmath13 , where each @xmath14 is a @xmath1-dimensional vector , the composite log - likelihood function is @xmath15 and the maximum composite likelihood estimator ( mcle ) is @xmath16 .
in addition to the computational simplicity , the composite likelihood function has many appealing theoretical properties . in particular , under some regularity conditions , @xmath17 is consistent and asymptotically normally distributed with variance equal to the inverse of the godambe information matrix : @xmath18 ( lindsay , 1988 ; varin , 2008 ; xu & reid , 2011 ) . here
@xmath19 is the sensitivity matrix , and @xmath20 is the variability matrix , with the composite score function @xmath21 . throughout this paper
we use @xmath22 to denote the fisher information matrix of the full likelihood function . given two composite likelihood functions @xmath23 and @xmath24 , @xmath24
is said to be _ more efficient _ than @xmath23 if @xmath24 has a greater godambe information matrix than @xmath23 in the sense of matrix inequality .
it is well known that the full likelihood function is more efficient than any other composite likelihood function under regularity conditions ( godambe , 1960 ; lindsay , 1988 ) , i.e. @xmath25 is non - negative definite .
in general , the second bartlett identity does not hold for composite likelihood functions , i.e. @xmath26 . after lindsay ( 1982 ) , we call a composite likelihood @xmath27 _ information - unbiased _ if @xmath28 , and _ information - biased _ , otherwise
. composite likelihood - based inferential tools have been developed for hypothesis testing ( chandler & bate , 2007 ; pace , salvan , & sartori , 2011 ) and model selection ( varin & vidoni , 2005 ; gao & song , 2010 ) .
information bias of a composite likelihood can make the resulting inference more difficult .
for example , if the composite likelihood is information - unbiased , the likelihood ratio statistic has the same asymptotic chi - square distribution as its full likelihood counterpart . on the other hand
, if it is information - biased the likelihood ratio statistic converges in distribution to a weighted sum of some independent @xmath29 random variables ( kent , 1982 ) .
adjustments have been proposed to the information - biased composite likelihood ratio statistic such that the adjusted statistic has an asymptotic chi - square distribution ( e.g. , chandler & bate , 2007 ; pace , salvan , & sartori , 2011 ) .
the full likelihood function is information - unbiased , but an information - unbiased composite likelihood is not necessarily fully efficient .
in fact , any component likelihood function @xmath5 is information - unbiased .
more generally , any composite likelihood function as the product of component likelihoods with mutually uncorrelated score functions is information - unbiased . as an example , consider a @xmath1-dimensional vector @xmath30 with density function @xmath31 , and defining @xmath32 .
it is easy to show that the covariance between the score function of @xmath33 and the score function of @xmath34 is zero for any @xmath35 .
hence any composite likelihood of the form @xmath36 where @xmath37 , is information - unbiased .
conversely , an information - biased composite likelihood function can be fully efficient .
the pairwise likelihood function for the equal - correlated multivariate normal model in section 2 is fully efficient when estimating the common variance @xmath38 and the correlation coefficient @xmath39 ( cox and reid , 2004 ) , but it is not information - unbiased ( pace , salvan , & sartori , 2011 ) .
a sufficient and necessary condition for a composite likelihood to be fully efficient is given in the following theorem .
theorem 1.suppose the full likelihood function @xmath40 has the score function @xmath41 and fisher information @xmath22 .
then , for any composite likelihood function @xmath42 with the score function @xmath43 , sensitivity matrix @xmath44 , variability matrix @xmath45 and godambe information @xmath46 , @xmath47 if and only if @xmath48 with probability @xmath49 for a constant vector @xmath50 with respect to the random vector @xmath51 . proofit is easy to show that @xmath52cov@xmath53 ( lindsay , 1988 ) .
as @xmath54var@xmath55 and @xmath56var@xmath57 , the result follows as the difference @xmath58 is the covariance matrix of @xmath59 .
theorem 1 with @xmath60 gives a sufficient condition for the maximum composite likelihood estimator to coincide with the mle ( kenne pagui , salvan & sartori , 2014 ) , which is satisfied by the pairwise likelihood for closed exponential family models ( mardia et al .
, 2009 ) . in particular
, the equicorrelated multivariate normal model with unknown variance @xmath38 and correlation coefficient @xmath39 belongs to the closed exponential family , and it has been shown that @xmath60 and @xmath61 or equivalently , @xmath62 for the pairwise likelihood function ( pace , salvan , & sartori , 2011 ) .
its application in more general exponential family models has been studied in kenne pagui , salvan , & sartori ( 2014b ) . in this paper
we explore the impact of information bias on the composite likelihood inference in more detail . in section 2
we show through the equicorrelated multivariate normal model that an information - biased composite likelihood may lead to less efficient estimates of the parameters of interest when the nuisance parameters are known .
a sufficient condition is also provided for the occurrence of such a paradoxical phenomenon .
we would expect that a more efficient composite likelihood can be obtained by incorporating additional independent component likelihoods or using higher dimensional component likelihoods .
however such strategies do not always work for information - biased composite likelihood functions , as shown in section 3 .
we conclude with a discussion in section 4 .
in the presence of nuisance parameters , it is well known that the maximum likelihood estimator of the parameter of interest will have a smaller asymptotic variance when the nuisance parameters are known .
it is easy to check that this also holds for information - unbiased composite likelihood functions .
suppose the @xmath3-dimensional parameter vector @xmath9 is partitioned as @xmath63 , where @xmath64 is a @xmath65-dimensional parameter vector of interest and @xmath66 is a @xmath67-dimensional nuisance parameter vector , @xmath68 .
the godambe information matrix of a information - unbiased composite likelihood is @xmath69 , and @xmath70 where @xmath71 is the @xmath72 submatrix of @xmath46 pertaining to @xmath64 , and @xmath73 the @xmath74 sub - matrix of @xmath46 pertaining to @xmath66 .
when @xmath66 is unknown , the asymptotic variance of the mcle of @xmath64 is given by @xmath75 ; when @xmath66 is known , the asymptotic variance of the mcle of @xmath64 can be shown to be @xmath76 .
since @xmath77 is a nonnegative matrix , we have @xmath78 .
however , the reverse relationship may be observed for an information - biased composite likelihood , which is illustrated through the equicorrelated multivariate normal model in the rest of this section . from previous section we know that the pairwise likelihood @xmath79 is information - biased for this model .
* example 1 .
* suppose @xmath80 are @xmath81 independent observations from the same @xmath1-dimensional multivariate normal distribution with zero mean and covariance matrix @xmath82 , where @xmath83 is identity matrix and @xmath84 is a @xmath85 matrix with all entries equal to @xmath49 .
the common correlation coefficient @xmath39 is the parameter of interest .
the equicorrelated multivariate normal model has been well studied to compare the efficiency of pairwise likelihood and full likelihood in different settings ( arnold & strauss , 1991 ; cox & reid , 2004 ; mardia et al . , 2009 ) : when @xmath38 is unknown , the maximum pairwise likelihood estimator of @xmath39 , denoted as @xmath86 , is identical to the mle of @xmath39 and hence fully efficient ; when @xmath38 is known , the maximum pairwise likelihood estimator , denoted as @xmath87 , is less efficient than the maximum likelihood estimator of @xmath39 .
here we are interested in comparing the asymptotic variances of @xmath86 and @xmath87 .
the asymptotic variance of @xmath87 is ( cox & reid , 2004 ) @xmath88 where @xmath89 .
the asymptotic variance of @xmath86 can be shown to be @xmath90 comparing the equations ( [ eq : known ] ) and ( [ eq : unknown ] ) , we find that as @xmath39 approaches its lower bound @xmath91 , @xmath92 decreases to zero while @xmath93 does not .
the ratio of the asymptotic variances , avar@xmath94 , as a function of @xmath39 is plotted in figure [ fig2 ] for @xmath95 .
we can see that when @xmath39 is positive , @xmath87 is more efficient than @xmath86 ; when @xmath96 , the opposite phenomenon is observed , and when @xmath39 approaches the lower bound @xmath97 , this ratio diverges to infinity .
we performed the comparisons for different @xmath1 and observed the same phenomenon .
= avar@xmath94 at @xmath95
. the vertical and horizontal dashed line denotes @xmath98 and @xmath99 respectively.,height=283 ] to see that information - biasedness is not a sufficient condition for the paradox to occur , we consider another information - biased composite likelihood function , the full conditional likelihood for the same model : @xmath100 where @xmath101 denotes the random vector excluding @xmath102 .
when @xmath38 is unknown , the maximum full conditional likelihood estimator of @xmath39 , @xmath103 is identical to @xmath104 and fully efficient ( mardia et al . , 2009 )
; when @xmath38 is known , the maximum full conditional likelihood estimator , @xmath105 is less efficient than the maximum likelihood estimator for @xmath106 . using the formula in mardia , hughes , & taylor ( 2007 ) ,
the ratio of the asymptotic variances , avar@xmath107 , as a function of @xmath39 is plotted in figure [ fig3.3 ] for @xmath95 .
we can see that the ratio is less than @xmath49 for all @xmath108 $ ] . at @xmath95 .
the horizontal dashed line denotes @xmath99.,height=302 ] for the inference based on unbiased estimating equations , henmi & eguchi ( 2004 ) provided a sufficient condition for the occurrence of paradox in the presence of nuisance parameters , which also applies to the information - biased composite likelihood since its score function is a special unbiased estimating equation : proposition 1.suppose the composite likelihood function @xmath42 is information - biased and @xmath63 with the maximum composite likelihood estimators @xmath109 .
we define @xmath110 , the mcle of @xmath64 when @xmath66 is known .
then , @xmath111 has a smaller asymptotic variance than @xmath110 if @xmath112 and @xmath111 are asymptotically independent but @xmath112 and @xmath113 are dependent . in the example of equicorrelated multivariate normal model , denote by @xmath114 the maximum pairwise likelihood estimator of @xmath38 , we can show that the asymptotic covariance between @xmath104 and @xmath114 is @xmath115 which goes to @xmath116 as @xmath39 approaches @xmath91 while the asymptotic covariance between @xmath87 and @xmath114 is not equal to zero at @xmath117 .
this may explain why the paradox occurs when @xmath39 is close to its lower bound @xmath91 .
in this section we consider the inference on a single parameter only , and the simple illustrative examples allow us to calculate the godambe information matrices or asymptotic variances analytically .
consider a set of information - unbiased composite likelihood functions @xmath118 with mutually uncorrelated score functions , it is easy to show that the product @xmath119 is also information - unbiased and has godambe information matrix @xmath120 where @xmath121 is the information matrix of @xmath122 .
however , as shown in the example below , if any composite likelihood @xmath122 is information - biased , then information additivity may not hold for the product of uncorrelated composite likelihoods .
* example 2 .
* suppose the random vector @xmath123 follows a normal distribution with mean vector @xmath124 and covariance matrix @xmath125 assume @xmath38 is known , @xmath126 and @xmath39 are unknown , and @xmath126 is the only parameter of interest .
suppose we use the independence likelihood function @xmath127 , which is free of the nuisance parameter @xmath39 , to estimate @xmath126 . to incorporate the information contained in the independent variable @xmath128
, we also consider the composite likelihood function @xmath129 .
it is easy to show that the maximum composite likelihood estimators for @xmath130 and @xmath131 are @xmath132 and @xmath133 with variances @xmath134 and @xmath135/n(1 + 2\sigma^2)^2 $ ] respectively .
we can compare the variances of the two maximum composite likelihood estimators directly .
for example if @xmath136 , the variance of @xmath137 is @xmath138 which is smaller than @xmath139 if and only if @xmath140 .
note that if @xmath141 this result is expected as @xmath142 determines @xmath126 exactly with @xmath143 ; but the dependence on @xmath38 of the range of @xmath39 over which @xmath128 degrades the inference is surprising ; as @xmath38 increases this range approaches @xmath144 .
intuitively , a composite likelihood with higher dimensional component likelihoods should achieve a higher efficiency , although it usually demands more computational cost . in this subsection
we focus on comparing the independence likelihood @xmath145 and the pairwise likelihood @xmath146 .
@xmath79 can be written as the product of @xmath147 and some pairwise conditional likelihood functions . under independence ,
@xmath148 is identical to the full likelihood , and @xmath149 , which is also fully efficient . for a multivariate normal model with continuous responses ,
zhao & joe ( 2005 ) proved that the maximum pairwise likelihood estimator of the regression coefficient has a smaller asymptotic variance than the maximum independence likelihood estimator .
in fact , within the family of information - unbiased composite likelihood functions pairwise likelihood is at least as efficient as independence likelihood : each bivariate density @xmath150 has a larger information matrix than @xmath151 and the total number of the bivariate densities in @xmath79 is @xmath152 when @xmath153 .
however , the two composite likelihoods are information - biased in general especially for complex dependent data and we may observe the reverse relationship .
a bivariate binary model was used in arnold & strauss ( 1991 ) to show that the pairwise conditional likelihood could be less efficient than the independence likelihood . for a bivariate model
the pairwise likelihood is the full likelihood and hence fully efficient .
here we consider a four dimensional binary model which has a complex dependence structure but also allows us to compare the ( asymptotic ) variances of different composite likelihood estimators analytically .
* example 3 .
* suppose @xmath154 follows a multinomial@xmath155 , where @xmath156 is a positive constant and @xmath157 .
the parameter @xmath9 controls both the mean and covariance structures , and we can change the value of @xmath156 to adjust the strength of dependence .
the value of @xmath158 is completely determined by @xmath159 . given a random sample of size @xmath81 from this model , we estimate @xmath9 based on the independent triplets @xmath160 , @xmath161 . the full likelihood for the model of @xmath123 is @xmath162 solving the score equation we get the maximum likelihood estimator of @xmath9 , @xmath163 .
the exact variance of @xmath164 is @xmath165 the independence likelihood function for the model of @xmath123 is @xmath166 and we can calculate its sensitivity matrix and variability matrix as @xmath167 the pairwise likelihood function is @xmath168 and we can calculate its sensitivity matrix and variability matrix as @xmath169 where @xmath170 , @xmath171 . for @xmath172 , the asymptotic variances of the maximum composite likelihood estimators for ( [ eq : eg2full ] ) , ( [ eq : eg2ind ] ) and ( [ eq : eg2pair ] ) multiplied by @xmath81 are plotted as a function of @xmath9 in figure [ fig3.1 ] .
we can see that when @xmath173 , the three estimators perform almost equally well ; when @xmath174 , the full likelihood becomes more efficient than the independence likelihood , and the independence likelihood estimator is more efficient than the pairwise likelihood estimator .
we also carried out the comparisons for different values of @xmath156 and found that at @xmath175 , both the independence likelihood and the pairwise likelihood are fully efficient , but when @xmath176 , the independence likelihood is more efficient than the pairwise likelihood and the ratio of asymptotic variances approaches @xmath49 when @xmath177 .
when @xmath178 , the pairwise likelihood estimator is more efficient than the independence likelihood estimator and the ratio of asymptotic variances approaches @xmath49 when @xmath179 .
this example suggests that in practical applications of composite likelihood inference , where the models will usually have more complex dependence structure and incomplete data ( e.g. , yi , zeng , & cook , 2011 ) , some care is required for the use of higher dimensional composite likelihood to obtain more efficient estimators .
a hybrid composite likelihood combining lower dimensional marginal and conditional likelihoods with different weights is suggested to guarantee the improvement of efficiency ( cox & reid , 2004 ; kenne pagui , salvan , & sartori , 2014a ) .
as a complement to the discussion on composite likelihood inference in reid ( 2012 ) , in this paper we explored the impact of information bias on the composite likelihood based inference in different scenarios .
an information - unbiased composite likelihood behaves somewhat like the ordinary likelihood but it can be very inefficient . on the other hand an information - biased likelihood brings extra difficulty to the computation and is more likely to exhibit undesirable inferential properties , although the information loss can be minimized with a set of carefully selected weights .
one way to avoid the paradoxical phenomenon in section 2 is to convert the composite score function @xmath180 to an unbiased estimating function by projecting ( henmi & eguchi , 2004 ; lindsay , yi , & sun , 2011 ) : @xmath181 where @xmath182 is the score function of full likelihood , @xmath183 ranges over all @xmath184 matrices , @xmath44 and @xmath45 are the sensitivity matrix and variability matrix .
it is easy to check that @xmath185 is information - unbiased .
since @xmath44 and @xmath45 are constant matrices , this projection does not change the point estimator of @xmath9 , and @xmath185 has the same godambe information as @xmath180 . in the equicorrelated multivariate normal model with @xmath186 ,
the score function of the pairwise likelihood is @xmath187 ( pace , salvan & sartori , 2011 ) . from equation ( [ eq : project ] ) , the projected estimating funtion of @xmath188 is equal to the score function of full likelihood , @xmath182 . in complex models ,
the required computation for the projected estimating function @xmath185 can be intractable and it may be a better idea to design a nuisance - parameter - free composite likelihood function carefully for practical use . as an example , a pairwise difference likelihood that eliminates nuisance parameters in a neyman
scott problem is described in hjort & varin ( 2008 ) . | does the asymptotic variance of the maximum composite likelihood estimator of a parameter of interest always decrease when the nuisance parameters are known ? will a composite likelihood necessarily become more efficient by incorporating additional independent component likelihoods , or by using component likelihoods with higher dimension ? in this note
we show through illustrative examples that the answer to both questions is no , and indeed the opposite direction might be observed .
the role of information bias is highlighted to understand the occurrence of these paradoxical phenomenon .
+ * key words : * pairwise likelihood ; estimating function ; bartlett s second identity ; godambe information matrix ; nuisance parameter . 2.0 |
ss433 is the only known persistent superaccretor in the galaxy a source of relativistic jets ( * ? ? ?
* for review ) .
this is a massive close binary , where the compact star is most probably a black hole @xcite .
ss433 intrinsic luminosity is estimated to be @xmath13erg / s , with its maximum located in non - observed uv region . in effect this is a very blue , but heavily absorbed object ; its estimated temperature depends on the accretion disc orientation @xcite , being @xmath14k when the disc is the most open to the observer .
it is important that almost all the observed radiation is formed in the supercritical accretion disc , and the donor star contributes less than 20% of the optical radiation .
the system s extreme luminosity suggests that the black hole s mass is @xmath15m@xmath16 .
such a big energy budget of the object is supported by a very well measured kinetic luminosity of the jets , @xmath17erg / s , both in direct x - ray and optical studies of the jets and in the studies of the jet - powered nebula w50 @xcite . at the same time
we know that practically all the energy at the accretion onto a relativistic star is released in x - rays .
this means that the observed radiation of ss433 was thermalised in the strong wind coming from the supercritical disc .
the wind from a supercritical disc was first predicted by @xcite and later confirmed in radiation - hydrodynamic simulations @xcite .
these ideas led to a prediction that ss433 , being observed face - on appears as an extremely bright x - ray source , and we may expect an appearance of a new type of x - ray sources in galaxies @xcite the face - on ss433 stars .
it is quite possible that the new type of x - ray sources , ulxs ( ultraluminous x - ray sources , @xcite ) are ss433-like objects observed nearly face - on .
their observed x - ray luminosities are @xmath18erg / s and they are certainly related to the massive star population .
the disc orientation in ss433 is about edge - on , the angle between the disc plane and the line of sight is @xmath19 , due to the precessional variations .
therefore we have no chance to observe the funnel in the supercritical disc directly .
the supercritical disc radiation is not isotropic .
if one takes into account the geometrical beaming in the disc funnel with the ( half ) opening angle @xmath20 @xcite and that a supercritical disc surface radiation is local eddington @xcite , one finds @xcite , that the observed x - ray luminosity of @xmath21erg / s is expected for the face - on ss433 star .
the recent data show @xcite that ulxs posses curvature and rather flat x - ray spectra , which are difficult to interpret with a single - component or any other simple model .
the supercritical accretion discs are expected to have flat ( @xmath22 ) x - ray spectrum @xcite , because the energy release @xmath23 in the discs must be @xmath24 ( @xmath25 ) to make the disc thick due to the radiation pressure .
the x - ray luminosity of ss433 is @xmath26erg / s ( at a distance of 5.5kpc @xcite ) , four orders of magnitude less than the bolometric luminosity .
it is believed that all the x - rays come from the cooling x - ray jets .
the jet may be easily accelerated by the radiation in the hydrodynamic funnel ( e.g. , * ? ? ?
* ) of the supercritical disc .
the jet velocity value , @xmath27 , and its unique stability , where the velocity does not depend on the activity state , indicate that the jet acceleration must be controlled by the line - locking mechanism @xcite .
the funnel has to be relatively transparent for the accelerating radiation , as it was confirmed in radiation - hydrodynamic simulatuons ( e.g. , * ? ? ?
these reasonings , however , bear a problem , why do not we observe any funnel radiation , missing four orders of magnitude in the x - rays ?
even a small amount of gas in the most outer part of the funnel , where we observe the x - ray jet base , may reflect and scatter some part of the direct @xmath13erg / s of the funnel radiation .
the structure of the funnel and the jet acceleration / collimation region is unknown .
apparently , the direct funnel radiation is entirely blocked for the observer .
the standard jet model @xcite suggests that all the observed x - ray radiation of ss433 is formed in the cooling x - ray jets .
however , an analysis of the latest xmm observations @xcite led the authors to a conclusion that the standard jet model can not produce the observed x - ray continuum . in the standard jet model
there are two conical anti - parallel jets , considered identical .
the jet s gas is optically thin , being in collisional ionization equilibrium and cooling adiabatically ( or adiabatically plus radiatively ) .
the jets are observed beginning from the jet base @xmath28 ( the distance between the base of the visible jet and the black hole , different for the red and blue jets ) , which depends on the precessional phase and on the eclipses by the donor star .
the gas temperature at the jet base @xmath29 is estimated from observations .
precessional and orbital variability of ss433 is well - known @xcite , the jets ( and both the accretion disc , and the accretion disc wind ) precess with a 164-day period and an amplitude of @xmath30 . at the precessional phase @xmath31 the disc is the most open to the observer , the angle between the jets and line of sight is @xmath32 ( we refer here to the approximate values because there are nutation - like and sporadic variabilities , both of about @xmath33 ) . the accretion disc rim ( or the opaque part of the wind )
is thick @xmath34 @xcite .
comparing the amplitudes of the precessional and orbital variabilities in different x - ray bands , @xcite found that both the donor star and the outer disc rim have about the same size , therefore , one may expect the same radius for the approaching jet base @xmath35 cm . in x - ray observations with asca and chandra @xcite , dozens of x - ray emission lines formed both in blue and red jets were resolved .
the lines are variable both in positions ( in accordance with the jet kinematic model ) and in intensities , the strongest are he- and h - like @xmath36 and @xmath37 iron lines .
this allowed the iron line diagnostics of the gas temperature at the jet bases by measuring the line ratio fexxv / fexxvi .
the temperature at the base of the jet was estimated as @xmath38 kev @xcite . in the latest xmm - newton observations
@xcite estimate the temperature of about @xmath39 kev , with the main uncertainty coming from the form of the underlying continuum . fitting the continuum with a thermal bremsstrahlung model ( ginga data ,
* ) gives a notably bigger value , @xmath40 kev .
the jet base temperature determined both in the line diagnostics and in the continuum fits , drops notably in eclipses , what confirms the cooling jet model . analysing the numerous asca data
, @xcite found that the farther part of the receding jet is weakened ( probably absorbed in the external gas located in the disc plane ) and nickel is highly overabundant ( @xmath15 times ) in the jets .
the last finding has been confirmed in the xmm observations @xcite .
note that some discrepancies may exist as the system ss433 is highly variable and its appearance depends strongly on the precessional phase .
@xcite found that the xmm ss433 spectrum can not be fitted with any simple continuum law .
the numerous lines point at the thermal model , but a strong continuum curvature indicates the comptonization .
the best formal model of the additional continuum component is a broken power law with a break at @xmath41 kev , where the low energy pl rises with energy ( @xmath42 ) , the high energy part is rather steep ( @xmath43 ) , however , the parameters are not usually well determined in the fits .
the overall continuum is too hard to be bremsstrahlung , the electron - electron bremsstrahlung can not fit it even for the temperatures of 40 kev .
a highly absorbed additional thermal component gives a too strong @xmath44 kev edge .
@xcite discuss a model with a single , very broad @xmath44 kev line , which may be formed in the innermost hot jet ( otherwise hidden ) and compton down - scattered by colder material .
these uncertainties in continuum make problems for the iron - line diagnostics of the jet base temperature .
practically in all the studies of ss433 x - ray spectra the jet mass loss @xmath45 and the jet kinetic luminosity @xmath46 were determined . to estimate @xmath47
one needs to adopt the jet ( half ) opening angle @xmath48 .
the mass loss rate derived from the x - ray line intensities depends strongly on the opening angle adopted , i. e. gas density at the jet base , because the emission measure is @xmath49 .
the kinetic luminosity values derived from x - ray spectra were quite diverse , up to @xmath50erg / s .
however , in two latest studies , one from the chandra data @xcite and the other from the xmm data @xcite , the values are @xmath51 and @xmath52erg / s respectively .
they are even closer to one another , matching different distances to ss433 adopted by these authors .
@xcite found @xmath53erg / s from the jet balmer line intensities in optical spectra .
the mass loss rate in the jets was estimated @xcite on the base of w50 nebula powered by the jets with approximately the same result , @xmath53erg / s .
thus the jet kinetic luminosity was measured with relatively good accuracy .
the jet opening angle was directly measured in the studies of x - ray and optical line widths @xcite .
@xcite found @xmath54 from the chandra spectra .
they assumed that the density is uniform through the jet cone s cross section and @xmath48 is the jet s cross sectional radius .
@xcite found @xmath55 from optical spectra , assuming that the emitting gas density is gaussian - distributed in the cone s cross section .
we give here the values of @xmath48 recalculated , using the @xcite definition .
such a coincidence in the x - ray and optical jet opening angles makes the jets pure ballistic from x - ray ( @xmath56 cm ) to optical ( @xmath57 cm ) regions . in optical spectra
the jet opening angle has been estimated from h@xmath58 line profiles during 8-day long observations , the nutation broading has been taken into account and possible jet sporadic activity was diminished .
in other chandra observations @xcite the opening angle has been found twice as big and it was greater in the higher temperature ( fe ) , than in the lower temperature ( si ) parts of the jets .
regarding comparatively short time scales of the x - ray observations and probable line broading due to electron scattering @xcite in the hottest parts of the jet , we may adopt @xmath59 as a good estimate of the jet opening angle .
in this paper we analyse x - ray spectra of ss433 using xmm data and the standard jet model .
we do not try to determine the jet kinetic luminosity , because the x - ray continuum is complex and can not be explained in the standard jet model @xcite .
instead we adopt these two main parameters
@xmath60erg / s ( @xmath61g / s ) and @xmath62 , as well established and known .
we find that the line fluxes can be well matched with the standard jet model and study the additional components of the x - ray spectrum of ss433 .
we used a public archive of the xmm - newton observatory .
there are 8 observations of ss433 in the archive , where the target was in the detector s centre .
they were carried out in april 2003 ( 4 observations at the precessional phase @xmath63 ) , october 2003 ( 2 observations at @xmath31 ) , and two single observations in march 2004 and april 2004 . in this paper
we concentrate on the study of x - ray jets and the thick accretion disc ( its wind ) .
then we have selected only those observations , where disc was not eclipsed by the donor and was the most open to the observer .
this is the observation taken on 19 october 2003 in orbit 707 ( @xmath64 ) in the orbital phase @xmath65 ( for orbital and precessional phases see @xcite ) . for comparison purposes we selected the observation from 6 april 2003 in orbit 609 , taken at the intermediate inclination ( @xmath66 ) of the disc and at about the same orbital phase @xmath67 .
the last observation is relatively short , hence we added the following observation taken in the same accretion disc precessional position , in orbit 610 ( @xmath68 ) for a control check .
these data allow us to study the jets and the disc wind at two different orientations of the disc / jets and at about the same orbital phase .
with such a strong mass transfer in the system , the jet and the accretion disc funnel visibility behind the wind photosphere ( the outer disc rim ) may be dependent on the orbital phase .
we use the epic pn observations to get the highest signal - to - noise data . during all the observations
the epic pn camera was operated in the small window mode with a medium filter .
the pn data were reprocessed using the xmmsas version 6.1 .
the actual life time of the detector after a correction for good time intervals was 11.4ksec in orbit 707 , 4.4ksec in orbit 609 and 5.6ksec in orbit 610 .
to model the jet spectrum we use the standard approach which was used by many authors to analyse the x - ray spectra of ss433 @xcite .
this standard multi - temperature model describes well both the jet line spectrum and its variability .
the jet gas moves in ballistic trajectories with a velocity @xmath69 .
in such a case the gas concentration can be written as follows : @xmath70 here @xmath28 is the visible jet base ( the distance between the base and the black hole ) , @xmath71 is the solid opening angle of the jet , @xmath45 the mass loss rate in the jet , @xmath72 the mass of a proton , @xmath73 molecular weight .
the gas temperature @xmath74 along the jet accounting for adiabatic and radiation cooling is determinated by the equation : @xmath75 where @xmath76 stands for specific energy losses by radiation per one particle .
introducing new variables @xmath77 , @xmath78 , @xmath79 one may write the equation ( [ e : thermo2 ] ) in a dimensionless form : @xmath80 where @xmath58 is a coefficient depending on physical parameters of the jet @xmath45 and @xmath28 : @xmath81 the introduced value @xmath82 is the total emissivity of the plasma , and radiation losses integrated over all energies , @xmath83 the value @xmath82 is tabulated in the model of hot , optically thin plasma apec / aped , which uses the atomic database atomdb v1.3.1 ( @xcite , @xcite , @xcite ) .
atomdb provides an improved spectral modelling capability through additional emission lines , accurate wavelengths for strong x - ray transitions , and new density - dependent calculations ( for more details see http://cxc.harvard.edu/atomdb/ ) .
relativistic boosting is taken into account in this model .
the model provides a possibility of separate analysis of both line and continuum emissivities . to calculate the multi - temperature jet spectrum we divide the jet in 20 equal parts in the section @xmath84 , the gas temperature at @xmath85 drops below 100ev , where the thermal instabilities begin to operate @xcite .
one needs to consider many ( at least 20 ) mono - temperatute parts of the jet , because the radiative cooling begins to dominate the adiabatic cooling at temperatures below @xmath86kev and has a strong peak at @xmath87kev .
the total spectrum of the jet is calculated using the formula : @xmath88 where @xmath89 is the plasma emissivity and @xmath90 is the spectrum normalisation of @xmath91-th part of the jet : @xmath92 where @xmath93 is a distance to ss433 ( 5.5kpc , * ? ? ?
the integration is carried out over the jet portion volumes @xmath94 .
the average gas temperature of the @xmath91-th part of the jet @xmath95 is estimated as follows : @xmath96 the main parameters which define the physical state of the jet are gas density and temperature @xmath97 at the visible jet base .
the density depends on the mass loss rate @xmath45 and the jet ( half ) opening angle @xmath48 . as we discussed in the introduction ,
the last two parameters were relatively well measured in ss433 in previous studies not only in x - ray range .
the opening angle was particularly very well measured in optical studies . to make our analysis more certain
we do not try to determine the jet kinetic luminosity ( the mass loss rate ) from the xmm spectra , the more so , that one has serious problems with an interpretation of continuum at @xmath44 kev @xcite .
therefore , we adopt the parameters @xmath98erg / s and @xmath99 as already known . from all previous x - ray studies we may expect the gas temperature at the base is in the range of @xmath100 kev ( see introduction ) , and the visible jet base @xmath35 cm .
the last estimate is quite substantial , it was argued above , that both the donor star and the outer disc rim have about the same size @xcite .
the partial eclipses of the x - ray jets by the donor observed in ss433 prove the estimate of the x - ray jet base @xmath28 . in the thermal jet model
we adopt the solar abundance of elements , however , nickel is 10 times overabundant @xcite .
section [ reflection ] discusses this in detail .
kev , is obtained in orbit 707 ( @xmath101 ) , a slightly lower temperature in orbit 610 ( @xmath63 ) ; in orbit 609 we find the temperature @xmath102 kev with notably bigger uncertainty . ,
in this section we estimate the jet parameters @xmath103 and @xmath28 using the fluxes and the flux ratio of the iron lines fexxvk@xmath104 and fexxvily@xmath104 @xcite and adopting a simple local pl continuum .
these fexxv and fexxvi lines formed in the approaching jet are the strongest and well resolved in the xmm spectra .
the resulting line fluxes depend on the underlying continuum model .
therefore we use the simplest continuum to analyse here how stable are the derived parameters depending on @xmath45 and on the errors of the measured line fluxes .
we do not examine here different continuum models , because @xcite have studied these xmm spectra using many versions of continuum and concluded that `` the high sensitivity and wide bandpass of the xmm - newton instruments rule out any of the simple continuum models used previously '' .
instead in the section below we find clear evidences for a certain additional continuum component ( reflection ) , study the whole spectrum with these two models , thermal jet and reflection , and correct the values of @xmath103 and @xmath28 .
( in the simple local continuum model ) and mass loss rate @xmath45 for orbit 707 observation ( @xmath105 ) .
the solid curve is for @xmath106g / s ( @xmath98erg / s ) , the dashed grey curve is for @xmath107 , the solid grey curve is for @xmath108g / s .
two horizontal dashed lines mark the temperature limits ( table1 ) . ,
width=264 ] [ cols="<,<,<,<,<,<,<",options="header " , ] we estimated the column density @xmath109 of the neutral is absorbing material in the range 0.82.0 kev . in this range a strong excess of radiation is observed ( fig.[f : whole ] ) , which does not practically depend on the thermal jet model parameters .
hereat we added a bb component , which might be an indication of the supercritical accretion disc funnel ( a wind ) radiation ( see section [ disc ] ) . in fig.[f : reflobs ] we show the observed spectrum of ss433 along with the model fits jet - plus - reflection in the four spectral ranges .
table[t : approx ] presents the model parameters .
the thermal jet model is identical in all ranges , yet the reflected component was fitted independently in each range ( however there is no reflection in low energies ) .
the blue jet parameters were found in the high energy range 7.012.0 kev matching the model fluxes and flux ratios of the he- and h - like iron lines to the observed values .
we do not see any need to use non - solar abundances of the elements , but ni is clearly overabundant .
@xcite found the ni abundance to be about 8 times solar .
our fits show that ni has to be overabundant by a factor 10 , since the ni line is inside the broad absorption feature ( fig.[f : reflobs]d ) .
the red jet parameters were found in the range 4.07.0 kev , where the red iron lines are seen clearly ( fig.[f : reflobs]c ) .
the jet parameters presented in table[t : approx ] are more accurate than those presented in table1 , where we used the simplest continuum model . in the precessional phase @xmath110
we find that the blue jet visible base is of @xmath111 cm and the gas temperature in this place is @xmath112 kev .
the red jet base is of @xmath113 cm , which is naturally explained by the obscuration of the red jet by the accretion disc ( the wind ) body in this orientation of the disc . in the precessional phase @xmath114 , where one may expect an obscuration of a bigger portion of the blue jet we find a lower temperature , @xmath115 kev .
however , @xmath116 is about the same in both precessional phases . the spectra taken at @xmath114 are notably noisier than the one taken at @xmath110 .
finding accurate jet parameters is not the main purpose of the present paper .
the results depend on the natural variability in the jet activity , besides that the physical picture may be more complex ( possible gas clumping , heating effects due to kinetic energy dissipation or the funnel radiation ) . however , the derived jet visible base @xmath117 cm is too small , while it has to be about @xmath118 cm to satisfy the observed eclipses by the donor star .
the inner hotter portions of the jet are partially eclipsed @xcite .
the jet s length is scaled ( [ e : fdm ] ) as @xmath119 , and one may scale the length to about the expected value of @xmath56 cm in the thermal jet model which we use .
the reflection model produces a very good fit in the high energy range 7.012.0 kev ( @xmath120 ) , it describes well both the whole continuum and the broad absorption feature detected by @xcite .
the feature is located between the blue shifted fexxvi and nixxvii lines and partly covers these lines ( fig.[f : reflobs]d ) .
the parameters of the reflection are found , the ionization parameter of the irradiated surface is @xmath121 and the illuminating radiation index is @xmath6 .
this means a flat ( @xmath122 ) spectrum of the incident radiation in this range . in the range 4.07.0
kev we have an equally good agreement between the observed spectrum and the model .
the reflection features at 6.26.8 kev are well produced .
however , in this range there is a small continuum divergence at @xmath123 kev , probably due to the fact that the illuminating radiation index changes with energy .
we find the index @xmath7 in the range , the ionization parameter is about the same as at higher energies , @xmath124 . in the range 2.04.0
kev we observe the same effect is stronger .
the model has a some deficit at 4 kev and an excess at 2 kev .
an inclusion of the reflected component is essential in the range , otherwise one can not produce the observed continuum at 34 kev .
the incident radiation has to have an index @xmath125 in the range .
this is confirmed by a too strong emission feature at @xmath126 kev in the reflected model .
similarly to the previous spectral range we may expect that the incident radiation index is changed with energy , and here a steeper incident spectrum is expected .
the model we use is restricted by the value @xmath127 .
we find then at @xmath127 the ionization parameter @xmath128 ( with a notable scatter ) . in the soft range 0.82.0 kev the reflected component is not detected .
even if there is a reflection in the soft range , it has other parameters than those found at higher energies . from the 7.012.0 kev band to the 2.04.0 kev band
, we found a gradual decrease of the spectral index of the illuminating radiation , from @xmath129 to @xmath8 at about the same ionization parameter , @xmath5 .
however , an additional component related to the thermal jet one , is indeed present in the soft energy range ( fig.[f : whole ] and fig.[f : reflobs]a ) . choosing the simplest way we add a bb component to describe the soft excess detected in the range .
the temperature of the bb component is @xmath130 kev and it is identical in both our cases , a ( without the red jet ) and b ( with the red jet ) .
the bb component dominates at energies @xmath131 kev and it is true again in both cases .
when we do not include the red jet in the fit , the bb luminosity is @xmath123 times smaller compared to the one with the red jet .
however , this is rather a result of the increased @xmath109 in the latter case ( b ) .
the soft @xmath132 kev radiation is very sensitive to the @xmath109 value , therefore we can not estimate the real luminosity in the soft excess from the xmm spectra . in both cases
we find the is extinction value to be about the same , it is ( table[t : approx ] ) @xmath133@xmath3 in case a and @xmath134@xmath3 in case b. the interstellar extinction in ss433 , found in optical and uv observations @xcite , is @xmath135 .
this corresponds @xcite to @xmath136@xmath3 , which is in good agreement with the extinction found in x - rays .
spectral residuals in the soft energy range are bigger than the ones we have in other ranges .
the most prominent negative feature which is not produced in the model fit is at 1.21.3 kev ( fig.[f : reflobs]a ) .
however , such a feature was predicted to be present in the spectra of ulxs , if they possess a supercritical accretion disc similar to the one in ss433 , but observed nearly face - on .
@xcite ( see their fig.2 ) have simulated spectra of the multi - colour funnel model ( mcf ) , where they introduced an lc edge of oviii , blueshifted to the ss433 jet velocity .
the absorption features arise inside the funnel .
the model spectra being reduced with standard xspec models with is absorption ( @xmath137@xmath3 ) produce fake broad emission / absorption features .
the strongest one is an absorption feature at 1.21.3 kev .
our model fits to the spectra taken at the precessional phase @xmath114 give us approximately the same results ( table[t : approx ] ) .
we find a little lower temperature and less luminosity of the soft bb component . in the sketch presented in fig.[f : geometry ] the probable reflecting region
is shown at @xmath138 at @xmath139 and @xmath140 at @xmath114 on the visible outer wall of the funnel .
if the soft excess is the visible funnel wall s own radiation ( plus a possible reflection of a soft funnel radiation ) , one may expect less luminosity of the wall due to the visibility conditions in the precessional phase @xmath114 .
in this section we first discuss the structure of the supercritical accretion disc wind based on the wind approach by @xcite , and estimate the wind and the jet funnel parameters for ss433 .
next we discuss the reflected spectrum s origin and possible interpretations of the soft excess .
erg / s ) in three energy ranges : 2.04.0 kev ( @xmath127 ) , 4.07.0 kev ( @xmath141 ) and 7.012.0 kev ( @xmath142 ) . the soft bb component ( @xmath143 kev , @xmath144erg / s ) detected in the 0.82.0 kev energy range
is shown by solid line .
there are two originally the same flat spectra after absorption in absori model with @xmath145@xmath3 , @xmath146 ( grey dash - dotted line ) and @xmath147@xmath3 , @xmath148 ( bold dash - dotted line ) .
there are two mcf ( multi - colour funnel model ) non - absorbed spectra with @xmath149 cm ( bold dotted line ) and @xmath150 cm ( grey dotted line ) .
, width=302 ] the limiting eddington luminosity is @xmath151erg / s , corresponding limiting mass accretion rate is @xmath152
g / s , where the thomson opacity for a gas with solar abundance is @xmath153 @xmath154/g , and the black hole mass @xmath155 is measured in 10 solar mass units . with an accretion rate at the outer disc boundary @xmath156 ,
the disc becomes supercritical at spherisation radius @xmath157 . in its outer parts
@xmath158 the disc is subcritical , @xmath159 , below the spherisation radius @xmath160 the accretion regime is locally critical , @xmath161 and there appears a strong wind .
the total mass loss in the wind below the spherisation radius is @xmath162 .
the gas is blown away by the radiation pressure and the accretion rate is decreased linearly with radius .
the outflow velocity is close to that of keplerian , @xmath163 , for the bulk of the wind it is @xmath164 .
following @xcite , the spherisation radius is ( also see @xcite ) @xmath165 this simple picture first drawn by @xcite nevertheless explains the main features that we observe in ss433 @xcite .
it is also confirmed in the main description in radiation hydrodynamic simulations @xcite .
the simple model does not account for advection effects in the disc regions close to a black hole , however , this does not change the main picture .
the @xmath166 stays the same if @xmath156 , even the emerging spectrum of the supercritical disc does not change notably when advection is included @xcite .
adopting the mass exchange rate in ss433 @xmath167/y @xcite we find the spherisation radius @xmath168 cm and the wind velocity @xmath169 km / s , where @xmath170 is measured in @xmath171/y units .
the wind velocity is directly measured in ss433 @xcite , it is @xmath172 km / s at the polar angle @xmath173 ( disc edge - on ) and increases sharply to @xmath174 km / s at @xmath175 when the disc is the most open to observer . inside the spherisation radius
the disc luminosity is locally eddington , which gives the total accretion luminosity @xcite for ss433 @xmath176erg / s . outside the funnel
the radiation is thermalised ( @xmath177 ) , the wind s photospheric radius and its temperature in the photosphere are estimated as @xmath178 both the size of the hot body and its temperature are quite close to those we observe in ss433 .
inside the funnel where the jets are formed , the photospheric radius ( the funnel bottom ) is @xmath179 cm , where @xmath180 is the funnel s solid angle and @xmath181 .
the ss433 jet funnel is not transparent down to the black hole .
below @xmath182 the funnel walls are probably not seen directly by a face - on observer .
a multicolour funnel model ( mcf ) has been proposed for ss433 @xcite to estimate the emerging x - ray spectrum .
they considered the opaque disc wind and estimated the gas cooling and its temperature inside the wind .
in the inner parts of the wind the radiation pressure dominates and @xmath25 , in the outer parts the gas pressure dominates and @xmath183 . the supercritical disc simulations @xcite show that the @xmath25 is valid from @xmath184 ( below which the advective energy transport is important ) to @xmath185 cm , approximately the outer boundary of the computational domain .
one may estimate the radiation trapping radius in the wind outflow comparing the dynamical time - scale @xmath186 and the photon diffusion time - scale @xmath187 , where @xmath188 , @xmath189 is thomson cross - section , and @xmath190 is proton mass . the critical trapping radius in the wind ( where @xmath191 ) is @xmath192
. one may expect @xcite that radiation is trapped within the wind up to @xmath193 . in this region
the radiation transfer is local . in outer regions
@xmath194 the radiation transfer inside the wind is global and radiation cools the wind ( @xmath183 ) .
this simple approach allows us to estimate ( see below ) the soft excess in the spectrum observed in the low energy range ( fig.[f : reflobs]a ) . adopting the wind temperature of @xmath195 k at the outer photosphere one may expect ( @xmath183 ) the temperature of @xmath196 kev at the level of spherisation radius .
photons entering the funnel from the funnel s walls have multiple scatterings on the walls before they may finally escape the funnel . the probability to escape from the infinite funnel after one scattering is @xmath197 .
the probability for a photon to remain in the funnel after n scatterings is @xmath198 .
if we consider that 90% of photons escaped the funnel ( @xmath199 ) , we find the average number of scatterings @xmath200 , when @xmath181 .
the number of scatterings in ss433 funnel ( @xmath201 ) is not significantly smaller than that in the infinite funnel . at the wall temperature of less than a few kev , as we may expect ,
the absorption is more probable than electron scattering .
this means that the photons are re - radiated back to the funnel with the local wall temperature @xmath202 .
the multicolour funnel has to have an integral spectrum @xcite with numerous edges , where the strongest edges are from the most abundant elements like iron and oxygen .
the underlying spectrum with the edges formed at about zero velocity is absolutely necessary to produce the observed stable jet outflow velocity @xmath27 in the line - locking mechanism @xcite .
both the reflected component which we detect in three energy ranges ( fig.[f : reflobs ] , table [ t : approx ] ) , and the bb component in the soft energy range 0.82.0 kev , may be formed at the outer visible part of the wall shown in fig .
[ f : geometry ] at @xmath203 . at the precessional phase @xmath31 we find @xmath204 cm ( table [ t : approx ] ) .
the ionization parameter ( [ e : ionpar ] ) found at the reflecting surface is @xmath5 and it is approximately the same in all the three energy ranges . as we discussed above , the multiple scattering in the funnel
is expected , so we may consider this value of @xmath205 as effective or averaged over all putative scatterings in the funnel .
a study of the additional spectral components is out of the scope of this paper , we only make rough geometrical estimates here .
we estimate an angle @xmath206 at which the reflecting medium is exposed to the incident radiation to produce the derived ionization parameter ( [ e : ionpar ] ) , @xmath207 .
using @xmath208 , we find @xmath209 for @xmath210 , where the wind velocity is measured in units of 2000 km / s and the funnel luminosity in @xmath211erg / s units .
this corresponds to the funnel depth @xmath212 cm , which is close to the spherisation radius .
the reflected component luminosity found in the energy range of 712 kev is @xmath213erg / s ( at a distance of 5.5 kpc ) , where the @xmath214 factor is ratio of the total cone surface at @xmath138 ( fig.[f : geometry ] ) to the visible cone surface . at the precessional phase @xmath31 , the ratio is @xmath215 , and the luminosity at @xmath138 is @xmath216erg / s . in the range of 712 kev
the reflection albedo is mainly geometrical and we might expect the angle @xmath217 , where @xmath218 is the incident flux in this energy range , if the reflecting surface is directly exposed to the hard radiation from its origin . assuming that the illuminating spectrum is flat , with total luminosity of @xmath13erg / s in the spectral range of 0.1 20 kev
, we find that its luminosity in the 712 kev range is @xmath219erg / s and the angle @xmath220 .
this gives us an estimate of the funnel depth , where the illuminating radiation is formed , @xmath221 cm . the funnel photosphere radius ,
@xmath222 cm , derived from the observed values ( @xmath223 ) , is notably greater than the last estimate of the radius where the hard illuminating radiation is formed .
indeed , the reflecting surface may not directly see the deepest funnel parts .
there have to be multiple scatterings in the funnel where the harder ( @xmath10 kev ) radiation may survive absorption . in fig.[f
: add ] we show the illuminating spectrum found above in the 4
12 kev energy range , arbitrarily scaled to the luminosity of @xmath211erg / s .
we show the spectral slope in the 2 4 kev range @xmath127 , but it was noted above that it has to be steeper , @xmath125 .
the figure also shows two originally flat absorbed spectra .
these two absorbed spectra illustrate two limiting cases between which one may expect to find true picture .
even very thick ( @xmath224@xmath3 ) medium of higher ionization ( @xmath146 ) can not depress the radiation spectrum below 3 4 kev .
lower ionization medium ( @xmath148 ) does absorb the radiation effectively .
therefore we expect an existence of an absorbing ( reflecting ) medium inside the funnel , which produces such a curved spectrum of radiation , illuminating the reflected surface
. note that having he- and h - like absorption edges at about zero velocity is a mandatory property of the funnel spectrum to produce the observed stable jet velocity ( @xmath225c ) due to the line - locking mechanism @xcite .
inspecting the results of the supercritical accretion disc simulations by @xcite we find that the walls at the level of @xmath226 cm consist of a low velocity gas with the temperature of @xmath227 k and the density of @xmath228@xmath229 .
the walls are the source of strong wind , they supply the funnel with the gas . the funnel walls photosphere with the temperature from @xmath230 k at @xmath231 cm to @xmath232 k at @xmath233 cm , may be the source of the edged radiation spectrum .
the nearly flat spectrum of the funnel ( fig.[f : reflobs],[f : add ] ) , which is expected to illuminate the reflecting surface , extends at least from 4 to 12 kev .
the supercritical accretion disc spectrum is expected to be flat @xcite , however , it is supposed to go up to a few kev only . the comptonization may extend and flatten the spectrum to higher energies .
at an accretion rate of @xmath234 of the eddington rate , the local emitted spectrum in the inner region of the disc is approximated by a diluted bb with the spectral hardening factor @xmath235 @xcite .
for the accretion rates close to the eddington limit , @xmath236 increases and may amount to 10 .
this may extend the disc spectrum to higher energies . inside the disc spherisation radius ,
the disc wind supplies the funnel with gas starting from very deep regions @xcite , this gas is directly accelerated by the funnel radiation .
the interaction of the gas accelerated to the velocities of @xmath237c with new portions of the walls wind creates the conditions quite suitable for comptonization .
these conditions might be about the same throughout the whole funnel length where the jet is formed , as the high - speed outflow has about the same velocity . in the integral observations a hard x - ray emission ( 20 100 kev ) with the total hard luminosity of
@xmath238erg / s has been detected @xcite , which may be well explained @xcite in terms of comptonized thermal plasma surrounding the visible jet base . despite the fact that the jet formation and collimation mechanism is not clear yet
, one may expect the comptonization is significant inside the funnel .
the observed spectrum , taken when the disc is the most open to the observer @xcite , is nearly flat in the energy range of 1020 kev .
this spectrum may be a continuation of the flat 712 kev spectrum found here in the reflecting model .
the soft excess detected in the ss433 spectra and introduced here as a bb component with the temperature of @xmath11 kev ( table[t : approx ] , fig.[f : add ] ) , may appear to be the direct radiation of the visible funnel wall at @xmath239 ( fig.[f : geometry ] ) .
when we take into account the factor @xmath215 , which is a ratio of the total cone surface at @xmath138 to the visible cone surface , we find the soft bb component luminosity @xmath144erg / s . the soft bb component is shown in fig.[f : add ] .
note that the luminosity of the soft component strongly depends on the adopted column density @xmath109 .
it is essential here that the soft component is necessary ( * ? ? ?
* and herein ) in order to understand the ss433 spectra within the framework of the thermal jet model . in fig.[f
: add ] we show two versions of the funnel s non - absorbed spectrum in the mcf model @xcite , at @xmath240 cm ( the value of @xmath241 derived above ) and @xmath242 cm .
deriving these spectra , we adopted the wind s photosphere radius and the temperature of @xmath243 cm , @xmath244 k respectively ( [ eq : rtphot_w ] ) , and @xmath183 as it was discussed above .
this illustrates that the mcf spectra might be responsible for the soft excess observed in ss433 .
nevertheless , one can say nothing else , exept that the additional soft component is present in ss433 spectrum .
the soft excess is observed in the spectra of ulxs @xcite , their temperature is @xmath245 kev and it contributes to the total ulx luminosity of @xmath234 .
the soft component luminosity in ulxs is about @xmath246erg / s . if the ulxs are nearly face - on ss433 stars , their intrinsic soft component luminosity has to be some less because of the geometrical beaming of their supercritical discs .
however , the soft x - ray luminosity is model dependent , it strongly depends on the adopted absorption value .
the goal of the present paper was to find indications of the supercritical accretion disc funnel in ss433 x - ray radiation .
the bolometric luminosity of the disc in ss433 is @xmath13erg / s and all the luminosity must be initially released in x - rays .
this is confirmed by the well - established kinetic luminosity of ss433 jets , which is @xmath17erg / s , and the fact that the jets are formed in the deepest places of the disc funnel close to the black hole .
the observed x - ray luminosity of ss433 is @xmath26erg / s and it is the radiation of the cooling x - ray jets .
both the orientation of ss433 and the visibility conditions do not allow us to see any deep areas of the funnel .
however , the strong mass loss of the accretion disc gives us a hope for detecting some indications of the funnel radiation .
we have analysed the xxm spectra of ss433 with a well - known standard model of the adiabatically cooling x - ray jets , taking into account cooling by radiation .
we confirm that the thermal jet model reproduces the emission line fluxes quite well .
when the disc is most open to the observer , the visible blue jet base is @xmath247 cm , the gas temperature at @xmath248 is @xmath112 kev . the is gas column density is @xmath2@xmath3 , which is in good agreement with the is extinction found in optical and uv observations .
we confirm the previous finding @xcite that the red jet is probably not seen ( blocked ) in the soft energy range of 0.8 2.0 kev .
however , both the model with blocked the red jet portions , and the model with the whole red jet visible ( but with some higher @xmath109 ) give the same result , that the thermal jet model alone can not explain the soft continuum .
we also confirm that nickel is highly overabundant in the jets @xcite .
we have adopted the luminosity @xmath60erg / s and the jet opening angle @xmath62 as well established and known .
however , the derived jet visible base @xmath248 is too small , it has to be @xmath56 cm to satisfy the observed eclipses by the donor star . using the scaling formula @xmath119
one may find that during these xmm observations the jet kinetic luminosity was @xmath1erg / s .
this scaling should not change the thermal jet spectrum .
we find that the thermal jet model alone can not reproduce the continuum radiation in the xmm spectral range .
therefore we use the thermal jet model together with the reflion ionized reflection model and find quite a good representation of the spectra . introducing the reflection not only explains the continuum curvature at the energies of @xmath249 kev and the fluorescent line of quasi - neutral iron , it reproduces the broad absorption feature at @xmath250 kev recently detected by @xcite .
we independently study three energy ranges ( 24 , 47 and 712 kev ) with the same thermal jet model to find the reflection parameters .
we find that the ionization parameter ( [ e : ionpar ] ) is about the same in these ranges , @xmath5 , which indicates a highly ionized reflection surface .
the illuminating radiation photon index changes from flat , @xmath6 , in the 712 kev range to @xmath7 in the range 47 kev and to @xmath8 in the range 24 kev .
we conclude that the additional reflected spectrum is an indication of the funnel radiation .
the illuminating radiation spectrum is flat in the range of 712 kev . with multiple scatterings in
the funnel the hard radiation may survive absorption .
the reflected luminosity in this 712 kev spectral range is @xmath251erg / s .
the softer ( @xmath252 kev ) part of the illuminating spectrum ( fig.[f : add ] ) carries a trace of absorption .
this is assumed to be due to the multiple scatterings in the funnel .
it is important to note that an existence of he- and h - like absorption edges at about zero velocity is a mandatory property of the funnel spectrum to be able to produce the observed jet velocity due to the line - locking mechanism .
the supercritical accretion disc spectrum is expected to be flat , however , it extends up to a few kev only .
comptonization may extend and flatten the spectrum to higher energies .
the interaction of the high velocity gas moving along the jet axis with the walls wind makes the conditions for the comptonization throughout the whole funnel length consistent .
direct evidences of the comptonized radiation have recently been found in the integral observations @xcite .
we have not found any evidences of the reflection in the soft 0.82.0 kev energy range .
the soft excess is detected in the ss433 spectra .
we suppose that in the soft x - rays we observe direct radiation of the visible funnel wall .
we represented this component as a black body radiation with a temperature of @xmath11 kev and luminosity of @xmath253erg / s .
the soft excess may be also fitted with a multicolour funnel ( mcf ) model .
both the soft x - ray luminosity and its derived spectrum strongly depend on the column density @xmath109 .
the soft component is necessary to add in order to understand the ss433 spectra within the framework of the thermal jet model .
the soft excess is observed in the spectra of ulxs with the soft component temperature of @xmath245 kev , which is identical with that found in ss433 . if the ulxs or some of them are nearly face - on versions of ss433 , one may adjust their soft x - ray components to the outer funnel walls radiation @xcite .
the authors thank victor doroshenko for useful comments and discussions , and anna zyazeva for the correction of this manuscript .
the research was supported by rfbr grant 07 - 02 - 00909 . | we have analysed the xmm - newton spectra of ss433 using a standard model of adiabatically and radiatively cooling x - ray jets .
the multi - temperature thermal jet model reproduces well the strongest observed emission line fluxes . fitting the he- and h - like iron line fluxes , we find that the visible blue jet base temperature is @xmath0 kev , the jet kinetic luminosity @xmath1erg / s and the absorbing column density @xmath2@xmath3 .
all these parameters are in line with the previous studies .
the thermal model alone can not reproduce the continuum radiation in the xmm spectral range , the fluorescent iron line and some broad spectral features . using the thermal jet - plus - reflection model
, we find a notable contribution of ionized reflection to the spectrum in the energy range from @xmath4 to 12 kev .
the reflecting surface is highly ionized ( @xmath5 ) , the illuminating radiation photon index changes from the flat spectrum ( @xmath6 ) in the 712 kev range to @xmath7 in the range of 47 kev , and to @xmath8 in the range of 24 kev .
we conclude that the reflected spectrum is an evidence of the supercritical disc funnel , where the illuminating radiation comes from deeper funnel regions , to be further reflected in the outer visible funnel walls ( @xmath9 cm ) . in the multiple scatterings in the funnel ,
the harder radiation @xmath10 kev may survive absorption , but softer radiation is absorbed , making the illuminating spectrum curved .
we have not found any evidences of reflection in the soft 0.82 kev energy range , instead , a soft excess is detected , that does not depend on the thermal jet model details .
however the soft component spectrum is basically unknown .
this soft component might prove to be the direct radiation of the visible funnel wall .
it is represented here either as black body radiation with the temperature of @xmath11 kev and luminosity of @xmath12erg / s , or with a multicolour funnel ( mcf ) model .
the soft spectral component has about the same parameters as those found in ulxs .
[ firstpage ] massive close binaries stellar mass black holes : individual : ss433 x - ray spectra : supercritical accretion discs ulxs . |
in this paper we study the behaviour near the origin of @xmath0 positive solutions @xmath1 and @xmath2 of the system @xmath16 where @xmath4 are continuous functions .
more precisely , we consider the following question . *
question 1*. for which continuous functions
@xmath17 do there exist continuous functions @xmath18 such that all @xmath0 positive solutions @xmath1 and
@xmath2 of the system satisfy @xmath19 @xmath20 and what are the optimal such @xmath21 and @xmath22 when they exist ? we call a function @xmath21 ( resp .
@xmath22 ) with the above properties a pointwise bound for @xmath23 ( resp .
@xmath24 ) as @xmath25 .
question 1 is motivated by the results on the single semilinear inequality @xmath26 and its higher order version @xmath27 which are discussed in @xcite . although the literature on semilinear elliptic systems is quite extensive , very little of it deals with semilinear inequalities .
we mention the work of bidaut - vron and grillot @xcite in which the following coupled inequalities are studied : @xmath28 and @xmath29 another related system of semilinear elliptic inequalities appears in @xcite ( see also @xcite ) and contains as a particular case the model @xmath30 our system ( 1.1 ) is different in nature from , and and its investigation completes the general picture of semilinear elliptic systems of inequalities
. in particular ( see theorem 3.7 ) , we will obtain pointwise bounds for positive solutions of the system @xmath31 which complement the studies of systems , or .
[ rem1 ] let @xmath32 since @xmath33 is positive and harmonic in @xmath34 , the functions @xmath35 are always positive solutions of . hence , any pointwise bound for positive solutions of must be at least as large as @xmath36 and whenever @xmath36 is such a bound for @xmath23 ( resp .
@xmath24 ) it is necessarily optimal . in this case
we say that @xmath23 ( resp .
@xmath24 ) is _ harmonically bounded _ at @xmath37 .
we shall see that whenever a pointwise bound for positive solutions of exists , then @xmath23 or @xmath24 ( or both ) are harmonically bounded at 0 .
our results reveal the fact that the optimal conditions for the existence of pointwise bounds for positive solutions of are related to the growth at infinity of the nonlinearities @xmath5 and @xmath6 . in dimension @xmath8 we prove that pointwise bounds exist if @xmath38 or @xmath10 grow at most linearly at infinity ( see theorems [ thm2.1 ] , [ thm2.2 ] and [ thm2.3 ] ) . in dimensions @xmath39 we will assume that @xmath5 and @xmath6 have a power type growth at infinity , namely @xmath40 @xmath41 with @xmath15 . in this setting , we will find ( see theorem [ thm3.4 ] ) that no pointwise bounds exist if the pair @xmath42 lies above the curve @xmath43 on the other hand , if @xmath42 lies below the curve then pointwise bounds for positive solutions of always exist and their optimal estimates depend on new subregions in the @xmath44 plane ( see theorems [ thm3.1 ] , [ thm3.2 ] , [ thm3.3 ] and [ thm3.4 ] ) .
we note that the curve lies below the sobolev hyperbola @xmath45 which separates the regions of existence and nonexistence for lane - emden systems : @xmath46 ( see @xcite ) .
our analysis of combines the brezis - lions representation formula for superharmonic functions ( see appendix [ seca ] ) , a moser type interation ( see lemma [ lem4.6 ] ) , and certain pointwise estimates ( see corollary [ cor1 ] ) for the nonlinear potential @xmath47 , @xmath48 , where @xmath49 is the newtonian potential operator over a ball in @xmath50 , @xmath51 , and @xmath6 is a nonnegative bounded function .
section [ nonlin ] in this work is concerned with various pointwise and integral estimates of nonlinear potentials of havin - mazya type and their connections with wolff potentials .
since the results in this section may be of independent interest , we state them in greater generality than is needed for our study of the system . in any dimension @xmath52
, we prove that our pointwise bounds for positive solutions of are optimal .
when these bounds are not given by @xmath36 , their optimality follows by constructing ( with the help of lemma [ lem4.1 ] ) solutions @xmath23 and @xmath24 of satisfying suitable coupled conditions on the union of a countable number of balls which cluster at the origin and are harmonic outside these balls . in this case , it is interesting to point out that although our optimal pointwise bounds are radially symmetric functions , these bounds are not achieved by radial solutions of , because nonnegative radial superharmonic functions in a punctured neighborhood of the origin are harmonically bounded as @xmath53 .
we also consider the following analog of question 1 when the singularity is at @xmath7 instead of at the origin . *
question 2*. for which continuous functions
@xmath17 do there exist continuous functions @xmath54 such that all @xmath0 positive solutions @xmath1 and @xmath2 of the system @xmath55 satisfy @xmath56 @xmath57 and what are the optimal such @xmath21 and @xmath22 when they exist ? this paper is organized as follows . in sections
[ sec2 ] and [ sec3 ] we state our main results in dimensions @xmath8 and @xmath39 respectively . in section [ nonlin ]
we obtain , using hedberg inequalities and wolff potential estimates , some new pointwise and integral bounds for nonlinear potentials of havin - mazya type . using these estimates , we collect in section [ sec4 ] some preliminary lemmas while sections [ sec5 ] and [ sec6 ] contain the proofs of our main results .
in this section we state our results for questions 1 and 2 when @xmath8 .
we say a continuous function @xmath58 is _ exponentially bounded _ at @xmath7 if @xmath59 where @xmath60 if @xmath17 are continuous functions then either 1 .
@xmath5 and @xmath6 are both exponentially bounded at @xmath7 ; 2 . neither @xmath5 nor @xmath6 is exponentially bounded at @xmath7 ; or 3 .
one and only one of the functions @xmath5 and @xmath6 is exponentially bounded at @xmath7 .
our result for question 1 when @xmath8 and @xmath5 and @xmath6 satisfy ( i ) ( resp .
( ii ) , ( iii ) ) is theorem [ thm2.1 ] ( resp .
[ thm2.2 ] , [ thm2.3 ] ) below . by the following theorem ,
if the functions @xmath5 and @xmath6 are both exponentially bounded at @xmath7 then all positive solutions @xmath23 and @xmath24 of the system are harmonically bounded at @xmath37 .
[ thm2.1 ] suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system @xmath61 @xmath62 in a punctured neighborhood of the origin in @xmath63 , where @xmath17 are continuous and exponentially bounded at @xmath7 .
then both @xmath23 and @xmath24 are harmonically bounded , that is @xmath64 and @xmath65 by remark [ rem1 ] , the bounds and are optimal . by the following theorem ,
it is _ essentially _ the case that if neither of the functions @xmath5 and @xmath6 is exponentially bounded at @xmath7 then neither of the positive solutions @xmath23 and @xmath24 of the system satisfies an apriori pointwise bound at @xmath37 . [ thm2.2 ] suppose @xmath17 are continuous functions satisfying @xmath66 let @xmath67 be a continuous function satisfying @xmath68
. then there exist @xmath0 positive solutions @xmath1 and @xmath2 of the system ( [ eq2.1 ] , [ eq2.2 ] ) in @xmath69 such that @xmath70 and @xmath71 by the following theorem , if at least one of the functions @xmath5 and @xmath6 is exponentially bounded at @xmath7 then at least one of the positive solutions @xmath23 and @xmath24 of the system is harmonically bounded at @xmath37 .
[ thm2.3 ] suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system @xmath72 in a punctured neighborhood of the origin in @xmath63 , where @xmath73 is continuous and exponentially bounded at @xmath7 .
then @xmath24 is harmonically bounded , that is @xmath74 if , in addition , @xmath75 in a punctured neighborhood of the origin , where @xmath58 is a continuous function satisfying @xmath76 for some @xmath77 then @xmath78 note that in theorems [ thm2.1][thm2.3 ] we impose no conditions on the growth of @xmath79 ( or @xmath80 ) as @xmath81 . by the following theorem ,
the bounds and for @xmath23 and @xmath24 in theorem [ thm2.3 ] are optimal .
[ thm2.4 ] suppose @xmath77 is a constant and @xmath82 is a continuous function satisfying @xmath83
. then there exist @xmath84 positive solutions @xmath1 and @xmath2 of the system @xmath85 such that @xmath86 and @xmath87 the following theorem generalizes theorems [ thm2.1 ] and [ thm2.3 ] by allowing @xmath23 and @xmath24 to be negative and allowing the right sides of ( [ eq2.1 ] , [ eq2.2 ] ) to depend on @xmath88 .
[ thm2.5 ] let @xmath89 and @xmath90 be @xmath0 solutions of the system @xmath91 @xmath92 in a punctured neighborhood of the origin in @xmath63 where @xmath93 and @xmath94 are positive constants
. then @xmath95 @xmath96 the analog of theorem [ thm2.5 ] when the singularity is at @xmath7 instead of at the origin is the following result .
[ thm2.6 ] let @xmath97 and @xmath98 be @xmath0 solutions of the system @xmath99 @xmath100 in the complement of a compact subset of @xmath63 where @xmath93 and @xmath94 are positive constants .
then @xmath101 apply the kelvin transform @xmath102 and then use theorem [ thm2.5 ] .
in this section we state our results for questions 1 and 2 when @xmath39 .
we will mainly be concerned with the case that the continuous functions @xmath17 in questions 1 and 2 satisfy @xmath103 @xmath104 for some nonnegative constants @xmath94 and @xmath105 .
we can assume without loss of generality that @xmath106 .
if @xmath94 and @xmath105 are nonnegative constants satisfying @xmath106 then @xmath107 belongs to one of the following four pointwise disjoint subsets of the @xmath108-plane : @xmath109 note that @xmath110 , @xmath111 and @xmath112 are two dimensional regions in the @xmath108-plane whereas @xmath113 is the curve separating @xmath111 and @xmath112 .
( see figure 1 . ) in this section we give a complete answer to question 1 when @xmath39 and the functions @xmath5 and @xmath6 satisfy ( [ eq3.1 ] , [ eq3.2 ] ) where @xmath114 .
the following theorem deals with the case that @xmath115 .
[ thm3.1 ] let @xmath17 be continuous functions satisfying ( [ eq3.1 ] , [ eq3.2 ] ) where @xmath116 suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system @xmath117 @xmath118 in a punctured neighborhood of the origin in @xmath119 . then both @xmath23 and @xmath24 are harmonically bounded , that is @xmath120 @xmath121 by remark [ rem1 ] , the bounds and are optimal .
the following two theorems deal with the case @xmath122 .
[ thm3.2 ] let @xmath17 be continuous functions satisfying ( [ eq3.1 ] , [ eq3.2 ] ) where @xmath123 suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system ( [ eq3.4 ] , [ eq3.5 ] ) in a punctured neighborhood of the origin in @xmath124
. then @xmath125 and @xmath126 by the following theorem the bounds and for @xmath23 and @xmath24 in theorem [ thm3.2 ] are optimal .
[ thm3.3 ] suppose @xmath94 and @xmath105 satisfy and @xmath82 is a continuous function satisfying @xmath127 .
then there exist @xmath84 positive solutions @xmath1 and @xmath2 of the system @xmath128 such that @xmath129 and @xmath130 the following theorem deals with the case that @xmath131 . in this case there
exist pointwise bounds for neither @xmath23 nor @xmath24 .
[ thm3.4 ] suppose @xmath94 and @xmath105 are positive constants satisfying @xmath132 let @xmath67 be a continuous function satisfying @xmath133 then there exist @xmath84 solutions @xmath1 and @xmath2 of the system @xmath134 such that @xmath135 and @xmath136 the following theorem can be viewed as the limiting case of theorem [ thm3.2 ] as @xmath137 .
[ thm3.5 ] let @xmath73 be a continuous function satisfying where @xmath138 suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system @xmath139 in a punctured neighborhood of the origin in @xmath50 , @xmath140 .
then @xmath24 is harmonically bounded , that is @xmath141 by remark [ rem1 ] , the bound is optimal . in theorem [ thm3.7 ] we will extend some of our results to the more general system @xmath142 using these extended results and the kelvin transform , we obtain the following theorem concerning pointwise bounds for positive solutions @xmath143 and @xmath144 of the system @xmath145 in the complement of a compact subset of @xmath146 , where @xmath147 note that @xmath94 and @xmath105 satisfy if and only if @xmath148 where @xmath110 and @xmath111 are defined at the beginning of this section and graphed in figure 1 . [ thm3.6 ]
let @xmath143 and @xmath144 be @xmath0 nonnegative solutions of the system in the complement of a compact subset of @xmath50 , @xmath39 , where @xmath94 and @xmath105 satisfy .
case a. : : if @xmath149 then as @xmath150 @xmath151 case b. : : if @xmath152 then as @xmath150 @xmath153 case c. : : if @xmath154 then as @xmath150 @xmath155 case d. : : suppose @xmath156 .
let @xmath157 and @xmath158 .
then @xmath159 and as @xmath150 @xmath160 [ thm3.7 ] let @xmath1 and @xmath2 be @xmath0 nonnegative solutions of the system @xmath161 @xmath162 in a punctured neighborhood of the origin in @xmath119 , where @xmath163 and @xmath94 and @xmath105 satisfy .
case a. : : suppose @xmath149 .
+ ( a1 ) ; ; if @xmath164 then as @xmath25 @xmath165 ( a2 ) ; ; if @xmath166 then as @xmath25 @xmath167 case b. : : suppose @xmath152 .
let @xmath168\lambda+\alpha\}\ ] ] + ( b1 ) ; ; if @xmath169 then as @xmath25 @xmath170 } \right ) , & \text { if } \beta > n-(n-2)\sigma .
\end{cases } \end{aligned}\ ] ] ( b2 ) ; ; if @xmath171 then as @xmath25 @xmath172 case c. : : suppose @xmath154 . + ( c1 ) ; ; if either + 1 . @xmath173 and @xmath174 or 2 .
@xmath175 and @xmath176 + then , as @xmath25 , @xmath23 and @xmath24 satisfy and , that is @xmath177 ( c2 ) ; ; if neither ( i ) nor ( ii ) holds then as @xmath25 @xmath178}\right ) , & \text { if } \beta < n-2\\ o\left ( \left(\frac{1}{|x|}\right)^{\frac{n-2}{n}(\beta\lambda+\alpha ) } \left(\log \frac{1}{|x|}\right)^{\frac{n-2}{n}\lambda } \right ) , & \text { if } \beta\geq n-2 \end{cases}\\ \label{eq3.30 } v(x)&=o\left ( \left(\frac{1}{|x|}\right)^{n-2}\right ) + o\left ( \left(\frac{1}{|x|}\right)^\beta \log \frac{1}{|x|}\right ) .
\end{aligned}\ ] ] case d. : : suppose @xmath156 .
let @xmath179 \quad\text{and } \quad b:=\frac{\alpha}{n}[(n-2)\sigma-2]+\beta.\ ] ] then @xmath180 . + ( d1 ) ; ; if either + 1 . @xmath181 and @xmath182 ; or 2 .
@xmath183 and @xmath184 + then , as @xmath185 , @xmath23 and @xmath24 satisfy and .
( d2 ) ; ; if neither ( i ) nor ( ii ) holds then as @xmath25 @xmath186 } \right ) , & \text { if } \frac{b}{1-a}<n-2\\ o\left ( \left(\frac{1}{|x|}\right)^{\frac{n-2}{n } ( \frac{b\lambda}{1-a}+\alpha+\varepsilon ) } \right ) , & \text { if } \frac{b}{1-a}\geq n-2 \end{cases}\ ] ] and @xmath187 for all @xmath157 .
in this section we are concerned with pointwise and integral estimates of certain nonlinear potentials using inequalities of hedberg type and wolff potential estimates ( see @xcite , @xcite ) . as a consequence , we will prove the following theorem .
[ thm1 ] let @xmath188 be the unit ball in @xmath50 , @xmath189 , and let @xmath190 then , for all nonnegative functions @xmath191 , @xmath192 and @xmath193 , and @xmath194 if @xmath195 and @xmath196 , where @xmath112 is a positive constant which does not depend on @xmath5 .
more precise pointwise estimates of @xmath197 in terms of wolff potentials @xmath198 along with their analogues for functions @xmath5 defined on the entire space @xmath50 , and riesz or bessel potentials in place of @xmath199 , will be discussed below ( see theorems [ thm2][thm4 ] ) .
we remark that if @xmath200 , then for all nonnegative @xmath201 , @xmath202 where @xmath112 is a positive constant which depends only on @xmath105 and @xmath203 .
there are similar pointwise estimates in the range @xmath204 under the additional assumption that @xmath205 is uniformly bounded , for instance , if @xmath206 .
these relations between nonlinear potentials @xmath207 and @xmath208 are due to havin and mazya , d. adams and meyers ( see @xcite , @xcite ) .
let @xmath209 be a nonnegative borel measure on @xmath50 .
for @xmath210 , the riesz potential @xmath211 of order @xmath212 is defined by @xmath213 for @xmath214 and @xmath215 , the wolff potential @xmath216 is defined by ( see @xcite , @xcite ) : @xmath217 there is also a nonhomogeneous version applicable for @xmath218 , @xmath219 where @xmath220 .
wolff potentials have numerous applications in analysis and pde ( see , for instance , @xcite , @xcite , @xcite , @xcite )
. we will also use the havin - mazya potential @xmath221 , where @xmath222 and @xmath215 , defined by : @xmath223 along with its nonhomogeneous analogue @xmath224 , where @xmath222 and @xmath218 , defined by : @xmath225 here bessel potentials @xmath226 with bessel kernels
@xmath227 , @xmath228 , are used in place of riesz potentials @xmath229 . clearly , @xmath230 , and hence @xmath231 for all @xmath232 .
note that the newtonian potential coincides with @xmath233 if @xmath234 , where @xmath235 and @xmath236 , we will denote the corresponding potentials by @xmath237 , @xmath238 , etc .
we will need the hardy - littlewood maximal function @xmath239 throughout this section @xmath240 , @xmath241 , @xmath242 , etc .
, will stand for constants which depend only on @xmath212 , @xmath243 , and @xmath203 .
the following pointwise estimates of nonlinear potentials are due to havin and mazya , and d. r. adams .
( see , for instance , ( * ? ? ?
10.4.2 ) and the references therein . )
[ thm2 ] let @xmath209 be a locally finite nonnegative borel measure on @xmath50 .
\(a ) if @xmath222 and @xmath244 , then , for some @xmath245 , @xmath246 ( b ) if @xmath247 and @xmath244 , then , for some @xmath245 , @xmath248 ( c ) if @xmath249 , @xmath250 , and @xmath251 , then @xmath252 ( d ) the above estimates with @xmath253 hold for the potential @xmath221 in place of @xmath254 if @xmath255 . theorem [ thm2 ] yields that the wolff potential @xmath256 is pointwise equivalent to the havin - mazya potential @xmath257 ( and @xmath258 is equivalent to @xmath254 if @xmath220 , up to a choice of @xmath240 ) , provided @xmath247 . in the range @xmath249 ( which excludes the critical case @xmath259 ) , the sharp upper estimate for @xmath221 fails , along with its counterpart for @xmath224 . however , there are natural substitutes under the additional assumption that the corresponding nonlinear potential is uniformly bounded .
the following theorem is due to adams and meyer ( see ( * ? ? ?
10.4.2 ) ) . [ thm3 ]
suppose @xmath260 for all @xmath261 .
\(a ) if @xmath262 and @xmath215 , then , for some @xmath245 , @xmath263 ( b ) if @xmath264 and @xmath250 , then , for some @xmath265 , @xmath266 ( c ) the above estimates with @xmath253 hold for the potential @xmath221 in place of @xmath254 .
we now deduce some pointwise bounds for nonlinear potentials .
[ thm4 ] let @xmath267 and @xmath268 .
then the following estimates hold .
\(a ) if @xmath269 , @xmath250 , and @xmath270 , where @xmath271 , then , for all @xmath232 , @xmath272 ( b ) if @xmath222 , @xmath250 , and @xmath273 , where @xmath274 , then , for all @xmath232 , @xmath275 ( c ) if @xmath276 , @xmath218 , and @xmath270 , where @xmath277 , then , for all @xmath232 , @xmath278 ( d ) if @xmath222 , @xmath279 , and @xmath273 , where @xmath277 , then , for all @xmath232 , @xmath280 suppose first that @xmath281 .
fix @xmath282 , and let @xmath283 , where @xmath235 and @xmath284 , @xmath285 . then by , @xmath286 is bounded above by a constant multiple of @xmath287 @xmath288 note that @xmath289 hence , @xmath290 to estimate @xmath291 , notice that by hlder s inequality with @xmath285 , @xmath292 hence , @xmath293 letting @xmath294 , and combining the preceding inequalities , we obtain @xmath295 minimizing the right - hand side over @xmath296 gives , with @xmath297 , @xmath298 as noted above , the preceding inequality holds for @xmath299 in place of @xmath300 if we set @xmath253 : @xmath301 @xmath302 provided @xmath303 , @xmath215 and @xmath304 .
this proves statement ( a ) of theorem [ thm4 ] . if @xmath305 , then noticing that @xmath306 for every @xmath232 , we deduce from this a cruder estimate : @xmath307 in the case @xmath308 , we have @xmath309 . using instead of
, we see that still holds .
therefore holds for all @xmath310 and @xmath304 .
if @xmath311 , then obviously @xmath312 . consequently , for @xmath311 follows from the corresponding estimate with @xmath313 , which proves statement ( b ) of theorem [ thm4 ] . to prove statement ( c ) , notice that in the case @xmath314 and @xmath303 , we clearly have , by looking at the cases @xmath180 and @xmath315 , @xmath316 hence yields @xmath317 for @xmath318 and @xmath303 , which proves statement ( c ) . in the case @xmath319 and @xmath308 , as in the estimates of @xmath299 above , we again use in place of , together with @xmath320 , to complete the proof of statement ( d )
. inequality can be deduced directly applying hedberg s inequality ( see @xcite , proposition 3.1.2(a ) ) to @xmath321 where @xmath322 , followed by sobolev s inequality .
theorem [ thm1 ] is immediate from statements ( b ) and ( d ) of theorem [ thm4 ] with @xmath323 and @xmath324 .
we will need the following corollary of theorem [ thm1 ] in the next section .
let @xmath325 be a nonnegative function where @xmath326 is a ball in @xmath50 , @xmath51 , and let @xmath327 then for @xmath328 we have @xmath329 where @xmath330 is a positive constant .
apply theorem [ thm1 ] with @xmath313 to the function @xmath331 .
in this section we provide some lemmas needed for the proofs of our results in sections [ sec2 ] and [ sec3 ] .
[ lem4.1 ] let @xmath332 be a continuous function such that @xmath333 .
let @xmath334 be a sequence in @xmath50 , where @xmath39 ( resp .
@xmath8 ) , such that @xmath335 and @xmath336 let @xmath337 be a sequence satisfying @xmath338 then there exist a positive constant @xmath339 and a positive function @xmath340 where @xmath341 ( resp .
@xmath342 ) such that @xmath343 @xmath344 @xmath345 @xmath346 let @xmath347 $ ] be a @xmath84 function whose support is @xmath348 .
define @xmath349 $ ] by @xmath350 where @xmath351 . then @xmath352 where @xmath353 .
let @xmath354 and @xmath355 since the functions @xmath356 have disjoint supports , @xmath357 and by , , and we have @xmath358 case i. : : suppose @xmath39 . then for @xmath359 and @xmath360 we have @xmath361 thus letting @xmath362 where @xmath363 is a fundamental solution of @xmath364 we have @xmath23 satisfies with @xmath365 and @xmath366
also @xmath367 and @xmath23 clearly satisfies and .
: : suppose @xmath8 .
then for @xmath359 and @xmath360 we have @xmath368 = i\varepsilon_j \log\frac{1}{r_j}. \end{aligned}\ ] ] thus letting @xmath369 we have @xmath23 satisfies with @xmath370 and @xmath371 also @xmath372 and @xmath23 clearly satisfies and .
[ lem4.2 ] if @xmath282 and @xmath373 , then @xmath374 for all @xmath375 where @xmath376 .
denote the left side of by @xmath377 .
let @xmath378 and @xmath379 . then @xmath380 [ lem4.3 ] let @xmath23 be a @xmath0 nonnegative superharmonic function in @xmath381 satisfying @xmath382 where @xmath383 and @xmath384 is a continuous nondecreasing function satisfying @xmath385 .
then @xmath386 let @xmath387 be a sequence which converges to the origin .
it suffices to prove with @xmath88 replaced with @xmath388 . by there exists @xmath389 such that @xmath390 thus @xmath391 define @xmath392 by @xmath393 by lemma [ a.1 ] .
thus @xmath394 and by @xmath395 hence by lemma [ a.1 ] we get @xmath396\\ & \quad + c\int_{|y - x_j|<r_j } \left(\log\frac{1}{|y - x_j|}\right ) \exp\left(ah\left(\log\frac{2}{|x_j|}\right ) \right)\,dy\\ & \leq c\left [ \log\frac{1}{|x_j|}+r^{2}_{j } \left(\log\frac{1}{r_j}\right ) \exp\left ( ah\left(\log\frac{2}{|x_j|}\right ) \right ) \right]\\ & \leq c\log\frac{1}{|x_j| } + o\left(h\left(\log\frac{2}{|x_j|}\right ) \right ) \quad \text{as } j\to\infty \end{aligned}\ ] ] by .
[ lem4.4 ] let @xmath23 be a @xmath0 nonnegative function in @xmath397 , satisfying @xmath398 where @xmath399 , @xmath400 , and @xmath401 are constants . 1 .
if @xmath402 then @xmath403 2 . if @xmath404 and @xmath405 then @xmath406 3 .
if @xmath402 , @xmath407 , and @xmath2 is a @xmath0 nonnegative solution of @xmath408 where @xmath409 and @xmath410 then as @xmath25 we have + v(x)= o ( ( ) ^n-2 ) + o ( ( ) ^[(n-2)-2]+ ) & if @xmath156 [ eq4.19 ] + o ( ( ) ^n-2 ) + o ( ( ) ^ ) & if @xmath154 [ eq4.20 ] .
let @xmath411 be a sequence which converges to the origin .
it suffices to prove , , , and with @xmath88 replaced with @xmath388 . for the proof of part ( i )
we can assume @xmath405 because increasing @xmath412 to @xmath203 weakens condition and does not change the estimate . by there exists @xmath389 such that @xmath413 thus @xmath414 define @xmath415 by @xmath416 by lemma [ a.1 ] .
then @xmath417 because @xmath405 and @xmath401 . hence by lemma [ a.1 ] and we have @xmath418\\ & \leq c\left [ \frac{1}{|x_j|^{n-2 } } + \int_{|y - x_j|<\frac{|x_j|}{2 } } \frac{-\delta u(y)\,dy}{|y - x_j|^{n-2 } } \right]\\ & \leq c \left [ \frac{1}{|x_j|^{n-2 } } + \int_{|y - x_j|<r_j } \frac{a\left ( \frac{1}{|x_j|}\right)^\gamma \left(\log \dfrac{1}{|x_j|}\right)^q}{|y - x_j|^{n-2}}\,dy\right]\\ & \leq c \left [ \left(\frac{1}{|x_j|}\right)^{n-2 } + \left(\frac{1}{|x_j|}\right)^\gamma \left(\log\frac{1}{|x_j|}\right)^q r^{2}_{j } \right]\\ & \leq c \left [ \left(\frac{1}{|x_j|}\right)^{n-2 } + o\left ( \left(\frac{1}{|x_j|}\right)^\gamma \left(\log\frac{1}{|x_j|}\right)^q \right)^{\frac{n-2}{n } } \right ] \quad\text{as } j\to\infty\end{aligned}\ ] ] by , which proves parts ( i ) and ( ii ) .
we now prove part ( iii ) .
for @xmath419 we have by and lemma [ a.1 ] that @xmath420^\sigma \\ & \leq c\left [ \frac{1}{|x_j|^{\sigma(n-2 ) } } + \left ( \left ( n_{b_{\frac{|x_j|}{2}}(x_j ) } ( -\delta u)\right ) ( x)\right)^\sigma \right ] \end{aligned}\ ] ] where @xmath421 .
thus by lemma [ a.1 ] @xmath422\\ & \leq c \left [ \frac{1}{|x_j|^{n-2 } } + \frac{|x_j|^{-\beta}}{|x_j|^{\sigma(n-2)-2 } } + |x_j|^{-\beta}(h(-\delta u))(x_j ) \right ] \end{aligned}\ ] ] where @xmath423 . case i. : : suppose @xmath424 . then using and with @xmath402 in corollary [ cor1 ]
we get @xmath425 thus follows from .
: : suppose @xmath426 .
then using , , and with @xmath402 in corollary [ cor1 ] we get @xmath427 thus follows from .
[ lem4.5 ] suppose @xmath1 and @xmath2 are @xmath0 nonnegative solutions of the system @xmath428 @xmath429 in a punctured neighborhood of the origin in @xmath50 , @xmath430 , where @xmath409 . 1 .
if @xmath431 then @xmath432 2 .
if @xmath105 and @xmath94 satisfy and @xmath433 in a punctured neighborhood of the origin , where @xmath434 , then for some @xmath407 we have @xmath435 choose @xmath436 such that @xmath1 and @xmath2 are @xmath0 nonnegative solutions of the system ( [ eq4.25 ] , [ eq4.26 ] ) in @xmath437 .
let @xmath438 be a sequence in @xmath50 such that @xmath439 it suffices to prove and with @xmath88 replaced with @xmath388 .
by lemma [ a.1 ] we have @xmath440 and , for @xmath419 , @xmath441\ ] ] @xmath442\ ] ] where @xmath443 does not depend on @xmath444 or @xmath88 . by , we have as @xmath445 that @xmath446 define @xmath447 by @xmath448 where @xmath449 . making the change of variables @xmath450 in , , and we get @xmath451 and @xmath452 \quad\text{for}\quad |\xi|<1\ ] ] @xmath453 \quad\text{for}\quad|\xi|<1.\ ] ] where @xmath454 .
we now prove part ( i ) .
if @xmath149 then part ( i ) follows from lemma [ lem4.4](i ) .
hence we can assume @xmath455 .
define @xmath456 and @xmath457 by @xmath458 it follows from and riesz potential estimates ( see ( * ? ? ?
* lemma 7.12 ) ) that @xmath459 in @xmath460 and hence @xmath461 thus by hlder s inequality @xmath462 where @xmath36 is given by . by and we have @xmath463 and for @xmath360 it follows from and that @xmath464 substituting in and using , we get @xmath465 which completes the proof of part ( i ) . next we prove part ( ii ) .
since increasing @xmath94 and/or @xmath105 weakens the conditions ( [ eq4.26 ] , [ eq4.28 ] ) on @xmath23 and @xmath24 we can assume instead of that @xmath466 since , for @xmath467 $ ] , @xmath468 and @xmath469 it follows from , and that for @xmath467 $ ] we have @xmath470 \quad\text{for } |\xi|<r\ ] ] and @xmath471 \quad\text{for } |\zeta|<2r\ ] ] where @xmath112 is independent of @xmath472 , @xmath473 , @xmath444 , and @xmath296 .
it therefore follows from ( [ eq4.26 ] , [ eq4.28 ] ) that for @xmath467 $ ] we have @xmath474 \right)^\lambda\\ & \leq cr^{-\alpha-(n-2)\lambda}_{j } \left [ \frac{1}{r^{(n-2)\lambda } } + \left ( ( n_{2r } g_j)(\xi)\right)^\lambda \right ] \quad \text { for } |\xi|<r , \end{aligned}\ ] ] and @xmath475 \right)^\sigma \\ & \leq cr^{-b}_{j } \left [ \frac{1}{r^{(n-2)\sigma } } + \left ( ( n_{4r } f_j)(\zeta ) \right)^\sigma \right ] \quad\text { for } |\zeta|<2r , \end{aligned}\ ] ] where @xmath476 .
thus for @xmath477 we have @xmath478(\xi)\right)^\lambda\\ & \leq cr^{(n - b)\lambda}_{j } \left [ r^{(2-(n-2)\sigma)\lambda } + \left ( ( m_{4r } f_j ) ( \xi)\right)^\lambda \right ] \end{aligned}\ ] ] where @xmath479 .
hence by there exists a positive constant @xmath93 which depends only on @xmath203 , @xmath212 , @xmath480 , @xmath94 , and @xmath105 such that @xmath481 by there exists @xmath482 such that @xmath483 to prove for some @xmath407 that holds with @xmath484 , it suffices by the definition of @xmath485 and @xmath486 to show for some @xmath457 that the sequence @xmath487 to prove and thereby complete the proof of lemma [ lem4.5](ii ) , we need the following result . [ lem4.6 ] suppose the sequence @xmath488 for some constants @xmath489 , @xmath490 , and @xmath467 $ ] .
let @xmath491 where @xmath93 is as in .
then there exists a constant @xmath492 such that the sequence @xmath493 provided @xmath494 $ ] and @xmath495 it follows from that @xmath496 we can assume @xmath497 for otherwise it follows from riesz potential estimates ( see ( * ? ? ?
* lemma 7.12 ) ) and that the sequence @xmath498 is bounded in @xmath499 and hence by we see that holds for all @xmath494 $ ] .
define @xmath500 by @xmath501 where @xmath502 is as in . by
, @xmath503 and by riesz potential estimates we have @xmath504 where @xmath505 . since , by , @xmath506 we have @xmath507 we can assume @xmath508 for otherwise by riesz potential estimates and we have @xmath509 which is bounded by . hence implies holds for all @xmath510 $ ] .
define @xmath511 and @xmath512 by @xmath513 by and , @xmath514 and by riesz potential estimates @xmath515 by .
it follows therefore from that @xmath516 which is a bounded sequence by . to complete the proof of lemma [ lem4.6 ] , it suffices to show @xmath517 for some @xmath518 because if holds for some @xmath519 satisfying then it clearly holds for all @xmath519 satisfying . by and we have @xmath520 case i. : : suppose @xmath521 . then by and @xmath522 case ii .
: : suppose @xmath523 .
then , by , @xmath524\\ & = \frac{(n-(2-\varepsilon))\lambda}{n } \left [ \frac{2-\varepsilon}{n-(2-\varepsilon)}+\frac{n}{n-(2-\varepsilon ) } \frac{1}{\lambda}-\sigma \right]\\ & = c_2(n,\lambda,\sigma)>0 \end{aligned}\ ] ] by . thus holds with @xmath525 .
this completes the proof of lemma [ lem4.6 ] .
we return now to the proof of lemma [ lem4.5](ii ) . by , the sequence @xmath526 is bounded in @xmath527 . starting with this fact and iterating lemma [ lem4.6 ] a finite number of times (
@xmath528 times is enough if @xmath529 ) we see that there exists @xmath530 and @xmath407 such that sequence @xmath531 is bounded in @xmath532 . in particular holds .
this completes the proof of lemma [ lem4.5](ii ) .
in this section we prove theorems [ thm2.1][thm2.5 ] . the following theorem with
@xmath533 immediately implies theorems [ thm2.1 ] and [ thm2.3 ] .
we stated theorems [ thm2.1 ] and [ thm2.3 ] separately in order to clearly highlight the differences between possibilities ( i ) and ( iii ) which are stated at the beginning of section [ sec2 ] .
[ thm5.1 ] suppose @xmath1 and @xmath2 are @xmath0 positive solutions of the system @xmath534 @xmath535 in a punctured neighborhood of the origin in @xmath63 , where @xmath73 is a continuous function satisfying @xmath536 then @xmath24 is harmonically bounded , that is @xmath537 for some constant @xmath389 . if , in addition , @xmath538 in a punctured neighborhood of the origin , where @xmath58 is a continuous function satisfying @xmath539 for some continuous nondecreasing function @xmath540 satisfying @xmath541 then @xmath542 for simplicity and to motivate theorem [ thm2.5 ] , we stated theorem [ thm2.3 ] for the special case @xmath533 rather than for more general @xmath543 as in theorem [ thm5.1 ] . also , the bound in theorem [ thm2.3 ] is optimal by theorem [ thm2.4 ] , whereas in general we can only show the bound in theorem [ thm5.1 ] is essentially optimal ( see theorem [ thm6.2 ] ) .
since @xmath23 is positive and superharmonic in a punctured neighborhood of the origin , there exists a constant @xmath544 such that @xmath545 in @xmath437 .
choose a positive constant @xmath546 such that @xmath547 for @xmath548 . then @xmath24 is a @xmath0 positive solution of @xmath549 since @xmath23 and @xmath24 are positive and superharmonic in @xmath437 , we have by lemma [ a.1 ] that @xmath550 and @xmath551 where @xmath552 , @xmath553 are constants and @xmath21,@xmath554 are harmonic functions .
suppose for contradiction there exists a sequence @xmath555 such that @xmath556 as @xmath445 and @xmath557 since , for @xmath558 , @xmath559 and similarly for @xmath24 , it follows from and that @xmath560 where @xmath112 does not depend on @xmath444 or @xmath88 . substituting @xmath484 in and using we get @xmath561
also , implies @xmath562 let @xmath563 and define @xmath564 by @xmath565 making the change of variables @xmath566 in , , and and using we get @xmath567 @xmath568 and @xmath569 where @xmath570 and @xmath112 does not depend on @xmath444 or @xmath472 .
let @xmath571 where @xmath572 then letting @xmath573 ) , it follows from that @xmath574 ( the idea of using jensen s inequality as above is due to brezis and merle @xcite . )
thus by and hlder s inequality @xmath575 hence , defining @xmath576 by @xmath577 it follows from and that @xmath578 by and we have @xmath579 and @xmath580 for fixed @xmath444 , think of @xmath581 as the density of a distribution of mass in @xmath582 satisfying , , and . by moving small pieces of this mass nearer to the origin in such a way that the new density ( which we again denote by @xmath581 ) does not violate , we will not change the total mass @xmath583 but @xmath584 will increase .
thus for some @xmath585 the functions @xmath586 satisfy , , and which , as elementary and explicit calculations show , is impossible because @xmath587 as @xmath445 .
this contradiction proves . since @xmath2 is positive and superharmonic , @xmath24 is bounded below in some punctured neighborhood of the origin by some constant @xmath588 .
hence by we have @xmath589 also by there exists a positive constant @xmath112 such that @xmath590 hence for @xmath591 small and positive we have by that @xmath592 where @xmath593 .
thus follows from lemma [ lem4.3 ] .
define @xmath594 by @xmath595 then @xmath596 is nondecreasing . by ,
@xmath597 as @xmath598 and there exists @xmath599 such that @xmath600 for @xmath601 .
thus @xmath602 define @xmath332 by @xmath603 and let @xmath604 , and @xmath110 be as in lemma [ lem4.1 ] . by holding @xmath388
fixed and decreasing @xmath485 we can assume @xmath605 @xmath606 and @xmath607 let @xmath608 . by lemma [ lem4.1 ] there exists a positive function @xmath340 which satisfies . by and we have @xmath609 which implies .
also for @xmath610 and @xmath611 it follows from , , , and that @xmath612 thus @xmath23 satisfies @xmath613 in @xmath614 . by , @xmath23 satisfies in @xmath615 .
thus @xmath23 satisfies in @xmath616 .
taking @xmath617 completes to proof of theorem [ thm2.2 ] .
the following theorem with @xmath533 , @xmath77 , immediately implies theorem [ thm2.4 ] .
[ thm6.2 ] suppose @xmath540 and @xmath82 are continuous nondecreasing functions satisfying @xmath618 then there exist @xmath84 positive solutions @xmath1 and @xmath2 of the system @xmath619 such that @xmath620 and @xmath621 let @xmath622 .
then @xmath24 satisfies @xmath623 and .
define @xmath332 by @xmath624 .
let @xmath625 be as in lemma [ lem4.1 ] and @xmath626 . by taking a subsequence if necessary
, it follows from @xmath627 that @xmath485 satisfies .
therefore , by lemma [ lem4.1 ] , there exists a positive function @xmath372 and a positive constant @xmath110 such that @xmath23 satisfies . thus @xmath23 satisfies @xmath627 in @xmath628 .
also for @xmath610 we have @xmath629 hence @xmath23 satisfies @xmath627 in @xmath630 .
finally @xmath631 which proves .
define functions @xmath23 and @xmath24 by @xmath632 then @xmath23 and @xmath24 are @xmath0 positive solutions of @xmath633 in a punctured neighborhood of the origin
. thus follows from theorem [ thm2.3 ] .
hence by @xmath634 thus follows from lemma [ lem4.3 ] .
in this section we prove theorems [ thm3.1][thm3.7 ] .
since increasing @xmath105 and/or @xmath94 weakens the conditions ( [ eq3.1 ] , [ eq3.2 ] ) , we can assume @xmath635 .
as in the first paragraph of the proof of theorem [ thm5.1 ] , there exist positive constants @xmath546 and @xmath636 such that @xmath23 and @xmath24 are positive solutions of the system @xmath637 let @xmath638 .
then in @xmath639 we have @xmath640 thus by ( * ? ? ?
* theorem 2.1 ) @xmath641 which proves and . as in the first paragraph of the proof of theorem [ thm5.1 ]
, we can assume the function @xmath6 is given by @xmath642 and then theorem [ thm3.5 ] follows immediately from lemma [ lem4.5](i ) with @xmath643 .
we prove theorem [ thm3.7 ] one case at a time .
case a. : : suppose @xmath149
. then by and lemma [ lem4.4](i ) applied to @xmath24 we have @xmath644 it follows therefore from that @xmath645 and hence by lemma [ lem4.4](i ) @xmath646 case a of theorem [ thm3.7 ] follows immediately from and .
the reasoning used to prove cases b , c , and d of theorem [ thm3.7 ] is as follows .
either @xmath23 satisfies @xmath647 or it does nt .
step i. : : if @xmath23 satisfies then we prove below that @xmath23 and @xmath24 satisfy and .
: : if @xmath23 does not satisfy then , for example , to prove theorem [ thm3.7 ] in case b , we prove below that the condition @xmath648 in ( b1 ) does not hold and @xmath23 and @xmath24 satisfy and .
these two steps complete the proof of case b as follows : if the condition @xmath648 in ( b1 ) holds then by step ii , @xmath23 satisfies and hence by step i , @xmath23 and @xmath24 satisfy ( [ eq3.25 ] , [ eq3.26 ] ) . on the other hand , if the condition @xmath171 in ( b2 ) holds then by steps i and ii , @xmath23 and @xmath24 satisfy either ( [ eq3.25 ] , [ eq3.26 ] ) or ( [ eq3.27 ] , [ eq3.28 ] ) .
but since ( [ eq3.25 ] , [ eq3.26 ] ) implies ( [ eq3.27 ] , [ eq3.28 ] ) , we have @xmath23 and @xmath24 satisfy ( [ eq3.27 ] , [ eq3.28 ] ) . similar reasoning will be used in cases c and d. step i. : : suppose @xmath23 satisfies .
then by lemma [ lem4.4](i ) with @xmath649 we see that @xmath23 satisfies as @xmath25 .
hence by , @xmath650 thus by lemma [ lem4.4](i ) applied to @xmath24 we have as @xmath25 that @xmath651 which implies @xmath24 satisfies as @xmath25 .
this completes the proof of step i. step ii . : : suppose @xmath652 by lemma [ lem4.5](ii ) @xmath653 for some @xmath654 .
we now complete the proof theorem [ thm3.7 ] by completing the proof of step ii one case at a time .
case b. : : suppose @xmath152 .
then by lemma [ lem4.5](i ) we have @xmath2 satisfies . hence by @xmath655\lambda+\alpha } \right).\ ] ] by the maximum @xmath656 of the two exponents on @xmath657 in is greater than @xmath203 . thus by lemma [ lem4.4](i ) , @xmath23 satisfies .
this completes the proof of step ii in case b. case c. : : suppose @xmath154 .
then by and lemma [ lem4.4](iii ) we have @xmath2 satisfies . hence by @xmath658 thus by neither ( i ) nor ( ii ) in the statement of case c holds . hence by lemma [ lem4.4](i),(ii ) , @xmath23 satisfies .
this completes the proof of step ii in case c. case d. : : suppose @xmath156 and @xmath93 and @xmath659 are defined by .
then by @xmath660>0.\ ] ] by and lemma [ lem4.4](iii ) we have @xmath661 for some @xmath662 . hence by @xmath663 and @xmath664 by .
thus by lemma [ lem4.4](iii ) @xmath24 satisfies with @xmath665 , that is @xmath666 where @xmath667+\beta=\frac{\alpha+p_0 \lambda}{n } [ ( n-2)\sigma-2]+\beta\\ & = p_0 a+b
. \end{aligned}\ ] ] by the sequence defined by @xmath668 decreases to @xmath669 .
thus after iterating a finite number of times the process of obtaining @xmath670 from @xmath671 and using we obtain as @xmath25 that @xmath24 satisfies for all @xmath157 . hence by @xmath672 for all @xmath157 . by the exponents on @xmath657 in ( 6.9 )
are greater than @xmath203 .
( that is neither ( i ) nor ( ii ) in the statement of case d hold . )
thus , by lemma [ lem4.4](i ) , @xmath23 satisfies for all @xmath157 .
this completes the proof of step ii in case d. since increasing @xmath105 weakens the condition on @xmath6 and since the bounds , do not depend on @xmath105 , we can assume without loss of generality that @xmath673 as in the first paragraph of the proof of theorem [ thm5.1 ] , there exists a constant @xmath599 such that @xmath23 and @xmath24 are @xmath0 positive solutions of @xmath674 in a punctured neighborhood of the origin in @xmath50 . by scaling
we can assume @xmath675 .
we now apply theorem [ thm3.7 ] , case d with @xmath676 .
let @xmath93 and @xmath659 be defined by
. then @xmath677 and @xmath678 .
thus neither ( i ) nor ( ii ) in theorem [ thm3.7 ] , case d , hold .
hence theorem [ thm3.2 ] follows from and .
let @xmath679 then @xmath24 satisfies @xmath623 and . define @xmath332 by @xmath624 .
let @xmath680 be as in lemma [ lem4.1 ] and @xmath681 . by taking a subsequence if necessary , @xmath485 satisfies .
therefore by lemma [ lem4.1 ] there exists a positive function @xmath367 and a positive constant @xmath339 such that @xmath23 satisfies .
thus @xmath23 satisfies @xmath627 in @xmath682 .
also for @xmath610 we have @xmath683 hence @xmath23 satisfies @xmath627 in @xmath684 .
finally , @xmath685 which proves .
it follows from that @xmath686 .
denote the problem by @xmath687 .
if @xmath688 and @xmath689 are constants and @xmath690 solves @xmath687 then clearly @xmath690 solves @xmath691 .
we can therefore assume @xmath692 since the first inequality in holds @xmath693 we see by that @xmath694 or , in other words , @xmath695 let @xmath603 and let @xmath696 , and @xmath339 be as in lemma [ lem4.1 ] .
define @xmath697 as a function of @xmath485 by @xmath698 then @xmath699 by decreasing @xmath485 ( and thereby decreasing @xmath356 ) we can assume @xmath700 @xmath701 by .
it follows from and that @xmath702 which implies @xmath703 and @xmath704 let @xmath82 be a continuous function such that @xmath705 . by lemma [ lem4.1 ] there exist positive functions @xmath706 such that @xmath23 satisfies and @xmath24 satisfies @xmath707 @xmath708 @xmath709 and @xmath710 theorem [ thm3.4 ] follows from ,
, , , @xmath627 , and @xmath623 .
let @xmath1 and @xmath2 be the kelvin transforms of @xmath143 and @xmath144 respectively . then @xmath711 @xmath712 and thus @xmath1 and @xmath2 are @xmath0 nonnegative solutions of the system ( [ eq3.22 ] , [ eq3.23 ] ) in a punctured neighborhood of the origin where @xmath713 using theorem [ thm3.7 ] we get the following results .
case a. : : suppose @xmath149 .
then @xmath714 , @xmath715 and @xmath716\\ & = \frac{n-2}{n}\left[\frac{2(n-2)}{n}\lambda+n+2\right]\geq(n-2)\frac{n+2}{n}>n-2 .
\end{aligned}\ ] ] thus by theorem [ thm3.7](a2 ) we have @xmath717}\right ) \quad\text { as } x\to0\\ & v(x)=o\left ( \left(\frac{1}{|x|}\right)^{(n-2)\left(1+\frac{2}{n}\right ) } \right ) \quad\text { as } x\to0 . \end{aligned}\ ] ] hence case a of theorem [ thm3.6 ] follows from .
case b. : : suppose @xmath152 .
then @xmath718+\alpha=\lambda n+n+2-(n-2)\lambda = n+2 + 2\lambda \end{aligned}\ ] ] and @xmath719 thus by theorem [ thm3.7](b2 ) we have @xmath720 hence case b of theorem [ thm3.6 ] follows from .
case c. : : suppose @xmath154
. then @xmath721 and @xmath722 thus by theorem [ thm3.7](c2 ) we have @xmath723 hence case c of theorem [ thm3.6 ] follows from .
case d. : : suppose @xmath156 .
let @xmath93 and @xmath659 be defined by . then by and direct calculation ( maple is helpful )
, we find + @xmath724>n-2\ ] ] + and @xmath725 by .
thus neither ( i ) nor ( ii ) in theorem [ thm3.7](d1 ) hold . also implies @xmath726.\ ] ] hence by theorem [ thm3.7](d2 ) we have @xmath727 thus case d of theorem [ thm3.6 ] follows from .
we use repeatedly the following special case of a result of brezis and lions @xcite . [
a.1 ] suppose @xmath23 is a @xmath0 nonnegative superharmonic function in @xmath728 , for some @xmath157
. then @xmath729 and for @xmath730 we have @xmath731 where @xmath36 is given by , @xmath732 and @xmath733 are constants , @xmath734 , and @xmath735 is harmonic . h. brezis and p - l . lions , a note on isolated singularities for linear elliptic equations , mathematical analysis and applications , part a , pp . 263-266 , adv . in math . suppl . stud .
, 7a , academic press , new york - london , 1981 .
e. mitidieri and s. pohozaev , a priori estimates and the absence of solutions of nonlinear partial differential equations and inequalities .
( russian ) tr . mat .
inst . steklova 234 ( 2001 ) , 1384 ; translation in proc .
steklov inst . math .
2001 , no . 3 ( 234 ) , 1-362 | we study the behavior near the origin of @xmath0 positive solutions @xmath1 and @xmath2 of the system @xmath3 where @xmath4 are continuous functions .
we provide optimal conditions on @xmath5 and @xmath6 at @xmath7 such that solutions of this system satisfy pointwise bounds near the origin . in dimension
@xmath8 we show that this property holds if @xmath9 or @xmath10 grow at most linearly at infinity . in dimension
@xmath11 and under the assumption @xmath12 , @xmath13 as @xmath14 , ( @xmath15 ) , we obtain a new critical curve that optimally describes the existence of such pointwise bounds .
our approach relies in part on sharp estimates of nonlinear potentials which appear naturally in this context .
_ keywords _ : pointwise bound ; semilinear elliptic system ; isolated singularity ; nonlinear potential estimate . |
the neutrinos were proposed by w. pauli in 1930 @xcite and first detected by c. cowan and f. reines in 1956 @xcite . however the questions on the type ( dirac or majorana ) and values of their masses still remain actual @xcite .
the most - studied and well established phenomena , which are governed by the neutrino masses , are the solar neutrino deficit @xcite and the atmospheric neutrino oscillations @xcite . in this paper
i show that the discussed phenomena can be explained within an economical model with three
democratic " majorana neutrinos , taking into account their coherence length .
if this theory is confirmed , it will provide the answers on the mentioned questions on the neutrino masses .
due to large number of ( anti)neutrino flux calculations and monte carlo methods in the market ( see , e.g. , refs .
[ 21 - 31 ] in ref .
@xcite and refs .
[ 15 - 21 ] in ref .
@xcite ) , and large number of discrepancies among them @xcite we will concentrate on the results , which do not strongly depend on the absolute normalization of the data samples . in the next section
the neutrino mass model is introduced , and the neutrino masses and mixing are derived .
the general formula for the neutrino oscillations and its applications to the atmospheric and solar neutrinos in the considered model are given in section [ sec : oscillations ] .
the predictions of this model for other neutrino experiments , and its further testing are outlined in section [ sec : other_experiments ] , which is followed by a conclusion .
consider the mass term for three left - handed majorana neutrinos @xmath0 where @xmath1 are the flavor indices , @xmath2 denotes charge conjugation , and @xmath3 is a democratic " mass matrix , which is invariant under the permutation group of three elements @xmath4 @xcite .
matrix @xmath5 can be considered , e.g. , as small perturbation . ]
the matrix @xmath6 has the eigenvalues @xmath7 and the corresponding eigenvectors @xmath8 which form the leptonic mixing matrix of tri - bimaximal @xcite type @xmath9 this matrix can be parametrized as @xmath10 where @xmath11 , @xmath12 , @xmath13 , @xmath14 , and we used nonstandard order of multiplication ( compare with ref .
@xcite ) of the two euler rotation matrixes @xmath15 the matrix @xmath16 diagonalizes the mass matrix @xmath6 as @xmath17 and makes the transformation from flavor to mass basis @xmath18 in which the mass term in eq . has a diagonal form @xmath19 now , using the definition @xmath20 with the majorana spinors @xmath21 , eq . can be rewritten as @xmath22 where @xmath23 , and @xmath24 are the sign factors , which can be absorbed by the transformation @xmath25 ( @xmath26 ) . on relation of the sign factors to the neutrino @xmath27 properties and mixing matrix see section 2.3.2 in ref .
@xcite and references therein .
the resulting neutrino mass spectrum @xmath28 has normal ordering and two degenerate values . in the following we assume small violation of this degeneracy , see ref .
@xcite for more details .
we should stress that the mixing matrix in eq .
is naturally formed by the eigenvectors of @xmath6 ( the last column of @xmath16 is the eigenvector , which corresponds to the heavier neutrino mass state ) , and is different from the popular tri - bimaximal and democratic " mixing patterns ( see ref .
@xcite and references therein ) , which fail to explain the solar neutrino data by the incoherence , as we do in section [ solar neutrinos : decoherence ] .
the momentum of a massive neutrino @xmath29 , which has the mass @xmath30 , can be expanded in powers of @xmath31 up to the first order as @xmath32 where @xmath33 is the vector , which depends on the production process , @xmath34 and @xmath35 are the momentum and energy of a neutrino in the massless approximation , respectively .
when the oscillation phase is measurable , i.e. , @xmath36 , the oscillation probability for the neutrino travelled space - time interval @xmath37 at the lowest order in @xmath38 can be written as @xcite @xmath39 e_\text{coh } \ , e_\text{loc}\end{aligned}\ ] ] with the coherence term @xmath40 \end{aligned}\ ] ] and the localization term @xmath41,\end{aligned}\ ] ] where @xmath42 and @xmath43 are the neutrino oscillation and coherence lengths , respectively ; @xmath44 is the neutrino mass splitting , and @xmath45 is the neutrino wave packet spatial size . in the limits @xmath46 and @xmath47 eq . can be rewritten as @xmath48 with @xmath49 . for the neutrino mass spectrum in eq . and mixing in eqs .
and we have @xmath50 where @xmath51 .
the oscillations among all the three flavor neutrinos have same size in vacuum . using the atmospheric neutrino mass splitting @xmath52 ev@xmath53 ( at 99.73% cl ) @xcite ,
we have @xmath54 . the flux of atmospheric neutrinos was measured by the super - kamiokande ( sk ) detector @xcite with high statistics for the neutrino energies in the range from 0.1 gev to @xmath55 gev .
using eq . , we have @xmath56 km . using the estimate @xmath57 @xcite , where @xmath58 mev and @xmath59 m are the mass and mean free path of muon , respectively
, we have @xmath60 m. hence eq .
is perfectly satisfied . using eq .
, one can see that eq .
is also well satisfied .
we notice that the difference between the @xmath61-like and @xmath62-like event distributions in the sk experiment @xcite can be explained by the matter effect on @xmath63 which travel through the earth .
indeed , @xmath64 in eq . and @xmath65 in eq .
can be uniformly written as @xmath66 with @xmath67 , and @xmath68 , where @xmath69 is the neutrino mixing angle in vacuum .
for the neutrinos propagating in the matter with the electron number density @xmath70 this probability will be modified as @xcite @xmath71 with @xmath72 and @xmath73 where the matter factor is given by @xmath74^{\frac{1}{2}},\end{aligned}\ ] ] and the mikheyev - smirnov - wolfenstein resonance density " is given by @xmath75}{e [ \text{mev } ] } \cos2\theta \quad \text{cm}^{-3 } \ , \text{n}_\text{a}\end{aligned}\ ] ] with @xmath76 and @xmath77 being fermi constant and avogadro number , respectively .
in particular , for the atmospheric neutrinos with energies @xmath78 gev we have @xmath79 , and using the mean electron number density in the earth core @xmath80 @xcite , we find the oscillation probabilities @xmath81 which are significantly suppressed with respect to @xmath82 in eq . .
the corresponding suppression in the earth mantle is less significant .
more detailed investigation of the earth s matter effect on the atmospheric neutrinos in the democratic model ( with possible small perturbations ) is in progress @xcite . in the muon neutrino survival probability versus the neutrino energy for the minos baseline @xmath83 km and chosen values of the electron number density @xmath70 . ]
the minos result for the amplitude of the muon neutrino oscillations @xcite , which is close to unity , is explained by the muon neutrino oscillations to both electron neutrino and tau neutrino , where the matter effect plays subdominant role . for the survival probability of the muon neutrino we have @xmath84 with the vacuum oscillation phase @xmath85 l[\text{km } ] } { e[\text{gev}]}.\end{aligned}\ ] ] eq .
can be effectively rewritten as @xmath86 with @xmath87 fig .
[ fig : minos ] shows the dependence of @xmath88 on the muon neutrino energy @xmath89 in the range from 2 gev to 10 gev for the minos baseline @xmath83 km , where the differences among the three lines demonstrate the relatively weak dependence on @xmath70 ( in the earth crust ) .
thousands of @xmath90 solar neutrinos were detected by sudbury neutrino observatory ( sno ) @xcite using the charged - current ( cc ) and neutral current ( nc ) reactions @xmath91 and @xmath92 respectively .
the ratio of the neutrino fluxes measured with cc and nc events is @xcite @xmath93 for the solar neutrinos with the energies @xmath94 mev the oscillations due to @xmath95 proceed in the matter of the sun as in vacuum , since the number density of electrons in the center of the sun is by a factor of 10 smaller than the resonance density " in eq .
the wave packet size of solar neutrinos , which have the energies @xmath96 mev , is @xmath97 m @xcite . using eq .
with @xmath95 , we have @xmath98 km . for the earth - sun distance
@xmath99 km we have @xmath100 ( this relation is satisfied for the neutrino mass splittings down to @xmath101 ev@xmath53 , which sets the lower bound for the violation of the degeneracy in the neutrino mass spectrum in eq . without spoiling the result in eq . .
) in this limit eq . is reduced to the simple incoherent form @xmath102 ( moreover the oscillations should be averaged out already due to the lack of the emitter localization @xcite . ) hence @xmath103 and , using eq . , we have @xmath104 which is in good agreement with the experimental data in eq . .
we remark that the suppression of @xmath63 flux in eq .
is independent of their energy .
this provides an efficient tool for examination of the solar model calculations , since the neutrino fluxes are quite sensitive to their parameter variations .
for example , the predictions of the solar models with late accretion for @xmath105b , @xmath106be and @xmath107 neutrino fluxes differ by 400% , 300% and 20% , respectively , see fig . 12 in ref
. @xcite .
we notice that the explanation of kamland experimental data by the neutrino oscillations with the mass splitting @xmath108 ev@xmath53 @xcite is questionable ev@xmath53 .
] , and at least one independent experiment is needed to verify the kamland result .
using eqs . and , the average mass , determined through the analysis of low energy beta decays , can be written as @xmath109 for @xmath110 ev we have @xmath111 ev , which is far below the present upper limit @xmath112 ev ( at 95% cl ) obtained in the troitsk neutrino mass " experiment @xcite , and below the sensitivity @xmath113 ev ( at 90% cl ) of ongoing katrin experiment @xcite .
however the new approaches such as the mare , echo , and project8 experiments , offer the promise to perform an independent measurement of the neutrino mass in the sub - ev region @xcite .
using eqs . and , the effective majorana mass in @xmath114 decay can be written as @xmath115 hence @xmath114 decay is strongly suppressed and practically unobservable in the discussed model , which can be checked to very good precision in possible future ton - scale experiments , e.g. , genius @xcite , lxe 10ty and hpxe 10ty @xcite . the t2k experiment reported indication of electron neutrino appearance from an accelerator - produced off - axis muon neutrino beam @xcite .
the minos collaboration also reported @xmath116 appearance signal @xcite .
the discussed t2k ( minos ) experiment observed 6 ( 62 ) candidate @xmath63 events against the expected 1.5 ( 50 ) events for the case of no @xmath116 oscillation , which resulted from their analysis , in particular , in the upper bounds @xmath117 at the 90% cl on the amplitude of the oscillation probability written as @xmath118 more data is expected from ongoing and coming experiments such as nova @xcite .
several reactor experiments recently reported indications of electron antineutrino disappearance @xcite with the ratios of observed to expected numbers of @xmath119 about @xmath120 , which yielded @xmath121 .
the discussed values of the amplitude @xmath122 are smaller than the predictions , which can be derived from the theory of democratic neutrino oscillations , introduced in section [ atmospheric neutrinos : oscillations ] .
however these results are strongly model dependent . in particular
, the uncertainties of the neutrino fluxes may have a significant effect , e.g. , reduction of @xmath123 by about 10% results in increase of @xmath122 by factor of 2.5 .
therefore we strongly encourage the neutrino physicists to reanalyze the data in the framework of the presented simple and natural theory .
we hope that this will bring the neutrino theory and experiment to a good agreement without introduction of sterile neutrinos or other exotics .
we introduced a simple model , which may explain the confirmed neutrino data , and resolve the long standing neutrino mass puzzles .
this model is based on the non - diagonal democratic " mass matrix of three majorana neutrinos , and predicts degenerate neutrino mass splittings , which govern the atmospheric neutrino oscillations , and result in the neutrino mass spectrum close to \{0.03 , 0.03 , 0.06}(ev ) .
the difference between @xmath61-like and @xmath62-like event distributions in the super - kamiokande is controlled by the matter effect .
the solar neutrino deficit is explained by their incoherence in the earth .
we outlined the predictions for the direct neutrino mass and the neutrinoless double beta decay experiments , and importance of the accelerator @xmath116 oscillation and the reactor @xmath119 disappearance experiments in verification of the considered model . definitely more detailed and careful analysis of the neutrino data with respect to the considered model
will be useful .
this work was supported in part by the us department of energy under the contract de - sc0007983 .
the author thanks lincoln wolfenstein , anatoly borisov and the prd referee for useful comments , and alexey petrov and gil paz for important questions .
w. pauli , camb .
monogr . part .
cosmol .
* 14 * , 1 ( 2000 ) . c. l. cowan , f. reines , f. b. harrison , h. w. kruse and a. d. mcguire , science * 124 * , 103 ( 1956 ) .
j. beringer _
_ [ particle data group ] , phys .
d * 86 * , 010001 ( 2012 ) .
r. davis , jr .
, d. s. harmer and k. c. hoffman , phys .
lett . * 20 * , 1205 ( 1968 ) .
y. ashie _ et al . _
[ super - kamiokande collaboration ] , phys .
d * 71 * , 112005 ( 2005 ) [ hep - ex/0501064 ] .
p. adamson _ et al .
_ [ minos collaboration ] , arxiv:1304.6335 [ hep - ex ] .
h. paukkunen and c. a. salgado , arxiv:1306.2432 [ hep - ph ] . c. zhang , x. qian and p. vogel , arxiv:1303.0900 [ nucl - ex ] .
g. mention , m. fechner , t. .lasserre , t. .a .
mueller , d. lhuillier , m. cribier and a. letourneau , phys .
d * 83 * , 073006 ( 2011 ) [ arxiv:1101.2755 [ hep - ex ] ] .
g. battistoni , a. ferrari , p. lipari , t. montaruli , p. r. sala and t. rancati , astropart . phys .
* 12 * , 315 ( 2000 ) [ hep - ph/9907408 ] .
h. harari , h. haut and j. weyers , phys .
b * 78 * , 459 ( 1978 ) .
h. fritzsch and z. -z .
xing , phys .
b * 372 * , 265 ( 1996 ) [ hep - ph/9509389 ] .
m. fukugita , m. tanimoto and t. yanagida , phys .
d * 57 * , 4429 ( 1998 ) [ hep - ph/9709388 ]
. m. fujii , k. hamaguchi and t. yanagida , phys .
d * 65 * , 115012 ( 2002 ) [ hep - ph/0202210 ] .
r. n. mohapatra and p. b. pal , world sci . lect
. notes phys .
* 60 * , 1 ( 1998 ) [ world sci .
notes phys .
* 72 * , 1 ( 2004 ) ] .
p. f. harrison and w. g. scott , phys .
b * 557 * , 76 ( 2003 ) [ hep - ph/0302025 ] .
a. nicolaidis , arxiv:1303.6479 [ hep - ph ] .
l. wolfenstein , phys .
d * 18 * , 958 ( 1978 ) .
p. f. harrison , d. h. perkins and w. g. scott , phys .
b * 530 * , 167 ( 2002 ) [ hep - ph/0202074 ] .
p. f. harrison and w. g. scott , phys .
b * 535 * , 163 ( 2002 ) [ hep - ph/0203209 ] .
z. -z .
xing , phys .
b * 533 * , 85 ( 2002 ) [ hep - ph/0204049 ] .
m. doi , t. kotani and e. takasugi , prog .
suppl . * 83 * , 1 ( 1985 ) .
d. zhuridov , arxiv:1405.5522 [ hep - ph ] .
s. k. garg and s. gupta , arxiv:1308.3054 [ hep - ph ] .
m. zralek , acta phys .
b * 29 * , 3925 ( 1998 ) [ hep - ph/9810543 ] . see chapter 8 in c. giunti and c. w. kim , `` fundamentals of neutrino physics and astrophysics , '' oxford , uk : univ .
( 2007 ) 710 p. y. ashie _ et al . _
[ super - kamiokande collaboration ] , phys .
. lett . * 93 * , 101801 ( 2004 ) [ hep - ex/0404034 ] .
k. abe _ et al . _
[ super - kamiokande collaboration ] , phys .
lett . * 97 * , 171801 ( 2006 ) [ hep - ex/0607059 ] .
y. farzan and a. y. smirnov , nucl .
b * 805 * , 356 ( 2008 ) [ arxiv:0803.0495 [ hep - ph ] ] .
l. wolfenstein , phys . rev .
d * 17 * , 2369 ( 1978 ) .
s. p. mikheev and a. y. smirnov , sov .
j. nucl .
phys . * 42 * , 913 ( 1985 ) [ yad .
* 42 * , 1441 ( 1985 ) ] .
v. d. barger , k. whisnant , s. pakvasa and r. j. n. phillips , phys .
d * 22 * , 2718 ( 1980 ) .
a. m. dziewonski and d. l. anderson , phys .
earth planet .
interiors * 25 * , 297 ( 1981 ) .
d. zhuridov , work in progress .
p. adamson _ et al .
_ [ minos collaboration ] , phys . rev .
lett . * 106 * , 181801 ( 2011 ) [ arxiv:1103.0340 [ hep - ex ] ] .
b. aharmim _ et al . _
[ sno collaboration ] , phys .
c * 72 * , 055502 ( 2005 ) [ nucl - ex/0502021 ] .
k. kiers , s. nussinov and n. weiss , phys .
d * 53 * , 537 ( 1996 ) [ hep - ph/9506271 ] .
v. n. gribov and b. pontecorvo , phys .
b * 28 * , 493 ( 1969 ) .
s. nussinov , phys .
b * 63 * , 201 ( 1976 ) .
a. m. serenelli , w. c. haxton and c. pena - garay , astrophys .
j. * 743 * , 24 ( 2011 ) [ arxiv:1104.1639 [ astro-ph.sr ] ] . t. araki _ et al .
_ [ kamland collaboration ] , phys .
* 94 * , 081801 ( 2005 ) [ hep - ex/0406035 ] .
_ et al . _
[ kamland collaboration ] , phys .
d * 83 * , 052002 ( 2011 ) [ arxiv:1009.4771 [ hep - ex ] ] .
a. gando _ et al .
_ [ kamland collaboration ] , arxiv:1303.4667 [ hep - ex ] . v.
m. lobashev _ et al .
_ , nucl .
suppl . *
91 * , 280 ( 2001 ) .
n. a. titov [ katrin collaboration ] , phys . atom .
nucl . * 67 * , 1953 ( 2004 ) .
j. a. formaggio [ katrin collaboration ] , aip conf .
proc . * 1441 * , 426 ( 2012 ) .
g. drexlin , v. hannen , s. mertens and c. weinheimer , adv .
high energy phys . * 2013 * , 293986 ( 2013 ) .
j. hellmig and h. v. klapdor - kleingrothaus , z. phys .
a * 359 * , 351 ( 1997 ) [ nucl - ex/9801004 ] .
j. j. gomez - cadenas , j. martin - albo , j. m. vidal and c. pena - garay , jcap * 1303 * , 043 ( 2013 ) [ arxiv:1301.2901 [ hep - ph ] ] .
z. -z .
xing , chin .
c * 36 * , 101 ( 2012 ) [ arxiv:1106.3244 [ hep - ph ] ] .
k. abe _ et al . _
[ t2k collaboration ] , phys .
lett . * 107 * , 041801 ( 2011 ) [ arxiv:1106.2822 [ hep - ex ] ] .
p. adamson _ et al .
_ [ minos collaboration ] , phys . rev .
lett . * 107 * , 181802 ( 2011 ) [ arxiv:1108.0015 [ hep - ex ] ] .
d. s. ayres _
[ nova collaboration ] , hep - ex/0503053 .
y. abe _ et al . _
[ double - chooz collaboration ] , phys .
lett . * 108 * , 131801 ( 2012 ) [ arxiv:1112.6353 [ hep - ex ] ] . f. p.
an _ et al . _
[ daya - bay collaboration ] , phys .
lett . * 108 * , 171803 ( 2012 ) [ arxiv:1203.1669 [ hep - ex ] ] .
j. k. ahn _
[ reno collaboration ] , phys .
* 108 * , 191802 ( 2012 ) [ arxiv:1204.0626 [ hep - ex ] ] . | it is shown that the well established and confirmed neutrino experimental results , such as atmospheric neutrino oscillations and solar neutrino deficit , can be easily explained for democratic " left - handed majorana neutrinos , taking into account the effect of incoherence . in this model
the absolute values of the neutrino masses can be extracted from the atmospheric neutrino oscillation data , which gives the mass spectrum close to \{0.03 ev , 0.03 ev , 0.06 ev}. the predictions of the considered model and its further testing are discussed . |
all stars with masses initially between @xmath08 and @xmath0150 eventually form a degenerate core that inevitably collapses to form a . much less certain
is its subsequent evolution , the potential formation of a , and the powering of a supernova ( sn ) explosion , sometimes associated with a long - duration burst ( lgrb ) .
the situation is deceptively simple and the outcome rests fundamentally on the solution to an energy problem .
an explosion or a fizzle depends on the efficiency with which the system can extract the prodigious gravitational energy released during collapse .
there are two principal forms of energy at disposal .
the first one is the gravitational - binding energy liberated by the collapsing star and in particular its degenerate core .
the second is its rotational energy ( actually drawn from gravitational energy ) , which is a function of the angular - momentum distribution and budget in the progenitor star . understanding how these two energy sources can be channeled to power relativistic and non - relativistic ejecta in core - collapse sne and
leave behind a neutron star , a fast - rotating pulsar , a magnetar , or a black hole , has been the subject of numerous studies and the source of much debate @xcite . thermal mev neutrinos are abundantly radiated from the optically - thick , dense , and hot , allowing its internal energy to be released on a diffusion timescale . in the neutrino mechanism for core - collapse sn explosions @xcite , the absorption in the infalling mantle of @xmath010% of this neutrino flux may alone lead to the revival of the stalled shock and the ejection of the progenitor envelope with an asymptotic kinetic energy of up to 1b ( @xmath1 ) @xcite .
but this generic mechanism should not , however , be the origin of the larger explosion energies of @xmath010b inferred for a small fraction of core - collapse sne . instead , their scarcity calls for exceptional circumstances , which seem intricately related to fast progenitor - core rotation @xcite .
it is probable that most stars contain some angular momentum at the time of death , either because they did not lose it completely through the combined effects of magnetic torques and stellar - wind mass loss @xcite , or because they gained it from a companion star in a binary system @xcite . as the envelope collapses , the rotational energy increases .
during this process , the inner core ( @xmath2 ) spins up by about three orders of magnitude and remains in solid body rotation , while the outer core develops a differentially - rotating profile @xcite .
the energy associated with rotation can be large ( @xmath3 ) and tapped by instabilities developing at the surface of the , in particular the magneto - rotational instability ( mri ; @xcite ) .
numerical simulations for fast - rotating progenitor stars suggest that the magneto - rotational mechanism of explosion is promising and offers a very attractive explanation for the existence of highly - energetic sne @xcite .
however , this mechanism relies fundamentally on the assumption that the mri can increase the magnetic field to the required values .
an assumption that has not yet been shown numerically in the full core - collapse context , although preliminary investigations in this sector are promising @xcite . hence , combined with the diversity of progenitor - core properties , these two mechanisms alone
, the neutrino and the magneto - rotational mechanism , may explain the diversity of core - collapse sne , potentially encompassing two orders of magnitude in explosion energy , from the low - luminosity sne ii - plateau @xcite to highly energetic sne ic @xcite .
a great puzzle is then to understand the necessary departures from this general core collapse scenario to produce an lgrb in addition to a sn explosion , as spectroscopically confirmed in , to date , six lgrb / sne pairs , ( for a recent compilation , see @xcite ) .
this requires that @xmath00.1 - 1b be injected into a low - mass , baryon - deficient collimated region ( at the origin of ) and that at the same time @xmath010b be injected quasi - isotropically to eject the progenitor envelope ( at the origin of the sn thermal emission observed in the optical ) .
the very low occurrence rate of lgrb / sn per core - collapse sn of @xmath41% @xcite calls for progenitor properties that are rarely encountered in star formation / evolution .
interestingly , an unambiguous diversity emerges among lgrb / sn observations , necessarily translating into a significant range for the inferred properties of the sn ejecta , with proposed masses and kinetic energies possibly varying by a factor of 510 for both @xcite .
unfortunately , a significant uncertainty is associated with such inferences .
for example , @xcite propose an ejecta mass of 11 with a total energy of 2050b for grb980425/sn1998bw , but @xcite reproduce the light curve with an ejecta of @xmath05 and a total energy of 22b .
such differences are not surprising since both spectra and light curves must be modeled simultaneously and with allowance for the complicated non - lte non - thermal and time - dependent effects controlling the radiative transfer .
the exceptionally fast ejecta expansion of hypernovae is expected to strengthen the time - dependent effects seen in standard " core - collapse sne @xcite while the large production of and significant mixing may sizably affect line - profile shapes from which the expansion rate is inferred @xcite .
two lgrb central - engine models are currently favored .
they suggest the key components for a successful lgrb / sn are a compact progenitor with a short light - crossing time of @xmath01s and fast rotation at the time of collapse .
one is the collapsar model @xcite : a fast - rotating progenitor fails to explode in its early post - bounce phase and instead forms a black hole , while the in - falling envelope eventually forms a keplerian disk feeding the hole on an accretion / viscous timescale comparable to that of the lgrb .
it is within this disk that the sn explosion is triggered and the synthesized .
the other model involves a proto - magnetar @xcite in which the lgrb is born after a successful sn explosion ( either by the neutrino or the magneto - rotational mechanism , although the latter seems more likely given the rapid rotation required for the magnetar ) and the ejection of the overlying envelope ( or at least the onset of the ejection of the inner envelope layers that clear the surface ) .
fast rotation in the permits the huge enhancement of the magnetic - field energy and stresses , which strengthen as the cools and contracts , eventually giving rise to relativistic ejecta ( @xmath010s after the onset of collapse , once the neutrino driven wind decays away ) . in both models , rotation is key to control the dynamics in a unique way .
it is also key to allow the simultaneous ejection of baryon - deficient material at relativistic speeds over a small solid angle and the quasi - spherical ejection of the progenitor envelope .
a critical difference between the collapsar model and the proto - magnetar model is that the collapsar has to form a black hole .
being so central to the model , it is legitimate to investigate what conditions black - hole formation in this context @xcite .
surprisingly , little has been done on this problem .
numerous simulations so far have focused on the early collapse phase and the revival of the sn shock , stopping too early to make any statement concerning black - hole formation . in 1d
, several studies have investigated the neutrino signal and progenitor dependence of non - rotating failed sne @xcite . @xcite
have performed 2d simulations of black - hole formation and the subsequent formation of an accretion disk , although with initial conditions that are incommensurate with currently suggested lgrb progenitors . in 3d
, @xcite performed fully general - relativistic simulations of black - hole formation , however using a simplified soft equation of state that favored it .
finally , other simulations have started from a pre - existing and investigated the powering of the jet at the origin of the lgrb @xcite , or the longer - term synthesis of in this unusual context @xcite .
no simulation has ever demonstrated from first principles , and thus convincingly , the validity of the collapsar model , i.e. , that the progenitors proposed for this model would indeed proceed through each and every necessary step : collapse , formation of a , failure of the shock revival , formation of a followed by that of a keplerian disk , and finally the powering of both the lgrb and the sn , including the synthesis of a generally large amount of by core - collapse sn standards .
this is an obvious shortcoming of all theoretical investigations on the collapsar model and its proposed progenitors .
similar gaps in the modeling of proto - magnetar driven lgrbs exist : the early magneto - rotational core - collapse sn evolution has been modeled in 2d @xcite and so has the phase in which relativistic outflows are driven @xcite , but the evolution connecting the two phases has not been modeled .
the robustness of the magneto - rotational explosion mechanism largely rests on the efficiency of angular - momentum transport , and in particular the extraction of the free - energy stored ( and replenished through accretion ) in differential rotation at the surface of the proto - neutron star .
the failure to extract this energy on short time scales may , however , facilitate black - hole formation , although it may also compromise energy extraction in the collapsar model .
these complicated issues require detailed modeling to build upon the promising results of @xcite and @xcite . to this day
, the collapsar model has been studied more extensively than the proto - magnetar model for lgrbs , so the later may look more promising in some ways in part because of the lesser scrutiny it has received . in this paper , we focus on one important aspect of the collapsar model to validate , or invalidate , the assumption , often made but so far never checked , that the lgrb progenitor models available in the literature indeed collapse to form a .
we do this by performing hydrodynamical simulations of the lgrb progenitor models of @xcite using the code ` gr1d ` @xcite .
this issue is critical for testing the potential of progenitor stars for producing lgrbs via the collapsar mechanism , but may also serve to diagnose an attractive channel for the formation of proto - magnetars .
such `` failed '' collapsars ( because they explode before forming a ) represent a serious alternative for the production of lgrbs , although they have their own caveats @xcite .
admittedly , the phenomenon of core collapse , bounce , and the events that follow are fundamentally multi - d .
we believe , however , that much can be learned from 1d simulations of the kind presented here .
for example , the mass - accretion rate onto the is largely determined by the angular - averaged density profile of the progenitor star , which we capture accurately .
our 1d exploration reveals the landscape of core properties at bounce and quantifies fundamental differences between progenitors . in the next section ,
we start by reviewing results from stellar - evolution models for lgrb progenitors .
we then describe our methodology for the ` gr1d ` simulations of lgrb progenitors available in the literature .
we make a short digression in section [ sect : rot ] to discuss the rotational properties of the collapsed cores of massive stars . in section [ sect_results ]
, we present our results before concluding in section [ sect_conc ] .
stellar - evolution calculations have been performed to investigate the mass , rotation , and metallicity requirements for producing fast - rotating pre - collapse stars .
both single- @xcite and binary - star @xcite evolutionary scenarios have been investigated .
fast rotation of the proto - stellar core is clearly critical to procure a large angular momentum to the star initially .
if the rotation rate attained is sufficiently large , the star may even evolve chemically homogeneously and avoid a supergiant phase , which is known to sap the core of its rotation through the effects of magnetic torques .
such fast - rotating chemically - homogeneous stars also naturally die as h - deficient he - poor wolf - rayet ( wr ) stars .
low metallicity quenches the stellar - wind mass loss rate , a condition that may be more important for a single - star scenario than for the binary - star scenario @xcite .
while the general outcome of these simulations is that it is possible to produce massive stars with a rapidly spinning core / envelope at death , it is difficult to compare the final properties of published models .
indeed , models are rarely evolved all the way to an iron core .
the treatments of mixing , mass loss , and angular - momentum loss / transport differ .
magnetic - fields may or may not be included and when they are the prescription may differ @xcite . furthermore , all these studies remain speculative about the outcome of collapse for such progenitors .
they argue for black hole formation and the formation of a disk based on order of magnitude estimates , rather than detailed numerical simulations . for a start , of all the lgrb progenitor simulations ,
only those of @xcite are evolved all the way to the onset of collapse . in simulations
halted well before , the iron - core mass is estimated from the co - core mass @xcite or is simply not considered in the discussion @xcite .
most studies consider a model viable for producing a collapsar based exclusively on the angular - momentum budget of the inner 3 , and perform no checks on the likelihood of forming a : its formation is deemed so obvious that the discussion of any alternate scenario is generally omitted .
differing in their criteria and approaches for selecting collapsar candidates , some studies may yield progenitor - mass ranges that do not even overlap : using an angular - momentum criterion , @xcite favor progenitor stars with a main - sequence mass below @xmath030 .
in contrast , @xcite , arguing for the need for both a large angular momentum , a large iron core , and a wo stellar type at death , favor progenitors with a main - sequence masses above @xmath035(magnetic fields are not treated in this work , though ) .
recent studies suggest that selecting collapsar progenitors based exclusively on a large angular - momentum budget may be too simplistic .
for example , the magneto - rotational explosion invoked to explain hypernovae derives its energy from this same large core angular momentum ( via the mri and the strongly differentially - rotating layers in the post - shock region ) .
this mechanism does not obviously accommodate the formation of a , as demonstrated by @xcite .
they simulated the collapse of the core and the development of a magneto - rotational explosion in model 35oc of @xcite and found that rotational energy of order 10b is readily available to launch a sn ejecta on a timescale of a few 100ms .
they furthermore found that accretion is easily shut off by the developing explosion and that the mass fails to grow to the instability threshold for black - hole formation .
however , the simulations of @xcite did not resolve the mri but instead used an equipartition ansatz to estimate the magnitude of the mri - amplified magnetic fields .
in reality , magnetic - field reconnection may , for example , compromise the dynamical potential of magnetic stresses , channeling magnetic energy into heat to be radiated away by neutrinos .
this and other alternatives have been studied by @xcite under the general form of viscous dissipation .
they find that the extra energy deposition can , in some cases , considerably alter the post - bounce dynamics and generate a vigorous explosion .
@xcite studied black - hole formation based on a variety of progenitor models characterized by different main - sequence mass , metallicity , and rotation rate .
they find that the outcome of collapse can be anticipated from the bounce compactness of the progenitor core structure and in particular that of the region inside 2.5 , which corresponds approximately to the maximum mass that a can have and remain in hydrostatic equilibrium .
they find that higher mass progenitors published in the literature do not always have larger iron cores and therefore that they are not necessarily more prone to black - hole formation .
they also reveal considerable diversity in progenitor core structure , even for the same main - sequence mass .
some stellar evolution studies obtain a monotonic increase of the iron - core mass ( or bounce compactness ; see fig .
9 of @xcite ) versus main - sequence mass ( @xcite ; see also @xcite ) , while some show an anti - correlation beyond @xmath040 @xcite .
the primary reason for this are differing prescriptions for rate and time of mass loss , one of the major uncertainties in massive star evolution ( see also the discussions in @xcite and @xcite ) .
l@cc|ccc|cccc|ccc|c 12sa & 0.003 & 0.000 & @xmath5 & ( 1.56 ) & ( 1.47 ) & ( 10.9 ) & @xmath5 & @xmath5 & @xmath5 & 0.167 & @xmath5 & @xmath5 & @xmath5 + 12sg & 0.239 & 0.198 & 2.728 & 2.33 & 2.12 & ( 7.57 ) & @xmath5 & @xmath5 & @xmath5 & 0.124 & 18.1 & 3.37 & 0.034 + 12sh & 0.141 & 0.144 & @xmath5 & ( 2.08 ) & ( 1.91 ) & ( 5.43 ) & @xmath5 & @xmath5 & @xmath5 & 0.148 & 25.5 & 3.35 & 0.008 + 12si & 0.075 & 0.208 & @xmath5 & ( 1.75 ) & ( 1.64 ) & ( 6.95 ) & @xmath5 & @xmath5 & @xmath5 & 0.153 & 24.7 & 3.17 & 0.010 + 12sj & 0.121 & 0.751 & @xmath5 & ( 2.05 ) & ( 1.90 ) & 6.77 & 0.470 & 91.6 & 2.27 & 0.166 & 4.10 & 3.53 & 0.305 + 12oa & 0.011 & 0.000 & @xmath5 & ( 1.52 ) & ( 1.44 ) & ( 11.9 ) & @xmath5 & @xmath5 & @xmath5 & 0.104 & @xmath5 & @xmath5 & @xmath5 + 12og & 0.029 & 0.149 & @xmath5 & ( 1.88 ) & ( 1.75 ) & 4.50 & 0.217 & 4.42@xmath6 & 7.32 & 0.179 & 22.5 & 3.17 & 0.006 + 12oh & 0.090 & 0.285 & @xmath5 & ( 1.84 ) & ( 1.71 ) & 7.62 & 0.210 & 710 . &
0.07 & 0.135 & 17.1 & 3.30 & 0.020 + 12oi & 0.095 & 1.061 & @xmath5 & ( 1.86 ) & ( 1.73 ) & 5.91 & 0.535 & 68.8 & 3.81 & 0.181 & 4.49 & 3.27 & 0.270 + 12ol & 0.076 & 0.299 & @xmath5 & ( 1.75 ) & ( 1.64 ) & 6.99 & 0.259 & 554 . & 0.36 & 0.136 & 19.7 & 3.61 & 0.017 + 12on & 0.170 & 1.709 & @xmath5 & ( 2.21 ) & ( 2.02 ) & 2.67 & 0.496 & 8.51 & 8.26 & 0.145 & 2.38 & 3.27 & 1.013 + 12ta & 0.008 & 0.000 & @xmath5 & ( 1.59 ) & ( 1.49 ) & ( 12.0 ) & @xmath5 & @xmath5 & @xmath5 & 0.117 & @xmath5 & @xmath5 & @xmath5 + 12tg & 0.034 & 0.148 & @xmath5 & ( 1.91 ) & ( 1.77 ) & 4.59 & 0.228 & 3.73@xmath6 & 7.35 & 0.182 & 24.9 & 3.43 & 0.007 + 12th & 0.107 & 1.042 & @xmath5 & ( 1.93 ) & ( 1.79 ) & 6.67 & 0.495 & 85.2 & 2.56 & 0.138 & 5.06 & 3.59 & 0.313 + 12ti & 0.145 & 1.323 & @xmath5 & ( 2.02 ) & ( 1.86 ) & 3.33 & 0.507 & 15.3 & 7.46 & 0.144 & 3.26 & 3.48 & 0.610 + 12tj & 0.517 & 1.281 & 0.853 & 2.51 & 2.37 & 2.97 & 0.640 & 2.64 & 8.58 & 0.191 & 1.10 & 4.31 & 3.286 + 16sa & 0.101 & 0.000 & @xmath5 & ( 1.88 ) & ( 1.74 ) & ( 14.6 ) & @xmath5 & @xmath5 & @xmath5 & 0.138 & @xmath5 & @xmath5 & @xmath5 + 16sg & 0.109 & 0.203 & @xmath5 & ( 1.91 ) & ( 1.77 ) & 9.79 & 0.404 & 2.71@xmath7 & 2.16 & 0.141 & 20.8 & 3.27 & 0.010 + 16sh & 0.081 & 0.341 & @xmath5 & ( 1.76 ) & ( 1.64 ) & ( 7.70 ) & @xmath5 & @xmath5 & @xmath5 & 0.182 & 16.8 & 2.68 & 0.019 + 16si & 0.380 & 0.189 & 1.132 & 2.38 & 2.20 & ( 9.85 ) & @xmath5 & @xmath5 & @xmath5 & 0.158 & 11.9 & 3.82 & 0.062 + 16sl & 0.075 & 0.207 & @xmath5 & ( 1.73 ) & ( 1.62 ) & ( 6.30 ) & @xmath5 & @xmath5 & @xmath5 & 0.130 & 28.4 & 3.31 & 0.007 + 16sm & 0.121 & 0.229 & @xmath5 & ( 2.02 ) & ( 1.85 ) & ( 8.31 ) & @xmath5 & @xmath5 & @xmath5 & 0.145 & 20.0 & 3.51 & 0.016 + 16sn & 0.496 & 0.455 & 0.777 & 2.42 & 2.27 & 9.45 & 0.508 & 40.6 & 1.77 & 0.187 & 3.25 & 3.69 & 0.451 + 16oa & 0.144 & 0.000 & @xmath5 & ( 2.15 ) & ( 1.96 ) & ( 15.8 ) & @xmath5 & @xmath5 & @xmath5 & 0.133 & @xmath5 & @xmath5 & @xmath5 + 16og & 0.193 & 0.176 & 3.437 & 2.33 & 2.11 & 7.16 & 0.230 & 4.37@xmath6 & 8.49 & 0.168 & 17.6 & 3.82 & 0.018 + 16oh & 0.185 & 0.248 & @xmath5 & ( 2.21 ) & ( 2.00 ) & 9.18 & 0.133 & 810 . & 7.9@xmath8 & 0.150 & 20.5 & 3.59 & 0.023 + 16oi & 0.344 & 0.733 & 1.449 & 2.37 & 2.19 & 7.10 & 0.553 & 18.2 & 5.11 & 0.152 & 3.21 & 4.06 & 0.700 + 16ol & 0.124 & 0.316 & @xmath5 & ( 2.02 ) & ( 1.86 ) & 8.63 & 0.200 & 593 . &
0.05 & 0.138 & 14.1 & 3.41 & 0.030 + 16om & 0.172 & 1.059 & @xmath5 & ( 2.17 ) & ( 1.98 ) & 5.64 & 0.590 & 24.5 & 6.31 & 0.177 & 2.60 & 3.02 & 0.480 + 16on & 0.357 & 1.382 & 1.458 & 2.40 & 2.22 & 3.36 & 0.582 & 4.70 & 10.8 & 0.162 & 1.40 & 3.59 & 2.082 + 16ta & 0.070 & 0.000 & @xmath5 & ( 1.76 ) & ( 1.64 ) & ( 16.0 ) & @xmath5 & @xmath5 & @xmath5 & 0.148 & @xmath5 & @xmath5 & @xmath5 + 16tg & 0.288 & 0.242 & 1.738 & 2.35 & 2.16 & 13.3 & 0.366 & 3.44@xmath9 & 2.41 & 0.174 & 10.4 & 4.02 & 0.083 + 16th & 0.434 & 0.598 & 0.958 & 2.41 & 2.25 & 8.01 & 0.511 & 23.7 & 3.44 & 0.151 & 2.57 & 3.80 & 0.599 + 16ti & 0.242 & 1.367 & 2.791 & 2.41 & 2.21 & 3.51 & 0.554 & 10.6 & 10.4 & 0.150 & 2.17 & 3.77 & 1.341 + 35oa & 0.178 & 0.289 & @xmath5 & ( 2.26 ) & ( 2.05 ) & ( 12.9 ) & @xmath5 & @xmath5 & @xmath5 & 0.153 & 14.6 & 3.53 & 0.044 + 35ob & 0.537 & 1.545 & 0.776 & 2.42 & 2.25 & 16.4 & 0.545 & 31.5 & 4.80 & 0.198 & 1.44 & 3.48 & 3.617 + 35oc & 0.458 & 1.980 & 0.972 & 2.43 & 2.29 & 4.44 & 0.622 & 4.84 & 23.6 & 0.162 & 1.10 & 4.49 & 7.521 + he16c & 0.137 & 0.133 & @xmath5 & ( 2.06 ) & ( 1.90 ) & ( 5.15 ) & @xmath5 & @xmath5 & @xmath5 & 0.134 & 25.5 & 3.11 & 0.007 + he16d & 0.283 & 0.440 & 1.706 & 2.35 & 2.16 & 8.65 & 0.367 & 117 .
& 0.88 & 0.171 & 4.74 & 3.65 & 0.206 + he16e & 0.129 & 1.428 & @xmath5 & ( 2.00 ) & ( 1.85 ) & 6.82 & 0.594 & 43.1 & 6.05 & 0.153 & 3.77 & 3.27 & 0.447 + he16f & 0.496 & 1.096 & 0.837 & 2.48 & 2.33 & 4.29 & 0.567 & 8.97 & 10.5 & 0.202 & 1.46 & 4.28 & 2.323 + he16h & 0.610 & 1.196 & 0.641 & 2.52 & 2.38 & 4.12 & 0.597 & 5.95 & 11.6 & 0.204 & 1.26 & 4.23 & 3.335 + he16k & 0.132 & 0.134 & @xmath5 & ( 2.04 ) & ( 1.88 ) & ( 5.16 ) & @xmath5 & @xmath5 & @xmath5 & 0.140 & 26.6 & 3.11 & 0.007 + he16l & 0.316 & 0.315 & 1.497 & 2.36 & 2.17 & 9.34 & 0.286 & 195 . & 0.24 & 0.165 & 6.73 & 3.73 & 0.116 + he16
m & 0.111 & 1.206 & @xmath5 & ( 1.91 ) & ( 1.77 ) & 10.4 & 0.532 & 94.0 & 2.61 & 0.134 & 4.42 & 3.23 & 0.348 + he16n & 0.198 & 1.203 & 3.424 & 2.36 & 2.15 & 7.81 & 0.604 & 35.9 & 7.15 & 0.154 & 2.59 & 3.84 & 0.970 + he16o & 0.298 & 1.209 & 1.891 & 2.39 & 2.20 & 6.77 & 0.594 & 25.0 & 8.86 & 0.127 & 1.98 & 3.61 & 1.414 + he16p & 0.573 & 1.038 & 0.672 & 2.49 & 2.35 & 6.42 & 0.584 & 14.2 & 9.46 & 0.235 & 1.58 & 4.16 & 2.523 [ tab : results ]
in this work , we use the open - source , spherically symmetric , general relativistic , eulerian hydrodynamics code ` gr1d ` @xcite .
rotation is included through a centrifugal - acceleration term in the momentum equation this is the most important dynamical feature of rotation relevant to core collapse . however , ` gr1d ` can not account for the associated deviations from spherical symmetry nor any angular - momentum redistribution .
we select the equation of state ( eos ) from @xcite characterized by a nuclear incompressibility of @xmath10mev ( hereafter referred to as the ls220 eos ) .
this eos provides the best match to both mass and mass - radius constraints from observations and nuclear theory @xcite . `
gr1d ` uses an efficient neutrino leakage / heating scheme that qualitatively reproduces the salient features of neutrino transport .
we refer the reader to @xcite for additional details on ` gr1d ` and our methodology . as described above , the only stellar - evolutionary models for lgrb progenitors that are evolved until the onset of collapse are those proposed by @xcite .
we thus focus on their model dataset for our investigation on the dynamics of the core - collapse sn engine and the potential formation of a black hole in the collapsar context . using ` kepler `
, @xcite investigated a rather narrow range of progenitor masses , but varied the initial rotation rate ( solid - body rotation is assumed initially ) and environmental metallicity from solar to 1% solar ( with an additional tunable factor as low as 0.1 for the metallicity - dependent mass loss rate , equivalent to a reduction in metallicity by a factor of 100 in their mass - loss prescription ) . arguing that the inferred mass of lgrb / sn ejecta known in 2006 is on the order of 10 , and
since higher - mass stars may lose too much angular momentum through stellar winds ( even at low metallicity ) , they focused primarily on lower - mass progenitors , with main - sequence masses of 12 and 16 , with the exception of one 35 model set .
we adopt the same nomenclature as for their 12- , 16- and 35-@xmath11 models .
it comprises the model s main - sequence mass , followed by a letter denoting the environmental metallicity ( ` s ' for solar , ` o ' for 10% solar , and ` t ' for 1% solar ) .
an additional letter is appended to individualize the models done with different wr mass - loss rate prescriptions , allowing or not for magnetic effects , and the total angular momentum of the star .
16-@xmath11 helium models are denoted by ` he16 ' followed by an individualizing capital letter . in this work ,
we simulate the collapse and post - bounce evolution with ` gr1d ` for all these progenitor models , with a primary focus on determining their ability to produce the key features of the collapsar model : a together with a keplerian disk . as we discuss in the following section ,
black - hole formation is not obviously guaranteed in any of these dying stars .
since lgrbs seem fundamentally related to rapid rotation , it is useful to summarize a few facts and concepts related to the gravitational collapse of rotating iron cores in massive stars .
first , it is reasonable to assume ( which is borne out by simulations , e.g. , @xcite ) that the iron core , in its pre - collapse state , will be approximately uniformly rotating .
such a solid - body rotation corresponds to the lowest energy state at fixed total angular momentum and will be assumed on a secular timescale by any rotating fluid that has some means to redistribute angular momentum . versus radius @xmath12 at both the pre - sn stage ( dashed lines ) and at core bounce ( solid lines ) for selected models of @xcite .
the inner homologously collapsing core maintains its initial uniform rotation throughout collapse .
[ fig : omega ] ] rotating core collapse , even for the high pre - collapse rotation rates of some of the potential lgrb progenitors that we consider in this study , proceeds qualitatively in a very similar fashion to non - rotating collapse as long as the ratio of the centrifugal acceleration @xmath13 to the gravitational acceleration @xmath14 , is small , @xmath15 due to angular - momentum conservation , the angular velocity behaves as @xmath16 .
@xmath17 stays constant for a collapsing mass shell , and , thus , the above ratio increases during collapse as @xmath18 and may potentially become large for small radii . in the case of @xmath19
, the collapsing rotating iron core will behave like a non - rotating core and separate into a subsonically collapsing inner core ( @xmath20 ) and a supersonically collapsing outer core ( @xmath21 ) .
the inner core exhibits a self - similar ( homologous ) velocity profile , @xmath22 , until core bounce and shock formation @xcite .
after core bounce , the inner core material forms the core of the and outer core material accumulates at its edge . the mass of the inner core at bounce is typically @xmath0@xmath23 for non - rotating cores , set by the eos of relativistic electrons , the trapped lepton number , core entropy , and gravity @xcite .
it increases monotonically , though slowly , with increasing pre - collapse rotation rate . for cores that reach @xmath24 ,
the mass of the inner core will be increased to @xmath25 @xcite .
since the inner core is collapsing homologously , we can introduce a homology parameter @xmath26 , so that @xmath27 where @xmath28 is the radius of a collapsing fluid element at time @xmath29 and @xmath30 is its initial radius .
this must hold for any mass shell within the inner core .
two collapsing fluid elements located initially at @xmath31 and @xmath32 , and rotating with a frequency @xmath33 and @xmath34 conserve angular momentum , @xmath35 .
homology implies @xmath36 , and @xmath37 , therefore @xmath38 since this property holds for any mass shell within the inner core , the rotational profile must be preserved under homologous collapse . in practice , because the pre - collapse cores of massive stars are always in solid - body rotation , so are the inner cores of the at bounce . outside of the homologously collapsing core the self similar relation of eq .
[ eq : selfsimilar ] does not hold .
most generally , gradients in the rotation rate develop due to the underlying density gradients in the hot postshock region and in the supersonically infalling region ahead of the accretion shock . based on these arguments
, we expect an early post - bounce rotational profile that is approximately uniform within the inner @xmath39 ( out to @xmath0@xmath40 in radius ) and strongly differential at larger radii .
this is confirmed by fig .
[ fig : omega ] which shows @xmath41 at bounce as obtained with ` gr1d ` for a variety of models considered in this study .
we also show the rotational profile at the onset of collapse .
this result is not entirely new , but has previously been pointed out by @xcite in the context of 1d and 2d rotating core collapse simulations .
since uniform rotation is the lowest energy state , the _ shear energy _ of differential rotation is to be interpreted as a free energy that will be tapped by any process ( e. g. , nonaxisymmetric rotational shear instabilities , viscosity , or the mri ) capable of redistributing angular momentum .
viscosity would lead to additional heating in the postshock region to enhance the neutrino mechanism @xcite while the mri action could strengthen the magnetic fields , driving bipolar outflows in the magneto - rotational mechanism @xcite . in our simulations
, we estimate the available free energy of differential rotation by computing the difference in rotational energy of the model in ` gr1d ` and the rotational energy of a uniformly spinning of the same angular momentum and moment of inertia , @xmath42 where from @xcite , for ` gr1d ` , @xmath43 and @xmath44 is the rotational energy , @xmath45 is the moment of inertia , and @xmath46 is the uniform rotation frequency , @xmath47 is the specific enthalpy , @xmath48 is the @xmath49 component of the metric , @xmath50 is the lorentz factor , and @xmath51 is the angular velocity . we take @xmath52 to be the radius where the matter density , @xmath53g@xmath54 .
we have performed simulations for the entire set of ` kepler ` models published in @xcite ] .
we first consider the rapidly spinning progenitors evolved without magnetic fields .
all these models have a dimensionless kerr spin ( @xmath55 ) at 3 greater than unity ( with the exception of model he16j , which has @xmath56 ) and are thus considered as promising collapsar candidates by @xcite .
unfortunately , when evolved with ` gr1d ` , the collapsing iron core of all such models halts its collapse and expands these models do not experience core bounce within a few seconds of evolution in ` gr1d ` .
we associate this problem with the neglect of the centrifugal acceleration in the momentum equation in ` kepler ` , an approximation that fails in the fastest rotating models .
this term is included in ` gr1d ` .
the mismatch suggests that their fastest models may be significantly affected by the addition of this term .
even if they did collapse , it is not clear that such extremely fast rotating cores would avoid a centrifugal bounce . in the remainder of this paper
, we thus limit our discussion to models evolved with magnetic fields and therefore subject to magnetic torques during their evolution . of the 46 models that fulfill this criterion
, we identify 4 additional models ( 12om , 16tj , 35od , and he16 g ) that do not collapse but instead expand when restarted with ` gr1d ` .
we exclude these as well from our study .
finally , for reference and completeness , we include the non - rotating models associated with each series ( 12sa , 12oa , 12ta , 16sa , 16oa , 16ta ) , making a total of 48 models .
each simulation is continued after core bounce until a forms or until a time of @xmath57s has passed , whichever comes first .
we present the results for these 48 models evolved with magnetic fields in table [ tab : results ] ( for the table layout , we group models in bundles first of increasing mass , then of decreasing metallicity , and finally in alphabetical order which generally corresponds to an increased initial rotation rate ) .
as advocated by @xcite , the bounce compactness is a robust quantity for diagnosing the propensity to black - hole formation , which is largely determined by the spatial extent encompassed by the 2.5 lagrangian mass coordinate at core bounce in the progenitor core . to avoid introducing biases associated with the non - uniform conditions in the progenitor simulations ( ` kepler ` models are not all evolved to the same central density on their collapse trajectory ) this compactness is unambiguously evaluated at the time of bounce .
the formal definition of this core compactness is @xmath58 where we take @xmath59 .
@xmath60 is the radial coordinate that encloses 2.5@xmath11 of baryonic material at the time of core bounce @xcite .
our simulations first demonstrate that most of the models have a small core compactness @xmath61 .
@xcite argues that a compactness of 0.45 represents a threshold value , for the neutrino mechanism , since above it an unrealistic neutrino - heating efficiency is required to prevent black - hole formation .
we further confirm this by determining the critical heating efficiency for the models in table [ tab : results ] via the same procedure as in @xcite , to which we refer the reader for full details .
we note that this criterion is for explosions via the neutrino mechanism and therefore neglects any magneto - rotational contribution to the powering of an explosion @xcite , so this threshold value is probably a lower limit .
we also stress that this criterion is based on spherically symmetric simulations with an efficient , but crude , approximation to neutrino transport .
therefore we suggest caution when attempting to interpret the outcome of a model based solely on the bounce compactness , since the adopted threshold value for bh formation is a semi - quantitative estimate based on where the slope of the required heating efficiency begins to clearly increase ( see the discussion in @xcite ) . when plotting the critical heating efficiency of the non - rotating solar metallicity stars from @xcite determined in @xcite , together with the generally fast - rotating progenitors of @xcite , we find that both datasets in fact overlap for the most part ( fig .
[ fig : eta_xi ] ) . in other words , in terms of compactness , most of these progenitors are similar to garden variety , low - mass , non - rotating , progenitors and do not seem to have any more reason to form a than , e.g. , the rsg progenitors expected to produce sne ii - plateau . as shown in table [ tab : results ] , provided no explosion is launched , half of these models have not formed a after 3.5s , and only @xmath015% do within @xmath01s .
we note that this result is not so surprising given the small iron - core mass ( @xmath62 ) of most @xcite models ( see their tables 1 and 2 ) .
the first conclusion from our work is therefore that most of the models presented here are rather unlikely to form a black hole and thus may fail in a very fundamental way to produce a collapsar , irrespective of their angular - momentum budget . versus bounce compactness @xmath61 for our ` gr1d ` simulations of the @xcite models , whose properties
are summarized in table 1 ( blue diamonds ) . for comparison
, we overplot the same quantity for the standard non - rotating core - collapse sn progenitor models of @xcite evolved at solar metallicity ( red squares ) . for the most part , the two distributions overlap , suggesting that the propensity to black - hole formation and explosion is comparable for both .
only models with the fastest rotation rates achieve a larger compactness in excess of 0.40.5 , but these may then be diverted from black - hole formation through an early magneto - rotational explosion .
[ fig : eta_xi ] ] within each sequence presented in table [ tab : results ] , the models that form a within 3.5s of core bounce , and thus at least in principle susceptible to form a collapsar , are the faster rotating ones characterized by very weak mass - loss rates .
these properties conspire to produce larger co cores , more typical of more massive stars that do not rotate . in the following discussion , we group these models into several categories .
the first category are models which obviously do not give rise to a lgrb , either by the standard type - i collapsar or the proto - magnetar mechanism , because they contain too little angular momentum . optimistically assuming a failed core - collapse sn , which is unlikely given the modest values of @xmath61 , models 12sg ( @xmath63 ) , 16og ( @xmath64 ) , 16si ( @xmath65 ) , and 16tg ( @xmath66 ) possess too little angular momentum in the remainder of the star to form a disk about the central black hole within @xmath67s of collapse .
this behavior is reflected by the stellar type at the time of death , i.e. a bsg star for model 12sg and a rsg star for models 16og and 16tg , only 16si is a wr star at the time of death .
quantitatively , this can be further inspected in table [ tab : results ] where we include the disk formation time , the black hole mass and spin at that time , and the mass exterior to the disk .
we define disk formation to be when the accreting material will first be supported at the about a black hole with the enclosed mass and angular momentum using the formulae of @xcite .
we estimate the disk formation time as twice the free fall time of the innermost mass element that reaches a keplerian velocity @xcite . @xmath68
^ 3}{8gm_\text{disk}}}\ ] ] where @xmath69 is the radius of the disk - forming lagrangian mass element in the pre - sn model .
if no such mass element exists , no disk will form . in this case
we include , instead of the enclosed black - hole mass , the total pre - sn stellar mass in parentheses . in the four models mentioned above , either no disk forms or the disk formation time is @xmath70s .
additionally , we can discuss the potential for these models to form a lgrb via the proto - magnetar model .
using eqs .
[ eq : freeenergy ] - [ eq : omegabar ] , we calculate the free energy available in differential rotation at 100ms after bounce .
we also calculate a reference spin period ( @xmath71 ) , measured at the onset of the neutrino driven explosion , by assuming solid - body rotation for the entire with the same total angular momentum and moment of inertia .
the corresponding values are given in table [ tab : results ] . in fig .
[ fig : freeenergy_period ] , we show the free energy available in differential rotation at 100ms and the spin period of the at the onset of explosion .
the total rotational energy of the , estimated as @xmath72 , will increase as the cools and contracts . in models 12sg
, 16og , 16si , and 16tg , which are contained within the green ( lightest shade ) box of fig .
[ fig : freeenergy_period ] , @xmath730.1b of free energy could be extracted from differential rotation via the mri and converted to explosion energy , much less than is needed for a magneto - rotational explosion .
also , the spin periods are @xmath7410ms , to @xmath75 @xcite , or roughly a factor of 2 , the spin periods are @xmath76ms . ] significantly larger than the @xmath77ms periods required for the proto - magnetar model to reproduce classical lgrb energies @xcite . , taken at the onset of explosion ( left axis , blue dots ; eq .
[ eq : omegabar ] ) and the free energy stored in differential rotation 100ms after bounce @xmath78 ( right axis , red stars ; eq . [ eq : freeenergy ] ) versus bounce compactness @xmath61 for all rotating models in table [ tab : results ] .
while models with a low bounce compactness show a diversity in core - rotation properties , those with a high bounce compactness systematically have short spin periods and a large budget of free energy stored in the differential rotation .
shaded boxes refer to specific groupings of models discussed in the text .
using @xmath79 as a black - hole formation criterion for non - rotating progenitors , we can qualitatively compare the reference spin periods of this figure to @xcite , who sketches the outcome of collapse as a function of progenitor spin and mass . from this
, one would predict that none of the lgrb progenitor models studied here formed black holes . ]
the second category are models with a larger angular - momentum budget but unfavorable bounce compactness .
although compact enough to lead to black - hole formation within 3.5s of core bounce , we find that the predicted critical heating efficiencies are similar to that expected for a standard 15 non - rotating rsg progenitor star @xcite .
these properties make them unlikely collapsar progenitors , but in contrast , make them ideal candidates for proto - magnetar formation , and perhaps lgrbs through that channel .
these models include 16oi ( @xmath80 ) , 16on ( @xmath81 ) , 16th ( @xmath82 ) , 16ti ( @xmath83 ) , he16d ( @xmath84 ) , he16l ( @xmath85 ) , he16n ( @xmath86 ) , and he16o ( @xmath87 ) and are contained in the orange ( medium shade ) box of fig .
[ fig : freeenergy_period ] .
in addition to having critical heating efficiencies similar to what is needed to explode typical low - mass massive stars , the free energy available in rotation is @xmath88(1b ) .
this energy may be converted to explosion energy via the magneto - rotational mechanism .
the spin period of these s is in the range 1 - 6ms , thus on the order of what is needed for the proto - magnetar model of lgrbs @xcite .
eventually , the fastest rotating progenitor models evolved with a strongly inhibited stellar - wind mass loss represent more suitable collapsar candidates , although each model has caveats .
this set is contained in the purple ( darkest shade ) box of fig .
[ fig : freeenergy_period ] and includes models 12tj ( @xmath89 ) , 16sn ( @xmath90 ) , 35ob ( @xmath91 ) , and 35oc ( @xmath92 ) .
model 12tj will form a 2.37 ( gravitational mass ) 0.85s after core bounce , followed by a keplerian disk after 2.64s , with a potential ejecta mass of 8.57 . however , much like the models in the previous category , model 12tj has @xmath03b of free energy available in rotation that may lead to a magneto - rotational explosion early - on , preventing collapsar formation .
this model is evolved at 1% solar metallicity , with an additional mass - loss rate scaling of 0.1 , equivalent to an overall evolution at @xmath93 solar metallicity , much below that observed for lgrb / sn sites .
we find that models he16f , he16h , and he16p have similar characteristics to model 12tj .
model 16sn forms a 2.27 0.78s after bounce .
being evolved at an effective metallicity of 0.01 solar , it has a lower angular - momentum budget at death and is thus more likely to avoid a magneto - rotational explosion .
however , it forms a keplerian disk only 40.6s after core bounce , with only 1.77 left over for the sn ejecta .
such characteristics might in fact be more amenable to reproduce recent observations of lgrb / sne characterized by a very early and narrow light - curve peak , as witnessed for example for grb100316d / sn 2010bh @xcite .
they may even explain why no sn is found in association with some nearby lgrbs @xcite .
finally , models 35ob and 35oc form a black hole within 0.78 and 0.97s of bounce , respectively .
model 35ob will accrete @xmath016.4 before a keplerian disk forms @xmath031.5s after the onset of collapse , 4.8 is then available for the sn ejecta . with the 35oc model , a disk forms very quickly after collapse , in 4.8s , and a significant amount of mass is exterior to the disk , @xmath94 , and thus much too large to accommodate inferred lgrb / sn ejecta masses .
however , the propensity to collapsar formation of the 35ob and 35oc model may be ill - founded if the mri is successful at powering a magneto - rotational explosion .
the free energy available in rotation is huge , i.e. on the order of 4 - 7.5b .
in fact , in the 2d magneto - hydrodynamic simulations of @xcite based on the 35oc model , it was found that , despite the large progenitor compactness , a magneto - rotational explosion was initiated @xmath0200ms after core bounce and that the mass decreased thereafter , never reaching the mass threshold for black - hole formation . in our models , the protoneutron stars in models 35ob and 35oc have @xmath95b of total rotational energy at the onset of explosion , amply matching the inferred energies of observed hypernovae .
these inferences are based on the assumption that energy extraction from the differentially - rotating layers at the proto - neutron star surface is efficient and can power an explosion . failing to do so , black - hole formation would result , although the question of energy extraction from the disk for the powering of a grb would then arise .
detailed multi - dimensional core - collapse simulations need to be carried out to investigate the efficiencies of magnetic / rotational / hydrodynamical instabilities for the transport of angular momentum and the extraction of rotational energy .
in this paper , we have performed 1d general - relativistic hydrodynamical simulations with ` gr1d ` of the collapse , bounce , and post - bounce phases of the lgrb candidates of @xcite to investigate their propensity to black - hole and disk formation .
we find that these progenitors are at odds with the proposed criteria for a collapsar progenitor or with the inferred properties of observed lgrb / sne , namely a h - deficient he - poor wr star with a massive iron core ( equivalent to a large compactness ) , a large angular momentum to form a keplerian disk soon after black - hole formation , an ejecta of @xmath010 , and an evolution at about 0.1 solar metallicity .
a critical aspect that we focus on in this study is the compactness of the progenitor cores at bounce , a quantity that helps diagnose the likelihood of black - hole formation .
we group the @xcite models in different categories according to their suitability for producing collapsars : 1 .
models with a dimensionless kerr spin parameter greater than unity at an enclosed mass of 3 , i.e. the models identified by @xcite as having the best potential for collapsar formation , fail to collapse when evolved with ` gr1d ` .
their cores are so fast spinning that the associated centrifugal acceleration leads them into expansion .
all these models are evolved until death in ` kepler ` without magnetic fields and centrifugal forces , which seems questionable given the unrealistically short spin periods at collapse .
models evolved with magnetic fields produce much lower rotation rates and most collapse with ` gr1d ` .
3 . of those evolved with magnetic fields , models with moderate rotation produce progenitors with a small compactness comparable to that characterizing the low - mass massive - star models proposed as progenitors of garden - variety core - collapse sne . a small fraction of these is endowed with sufficient angular momentum to make a proto - magnetar , and thus a potential channel for producing lgrbs .
a few models ( 12tj , 16sn ) with the fastest rotation possess a large compactness favorable for black - hole formation and sufficient angular momentum for the formation of a keplerian disk , but they require evolution at metallicities in the range 0.0001 - 0.01z@xmath96 , significantly lower than the metallicity of a few tenths solar or even higher at which these lgrb / sne are found @xcite .
although in many respects very attractive for forming a collapsar ( if we ignore its huge core angular momentum ) , model 35oc is characterized by a large envelope mass of @xmath023 , which is a factor 2 - 10 times larger than the inferred ejecta mass of lgrb / sne discovered so far ( for a summary , see @xcite ) . we note that these models have a large angular momentum in the core , as models in the previous category , and may thus experience a magneto - rotational explosion preventing collapsar formation .
our quantitative study thus spells out the various shortcomings of these progenitor stars for producing collapsars . even in those models that have the right compactness for black - hole formation and sufficient angular momentum for disk formation , it is still unresolved today how they would avoid the magneto - rotational mechanism of explosion that is used to explain hypernovae @xcite the difficulty of forming a and avoiding a magneto - rotational explosion in fast - rotating cores , at least in the models of @xcite , lends credence to the proto - magnetar model of lgrb / sne .
uncertainties in mass loss at low metallicity , and in particular during transient phases of dynamical mass loss as observed in some luminous blue variable stars , is an issue , since it may completely dominate the mass lost in the form of a weaker , but secular , steady - state wind @xcite .
this uncertain mass - loss rate plagues more severely the evolution of higher - mass stars , since 15 - 20 stars stay further away from the eddington limit , and overall lose little mass , even at solar metallicity . by what mechanism ,
at what rate , and during what phases a 100 star loses mass ( and angular momentum ) is much less well known and this directly conditions the final mass and iron core mass at collapse .
overall , this suggests that studies of collapsar progenitors would benefit from a second look .
angular momentum is key in the current collapsar and proto - magnetar models , but there is a stiff requirement on the progenitor compactness to speculate on its propensity for forming a , and thus for producing a lgrb through one or the other channel .
a major step forward in resolving those issues would be to conduct massive - star evolution with rotation , centrifugal force , and magnetic fields always all the way to the formation of a degenerate neutronized core on the verge of collapse .
this would allow a straightforward comparison of results between groups , and an easy determination of the compactness using ` gr1d ` to test the suitability of the core for black - hole formation .
the ultimate check on the collapsar model requires multi - dimensional simulations covering the whole evolution from progenitor collapse , bounce , failed explosion during the phase , formation of a followed by the formation of a keplerian disk , and the powering of a @xmath010b sn ejecta . as we emphasize , black - hole formation is perhaps one of the most difficult steps in this sequence of events , and in that respect , renders the proto - magnetar channel quite attractive for the production of hypernovae and lgrbs .
the diversity of lgrb / sne , the existence of sn - less lgrbs and of lgrb - less hypernovae , may in fact call for a variety of formation channels for these rare events , including both collapsars and proto - magnetars .
we acknowledge fruitful discussions with r. hirschi , a. beloborodov , and t. piro .
we also thank stan woosley for his comments on a draft version of this paper .
this research is supported in part by the national science foundation under grand nos .
ast-0855535 and oci-0905046 and by the sherman fairchild foundation .
is supported in part by a post - graduate fellowship from the natural sciences and engineering research council of canada ( nserc ) .
the computations were performed at caltech s center for advanced computing research on the cluster `` zwicky '' funded through nsf grant no .
phy-0960291 and the sherman fairchild foundation . | we present a quantitative study on the properties at death of fast - rotating massive stars evolved at low - metallicity , objects that are proposed as likely progenitors of long - duration bursts ( lgrbs ) .
we perform 1d+rotation stellar - collapse simulations on the progenitor models of @xcite and critically assess their potential for the formation of a black hole and a keplerian disk ( namely a collapsar ) or a proto - magnetar .
we note that theoretical uncertainties in the treatment of magnetic fields and the approximate handling of rotation compromises the accuracy of stellar - evolution models .
we find that only the fastest rotating progenitors achieve sufficient compactness for black - hole formation while the bulk of models possess a core density structure typical of garden - variety core - collapse supernova ( sn ) progenitors evolved without rotation and at solar metallicity . of the models that do have sufficient compactness for black - hole formation ,
most of them also retain a large amount of angular momentum in the core , making them prone to a magneto - rotational explosion , therefore preferentially leaving behind a proto - magnetar .
a large progenitor angular - momentum budget is often the sole criterion invoked in the community today to assess the suitability for producing a collapsar .
this simplification ignores equally important considerations such as the core compactness , which conditions black - hole formation , the core angular momentum , which may foster a magneto - rotational explosion preventing black - hole formation , or the metallicity and the residual envelope mass which must be compatible with inferences from observed lgrb / sne .
our study suggests that black - hole formation is non trivial , that there is room for accommodating both collapsars and proto - magnetars as lgrb progenitors , although proto - magnetars seem much more easily produced by current stellar - evolutionary models . |
there is class of high - energy and momentum transfer ( semi ) exclusive nuclear reactions which suits very well for studies of nuclear qcd .
there are several common features in theoretical approaches for studying these reactions : ( a ) one can clearly identify a hard sub - process ; ( b ) which can be factorized from the soft nuclear part of the reaction ; ( c ) the subprocess is hard enough for pqcd to be applicable ; ( d ) soft part of the reaction could be expressed through the measurable / extractable quantities such as partonic distribution functions , hadron - hadron scattering amplitudes , form - factors and calculable nuclear wave functions . if all above requirements are achieved then such theoretical approaches may yield ( sometimes ) parameter free predictions ( see e.g. @xcite ) and practically always will allow us to study the aspects of strong interaction dynamics which otherwise can not be investigated without using nuclear targets .
several of such reactions which satisfy above criteria are , ( i ) semiinclusive deep - inelastic nuclear reactions aimed at studies of nuclear modifications of pdfs ( emc effects)(see e.g. @xcite ) ; ( ii ) high momentum transfer elastic scattering off few - nucleon systems that can be used to study the onset of quark degrees of freedom in strongly correlated few nucleon systems@xcite , ( iii ) electrodisintegration of few - nucleon systems in the kinematics dominated by final state interaction that is used to probe the dominance of point like configurations in the hadronic wave function at large @xmath2 ( color transparency / coherence phenomena ) @xcite ; ( iv ) dis nuclear scattering at @xmath3 as a framework for investigation of the mechanism of generation of super - fast quarks@xcite ( v ) as well as , high - energy break - up of two nucleons in nuclei which can be used to probe the dynamics of strong interaction in two - nucleon system at intermediate to short distances@xcite .
in this presentation we focus on the last class ( v ) of the reactions , in which high energy photon produces two energetic nucleons which equally share the initial energy of the photon
. these reaction kinematically corresponds to the break - up of 2n system at @xmath4 angle in the @xmath5 @xmath6 center of mass reference frame .
due to completely symmetric configuration , photon predominantly probes the structure and the dynamics of the exchanged particle in the nn system ( fig.1 ) .
high energy of photon in this case provides necessary resolution to probe the qcd content of nn interaction whether it proceeds through the @xmath7 exchange ( fig.1b ) , quark interchange ( fig.1c ) or gluon exchange ( fig.1d ) .
effectiveness of these processes in probing qcd aspects of nuclear interaction can be seen ( for @xmath8 reactions ) from the following kinematical considerations@xcite ) in which : @xmath9 simple estimate shows that already at @xmath10 gev the invariant momentum transfer @xmath11 gev@xmath12 and invariant mass of the @xmath13 system @xmath14 gev .
these are conditions for which one expects an onset of quark degrees of freedom in the dynamics of strong interaction@xcite .
one of the first theoretical predictions for high energy and large cm angle @xmath8 reactions was the prediction of the @xmath15 dependence of the differential cross section based on the quark counting rule@xcite , which yields:@xmath16 .
this prediction was experimentally confirmed already starting at @xmath17 gev for several set of experiments at slac@xcite and jefferson lab@xcite .
the quark counting predictions are based on the hypothesis that the fock states with minimal number of partonic constituents dominate in two - body large angle hard collisions@xcite .
although successful in describing energy dependences of the number of hard processes , this hypothesis does not allow to make calculation of the absolute values of cross sections .
especially for reactions involving baryons , calculations within perturbative qcd underestimate the measured cross sections by orders of magnitude ( see e.g.@xcite ) .
this may be an indication that in the accessible range of energies bulk of the interaction is in the domain of nonperturbative qcd@xcite .
however , the main problem is that even if we fully realize the importance of nonperturbative interactions the theoretical methods of calculations in the nonperturbative domain are very restricted .
the underlying assumption in hard rescattering model ( hrm)@xcite is that high energy photodisintegration of two - nucleon system proceeds through the two stages in which an absorption of photon by a quark of one nucleon is followed by a high - momentum transfer ( hard ) rescattering with a quark from the second nucleon .
the latter rescattering produces a final two nucleon state with large relative momenta .
a typical diagram representing such a scenario is presented in fig.2 . analyzing the type of diagrams as in fig.2 allows us to do the following observations : * the dominant contribution comes from the soft vertices of @xmath18 transition , while quark rescattering proceeds trough the hard gluon exchange , * the @xmath18 transition can be evaluated through the conventional deuteron wave functions , * the structure of hard quark interchange interaction in the rescattering part of the reaction is similar to that of hard nn scattering , * as a result the sum of the multitude of diagrams with incalculable nonperturbative part of the interaction can be expressed through the experimentally measured amplitude of hard @xmath13 scattering .
based on these observations , calculation of the @xmath19 amplitude yields@xcite @xmath20 where @xmath21 is high momentum transfer elastic @xmath22 scattering amplitude , @xmath23 ( @xmath24 ) represents the helicity wave function of nucleon and @xmath25 is the helicity of incoming photon . based on eq.([ampl ] ) one obtains the following expression for the differential cross section of @xmath19 reaction : @xmath26 where @xmath27 and @xmath28 .
the interesting properly of the function @xmath29 is that @xmath30
. therefore for @xmath4 cm scattering hrm prediction is parameter free . since @xmath22 cross section in high momentum transfer behaves like @xmath31 , eq.([crs_gdpn ] )
predicts same @xmath32 dependence as quark counting rule without requiring an onset of pqcd regime . also , due to angular dependence of the @xmath22 amplitude
, hrm predicts an angular distribution being not symmetric around @xmath4 cm .
these predictions agree reasonably well with the experimental data@xcite(see e.g. fig.3 ) .
cm scattering ( top ) and angular dependence of the cross section at @xmath33 gev ( bottom).,width=288,height=192 ]
with all its success and accuracies yet to be improved hrm is only one of the approaches in describing hard photodisintegration reactions .
other models such as reduced nuclear amplitude ( rna ) formalism@xcite and quark - gluon string ( qgs ) model@xcite describe many features of hard photodisintegration reaction , with qgs being rather successful in describing lower energy data .
however rna and qgs require an absolute normalization .
recently it was suggested@xcite that the break - up of pp pair from @xmath1 will further advance our understanding of the dynamics of hard photodisintegration of two - nucleon systems and allow further discrimination between above mentioned models . within hrm the typical diagram describing two - proton break - up
is shown in fig.4 .
.,width=326,height=134 ] hrm calculation similar to that of deuteron break - up reaction yields : @xmath34 where @xmath35 and @xmath36 with @xmath37 being @xmath38 helicity amplitudes .
since the cross section of @xmath38 scattering enters in eq.([pp ] ) one of the interesting predictions of hrm is the possibility of observation of energy oscillations at @xmath4 cm scattering similar to one observed in elastic @xmath38 scattering
. another interesting feature of two - proton break - up reactions is the fact that at lower energies this reaction is three - step@xcite rather than two - step process .
hrm in fact predicts that with an increase of energy due to the onset of quark - interchange ( rather than meson - exchange ) mechanism the two - body processes will dominate the cross section .
the pioneering experiment of high energy @xmath38 break - up reaction@xcite was recently performed at jefferson lab which may shed new light on many issues of hard rescattering processes .
in addition to the cross section measurements , polarization observables may provide a new insight into the dynamics of hard photodisintegration .
original motivation for polarization measurements in high energy photodisintegration reaction was the expectation that the onset of the pqcd regime in the reaction dynamics will be accompanied by an observation of the helicity conservation in polarized reaction .
both energy and angular distributions of several polarization observables have been measured at jlab@xcite .
although the energy range covered was rather restricted , it provided an interesting insight into the structure of hrm .
one of the unique features of hrm is that the struck quark carries the helicity of incoming photon . as a result one of the final nucleons
will carry the bulk of the polarization of incident photon ( see e.g. eq.([ampl ] ) ) .
thus in hrm photon plays as a helicity selector for the final nucleons .
this yields a prediction for large asymmetry@xcite ( @xmath39 ) for the longitudinal polarization of outgoing nucleons .
in ref.@xcite we predicted a sizable asymmetry for @xmath39 even though the existing data@xcite ( with rather large errors ) at that time were indicating on vanishing values of @xmath39 . for @xmath40
gev.,width=364,height=192 ] however , recent data@xcite , appears to confirm hrm prediction for large values of @xmath39 ( see fig.5 )
. it will be interesting also to check the other hrm prediction that @xmath39 will continue to approach to unity with an increase of photon energy at @xmath4 cm scattering .
it is very interesting that above described helicity selection mechanism of hrm predicts ( _ an opposite _ ) vanishing value of @xmath39 for two - body break up of proton pair from @xmath1 .
this follows from the fact that the dominant part of the amplitude which represents two final state nucleons polarized in same direction are proportional to the nuclear ground state wave function with two initial nucleons having same helicities .
due to pauli principle this part of the amplitude is strongly suppressed for the proton - pair in @xmath1 target .
no such suppression exists for @xmath22 break up reactions .
there is an accumulating evidence that hard rescattering mechanism explains the underlying dynamics of high energy and large cm angle break - up of a nucleon pair from @xmath0 and @xmath1 targets .
one of the important features of hrm is that its prediction of unpolarized cross section at @xmath4 center of mass photodisintegration of deuteron is parameter free and no further adjustments are required . a new venue for advancing our understanding of the dynamics of hard break - up reactions could be an extension of these studies to the kinematics in which two excited baryonic states ( like @xmath41-isobars ) are produced at large center of mass angles of @xmath5 @xmath13 system .
frankfurt , g.a .
miller , m. m. sargsian and m. i. strikman , phys .
lett . * 84 * , 3045 ( 2000 ) .
m. m. sargsian _ et al .
_ , j. phys .
g * 29 * , r1 ( 2003 ) [ arxiv : nucl - th/0210025 ] .
s.j . brodsky and b.t .
chertok , phys .
lett . * 37 * , 269 ( 1976 ) .
l. l. frankfurt _ et al .
_ , phys . rev .
c * 56 * , 1124 ( 1997)[arxiv : nucl - th/9603018 ] . m. m. sargsian , int .
j. mod . phys .
e * 10 * , 405 ( 2001 ) [ arxiv : nucl - th/0110053 ] .
l. l. frankfurt and m. i. strikman , phys .
rept . * 160 * , 235 ( 1988 ) .
j. napolitano _ et al .
_ , phys . rev
. lett . * 61 * , 2530 ( 1988 ) ; s.j .
freedman _ et al .
_ , phys . rev . *
c48 * , 1864 ( 1993 ) .
belz _ et al .
lett . * 74 * , 646 ( 1995 ) . c. bochna
_ , phys . rev . lett . * 81 * , 4576 ( 1998 ) .
schulte _ et al .
* 87 * , 102302 ( 2001 ) .
e. c. schulte _ et al .
_ , phys .
c * 66 * , 042201 ( 2002 ) .
s.j . brodsky and g.r .
farrar , phys .
* 31 * , 1153 ; v. matveev , r.m .
muradyan and a.n .
tavkhelidze , lett .
nuovo cimento * 7 * , 719 ( 1973 ) . n. isgur and c.h .
llewellyn smith , phys .
* 52 * , ( 1984 ) 1080 .
a. radyushkin , acta phys .
* b15 * , 403 ( 1984 ) . m. m. sargsian , phys .
b * 587 * , 41 ( 2004 ) [ arxiv : nucl - th/0309054 ] .
m. m. sargsian , arxiv : nucl - th/0208027 . s. j. brodsky _ et al .
_ , phys .
b * 578 * , 69 ( 2004 ) [ arxiv : nucl - th/0305068 ] .
s.j . brodsky and j.r .
hiller , phys .
c * 28 * , 475 ( 1983 ) .
l. a. kondratyuk _ et al .
_ , phys .
c * 48 * , 2491 ( 1993 ) ; v. y. grishina , _ et al .
j. a * 10 * , 355 ( 2001 ) . | * abstract * we discus recent developments in theory of high energy two - body break - up reactions of few - nucleon systems . the characteristics of these reactions are such that the hard two - body quasielastic subprocess can be clearly separated from the accompanying soft subprocesses .
we discuss in details the hard rescattering model ( hrm ) in which hard photodisintegration develops in two stages . at first , photon knocks - out an energetic quark which rescatters subsequently with a quark of the other nucleon .
the latter provides a mechanism of sharing the initial high momentum of the photon by the outgoing two nucleons . within hrm
we discuss hard break - up reactions involving @xmath0 and @xmath1 targets .
another development of hrm is the prediction of new helicity selection mechanism for hard two - body reactions , which was apparently confirmed in the recent jlab experiment . |
the concept of the @xmath0-matrix is deeply rooted in the concept of minkowski spacetime and its poincare symmetry .
the poincare group on the minkowski spacetime allows us to define particle four - momentum and spin .
this in turn allows us to consider particle scattering and decay of incident states into outgoing states .
the resulting probability amplitudes are summarized as @xmath0-matrix elements .
the conservation of probability imposes unitarity of the @xmath0-matrix .
the @xmath0-matrix commutes with the generators of the poincare group . as a result of this symmetry
the total four - momentum and total angular momentum are conserved in particle scattering and decays .
internal symmetries of the @xmath0-matrix impose additional conservation laws . in @xmath0-matrix theory particle scattering and decays
are isolated and time - reversible quantum events in minkowski spacetime .
the reason for this is that minkowski spacetime is empty .
there is no quantum environment in minkowski spacetime with which the scattering and decay processes could interact .
another reason is that the scattering and decay processes do not interact with the minkowski spacetime itself since it has no quantum structure that , in effect , could present itself as a quantum environment .
particle scattering and decay processes take place , in fact , in real physical universe .
suppose that there exists a quantum environment in the universe that interacts with particle processes .
such interaction can not originate in the known interactions of the standard model .
these interactions would lead to observable violation of the conservation laws and render the dynamics of particle interaction inaccessible to experiment .
the interaction of particle processes with the quantum environment must originate from the outside of the standard model .
if the quantum environment and this new kind of interaction are to be an integral part of the nature , then they must be fully consistent with the conservation laws and unitarity of the standard model .
there exists such an interaction in the nature .
it is the pure dephasing interaction between a quantum system @xmath0 and a quantum environment @xmath12 which is a non - dissipative interaction that affects only the phase(s ) of the quantum system @xmath0 . in general
, its effect is to change the quantum information content of the quantumm system @xmath0 . unlike the force of gravity , which has undetectable effects on particle processes
, the pure dephasing interaction could be observable .
the key to observing dephasing interaction , and thus the quantum environment , is the @xmath0-matrix unitarity .
the evolution of an isolated initial state @xmath13 into isolated final states @xmath1 is governed by the unitary evolution law @xmath14 evolution of an initial state @xmath13 interacting with a quantum environment @xmath12 is governed by a non - unitary evoluion law given by kraus representation @xmath15 where @xmath16 are unitary or non - unitary kraus operators @xcite . the necessary and sufficient condition for the non - unitary evolution law to be consistent with the unitary evolution law
is that the interaction with the environment be a final state interaction which involves only the produced state @xmath1
. the non - unitary evolution of the @xmath0-matrix final state @xmath1 yields the observed non - unitary final state @xmath2 @xmath17 the effect of the pure dephasing interaction with the environment thus shall be to modify the phases of the @xmath0-matrix amplitudes in the observed amplitudes of the new state @xmath2 .
the existence of the quantum environment will manifests itself in the difference between the phases of the observed and @xmath0-matrix amplitudes and therefore in an apparent violation of the unitary evolution law . to gain information about the phases of the @xmath0-matrix
amplitudes we shall use the fact that the @xmath0-matrix unitary evolution law evolves pure initial states @xmath13 into pure final states @xmath1 in exclusive processes . in two - body processes such as @xmath18
this condition imposes no constraints on the amplitudes and thus no specific information about their phases .
the situation is different in @xmath5 and similar production processes . in this work
we aim to test the unitary evolution law in the pion production @xmath19 measured at cern on polarized target at 17.2 gev / c @xcite .
we develop the necessary spin formalism and show that the purity of the final state density matrix in @xmath5 processes is controlled by the recoil nucleon polarization .
evolution of pure initial states to pure final states imposes 9 constraints on 16 angular intensities describing the final state . using @xmath3-parity conservation
we show that all constraints are identities provided that the partial wave amplitudes satisfy the conditions @xmath20 where @xmath21 and @xmath22 are parity conserving nucleon transversity unnatural and natural exchange amplitutes with dipion spin , helicity and nucleon transversity @xmath23 and @xmath24 , respectively .
the conditions ( 1.4 ) imply that the relative phases between any two unnatural or two natural exchange amplitudes of the same transversity @xmath25 as well as between any unnatural and natural exchange amplitudes of opposite transversities must be @xmath6 or @xmath7 . as a result ,
the unnatural and natural amplitudes share a common phase @xmath26 and @xmath27 , respectively . because strong , electromagnetic and weak interactions do not mix particles with different spins , self - consistency requires that there be no mixing of resonances of different spins in any partial wave amplitudes with such unitary phases .
the elegant simplicity and uniqueness of these predictions renders the test of unitary evolution law possible using the existing cern measurements on polarized target at dipion masses below 1080 mev where @xmath0- and @xmath3-wave amplitudes dominate .
we found that the measurements of density matrices @xmath8 and @xmath9 yield a solution with @xmath10 mixing in the @xmath0-wave .
but this solution is entirely excluded by the measured data on density matrix @xmath11 within at least five standard deviations .
this result shows that the data require @xmath10 mixing in the @xmath0-wave but reject unitary phases in an apparent violation of the unitary evolution law .
all previous amplitude analyses of @xmath28 at 17.2 gev / c @xcite and at 1.78 gev / c @xcite as well as of @xmath29 at 5.98 and 11.85 gev / c @xcite found non - unitary relative phases of all transversity amplitudes .
recent amplitude analysis of @xmath0- and @xmath3-wave subsystem in @xmath28 at 17.2 gev / c established that the width of @xmath30 resonance peak observed in all @xmath3-wave amplitudes does not depend on its helicity @xmath31 @xcite as required by the rotational / lorentz symmetry of the @xmath0-matrix .
furthermore , the non - unitary relative phases of the @xmath0-and @xmath3-wave amplitudes are near the unitary values .
these findings show that the observed non - unitary phases are consistent with the @xmath0-matrix unitary evolution law for the production process .
this is possible if the non - unitary phases arise from a pure dephasing interaction of the produced @xmath0-matrix final state @xmath1 with a quantum environment with .
the contrast between the predicted unitary relative phases and the observed non - unitary phases therefore presents an unambigous evidence for the non - unitary evolution of the produced final state and supports the hypothesis of the existence of the quantum environment and its pure dephasing interaction with particle scattering processes . in a sequel paper @xcite
we show that the consistency of the pure dephasing interaction with the standard model in @xmath5 processes requires that it be a dipion spin mixing interaction the effect of which is the mixing of @xmath0-matrix partial wave amplitudes to form observable amplitudes .
the theory predicts @xmath10 mixing in the @xmath0-and @xmath3-wave amplitudes in @xmath32 . the predicted moduli and relative phases of the mixed amplitudes are in an excellent qualitative agreement with the experimental results @xcite .
the evidence for @xmath10 mixing dates back to 1960 s @xcite and was later confirmed in all amplitude analyses of @xmath28 and @xmath29 on polarized targets .
the issue of the experimental test of the unitary evolution law was first raised in 1974 by marinov who suggested to describe the time evolution of @xmath33 system using a model of a non - unitary evolution from pure states into mixed states @xcite .
he found that such evolution is time - irreversible and violates @xmath34 symmetry and proposed to make a complete measurement of matrix elements of @xmath35 to test the unitary evolution law . in 1980
wald showed rigorously that any scattering process of particles that evolves pure initial state into a mixed final state is not an invertible process and therefore it is time - irreversible and violates @xmath34 symmetry @xcite .
he suggested that such non - unitary processes will occur in curved space - time , and that quantum gravity violates @xmath34 symmetry and time - reversal invariance . in 1982
hawking pointed out that particle scattering does not take place in a structureless continuum of the minkowski space - time but in an environment of quantum space - time fluctuations and suggested that pure initial states of interacting particles will evolve into mixed final states due to the interaction of the particle scattering process with quantum fluctuations of the space - time metric - at any energy @xcite .
hawking questioned the universal validity of the unitary time evolution in the presence of metric fluctuations , and suggested that initial and final state density matrices @xmath36 and @xmath37 are connected by a non - unitary evolution law described by a linear but non - unitary and non - invertible superscattering operator @xmath38 to avoid negative probabilities the mapping ( 1.5 ) must be completely positive .
a linear mapping ( 1.5 ) is completely positive if and only if it has the form of the kraus representation ( 1.2 ) @xcite .
recently unruh and wald @xcite and oppenheim and reznik @xcite demonstrated the feasibility of non - unitary evolution in quantum field theories .
hawking s ideas inspired attempts to test unitary evolution law experimentally .
the efforts focused mainly on non - unitary time evolution of neutral kaons @xmath33 system using lindblad type evolution laws @xcite and led to the predictions of @xmath34 violation and a modification of epr correlations @xcite . during the recent years experiments with neutral kaons
have yielded sensitive results on violations of @xmath34 symmetry , time reversal invariance and entanglement of kaon pairs @xcite .
so far these experiments did not provide a conclusive confirmation of a non - unitary evolution of free neutral kaon systems .
these observations can be understood as the result of the absence of the diparticle spin mixing in the time evolution of the @xmath33 system .
the paper is organized as follows . in section ii .
we briefly discuss the evolution of pure initial states by unitary and non - unitary evolution laws . in section iii .
we show that @xmath0-matrix unitary evolution implies evolution from pure initial states into pure final states in exclusive processes . in sections iv
we review and develop the spin formalism necessary to derive and discuss the constraints imposed by the unitary evolution law on the nucleon transversity amplitudes .
we emphasize the coherent superposition of diparticle states and the constraints imposed by the conservation of @xmath3-parity on partial wave amplitudes .
final state density matrix is defined and its properties derived in some detail . this material will be used also in the sequel paper . in section vii .
we present the unitary constraints on angular intensities from which we derive the unitary condition ( 1.4 ) and a self - consistent set of unitary phases . in section viii .
we present a unique unitary solution for @xmath0- and @xmath3-amplitudes below 1080 mev and show it is excluded by the data on observables @xmath11 .
this conclusion does not change when we include @xmath39-waves in the analysis .
we discuss the evidence for the quantum environment and the dephasing interaction in section ix . and a physical interpretation of the unitary and non - unitary relative phases in section x .. we present our conclusions and outlook in the section xi .. the appendix provides an outline of the proof of the unitary condition ( 1.4 ) .
let @xmath40 be a hilbert space of vector states with an orthonormal basis @xmath41 where @xmath42 .
let @xmath43 be the hilbert - schmidt space of linear operators on @xmath40 with an orthonormal basis @xmath44 .
density matrix @xmath45 is a hermitian and positive operator @xmath46 with @xmath47 .
density matrix @xmath48 represents a pure state if it satisfies condition @xmath49 .
a quantum state satisfies this purity condition if and only if its density matrix has the form @xmath50 where @xmath51 is a vector in @xmath40 .
let @xmath52 be a linear operator on @xmath40 and @xmath13 an arbitrary density matrix in @xmath43 .
then @xmath52 defines a mapping - or an evolution law - of the state @xmath13 into a state @xmath35 @xmath53 let @xmath54 be a pure state and @xmath55 .
then @xmath56 is also a pure state .
if @xmath52 is a unitary operator the mapping is a trace conserving and invertible unitary evolution .
otherwise it is a simple non - unitary evolution with normalized density matrix @xmath57 . to be physically meaningful the mappings from initial to final states must preserve the positivity of all probabilities which requires that they be completely positive . the necessary and sufficient condition for an evolution law to be completely positive is that it has the form called kraus representation @xcite @xmath58 where @xmath16 are unitary or non - unitary kraus operators . for trace preserving evolution
they satisfy completness relation @xmath59 kraus representation ( 2.2 ) describes a general completely positive , non - unitary and non - invertible evolution .
let @xmath54 be a pure initial state .
then the final state is a mixed state @xmath60 a special case of ( 2.2 ) is the evolution law ( 2.1 ) which evolves pure initial states into pure final states .
it describes evolution of isolated quantum systems while the kraus representation describes a non - unitary evolution of open quantum systems .
kraus representation arises from a unitary evolution law governing the co - evolution of the quantum system with its quantum environment after the interacting degrees of freedom between the two systems have been traced out in their joint final state density matrix @xcite .
the dissipative dephasing interactions exchange not only phases but also four - momentum and/or angular momentum . there is no exchange of four - momentum and/or angular momentum in pure dephasing interactions of the quantum system with its quantum environment .
unitary @xmath0-matrix evolves an arbitrary initial state @xmath13 into a final state @xmath61 any pure initial state @xmath13 can be written in the form @xmath54 .
the evolution operator @xmath0 brings the state vector @xmath62 to a state vector @xmath63 so that @xmath50 is a pure state . using a completness relation @xmath64
the state vector @xmath51 has an explicit form @xmath65 where the first sum is over all allowed final states @xmath66 , the second sum is over final state spins @xmath67 and the integration is over the entire phase space of final state momenta @xmath68 .
the symbol @xmath69 labels the quantum numbers of the state @xmath66 .
the density matrix @xmath50 has an explicit form @xmath70 where latexmath:[\ ] ] and similarly for @xmath617 .
the comparison of ( 10.9 ) and ( 10.8 ) suggests the real and imaginary parts of these partial wave amplitudes are both akin to complex standing waves on strings open and closed at both ends .
this similarity suggests to consider the @xmath0-matrix partial wave amplitudes as representations of some kind of complex dynamical ( vibrational ) modes confined to a finite dynamical region of a transient compound state .
the modes are associated with the produced partial waves @xmath650 . in empty space
these partial waves propagate without a change leaving the @xmath0-matrix amplitudes intact . in the presence of a quantum environment
the propagating partial waves @xmath650 are modified by the environment .
this results in the modification of the @xmath0-matrix partial wave amplitudes into the observed partial wave amplitudes with the ensuing non - unitary relative phases .
the quantum environment is responsible not only for the generation of the non - unitary phases but also for the observed @xmath10 mixing @xcite .
we have presented a spin formalism that allowed us to construct the full final state density matrix @xmath1 in @xmath5 processes in @xmath0-matrix theory .
we have shown that the unitary evolution law imposes specific constrains on the relative phases of the transversity amplitudes @xmath651 and @xmath652 .
the contrast between these predicted phases and the observed phases in all amplitude analyses of the pion production process presents an apparent violation of the unitary evolution law .
previous attempts to test unitary evolution law were framed as the test of the quantum mechanics itself .
the central idea of this work is that an apparent violation of the unitary evolution law does not signal the breakdown of quantum theory .
rather it is an unambigous evidence for a non - unitary evolution of the produced final state @xmath1 to the observed final state @xmath2 generated by a new interaction of the state @xmath1 with a quantum environment . to render the production mechanism accessible to experimental observation and to be consistent with the standard model
this new interaction must be a pure dephasing interaction .
this new non - standard dynamics represents a new physics beyond the standard model .
we identify the quantum environment with dark neutrinos which form a distinct component of dark matter .
this interpretation endows the quantum environment with a physical and material identity which connects it to the physics of the dark sector of the universe .
a part of the quantum environment are also the active and light sterile neutrinos backgrounds .
we elaborate on this model of the quantum environment in ref .
@xcite .
the spin formalism developed in this work and its consequences for the unitary phases of partial wave amplitudes apply equally well to a number of other meson production processes such as @xmath653 , @xmath654 , @xmath655 and others .
these processes could be measured on polarized target and the measurements with @xmath656 would also allow measurements of recoil @xmath656 polarization by its weak decays .
modern polarized targets reach high values of polarization and enable to select an arbitrary direction of the polarization vector @xcite .
such experiments would provide new and independent tests of the unitary evolution law and advance our understanding of the quantum environment and its pure dephasing interactions with particle scattering processes .
the unitary evolution constraints ( 7.12 ) on parity conserving transversity amplitudes follow from the constraint @xmath382 which we can write in the form @xmath657 expansions of the intensities involve @xmath658 and @xmath659 on the l.h.s . and r.h.s .
of ( a1 ) , respectively . using the expressions for @xmath660 in the table i. in the intensities ( 6.32 ) and ( 6.33 ) , relabeling of some terms , parity relations ( 5.16 ) , and a relation for spherical harmonics @xmath280 , the intensities on l.h.s . of ( a1 ) read @xmath661 @xmath662 @xmath663 @xmath664 @xmath665 @xmath666 where @xmath388 for @xmath389 and @xmath390 for @xmath391 . using these expressions the l.h.s .
of ( a1 ) reads @xmath667 @xmath668 where @xmath669 and where we used some relabeling and identities @xmath670 @xmath671 after a similar procedure the r.h.s . of ( a1 ) can be brought to the form @xmath672 since @xmath673 the equation ( a5 ) immediately implies the conditions ( 7.12 ) .
k. rybicki , _ data tables of spin moments measured in @xmath32 on polarized target at 17.2 gev / c for dipion masses 580 - 1600 mev and four - momentum transfers @xmath674=0.01 - 0.20 @xmath675 _ , private communication , 1996 . | unitary evolution law describes isolated particle scattering processes in an empty minkowski spacetime .
we put forward a hypothesis that the physical universe includes a quantum environment that interacts with some particle scattering and decay processes .
while the scattering process is governed by the @xmath0-matrix dynamics and its conservation laws and unitarity , the interaction with the environment evolves the produced final state @xmath1 to the observed state @xmath2 . to be consistent with the standard model
this new interaction must be a pure dephasing interaction .
governed by a non - unitary evolution law , it modifies the phases of the @xmath0-matrix amplitudes and can give rise to mixing of such amplitudes to form observed amplitudes .
we present the first test of unitary evolution law in particle scattering .
conservation of @xmath3-parity in strong interactions imposes constraints on partial wave helicity and nucleon transversity amplitudes in @xmath4 processes .
an independent set of constraints on these amplitudes is imposed by the @xmath0-matrix unitary evolution law .
the unitary evolution evolves pure initial states into pure final states leading to 9 independent constraints on 16 components of angular intensities in @xmath5 processes .
when expressed in terms of parity conserving transversity amplitudes , all 9 constraints are identities provided a single constraint on the transversity amplitudes holds true .
the constraint implies that relative phases between transversity amplitudes of the same naturality and transversity must be @xmath6 or @xmath7 . assuming a self - consistent set of these unitary phases we use the cern data on spin observables @xmath8 and @xmath9 to determine a unique solution for the @xmath0- and @xmath3-wave moduli below 1080 mev .
the data require @xmath10 mixing in the @xmath0-wave but this unitary solution is excluded by data on observables @xmath11 within at least 5 standard deviations .
all previous amplitude analyses of @xmath5 processes found non - unitary relative phases in an apparent violation of the unitary evolution law .
the contrast between the predicted unitary relative phases and the observed non - unitary phases presents an unambigous evidence for the non - unitary evolution of the produced final state and supports the hypothesis of the existence of a quantum environment and its pure dephasing interaction with particle scattering processes . |
when surfactant molecules , or amphiphiles which consist of molecules that combine both polar and non - polar parts , are dissolved in a single solvent such as water , these molecules tend to form bilayer membranes where the hydrocarbon parts of each monolayer are aggregated in the middle of the bilayer to reduce the contact between the water and the non - polar parts of the molecule due to the hydrophobic interactions of the hydrocarbon chains .
these membranes are not covalenly bonded , but rather stabilized by the weaker hydrophobic interaction .
these membranes can have characteristic sizes which may be much larger than those of a single molecule .
the sizes and shapes of these membranes can change as a function of temperature , salinity , and/or surfactant concentration . depending on the physical conditions of the system
, these membranes can form random extended surfaces , regular periodic structures , or closed vesicles separating an interior from an exterior @xcite . furthermore ,
some bilayer membranes are prototypes of biological systems , although it should be noted that true biological membranes have several components and greater complexity . in a wide class of membranes
, molecules move freely within the membrane forming a @xmath11-dimensional fluid offering little resistance to changes in membrane shape .
such membranes are physical examples of random surfaces , which can undergo violent shape changes .
molecules in membranes can also exhibit varying degrees of orientational and positional order including tilt(smectic - c ) , hexatic , and crystalline orderings similar to those found in free standing liquid crystal films @xcite .
these ordered membranes provide fascinating laboratories for the study of the coupling between order and geometry , analogous in many ways to the coupling between matter and geometry in general relativity .
the underlying cause of this coupling is as follows .
a field describing orientational order can not be parallel everywhere if it is forced to lie on a surface , such as that of a sphere , that curves in two directions and thus has nonzero gaussian curvature .
it could lower its energy by flattening the surface .
an additional complication arises when order develops on closed surfaces .
a closed surface can be classified according to its genus @xmath12 : the number of handles .
orientational order on a closed surface necessarily has topological defects ( vortices ) with total strength ( vorticity ) equal to the euler characteristic @xmath13 of the surface @xcite .
tangent - plane order on a sphere ( torus ) will have vorticity @xmath11 @xmath14 , since a sphere ( torus ) has genus @xmath15 ( @xmath16 ) , respectively .
the continuous development of vector order on a deformable surface of genus zero will be accompanied by a continuous change from spherical to ellipsoidal shape @xcite .
since vortices are energetically costly , it may be favorable for a closed physical membrane to transform into an open cylindrical structure when tangent - plane order develops in response to changes in temperature or other control variables @xcite .
indeed , there are a number of experimental examples of shape changes that may be explained by the development of tangent - plane order .
orientational orders for smectic - c and hexatic liquid crystals are described by an @xmath0-atic order parameter @xmath17 \rangle$ ] with @xmath18 and @xmath19 , respectively . in this paper
, we present the existence of the kt transition of @xmath0-atic order on a closed surface of genus zero and investigate the development of @xmath0-atic order below the kt transition temperature and the concomitant change in shape from spherical to non - spherical .
an @xmath0-atic order parameter can have vortices of strength @xmath8 , and , since it is generally favorable to form vortices of minimum strength , we expect @xmath7 maximally separated vortices of strength @xmath8 to be present in the ordered phase ( low temperature phase ) .
thus for @xmath18 , we expect two antipodal vortices , and respectively for @xmath20 , @xmath21 and @xmath19 , we expect vortices at the vertices of a tetrahedron , an octahedron , and an icosahedron
. indeed calculations on a rigid sphere confirm this conjecture @xcite .
we also expect the vesicle shape to change from spherical to ellipsoidal , tetrahedral , octahedral , or icosahedral in the four cases above .
our calculations are based on a phenomenological hamiltonian for a complex order parameter field whose coupling to shape occurs via a covariant derivative and via changes in the metric tensor @xcite .
the model is almost identical to the landau - ginzburg theory of superconductivity except that vorticity is fixed by surface topology rather than energetically determined by an external magnetic field .
the ordering transition we find is very similar to the transition from a normal metal to the abrikosov vortex lattice in a superconductor , and indeed our analysis follows very closely that of abrikosov @xcite .
we find the order parameter , which is necessarily spatially inhomogeneous because of the topological constraint on the total vorticity , from among a highly degenerate set of functions that diagonalize the harmonic hamiltonian on a rigid sphere .
this degenerate set has exactly @xmath7 zeros at arbitrary positions on the sphere and is very similar in form to fractional quantum hall wave functions @xcite .
this paper is organized as follows . in sec .
ii , we review fluid membranes in some detail and describe their free energy as a function of curvature .
we introduce a tangent - plane orientational order parameter and its coupling to shape changes in sec .
the model hamiltonian for a deformable spherical membrane with @xmath0-atic order is presented in the normal gauge representation and the existence of the kt transition is described in sec .
iv . minimization scheme and the calculated shapes for temperatures below the kt transition temperature to the ordered @xmath0-atic phase for @xmath22 , and @xmath23 are given in sec .
v , and we give concluding remarks in sec . vi .
in this paper , we use the term membrane to denote a thin film of one material that separates two similar materials ( bilayer membrane ) .
we focus on fluid membranes and their possible degrees of orientational order . in fluid memebranes
, there is no in - plane shear modulus and the only in - plane deformations are compressions / expansions . an @xmath0-atic membrane ,
although fluid , will sustain a local orientational order defined modulo rotations by @xmath24 , which is described by a local order parameter @xmath25 $ ] .
although fluid membranes can be composed of many different types of chemical and molecular species , their behavior ( shapes , fluctuations , thermodynamics ) can be understood from a unified point of view that considers their free energy of deformation .
if the membrane were constrained to lie in a plane , the only relevant energy would be the compression of the molecules , that is , change of the average area per molecule .
this is analogous to sound waves in a 3-dimensional fluid ; there is no low - frequency reponse of the system to shear .
however , since the membrane can also deform in the normal direction , there is an additional set of modes describing the conformations of the film .
these out - of - plane deformations are known as bending or curvature modes and the free energy associated with such modes is known as the curvature free energy . while a general deformation of the membrane involves changes in both volume and curvature , we shall see that the lowest energy deformations usually involve only curvature . in most systems , changing the average volume of the membrane is a higher energy process and hence is less important when effects involving the thermal behavior of the membrane are considered . for a membrane with finite thickness , we denote as pure curvature deformations those perturbations of the membrane that do not change the overall membrane volume , but where there may be local stretching and compression of different parts of the film .
the location of the interface within a membrane can be chosen to lowest order in the curvature energy so that the surface which defines the interface undergoes no stretching or compression ( neutral surface ) . for long - wavelength curvature deformations whose wave lengths are much larger than the membrane thickness , the exact position of the neutral surface within the membrane is not crucial .
the membrane is characterized by the area per molecule , @xmath26 , its thickness , @xmath27 , and its curvature . for simplicity , we assume that the equation of state of the membrane determines the thickness as a function of the area per molecule .
a simple example is the case of an incompressible molecule where the product of @xmath28 is constrained to equal the molecular volume so that @xmath29 .
we thus take the curved membrane to be characterized only by the area per molecule and its curvature . consider a locally flat , isolated interface .
saturation occurs when the free energy achieves a minimum with respect to @xmath26 ; @xmath30 where @xmath31 is the free energy per molecule for a flat layer and @xmath26 is the area per molecule .
the free energy per molecule is minimized when @xmath32 , the optimal value of the area per molecule which arises from a balance of terms such as the entropy and the tension terms or attractions .
the entropy favors a larger area per molecule because of the greater number of center - of - mass positions and chain conformations , while the tension terms and attractions favor a small value of @xmath26 to reduce the contact of the hydrocarbon chains with the water .
there can be deviations in the area per molecule from this optimal value and the energy cost of such a compression or expansion is @xmath33 where the primes signify a derivative with respect to @xmath26 . however , these deformations are typically of higher energy than the curvature deformations .
a membrane can change its shape or size with a much lower free energy cost than that required to compress or expand it .
it is important to remember , therefore , that for insoluble amphiphiles it is the saturation of the membrane and the minimization of the area per molecule that permits the usual surface - tension term to be neglected ; @xmath34 .
the surface tension is no longer relevant since the molecules adjust their area to optimize the free energy and it is the curvature energy that determines the properties of the membrane . to construct the effective free energy for fluid membranes , it is convenient to introduce general curvilinear coordinates @xmath35 and the metric structure .
defining locally a system of coordinates @xmath36 on the membrane and denoting by @xmath37 the position of the point @xmath36 in bulk @xmath38-dimensional euclidean space , the metric tensor @xmath39 induced by the embedding is @xmath40 where @xmath41 is a tangent vector to the surface , and the element of area is @xmath42 the extrinsic curvature tensor @xmath43 is defined by @xmath44 where @xmath45 is a covariant derivative with respects to the metric @xmath46 . in general
, @xmath43 is normal to the surface [ 3 ] .
therefore , in the particular case of a surface in @xmath47 , @xmath43 is proportional to the unit normal vector @xmath48 to the surface and it is written as @xmath49 where @xmath50 is a symmetric tensor .
we can dicuss the curvature energy using symmetry considerations .
the free energy must be only a function of the field @xmath37 , invariant under displacements and rotations in @xmath51 , and reparametrization - invariant .
expanding in local terms involving more and more derivatives and keeping only the terms relevant by naive power counting , the most general form of the curvature energy @xmath52 up to quadratic order in the curvatures can be written in terms of the mean and gaussian curvatures and has only two terms @xmath53 here , @xmath54 is the mean curvature , @xmath55 , and @xmath56 is the gaussian curvature , @xmath57 .
this form for the curvature energy was discussed by helfrich and canham and states that the mean curvature that minimizes the energy has a value @xmath58 , termed the spontaneous curvature of the membrane @xcite .
the energy cost of deviating from the spontaneous curvature is the bending or curvature modulus , @xmath5 .
the parameter @xmath59 , known as the saddle - splay modulus , measures the energy cost of saddle - like deformation .
the spontaneous curvature describes the tendency of the bilayer membrane to bend .
it is viewed to arise from the fact that the two layers of the bilayer do not have areas per molecule which are exactly equal , nor are their curvatures equal and opposite .
this can result in a vesicle that is more stable than a flat , lamellar bilayer since it may be that the favorably curved outer layer has a smaller value of the area per molecule and hence more molecules in it than the inner layer .
the bending moduli , @xmath5 and @xmath59 , arise from the elastic constants determined by the head - head and tail - tail interactions .
it is expected that these moduli are senstively dependent on the surfactant chain length but only weakly dependent on the head - head interaction strength .
the parameters @xmath58 , @xmath5 , and @xmath59 can be derived from a simple microscopic model that incorporates both the change in the area per molecule and the curvature .
we note that a stable membrane will always have @xmath60
. however , the sign of @xmath59 can be either positive or negative .
membranes that prefer isotropic shapes where the gaussian curvature @xmath61 such as spheres or planes will have @xmath62 , while membranes that prefer saddle shapes where the gaussian curvature @xmath63 will have @xmath64 .
one can show that the requirement of a positive - definite quadratic term implies that membranes are only stable if @xmath65 ; otherwise higher order curvature terms are needed to stabilize the system .
we consider the spherical membrane whose topology is fixed .
thus the gaussian curvature energy gives constant contribution by gauss - bonnet theorem @xmath66 .
in a fluid membrane , molecules can flow freely to adapt themselves to any particular shape of the surface .
if correlations among the molecular positions of the molecules forming the membrane exist , the molecules may exhibit in - plane crystalline order and form a kind of two - dimensional solid . on the other hand
, the molecules may exhibit a weaker order in which the orientations are correlated at long distance scale .
orientational order means that to each point on the membrane is associated a preferred direction whithin the tangent plane of the membrane .
for example at high temperature , the stable phase of the membrane is generally the liquid phase with no translational or orientational order if we are considering the hexatic order , or the smectic - a phase in which molecular axes are normal to the surface for the vector order . at lower temperature
, the membrane can condense into an hexatic phase in which there is quasi - long - range 6-fold bond - angle order or into a smectic - c phase in which molecules tilt relative to the surface normal . to describe sm - c ( vector ) and/or hexatic order
, we introduce at each point @xmath37 on the membrane a unit vector @xmath67 in the tangent plane of the membrane . for sm - c order , @xmath67 is a true vector , invariant under rotations of @xmath68 ( p is an integer ) about the unit surface normal @xmath69 erected at @xmath37 . for hexatic order , rotations of @xmath67 by @xmath70 about @xmath69 lead to physically equivalent states .
more generally , we consider `` @xmath0-atic '' order in which rotations of @xmath67 by @xmath71 produce physically equivalent states .
a two - dimensional nematic with an in - plane symmetric traceless tensor order parameter is an example of a @xmath11-atic .
although we know of no physical realizations of other @xmath0-atics yet , we find it instructive to consider how the development of such order affects morphological changes in spherical vesicles .
to describe tangent - plane @xmath0-atic order , we introduce orthonormal unit vectors @xmath72 and @xmath73 at each point on the membrane .
@xmath74 defines a local angle @xmath75 .
@xmath0-atic order is then described by the order parameter @xmath76 $ ] , which can be related to the @xmath0-th rank symmetric traceless tensor constructed from the unit vector @xmath77 .
the @xmath0-th rank symmetric traceless tensors are the @xmath0-th rank spherical tensors .
they are listed in table [ n - tensor ] for @xmath78 and 6 . in general
, there are @xmath79 components in the @xmath0-th rank spherical tensor @xmath80 since @xmath81 can be either @xmath16 or @xmath11 . by permutational symmetry , there are only @xmath82 possible independent components .
however , there are @xmath83 additional traceless conditions .
hence , there are only @xmath11 independent components in @xmath80 .
the linear combination of these @xmath0-th rank spherical tensors @xmath84 where @xmath85 is the number of @xmath86 in @xmath87 becomes @xmath88 where @xmath89 and @xmath90 .
thus @xmath0-atic order parameter is described by @xmath91 note that since @xmath92 depends on the choice of orthonormal vectors @xmath72 and @xmath73 , the order parameter @xmath93 does as well .
this means that any spatial derivatives in a phenomenological hamiltonian for @xmath9 must be covariant derivatives .
we now describe the hamiltonian of the membrane with orientational order in a reparametrization - invariant way . for a flat membrane
the vector hamiltonian corresponds to the usual xy model @xmath94 using the fact that any spatial derivatives must be covariant derivatives and @xmath95 is the tangential component of @xmath96 , one can show that the only possible free energy term for @xmath77 which respects this @xmath97 symmetry is @xmath98 the dimensionless coupling constant @xmath99 is the vector - order stiffness and measures the strength of the coupling between the orientations of neighboring vectors .
we then generalize this xy - like hamiltonian for the vector order to the @xmath0-atic hamiltonian using the @xmath0-th rank spherical tensors introduced in table [ n - tensor ] .
in general , the @xmath0-atic hamiltonian can be written as @xmath100 where @xmath4 is the @xmath0-atic rigidity . using the spherical tensors in table 1 , we find @xmath101 and @xmath102 for @xmath103 .
hence , the hamiltonian for the @xmath0-atic order becomes @xmath104 in this form of @xmath105 , we neglected all terms that are irrelevant at large distance by power counting .
other terms such as @xmath106 which couple @xmath77 to the principal directions of curvature of the membrane are not invariant under the global rotation by @xmath24 of @xmath77 , and are irrelevant at large distance .
thus , this free energy has a full @xmath107 rotational symmetry .
this is similar to the fact a 2-dimensional crystal with hexagonal or triangular structure has isotropic elastic properties at large distance scale . for @xmath0-atics with @xmath108 ,
there is only one elastic constant @xmath4 . for @xmath18 or @xmath20
, there are in general two elastic constants . for simplicity
, we will consider the single elastic constant approximation for all @xmath0-atics . to describe this hamiltonian in terms of a local angular order parameter @xmath109 , at each point
we introduce two orthonormal vectors @xmath110 tangent to the membrane .
this is equivalent to introduce a 2-_bein _ @xmath111 compatible with the induced metric @xmath39 . in components ,
@xmath112 and the orthonormality @xmath113 implies @xmath114 the angular order parameter is frustrated by the rotation of tangent vectors that occurs under parallel transport on a curved surface .
the amount of frustration is given by the gauge field @xmath115 , _ i.e. _ the covariant derivative of @xmath116 in direction @xmath117 defines the gauge field @xmath115 . under parallel transport in direction @xmath118 , each @xmath116
is rotated by an angle @xmath119 .
thus the gauge field @xmath115 is defined by @xmath120 where @xmath121 is the antisymmetric tensor with @xmath122 and @xmath123 is called the spin - connection and describes how the basis vector @xmath116 rotates under parallel transport according to the gaussian curvature @xmath56 of the surface .
in fact , @xmath115 is related to @xmath56 .
the curl of the gauge field @xmath115 is the gaussian curvature ; @xmath124 where @xmath125 is the antisymmetric tensor defined via @xmath126 the covariant derivative of @xmath127 writes @xmath128 where @xmath129 satisfying @xmath130
. then the @xmath0-atic hamiltonian writes @xmath131 this form of the free energy is invariant under local transformations @xmath132 , @xmath133 .
this gauge invariance corresponds to a local rotation of the reference frame @xmath116 . in terms of the complex order parameter @xmath9 ,
the hamiltonian for @xmath0-atic order becomes @xmath134 where @xmath135 . in this description ,
all @xmath0-atics have the same long - wavelength elastic energy .
their properties can differ however , because their topological excitations are characterized by different winding numbers . since we assume that in the disordered state the membrane forms a spherical shape and there is no topological deformation of the membrane shape , order in @xmath9 is necessarily accompanied by the topological excitations called vortices . as we can see in ref .
@xcite , the total winding vorticity of a vector field on a closed surface with @xmath136 handles must be @xmath137 .
thus , a vector field on the surface of a sphere @xmath138 has total vorticity 2 .
the minimum winding number disclination for an @xmath0-atic is @xmath8 .
the energy of an individual disclination on both flat and curved surfaces is proportional to the square of its winding number .
it is therefore always favorable to form disclinations with the lowest possible winding number .
in addition , disclinations with the same sign repel each other .
these considerations imply that the ground state of a sphere with surface @xmath0-atic order will have @xmath7 maximally separated disclinations of winding number @xmath8 .
in this section , we present that in a certain region in the parameter space @xmath139 there is an effective kt transition .
ovrut and thomas have discussed the structure of the kt transition of a vortex - monopole coulomb gas on a rigid sphere and have shown it is the same as in the planar case . they have shown the kt transition temperature on a rigid sphere is the same as that on the euclidean plane , _
i.e. _ @xmath140 , where @xmath141 is the square of the vortex charge in their discription @xcite .
recently , we have extended the study of the structure of the kt transition to the hexatic hamiltonian on the fluctuating spherical surfaces and have shown the hexatic hamiltonian undergoes the effective kt transition at the same tenperature as in the planar case @xcite . following the procedure in ref .
@xcite , we generalize a study of the kt transition to the @xmath0-atic hamiltonian and present the region in the parameter space @xmath139 where the kt transition exists . from the previous sections , the @xmath0-atic hamiltonian on the fluctuating surfaces writes @xmath142
; @xmath143 we investigate the effect of thermal shape fluctuations of a genus zero surface on the kt transition in the low temperature limit @xmath144 . in this limit
, we can parametrize the surface by its radius vector as a function of standard polar coordinates @xmath145 ; @xmath146 where @xmath147 is the radial unit vector and @xmath148 measures deviation from sphericity with radius @xmath149 .
this parametrization is a `` normal gauge '' . to make the equations simple , we map this parametrization onto the unit sphere parametrization with @xmath150
. since we are interested in the limit @xmath144 and in this limit @xmath151 dominates , we first minimize @xmath151 over the shape fluctuation field @xmath152 which gives @xmath153 and then we minimize @xmath105 over @xmath109 with @xmath154 and find @xmath155 where the superscript@xmath156 stands for the rigid sphere with @xmath154 . in ref .
@xcite , lubensky and prost show that in the ground state @xmath7 disclinations of strength @xmath24 are arranged at the vertices of polyhedron inscribed in the sphere .
a disclination at @xmath157 with strength @xmath158 gives rise to a singular contribution @xmath159 to @xmath160 satisfying @xmath161 where @xmath162 is a contour enclosing @xmath163 .
thus , in general @xmath164 where @xmath165 is nonsingular , @xmath166 , and @xmath167 where @xmath168 where is the disclination density in the ground state and @xmath169 s are the coordinates of the vertices of icosahedron .
since @xmath170 satisfies @xmath171 , it is divergence - less and purely transverse .
accordingly @xmath170 can be written in terms of the curl of scalar fields and by applying the operator @xmath172 to @xmath173 , we find these scalar fields to be related to the gaussian curvature @xmath174 of the rigid sphere and the ground state disclination density on the rigid sphere , @xmath175 where @xmath174 is a gaussian curvature of the rigid sphere and @xmath176 is the disclination density in the ground state . now taking into account the bond angle fluctuations around @xmath160 and the shape fluctuations around the sphere , @xmath177 the full hamiltonian writes @xmath178 @xmath179 the angle fluctuation field @xmath180 can also have disclinations of strength @xmath181 where @xmath85 is an integer , due to the thermal fluctuation @xcite .
thus , @xmath182 can be decomposed into singular and nonsingular parts ; @xmath183 where @xmath184 is nonsingular , @xmath185 and @xmath186 where @xmath187 is the thermally - excited disclination density with disclinations of strength @xmath158 at @xmath169 .
the vector @xmath188 can always be chosen so that it is purely transverse , _
i.e. _ @xmath189 . in the hexatic hamiltonian
, @xmath182 always occurs in the combination @xmath190 . the spin - connection @xmath191 can and will in general have both a longitudinal and a transverse component . however , one can always redefine @xmath184 to include the longitudinal part of @xmath191
. this amounts to choosing locally rotated orthonormal vectors @xmath192 and @xmath193 so that @xmath194 .
thus we may take both @xmath188 and @xmath191 to be transverse and the hexatic hamiltonian @xmath195 can be decomposed into a regular longitudinal part , @xmath196 and a transverse part , @xmath197 where the cross term @xmath198 is dropped since @xmath199 .
it costs an energy @xmath200 to create the core of a disclination of strength @xmath85 .
( we assume for the moment that the core energies of the positive and negative disclinations are the same .
see , however , refs .
thus , partition sums should be weighted by a factor @xmath201 for each disclination of strength @xmath85 . since @xmath202 , we may at low temperature
restrict our attention to configurations in which only configurations of strength @xmath203 appear .
let @xmath204 be the number of disclinations of strength @xmath203 and let @xmath205 be the coordinate of the core of the disclination with strength @xmath203 labeled by @xmath206 .
the hexatic membrane partition function can then be written as @xmath207 where @xmath208 , and @xmath209 .
@xmath210 depends on all of the disclination coordinates @xmath211 where @xmath212 and @xmath213 is the sum over all possible disclination distribution with the topological constraint @xcite ; @xmath214 where @xmath215 is a molecular solid angle .
the kronecker factor @xmath216 in @xmath213 imposes the topological constraint that the total disclination strength on a sphere is 2 since with @xmath217 we have @xmath7 ground state disclinations with the strength @xmath8 giving the total disclination strength @xmath218 . the hexatic model of eq .
( [ eq : partition ] ) can easily be converted to a coulomb gas model using @xmath219 which follows from eq .
( [ curl - spin ] ) and eq .
( [ eq : vortex ] ) where @xmath220 is the deviation of the gaussian curvature from the rigid sphere .
the quantity @xmath221 is a `` charge '' density with contributions arising both from disclinations and from gaussian curvature
. equation ( [ eq : curl - vel ] ) implies @xmath222 where we used @xmath223 and @xmath224 is the laplacian on a surface with metric tensor @xmath46 acting on a scalar .
recall [ eq .
( [ antisym ] ) ] that @xmath225 rotates a vector by @xmath226 so that @xmath227 is perpendicular to @xmath228 and is thus manifestly transverse .
using eq .
( [ eq : va - aa ] ) in eq .
( [ eq : h - perp ] ) , we obtain @xmath229 where @xmath230 is the coulomb hamiltonian associated with the charge @xmath231 .
since the longitudinal variable @xmath184 appears only quadratically in @xmath232 , the trace over @xmath184 can be done directly giving the liouville action @xcite arising from the conformal anomaly ; @xmath233 where @xmath234 the coulomb gas partition function can thus be written @xmath235 the coulomb gas model can be converted following standard procedures into a sine - gordon model .
the first step is to carry out a hubbard - stratonovich transformation on @xmath236 : @xmath237 where the liouville factor @xmath238 is needed to ensure that @xmath239 be one when @xmath240 . inserting this in eq .
( [ eq : tr - hs ] ) , we obtain @xmath241 where @xmath242 the only dependence on disclinations is now in the term linear in @xmath0 .
thus to carry out @xmath213 , we need only to evaluate @xmath243 -\omega \ } } \right)^{n_{+ } } \left ( y \int\frac{d\omega}{a^{2 } } e^{-i \ { 2\pi [ \phi(\omega)/n ] -\omega \ } } \right)^{n_{- } } \nonumber \\ & = & \int \frac{d\omega}{2\pi } e^{(2y / a^{2})\int d\omega \cos[2\pi(\phi / n)-\omega]}.\end{aligned}\ ] ] thus @xmath244 } e^{-i\int d\omega \phi \delta k}. \label{eq : tr - sg}\ ] ] we can now change variables , letting @xmath245 .
the term linear in the gaussian curvature then becomes @xmath246 where we used @xmath247 .
the integral over @xmath248 in eq .
( [ eq : tr - sg ] ) is now trivial , and dropping the prime we obtain @xmath249 where @xmath250 is the sine - gordon action on a fluctuating surface of genus zero .
the first two terms of this action are the gradient and cosine energies present on a rigid sphere .
the final term provides the principal coupling between @xmath251 and fluctuations in the metric .
it is analogous to the dilaton coupling @xcite of string theory though here the coupling constant is imaginary rather than real .
note that the liouville action is not explicitly present in eq .
( [ eq : par - sg ] ) . in the regime @xmath144 , we can truncate the higher order terms in @xmath152 . in the normal gauge ,
the partition function becomes @xmath252\end{aligned}\ ] ] where @xmath253 and we used @xmath254 . to lowest order in @xmath152 , the shape fluctuation field @xmath152 is linearly coupled to the scalar field @xmath251 which is the conjugate field to the disclinations . in ref .
@xcite , we have shown the similar coupling in the fluctuating flat membrane is quadratic in the shape fluctuation field .
integrating over the shape fluctuation field @xmath152 gives the effective hamiltonian for the conjugate field to the disclinations @xmath255\end{aligned}\ ] ] with @xmath256 .
this is the massive sine - gordon theory .
the shape fluctuations induce the mass term for @xmath251 field and screen the coulombic interaction between the disclinations giving the yukawa interaction between them .
this partition function is equivalent to that of the yukawa gas hamiltonian on the rigid sphere with radius @xmath149 ; @xmath257 where @xmath258 where @xmath259 is the legendre polynomial with degree @xmath260 and @xmath261 is the angle between two disclinations at @xmath169 and @xmath262 . for @xmath263 , degree of legendre polynomial , @xmath264 , is real and the length scale introduced by @xmath265 is larger than the system size , @xmath266 after recovering the original length scale by mapping the unit sphere back to the sphere with the radius @xmath149 . on the other hand ,
if @xmath267 , @xmath264 is a complex number , @xmath268 where @xmath269 and @xmath270 .
the length scale introduced by @xmath271 may be interpreted as the debye screening length arising from shape fluctuations .
the interaction energy between two disclinations @xmath206 and @xmath272 at positions @xmath169 and @xmath262 with strength @xmath158 and @xmath273 is given by @xmath274 where @xmath275 is the chordal distance between two disclinations on the sphere with radius @xmath149 .
the interaction @xmath276 has the following limiting forms : @xmath277 following the analogy of the 2d coulomb gas , when the screening length is much larger than the system size , @xmath278 , the induced mass term arising from shape fluctuations is irrelavent for the kt transition and the kt transition temperature is given by @xmath279 for @xmath280 .
however , for @xmath281 , the screening length is shorter than the system size , @xmath282 , and the mass term is relavent for the kt transition and changes the universality class of the system .
there is no kt transition at non - zero temperature .
the disclinations are always unbound at non - zero temperature and the kt transition temperature vanishes .
thus , when @xmath6 , the effect of shape fluctuations is irrelavent and the kt transition occurs at the finite temperature .
for @xmath6 , there exists a kt transition at the finite temperature @xmath284 . to describe the phase transition and the corresponding shape changes
, we introduce the magnitude of the complex @xmath0-atic order parameter as in @xmath285 and consider the simplest long - wavelength landau - ginzburg hamiltonian for the @xmath0-atic order parameter @xmath9 ; @xmath286 this landau - ginzburg @xmath0-atic hamiltonian is similar to the landau - ginzburg hamiltonian for a superconductor in an external magnetic field , @xmath287,\ ] ] where @xmath288 .
both have a complex order parameter @xmath9 with covariant derivatives providing a coupling between @xmath9 and a vector potential " @xmath289 or @xmath290 . in a magnetic field ,
the superconductor can undergo a second order mean - field transition from a normal metal to the abrikosov vortex lattice phase with a finite density of vortices determined energetically by temperature and the magnetic field @xmath291 .
the magnetic field is conjugate to the vortex number @xmath292 since @xmath293 , where @xmath294 is the length of the sample along @xmath291 and @xmath295 is the flux quantum . on a closed surface with @xmath0-atic order
, there is a second - order mean - field transition to a state with vortex number determined by topology rather than conjugate external field .
thus , the mean - field @xmath0-atic transition on a closed surface is analogous to the transition to an abrikosov phase with a fixed number of vortices rather than fixed field conjugate to vortex number .
however , a meissner phase with zero vortices does not exist because we do not have an analogue for magnetic intensity @xmath291 , rather topology fixes vortex number .
the complete hamiltonian for @xmath0-atic order on a deformable surface is @xmath296 .
the field @xmath148 in @xmath297 can be expanded in spherical harmonics .
any isotropic change in @xmath298 can be described by @xmath299 .
in addition , uniform translations , which change neither the shape nor the energy of the vesicle , correspond to distortions in @xmath152 with @xmath300 and can be absorbed by fixing the center of mass of the vesicle .
these considerations imply that @xmath152 will have no @xmath301 or @xmath300 components in spherical harmonic expansion : @xmath302 the shape and size of the vesicle are determined entirely by the parameters @xmath299 and @xmath303 . the reduced tensors @xmath304 and @xmath305 do not depend on @xmath299 .
therefore , @xmath299 can be expressed as a function of @xmath306 and @xmath303 via the relation @xmath307
\nonumber \\ & = & 4\pi r_{0}^{2 } \left [ 1 + \frac{1}{2\cdot 4\pi}\sum_{l=2}^{\infty } \sum_{m}(l(l+1)+2)|\rho_{lm}|^{2 } + o(\rho^{4 } ) \right ] .\end{aligned}\ ] ] in the disordered phase above @xmath283 , @xmath308 and @xmath309 .
we will use the hamiltonian @xmath310 expressed in terms of reduced parameters and the constant area @xmath306 in our calculations of shape changes below the second - order disordered - to-@xmath0-atic transition .
using the spherical polar coordinates with origin at the center of mass of the vesicle , the metric tensor @xmath46 can be written as @xmath311 and the curvature tensor @xmath50 is @xmath312 where @xmath313 .
the hamiltonian is a functional of @xmath314 and @xmath148 at the fixed area @xmath306 . to find the equilibrium form of @xmath314 and @xmath148
, we need to minimize @xmath310 over @xmath314 and @xmath148 .
there is a considerable simplification if we restrict our attention to the neighborhood of the transition temperature @xmath283 .
near @xmath283 , to order @xmath315 , the variation of the curvature energy is of order @xmath316 .
when this is comparable to the landau - ginzburg free energy for @xmath9 , we only need to keep couplings of order @xmath317 . to this order
, the landau - ginzburg hamiltonian for @xmath9 becomes @xmath318 where @xmath319 and @xmath320 . in this formula , @xmath321 has been evaluated at @xmath322 , @xmath323 and @xmath324 . in terms of @xmath152 ,
the curvature energy can be written as @xmath325 where @xmath326 hence , to order @xmath315 , we can write @xmath310 as @xmath327 with @xmath328 which does not depend on @xmath148 and @xmath329 which contains @xmath148 .
this can be rewritten as @xmath330 with @xmath331 @xmath332 minimizing over @xmath148 , we obtain @xmath333 where @xmath334 satisfies @xmath335 and @xmath336 is the dirac delta function in spherical harmonics space with @xmath337 . by substituting this for @xmath148 into @xmath310 , we have @xmath338 where @xmath339 before proceeding with our analysis of the @xmath0-atic transition on a deformable sphere , it is useful to recall abrikosov s calculation of the transition to the vortex state .
the first step is to calculate the eigenfunctions of the harmonic part of @xmath340 when @xmath341 .
these can be divided into highly degenerate sets separated by an energy gap , @xmath342 . in the landau gauge with @xmath343 ,
the eigenfunctions in the lowest energy manifold are @xmath344 where @xmath345 .
the order parameter @xmath346 of the ordered state is expressed as a linear combination @xmath347 , where the complex parameters @xmath348 are determined by minimization of @xmath340 . with this analogy in mind , we will follow abrikosov s treatment of the superconducting transition near @xmath349 to study the development of @xmath0-atic order on a sphere .
we first diagonalize @xmath310 in the harmonic level when @xmath308 and @xmath350 , @xmath351 , that is , we determine the functions @xmath9 that satisfy @xmath352 , for @xmath308 , @xmath353 .
nonlinearities arise from the @xmath354 term in landau - ginzburg hamiltonian for @xmath9 and from the fact that @xmath115 and @xmath46 depend nonlinearly on the deviation from the ideal spherical shape . to minimize @xmath355
, first we seek the lowest energy configurations of the operators corresponding to the harmonic terms only .
then , we linearly combine these eigenfunctions to get the function which has the lowest energy for @xmath355 . that is , the manifold of the lowest energy eigenstates of the harmonic terms is highly degenerate .
nonlinear terms pick out the lowest energy state which is the linear combination of the degenerate eigenstates of the harmonic terms and resolve the degeneracy .
we divide @xmath355 into the harmonic part , @xmath356 , and the nonlinear part , @xmath357 : @xmath358 the differential operator for @xmath356 is @xmath359 where @xmath360 in the spherical polar representation , @xmath361 has the form @xmath362.\ ] ] one possible form of the eigenfunctions of @xmath361 has the structure of @xmath363 .
@xmath364 the complete spectrum and the eigenfunctions are derived in the appendix . introducing the projection representation ,
we can reparametrize the sphere using a @xmath365 coordinate system defined by @xmath366 the inverse transformation gives @xmath367 the corresponding metric tensor , spin connection , and the differential operator , @xmath361 , in this representation can be written as follows @xmath368 @xmath369 @xmath370.\ ] ] the eigenfunction @xmath363 can be written as @xmath371 this eigenfunction has vortices at @xmath372 and @xmath373 . now
, the vortices can be arbitrarily placed by multiplying @xmath374 which is a function of @xmath375 alone . from the differential operator @xmath361 , we see @xmath376
thus , the full eigenfunctions of @xmath361 are the product of @xmath377 and the arbitrary function @xmath374 . in particular , for @xmath18
, we have @xmath378 by taking @xmath379 , @xmath380 which means that @xmath9 has the vortices at @xmath381 and @xmath382 and that @xmath9 goes to @xmath383 as @xmath375 goes 0 , @xmath9 goes to @xmath384 as @xmath375 goes to @xmath385 .
this means there is an enormous degeneracy in the lowest energy manifold of the harmonic terms . in the lowest energy manifold of the harmonic terms , @xmath9 will have exactly @xmath7 zeros specifying the vortex positions , each of strength @xmath8 . in view of this vortex strength and distribution ,
we choose @xmath374 as @xmath386 where @xmath387 are arbitrary and @xmath388 if @xmath389 . by minimizing @xmath355 with the degenerate eigenfunctions of @xmath361 ,
@xmath390 the non - linear terms pick out the set @xmath391 that gives the lowest energy for @xmath355 .
this set @xmath392 also gives @xmath393 , vortex positions , and the shape change @xmath298 of the membrane . in terms of
spherical coordinates , @xmath394 and @xmath395 where @xmath396 , @xmath397 specifies the position of the @xmath206-th zero of @xmath9 , and @xmath398 is the normalization factor which makes @xmath399 since @xmath9 of eq .
( [ psi - eigen ] ) is the most general function in the lowest energy manifold , the order parameter and vesicle shape just below the transition temperature are obtained by minimizing @xmath355 over @xmath400 and the positions of zeros .
inserting this expression for @xmath9 , the free energy becomes @xmath401 where @xmath402 and @xmath403 \nonumber \\ & \equiv & \psi_{0}^{2}\phi(\omega).\end{aligned}\ ] ] minimization over @xmath400 leads to @xmath404/u[\{\omega_{i}\}]$ ] and the effective free energy density @xmath405 = -\frac{1}{2}\frac{r^{2}[\{\omega_{i}\ } ] } { u[\{\omega_{i}\}]},\ ] ] depending only on the positions @xmath406 of the zeros of @xmath407 where @xmath408 & = & ( r - r_{c})\int d\omega | p(\omega ) |^{2 } \\ u[\{\omega_{i}\ } ] & = & u\int d\omega | p(\omega ) |^{4 } - \frac{4\pi } { \kappa{\cal a}}\sum_{l=2}^{\infty}\sum_{m =- l}^{l } \frac{|\tilde{\phi}_{lm}|^{2}}{(l-1)^{2}(l+2)^{2}}.\end{aligned}\ ] ] the next step is to minimize @xmath409 $ ] over @xmath169 .
shape changes are described by @xmath410 where @xmath411 with @xmath406 that minimize @xmath409 $ ] . in the disordered phase above the transition temperature , @xmath412 .
hence , @xmath308 and @xmath323 .
the vesicle shape is spherical . in the ordered phase below the transition temperature
, we have been able to evaluate @xmath409 $ ] analytically for @xmath18 and @xmath20 . for these two cases ,
we find as expected that the minimum energy configurations are , respectively , those with zeros of @xmath314 at opposite poles and at the vertices of a tetrahedron .
the shape function @xmath413 associated with @xmath0-atic order is proportional to @xmath414 to the order of our calculations . in general , the legendre decomposition of @xmath415 will contain legendre polynomials of order @xmath7 . for @xmath18 and @xmath20
, we find @xmath416 \nonumber \\ & & + \frac{4}{3645}\sqrt{\pi } \left [ y_{4}^{0}(\omega ) + \sqrt{\frac{10}{7}}(y_{4}^{3}(\omega ) -y_{4}^{-3}(\omega ) ) \right ] .
\label{shape2}\end{aligned}\ ] ] figure [ fig12 ] shows the shapes described by these functions .
the transformations from the initial spherical shape to distorted shapes occur continuously .
our calculations for the shape are valid to order @xmath417 .
as temperature is lowered , higher - order terms in @xmath417 and higher - order spherical harmonics are needed to describe the equilibrium shape . in ref .
@xcite , a variational function for @xmath9 and spherical harmonics up to order @xmath418 were used to calculate the shape for @xmath18 for temperatures well below the transition .
we use symmetry considerations to treat @xmath419 and @xmath23 . for @xmath21 and @xmath19 , we expect the zeros of @xmath314 to lie , respectively , at the vertices of an octahedron and an icosahedron .
we find the shape functions for @xmath21 and @xmath19 @xmath420 \nonumber \\ & & + \frac{1}{60060}\sqrt{13\pi } \left [ y_{6}^{0}(\omega ) - \sqrt{\frac{7}{2}}(y_{6}^{4}(\omega ) + y_{6}^{-4}(\omega ) ) \right ] , \\
\bar{\rho}^{(6)}(\omega ) & = & \frac{15488}{275559375 } \sqrt{13\pi } \left [ y_{6}^{0}(\omega ) + \sqrt{\frac{7}{11 } } ( y_{6}^{5}(\omega ) - y_{6}^{-5}(\omega ) ) \right ] \nonumber \\ & & + \frac{256}{692803125}\sqrt{21\pi}\left [ y_{10}^{0}(\omega ) + \sqrt{\frac{33}{13 } } ( y_{10}^{5}(\omega ) - y_{10}^{-5}(\omega ) ) + \frac{187}{247 } ( y_{10}^{10}(\omega ) + y_{10}^{-10}(\omega ) ) \right ] \nonumber \\ & & - \frac{512}{6834953125 } \sqrt{\pi}\left [ y_{12}^{0}(\omega ) - \frac{1}{3}\sqrt{\frac{286}{119 } } ( y_{12}^{5}(\omega ) - y_{12}^{-5}(\omega ) ) + \frac{247}{357 } ( y_{12}^{10}(\omega ) + y_{12}^{-10}(\omega ) ) \right ] \end{aligned}\ ] ] figure [ fig36 ] shows the shapes described by these functions . for @xmath421 ,
following the calculations of palffy - muhoray @xcite , we expect the zeros of @xmath314 to lie at the vertices of a distorted cube obtained by rotating its top face about its fourfold axis by @xmath422 and compressing opposite faces . for this case , we minimized the energy over a single parameter determining the separation between opposite rotated faces and obtain @xmath423 .
\end{aligned}\ ] ] figure [ fig4 ] shows the shape for @xmath421 .
we have presented the general @xmath0-atic hamiltonian in terms of the @xmath0-th rank spherical tensors and have shown there exists an effective kt transition for @xmath6 .
also we have presented here an analysis of the mean - field transition to @xmath0-atic order on a fixed - area surface of genus zero and the corresponding continuous shape changes below the kt transition temperature in this parameter region .
we are really dealing with two kinds of order : @xmath0-atic order and the positional order of vortices . in mean - field theory , these two kinds of order develop simultaneously .
our analysis is very similar to that of abrikosov for the transition from a normal metal to a vortex lattice in a type ii superconductor at @xmath349 and ignores the effects of fluctuations . in superconductors ,
fluctuations drive the normal - to - superconducting transition first order in a nonvanishing magnetic field @xcite . in the abrikosov phase ,
fluctuations of the vortex lattice destroy superconductivity but not long - range periodic order . in two - dimensions ,
screening of vortices drives the kosterlitz - thouless transition in an infinite superconductor in zero field to zero
. both of the above effects may be important for @xmath0-atic order on a sphere .
we have been able to find in the parameter region @xmath424 , the effect of shape fluctuations is irrelavent and will not lead to qualitative changes in our results . however , outside this region , _
i.e. _ for @xmath3 , shape fluctuations are relavent .
the interaction between the vortices is screened to become the massive sine - gordon theory and hence is equivalent to a neutral yukawa gas on a sphere .
consequently , this screening effect derives the kosterlitz - thouless transition temperature to zero .
the vortices are always unbound for a non - zero temperature . in ref .
@xcite , evans has discussed the shape changes of the @xmath0-atic hamiltonian in the opposite limit ( high temperature limit ) to the mean - field limit described in this paper . using the lowest landau level approximation ,
the effect of thermal fluctuations are discussed .
although we agree with him in that the amplitudes of deformations have zero thermal average , since the lowest landau levels include only the ground state vortices we are not sure if this approximation is valid in the high temperature phase where thermally excited vortex - antivortex pairs are always unbound .
thus we believe the effects of fluctuations to the shape changes in the disordered state deserve further investigation .
we present the complete spectrum of the differential operator , @xmath361 , for the harmonic part , @xmath356 . the operator @xmath361 in the spherical polar representation , @xmath426 is similar to the differential operator @xmath427 for the superconductivity where @xmath428 and @xmath429 the schrdinger equation on a sphere in the field of a dirac magnetic monopole of strength @xmath12 at the origin plus an infinitely long and thin solenoid carrying flux @xmath430 along the @xmath375 axis is described by @xmath431 with @xmath432 the monopole has been chosen to have a dirac string singularity along the negative @xmath375 axis . setting @xmath433 , @xmath434 , we find @xmath435 and @xmath427 is exactly identical to @xmath361 .
hence , the differential operator , @xmath361 , is equivalent to the hamiltonian describing the field of a dirac magnetic monopole of strength @xmath0 at the origin plus an infinitely long and thin solenoid carrying flux @xmath436 along the positive @xmath375 axis . following the method in ref .
@xcite , we define the formal differential operator @xmath437 with components @xmath438 \\ j_{z } & = & -i(\partial_{\phi } + in ) -n\end{aligned}\ ] ] which obeys the formal commutation relations @xmath439 = 0 \ ; ; \;\ ; [ j_{i},j_{j}]=i\varepsilon_{ijk}j_{k},\ ] ] for @xmath440 .
we find @xmath441 let us denote @xmath442 where we set @xmath443 p(\cos\theta ) \ ; ; \;\ ; m=0,\pm 1,\pm 2 , \cdots .\ ] ] then @xmath444 the eigenvalues of @xmath445 are found to be given by @xmath446 with @xmath447 the corresponding orthonormal eigenfunctions are @xmath448 where @xmath449 denote jacobi polynomials , and @xmath450 the complete spectrum , @xmath451 , of the operator @xmath361 is @xmath452 for @xmath453 , @xmath454 for @xmath455 , @xmath456 thus , @xmath457 the ground state energy is generated when @xmath458 , @xmath459 . @xmath460
the corresponding ground state orthonormal eigenfunction has the form @xmath461 which is the same as that in eq .
( [ psi - eigen ] ) . | we present the existence of the kosterlitz - thouless ( kt ) transition for @xmath0-atic tangent - plane order on a deformable spherical surface and investigate the development of quasi - long range @xmath0-atic order and the continuous shape changes of a deformable spherical surface below the kt transition in the low temperature ( mean field ) limit .
the @xmath0-atic order parameter @xmath1 $ ] describes , respectively , vector , nematic , and hexatic order for @xmath2 and 6 .
we derive a coulomb gas hamiltonian to describe it .
we then convert it into the sine - gordon hamiltonian and find a linear coupling between a scalar field and the gaussian curvature on the fluctuating spherical surface .
after integrating over the shape fluctuations , we obtain the massive sine - gordon hamiltonian , where the interaction between vortices is screened .
we find , for @xmath3 where @xmath4 and @xmath5 are @xmath0-atic and bending rigidity , respectively , the kt transition is supressed altogether . on the other hand ,
if @xmath6 , there is an effective kt transition . in the ordered phase
, tangent - plane @xmath0-atic order expels the gaussian curvature .
in addition , the total vorticity of orientational order on a surface of genus zero is two .
thus , the ordered phase of an @xmath0-atic on such a surface will have @xmath7 vortices of strength @xmath8
, @xmath7 zeros in its order parameter , and a nonspherical equilibrium shape .
our calculations are based on a phenomenological model with a gauge - like coupling between @xmath9 and curvature , and our analysis follows closely the abrikosov treatment of a type ii superconductor just below @xmath10 . |
unraveling the pattern of soft and collinear divergences in scattering amplitudes is a critical endeavor to advance our understanding of gauge theories in general and to assist in concrete computations for collider phenomenology , e.g. , in massless gauge theories such as quantum chromodynamics ( qcd ) .
these studies have a long history ( see e.g. the early review @xcite ) and have contributed to our knowledge of the universal infrared ( ir ) structure of gauge theory amplitudes .
based on the concepts of soft and collinear factorization , non - abelian exponentiation , and the study of collinear limits , significant information about scattering amplitudes is available to all orders in perturbation theory . for precision predictions at modern colliders , especially within qcd and
including higher order quantum corrections , these insights are of great practical importance @xcite .
quite generally , factorization implies the separation of scales in a given scattering reaction , i.e. the process - dependent hard scale @xmath9 from those governing the soft and collinear limit , defined for instance by the masses @xmath10 of the scattering particles with @xmath11 or by the regions of soft momenta .
note that the soft and collinear singularities of massless particles ( gauge bosons ) strictly require the definition of a regulator , which is conveniently performed in @xmath12 dimensions . as an immediate consequence of factorization ,
evolution equations emerge , which depend on the kinematics of the specific process and on the chosen regulator .
their solution leads to non - abelian exponentiation , a result which also arises from an effective field theory formulation based on the ultra - violet ( uv ) renormalization properties of effective operators and their anomalous dimensions ; see e.g. ref .
moreover , for scattering amplitudes in gauge theories , the underlying factorization imposes strong constraints on the anomalous dimensions and the all - order structure of the ir singularities @xcite . in the present paper
we will specialize our investigations in a number of ways .
first , we choose to work in the @xmath0 supersymmetric yang - mills ( sym ) theory , which is the simplest non - abelian gauge theory in four dimensions due to the vanishing of the four - dimensional @xmath13 function . in our study
, we are concerned with form factors and scattering amplitudes in this theory , which allows us to study their ir singularities without interference from uv divergences .
second , we will be working in the so - called planar limit and , for scattering amplitudes @xmath14 of @xmath1 external particles , we will assume color ordering .
our main focus is on the study of different kinematical regimes , i.e. scattering amplitudes of massless and massive particles and associated form factors , using different regulator schemes .
the general property of factorization prompts us to ask whether one can delineate a well - defined finite part of @xmath14 independent of the chosen ir regularization . while reasoning along these lines has already been employed in the derivation of radiative corrections for heavy - quark hadroproduction at two loops in qcd @xcite ( see also refs .
@xcite ) , this issue is more generally related to the important question whether physical observables in theories with massless particles are independent of the regulator by construction ; see e.g. ref .
; see e.g. the discussion in ref .
@xcite . to that end , in this paper we specifically compare dimensional and massive regularizations schemes for @xmath1-particle scattering amplitudes @xmath14 in @xmath0 sym theory , an ideal testing ground for these questions due to the simplicity of its loop expansion .
sym theory , ir - safe inclusive differential cross - sections were studied in ref .
] in - depth studies of the latter may lead to new insights for gauge theories with massive particles which will eventually also be of interest for applications in collider phenomenology .
let us start thus , for simplicity , with maximally - helicity - violating ( mhv ) scattering amplitudes .
factorization implies that the color - ordered amplitude @xmath15 of @xmath1 particles can be written as ( see e.g. ref .
@xcite ) @xmath16 here @xmath17 and @xmath18 are `` soft '' and `` jet '' functions , respectively , and @xmath19 is an ir - finite `` hard function '' . in general , @xmath20 and @xmath19 are vectors in a space of possible color structures , and @xmath17 is a matrix .
in the planar limit , @xmath17 is proportional to the identity matrix , and one can combine @xmath21 and @xmath22 into a product of `` wedge '' functions @xmath23 that depend only on two adjacent particles @xmath24 and @xmath25 of the color - ordered amplitude @xcite @xmath26 as will be detailed below , the factorization ( [ intro - factorization2 ] ) holds not only in dimensional regularization , but also in cases where masses are used to partially or fully regulate the ir divergences .
the wedge functions @xmath27 satisfy renormalization group equations which imply that they exponentiate .
the factorization ( [ intro - factorization2 ] ) fixes the hard function @xmath28 only up to finite pieces , but with a suitable definition of the wedge function , we suggest that @xmath29 can be used to define a regulator - independent finite part of the amplitude . for regulators that leave the external particles massless , such as dimensional regularization in @xmath30 dimensions or the common - mass higgs regulator described below , each wedge has half the ir divergences of a sudakov form factor @xmath31 ( see e.g. refs .
@xcite ) , so it is natural to define @xmath32 @xcite .
we show in this paper that , with this definition , @xmath33 is identical for both of these regulators through two - loop order .
we also analyze a refined version of the higgs regulator with differential masses , described below . in this case , the external particles have distinct masses , and so the wedge function can not simply be defined in terms of a sudakov form factor .
instead , we define the one - loop wedge function in terms of a certain ir - divergent triangle diagram , and then use extended dual conformal invariance to extend this to an all - loop expression for the sum of wedge functions . with this choice for the ir - divergent wedge function , we establish that the ir - finite hard function @xmath33 takes precisely the same form for the differential - mass higgs regulator as for the common - mass regulator .
we lack , however , an explicit operator definition for the wedge function in this case . for @xmath0 sym theory , the regulator - independent hard function @xmath34 takes the simple form @xmath35 due to the conjectured duality between the finite part of the mhv scattering amplitudes and the ( uv renormalized ) expectation value of certain cusped wilson loops ( see refs .
@xcite for reviews ) . in , @xmath36 is the cusp anomalous dimension @xcite , for which a prediction to all orders in the coupling constant @xmath37 exists @xcite , and @xmath38 and @xmath39 are kinematic - independent functions .
the amplitude is expressed as a function of the dual or region momenta @xmath40 , which are defined by p_i^ = x^_i - x_i+1^ with @xmath41 , and @xmath42 .
the first three terms on the r.h.s . of
, whose kinematical dependence is determined solely by the one - loop contribution @xmath43 , constitute the abdk / bds ansatz @xcite .
the _ a priori _ undetermined remainder function @xmath44 contains the only non - trivial , i.e. loop - dependent , kinematical dependence
. follows from a conformal ward identity for the dual wilson loop @xcite . the first term on the r.h.s .
of provides a particular solution to this ward identity .
the remainder function @xmath45 is the general homogeneous solution to the ward identity , and depends only on dual conformal cross - ratios , which take the form @xmath46 . due to the absence of dual conformal cross - ratios for @xmath47 and @xmath48 , the remainder functions @xmath49 and @xmath50 vanish and therefore the corresponding hard functions @xmath51 and @xmath52 are completely determined by their one - loop value and the kinematic - independent functions @xmath53 , and @xmath39 . for @xmath54 ,
dual conformal cross - ratios can be built , and the remainder function is known to be non - zero starting from two loops and @xmath55 external particles @xcite .
its higher - loop and higher - point form is under intense investigation ; see e.g. refs .
@xcite .
the planar mhv @xmath1-point amplitude for @xmath0 sym theory has been studied using dimensional regularization and also using an alternative massive ir regulator @xcite .
the latter is motivated by the ads / cft correspondence and consists of computing scattering amplitudes on the coulomb branch of @xmath0 sym theory , i.e. giving a non - trivial vacuum expectation value to some of the scalars .
one can achieve a situation where the propagators on the perimeter of any loop diagram are massive , thereby regulating the ir divergences .
the simplest case , the `` common mass higgs regulator '' in which only one mass @xmath56 is introduced , corresponds to the breaking of the @xmath57 gauge group to @xmath58 , with fields in the adjoint representation of @xmath59 remaining massless .
in the more general `` differential - mass higgs regulator '' , one breaks the gauge group further to @xmath60 , thereby introducing various masses @xmath61 , @xmath62 .
fields in the adjoint of the broken @xmath59 , which appear as external states in the scattering amplitudes , now have nonzero masses @xmath63 .
we use a decomposition of the one - loop mhv @xmath1-point amplitude into a sum of ir - divergent triangle diagrams and ir - finite six - dimensional box integrals to define the sum of one - loop wedge functions as , @xmath10 , and @xmath64 ; see . ] [ sumwedgesoneloop ] _
i=1^n ^ ( ) |_one - loop = _ i=1^n in the uniform small mass limit ( i.e. , @xmath65 , with @xmath66 fixed and @xmath67 ) .
the one - loop hard function @xmath68 is then expressed in terms of ir - finite quantities , and thus is manifestly regulator - independent .
a key point is that the massive regulator is closely connected to dual conformal symmetry .
the higgs masses can be interpreted within the ads / cft duality as the radial coordinates in a @xmath69-dual ads@xmath70 space .
while the isometries of this space yield the usual dual conformal transformations for zero masses , they define a different realization of this symmetry for finite masses , dubbed `` extended dual conformal symmetry '' @xcite .
since no further regulator is needed in the massive setup , the extended dual conformal symmetry is expected to be an exact symmetry of the planar amplitudes .
recently , it was shown that tree - level amplitudes on the coulomb branch of @xmath0 sym and also all cuts of planar loop amplitudes do indeed have this extended symmetry @xcite . together with the expected cut - constructibility of @xmath0 sym this
then proves the extended dual conformal symmetry property conjectured in ref .
@xcite .
what we wish to emphasize is that , while planar amplitudes have extended dual conformal symmetry , the wedge functions and regulator - independent hard functions separately do not .
this is not surprising , since extended dual conformal transformations act on the masses @xmath10 ( as well as the dual variables @xmath71 ) , whereas the hard function is , by definition , independent of the masses in the uniform small mass limit .
nevertheless , extended dual conformal symmetry can be used to determine the all - loop structure of the ir divergences of scattering amplitudes in the case of the differential - mass higgs regulator . assuming the validity of for mhv scattering amplitudes , together with , we obtain the following expression for the ir - divergent pieces in the differential - mass setup [ sumwedgesintro ] & & _ i=1^n ( ) + & & = _ i=1^n valid for uniform small masses . having deduced the form of the ir divergences of the amplitude for the differential - mass regulator
, we turn the argument around and use together with extended dual conformal symmetry to deduce the anomalous dual conformal ward identity , from which the all - loop result ( [ solution - ward - identity ] ) follows .
hence , a derivation of from first principles would constitute a proof of without having to rely on the scattering amplitude / wilson loop duality .
it would be very interesting to understand the origin of from a renormalization group approach .
a first step could be to find a suitable operator definition for the wedge function in the differential - mass regulator case .
we leave these questions for future work .
this paper is organized as follows . in section [ sect - bds ]
, we review the form of color - ordered mhv scattering amplitudes in planar @xmath0 sym in dimensional regularization and in the massive regularization of ref .
@xcite . in section [
sect - finite ] we discuss factorization and exponentiation properties of scattering amplitudes and form factors .
we propose a definition , involving sudakov form factors , for a regulator - independent hard function that can be computed from the ir - divergent scattering amplitudes .
we review the result for the form factors up to two loops in dimensional regularization , and compute the analogous quantities to two - loop order in the massive regularization .
we then show that the ( logarithm of the ) hard function defined earlier is the same in both cases . in section [ sect - diff ] ,
we discuss a more general differential - mass higgs regularization , and compute the ir - divergent terms of the one - loop amplitude in this regularization .
we then use extended dual conformal symmetry to derive the all - loop form of the ir - divergent terms , and discuss their relation to the dual conformal ward identity .
section [ sect - discussion ] contains our conclusions , and two appendices contain technical details used in the paper .
[ sect - bds ] in this section , we briefly review the form of color - ordered mhv amplitudes in planar @xmath0 sym theory in different regularization schemes . the all - loop - order @xmath1-point amplitude is given by the tree - level amplitude times a helicity - independent function @xmath72 , which we expand in the t hooft parameter [ expansion ] m_n = 1 + a^m_n^,a = g^2 n 8 ^2 ( 4 e^- ) ^ where @xmath73 .
loop - level amplitudes are uv - finite but suffer from ir divergences which can be regulated using either dimensional regularization in @xmath74 dimensions , or a higgs regulator in four dimensions .
we discuss each of these in turn . in dimensional regularization , the @xmath1-point amplitude takes the form @xcite [ bds ] m_n & = & a^_i=1^n ( ^2 ) ^ + & & + ( a)4 f_n^ ( ) + n f(a ) + c(a ) + r_n ( , a ) + ( ) . the momentum dependence of the amplitude is expressed in terms of dual variables @xmath71 defined via @xmath75 , where @xmath76 are the momenta of the external states ; we also define @xmath77 .
the terms on the first line of are ir - divergent and are specified in terms of the cusp and collinear anomalous dimensions @xcite [ defgamma ] ( a ) & = & _ = 1^ ^a^= _
= 1^ 4 f_0^a^= 4a - 4_2 a^2 + 22 _ 4 a^3 + ( a^4 ) , + [ defg ] _
0(a ) & = & _ = 1^_0^a^= _ = 1^ a^= - _ 3 a^2 + ( 4 _ 5 + _ 2 _ 3 ) a^3 + ( a^4 ) .
the terms on the second line of are ir - finite and are determined by the finite part of the one - loop amplitude [ finiteoneloop ] f_n^m_n^ + _
i=1^n ( ^2 ) ^ as well as the constants @xcite [ deff ] f(a ) & = & - a^= a^2 + ( - _ 6 - _ 3 ^ 2 ) a^3 + ( a^4 ) , + [ defc ] c(a ) & = & - a^2 + a^3 + ( a^4 ) , and a remainder function @xmath78 potentially contributing beginning at two loops . the original proposal by bern , dixon , and smirnov @xcite conjectured that holds with @xmath79 .
explicit calculations bore this out for @xmath47 ( through four loops ) @xcite and @xmath48 ( through two loops ) @xcite , but the two - loop calculation for @xmath55 @xcite revealed the necessity for a non - constant function @xmath80 .
explicit expressions for are given in ref .
@xcite . for @xmath47 and @xmath48
, they are [ deff4 ] f_4^&= & ^2 ( x^2_13 x^2_24 ) + , + [ deff5 ] f_5^&= & - _
i=1 ^ 5 ( ) ( ) + .
the four- , five- , and six - point functions have also been computed @xcite using the common - mass higgs regulator described in the introduction .
these amplitudes exhibit an exponentiation similar to which motivated the following analog for higgs - regulated @xmath1-point amplitudes @xcite [ bdshiggs ] _
n & = & _ i=1^n + & & + _ n^(1 ) ( ) + n ( a ) + ( a ) + _ n ( , a ) + ( m^2 )
. the terms on the first line of are ir - divergent .
the cusp anomalous dimension ( [ defgamma ] ) is independent of regularization scheme , but the analog of the collinear anomalous dimension is given by [ defgt ] _ 0(a ) = - _ 3 a^2 + ( _ 5 - _ 2 _ 3 ) a^3 + ( a^4 ) .
the terms on the second line of are ir - finite and are determined by the finite part of the one - loop amplitude _ n^m_n^ + _ i=1^n ^2 ( ) as well as the constants @xcite [ defft ] ( a ) & = & a^2 + o(a^3 ) , + [ defct ] ( a ) & = & - a^2 + o(a^3 ) . and a remainder function @xmath81 . as in the case of dimensional regularization
, the remainder function vanishes for four- and five - point amplitudes . for @xmath55 ,
it was shown @xcite that the two - loop remainder function @xmath82 in the higgs - regulated amplitude is precisely equal to its value @xmath83 in dimensional regularization , and this is expected to hold generally .
the one - loop amplitudes may be evaluated to show @xcite [ deff4 t ] _ 4^&= & ^2 ( x^2_13 x^2_24 ) + = f_4 ^ - , + [ deff5 t ] _ 5^&= & - _
i=1 ^ 5 ( ) ( ) + = f_5 ^ - , and more generally @xcite [ fft ] _
n^= f_n^- .
[ sect - finite ] comparing the known expressions for higgs - regulated amplitudes ( [ bdshiggs ] ) with those for dimensionally - regulated ones ( [ bds ] ) , one observes that the ir - finite parts of the amplitudes are equal in both regularizations , up to constants . in this section ,
we make the connection more precise by introducing a regulator - independent expression for the finite part of the amplitude . in a planar theory , the factorization ( see e.g. ref .
@xcite ) of color - ordered amplitudes takes the specific form @xcite [ eq : defmn ] m_n = h_n ( ) where @xmath84 is an ir - divergent `` wedge function '' depending only on @xmath85 and resulting from the exchange of soft gluons in the wedge between the @xmath86th and @xmath25th external particles , and @xmath87 is an ir - finite hard function . with a suitable definition for @xmath88 , we can use [ finite ] h_n = m_n - _ i=1^n ( ) to define the ir - finite part of the amplitude .
the forms of both @xmath72 and @xmath89 will depend on the specific regulator , but we will find that @xmath33 is regulator - independent . in dimensional regularization , the wedge function can be defined as the square root of the gluon form factor @xcite .
form factors in @xmath0 sym have been studied at strong coupling @xcite , at one loop @xcite and at two loops @xcite , while three - loop results can be inferred from the respective qcd computations @xcite using the principle of maximal transcendentality ; see e.g. ref .
@xcite . in @xmath0 sym we can equivalently use the form factor [ formfactor ] ( q^2 )
= j , p_i | _ ij ( q ) | i , p_i-1 for scalars @xmath90 coupling to the operator _
ij = with @xmath91 .
the operator @xmath92 belongs to the stress - energy multiplet of @xmath0 sym and is not uv renormalized .
this form factor has been computed to two loops in dimensional regularization @xcite 3.5 cm 3.5 cm 3.5 cm ( q^2 ) = 1 + a q^2 + a^2 ( q^2)^2 + where @xmath93 represent the scalar integrals shown in fig .
[ fig - massless ] .
( despite its apparent non - planarity , integral @xmath94 is actually leading order in the @xmath95 expansion as is clear from the corresponding double - line diagram in fig .
[ fig - doubleline ] . )
the explicit expressions for these integrals given in appendix [ app - integrals ] reveal that the form factor exponentiates to two - loop order @xmath96 + a^2 \left(\mu^2 \over q^2\right)^{2\eps } \left [ { \pi^2 \over 24 \eps^2 } + { \zeta_3\over 4\eps } + \co(\eps ) \right ] + \co(a^3)\,.\ ] ] can be promoted to all orders in perturbation theory since the momentum dependence of @xmath97 is governed by an evolution equation ( see e.g. refs .
@xcite ) . in @xmath12 dimensions ,
the following factorization ansatz holds : @xmath98 all dependence of @xmath97 on the hard momentum @xmath9 , which is taken to be euclidean ( @xmath99 in our mostly - plus metric convention ) here and in the sequel , rests inside the function @xmath100 .
the latter is finite in four dimensions and can be considered as a suitable continuation of the collinear anomalous dimension ( [ defg ] ) to @xmath12 dimensions .
@xmath101 on the other hand serves as a pure counterterm .
renormalization group invariance of @xmath97 yields @xmath102 i.e. , @xmath103 and @xmath104 are both governed by the cusp anomalous dimension @xmath36 of . the solutions of can be conveniently given with the help of the @xmath74-dimensional continuation @xmath105 of the t hooft parameter defined in @xmath106 which exhibits scale dependence on dimensional grounds and vanishes in the ir for @xmath12 with @xmath107 .
using and exploiting the fact that @xmath104 has no explicit scale dependence , which allows one to express it entirely through @xmath36 , one arrives at the following all - order expression for @xmath97 @xmath108 where the explicit solution of for the counterterm function @xmath104 has been used @xmath109 the double poles @xmath110 at each loop order are generated by the two @xmath111- and @xmath112-integrations over @xmath113 , while the single poles in @xmath114 arise from the outer @xmath112-integration over @xmath103 .
explicit computation , e.g. along the lines of refs .
@xcite , yields @xmath115 \left ( \mu^2 \over q^2 \right)^{\ell \eps}\ ] ] where the boundary condition for @xmath103 has been chosen as @xmath116 which is consistent with .
@xmath117 is given in and @xmath118 can be read off from as @xmath119 the exponentiation of proceeds trivially with the help of the boundary condition for @xmath97 in @xmath74 dimensions , i.e. , @xmath120 , which is implicit also in our choice for @xmath121 in .
note also that the all - order result for the form factor in applies literally to theories with less supersymmetry , e.g. , to qcd .
there , the coupling constant @xmath122 has to be read as the strong coupling @xmath123 continued to @xmath74-dimensions and the respective qcd expressions for the anomalous dimensions @xmath36 and @xmath117 are related to by the principle of maximal transcendentality .
moreover , @xmath121 admits a further decomposition @xcite into three terms : a universal ( spin - independent ) eikonal anomalous dimension , ( twice ) the coefficient of the @xmath124-term in the collinear evolution kernel , and a process - dependent term accounting for the running coupling in the coefficient function of the hard scattering .
the latter is proportional to the qcd @xmath13-function and is , of course , absent in @xmath125 sym .
we now introduce the wedge function as announced above . defining @xmath126
, we see from that [ eq : logw ] w(q^2 ) = a^ ( ^2 q^2 ) ^ , where [ defw ] w(a ) = a + ( a^3 ) and @xmath36 and @xmath117 are given in . with at our disposal , and using exponentiation of the @xmath1-point amplitude @xmath72 in , we can now define the finite remainder @xmath28 via [ eq : loghn ] h_n = ( a)4 f_n^ ( ) + n + c(a ) + r_n ( , a ) + ( ) thus the ir - divergent pieces of the wedge function remove all the ir divergences of the @xmath1-point amplitude . at one loop ,
gives [ hf ] h_n^= f_n^- allowing us to rewrite it in its final form as [ bdsfinite ] h_n = ( a)4 h_n^ ( ) + n d(a ) + c(a ) + r_n ( , a ) + ( ) where [ defd ] d(a ) = f(a ) - w(a ) + ( a ) = - a^2 + ( a^3 ) with @xmath39 given by .
using renormalization group arguments similar to the derivation of for the form factor @xmath97 , it is obvious that @xmath127 , @xmath128 , and therefore also @xmath33 in can be expressed to all orders via ( double-)integrals over the respective anomalous dimensions ; see e.g. refs . @xcite .
3.5 cm 3.5 cm 3.5 cm 3.5 cm now we turn to the study of the higgs - regulated @xmath0 sym form factor @xmath129 .
we will assume that it is given by the same scalar integrals as in fig .
[ fig - massless ] except that some of the internal legs are now massive [ deftphi ] ( q^2 ) = 1 + a q^2 + a^2 ( q^2)^2 + there are several different mass assignments for the two - loop integrals ( see fig .
[ fig - massive ] ) , as can be seen from the double - line representation in fig .
[ fig - doubleline ] .
we have computed these integrals ( see appendix [ app - integrals ] ) and they reveal that the higgs - regulated form factor exponentiates to two - loop order @xmath130 + a^2 \left [ { \pi^2 \over 12 } \log^2 \left ( \frac{q^2}{m^2 } \right ) + \zeta_3 \log \left ( \frac{q^2}{m^2 } \right ) + { \pi^2 \over 45 } + \co(m^2 ) \right ] + \co(a^3 ) \ , . { \nonumber}\end{aligned}\ ] ] the all - loop - order generalization of relies on the same factorization ansatz discussed before and the separation of scales , i.e. @xmath131 , so that the momentum dependence of @xmath97 is described by the evolution equation @xcite @xmath132 and , consistent with , the renormalization group equation for @xmath133 now reads @xmath134 where the limiting procedure @xmath135 indicates that we neglect any power suppressed terms of @xmath136 .
it is subject to the following solution @xmath137 with a non - vanishing boundary condition @xmath138 , since the ir sector has been altered in contrast to .
the solution to the higgs - regularized form factor @xmath139 then becomes @xmath140 where the integration range is naturally cut off in the ir at @xmath141 , i.e. at the mass scale set by the higgs regulator , and the function @xmath142 has been introduced to match the fixed - order computation in .
note that the evolution equation for the higgs - regulated form factor in @xmath0 sym depends only on the sum of @xmath138 and @xmath143 . to agree with the fixed - order computation in , we choose ( cf . ) ( 1,a ) + ( 1,a ) = - 2 _ 0(a ) leading to the solution of @xmath144 the exponentiation of requires further matching conditions for @xmath139 to be obtained from explicit @xmath145-loop computations .
a few comments are in order here .
first , matching to fixed - order computations could , in principle , also impose the condition @xmath146 , i.e. demand that the collinear anomalous dimensions coincide .
this would proceed at the expense of a non - zero result for @xmath138 .
next , the higgs - regulated form factor is finite , so that can be evaluated in four dimensions . in theories with broken supersymmetry , e.g. qcd with massive quarks ,
collinear singularities are regulated by the heavy quark masses , whereas all soft gluon divergences require dimensional regularization .
in such a case , the analogous functions @xmath104 and @xmath103 have a clear physical interpretation and are independent ( see e.g. ref .
for example , the ( electric ) form factor of a massive quark - anti - quark pair in qcd is known to two loops @xcite , and the analogs of collinear anomalous dimensions naturally coincide in this case , i.e. @xmath146 . in the case of the common - mass higgs regulator ( so that the external states are massless ) , the wedge function can again be defined as the square root of the form factor ( [ eq : phihiggs ] ) so that @xmath147 where [ defwt ] ( a ) = a^2 + ( a^3 ) and @xmath36 and @xmath148 were given in .
we now use in to define the finite part of the @xmath1-point amplitude [ eq : logth ] _
n & = & _ n - _ i=1^n ( ) + & = & ( a)4 _ n^+ n + ( a ) + _ n + ( m^2 ) again finding that the ir - divergent pieces of the wedge function precisely remove the ir divergences of the @xmath1-point amplitude . at one loop , gives [ htft ] _
n^ so that we can rewrite the finite part of the amplitude in its final form [ bdshiggsfinite ] _
n = ( a)4 _ n^+ n ( a ) + ( a ) + _ n ( ) + ( ) where [ defdt ] ( a ) = ( a ) - ( a ) = - a^2 + ( a^3 ) with @xmath149 given by . in complete analogy to the previous discussion
, it is evidently possible to exploit renormalization group properties to provide expressions for @xmath150 , @xmath151 , and hence @xmath152 in in terms of ( double-)integrals over the anomalous dimensions similar to . by comparing the results of the last two subsections
, we can see that @xmath153 as we have defined it in is a good candidate for a regularization - independent ir - finite quantity characterizing the planar mhv @xmath1-point amplitude .
the one - loop hard functions are identical in both dimensional and higgs regularization ( cf . eqs .
( [ fft ] ) , ( [ hf ] ) , and ( [ htft ] ) ) h_n^ ( ) = _ n^ ( ) .
[ equalh ] moreover , calculations through two loops show the equality of the kinematic - independent functions appearing in the @xmath1-point amplitude ( [ bdsfinite ] ) and ( [ bdshiggsfinite ] ) [ equality ] c(a ) = ( a ) , d(a ) = ( a ) ( cf . and ) .
the regulator independence of @xmath39 was previously observed in ref . @xcite .
if holds to all loops , then the regulator independence of the four- and five - point hard functions necessarily follows . for @xmath55 at two loops , agreement between the remainder function in dimensional regularization and massive regularization was observed in ref .
@xcite , and this agreement is expected for all @xmath1 .
we thus expect the hard function @xmath33 to be regularization - independent for all @xmath1-point functions , that is _
n ( ) = h_n ( ) for dimensional and higgs regularizations .
[ sect - diff ] in section [ sect - finite ] , we defined an ir - finite hard function @xmath33 for the @xmath1-point amplitude , and showed ( through two loops ) that it has the same form ( including constants ) for dimensional regularization and for a common - mass higgs regulator . in this section
, we generalize our discussion to a more general class of regulators , , the higgs regulator with arbitrary distinct masses .
this is also interesting from the point of view of collider phenomenology . in qcd , amplitudes with different masses have been considered to two loops for the heavy - to - light transitions , i.e. , the ( axial)-vector form factor with one massive and one massless quark @xcite .
also , electroweak logarithms in four - fermion processes at high energy arising from loop corrections with massive @xmath154- and @xmath155-gauge bosons have been considered to two loops ( see e.g. ref .
@xcite ) .
recall that breaking the @xmath57 symmetry of @xmath0 sym theory to @xmath156 by assigning distinct vacuum expectation values to one of the scalar fields results in non - zero masses @xmath157 for the off - diagonal adjoint fields and
distinct masses @xmath10 for the internal propagators of the scalar integrals that characterize loop amplitudes .
( in fact , extended dual conformal invariance requires the freedom to vary the masses . ) one can then define a differential - mass higgs regulator by taking all the masses @xmath10 to zero .
more precisely , if @xmath158 , the `` uniform small mass limit '' is defined as the limit @xmath135 with @xmath159 held fixed .
regulator independence means that the result does not depend on the choice of @xmath159 . because the external legs of the @xmath1-point amplitude now have distinct masses @xmath160 , it is no longer possible to define the wedge function @xmath84 as the square root of a form factor as we did in sec .
[ sect - finite ] .
in fact , it is not obvious what the operational definition of @xmath88 for the differential - mass higgs regulator should be , and we leave this question to the future . for
now , we adopt a different approach by decomposing the one - loop @xmath1-point amplitude into an ir - divergent and a manifestly regulator - independent ir - finite piece , and then defining the one - loop wedge function in terms of the former .
the extended dual conformal invariance of the bds ansatz then allows us to generalize this to an all - loop wedge function .
as is well known , the one - loop mhv @xmath1-point amplitude in dimensional regularization can be written as a sum of two - mass - easy ( and one - mass . ] ) scalar box integrals @xcite [ sumoftwomasseasy ] m_n^= - 1 8 _ i=1^n _ j = i+2^i+n-2 .
we will assume that the amplitude on the coulomb branch is given , at least up to @xmath136 , by the same set of integrals , with the mass configuration dictated by dual conformal symmetry .
the two - mass - easy diagram in fig .
[ fig-2me ] corresponds to the integral [ 2me ] = where @xmath161 with @xmath162 .
later , we will take the uniform small mass limit , so henceforth we drop all mass dependence from the numerators , as those terms would only contribute at @xmath163 .
it is known that one can decompose into a sum of ir - divergent triangle integrals and an ir - finite six - dimensional integral ( see e.g. ref .
@xcite ) @xmath164 where |_tri & = & [ tri ] and @xmath165 with @xmath166 and @xmath167 the two solutions of the equations @xmath168 . in appendix [ app - decomp ] of this paper
, we review the derivation of this decomposition using twistor methods .
since the six - dimensional integral ( [ 2me6d ] ) is ir - finite , it is independent of which ir - regulator we employ to regulate the @xmath1-point amplitude . therefore in the decomposition of the differential - mass higgs - regulated amplitude [ mn1wedges ] m_n^= _ i=1^n ^ ( ) + h_n^ ( ) + ( m^2 ) a natural candidate for the regulator - independent hard function is h_n^= - 1 8 _ i=1^n _ j = i+2^i+n-2 .
moreover , the sum of the one - loop wedge functions in will then be given by the sum of triangle diagrams ( [ tri ] ) that contribute to the @xmath1-point amplitude .
most of the triangle diagrams in this sum cancel , leaving _
i=1^n ^ ( ) & = & - _ i=1^n .
this suggests the following one - loop expression for the wedge function , @xmath10 , and @xmath64 . ]
[ wedge ] ^ ( ) = - although we could in principle add a contribution that vanishes upon summing over @xmath25 .
reduces to our previous definition @xmath126 when @xmath169 .
evaluating in the uniform small mass limit , we obtain [ oneloopwedge ] ^ ( ) = - ^2 ( x_i-1,i+1 ^ 2 m_i^2 ) - _ 2 ( 1 - ) - _ 2 ( 1 - ) + ( m^2 ) . substituting into and using the identity @xmath170 , we finally obtain the differential - mass higgs - regulated @xmath1-point amplitude [ decomp ] m_n^ & = & - _ i=1^n ^2 ( x_i-1,i+1 ^ 2 m_i^2 ) + _
i=1^n ^2 ( m_i+1 ^ 2 m_i^2 ) + h_n^ ( ) + ( m^2 ) where , as discussed above , @xmath171 is ir - finite and regulator - independent ; in particular , it does not depend on @xmath159 in the uniform small mass limit @xmath135 , where @xmath172 with @xmath159 fixed .
note that although @xmath173 has extended dual conformal invariance , the decomposition into @xmath174 and the specific ir - divergent pieces in breaks this symmetry .
this is not surprising , as triangle integrals manifestly violate dual conformal symmetry . in the previous subsection
, we derived an expression ( [ oneloopwedge ] ) for the one - loop wedge function valid for the differential - mass higgs regulator .
we now use the extended dual conformal invariance of the amplitude and the bds ansatz to derive the explicit form of the wedge function at higher loops .
recall that extended dual conformal invariance implies @xcite that the amplitude can only be a function of [ uij ] = m_i m_j + ( m_i - m_j)^2 . for the common - mass higgs regulator
, this reduces to @xmath175 . hence , assuming the validity of the all - loop expression ( [ bdshiggs ] ) for the common - mass higgs - regulated amplitude , its unique generalization is obtained by replacing @xmath176 with @xmath177 everywhere to obtain _
n & = & _ i=1^n + & & + h_n^(1 ) ( m^2/ ) +
n ( a ) + ( a ) + _ n ( m^2/ , a ) + ( m^2 ) where we have also used . in the uniform small mass limit , we can neglect @xmath178 relative to @xmath176 in so that @xmath179 becomes @xmath180 yielding [ bdsdiff ] _
n & = & _ i=1^n + & & + h_n^ ( m^2 m_i m_j ) +
n ( a ) + ( a ) + _ n ( m^2 m_i m_j , a ) + ( m^2 ) .
the apparent dependence of @xmath181 on @xmath10 is illusory since the mass dependence cancels out in the dual conformal cross ratios on which @xmath181 only depends , so that @xmath182 .
there does , however , remain some dependence on @xmath10 in @xmath183 . applying the same reasoning as above to the one - loop amplitude ,
we obtain [ diff ] m_n^ = - _ i=1^n ^2 ( m_i-1 m_i+1 ) + h_n^ ( ) + ( m^2 ) .
now since are both valid expressions for the differentially - regulated one - loop amplitude , we deduce that [ onelooprln ] h_n^ ( ) & = & h_n^ ( ) + _ i=1^n + & & + _
i=1^n ^2 ( m_i+1 ^ 2 m_i^2 ) + ( m^2 ) . substituting into and using ,
we obtain terms disappears on performing the sum over @xmath25 . ]
[ bdsdifftwo ] _
n & = & _ i=1^n + & & + h_n^(1 ) ( ) + n ( a ) + ( a ) + _ n ( , a ) + ( m^2 ) where now only the terms in the sum on the first line depend on the regulator , while the pieces on the second line are all regulator - independent . recalling that m_n = _ i=1^n ( ) + h_n [ decom ] we deduce from that the all - order expression for the sum of wedge functions in differential - mass higgs regularization is . ]
[ sumwedge ] & & _ i=1^n ( ) + & & = _ i=1^n and the regulator - independent ir - finite piece is , as before h_n = ( a)4 h_n^ ( ) + n
d(a ) + c(a ) + r_n ( , a ) where we have dropped all the tildes .
we observe again that , although the amplitude @xmath128 has extended dual conformal invariance , the separate terms in the decomposition ( [ decom ] ) do not .
[ sect - anom ] in the previous section , we obtained the all - loop expression ( [ sumwedge ] ) for the sum of wedge functions by assuming . in this section , we show inversely that follows from .
therefore , it would be interesting to have a first - principles derivation of .
the @xmath1-point amplitude has exact extended dual conformal symmetry , and so is annihilated by the generator of dual special conformal transformations ^ m_n
_ i=1^n m_n = 0 .
[ exact ] in ref .
@xcite , it was suggested that the ir - divergent properties of the higgs - regulated amplitude provides a relation between this exact ward identity and the anomalous dual conformal ward identity for the ir - finite part of the @xmath1-point amplitude that was originally derived in a wilson loop context @xcite .
we will see that this is indeed the case .
as we have seen , the @xmath1-point amplitude can be written as m_n = _ i=1^n ( ) + h_n . using the expression ( [ sumwedge ] ) for the sum of the wedge functions
, one can easily show that ^ _ i=1^n ( ) = - _ i=1^n ( ) which by virtue of implies that ^ h_n = _ i=1^n ( ) . but @xmath33 is regulator - independent , i.e. , has no dependence on @xmath10 in the uniform small mass limit , so the @xmath56-dependent pieces in @xmath184 drop out when acting on @xmath33 and we have @xmath185 \log h_n = \frac{\gamma(a ) } { 4 } \sum_{i=1}^{n } \left [ x_{i , i+1}^{\mu } \ , \log \frac{x^2_{i , i+2}}{x^2_{i-1,i+1 } } \right ] \label{anom}\ ] ] which is precisely the anomalous dual conformal ward identity @xcite .
this in turn implies .
we thus see that the decomposition of the amplitude into contributions which separately do not possess extended dual conformal invariance was necessary to obtain the anomalous dual conformal ward identity for the finite ( regulator - independent ) part of the amplitude .
[ sect - discussion ] in this paper , we have given a prescription for defining an unambiguous , regulator - independent ir - finite part of the mhv @xmath1-point scattering amplitude in planar @xmath0 sym theory .
this prescription involves the definition of an ir - divergent wedge function associated with a pair of adjacent external legs of the amplitude .
the ir - finite part of the amplitude is then defined as the quotient of the @xmath186point amplitude by the product of wedge functions , cf . .
for regulators that leave the external legs massless ( e.g. , dimensional regularization or the common - mass higgs regulator ) , the wedge function can be naturally defined in terms of a form factor @xmath97 which has the same ir - divergences .
computation of this form factor in dimensional regularization and in the common - mass higgs regularization through two loops shows that the ir - finite part of the amplitude is identical for these two regularizations . for the more general differential - mass higgs regulator , which gives ( small ) masses to the external legs , a wedge function that results in a regulator - independent hard function
can still be calculated , but an operator definition in this case is still lacking .
we remark that the idea of defining a regulator - independent finite hard function can also be applied to other objects , e.g. wilson loops .
this is particularly interesting in the context of the wilson loop / scattering amplitudes duality , since the two objects have different types of divergences , viz . , uv and ir divergences respectively .
although these divergences are related , defining hard functions for both objects could be useful for stating the duality in a regulator - independent way .
there exist in the literature other procedures for removing the ir divergences of scattering amplitudes .
for example , for non - mhv amplitudes , one can define an ir - finite ratio function @xcite by factoring out the entire mhv amplitude , using the universality of ir divergences , i.e. that they do not depend on the helicity configuration .
another example involves mhv amplitudes with @xmath187 external legs . since the four- and five - point amplitudes ( or , equivalently , wilson loops )
are known up to kinematic - independent functions , they can be used to remove the divergences of higher - point amplitudes by defining suitable ratios @xcite .
this latter procedure preserves dual conformal symmetry .
the hard functions defined in this paper are not dual conformal invariant ; they have the advantage , however , of allowing us to study the @xmath47 and @xmath48 cases as well .
in particular , it would be interesting to understand better the systematics of how the bes equation @xcite for @xmath36 arises from the loop expansion of the four - point amplitude .
the breaking of dual conformal invariance by the hard function also implies an intimate connection between the anomalous dual conformal ward identity it satisfies and the ir divergences ( wedge functions ) of differential - mass regulated amplitudes .
a first - principles derivation of the latter would therefore be most interesting .
finally , we believe that , although our investigation has been specialized to @xmath0 sym theory , the insight into the interplay between regulator , kinematics , and soft and collinear momentum configurations applies to many gauge field theories , including those with broken supersymmetry , such as qcd , and also to electroweak radiative corrections in the standard model .
it is a pleasure to thank z. bern , j. drummond , and g. korchemsky for interesting discussions .
all of the authors also wish to express their gratitude to the kavli institute for theoretical physics , where part of this work was carried out during the `` the harmony of scattering amplitudes '' program in spring 2011 .
this research was supported in part by the european - union funded network _
contract no .
pitn - ga-2010 - 264564 , as well as the national science foundation under grant nos .
phy05 - 51164 and phy07 - 56518 .
[ app - integrals ] in this appendix , we list the results for various massless and massive three - point integrals that contribute to the form factors computed in this paper .
we use the mostly - plus metric , the propagators are of the form @xmath188 , and the measure of each internal loop momentum is multiplied by a factor of @xmath189 .
the massless integrals shown in fig .
[ fig - massless ] are dimensionally - regulated , giving rise to the following laurent expansions @xcite i_b & = & ( ^2 ^-)^ = 1 q^2 ( ^2 q^2)^ , + i_c & = & 1 ( q^2)^2 ( ^2 q^2)^2 , + i_d & = & 1 ( q^2)^2 ( ^2 q^2)^2 , where @xmath190 . the integrals shown in fig .
[ fig - massive ] use a common - mass higgs regulator , and can be evaluated to give _
b & = & = 1 q^2 , + _ c1 & = & 1 ( q^2)^2 , + _ c2 & = & 1 ( q^2)^2 , + _ d & = & 1 ( q^2)^2 .
[ app - decomp ] we begin by rewriting the dual conformal invariant integral of equation ( [ 2me ] ) in terms of momentum twistors @xcite ; see ref .
@xcite for a pedagogical introduction to this topic .
a point @xmath71 in dual space corresponds to a ( projective ) line @xmath191 in momentum twistor space .
the invariant @xmath176 can be expressed as [ fourbracket ] = where the twistor four - bracket is = _ abcd z^a_z^b_z^c_z^d_. we introduce the infinity bitwistor @xmath192 , which when contracted with @xmath193 gives the two - bracket = _
abcd i^ab z^c_z^d_. finally , we introduce a modified mass - regulated four - bracket [ modified ] _ i + m_i^2 . rewriting using , we obtain [ 2metwistor ] = + ( m^2 ) where n_ij - and where @xmath194 .
we now decompose the two - mass - easy integral into a sum of ir - divergent and ir - finite contributions using a twistor identity . to derive this identity , consider the infinity bitwistor @xmath192 and expand it in the basis spanned by the six simple bitwistors [ infinity ] i^ab = c_i-1,i z^a_i-1 z^b_i + c_i , i+1 z^a_i z^b_i+1 + c_j-1,j z^a_j-1 z^b_j + c_j , j+1 z^a_j z^b_j+1 + c_i , j z^a_i z^b_j + c _ , z^a _ z^b _ where @xmath195 denotes the line in momentum twistor space formed by the intersection of @xmath196 and @xmath197 . contracting with @xmath198
, we obtain the identity [ iden ] & = & c_i-1,i + c_i , i+1 + c_j-1,j + & & + c_j , j+1 + c_i , j + c _ , .
we multiply and divide the integrand of by @xmath199 , using to rewrite the numerator .
the first four terms will each cancel one of the propagators as @xmath200 and then drop the @xmath201 piece . ] resulting in four triangle integrals , whereas the last two terms remain box integrals = be determined by contracting with each of the basis elements in turn , e.g. , = c_j-1,j + c_j , j+1 and five other equations .
thus we find |_tri & = & + & = & and [ magic ] |_box & = & + & = & where @xmath166 and @xmath167 are the two solutions of the equations @xmath168 , and where in we have used the identities n_ij = = .
the presence of two - brackets in @xmath202 and @xmath203 indicate that these expressions are not individually dual conformal invariant .
this is not surprising , as scalar triangle diagrams violate dual conformal invariance .
we observe that the integrands in contain `` magic '' numerators , which render the resulting integrals ir - finite .
hence we have dropped the mass dependence in the denominator .
one can show that this integral is in fact equivalent to the scalar two - mass easy integral in six dimensions @xcite |_box = .
since this integral is ir - finite , it is manifestly independent of which ir regulator is used to regulate the amplitude . in particular , it has no dependence on @xmath159 in the uniform small mass limit of the differential - mass higgs regulator introduced in sec .
[ sect - diff ] . | the ir - divergent scattering amplitudes of @xmath0 supersymmetric yang - mills theory can be regulated in a variety of ways , including dimensional regularization and massive ( or higgs ) regularization .
the ir - finite part of an amplitude in different regularizations generally differs by an additive constant at each loop order , due to the ambiguity in separating finite and divergent contributions .
we give a prescription for defining an unambiguous , regulator - independent finite part of the amplitude by factoring off a product of ir - divergent `` wedge '' functions .
for the cases of dimensional regularization and the common - mass higgs regulator , we define the wedge function in terms of a form factor , and demonstrate the regularization independence of the @xmath1-point amplitude through two loops .
we also deduce the form of the wedge function for the more general differential - mass higgs regulator , although we lack an explicit operator definition in this case . finally , using extended dual conformal symmetry , we demonstrate the link between the differential - mass wedge function and the anomalous dual conformal ward identity for the finite part of the scattering amplitude . hu - ep-11 - 41 + desy-11 - 148 + bow - ph-150 + nsf - kitp-11 - 197 + * form factors and scattering amplitudes in @xmath0 sym + in dimensional and massive regularizations * johannes m. henn@xmath2 , sven moch@xmath3 , and stephen g. naculich@xmath4 @xmath5_institut fr physik + humboldt - universitt zu berlin , newtonstrae15 , d-12489 berlin , germany _ @xmath6 _ deutsches elektronensynchrotron desy , + platanenallee 6 , d-15738 zeuthen , germany
_ @xmath7 _ department of physics + bowdoin college , brunswick , me 04011 , usa _
@xmath8 _ kavli institute for theoretical physics + university of california , santa barbara , ca 93106 , usa _ |
at the tevatron @xmath2 mesons are copiously produced in @xmath5 collisions at a center of mass energy of @xmath6 tev .
having recorded datasets corresponding to an integrated luminosity of more than 5 fb@xmath7 , the two tevatron experiments cdf and d0 are able to perform detailed studies of @xmath2 mesons .
essential for these analyses are triggers that efficiently select events with @xmath2 mesons .
they are identified by decays to @xmath8 mesons with @xmath9 , semileptonic decays to muons ( d0 ) , or hadronic decays with a displaced decay vertex ( cdf ) .
the @xmath2 mesons are a promising place to search for physics beyond the standard model ( sm ) because several flavor changing neutral current ( fcnc ) processes can be measured in the @xmath2 system .
as there are no fcnc processes at tree level in the sm , new physics contributions may be sizable .
two rare fcnc decay modes of the @xmath2 that are analyzed at the tevatron are @xmath3 and @xmath10 .
another interesting fcnc process is the oscillation of @xmath2 mesons .
it is described by a schrdinger equation @xmath11 where @xmath12 and @xmath13 are @xmath14 hermitian matrices .
the solution are two eigenstates with defined masses , @xmath15 and @xmath16 , and decay widths , @xmath17 and @xmath18 , respectively . while the measured oscillation frequency , @xmath19 , agrees well with the ( less precise ) sm prediction , there may still be sizable new physics contributions to the phase @xmath20 .
this would result in a violation of the _ cp _ symmetry in the @xmath2 system , that is expected to be conserved to very good approximation in the sm .
in the sm the fcnc process @xmath3 is predicted to have a branching ratio of @xmath22 @xcite .
a significant enhancement of the branching ratio is expected in several new physics models .
for example in the minimal super - symmetric standard model the @xmath2 branching ratio is proportional to @xmath23 where @xmath24 is the ratio between the vacuum expectation values of the two neutral higgs fields . in @xmath25-parity violating super - symmetric models an enhancement is possible even at low values of @xmath24 .
both tevatron experiments optimize the selection of @xmath3 candidates using simulated signal events and background events from mass sidebands .
while d0 combines the discriminant variables in a boosted decision tree , cdf uses a neural network ( nn ) .
the combinatorial background is estimated by a fit to the mass sidebands . the contribution from decays of @xmath26 mesons to two light hadrons , which could peak in the signal mass region ,
was estimated to be an order of magnitude lower than the combinatorial background . to obtain an absolute branching ratio
the number of signal events is normalized to the high - statistics @xmath27 mode . for the limit calculation cdf
splits the data sample in three bins in nn output and five bins in mass which improves the sensitivity by 15% compared to using just one bin .
the d0 analysis is performed on three separate data samples corresponding to three different detector configurations .
cdf does not see a significant excess ( fig .
[ fig : bsmumu ] left ) in a data sample of 2 fb@xmath7 and sets a 90% confidence level ( cl ) limit at @xmath28 @xcite .
d0 has not yet unblinded the signal region ( fig .
[ fig : bsmumu ] right ) and quotes an expected upper limit at 90% cl of @xmath29 for a data sample of 5 fb@xmath7 @xcite .
candidates measured by cdf in three bins of neural network output ( left ) and by d0 in three run ranges ( right ) . ] candidates measured by cdf in three bins of neural network output ( left ) and by d0 in three run ranges ( right).,title="fig : " ] candidates measured by cdf in three bins of neural network output ( left ) and by d0 in three run ranges ( right).,title="fig : " ] candidates measured by cdf in three bins of neural network output ( left ) and by d0 in three run ranges ( right).,title="fig : " ]
current measurements of the fcnc quark level process @xmath31 in @xmath0 and @xmath1 decays already provide stringent constraints on new physics models . at the tevatron
this process can be analyzed in @xmath2 decays if the photon is virtual and decays to @xmath32 . at hadron level
the decay @xmath10 is reconstructed . in the sm
it is predicted to have a branching ratio of the order @xmath33 @xcite
. decays to @xmath34 and @xmath35 are excluded from the measurement .
however , the former one is used as a normalization channel . in a data sample of 0.92
fb@xmath7 cdf sees an excess with a significance of 2.4 standard deviations ( fig .
[ fig : bsmumuphi ] left ) . assuming this excess comes from @xmath10 decays , a branching ratio of @xmath36 is measured , well consistent with the sm expectation . since the signal is not significant cdf quotes a 95% cl limit on the relative branching ratio of @xmath37 , corresponding to a limit on the absolute branching ratio of @xmath38 @xcite .
d0 observes no events in the signal region in a data sample of 0.45 fb@xmath7 ( fig .
[ fig : bsmumuphi ] right ) .
the derived 95% cl upper limits are @xmath39 and @xmath40 @xcite .
candidates measured by cdf ( left ) and d0 ( right ) . ] candidates measured by cdf ( left ) and d0 ( right ) . ]
_ cp _ violation in mixing can be observed by measuring decay rates to flavor specific final states , @xmath42 and @xmath43 , that do not exhibit direct _ cp _ violation .
the asymmetry @xmath44/ [ \gamma(\bar{b}^0_s \rightarrow f)+\gamma(b^0_s \rightarrow \bar{f } ] = ( \delta\gamma/\delta m ) \tan\phi$ ] with @xmath45 is expected to be @xmath46 in the sm @xcite . since it is often measured in semileptonic
decays it is called semileptonic _ cp _ asymmetry .
d0 measures the asymmetry in @xmath47 decays where the @xmath48 is reconstructed in the decays @xmath49 , @xmath50 ( fig . [ fig : asl ] left ) and @xmath51 , @xmath52 .
the time evolution of @xmath2 and @xmath53 mesons is fitted in an unbinned maximum likelihood fit where information about the production flavor is determined from decay products of the second @xmath54 hadron produced in the collision ( opposite side flavor tagging ) . as the final state is not reconstructed exclusively , the missing momentum is corrected for using simulation . in a data sample of 5 fb@xmath7 d0 measures an asymmetry of @xmath55 @xcite .
this value agrees well with the sm prediction and previous measurements @xcite as can be seen in fig .
[ fig : asl ] right .
candidates measured by d0 ( left ) and comparison of @xmath56 measurements @xcite ( right ) . ] candidates measured by d0 ( left ) and comparison of @xmath56 measurements @xcite ( right ) . ]
_ cp _ violation in the interference of decays with and without mixing can be observed in decays to final states that are accessible to @xmath2 and @xmath53 . for the decay to @xmath34
the phase @xmath58 between both processes is in the sm given by @xmath59 @xcite .
contributions from new physics could enlarge it in the same way as the phase @xmath60 .
as @xmath8 and @xmath60 are vector mesons the decay of the pseudoscalar @xmath2 can proceed via angular momentum @xmath61 or 2 , resulting in a _
parity of @xmath62 .
an angular analysis allows to identify the _
cp _ eigenstates .
information about _ cp _ violation can be extracted from a comparison to the mass eigenstates identified by their defined lifetime .
thus the mixture of final _ cp _ eigenstates is advantageous if @xmath63 is significantly different from zero .
further sensitivity is achieved by identifying the production flavor ( flavor tagging ) .
it allows to measure the asymmetry @xmath64/ [ \gamma(\bar{b}^0_s \rightarrow f)+\gamma(b^0_s \rightarrow f ) ] = \pm\sin(2\beta_s)\sin(\delta m\ , t)$ ] where the sign depends on the _ cp _ parity of the final state . as can be seen from the equation this approach requires to resolve the fast @xmath2 oscillations .
both tevatron experiments analyze a data sample of 2.8 fb@xmath7 each .
cdf selects about 3200 @xmath65 signal events and measures the 68% and 95% cl region in the @xmath58-@xmath63 plane shown in fig .
[ fig : jpsiphi ] left @xcite . because of unknown strong phases the result exhibits a symmetry with respect to @xmath66 and @xmath67 .
the one - dimensional confidence region for @xmath58 is @xmath68 $ ] at 68% cl .
d0 selects approximately 2000 signal events and obtains the confidence regions shown in fig . [
fig : jpsiphi ] right @xcite . from the fit the value @xmath69 is determined . in this result
the strong phases are constrained to values measured in @xmath70 decays , which removes the ambiguity . in ref .
@xcite d0 recently presented an updated preliminary result without constraints on strong phases .
as can be seen in fig .
[ fig : jpsiphi ] both results are consistent with each other .
the consistency with the sm , measured by the @xmath71-value , is 7% ( 1.8@xmath72 ) for cdf and 8.5% ( 1.7@xmath72 ) for d0 . the green band in fig . [ fig : jpsiphi ]
is given by @xmath73 and shows the region accessible by new physics models that have the same value of @xmath74 , which is dominated by tree level processes , as predicted in the sm .
-@xmath63 plane measured by cdf ( left ) and 68% and 90% cl regions in the @xmath75-@xmath63 plane measured by d0 ( right ) where @xmath76 . the cross shows the @xmath75 and @xmath63 values measured by d0 . ]
-@xmath63 plane measured by cdf ( left ) and 68% and 90% cl regions in the @xmath75-@xmath63 plane measured by d0 ( right ) where @xmath76 .
the cross shows the @xmath75 and @xmath63 values measured by d0 . ]
no evidence for physics beyond the sm was found so far in the @xmath2 system .
the limits on @xmath77 place stringent constraints on new physics models and a further reduction of their parameter space can be expected from updated measurements .
first evidence or even observation of @xmath4 seems within reach of the tevatron . while the @xmath78 measurements agree with the sm within uncertainties , some deviation from the sm is observed in decays to @xmath34 .
more data is available and will be taken , which gives the tevatron a realistic chance to observe new physics if _
cp _ violation in the @xmath2 system is large .
99 slides : ` http://indico.cern.ch/contributiondisplay.py?contribid=80&sessionid=2&confid=53294 ` .
m. blanke , a. j. buras , d. guadagnoli and c. tarantino , jhep * 0610 * , 003 ( 2006 ) .
t. aaltonen _ et al . _
[ cdf collaboration ] , phys .
lett . * 100 * , 101802 ( 2008 ) .
d0 collaboration , d0 conference note 5906 , + ` http://www-d0.fnal.gov/run2physics/www/results/prelim/b/b56/b56.pdf ` c. q. geng and c. c. liu , j. phys .
g * 29 * , 1103 ( 2003 ) .
t. aaltonen _ et al . _
[ cdf collaboration ] , phys .
d * 79 * , 011104 ( 2009 ) .
v. m. abazov _ et al . _
[ d0 collaboration ] , phys .
d * 74 * , 031107 ( 2006 ) .
a. lenz and u. nierste , jhep * 0706 * , 072 ( 2007 ) . v. abazov
_ et al . _
[ d0 collaboration ] , arxiv:0904.3907 [ hep - ex ] . v.
m. abazov _ et al . _
[ d0 collaboration ] , phys .
lett . * 98 * , 151801 ( 2007 ) .
cdf collaboration , cdf public note 9015 , + ` http://www-cdf.fnal.gov/physics/new/bottom/070816.blessed-acp-bsemil/public-acp-bsemil.ps ` .
cdf collaboration , cdf public note 9458 , + ` http://www-cdf.fnal.gov/physics/new/bottom/080724.blessed-tagged_bsjpsiphi_update_prelim/public_note.pdf `
. v. m. abazov _ et al .
_ [ d0 collaboration ] , phys .
* 101 * , 241801 ( 2008 ) .
d0 collaboration , d0 conference note 5933 , + ` http://www-d0.fnal.gov/run2physics/www/results/prelim/b/b58/b58.pdf ` . | while @xmath0 and @xmath1 mesons are well studied at b - factories , large samples of @xmath2 mesons are only available at the tevatron so far .
since the @xmath2 meson consists of quarks of the second and third generation it provides a complementary probe for searches for new physics effects .
results of the cdf and d0 experiments on the decays @xmath3 and @xmath4 as well as on _ cp _ violation in @xmath2 mixing and in interference between mixing and decay are presented . |
the understanding of spin - dependent effects in inclusive hadron production processes in the framework of qcd is still far from being satisfactory , despite significant experimental and theoretical progress over the past few years . in particular
, the study of hyperon polarization ( @xmath13 ) and the analyzing power ( @xmath11 ) could provide invaluable and completely new insight into the field of `` spin physics '' and , in addition , might also yield a better understanding of the hadronization process . in this paper we will study the existing data for one measured spin - dependent quantity ( transverse hyperon polarization @xmath13 in inclusive reactions @xmath14 ) from an empirical point of view in collisions of unpolarized protons , antiprotons , @xmath15 , @xmath16 or hyperons with protons or nuclei .
experiments on hyperon production performed during more than two decades since the first polarization observation @xcite have shown that the hyperon polarization is significant in a wide range of beam energies . almost all the existing data ( with an equivalent proton momentum on a fixed target from 4 up to 2049 gev / c ) at medium and high energies are used for the analysis .
the study of @xmath9 hyperon polarization in @xmath17 and @xmath18 collisions has been carried out in many experiments and revealed an approximate scaling behaviour as a function of @xmath7 variable at fixed @xmath0 , as well as a function of @xmath0 at fixed @xmath7 @xcite-@xcite .
further investigations of the hyperon and proton polarization using different beams and targets have shown its dependence on the hyperon and the beam flavors , as well as on the target atomic weight @xcite-@xcite .
it has to be mentioned that a large value of the hyperon polarization represents a significant problem for the existing strong interaction theory which predicts , in the framework of pertubative quantum chromodynamics ( pqcd ) , the vanishing of the polarization at high @xmath0 and energy @xcite . for the review of the hyperon polarization data and the existing models
see @xcite .
recently a new scaling law has been proposed for a hadron production analyzing power ( @xmath11 ) in reactions @xmath19 where @xmath20 , @xmath21 and @xmath22 are hadrons , and the hadron @xmath20 is transversely polarized @xcite .
since the hyperon polarization and the hadron analyzing power could be closely related @xcite , it is reasonable to expect that similar scaling properties take place for the inclusive hyperon production in collisions of unpolarized hadron beams with protons or nuclei . according to ref .
@xcite @xmath11 at high energies ( @xmath23 gev ) and high transverse momentum @xmath0 ( @xmath24 gev / c ) can be described by a simple function of two variables , @xmath0 and @xmath25 : @xmath26 where @xmath27 here , @xmath28 , @xmath29 and @xmath30 are four - momenta of the produced hadron , the target hadron and the beam hadron , respectively .
the energies @xmath31 ( produced hadron ) and @xmath32 ( beam hadron ) are measured in a reference frame , where a polarized ( @xmath33 or @xmath34 ) particle strikes an unpolarized target ( @xmath33 or @xmath35 ) which is at rest .
there are also alternative expressions for the scaling variable @xmath25 , which are close to ( [ eq : xa ] ) numerically at the high beam and secondary hadron energies @xcite .
in particular , neglecting masses in ( [ eq : xa ] ) , we have in the c.m .
frame @xmath36 where @xmath37 and @xmath38 .
here @xmath39 , @xmath40 and @xmath41 is momentum of the produced hadron , its longitudinal component and the maximum possible value of it , respectively , all in the c.m .
reference frame .
( [ eq : fgx ] ) and ( [ eq : xfr ] ) describe @xmath11 for most of the hadron production reactions almost as well as eqs .
( [ eq : fgx ] ) and ( [ eq : xa ] ) @xcite .
the only exception is the @xmath11 for @xmath9 hyperon production in @xmath42 collisions , for which the variable ( [ eq : xfr ] ) is preferable since it gives a smaller @xmath43 ( @xmath44 for ( [ eq : xfr ] ) vs 39.4/44 for ( [ eq : xa ] ) ) , where @xmath45 is the number of degrees of freedom in a fit .
it is interesting that @xmath9 hyperon polarization is also better described by eq .
( [ eq : fgx ] ) if @xmath25 is given by ( [ eq : xfr ] ) .
it has to be mentioned that the variable @xmath46 was used to describe the scaling properties of the secondary hadron ( @xmath47 ) spectra in the region @xmath48 @xcite .
a natural generalization of ( [ eq : xfr ] ) is given by a linear function of @xmath49 and @xmath7 variables with relative weights which could be determined from the data fit : @xmath50 where @xmath51 . for @xmath52 the variable @xmath53 in ( [ eq : xfra ] ) is close to @xmath54 in the beam fragmentation region ( @xmath55 ) and @xmath56 in the central region ( @xmath57 ) .
the variable @xmath58 have similar properties in the backward hemisphere ( @xmath59 ) . for @xmath52 variable @xmath53
is identical to @xmath25 ( eq.([eq : xfr ] ) ) .
the hyperon polarization features are different in some respects from the analyzing power features listed in @xcite .
in particular , we have to take into account that the polarization of hyperons produced in @xmath17 collisions is antisymmetric in @xmath7 by virtue of rotational invariance @xcite : @xmath60 from the relation ( [ eq : pasym ] ) we have also for the hyperon polarization @xmath61 for @xmath17 or other symmetric initial states in contrast to the analyzing power , which could be different from zero even at @xmath62 , see @xcite .
the initial state @xmath63 is not symmetric with respect to the @xmath64 rotation since one proton is polarized . from a dimensional analysis alone ,
the polarization @xmath6 admits two types of contributions , which are functions of @xmath65 and @xmath66 : @xmath67 , \label{eq : gut}\ ] ] where @xmath68 , @xmath69 and @xmath70 are mandelstam variables : @xmath71 where in the last column the hadron masses are neglected .
it is easy to show that at high energies the following approximations are valid : @xmath72 @xmath73 where @xmath1 are given by eq .
( [ eq : xfra ] ) with @xmath52 .
relations , similar to ( [ eq : xad1ef])-([eq : xad2ef ] ) , have been used in refs .
using eqs .
( 7 ) - ( [ eq : xad2ef ] ) it is easy to show that @xmath74 , where @xmath25 is given by eq .
( [ eq : xa ] ) . the expression , which takes into account ( [ eq : pasym])-([eq : xad2ef ] ) and other features of @xmath13 is given below : @xmath75 , \label{eq : pol}\ ] ] where an additional phase @xmath76 and a signature parameter @xmath77 are introduced to take into account a possible violation of ( [ eq : pasym ] ) for collisions , different from @xmath17 one ( for @xmath17 collisions @xmath78 and @xmath79 ) .
it is assumed below that @xmath80 for @xmath18 collisions , while @xmath76 could be different from zero .
the functions @xmath81 and @xmath82 are determined below from the experimental data .
exactly the same approach can be used to derive an expression for the @xmath83 .
the difference is that in the last case @xmath84 in ( [ eq : gut ] ) , and @xmath85 dominates , at least for the @xmath86 reaction , as the data fit show , over @xmath87 .
for the process @xmath88 we use an expression @xmath89 for the @xmath90 , where @xmath91 is a fit parameter and @xmath92 is measured in gev / c .
the other processes may require a different expression .
the exact shape of the @xmath90 should be measured in future experiments . for the @xmath82 we use an expression @xmath93 , \label{eq : gx}\ ] ]
similar to the one , used in @xcite , which allows one to reproduce such features , as an approximate linear dependence of @xmath94 on the @xmath7 , and non - linear , sometimes oscillating behaviour of the hyperon polarization for other reactions .
the difference of eq .
( [ eq : gx ] ) from the corresponding expression in @xcite is that the former has an additional factor @xmath95 , that makes parameters @xmath22 and @xmath8 less correlated and reflects the tendency of hyperon polarization magnitude decrease with @xmath8 rise .
the parameters @xmath96 , @xmath22 , @xmath97 , @xmath76 , @xmath77 and @xmath8 are determined from the data .
the @xmath9 polarization data fits indicate that both parameters @xmath97 and @xmath76 can depend on @xmath0 and only at high enough @xmath98 gev / c are compatible with @xmath0-independent constant values : @xmath99 @xmath100 where the target dependent factor @xmath101 is introduced to insure @xmath102 for @xmath17 collisions , as required by relation ( [ eq : pasym ] ) . for a neutron target @xmath103 , so the phase @xmath76 is the only parameter in the above equations which makes the hyperon polarization for the proton target different from that of the neutron one .
this feature of eq .
( [ eq : x2 ] ) can be used to estimate the difference in the polarization of hyperons produced on the proton and the neutron targets .
the factor @xmath104 takes into account a possible atomic weight dependence of the hyperon polarization .
the @xmath105 could be a constant or a function of kinematic variables : @xmath106 where @xmath107 and @xmath108 are fit parameters . in the beam fragmentation region @xmath109 , and @xmath110 , while @xmath111 .
so , in this region the dependence of eq .
( [ eq : pol ] ) on @xmath7 is determined , mainly , by the first term @xmath112 . the second term in ( [ eq : pol ] )
, @xmath113 has a weak @xmath7 dependence . in the target fragmentation region
@xmath7 dependence is determined , mainly , by the @xmath114 .
( [ eq : pol ] ) for @xmath115 can be expressed using ( [ eq : gx ] ) in a different way , with the explicit @xmath49 and @xmath7 dependences : @xmath116 \sin [ \omega(x_{f}w_{2 } - x_{2 } ) ] , & if $ \sigma = + 1 $ ; \cr \sin [ \omega(x_{r}w_{1 } - x_{1 } ) ] \cos [ \omega(x_{f}w_{2 } - x_{2 } ) ] , & if $ \sigma = -1$. \cr } \label{eq : pol1}\ ] ] eq .
( [ eq : pol ] ) has a more general form , but it coincides with ( [ eq : pol1 ] ) if we choose eq .
( [ eq : gx ] ) for @xmath82 and set @xmath115 .
the magnitude of the hyperon polarization is about @xmath117 .
it is assumed here that @xmath81 is equal unity at its maximum . in case of the analyzing power measurements we have a non - zero contribution to the left - right asymmetry from the polarized hadron ( beam or target ) only and the feature ( [ eq : pasym ] )
is not valid .
naively , we may say that the first term , @xmath118 , in ( [ eq : pol ] ) , corresponding to a polarized beam , gives the main contribution to the analyzing power and , as a result , we have eq .
( [ eq : fgx ] ) for it .
for the case of @xmath9 hyperon polarization in the @xmath17 collisions both terms in ( [ eq : pol ] ) , corresponding to the beam fragmentation and the target fragmentation , have non - zero contributions and cancel each other at @xmath62 , in accordance to ( [ eq : pasym ] ) .
relation ( [ eq : pasym ] ) is not valid in general for collisions of different hadrons , like @xmath119 or @xmath120 collisions . in these cases we have to use different functions @xmath85 and @xmath87 of @xmath53 and @xmath58 , respectively . in particular
, parameter @xmath76 can be different from zero and @xmath77 can be different from unity .
it should be also mentioned , that in the case of a linear function @xmath82 and @xmath80 the @xmath49 terms are cancelled in ( [ eq : pol ] ) .
the use of @xmath121 in ( [ eq : gx ] ) makes @xmath82 non - linear , that prevents complete cancellation of @xmath49 terms in ( [ eq : pol ] ) .
a linear case corresponds to the limit @xmath122 in the above equations , while the data fits give for @xmath88 process @xmath123 , indicating a rather non - linear behaviour of @xmath82 at large values of arguments . as we will see in the following sections the @xmath8 can be expressed for different reactions via a linear combination of terms , which are functions of quantum numbers characterizing the particular reaction ( see eq .
( [ eq : omega ] ) ) .
( [ eq : omega ] ) is used below to fix the @xmath8 in eq .
( [ eq : pol ] ) in data fits that allows us to estimate the other parameters with a better accuracy .
it follows from eqs .
( [ eq : pol ] ) - ( [ eq : alpha ] ) that for the hadron - proton collisions the hyperon polarization @xmath13 and the analyzing power @xmath11 obey the helmholtz equation : @xmath124 eq .
( [ eq : helm ] ) reflects the non - linearity of the @xmath13 dependence and its oscillation as a function of scaling variables @xmath1 , which is characterized by the @xmath8 parameter .
most of the hyperon production experiments have been performed on nuclear targets , where relation ( [ eq : pasym ] ) may not be exactly valid .
also , practically all the available data are concentrated in the forward hemisphere .
we can not exclude that future measurements in the backward hemisphere ( nuclear target fragmentation region ) will show some deviation from ( [ eq : alpha ] ) , which will require a correction of the polarization @xmath35-dependence .
an estimate of nuclear effects in the forward hemisphere will be given below on the base of existing data .
we assume also for simplicity that the @xmath8 parameter is the same for the forward and the backward hemispheres and reflects the dynamics of strong interaction .
the polarization of @xmath125 hyperons ( @xmath94 ) in @xmath126 collisions has been measured in many experiments @xcite-@xcite .
the data include proton collisions with protons and different nuclei ( @xmath127 and @xmath128 ) .
we assume here that for the positive polarization and scattering to the left the hyperon spin is directed along the unit vector @xmath129 , which is the normal to the production plane . here
, @xmath130 and @xmath131 is a momentum of the beam hadron and that of the produced hyperon , respectively . the fit parameters for eqs .
( [ eq : pol])-([eq : alpha ] ) are shown in table [ t1_lambda ] for different fit conditions .
the fits have been performed for the data @xcite-@xcite . in the fit # 1
all parameters are free .
the use of @xmath0-dependent phases @xmath97 and @xmath76 in eqs .
( [ eq : x1])-([eq : x2 ] ) improves the fit quality significantly ( @xmath132 changed from 1.81 to 1.26 ) .
the dependence of phases @xmath97 and @xmath76 on @xmath0 is shown in fig . [ x1_x2_pt ] . as in the case of polarization dependence on @xmath0
( see fig .
[ lam_pa_pt ] ) the phases @xmath97 and @xmath76 are described well by the terms @xmath133 $ ] with a plateau for @xmath0 above 1 - 2 gev / c .
the absolute value of @xmath76 is small in accordance with the approximate validity of the feature ( [ eq : pasym ] ) for @xmath18 collisions .
the values of @xmath134 and @xmath135 for the fit # 1 are close to the expected ones ( 0.5 and 1 , respectively ) with @xmath136 .
so , the fit # 1 supports the choice of scaling variables in the form @xmath137 and an approximate rotational asymmetry ( [ eq : pasym ] ) for the polarization in @xmath18 collisions , though the experimental data exist only in the forward hemisphere .
the parameter @xmath8 in the fit # 1 is equal to @xmath138 .
the large value of @xmath8 excludes the possibility of a linear dependence of polarization on @xmath1 or @xmath7 variables .
the fit # 2 is made with four parameters fixed ( @xmath139 , @xmath80 , @xmath140 and @xmath141 ) that simplifies eq .
( [ eq : pol ] ) and allows one to determine the other parameters with a better accuracy .
an interesting feature of the above fits is that at the first approximation the @xmath105 is proportional to @xmath54 .
the higher is a hyperon momentum the more its polarization is attenuated by the interactions with a nuclear target . for a heavy nuclear target and a large @xmath7
we expect a significant attenuation of the polarization in comparison with the @xmath17 collisions case .
in particular , for @xmath142 we have @xmath143 and the polarization on a lead target is reduced by a factor @xmath144 . on a beryllium target
a corresponding factor is 0.79 .
a similar order of magnitude for the polarization degradation in complex nuclei can be found in @xcite , where it is also shown that the polarization degradation is @xmath0-independent .
ccccccc + fit # & 1 & 2 + + @xmath22 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath8 & [email protected] & @xmath146 & + [ 0.1 cm ] @xmath77 & [email protected] & 1.0 & + [ 0.1 cm ] @xmath96 & [email protected] & 0.5 & + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath148 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath149 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath151 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath152 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath153 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath107 & [email protected] & 0.0 & + [ 0.1 cm ] @xmath108 & [email protected] & [email protected] & + [ 0.1 cm ] @xmath132 & 325.7/256 & 326.0/260 + [ 0.1 cm ] [ t1_lambda ] the @xmath0 dependence of @xmath94 is described by the function @xmath81 , which is sharply rising for @xmath154 gev / c and is practically constant for higher @xmath0 .
the experimental dependence of @xmath94 on @xmath0 is illustrated in fig .
[ lam_pa_pt ] , where the ratio of @xmath94 and @xmath155 is shown , and @xmath156 the ratio is assumed to be a function of @xmath0 only . an additional cut @xmath157 is used for the data shown in fig .
[ lam_pa_pt ] to exclude the points with large fractional errors of the ratio . the data in fig .
[ lam_pa_pt ] show the independence of the ratio on the beam energy , @xmath7 and the target type , and confirms the scaling behaviour and factorization of the @xmath0 and @xmath1 dependencies , assumed in ( [ eq : pol ] ) .
new measurements are desirable for @xmath158 gev / c to clarify the @xmath94 and @xmath81 behaviour at high @xmath0 . the dependence of @xmath94 on @xmath7 is illustrated in fig .
[ lam_be_xf ] , where the most precise polarization data , measured in @xmath159 collisions at 400 and 800 gev / c are shown .
the fitting curves reproduce the @xmath7 dependence , as well as small variations of it with @xmath0 .
the attenuation of the @xmath9 polarization on nuclear targets and the dependence of the attenuation on @xmath7 are illustrated in fig .
[ lam_adep_xf ] .
the measured polarization is divided by @xmath160 and the corresponding ratio is plotted in fig .
[ lam_adep_xf ] vs @xmath7 .
the data for the @xmath17 @xcite , @xmath159 @xcite , and @xmath161 @xcite collisions show a decrease of polarization on @xmath162 and @xmath163 targets with @xmath7 rise in comparison with the proton - proton collisions case .
the existing @xmath164 data on medium and heavy nuclear targets are limited in terms of their accuracy and kinematic range .
additional measurements on medium and heavy nuclear targets for positive and negative @xmath7 are desirable to confirm and clarify the polarization attenuation effects , shown in fig .
[ lam_adep_xf ]
. it will be interesting also to measure the @xmath35-dependence of the analyzing power @xmath83 and compare it with that of the hyperon polarization . the general agreement between the data and the fits # 1 - # 2
can be considered a good one , taking into account statistical and possible systematic errors of the data in different experiments .
we may conclude that @xmath9 polarization reveals a scaling behaviour , when it is presented in a forward hemisphere as a function of two scaling variables @xmath1 ( or @xmath49 and @xmath7 ) and @xmath0 .
in this section the polarization of @xmath165 and @xmath166 hyperons in @xmath17 and @xmath18 collisions is analyzed .
the corresponding fit parameters are presented in tables [ t2_sigma ] and [ t2_xi ] .
the fits have been performed using eqs .
( [ eq : pol])-([eq : alpha ] ) , and eq .
@xmath167 for the function @xmath81 , which is found to be more appropriate for an approximation of the polarization dependence on @xmath0 . for this section and all the following the parameters @xmath107 and @xmath96 are fixed ( @xmath141 and @xmath140 ) .
the polarization of @xmath168 hyperons ( @xmath169 ) in @xmath18 collisions has been measured on a @xmath170 target at 800 gev / c @xcite and on a @xmath170 and a @xmath162 targets at 400 gev @xcite .
a fit # 1 is made with some parameters fixed due to a limited data statistics .
the parameter @xmath171 is compatible with the value of @xmath8 for the case of @xmath9 polarization discussed in the previous section .
the fit # 2 is done with the value of @xmath139 , which follows from eq .
( [ eq : omega ] ) .
the dependence of @xmath169 on @xmath0 is illustrated in fig .
[ sigp_pa_pt ] , where a ratio of @xmath169 and @xmath172 is plotted .
the curve corresponds to the function @xmath81 in eq .
( [ eq : f2 ] ) with @xmath173 gev / c ( fit # 2 ) .
the function @xmath81 is well approximated by a linear dependence for @xmath174 gev / c and a plateau for higher @xmath0 ( see fig .
[ sigp_pa_pt ] ) , but additional measurements are desirable for higher @xmath0 and different @xmath7 .
it was stated in @xcite that the @xmath168 polarization at 800 gev / c decreases as a function of @xmath0 at fixed @xmath7 .
the results of the fit # 2 indicate that such unusual @xmath0 dependence is probably due to the @xmath0 dependence of the parameter @xmath76 in eq .
( [ eq : x2 ] ) . as we can see from eq .
( [ eq : pol1 ] ) , the polarization ( for @xmath80 ) is proportional to @xmath175 \approx \sin [ 3(x_{f}/2 - x_{2 } ) ] $ ] .
since @xmath176 and @xmath177 , the value of @xmath76 is about 0.09 for @xmath178 gev / c , it starts to grow fast for @xmath179 gev / c , and has a plateau @xmath180 for higher @xmath0 values .
the value of @xmath7 in @xcite is about 0.46 , so @xmath181 $ ] and decreases with @xmath0 rise for @xmath182 gev / c .
this decrease of @xmath183 is not compensated by a corresponding increase of the @xmath81 , that leads to the observed polarization decrease with @xmath0 rise .
if we take @xmath184 , such effect as a decrease of polarization with @xmath0 increasing is not expected .
the results of three other @xmath168 polarization measurements , which have typical @xmath185 , do not reveal a decrease of @xmath169 with @xmath0 rise @xcite .
cccccccc + fit # & 1 & 2 & 3 & 4 + + @xmath186 & @xmath187 & @xmath187 & @xmath188 & @xmath188 + [ 0.1 cm ] @xmath22 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & [email protected] & @xmath146 & [email protected] & 6.090 + [ 0.1 cm ] @xmath77 & 1.0 & 1.0 & 1.0 & 1.0 + [ 0.1 cm ] @xmath189 & [email protected] & [email protected] & 0.66 & 0.66 + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & 0.25 & [email protected] + [ 0.1 cm ] @xmath148 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath149 & [email protected] & 5@xmath14519 & 0.0 & 0.0 + [ 0.1 cm ] @xmath150 & 0.337@xmath145 0.015 & 0.341@xmath145 0.013 & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath152 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath108 & [email protected] & [email protected] & -0.26 & -0.26 + [ 0.1 cm ] @xmath132 & 33.0/17 & 33.6/18 & 0.06/3 & 0.06/3 + [ 0.1 cm ] [ t2_sigma ] cccccccc + fit # & 1 & 2 & 3 & 4 + + @xmath186 & @xmath190 & @xmath190 & @xmath191 & @xmath191 + [ 0.1 cm ] @xmath22 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & [email protected] & 6.090 & @xmath192 & 6.090 + [ 0.1 cm ] @xmath77 & 1.0 & 1.0 & 1.0 & 1.0 + [ 0.1 cm ] @xmath189 & [email protected] & [email protected] & 0.5 & 0.5 + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath148 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath149 & 2 @xmath14512 & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & -0.2 & [email protected] + [ 0.1 cm ] @xmath151 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath152 & [email protected] & [email protected] & 0.0 & 0.0 + [ 0.1 cm ] @xmath108 & [email protected] & [email protected] & -0.12 & -0.12 + [ 0.1 cm ] @xmath132 & 71.0/46 & 71.0/47 & 5.7/13 & 5.7/13 + [ 0.1 cm ] [ t2_xi ] it was also stated in @xcite that an energy dependence of @xmath169 is observed by comparing the results at 800 gev / c ( @xmath170 target ) @xcite , at 400 gev / c ( @xmath162 target ) @xcite and 400 gev / c ( @xmath170 target ) @xcite .
the observed difference is really due to the different targets ( @xmath170 vs @xmath162 ) and slightly different @xmath7 values used for this comparison . in the case of comparison of the data on @xmath162 and @xmath170
targets a strong @xmath35-dependence is the reason of a higher polarization on the @xmath162 target , since @xmath193 ( see fit # 1 ) . for the data on @xmath170 target @xcite the corresponding @xmath194 and @xmath195 , while the 800 gev / c data are measured at @xmath196 and @xmath197 .
the data @xcite show that the @xmath198 increases by 0.05 or more , when @xmath7 is increased from 0.47 to 0.52 .
so , the expected @xmath169 for 800 gev / c and @xmath199 is about @xmath200 , which is compatible with the measured value @xmath195 at 400 gev / c .
the energy independence of the polarization is confirmed by fig .
[ sigp_pa_pt ] , where all the data points at two beam energies and two different targets are approximated well by a single function @xmath81 and are compatible with the scaling law ( [ eq : pol ] ) , described using @xmath0 and two scaling variables @xmath53 and @xmath58 .
the data fit indicates also that @xmath168 polarization decreases on a nuclear target ( @xmath201 ) .
the polarization attenuation is more significant at high @xmath54 values , similar to that of for the @xmath9 hyperons .
we may conclude that the use of two scaling variables ( @xmath53 and @xmath58 ) in the form of eq .
( [ eq : pol ] ) instead of one ( @xmath7 ) resolves the problem of energy dependence of the @xmath168 polarization and the problem of its unusual @xmath0 dependence ( see details in @xcite ) , and presents the existing data in the energy independent way .
the polarization of @xmath202 ( @xmath203 ) has been measured at 400 gev / c in @xmath204 @xcite and @xmath159 @xcite collisions .
since just a few points in @xmath7 and @xmath0 has been measured , some of the fit # 3 parameters are fixed , see table [ t2_sigma ] .
in addition , for the fit # 4 the @xmath8 parameter was fixed : @xmath205 . according to the fit the @xmath203 has a maximum near @xmath206 with corresponding @xmath207 .
the polarization of @xmath208 ( @xmath209 ) produced in @xmath159 has been measured at 28.5 gev / c @xcite and at 18.5 gev / c @xcite .
only two data points are available from these two experiments . at 28.5 gev
/ c the value of @xmath209 is @xmath210 ( @xmath211 , @xmath212 gev / c ) , and at 18.5 gev / c it is @xmath213 ( @xmath214 , @xmath215 gev / c ) , which is consistent with the @xmath216 polarization in the same kinematic area . the polarization of @xmath217 ( @xmath218 ) has been measured at 800 gev / c in @xmath159 @xcite collisions , and at 400 gev / c in @xmath204 @xcite and @xmath159 @xcite collisions . the fit # 1 is made with some parameters fixed .
the value of parameter @xmath8 is @xmath219 . in the fit # 2
the @xmath205 is used and the corresponding curve is shown in fig .
[ xim_pa_xf ] for 400 gev / c @xmath159 collisions and @xmath220 gev / c .
the data shown in fig .
[ xim_pa_xf ] have an additional cut @xmath221 gev / c to reduce the smearing of the data points due to the @xmath0 dependence . the fit # 2 and fig .
[ xim_pa_xf ] data indicate a local maximum in the absolute value of polarization at @xmath222 .
the @xmath35-dependence of the @xmath218 is not significant ( @xmath223 ) .
the polarization of @xmath224 ( @xmath225 ) has been measured at 400 gev / c in @xmath159 @xcite collisions .
the data are presented in fig .
[ xi0_pa_xf ] vs @xmath7 . due to a small number of experimental points some of the fit parameters were fixed near the values which give the best @xmath43 .
the fit # 3 is done with the @xmath8 free and the fit # 4 - with @xmath205 . the fit # 4 and fig . [ xi0_pa_xf ]
data indicate a local maximum in the absolute value of polarization near @xmath196 .
the fits # 2 and # 4 indicate a possible decrease in the absolute value of @xmath225 and @xmath218 for @xmath226 due to a high value of @xmath207 .
this feature makes these reactions different from the @xmath9 production in @xmath18 collisions , where we see a steady increase of polarization with @xmath7 rise .
an interesting observation follows from the analysis of the scaling properties of hyperon ( @xmath9 , @xmath216 , @xmath217 and @xmath224 ) polarization in @xmath126 collisions - the parameter @xmath8 ( `` polarization oscillation frequency '' ) in eqs .
( [ eq : pol])-([eq : f2 ] ) , which describes the rate of change of the polarization with @xmath1 increase , is related with the number of sea quarks ( @xmath227 ) , picked up from the sea during a hyperon production : @xmath228 .
in this section we analyze the hyperon polarization data in @xmath230 collisions .
the corresponding fit parameters are presented in table [ t5_km ] .
since the initial state ( @xmath230 ) is not symmetric vs the rotation transformation and there are no measurements on nuclear targets , some modification of eqs .
( [ eq : pol])-([eq : f2 ] ) is introduced . in particular , the parameters @xmath107 , @xmath108 are fixed at zero values , @xmath140 , and the factor @xmath101 in eq .
( [ eq : x2 ] ) is omitted : @xmath231 the function @xmath232 is given by eq .
( [ eq : f2 ] ) .
the @xmath77 parameter in ( [ eq : pol ] ) could be different from unity due to an asymmetric initial state .
the polarization of @xmath9 ( @xmath94 ) hyperons in @xmath230 collisions has been measured at 176 gev / c @xcite , at 32 gev / c @xcite , at 14.3 gev / c @xcite , and at 10 and 16 gev / c @xcite .
the data with lower beam momenta are not used for this analysis due to the energy dependence of the polarization below 9 gev / c at fixed negative @xmath7 @xcite .
the interesting feature of these @xmath233 beam data is that they include both , the beam and the target fragmentation regions .
that feature allows us to estimate the @xmath77 parameter in eq .
( [ eq : pol ] ) and check that the scaling behaviour is valid for both hemispheres .
only the data with @xmath234 are used for this analysis to avoid a large contribution of exclusive channels near the kinematic limits .
the data fits ( # 1 and # 2 ) have been performed with some parameters fixed due to a limited statistics . in the fit # 1 the parameter @xmath8 is free , and in the fit # 2 its value is fixed at @xmath235 .
both fits have @xmath132 about 1.6 and indicate the maximum near @xmath142 and almost @xmath7-independent polarization for the negative @xmath7 region .
the @xmath77 parameter is @xmath236 in contrast to the @xmath17 data .
the @xmath0 dependence of the @xmath94 is compatible with a linear rise of the @xmath81 ( see eq .
( [ eq : f2 ] ) ) for @xmath237 gev / c .
cccccccc + fit # & 1 & 2 & 3 & 4 + + @xmath186 & @xmath9 & @xmath9 & @xmath217 & @xmath217 + [ 0.1 cm ] @xmath22 & 4.18 @xmath1450.47 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & 3.53 @xmath1450.22 & 3.58 & [email protected] & 3.58 + [ 0.1 cm ] @xmath77 & -0.77 @xmath1450.12 & -0.77 @xmath1450.11 & [email protected] & [email protected] + [ 0.1 cm ] @xmath189 & 1.0 & 1.0 & 1.0 & 1.0 + [ 0.1 cm ] @xmath147 & [email protected] & [email protected]&[email protected] & [email protected] + [ 0.1 cm ] @xmath148 & 0.81 @xmath1450.21 & 0.80 @xmath1450.20 & -0.01 @xmath1450.13 & [email protected] + [ 0.1 cm ] @xmath149 & 26 & 26 & 26 & 26 + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & 0.18 @xmath1450.13 & 0.19 @xmath1450.11 & [email protected] & [email protected] + [ 0.1 cm ] @xmath152 & 7 & 7 & 27 & 27 + [ 0.1 cm ] @xmath132 & 88.8/55 & 88.9/56 & 12.9/11 & 13.3/12 + [ 0.1 cm ] [ t5_km ] the polarization of @xmath217 ( @xmath238 ) hyperons in @xmath230 has been studied at 5 gev / c @xcite and at 4.2 gev / c @xcite .
since the polarization in @xcite is integrated over one of the variables ( @xmath7 or @xmath0 ) , we have to assign mean values for these integrated variables . as estimates of these mean values
we take for the analysis @xmath239 , and @xmath240 gev / c , respectively .
only data with @xmath241 are used for the analysis to exclude the resonance region , in particular in the extreme forward region ( @xmath55 ) which is dominated by the baryon exchange process @xmath242 .
the fit # 3 is made with the @xmath8 parameter free , and the fit # 4 is made for @xmath235 . as in the case of the @xmath9 polarization the @xmath77 parameter is negative , but with a much larger uncertainty .
additional measurements are desirable for @xmath243 gev / c where polarization could reach high values . both , the @xmath7 and the @xmath0 dependencies of @xmath238 are similar to the case of the @xmath9 polarization in @xmath230 collisions .
the @xmath229 polarization in @xmath230 collisions has been measured at 32 gev / c @xcite . since just two @xmath7
points are available , only the magnitude of polarization is estimated which is about 0.4 .
the hyperon polarization data for @xmath244 collisions have been analyzed using eqs .
( [ eq : pol])- ( [ eq : x2n ] ) .
only the data with @xmath245 are used for the analysis .
the fit parameters are presented in table [ tab_kp ] .
the @xmath9 polarization in @xmath244 collisions has been measured at 8.2 and 16 gev / c @xcite , at 13 gev / c @xcite , and at 32 gev / c @xcite .
the fit # 1 is made with the @xmath8 parameter free , and the fit # 2 - with the @xmath235 .
the polarization does not depend on energy and has a maximum near @xmath246 .
the @xmath229 polarization in @xmath244 collisions has been measured at 32 and 70 gev / c @xcite , at 8.2 and 16 gev / c @xcite , at 32 gev / c @xcite , and at 13 gev / c @xcite .
the data @xcite for 32 and 70 gev / c are combined for this analysis and an average momentum 50 gev / c is assigned to it .
the fit # 3 and the fit # 4 have been performed with @xmath8 parameter free and fixed ( @xmath247 ) , respectively .
the fits result in a negative @xmath248 , similar to the @xmath249 case .
the fits indicate the existence of significant phases @xmath97 and @xmath76 , which depend on @xmath0 . the hyperon polarization observed in this reaction is higher than in any other one .
cccccc + fit # & 1 & 2 & 3 & 4 + + @xmath186 & @xmath9 & @xmath9 & @xmath229 & @xmath229 + [ 0.1 cm ] @xmath22 & -3.0 @xmath1451.9 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & 4.4 @xmath1451.5 & 3.58 & [email protected] & 5.55 + [ 0.1 cm ] @xmath77 & 0.30 @xmath1450.18 & 0.36 @xmath1450.28 & [email protected] & [email protected] + [ 0.1 cm ] @xmath189 & 0.4 & 0.4 & 0.4 & 0.4 + [ 0.1 cm ] @xmath147 & -0.19 @xmath145 0.13 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath148 & 0.0 & 0.0 & [email protected] & [email protected] + [ 0.1 cm ] @xmath149 & 0.0 & 0.0 & 17@xmath14528 & 17@xmath14520 + [ 0.1 cm ] @xmath150 & -0.53 @xmath1450.31 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & 0.0 & 0.0 & [email protected] & [email protected] + [ 0.1 cm ] @xmath152 & 0.0 & 0.0 & [email protected] & [email protected] + [ 0.1 cm ] @xmath132 & 9.6/12 & 9.8/13 & 12.4/19 & 13.0/20 + [ 0.1 cm ] [ tab_kp ]
the hyperon polarization data for @xmath252 collisions have been analyzed using eqs .
( [ eq : pol])- ( [ eq : x2n ] ) .
the fit parameters are presented in table [ tab_pi ] .
a cut @xmath245 is used to reduce exclusive reaction contributions .
the @xmath9 polarization in @xmath250 collisions has been measured at 3.95 gev / c @xcite , at 6 gev / c @xcite , at 15 gev / c @xcite , at 16.1 gev / c @xcite and at 18.5 gev / c @xcite .
the fit # 1 is made with the @xmath8 parameter free and the fit # 2 is performed with the @xmath235 .
the polarization is positive in the target fragmentation region and is negative or near zero for positive @xmath7 .
the @xmath9 polarization in @xmath251 collisions has been measured at 18.5 gev / c @xcite .
the fit # 3 is made with the @xmath8 parameter free and the fit # 4 is made with the @xmath235 .
the polarization magnitude for @xmath9 hyperons produced in @xmath252 collisions is smaller than it is in the @xmath17 collisions .
cccccc + fit # & 1 & 2 & 3 & 4 + + @xmath253 & @xmath254 & @xmath254 & @xmath255 & @xmath255 + [ 0.1 cm ] @xmath22 & -1.8 @xmath145 0.5 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & 4.33 @xmath145 0.94 & 3.58 & [email protected] & 3.58 + [ 0.1 cm ] @xmath77 & 1.0 & 1.0 & 1.5 & 1.5 + [ 0.1 cm ] @xmath189 & 1.0 & 1.0 & 1.0 & 1.0 + [ 0.1 cm ] @xmath147 & 0.564 @xmath145 0.088 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath148 & [email protected] & 2.5 @xmath1451.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath149 & 50@xmath14535 & 49@xmath14516 & 0.0 & 0.0 + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & 0.45 @xmath145 0.47 & 0.50 @xmath1450.43 & 0.9 & 0.9 + [ 0.1 cm ] @xmath152 & 12@xmath14511 & 13@xmath14511 & 1.0 & 1.0 + [ 0.1 cm ] @xmath132 & 20.9/19 & 21.2/20 & 0.40/3 & 0.42/4 + [ 0.1 cm ] [ tab_pi ]
the reactions presented in this section can be considered as exotic ones due to a very unusual behaviour of the corresponding polarization with the rise of scaling variables .
this behaviour can not be predicted by the existing models .
most of the theoretical models predict zero polarization for antihyperon production in @xmath126 collision since they do not have valence quarks common with the beam or target hadrons .
the recent experiments have revealed non - zero polarization for @xmath256 @xcite and @xmath257 @xcite hyperons .
other antihyperons also indicate non - zero polarizations with small but not negligible magnitudes .
the fit parameters are presented in table [ t7_anti ] .
the @xmath105 parameter in eq .
( [ eq : pol ] ) is set to zero due to a limited statistics and too few used targets .
the @xmath229 polarization have been measured in @xmath159 collisions at 400 gev / c @xcite and at 800 gev / c @xcite .
although the magnitude of the polarization is very low ( @xmath258 ) , the fit indicates an oscillation of the polarization as a function of @xmath7 with good @xmath259 .
the data fits have been performed for @xmath8 parameter free ( fit # 1 ) and @xmath260 ( fit # 2 ) . due to a limited statistics and kinematic range of the data ,
the @xmath148 parameter is fixed at the same value ( 0.31 ) as it is for @xmath9 polarization in @xmath18 collisions .
as we will see below the large value of @xmath261 is typical for such exotic process ( @xmath262 ) which corresponds to the @xmath263 exchange , where @xmath264 is a baryon number .
the @xmath256 polarization have been measured at 800 gev / c in collisions of a proton beam with a @xmath162 target @xcite .
the data fit has been performed @xmath265 ( fit # 3 ) , which follows from eq .
( [ eq : omega ] ) . at the same time
this value of @xmath8 is a minimal one which describes the unusual polarization dependence on @xmath7 with the middle point closer to zero than the other ones .
the fit function indicates that polarization magnitude could be about 0.18 .
if such an unusual behaviour will be confirmed by future experiments that reaction will be a large challenge to the strong interaction theory . much more data are desirable since we have only 3 data points for this reaction .
the polarization of @xmath229 ( @xmath266 ) hyperons in @xmath120 collisions has been studied at 176 gev / c @xcite .
the fit # 4 is made with @xmath8 parameter free , and the fit # 5 is made for @xmath267 .
the dependence of @xmath266 on @xmath7 is shown in fig .
[ lab_pm_xf ] along with fit # 5 predictions for @xmath268 gev / c and @xmath220 gev / c , respectively .
the predictions are made for 176 gev / c @xmath120 collisions .
both fits indicate local maximums of @xmath269 at @xmath270 and at @xmath271 and an oscillation of @xmath266 as a function of @xmath7 ( @xmath272 ) . an interesting feature of the data is that almost a full period of the oscillations is covered by the data .
this observation needs additional conformation in different kinematic regions due to a limited statistics and the kinematic range covered by the data .
additional measurements are desirable to clarify the @xmath0 dependence of the @xmath266 . the large values of @xmath8 have been found also for @xmath273 and for @xmath274 reactions considered above .
the dependence of the @xmath8 parameter on quantum numbers of the hadrons participating in the reaction will be discussed in details below .
cccccccc + fit # & 1 & 2 & 3 & 4 & 5 + + @xmath275 & @xmath276 & @xmath276 & @xmath277 & @xmath278 & @xmath279 + [ 0.1 cm ] @xmath22 & -0.47 @xmath1450.23 & [email protected] & -16@xmath14512 & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & 18.5 @xmath1455.7 & @xmath280 & 65.19 & [email protected] & @xmath281 + [ 0.1 cm ] @xmath77 & 1 & 1 & 1 & -1 & -1 + [ 0.1 cm ] @xmath189 & 0.34 & 0.34 & 0.4 & [email protected] & [email protected] + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath148 & 0.31 & 0.31 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath149 & [email protected] & [email protected] & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath150 & 0.309 @xmath1450.072 & 0.322 @xmath1450.077 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & [email protected] & 0.10 @xmath1450.16 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath152 & 1.27 & 1.27 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath132 & 12.9/16 & 13.3/17 & 0.00/0 & 0.74/2 & 1.08/3 + [ 0.1 cm ] [ t7_anti ] the @xmath257 polarization has been measured in the @xmath204 collisions at 800 gev / c @xcite .
the fit was not made since only four data points have been measured at fixed @xmath282 .
the sign of polarization is positive and its magnitude is about @xmath283 .
in this section we analyze the polarization of @xmath217 and @xmath284 hyperons produced by a neutral unpolarized beam containing hyperons and the @xmath217 polarization produced by @xmath202 beam . in addition , the @xmath284 and the proton polarizations in @xmath126 collisions are analyzed and compared with the hyperon polarization in different reactions .
the @xmath105 parameter in eq .
( [ eq : pol ] ) is set to zero due to a limited number of used targets and statistics . in this subsection
we consider the data on @xmath217 polarization which have been measured using a neutral unpolarized high energy beam containing @xmath9 and @xmath224 hyperons @xcite .
the primary 800 gev / c proton beam was used to produce a neutral strangeness containing beam , which in its turn interacted with a @xmath162 target .
the average momentum of produced @xmath217 and @xmath284 hyperons was about 395 gev / c .
this value of momentum was used to estimate the momentum of the neutral beam .
it is assumed in this analysis that the ratio ( @xmath285 ) of the neutral beam momentum to the 800 gev / c primary momentum is the same as the ratio of @xmath217 momentum to the neutral beam momentum .
these relations give @xmath286 and the neutral beam momentum about @xmath287 gev / c . using this assumption
we performed a fit of the data @xcite .
the fit parameters are presented in table [ tab_str1 ] .
the polarization of @xmath217 was fitted with @xmath8 parameter free ( fit # 1 ) and with @xmath288 ( fit # 2 ) .
the @xmath7 dependence of polarization is shown in fig .
[ xim_xi0_xf ] along with the fit # 2 predictions for @xmath268 gev / c and @xmath220 gev / c .
though the magnitude of polarization ( @xmath289 ) is about 0.026 only , a clear @xmath7 dependence is seen , which is consistent with the oscillation behaviour , predicted by eqs .
( [ eq : pol])-([eq : x2n ] ) .
the frequency parameter @xmath290 is very large and more than a full period of oscillation is seen in fig .
[ xim_xi0_xf ] .
the @xmath284 polarization has been measured using unpolarized neutral beam described in the previous subsection @xcite .
the polarization of @xmath284 was fitted with @xmath205 ( fit # 3 ) .
the magnitude of @xmath284 polarization is about 0.11 .
the @xmath284 polarization has been measured using 800 gev / c @xmath159 collisions for @xmath291 and @xmath292 gev / c @xcite .
the mean value of polarization in this kinematic range is @xmath293 , but the dependence of it on @xmath7 clearly indicates its oscillation as a function of @xmath7 ( see fig . [ omeg_pp_xf ] ) .
the fit # 4 is performed ( see table [ tab_str1 ] ) with @xmath8 parameter fixed at 22.27 as is predicted by eq .
( [ eq : omega ] ) .
it has to be mentioned that this large value of @xmath8 was estimated using the other reactions and at the same time it is quite consistent with the observed @xmath7 dependence for the @xmath294 reaction .
the magnitude of polarization oscillation is about @xmath295 .
ccccccc + fit # & 1 & 2 & 3 & 4 + + @xmath275 & @xmath296 & @xmath296 & @xmath297 & @xmath298 + [ 0.1 cm ] @xmath22 & 2.37 @xmath1450.44 & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath8 & 46.0 @xmath1453.7 & @xmath299 & @xmath300 & 22.27 + [ 0.1 cm ] @xmath77 & -1.0 & -1.0 & -1.0 & 1.0 + [ 0.1 cm ] @xmath189 & 1.0 & 1.0 & 1.0 & 0.39 + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath148 & 0.0 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath149 & 0.0 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & 0.0 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath152 & 0.0 & 0.0 & 0.0 & 0.0 + [ 0.1 cm ] @xmath132 & 13.75/13 & 13.75/14 & 0.04/1 & 0.01/2
+ [ 0.1 cm ] [ tab_str1 ] the polarization of @xmath217 has been measured using 330 gev / c @xmath202 beam on carbon and copper targets @xcite .
the polarization of @xmath217 for a sample , combining the @xmath302 and the @xmath170 target measurements , was fitted with free @xmath8 parameter ( fit # 1 ) and with @xmath139 ( fit # 2 ) .
the fit parameters are presented in table 9 .
the @xmath217 polarization is negative and its magnitude is about 0.4 .
cccccc + fit # & 1 & 2 & 3 & 4 + + @xmath275 & @xmath303 & @xmath303 & @xmath304 & @xmath304 + [ 0.1 cm ] @xmath22 & [email protected] & [email protected] & 0.74 @xmath1450.14 & [email protected] + [ 0.1 cm ] @xmath8 & [email protected] & @xmath146 & 6.12 @xmath1450.21 & 6.090 + [ 0.1 cm ] @xmath77 & [email protected] & [email protected] & 1.0 & 1.0 + [ 0.1 cm ] @xmath189 & 1.0 & 1.0 & 0.4 & 0.4 + [ 0.1 cm ] @xmath147 & [email protected] & [email protected] & 0.647 @xmath1450.014 & [email protected] + [ 0.1 cm ] @xmath148 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath149 & 4.6 & 4.6 & 1.75 & 1.75 + [ 0.1 cm ] @xmath150 & [email protected] & [email protected] & [email protected] & [email protected] + [ 0.1 cm ] @xmath151 & [email protected] & [email protected] & 1.75 & 1.75 + [ 0.1 cm ] @xmath152 & 1.3 & 1.3 & 1.75 & 1.75 + [ 0.1 cm ] @xmath132 & 10.7/17 & 10.7/18 & 35.9/31 & 35.9/32 + [ 0.1 cm ] [ tab_str2 ] it is interesting to compare the hyperon polarization data with the proton polarization data in @xmath305 collisions .
the data have been measured at 100 , 200 , 300 and 400 gev / c in collisions of a proton beam with protons and a carbon target @xcite .
the data fits have been performed for @xmath8 parameter free ( fit # 3 ) and @xmath205 ( fit # 4 ) .
the value of @xmath306 is consistent with the corresponding values of @xmath8 for @xmath307 reactions , indicating a similar mechanism for the polarization origin .
the @xmath308 dependence of the polarization is mainly due to a significant @xmath308 dependence of the phases @xmath97 and @xmath76 .
the data are consistent with the scaling described by eq .
( [ eq : pol ] ) with @xmath81 given by eq .
( [ eq : f2 ] ) .
there is a very interesting feature of the hyperon polarization , related with the value of the parameter @xmath8 in eqs .
( [ eq : pol])-([eq : x2 ] ) .
the results of the data fits indicate ( see table [ pol_max ] ) that for the hyperon and antihyperon production in @xmath126 , @xmath309 , @xmath252 , @xmath310 , @xmath311 , and @xmath120 collisions the @xmath8 parameter depends on flavors of the projectile , the produced hyperon and the target .
it can be expressed for the reaction @xmath312 using the formula @xmath313 where @xmath314 depends on the quark content of hadrons participating in the reaction , and @xmath315 are fit parameters .
in particular , @xmath316 , \label{eq : q1}\ ] ] @xmath317 \delta(b_{a\bar{c } } ) , \label{eq : q2}\ ] ] @xmath318 \delta(b_{b\bar{c } } ) \ } , \label{eq : q3}\ ] ] @xmath319\delta[n_{q}(a\bar{c})-2 ] + \delta[n_{q}^{s}(a\bar{c})-2]\ } , \label{eq : q4}\ ] ] @xmath320 where @xmath321 and @xmath322 in the above formulas @xmath323 is the minimal number of quarks in the @xmath324 system ( @xmath325 pairs of the same flavor are cancelled , see quark diagram in fig .
[ diag1 ] ) .
the @xmath326 denotes additional ( extra ) quarks and antiquarks , produced in a process above the minimal level when a higher order quark level diagram is used , as in the case of the process @xmath327 . in the last mentioned process @xmath328 , and we assume that the inclusive protons are produced via a single valence quark fragmentation .
so , four additional quarks and antiquarks are produced in this process , similar to the case of @xmath329 process , shown in fig .
[ diag1 ] . for all other processes presented in table [ pol_max ] the @xmath330 and @xmath331 .
the sum @xmath332 $ ] can also be considered as the number of spectator quarks in a process @xmath333 .
we may conclude from the analysis of 22 reactions presented in table 10 that non - zero number of spectator quark @xmath334 is always needed ( which means quark exchange diagram ) to generate a non - zero baryon polarization .
the @xmath335 is the net number of @xmath70 quarks or antiquarks in the @xmath324 system .
the @xmath336 and @xmath337 are the baryon numbers of the @xmath324 and the @xmath338 systems , respectively .
the @xmath339 , @xmath340 and @xmath341 are the corresponding baryon numbers , and @xmath342 is the number of @xmath70 quarks in a produced hadron @xmath22 .
the value @xmath323 is used in the constituent interchange model ( cim ) @xcite which predicts a cross section at @xmath343 of the form @xmath344 the first term of eq .
( [ eq : omega ] ) contributes to the processes @xmath345 , where @xmath264 means baryon ( reactions # 1 - # 5 ) , the second one takes into account the specific properties of the meson induced reactions # 6 - # 12 , the third term is important for the antibaryon production , the @xmath346 takes into account very high `` oscillation frequency '' data for the ( anti)hyperons , containing @xmath347 @xmath70 quarks when the beam hadron has @xmath347 ( @xmath348 ) @xmath70 quarks ( see reactions # 15 - 16 ) . the term @xmath349 is introduced to take into account the reactions @xmath350 , for which all other terms are zero
the preliminary analysis of analyzing power for these reactions gives @xmath351 ( see also @xcite ) .
a detailed study of the meson induced reactions will be made in a separate paper .
the reactions in the table [ pol_max ] can be classified according to the set of @xmath314 values .
the most simple is the case of baryon production in @xmath126 collisions when only the @xmath352 term is non - zero and the @xmath353 is related with the number of spectator quarks as is illustrated by a quark diagram in fig .
[ diag1 ] .
we may suggest that the mean colour field created by the spectator quarks in these reactions is the reason of polarization oscillation as a function of @xmath7 .
the @xmath8 values presented in table [ pol_max ] were fitted using eq .
( [ eq : omega ] ) and the parameters @xmath354 are shown in table [ omega_fit ] .
the reactions # 11,15,18 and 20 - 22 were not used in the fit , since they have too few data points .
if the inclusive protons are produced via the single beam quark fragmentation similar to the reaction # 2 , then for the number of spectator quarks @xmath355 the predicted value @xmath356 is in a good agreement with the estimated value @xmath357 .
the dependence of the @xmath8 parameter vs @xmath358 is shown in fig .
[ ome_vs_omeq ] for 15 different reactions , presented in table [ pol_max ] .
the arrows indicate also the predictions of the @xmath359 for the analyzing power in the reactions @xmath360 , @xmath361 in the reactions @xmath362 , and @xmath363 for polarization in the reaction @xmath364 .
cccccccccc + + # & reaction & @xmath365 & @xmath366 & @xmath367 & @xmath368 & @xmath8 & @xmath358 & @xmath369 & @xmath370 + + 1 & @xmath371 & 2 & 0 & 0 & 0 & [email protected] & 3.045 & [email protected] & -0.275 + [ 0.1 cm ] 2 & @xmath372 & 4 & 0 & 0 & 0 & [email protected] & 6.090 & [email protected] & -0.138 + [ 0.1 cm ] 3 & @xmath373 & 4 & 0 & 0 & 0 & [email protected] & 6.090 & [email protected] & -0.138 + [ 0.1 cm ] 4 & @xmath374 & 2 & 0 & 0 & 0 & [email protected] & 3.045 & [email protected] & 0.275 + [ 0.1 cm ] 5 & @xmath375 & 4 & 0 & 0 & 0 & @xmath3764.0 & 6.090 & [email protected] & 0.275 + [ 0.1 cm ] 6 & @xmath249 & 3 & -1 & 0 & 0 & @xmath3770.22 & 3.58 & [email protected] & 0.604 + [ 0.1 cm ] 7 & @xmath378 & 3 & -1 & 0 & 0 & @xmath3790.88 & 3.58 & [email protected] & 0.604 + [ 0.1 cm ] 8 & @xmath380 & 3 & -1 & 0 & 0 & @xmath3811.5 & 3.58 & @xmath3820.15 & -0.052 + [ 0.1 cm ] 9 & @xmath383 & 3 & -1 & 0 & 0 & @xmath3840.94 & 3.58 & @xmath3820.14 & -0.178 + [ 0.1 cm ] 10 & @xmath385 & 3 & -1 & 0 & 0 & @xmath3866.5 & 3.58 & @xmath3870.17 & -0.178 + [ 0.1 cm ] 11 & @xmath388 & 3 & 1 & 0 & 0 & & 5.55 & [email protected] & 0.302 + [ 0.1 cm ] 12 & @xmath389 & 3 & 1 & 0 & 0 & @xmath3811.9 & 5.55 & [email protected] & 0.604 + [ 0.1 cm ] 13 & @xmath390 & 6 & 2 & 2 & 0 & @xmath3915.7 & 22.27 & @xmath3920.010 & -0.025 + [ 0.1 cm ] 14 & @xmath393 & 2 & 0 & 2 & 0 & [email protected] & 14.21 & @xmath394 0.046 & 0.194 + [ 0.1 cm ] 15 & @xmath395 & 6 & 2 & 2 & 1 & & 65.19 & @xmath3960.03 & -0.188 + [ 0.1 cm ] 16 & @xmath397 & 2 & 0 & 0 & 1 & [email protected] & 45.97 & @xmath3980.005 & 0.022 + [ 0.1 cm ] 17 & @xmath399 & 2 & 0 & 0 & 0 & [email protected] & 3.045 & [email protected] & -0.275 + [ 0.1 cm ] 18 & @xmath400 & 4 & 0 & 0 & 0 & & 6.090 & [email protected] & 0.132 + [ 0.1 cm ] 19 & @xmath401 & 4 & 0 & 0 & 0 & [email protected] & 6.090 & @xmath4020.014 & 0.072 + [ 0.1 cm ] 20 & @xmath403 & 6 & 2 & 2 & 0 & & 22.27 & [email protected] & 0.031 + [ 0.1 cm ] 21 & @xmath404 & 6 & 2 & 2 & 0 & & 22.27 & [email protected] & 0.092 + [ 0.1 cm ] 22 & @xmath405 & 2 & 0 & 0 & 0 & & 3.045 & [email protected] & 0.275 + [ 0.1 cm ] [ pol_max ] the line in fig .
[ ome_vs_omeq ] shows the result of the fit @xmath406 with @xmath407 and @xmath408 .
this figure confirms a strong correlation of the @xmath8 parameter with the quantum numbers which characterize the reaction .
the magnitude of hyperon polarization ( @xmath369 ) varies significantly with the hyperon , the projectile and the target flavors and the target atomic weight .
the values of @xmath369 for different reactions were estimated for @xmath409 , where most of the data have been measured , using eqs .
( [ eq : pol])-([eq : x2n ] ) and fit parameters , taken from tables 1 - 9 .
the sign of the @xmath369 is indicated the same as that for the experimental data if it does not vary in the mentioned above region , and @xmath145 - otherwise .
the results are presented in table 10 for 22 different reactions .
it is easy to notice from table [ pol_max ] that there is a correlation between the value of @xmath369 and the @xmath8 parameter for the corresponding reactions .
the product of these two values , @xmath410 varies much less than each value separately .
.the fit parameters of eqs .
( 21 ) and ( 32 ) .
[ cols="^,^,^,^,^,^,^,^,^ " , ] [ omega_fit ] the @xmath77 parameter was found to be consistent with @xmath411 where @xmath412 where @xmath413 , @xmath414 , and @xmath415 are hypercharges ( @xmath416 ) of @xmath20 , @xmath22 , and strangeness change , respectively .
the @xmath417 is the number of quarks with parallel spins in the hadron @xmath22 .
it is assumed here , in accordance with the @xmath418 quark model @xcite , that @xmath419 for the @xmath284 , @xmath420 for the @xmath166 , @xmath421 , protons , neutrons , and @xmath422 for the @xmath9 .
the magnitude @xmath369 of the polarization can be approximated by @xmath423 where the value of @xmath424 is given by eq .
@xmath425 where @xmath426 @xmath427 + 2\delta(s_{c}+3)\delta(s_{a}+1 ) \label{eq : q2}\ ] ] @xmath428 + \delta[n_{q}(a\bar{c})-6 ] \delta(b_{b\bar{c } } ) \ } , \label{eq : q3}\ ] ] @xmath429\delta[n_{q}(a\bar{c})-2 ] + \delta[n_{q}^{s}(a\bar{c})-2]\ } , \label{eq : q4}\ ] ] @xmath430 the @xmath431 and @xmath432 are the fit parameters , shown in table 11 .
the fit of @xmath433 using @xmath434 ( the sign of @xmath369 was not taken into account ) was made for the reactions # 1 - # 22 .
the values of @xmath435 are presented in table [ pol_max ] with the sign in the region @xmath436 given by eq .
( [ eq : p_q ] ) .
the magnitude of polarization in the reaction @xmath437 is @xmath438 , see eq .
( [ eq : p_q ] ) . for the @xmath217 and @xmath224 hyperon production in @xmath17 collisions the magnitude of polarization is @xmath439 due to the factor @xmath440 in eq .
( [ eq : phi_q ] ) .
the @xmath189 parameter was found to be consistent with an approximation @xmath441 , \label{eq : pt0}\ ] ] where @xmath442 gev / c .
( [ eq : omega ] ) and ( [ eq : p_q ] ) should have their explanations in the models of strong interaction and ultimately in the qcd , perhaps , in the non - pertubative approach .
the value of parameter @xmath8 can also be estimated for the processes in which the analyzing power is measured . that requires additional precise measurements in a wide range of @xmath7 in a kinematic area where the @xmath25 scaling is fulfilled ( @xmath443 1 gev / c and/or the beam energy @xmath23 gev ) .
the reactions with meson beams are of special interest because of a high value of the @xmath8 parameter and a possibility of `` oscillation '' of the analyzing power as a function of @xmath7 ( compare figs .
[ lab_pm_xf]- [ omeg_pp_xf ] and the preliminary results in @xcite ) .
if we assume that eqs .
( [ eq : pol])-([eq : omega ] ) are valid also for the analyzing power @xmath83 , we can make predictions of the @xmath8 for different reactions .
in particular , we expect @xmath359 for the @xmath444 , @xmath445 , @xmath446 , @xmath447 , @xmath448 , @xmath8 production in @xmath17 and @xmath120 collisions ; @xmath139 for the @xmath9 production in @xmath17 collisions ; @xmath205 for the proton production in @xmath17 collisions , @xmath361 for @xmath362 , and @xmath260 for the antiproton and @xmath229 production in @xmath17 collisions .
these predictions include about two tens different reactions and a preliminary analysis confirms these predictions .
we may conclude that eq .
( [ eq : omega ] ) has a high predictive power which allows one to predict the @xmath8 parameter for about four tens different spin - dependent processes .
it is highly unlikely that the @xmath8 value in so many different processes is described by the same eq .
( [ eq : omega ] ) for accidental reasons .
moreover , within the experimental errors we may fix two out of five parameters : @xmath449 , and @xmath450 .
( [ eq : omega ] ) also predicts that in some cases a target particle plays a significant role in the dynamics of a hyperon production .
this is the case when the baryon number @xmath337 and @xmath451 are different from zero .
the scaling behaviour of the @xmath83 and @xmath6 indicates that the corresponding processes take place at the quark or parton level . in case of the discussed above scaling for the analyzing power and for the hyperon polarization
such constituents could be the constituent quarks or the current quarks ( see related discussion in @xcite ) . in order to resolve quark degrees of freedom inside a hadron ,
the transverse momentum @xmath0 in a process should be higher than @xmath189 .
that could be the reason for the @xmath0 dependence of polarization ( [ eq : f2 ] ) .
the results of the previous sections indicate that the use of two scaling variables @xmath452 is essential for the universal energy independent description of the existing hyperon polarization data in inclusive reactions .
the scaling variables transform into each other under rotation transformation around the normal to the scattering plane and allow one to satisfy in a natural way the feature ( [ eq : pasym ] ) , which is also related with the rotation invariance .
the variables @xmath1 treat on a more equal basis the data for the central and the fragmentation regions as well as the transverse and the longitudinal momentum components of the produced hyperon .
the form of eq .
( [ eq : pol ] ) is chosen from the dimensional analysis , the rotation invariance requirement and it is also motivated by the existing data for both the hyperon polarization and the analyzing power in inclusive reactions for hadron - hadron collisions .
the derived formulas are applicable for both the @xmath83 and the @xmath6 data approximation and reflect the scaling properties of these two classes of processes at high enough energies and @xmath0 .
the specific range of energies and @xmath0 at which the scaling properties are valid could depend on the process type .
the existence of the above scaling implies that large spin effects will survive at high energies and do not depend directly on the beam energy .
these properties of the @xmath83 and the @xmath6 are waiting for their explanation in the strong interaction theory in general and in the quantum chromodynamics ( qcd ) in particular .
it is important to take into account in the future models a deep analogy between the hyperon polarization and the analyzing power , which follows from the results of this study and @xcite .
below we list some common features of the @xmath83 and the @xmath6 , which indicate a common nature of both phenomena .
features of the analyzing power in inclusive hadron production by polarized protons and antiprotons and features of the hyperon polarization in inclusive hadron - nucleon collisions : * scaling behaviour for the hyperon polarization and the analyzing power as a function of @xmath453 and @xmath308 in the beam fragmentation region , and in the central region . *
the @xmath6 and the @xmath454 are approximated by a product of functions of @xmath0 and @xmath1 : @xmath455 $ ] , where the @xmath76 could depend on @xmath0 and a reaction type . *
the function @xmath81 rises with @xmath0 at small @xmath456 and have a plateau or decrease above @xmath189 , where @xmath457 0.4 - 2 gev / c depends on a reaction type . *
the function @xmath82 is proportional to @xmath458 $ ] , where @xmath97 could depend on @xmath0 .
the @xmath8 is a universal function of quantum numbers characterizing the reaction , common for the @xmath6 and @xmath454 . *
the @xmath6 and the @xmath454 are zero at @xmath308 = 0 due to the absence of a preferable direction , such as a normal to the scattering plane .
this implies @xmath459 .
* the sign and the magnitude of the @xmath6 and the @xmath454 depend on the projectile , the target , and the produced hadron flavors .
we may conclude that a close similarity of the kinematic behaviour of @xmath6 and @xmath454 and a universal dependence of the @xmath8 parameter on quantum numbers indicate on a common origin of spin effects in these two classes of spin - dependent reactions .
the polarization of hadrons is a pure quantum mechanics effect related with the interference of spin - flip ( @xmath460 ) and spin - nonflip ( @xmath461 ) amplitudes .
the transverse hadron polarization can be expressed via the @xmath461 and @xmath460 amplitudes as @xcite @xmath462 taking into account eqs .
( [ eq : pol])- ( [ eq : gx ] ) we can choose amplitudes @xmath461 and @xmath460 as @xmath463 + ( \sqrt{\sigma})^{*}exp[-i\omega ( x_{a- } - x_{1 } + x_{2})/2 ] \bigr\ } , \label{eq : f}\ ] ] @xmath464 - \sqrt{\sigma}exp[i\omega(x_{a- } - x_{1 } + x_{2})/2 ] \bigr\ } , \label{eq : g}\ ] ] where @xmath465 and @xmath466 are the functions of kinematic variables ( @xmath7 , @xmath0 , @xmath467 ) with zero relative phases , which have to satisfy the constrains followed from both the polarization and the cross section data . the generalized optical theorem predicts the following relation for a cross section : @xmath468 where @xmath469 is the corresponding unpolarized inclusive cross section @xcite , which can be used together with eq .
( [ eq : im_fg ] ) to fix the @xmath465 and @xmath466 . to have a non - zero value of the @xmath6 both amplitudes have to be non - zero and the phase difference @xmath470 between spin - flip and spin - nonflip amplitudes has to be non - zero too . for the cases of @xmath471
the following phase differences are expected from eqs .
( [ eq : f])-([eq : g ] ) : @xmath472 as we can see from eq .
( [ eq : dfg ] ) , the variables @xmath49 and @xmath8 play an important role in the hyperon polarization phenomena since they determinate the phase difference between the spin - flip and non - flip amplitudes .
the higher is the @xmath8 value , the larger the @xmath470 change rate with the @xmath49 increase .
( [ eq : omega ] ) for @xmath358 can be considered as a sum of effective `` charges '' , which create a mean field and lead to the change of the phase difference @xmath470 . in the lowest - order pertubative @xmath473
all amplitudes are relatively real .
this tends to rule out polarization in the hard scattering of partons , which seems to be well described in the low - order @xmath473 .
the observation of undiminished polarization near @xmath474 gev / c implies that either pertubative @xmath473 does not apply or that another mechanism is responsible , such as interference of exited states or the fragmentation process @xcite .
several phenomenological models have been proposed to explain the hyperon polarization data and the analyzing power data ( see recent review in @xcite ) .
some of the models have the features that allow one to understand , at least at a qualitative level , the analogy between the @xmath83 and the @xmath6 discussed above .
one class of such models assumes that an orbital motion of valence quarks and surface effects are responsible for the correlation between the quark spin direction and transverse motion of produced hadron @xcite . by taking the @xmath9 s containing two , one or zero valence quark(s ) of a beam proton into account ,
the model predicts the sign and the @xmath7 dependence of @xmath9 s polarization .
the model predicts correctly the sign and the magnitude of @xmath202 , @xmath166 polarization in @xmath17 collisions , and @xmath9 polarization in @xmath230 collisions .
it also predicts a smaller magnitude for @xmath9 s produced by @xmath16 beams .
the authors of the model claim an analogy of mechanisms , which lead to non - zero hyperon polarization and analyzing power .
the model does not explain the specific features of antihyperon polarization .
a separate approach was developed by troshin and tyurin , which assumes the rotation of a quark - antiquark cloud inside constituent quarks @xcite .
the main role belongs to the orbital angular momentum and polarization of the strange quark - antiquark pairs in the internal structure of constituent quarks .
the hyperons are produced in two stages .
at the first stage the overlapping and interaction of peripheral clouds occur which results in massive quark appearance and a mean field is generated .
constituent quarks located in the central part of hadron are supposed to scatter in a quasi - independent way by this mean field . at the second stage
two mechanisms take place : recombination of the constituent quarks with a virtual massive strange quark ( soft interaction ) into a hyperon or a scattering of a constituent quark in the mean field , excitation of this constituent quark , appearance of a strange quark as a result of decay of the constituent quark and a subsequent fragmentation of a strange quark into a hyperon ( high @xmath0 s hard interaction ) .
the resulting expression at @xmath475 gev / c when hard interactions dominate is @xmath476 , \label{eq : trosh}\ ] ] where @xmath477 is the polarization of the constituent quark @xmath478 which arises due to multiple scattering in the mean field and @xmath479 is the mean value of an internal angular momentum inside the constituent quark .
thus , in this model the polarization of a strange quark is the result of multiple scattering of a parent constituent quark , the correlation between the polarization of a strange quark and the polarization of the constituent quark and a local compensation of a spin and an orbital angular momentum of a strange quark .
the simplest possible @xmath480 dependence of @xmath477 is taken @xmath481 where @xmath482 .
the model predicts the negative sign and @xmath7 dependence of the @xmath9 polarization
( [ eq : trosh ] ) resembles eq .
( [ eq : gx ] ) , especially in the beam fragmentation region , with effective @xmath483 . eqs .
( [ eq : trosh ] ) , ( [ eq : trosh1 ] ) predicts a scaling behaviour of the hyperon polarization .
the concept of the mean field , generated by quarks , which leads to a hyperon polarization and to its change with @xmath480 rise similar to eq .
[ eq : gx ] is also in consent with the analysis , presented above .
there are no predictions for other hyperons , though the authors assume zero polarization in inclusive process @xmath484 due to a low probability of multiple scattering in the mean field in comparison with a single scattering .
a single scattering does not polarize quarks and protons appear unpolarized in the final state since a single scattering is dominant in this process .
there are several semiclassical models , which provide simple arguments for a qualitative description of the hyperon polarization , but since they fully ignore the relevance of the phase difference , which is crucial , they are unable to make solid quantitative predictions . in the recombination model
@xcite a dynamical reason for the spin - momentum correlation is explained by the effect of thomas precession @xcite .
the effect arises when the direction of the force acting on a quark does not coincide with the direction of its motion .
it leads to a rotation of the quark spin and could be the reason of the discussed above `` oscillation '' of polarization or analyzing power as a function of @xmath7 .
the thomas frequency is an inverse function of a quark mass @xmath485 where @xmath486 is the strange quark velocity , @xmath487 - the force , @xmath488 - the strange quark mass , and @xmath489 .
an additional term will appear in the effective hamiltonian which describes the recombination process @xmath490 where @xmath491 is a spin of the quark . within the old - fashioned perturbation theory
the final expression for the @xmath9 polarization is @xmath492 where it is assumed that a recombination time @xmath493 , the average momentum of the quark is @xmath494 and @xmath495 gev@xmath496 is a distance scale of the order of the proton radius .
the @xmath497 gev / c@xmath498 is an effective mass and the @xmath499 is a fraction of a proton momentum ( @xmath500 ) which carries the @xmath70 quark @xcite .
these assumptions lead to a quark mass cancellation in the polarization formula ( [ eq : thom3 ] ) and a scaling behaviour of the @xmath9 polarization .
the model gives the right sign and a good approximation of the @xmath7 dependence for the @xmath9 polarization .
there are also many predictions for hyperon polarization in other reactions .
they are based on some rules which are formulated within the framework of the recombination model . in particular , there is a statement that the effect of recombination of the partons in the proton as they are transferred into the outgoing hadron may be different depending on whether they are accelerated ( as are the slow sea partons ) or decelerated ( as are the fast valence partons ) .
it is also a statement that two partons with similar wave functions in the proton may interact with themselves differently not as they interact with a parton whose wave function is different .
this results in a simple rule : slow partons preferentially recombine with their spins down in the scattering plane while fast partons recombine with their spins up .
the model predicts correctly the polarization sign but not the magnitude for some of the reaction .
so , the model predicts the same polarization magnitude for @xmath501 as it is for @xmath502 , but we know from the above analysis that the polarization magnitudes for @xmath503 are two times smaller .
similarly , for @xmath249 process the model predicts the same magnitude as for @xmath501 , while the measured value is two times larger . since the polarization in the model is essentially kinematic at the quark level all the antibaryons should have zero polarization .
as we have seen in the above analysis , the experimental situation is much more complicated .
there is another estimate of the hadronization time which follows from the analysis of @xmath35-dependence of hadron production @xmath504 , where @xmath505 gev @xcite .
the use of this @xmath506 estimate results in a different expression for the @xmath9 polarization @xmath507 which is an inverse function of the @xmath70 quark mass @xmath488 .
this example shows that the hyperon polarization is sensitive to the details of the hadronization process .
these results seem to imply that the origin of hyperon polarization is closely related with the confinement mechanism .
since the thomas precession frequency is an inverse function of the quark mass @xmath488 , this may be the reason for a large variation of the @xmath358 parameter for different reactions . the ratio @xmath508 is estimated to be from 17 to 25 with a mean about 21 @xcite .
the same order of magnitude is given by the ratio @xmath509 of the @xmath358 parameters for the processes @xmath510 and @xmath501 , in which @xmath70 and @xmath511 quarks play an important role .
it is interesting to estimate the total rotation angle of a quark spin using the above approximations @xmath512 taking typical @xmath142 , @xmath513 , @xmath514 gev / c , and @xmath515 mev / c@xmath516 , we have @xmath517 , while for the @xmath518 quark with @xmath519 mev / c@xmath516 the @xmath520 .
so , the total rotation angle of a quark spin due to the thomas precession could be rather large and varies in the same range as that of the @xmath358 . in case of @xmath217 and @xmath224 production in @xmath17 collisions the effective field created by the spectator quarks @xmath521
is expected to be two times higher than for @xmath9 production and the corresponding rotation angle will be also two times larger , in agreement with the @xmath8 value for @xmath217 and @xmath224 production .
another explanation of spin - momentum correlation follows from a picture of a colour flux tube , which emerges after the collision between an outgoing quark and the rest of hadronic system @xcite .
the @xmath418 wave function is assumed for hadrons , in particular , for @xmath9 the @xmath522 system is in a singlet state , so the @xmath9 polarization is that of the @xmath70 quark .
an outgoing @xmath523 diquark with spin @xmath524 and isospin @xmath525 stretches the colour field and a @xmath526 pair is produced .
it is assumed that the @xmath70 quark has @xmath0 which must be locally compensated by that of the @xmath527 quark . as a result
, the @xmath526 pair has an orbital momentum which is assumed to be balanced by the spin of the @xmath526 pair .
the model predicts a negative @xmath9 polarization in @xmath17 collisions but can not predict its magnitude or @xmath7 dependence .
the @xmath0 dependence of the polarization is linear .
the model needs additional assumptions to explain the polarization in other reactions and fails to explain the antihyperon polarization .
we propose a simple toy model which uses an analogy with the optics .
let us consider the @xmath217 production in a collision of two protons ( @xmath20 and @xmath21 ) in their c.m .
reference frame .
the proton s longitudinal size is about @xmath528 , where @xmath529 fm is a proton radius and @xmath530 . in an optical picture
the phase can be related with the number of scattering centers @xcite .
we assume here that a hadron can be characterized by an effective refractive index ( @xmath531 ) which leads to a phase difference @xmath532 between spin - flip and spin - nonflip quark scattering amplitudes , where @xmath518 is a total path length inside a proton and @xmath533 is a quark momentum .
it is assumed here for the sake of simplicity that a quark from the proton @xmath20 passes on average half of the proton @xmath21 thickness and then changes its angle due to a scattering in the proton @xmath21 .
the second part of its way inside the proton is approximately by a factor of @xmath534 larger than that before the scattering ( we consider here not too large scattering angles ) .
this results in a phase difference @xmath535 where @xmath536 fm is the proton compton wavelength .
( [ eq : phase ] ) can be rewritten as @xmath537 , where @xmath538 and @xmath539 is the mean fraction of a quark momentum which is carried by the produced hyperon . a similar consideration of the proton @xmath21 quark scattering inside the proton @xmath20 results in a @xmath540 and the total contribution into the @xmath217 polarization is @xmath541 , \label{eq : pol_eff}\ ] ] which is very similar to eq .
( [ eq : pol ] ) .
the averaging over the transverse quark coordinates inside a proton is not taken into account for the sake of simplicity . a more careful consideration of a quark path length after the scattering removes the singularity @xmath542 in eq .
( [ eq : pol_eff ] ) since the path length is limited at @xmath543 by the @xmath544 .
the @xmath0 dependence of a hyperon polarization is also not taken into account in eq .
( [ eq : pol_eff ] ) since we assume that @xmath0 is high enough to resolve quarks inside the hadron structure .
the condition for that is @xmath545 gev / c , where @xmath546 is a constituent quark radius .
comparison of eqs .
( [ eq : omeg_eff ] ) and ( [ eq : omega ] ) assumes that @xmath547 .
we may learn from this toy model that the hyperon polarization oscillation is probably related with a corresponding scattering amplitude phase change due to the hadron mean field generated during hadron interaction .
the field strength is proportional to the number of scattering centers @xmath323 in agreement with the above analysis of experimental data .
the scaling variables @xmath53 and @xmath58 arise in this model from the consideration of geometrical and relativistic properties of interacting hadrons and the assumption that the phase difference @xmath548 is proportional to the quark path length in the mean hadron field .
the analysis of experimental data on the hyperon and antihyperon polarization has been made .
it is shown that the existing ( anti)hyperon polarization data in inclusive reactions for @xmath126 , @xmath252 , @xmath309 , @xmath120 and hyperon - nucleon collisions can be described by a function of @xmath0 and two scaling variables @xmath137 : @xmath549 $ ] . the function @xmath82 is approximated in the scaling limit by a simple expression @xmath550 $ ] which results in an oscillation of polarization in case of large @xmath8 value .
functions @xmath81 , @xmath97 , and @xmath76 depend on @xmath0 and can be approximated by constants above 1 - 2 gev / c .
it is found that the @xmath8 parameter is consistent with an universal dependence on quark composition ( eq .
( [ eq : omega ] ) ) for about two tens measured ( anti)baryon production processes .
in addition , the preliminary analysis indicates the validity of the same eq .
( [ eq : omega ] ) for about two dozen spin - dependent processes in which analyzing power of produced hadrons was measured .
universal character of the @xmath8 dependence on quantum numbers allows one to predict the @xmath8 values for a large number of processes in which analyzing power or hadron polarization is measured .
it indicates also that we deal with a physical case and not just with data description .
the polarization of @xmath229 in @xmath120 collisions is fitted well by eq .
( [ eq : pol ] ) and indicates an oscillation of it as a function of @xmath7 .
similar oscillations with large @xmath8 parameter are also seen for other processes , including @xmath296 and @xmath298 .
the data fits indicate also that the @xmath9 hyperon polarization decreases on nuclear targets according to the law @xmath556 , with @xmath557 .
this effect may be related with the rescattering of polarized @xmath70 quarks in the nuclear matter before formation of a hyperon is over .
since the formation length is proportional to the final hyperon momentum , we expect a rise of @xmath70 quark rescattering probability with @xmath54 increase .
due to the similarity between the hyperon polarization and the analyzing power features we expect that the @xmath35-dependence of the analyzing power is also described by the law @xmath558 with @xmath559 .
the polarization sign and its magnitude depend on quark composition of hadrons participating in the reaction and can be predicted using the proposed formulas . in most cases
the processes with large @xmath8 value have a small magnitude which makes it difficult to measure .
there is an analogy between the scaling properties of polarization for hyperons , produced in collisions of unpolarized hadrons , and the scaling properties of the analyzing power of hadrons , produced in the collisions of polarized protons ( antiprotons ) with hadrons .
this analogy between the analyzing power and the hyperon polarization indicates on a common origin of both phenomena .
the possible mechanism may be related with the confinment forces .
these forces are created by the mean colour field generated after the initial hard scattering of quarks .
the quark spin rotation in the mean field could lead to the phase difference between spin - flip and spin - nonflip amplitudes which results in the hyperon polarization .
99 g. bunce et al .
* 36 * ( 1976 ) 1113 .
f. abe et al . , _ phys .
d34 * ( 1986 ) 1950 . f. abe et al . , _ phys . rev .
* 50 * ( 1983 ) 1102 .
bonner et al . , _ phys . rev .
_ * d38 * ( 1988 ) 729 .
dukes et al . ,
_ phys.lett .
_ * b193 * ( 1987 ) 135 .
k. raychaudhuri et al . , _ phys .
* 90 b * ( 1980 ) 319 .
gourlay et al . , _ phys .
lett . _ * 56 * ( 1986 ) 2244 . c. wilkinson et al . , _ phys .
* 46 * ( 1981 ) 803 .
k. heller et al . , _ phys . rev .
* 51 * ( 1983 ) 2025 .
k. heller et al . , _ phys .
* 41 * ( 1978 ) 607 . b. lundberg et al . , _ phys .
* d 40 * ( 1989 ) 3557 .
ramberg et al . , _ phys .
* 338 b * ( 1994 ) 403 .
a.m. smith et al .
* 185 * ( 1987 ) 209 .
a. morelos et al . , _ phys .
* 71 * ( 1993 ) 2172 . c. wilkinson et al .
* 58 * ( 1987 ) 855 . c. ankenbrandt et al .
* 51 * ( 1983 ) 863 .
wah et al . , _ phys .
* 55 * ( 1985 ) 2551 .
l. deck et al .
_ * d 28 * ( 1983 ) 1 .
bonner et al . , _ phys .
* 62 * ( 1989 ) 1591 .
j. duryea et al . , _ phys . rev .
* 67 * ( 1991 ) 1193 .
p.m. ho et al .
_ * 65 * ( 1990 ) 1713 .
trost et al . , _ phys .
* d 40 * ( 1989 ) 1703 .
r. rameika et al .
* d 33 * ( 1986 ) 3172 .
m.l.faccini-turleur et al . , _
z. phys . _
* c 1 * ( 1979 ) 19 .
a.borg et al . , _
nuovo cimento _ * 22 a * ( 1974 ) 559 .
h. abramowicz et al . , _ nucl .
* b 105 * ( 1976 ) 222 .
h. grassler et al . , _ nucl .
* b 136 * ( 1978 ) 386 .
m. baubillier et al . , _ nucl .
* b 148 * ( 1979 ) 18 .
j. bensinger et al .
* b 252 * ( 1985 ) 561 .
ganguli et al . , _ nucl .
* b 128 * ( 1977 ) 408 .
chliapnikov et al . , _ nucl .
* b 112 * ( 1976 ) 1 . w. barletta et al . , _ nucl .
* b 51 * ( 1973 ) 499 .
ajinenko et al . , _ phys .
_ * 121 b * ( 1983 ) 183 .
b. adeva et al . , _ z. phys . _
* c 26 * ( 1984 ) 359 .
r. sugahara et al .
* b 156 * ( 1979 ) 237 .
f. barreiro et al . , _ phys .
_ * d 17 * ( 1978 ) 669 .
j. bensinger et al .
lett . _ * 50 * ( 1983 ) 313 .
stuntebeck et al . , _ phys .
* d 9 * ( 1974 ) 608 .
k. heller , in proceeding of the 12th international symposium on high - energy spin physics , september 10 - 14 , 1996 .
amsterdam , the netherlands . ed . by c.w .
de jager et al .
world sci .
, singapore . ( p. 23 ) 1996 .
luk et al . , _ phys .
* 70 * ( 1993 ) 900 .
adamovich et al . , _ z. phys .
_ * a 350 * ( 1995 ) 379 .
polvado et al . , _ phys .
* 42 * ( 1979 ) 1325 .
kane , j. pumplin and w. repko , _ phys .
* 41 * ( 1978 ) 1689 .
j.lach , hyperon polarization and magnetic moments .
preprint fermilab - conf-93/381 .
j.soffer , is the riddle of the hyperon polarization solved ? , ( marseille , cpt ) .
preprint cpt-99-p-3898 , sep .
invited talk at hyperon 99 : hyperon physics symposium , batavia , illinois , 27 - 29 sep .
p. 121 ; hep - ph/9911373 ( 1999 ) .
abramov , a new scaling for single - spin asymmetry in meson and baryon hadroproduction .
ihep preprint 98 - 84 , protvino , 1998 ; hep - ph/0110152 .
abramov , _ eur .
j. _ * c 14 * ( 2000 ) 427 ; doi 10.1007/s100529900355 .
liang zuo - tang and c.boros , _ phys .
* 79 * ( 1997 ) 3608 .
bushnin et al . , _ phys .
* 29 * ( 1969 ) 48 .
j. felix et al .
* 76 * ( 1996 ) 22 .
r. blankenbeckler and s.j .
brodsky , _ phys .
rev . _ * d 10 * ( 1974 ) 2973 .
j. qui and g. sterman , preprint itp - sb-98 - 28 ; _ phys .
rev . _ * d 59 * ( 1999 ) 014004 ; hep - ph/9806356 .
single transverse spin asymmetries in hadronic pion production .
j. franklin , _ phys .
* 172 * ( 1968 ) 1807 . c. boros , liang zuo - tang , _ phys .
* d 57 * ( 1998 ) 4491 .
ryskin , polarization phenomena and confinement forces , in proc . of the int .
on quark confinement and the hadron spectrum , como , italy , 20 - 24 june 1994 . edited by n.brambilla and g.m.prosperi .
river edge , n.j . , world scientific , 1995 , p. 261 .
troshin and n.e .
tyurin , _ phys .
rev . _ * d 55 * ( 1997 ) 1265 .
troshin and n.e .
tyurin , _ phys .
* d 52 * ( 1995 ) 3862 .
degrand , h. miettinen , _ phys .
* d 24 * ( 1981 ) 2419 .
degrand et al .
* d 32 * ( 1985 ) 2445 .
thomas , _ philos . mag . _ * 3 * ( 1927 ) 1 .
logunov , _ on tomas precession .
_ ihep preprint 98 - 85 , protvino , 1998 .
abramov , _ yad . fiz . _
* 44 * ( 1986 ) 1318 [ _ sov .
* 44 * ( 1986 ) 856 ] .
d.e.groom et al . , _ eur .
j. _ * c 15 * ( 2000 ) 1 .
b. anderson , g. gustafson and g. ingelman , _ phys .
* b 85 * ( 1979 ) 417 ; _ phys . rep . _
* 97 * ( 1983 ) 31 .
ryskin , _ yad . fiz . _ * 48 * ( 1988 ) 1114 [ _ sov . j. nucl
. phys . _ * 48 * ( 1988 ) 708 ] .
troshin , n.e .
tyurin , _ uspechi fiz . nauk _
* 164 * ( 1994 ) 1073 . | experimental data on the polarization of hyperons , inclusively produced in hadron - hadron collisions , have been analyzed .
it is shown that the existing data can be described by a function of transverse momentum ( @xmath0 ) and two scaling variables @xmath1 = @xmath2 : @xmath3 $ ] .
the function @xmath4 is proportional to @xmath5/\omega$ ] , which results for some reactions in an oscillation of @xmath6 as a function of @xmath7 .
the @xmath8 , as well as the magnitude and the sign of the hyperon polarization depend on quark composition of hadrons participating in a reaction .
the atomic weight dependence of the @xmath9 hyperon polarization is characterized by the parameter @xmath10 .
there is an analogy between the scaling properties of the hyperon polarization and the analyzing power ( @xmath11 ) in hadron production reactions .
this new scaling law allows one to predict hyperon polarization for reactions and kinematic regions , yet unexplored in experiments and to confront these predictions with future experiments and various models .
preprint ihep 2001 - 13 + hep - ph/0111128 * * * * * * v.v.abramov@xmath12 0.5 cm experimental physics department , institute for high energy physics , p.o .
box 35 , protvino , 142281 moscow region , russia @xmath12e - mail : [email protected] _ keywords : _ inclusive reaction ; polarization ; asymmetry ; spin ; qcd _ pacs : _ 13.85.ni ; 13.88.+e ; 12.38.qk |
ultracold atom physics has undergone spectacular development in the last two decades . during the last years , after the first theoretical proposal to simulate the bose - hubbard model with ultracold bosons in an optical lattice @xcite and its successful experimental realization @xcite , considerable theoretical and experimental progress has been made on degenerate quantum gases in optical lattices @xcite .
one major common goal shared by these efforts is to simulate typical systems or minimal models of condensed matter physics , e.g. the hubbard model , using ultracold atoms confined in optical lattices @xcite .
it is expected that these investigations can give key insights into unresolved open questions in condensed matter physics , e.g. the mechanism of high-@xmath0 superconductivity .
usually , interactions in cold gases are isotropic and short ranged .
naturally one can ask the question which type of new physics would arise in these systems if the atoms have additional long - range anisotropic interactions . as a matter of fact
, a wealth of interesting physics has already been revealed by theoretical and experimental investigations of ultracold dipolar bose gases both in harmonic traps @xcite and optical lattices @xcite .
in contrast to dipolar bose gases , due to the pauli exclusion principle , dipolar effects in fermionic systems can manifest themselves only if the dipolar interaction energy is at least of the order of the fermi energy .
although there have been already some theoretical investigations on these systems @xcite , this condition could hardly be reached experimentally until the recent successful realization of a degenerate quantum gas of fermionic @xmath1k@xmath2rb polar molecules @xcite . with this achievement ,
the door towards exploring the many - body physics originating from dipole - dipole interactions in fermionic systems has been opened , which has already lead to a number of theoretical investigations on these systems @xcite . motivated by experimental progress on ultracold polar molecules , here we investigate the low temperature quantum phases of a system of cold polar fermions loaded into a 2d square optical lattice .
the rest part of this paper is organized as follows . in sec .
ii , we introduce the system and the model studied here . in sec .
iii , we describe the theoretical approach used in our investigation . in sec .
iv , the main section of this paper , we present a detailed discussion of the quantum phases at both zero and finite temperature . finally , we conclude in sec . v.
motivated by ongoing experiments , we focus our investigations on possible setups based on @xmath1k@xmath2rb molecules which have a permanent electric dipole moment of @xmath3 debye ( d ) , where @xmath4 .
however the effective dipole moment in the laboratory frame is zero in the absence of an external electric field .
when an external electric field is applied the molecules align with the field and have an induced dipole moment @xmath5 which increases with the strength of the external field . currently , the experimentally accessible range of @xmath5 is @xmath6 $ ] @xcite . in this work
we assume the external electric field is oriented perpendicular to the optical lattice plane .
loading the molecules into an optical lattice , the physics of this system for sufficiently low filling can be captured by an extended hubbard model within the lowest band approximation,@xmath7 here , @xmath8 denotes nearest - neighbour sites , @xmath9(@xmath10 ) is the creation ( annihilation ) operator of a fermionic molecule with spin @xmath11 on site @xmath12 in the wannier representation , and @xmath13 is the particle number operator . the first term in eq .
( @xmath14 ) describes the kinetic energy with @xmath15 being the hopping amplitude ; the second term represents the onsite interaction @xmath16 between molecules with opposite spin ; the third term originates from the long - range dipole - dipole interactions , where @xmath17 characterizes the dipole - dipole interaction strength with @xmath18 and @xmath19 being the lattice constant and the vacuum permittivity respectively , and @xmath20 is the dimensionless position vector of lattice site @xmath12 ; finally , in the last term @xmath21 denotes the chemical potential . in our calculations
we assume a 2d optical lattice created by laser beams with wavelength @xmath22 as used in the experimental setup of ref .
we consider the following parameters : the height of the lattice potential in the direction perpendicular to the lattice plane @xmath23 and parallel to the plane @xmath24 are chosen as @xmath25 , @xmath26 , where @xmath27 is the recoil energy , @xmath28 is the mass of molecule and @xmath29 is planck s constant . under these conditions , the dipolar interaction strength is in the range @xmath30 $ ] , and the hopping amplitude is approximately @xmath31 .
therefore the ratio between the dipolar interaction strength and the hopping amplitude @xmath32 is approximately in the range @xmath33 $ ] , which we assume in our calculations . a good estimate of the strength of the onsite interaction @xmath16 requires a detailed study of all types of short - range interactions between the polar molecules , which is beyond the scope of this work .
since the main aim of this work is to investigate the possibility of observing a supersolid phase of polar molecules , we assume that the onsite interaction is attractive , i.e. , @xmath34 .
moreover , we notice that since large attractive on - site interactions will make the system unstable , the region of @xmath35 investigated in this work is restricted to @xmath36 $ ] . on the route towards observing the supersolid phase in experiments
, the temperature plays a dominant role . in current experiments with polar molecules in a harmonic trap ,
the lowest temperature which can be reached is of the order of the fermi temperature @xmath37 @xcite . also , when the lattice potential is ramped up , generally , the molecules will be heated in this process , which results in even higher temperatures . in the following ,
the system is studied at both zero and finite temperatures , and the temperature region where a supersolid phase can be possibly observed is determined .
before discussing our results in detail , we give a brief description of our theoretical approach . in polar molecule systems , we have to consider not only the on - site attraction of the molecules but also the inter - site repulsion due to the dipolar interaction . for treating the on - site interactions , the dynamical mean - field theory
( dmft ) @xcite is well suited since it is non - perturbative , captures the local quantum fluctuation exactly , and gives exact results in the infinite - dimensional limit . for long - ranged interactions in fermionic system , on the other hand ,
there are , to our knowledge , up to now only few efficient ways to treat them , although dmft has already been generalized to the so called cellular dmft in order to treat short - range inter - site interactions @xcite .
we include long - range interactions as follows .
first we notice that in the high - dimensional limit inter - site interactions only contribute on the hartree level @xcite . in other words ,
the hartree term of the inter - site interaction will dominate as the spatial dimension of the system increases .
this motivates us to keep only the hartree contribution of the inter - site interaction in the hamiltonian as an approximation to the original hamiltonian ( [ eq : hami ] ) , i.e. @xmath38 our dmft hamiltonian therefore takes the form @xmath39 where@xmath40 and we have rescaled the chemical potential @xmath21 in ( [ eq : hf_hamiltonian ] ) according to the trivial constant term in ( [ eq : hf_deouple ] ) .
we investigate this system using real - space dmft ( r - dmft)@xcite , which is an extension of dmft to a position - dependent self - energy and fully captures the inhomogeneity of the system . within dmft
/ r - dmft , the physics on each lattice site is determined from a local effective action which can be captured by an effective anderson impurity model @xcite . in this work ,
we use exact diagonalization ( ed ) @xcite of the anderson impurity hamiltonian to solve the local action .
details of the r - dmft method can be found in previous works @xcite . within each r - dmft iteration ,
the hartree contributions @xmath41 are calculated and the hamiltonian ( [ eq : hf_hamiltonian ] ) is updated according to the new values @xmath41 .
this modified r - dmft iteration is repeated until a convergent solution is obtained .
at zero temperature , the main properties of the system are summarized in the phase diagram in fig . [
flo : pd_12x12 ] , which shows the phase boundary between the supersolid ( ss ) phase and the homogeneous singlet superfluid phase ( hsf ) at different onsite interaction strengths @xmath16 in terms of the filling factor @xmath42 and the dipolar interaction strength @xmath43 , where @xmath44 is the number of lattice sites .
this phase diagram is based on calculations for a @xmath45 square lattice .
we will denote the phase boundary between the supersolid ( ss ) phase and the homogeneous singlet superfluid phase the _ ss - hsf boundary _ in the following . along the vertical line at half filling ( @xmath46 ) in fig .
[ flo : pd_12x12 ] , irrespective of the strengths of on - site and inter - site interactions , the system is always in an incompressible charge density wave ( cdw ) phase , which is characterized both by vanishing compressibility @xmath47 and a finite density modulation in real space .
more specifically , the density distribution over the lattice has a checkerboard ( cb ) structure , which is characterized by @xmath48@xmath49 with @xmath50 and @xmath51 . in our calculations ,
we choose @xmath52 as the cdw order parameter and we will denote this incompressible cdw phase as `` cb solid '' . away from half filling but still in the region bounded by the ss - hsf boundary and the vertical line of half filling , a supersolid phase is obtained by doping the cb solid with either particles or vacancies , which is characterized by the coexistence of the singlet pairing order parameter @xmath53 and the cdw order @xmath52 . as we can see from fig . [
flo : pd_12x12 ] , at fixed on - site interaction strength @xmath16 , the filling region in which the system remains supersolid increases with the strength of the dipolar interaction .
this can be understood by the following simple argument .
we notice that near the ss - hsf boundary the dominant contribution to the inter - site interaction energy between two nearest neighbor sites @xmath54 is positive and approximately proportional to @xmath55 , therefore a larger @xmath43 will make the system more likely to favor a modulated density distribution since it can lower the intersite interaction energy in this way . on the other hand , at fixed dipolar interaction strength @xmath43 , we can see from fig .
[ flo : pd_12x12 ] that a larger on - site attraction strength @xmath35 corresponds to a larger supersolid filling factor region .
this can be understood by a similar argument as above .
we notice that near the ss - hsf boundary the total on - site interaction energy of two neighboring sites is approximately proportional to @xmath56 , one can thus easily see that for large @xmath35 the system will favor a large density imbalance between neighboring sites which results in a modulated density distribution . concerning the proposed experimental setup with @xmath1k@xmath2rb molecules in an optical lattice , according to the discussions above , the supersolid phase will be observed more easily in an interaction region with relatively large dipolar interaction and onsite attraction , since this will stabilize the supersolid in a large filling factor region , e.g. in the range @xmath57 for interaction strengths @xmath58 and @xmath59 . on the right - hand side of the ss - hsf boundary ,
the system is in a homogeneous singlet superfluid phase which is characterized by a uniform distribution of both the density @xmath60 and the singlet paring order parameter @xmath61 .
we note that in the classical limit of zero hopping , the long - range interacting model exhibits a devil s staircase of various solid phases @xcite .
therefore , for much larger @xmath43 , which far exceeds the region investigated in this work , one would expect to find other types of incompressible cdw phases rather than cb solid and possibly new ss phases similar to those found for dipolar bose gases in optical lattices @xcite .
a detailed investigation of this `` devil s staircase '' in the quantum case is beyond the scope of this current work . and the strength of dipolar interaction @xmath43 , obtained from calculations on a @xmath45 square lattice .
different curves indicate the phase boundaries between supersolid ( ss ) and homogeneous singlet superfluid ( hsf ) at different onsite interaction strengths @xmath16 . on the left - hand side of each curve
the system is in the ss phase ( except exactly at half filling ( @xmath62 ) ) while the right - hand side region indicates the homogeneous singlet superconducting phase . from left to right , different curves correspond to @xmath63 , and @xmath64 respectively.,width=288 ] recently , there has been considerable progress in single - site addressability in optical lattices using electron and optical microscopy which allows for a direct , _ in situ , _ experimental observation of particle positions and density - density correlations of the system @xcite . in fig .
[ flo:12x12_nf_sc_distribution ] we show three sets of snapshots of the density and pairing order parameter distributions of the system for the different phases discussed previously .
we observe that the system develops the @xmath65 density modulation in both the cb solid and ss phase .
moreover in the ss phase the amplitude of the pairing order parameter @xmath61 also has a similar spatial checkerboard modulation . in the hsf phase both the density and the pairing order parameter are constant over the lattice .
we remark here that although the distribution of density and pairing order parameter is obtained in the absence of the external trapping potential , the typical feature of these phases can nevertheless be observed at the center of a shallow harmonic trap where a local density approximation is valid .
( left panel ) and singlet pairing order parameter @xmath61 ( right panel ) distribution of the system ( @xmath66 ) . in descending order , the filling factor has the values @xmath67 and @xmath68 respectively , corresponding to the checkerboard solid , supersolid , and homogeneous singlet superfluid . , width=336 ] ( left panel ) and singlet pairing order parameter @xmath61 ( right panel ) distribution of the system ( @xmath66 ) . in descending order , the filling factor has the values @xmath67 and @xmath68 respectively , corresponding to the checkerboard solid , supersolid , and homogeneous singlet superfluid . , width=336 ] ( left panel ) and singlet pairing order parameter @xmath61 ( right panel ) distribution of the system ( @xmath66 ) . in descending order
, the filling factor has the values @xmath67 and @xmath68 respectively , corresponding to the checkerboard solid , supersolid , and homogeneous singlet superfluid .
, width=336 ] in this subsection we investigate finite temperature effects and calculate the critical temperature of the supersolid which could be useful as a guide for future experiments .
first , we study the melting process of the supersolid to a normal phase with increasing @xmath69 at fixed filling .
[ flo : ss_melt_u_-4d_nf_1d21 ] shows the temperature dependence of the cdw order parameter @xmath52 and the average value of singlet pairing order parameter @xmath70 , which is used to characterize the superfluidity of the system , for @xmath71 , @xmath58 and filling factor @xmath72 .
the ss melts into a normal phase via two successive transitions .
first , it melts into a phase with zero paring order parameter @xmath61 but finite @xmath52 at the temperature @xmath73 , where @xmath74 is the boltzmann constant .
we will denote this phase cdw in the following discussion .
one interesting behavior to be noted in this step of the melting process is that as @xmath69 increases the cdw order parameter keeps growing to a maximum until the paring order @xmath75 decreases to zero , which is quite different from the melting processes of bosonic supersolids investigated in ref .
the physical reason for this behavior is the competition between pairing and cdw order in the system .
similar phenomena have already been observed in previous studies of condensed matter systems , see e.g. ref .
@xcite . to clarify this point , in the vicinity of the phase transition where both order parameters are small
, we calculate the ginzburg - landau ( gl ) free energy of the system up to the 4th order : @xmath76 here @xmath77 are the gl coefficients where a positive @xmath78 indicates competition between the two types of order . a detailed calculation ( see appendix a ) shows that @xmath78 is indeed positive , thus justifying the physical picture given above . upon further increase of the temperature @xmath69 ,
the cdw order parameter @xmath52 decreases to zero at @xmath79 . , @xmath58 , @xmath72 ) .
squares and circles correspond to the temperature dependence of the cdw order @xmath52 and the average value of the pairing order parameters @xmath75 , which is used to characterize the superfluidity of the system , respectively.,width=288 ] moreover , we map out a phase diagram of the system with respect to temperature @xmath69 and the dipolar interaction strength @xmath43 for onsite attraction @xmath71 at fixed filling @xmath72 , see fig . [
flo : finite_temperature_pd ] . when we lower the system temperature , for weak dipolar interaction ( @xmath80 ) we observe a transition from the normal to the hsf phase . while in a very narrow region @xmath81
, we observe first a transition between a normal phase ( white region ) and a cdw phase ( yellow region ) , then a transition into the ss phase ( green region ) , and finally a transition into the hsf ( blue region ) which is due to the competition between cdw order and pairing as discussed above . at larger dipolar interaction strength ( @xmath82 ) , we observe first a transition between the normal phase and the cdw phase , then a transition into the supersolid . as can be seen from fig .
[ flo : finite_temperature_pd ] , in order to observe the supersolid for these parameters , @xmath43 should be larger than @xmath83 , with a maximum critical temperature of about @xmath84 which is approximately @xmath85 in comparison to the lowest temperatures reached in current experiments with polar molecules in a harmonic trap , which are of order @xmath37 @xcite , this means that the temperature still has to be lowered by one order of magnitude in order to observe the supersolid . but given major theoretical and experimental efforts in lowering the temperature of cold atoms and molecules in optical lattices @xcite , we expect this supersolid phase will be accessible experimentally in the near future . and filling @xmath86 obtained from calculations on a @xmath45 square lattice .
see the text for details.,width=288 ]
in conclusion , we have shown that when fermionic polar molecules are loaded into a 2d square lattice , this system will exhibit a supersolid phase which can be observed in an experimental setup based on @xmath1k@xmath2rb molecules provided the lowest temperature of current experiments can be lowered by one order of magnitude .
we acknowledge useful discussions with e. altman , i. bloch and e. demler . l.
he thanks i. titvinidze and a. privitera for many helpful discussions and technical support at the beginning of this work .
this work was supported by the german science foundation dfg via the dip project bl 574/10 - 1 .
in this appendix , we give the detailed calculation of the gl coefficients from the microscopic hamiltonian ( [ eq : hami ] ) in the bloch representation , i.e. , @xmath87 where @xmath88 and @xmath89 is fourier transformation of the hopping matrix . the partition function of the system can be written in the path - integral form of @xmath90}$ ] where the action of the system is @xmath91\nonumber \\ & = & \sum_{k}\sum_{\sigma}(-i\omega_{n}+\varepsilon(\mathbf{k})-\mu)\bar{c}_{k\sigma}c_{k\sigma}\nonumber \\ & & + \frac{u}{\beta n}\sum_{k_{1}k_{2}q}\bar{c}_{k_{1}+q\uparrow}\bar{c}_{k_{2}-q\downarrow}c_{k_{2}\downarrow}c_{k_{1}\uparrow}\\ & & + \frac{1}{2}\frac{1}{\beta n}\sum_{q}v(\mathbf{q})\sum_{kk'\sigma\sigma'}\bar{c}_{k+q,\sigma}c_{k,\sigma}\bar{c}_{k'-q,\sigma'}c_{k',\sigma'}\nonumber \end{aligned}\ ] ] where @xmath92 is fermionic matsubara frequency with @xmath93 being the inverse temperature and @xmath94 are three - momentum comprising a vectorial momentum in 2-dimensional space and a fermionic matsubara frequency .
now by performing two hubbard - stratonovich transformations@xmath95\nonumber \\ & = & \int\mathcal{d}(\chi,\bar{\chi})\exp\left\ { -\sum_{q}\frac{\beta n}{|u|}|\chi_{q}|^{2}\right.\\ & & \left.+\bar{\chi}_{q}\left(\sum_{k}c_{k\downarrow}c_{q - k\uparrow}\right)+\left(\sum_{k}\bar{c}_{q - k\uparrow}\bar{c}_{k\downarrow}\right)\chi_{q}\right\ } \,,\nonumber \end{aligned}\ ] ] @xmath96\nonumber \\ & = & \int\mathcal{d}\phi\exp[-\sum_{q}\frac{\beta n}{2}\phi_{q}v^{-1}(\mathbf{q})\phi_{-q}\nonumber \\ & & -i\sum_{q}\phi_{q}\sum_{k\sigma}\bar{c}_{k - q,\sigma}c_{k,\sigma}]\,,\end{aligned}\ ] ] we introduce two auxiliary fields @xmath97 and @xmath98 which separately relate to the singlet pairing order and charge density wave order in the following way:@xmath99 after integrating out the fermionic degrees of freedom @xmath100 and @xmath101 , the partition function becomes @xmath102}$ ] where@xmath103 & = & \sum_{q}\left(\frac{\beta n}{|u|}|\chi_{q}|^{2}+\frac{\beta n}{2}\phi_{q}v^{-1}(\mathbf{q})\phi_{-q}\right)\nonumber \\ & & -\mathrm{tr}\ln\hat{\mathcal{g}}^{-1}\,.\end{aligned}\ ] ] the matrix @xmath104 have the following structure , @xmath105 where matrices @xmath106 , @xmath107 , @xmath108 , @xmath109 , and @xmath110 are given by @xmath111 , @xmath112 , @xmath113 , @xmath114 , and @xmath115 . before presenting a further analysis of the effective action @xmath116 $ ] in terms of the specific order parameter modes which we are interested in
, one thing to be noted is that the zero mode of @xmath98 , i.e. @xmath117 , corresponding to the particle density of the system , always has a non - vanishing contribution to the effective action @xmath116 $ ] independently of the system s parameters ( interaction strength , temperature , etc ) .
but on the other hand , we also note that the effect of @xmath117 is just to renormalize the chemical potential @xmath21 .
thus in the following analysis we simply neglect @xmath117 in @xmath116 $ ] by replacing the bare chemical potential @xmath21 with a renormalized one @xmath118 which can be determined from the particle density of the system . to simplify the notation , we further define@xmath119 @xmath120 we note that @xmath121,\label{eq : trace_ln}\ ] ] where the first term just contributes a trivial constant . in the vicinity of the phase transition when both the order parameter fields @xmath122 and @xmath123 are small
, we can expand the second term in ( [ eq : trace_ln ] ) in terms of @xmath124 and @xmath125 , i.e. , @xmath126=\sum_{n=1}\frac{(-1)^{n+1}}{n}\mathrm{tr}(\hat{\mathcal{g}}_{0}\hat{\lambda}+\hat{\mathcal{g}}_{0}\hat{\phi})^{n}.\label{eq : tr_ln_expansion}\ ] ] odd order terms of either @xmath127 or @xmath128 in the above expansion vanish since the effective action @xmath116 $ ] preserves the symmetry of the unordered phase .
since we are concerned with finite temperature phase transitions , we neglect the quantum fluctuations by treating @xmath97 and @xmath98 to be independent of matsubara frequency , i.e. , we focus on the zero matsubara frequency component of @xmath97 and @xmath98 .
furthermore we assume that @xmath129 , @xmath130 and their conjugate dominate the effective interaction @xmath116 $ ] , where @xmath51 . from the above expansion formula ( [ eq : tr_ln_expansion ] ) , the ginzburg - landau coefficients @xmath77 can be calculated explicitly . in the following ,
we calculate the gl coefficient @xmath78 which we are most interested in . expanding @xmath116 $ ] to the 4th order in @xmath129 and @xmath130 , the cross term between @xmath129 and @xmath130 is given by @xmath131 + 4\mathrm{tr}\left[\left(\hat{\mathcal{g}}_{0}\hat{\lambda}\right)^{2}\left(\hat{\mathcal{g}}_{0}\hat{\phi}\right)^{2}\right]\right\ } \nonumber \\ & = & \sum_{\mathbf{k},i\omega_{n}}\left[\frac{4}{(\xi^{2}(\mathbf{k})+\omega_{n}^{2})(i\omega_{n}-\xi(\mathbf{k}))(i\omega_{n}-\xi(\mathbf{k+q}))}\right.\nonumber \\ & & \left.+\frac{2}{(\xi^{2}(\mathbf{k})+\omega_{n}^{2})(\xi^{2}(\mathbf{k}+\mathbf{q})+\omega_{n}^{2})}\right]|\chi_{(\mathbf{0},0)}|^{2}|\phi_{(\mathbf{q},0)}|^{2}\nonumber \\ & = & c_{\chi\phi}\,|\chi_{(\mathbf{0},0)}|^{2}|\phi_{(\mathbf{q},0)}|^{2}\end{aligned}\ ] ] where @xmath132 .
it accounts for the competition between singlet pairing and cdw order . from ( [ eq : pairing_field ] ) and ( [ eq : cdw_field ] ) we easily obtain @xmath133 and @xmath134 which gives @xmath135 after identifying the effective action @xmath116 $ ] with the ginzburg - landau free energy . @xmath137\nonumber \\ & = & ( \frac{-\beta}{2\pi i})\ointctrclockwise dz\frac{1}{e^{\beta z}+1}\left[\frac{4}{(z-\xi(\mathbf{k}))^{2}(z+\xi(\mathbf{k}))(z-\xi(\mathbf{k}+\mathbf{q}))}+\frac{2}{(z^{2}-\xi^{2}(\mathbf{k}))(z^{2}-\xi^{2}(\mathbf{k}+\mathbf{q}))}\right]\nonumber \\ & = & \beta\left[\frac{1}{\xi(\mathbf{k})+\xi(\mathbf{k}+\mathbf{q})}\left(\frac{4}{(1+e^{\beta\xi(\mathbf{k}+\mathbf{q})})(\xi(\mathbf{k}+\mathbf{q})-\xi(\mathbf{k}))^{2}}-\frac{e^{\beta\xi(\mathbf{k})}}{(1+e^{\beta\xi(\mathbf{k})})\xi^{2}(\mathbf{k})}\right)\right.\nonumber \\ & & + \frac{\xi(\mathbf{k}+\mathbf{q})-3\xi(\mathbf{k})+e^{\beta\xi(\mathbf{k})}(\xi(\mathbf{k}+\mathbf{q})+\xi(\mathbf{k})(-3 + 2\beta(\xi(\mathbf{k}+\mathbf{q})-\xi(\mathbf{k}))))}{(1+e^{\beta\xi(\mathbf{k})})^{2}\xi^{2}(\mathbf{k})(\xi(\mathbf{k}+\mathbf{q})-\xi(\mathbf{k}))^{2}}\nonumber \\ & & \left.+\frac{1}{\xi^{2}(\mathbf{k})-\xi^{2}(\mathbf{k}+\mathbf{q})}\left(\frac{\tanh\frac{\xi(\mathbf{k}+\mathbf{q})}{2}}{\xi(\mathbf{k}+\mathbf{q})}-\frac{\tanh\frac{\xi(\mathbf{k})}{2}}{\xi(\mathbf{k})}\right)\right].\end{aligned}\ ] ] the remaining summation over momentum @xmath138 can be easily replaced by an integral over the density of states @xmath139 and evaluated numerically . as we can see from fig .
[ flo : gl4_vs_temperatur_rho_0d6 ] , which shows the temperature dependence of @xmath136 when @xmath140 at filling @xmath86 , @xmath136 is positive in this temperature region .
therefore , since the gl coefficient @xmath78 has the same sign as @xmath136 , we now clearly see that singlet pairing and cdw order indeed compete with each other in the parameter region under investigation . | we study a system of ultra - cold fermionic polar molecules in a two - dimensional square lattice interacting via both the long - ranged dipole - dipole interaction and a short - ranged on - site attractive interaction .
singlet superfluid , charge density wave , and supersolid phases are found to exist in the system .
we map out the zero temperature phase diagram and find that the supersolid phase is considerably stabilized by the dipole - dipole interaction and thus can exist over a large region of filling factors .
we study the melting of the supersolid phase with increasing temperature , map out a finite temperature phase diagram of the system at fixed filling , and determine the parameter region where the supersolid phase can possibly be observed in experiments . |
quasars are highly energetic sources located at the centers of galaxies created by accretion of matter onto supermassive black holes .
they are the most luminous subclass of active galactic nuclei ( agn ) and one of the most luminous classes of objects in the universe . through accurate measurements of the quasar luminosity function and its evolution ,
the formation history of supermassive black holes can been studied in detail ( @xcite and references therein ) . for studies of baryonic acoustic oscillations ( bao ) and the ly-@xmath15 forest ,
quasars act as tracers and backlights of matter clustering ( e.g. @xcite ) .
all these studies benefit from efficient and/or complete selection of large quasar samples .
@xcite pioneered quasar selection through the use of multi - color imaging data .
because quasars are not characterized by a single temperature like stars , they usually occupy different regions of color space .
for instance , most quasars are bluer in the uv / visible and redder in the infrared . by virtue of this , color selection methods using uv , visible , and mid - infrared photometry are one of the easiest and most common ways to select quasars at various redshifts ( e.g. @xcite ) .
nearly all quasars show @xmath1410%20% stochastic variability at uv and visible wavelengths over timescales of many months to years @xcite . since only a small fraction of stars are variable at this level , and many of these variable stars are periodic , variability data strongly separates quasars from stars .
models of variability such as a power - law @xcite or a damped random walk @xcite are usually used to identify the quasars . with the emergence of time - domain surveys such as pan - starrs @xcite and the future
large synoptic sky telescope @xcite , quasar selection based on variability will increase in significance and potentially help to fill the selection gaps of color selection techniques .
in addition to color and variability selection , more sophisticated selection methods have been developed , such as full multi - wavelength sed fitting @xcite , kernel density estimation @xcite , the likelihood method @xcite , neural networks @xcite , and extreme deconvolution @xcite .
these statistical methods model the underlying flux distribution of quasars based on empirical data and then assign probabilities that sources are quasars .
the modeled phase space is expandable and can include variability in addition to flux .
statistical quasar selection methods have proven to be very efficient , as they incorporate multi - dimensional information on quasar properties @xcite .
we aim to quantify various quasar selection methods for luminous and point source quasars such as color selection , probabilistic selection , variability selection , and combinations of the selection methods in the dark energy survey ( des ) supernova fields .
the dark energy survey ( des ) is a 5000 deg@xmath13 _
_ grizy-__band survey of the southern sky to probe the nature of dark energy @xcite .
des includes ten fields with a total area of @xmath14 30 deg@xmath13 that are surveyed at a higher cadence to search for type ia supernovae , and are known as the des supernova fields .
we also present a catalog of spectroscopically - confirmed quasars observed by ozdes @xcite in the des supernova fields .
ozdes is a complementary spectroscopic survey that targets the supernova fields to obtain redshifts for supernovae host galaxies and to conduct quasar reverberation mapping experiment @xcite and other projects .
the outline of the paper is as follows . in [ sec : data ] , we describe our photometric data sets and their corresponding surveys .
these surveys include the dark energy survey ( des ) , the vista hemisphere survey ( vhs ; @xcite ) , and wise @xcite .
we introduce the sdss stripe 82 quasar catalog from @xcite in [ sec : peters_cat ] that we use to evaluate the quasar selection methods . in [ sec : shape_selection ] , we investigate selecting quasars as point sources .
we evaluate the selection completeness and efficiency of various visible and ir color selection methods in [ sec : color_selection ] .
we analyze the ` xdqsoz ` probabilistic quasar selection algorithm @xcite in [ sec : xdqso ] . in [ sec : variability ] , we consider a @xmath4-based variability selection method .
we present the ozdes survey and the ozdes quasar catalog in [ sec : ozdes_qso_cat ] .
finally , we close in [ sec : discussion ] with discussions and conclusions . throughout the paper
, we adopted a flat @xmath16cdm cosmology with @xmath17 = 70 km s@xmath18mpc@xmath18 and @xmath19=0.3 . unless stated otherwise , all visible magnitudes refer to the des magnitudes .
the ab magnitude system is used when quoting sdss and des magnitudes , while the vega magnitude system is used for wise and vhs magnitudes .
the dark energy survey ( des ) is a wide - area 5000 deg@xmath13 survey of the southern hemisphere in the _ grizy _ bands @xcite . using the dark energy camera ( decam ; @xcite ) at the 4 m blanco telescope at the cerro tololo inter - american observatory , des aims to probe the nature of dark energy using four different astrophysical probes : type ia supernovae , baryonic acoustic oscillations ( bao ) , galaxy clusters , and weak lensing .
the planned 5@xmath20 point source depths of the survey are _ g _ = 26.5 mag , _ r _ = 26 mag , _ i _ = 25.3 mag , _ z _ = 24.7 mag , and _ y _ = 23 mag @xcite
. des is covering much more area than other surveys of similar depth ( e.g. the noao deep wide - field survey , @xcite ) and is much deeper than other surveys of larger area ( e.g. sdss and pan - starrs ) , so it is well suited to identifying new quasars .
the survey finished its third season of operation in february 2016 @xcite and recently started its fourth observing season in august 2016 .
des conducts a multi - epoch supernova survey of two deep fields ( c3 and x3 ) and eight shallow fields to search for type ia supernovae @xcite .
each supernova field subtends @xmath14 3 deg@xmath13 , or 30 deg@xmath13 total .
the supernova survey has a mean cadence of @xmath14 7 days in the _ griz _ bands .
while the main des survey is conducted under good seeing conditions ( at a median fwhm @xmath14 09 ) , it switches to imaging the supernova fields when the seeing increases to @xmath21 11 or when the supernova fields have not been observed for more than a week .
the supernova component is expected to comprise roughly one - third or @xmath14 1300 hours of the total des observing time .
table [ tab : des_sn ] shows the names and field centers of the ten supernova fields .
most of the supernova fields are also well - studied by other surveys such as the chandra deep field - south ( cdfs ; @xcite ) and the vimos - vlt deep survey ( vvds ; @xcite ) .
these fields are optimal for studying variability - based quasar selection since they have more photometric epochs than the main des wide - field survey . by the end of the 5-year survey ,
all ten supernova fields are expected to have more than ten times the exposure time of the wide - field survey , more than 100 epochs , and reach an expected 5@xmath20 point source depth in the _ griz _ bands of @xmath14 28 mag for the deep supernova fields and @xmath14 26.5 mag for the shallow fields @xcite . for our work
, we used the coadded catalog _
y1a1_coadd_objects_dfull _ for the des year one ( y1 , begun in 2013b ) observations of the supernova fields .
the catalog combines all available exposures ( 300 or more ) from the des y1 operation and some from the science verification phase that have sufficient image quality ( typical seeing fwhm @xmath7 11 , or 125 in some cases ) .
it is typically @xmath14 2 magnitudes deeper than the des y1 wide field coadded catalog .
the catalog includes the weighted average of many photometric quantities from single - epoch des data .
for example , _ wavgcalib_mag_psf _ and _ wavg_magerr_psf _ are the weighted average of the single - epoch _ mag_psf _ and _ magerr_psf _ values . the weighted averages are often found to be more accurate than the coadded quantities derived from sextractor @xcite , especially in areas where the number of epochs is small or where the coadded psfs are not well - fit by psfex @xcite .
for the rest of this paper , we used the weighted average quantities whenever referring to des photometry , such as _ wavgcalib_mag_psf_[grizy ] _ and _ wavg_magerr_psf_[grizy ] _ for the magnitudes and magnitude errors .
[ t ] [ tab : des_sn ] .des supernova fields [ cols="^ , < , < , < " , ] quasars with non - detections in des bands , no wise matches , or no vhs matches have entries of 99 for their magnitudes and errors . the naming schemes for the photometry follow the catalogs from which they are derived ; see the relevant sections as indicated for more information
. there are 44 columns in the catalog .
[ fig : ozdes_miz ]
large samples of quasars are needed to measure quasar luminosity functions and their evolution , particularly at lower luminosities , to estimate quasar host halo masses with weak lensing , and to study baryonic acoustic oscillations ( bao ) with the ly-@xmath15 forest , most notably with the dark energy spectroscopic instrument ( desi ) .
we used the @xcite quasar catalog for the sdss stripe 82 region that overlaps the des supernova fields , which we refer to as the p15-s1s2 quasar catalog , to analyze the completeness and efficiency of quasar color selection , probabilistic selection , variability selection , and combinations of colors and probabilistic selection methods with variability at the bright ( @xmath6 @xmath7 19.8 mag ) and faint ( @xmath6 @xmath7 22 mag ) magnitude limits .
we used only point sources that satisfy our data quality and photometric error cuts .
the sources used to investigate each method also need to meet the distinct detection thresholds for each selection method because of their dependence on the external wise and vhs catalogs .
the results of our analyses are summarized in table [ tab : summary ] and figure [ fig : summary_peters ] .
figure [ fig : normalization ] shows the fraction of sources that meet the selection - dependent thresholds ( i.e @xmath22 from table [ tab : summary ] ) for each selection method as a function of the des @xmath6-band magnitude .
the completeness and efficiency values both have statistical poissonian uncertainties due to the actual number of quasars in this study , and systematic uncertainties due to cosmic variance in the quasar population and the characteristics of the des data for the s1 and s2 fields . for a typical sample size of 300 ( 50 ) quasars in the faint ( bright ) samples used in this study , the statistical uncertainties are on the order of 8% ( 18% @xmath23 22% ) and these serve as guidelines to compare the relative performance of the selection methods .
these percentages correspond to one - sided , 90% confidence limits .
we calculated the surface densities of quasars and all candidates ( quasars and non - quasars ) selected by each selection method , also shown in table [ tab : summary ] .
as the area of stripe 82 is more precisely known than the overlap between des s1 and s2 fields with stripe 82 , and partly because cosmic variance is smaller over a larger region , the surface densities are calculated using the overall p15 quasar surface densities , rather than using the sources in the des s1 and s2 fields alone . for their `` good '' quasar candidates ( [ sec : peters_cat ] ) , the p15 catalog has a surface density @xmath24 = 132.79 deg@xmath25 at @xmath26 @xmath7 22 mag and @xmath27 = 24.36 deg@xmath25 at @xmath26 @xmath7 19.9 mag .
the surface density of quasars @xmath28 and all candidates @xmath29 ( quasars and non - quasars ) selected by a selection method is then @xmath28 = @xmath30 @xmath22 @xmath31 and @xmath29 = @xmath30 @xmath22 @xmath31/@xmath32 = @xmath28/@xmath32 , where @xmath30 refers to either @xmath33 or @xmath34 , @xmath22 is from table [ tab : summary ] , @xmath31 is the completeness , and @xmath32 is the efficiency . the surface density of selected non - quasars is simply @xmath29 @xmath23 @xmath28 .
while variability selection gives the highest surface density , the selection efficiency is one of the lowest and this results in a significant fraction of contaminants . among individual selection methods ,
@xmath11 returns the highest quasar surface density , although it has a lower efficiency than @xmath12 and ` xdqsoz ` . for combined selection methods , ` xdqsoz`+variability is more efficient than @xmath12+variability and @xmath11+variability but also returns lower quasar surface density .
the @xmath11+variability selection gives the highest quasar surface density among the three combined selection methods .
taking both the surface densities of selected quasars and selection efficiencies into account , ` xdqsoz ` selection alone and a combination of color or ` xdqsoz ` with variability result in relatively high surface densities of quasars and a modest amount of contamination .
the most significant drawback for both the @xmath12 and ` xdqsoz ` methods is that the wise data are only available for 55% of the quasars with @xmath6 @xmath7 22 mag , as both @xmath8 and @xmath9 detections are required .
while the @xmath11 selection method has lower efficiency , it will produce a higher surface density of quasars because a @xmath9 detection is not required .
since we used only photometry from the first year of des operations , we employed a simple multi - band @xmath4 to detect variability . variability selection based on more sophisticated quasar variability models , such as the damped random walk model , is deferred to the future when substantially more epochs will be available .
the combination of the @xmath12 , @xmath11 , or ` xdqsoz ` methods with variability improves the efficiency of quasar selection ( fewer false positives ) at slightly lower completeness . at the bright end ,
@xmath12+variability and ` xdqsoz`+variability can be applied in the des supernova fields , but the fraction of sources with wise detections will diminish for fainter sources .
the @xmath11 or @xmath11+variability selection method can be used instead , particularly because they both return high quasar surface densities , and in the case of @xmath11+variability , with good selection efficiency .
the depth of wise data is less of an issue for the shallower fields .
although variability selection will not be as useful in these fields because of the small number of epochs , @xmath12 , @xmath11 or ` xdqsoz ` selection would be a good alternative .
we also presented the ozdes quasar catalog of 1,263 spectroscopically - confirmed quasars in the 30 deg@xmath13 des supernova fields brighter than _
_ i__=22 mag .
the catalog includes all the quasars selected from the des / ozdes reverberation mapping project with good quality ozdes redshifts , m@xmath35 @xmath7 @xmath2322 mag , and visually confirmed emission and absorption lines .
the ozdes quasar catalog is not homogeneous or complete , although its completeness will improve as the ozdes observations continue . * 13l & & & & & ( deg@xmath25 ) & ( deg@xmath25 ) + ( lr)2 - 3 ( lr)4 - 5 ( lr)6 - 7 ( lr)8 - 9 ( lr)10 - 11 ( lr)12 - 13 selection & bright & total & bright & total & bright & total & bright & total & bright & total & bright & total + & & & & & & & & & ( @xmath36=24.4 ) & ( @xmath36=132.8 ) & & + @xmath12 & 0.91 & 0.93 & 0.71 & 0.54 & 0.65 & 0.50 & 0.96 & 0.55 & 21.3 & 67.9 & 30.0 & 125.8 + @xmath11 & 0.87 & 0.91 & 0.61 & 0.49 & 0.53 & 0.44 & 0.98 & 0.72 & 20.8 & 87.0 & 34.1 & 177.6 + @xmath37 & 0.89 & 0.92 & 0.63 & 0.44 & 0.56 & 0.41 & 0.99 & 0.52 & 21.5 & 63.5 & 34.1 & 144.4 + ` xdqsoz ` & 0.92 & 0.90 & 0.73 & 0.60 & 0.67 & 0.54 & 0.96 & 0.55 & 21.5 & 65.7 & 29.5 & 109.6 + variability & 0.86 & 0.88 & 0.088 & 0.11 & 0.075 & 0.10 & 0.86 & 0.90 & 18.0 & 105.2 & 204.7 & 956.1 + @xmath12@xmath38var & 0.81 & 0.88 & 0.83 & 0.63 & 0.68 & 0.55 & 0.82 & 0.49 & 16.2 & 57.3 & 19.5 & 90.9 + @xmath11@xmath38var & 0.78 & 0.86 & 0.80 & 0.60 & 0.63 & 0.51 & 0.84 & 0.65 & 16.0 & 74.2 & 20.0 & 123.7 + ` xdqsoz`@xmath38var & 0.83 & 0.85 & 0.83 & 0.68 & 0.68 & 0.57 & 0.82 & 0.49 & 16.6 & 55.3 & 20.0 & 81.3 + the bright sample refers to sources with @xmath6 @xmath7 19.8 mag and the total sample refers to sources with @xmath6 @xmath7 22 mag .
completeness and efficiency are defined at the end of [ sec : peters_cat ] .
`` @xmath39 '' is the product of completeness and efficiency .
`` @xmath22 '' is the fraction of quasars that meet the selection - dependent detection threshold(s ) .
this fraction is relative to all quasars that satisfy our point source , flags , and photometric errors cuts .
since both the @xmath12 and ` xdqsoz ` methods have the same des and wise detection requirements , their @xmath22 values are the same .
the statistical uncertainties in the number of quasars for a typical sample size of 300 ( 50 ) in the faint ( bright ) sample used in this study are on the order of 8% ( 18% @xmath2322% ) . for more discussion , see [ sec : discussion ] .
the last four columns show the surface densities of selected quasars ( @xmath28 ) and all candidates ( @xmath29 ) for each method .
the surface densities are calculated based on the overall p15 quasar surface density ( @xmath36 ) , @xmath22 , and completeness / efficiency as detailed in [ sec : discussion ] .
the surface density of the selected non - quasars is simply @xmath29 @xmath23 @xmath28 .
[ fig : summary_peters ]
funding for the des projects has been provided by the u.s .
department of energy , the u.s
. national science foundation , the ministry of science and education of spain , the science and technology facilities council of the united kingdom , the higher education funding council for england , the national center for supercomputing applications at the university of illinois at urbana - champaign , the kavli institute of cosmological physics at the university of chicago , the center for cosmology and astro - particle physics at the ohio state university , the mitchell institute for fundamental physics and astronomy at texas a&m university , financiadora de estudos e projetos , fundao carlos chagas filho de amparo pesquisa do estado do rio de janeiro , conselho nacional de desenvolvimento cientfico e tecnolgico and the ministrio da cincia , tecnologia e inovao , the deutsche forschungsgemeinschaft and the collaborating institutions in the dark energy survey .
the collaborating institutions are argonne national laboratory , the university of california at santa cruz , the university of cambridge , centro de investigaciones energticas , medioambientales y tecnolgicas - madrid , the university of chicago , university college london , the des - brazil consortium , the university of edinburgh , the eidgenssische technische hochschule ( eth ) zrich , fermi national accelerator laboratory , the university of illinois at urbana - champaign , the institut de cincies de lespai ( ieec / csic ) , the institut de fsica daltes energies , lawrence berkeley national laboratory , the ludwig - maximilians universitt mnchen and the associated excellence cluster universe , the university of michigan , the national optical astronomy observatory , the university of nottingham , the ohio state university , the university of pennsylvania , the university of portsmouth , slac national accelerator laboratory , stanford university , the university of sussex , texas a&m university , and the ozdes membership consortium .
the des data management system is supported by the national science foundation under grant number ast-1138766 .
the des participants from spanish institutions are partially supported by mineco under grants aya2012 - 39559 , esp2013 - 48274 , fpa2013 - 47986 , and centro de excelencia severo ochoa sev-2012 - 0234 .
research leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp7/2007 - 2013 ) including erc grant agreements 240672 , 291329 , and 306478 .
this work is based in part on data obtained at the australian astronomical observatory through program a/2013b/012 .
this research made use of astropy , a community - developed core python package for astronomy ( astropy collaboration , 2013 ) .
@xmath40 department of astronomy , the ohio state university , columbus , oh 43210 , usa + @xmath13 center for cosmology and astro - particle physics , the ohio state university , columbus , oh 43210 , usa + @xmath41 institute of astronomy , university of cambridge , madingley road , cambridge cb3 0 ha , uk + @xmath42 kavli institute for cosmology , university of cambridge , madingley road , cambridge cb3 0 ha , uk + @xmath43 australian astronomical observatory , north ryde , nsw 2113 , australia + @xmath44 school of mathematics and physics , university of queensland , brisbane , qld 4072 , australia + @xmath45 the research school of astronomy and astrophysics , australian national university , act 2601 , australia + @xmath46 centre for astrophysics & supercomputing , swinburne university of technology , victoria 3122 , australia + @xmath47 fermi national accelerator laboratory , p. o. box 500 , batavia , il 60510 , usa + @xmath48 research school of astronomy and astrophysics , the australian national university , canberra , act 2611 , australia + @xmath49 inaf- osservatorio astronomico di torino - strada osservatorio 20 , pino torinese , 10020 , italy + @xmath50 school of physics and astronomy , university of southampton , southampton , so17 1bj , uk + @xmath51 sydney institute for astronomy , school of physics , a28 , the university of sydney , nsw 2006 , australia + @xmath52 cerro tololo inter - american observatory , national optical astronomy observatory , casilla 603 , la serena , chile + @xmath53 department of physics & astronomy , university college london , gower street , london , wc1e 6bt , uk + @xmath54 department of physics and electronics , rhodes university , po box 94 , grahamstown , 6140 , south africa + @xmath55 cnrs , umr 7095 , institut dastrophysique de paris , f-75014 , paris , france + @xmath56 sorbonne universits , upmc univ paris 06 , umr 7095 , institut dastrophysique de paris , f-75014 , paris , france + @xmath57 laboratrio interinstitucional de e - astronomia - linea , rua gal .
jos cristino 77 , rio de janeiro , rj - 20921 - 400 , brazil + @xmath58 observatrio nacional , rua gal .
jos cristino 77 , rio de janeiro , rj - 20921 - 400 , brazil + @xmath59 department of astronomy , university of illinois , 1002 w. green street , urbana , il 61801 , usa + @xmath60 national center for supercomputing applications , 1205 west clark st . , urbana , il 61801 , usa + @xmath61 institut de cincies de lespai , ieec - csic , campus uab , carrer de can magrans , s / n , 08193 bellaterra , barcelona , spain + @xmath62 institut de fsica daltes energies ( ifae ) , the barcelona institute of science and technology , campus uab , 08193 bellaterra ( barcelona ) spain + @xmath63 kavli institute for particle astrophysics & cosmology , p. o. box 2450 , stanford university , stanford , ca 94305 , usa + @xmath64 george p. and cynthia woods
mitchell institute for fundamental physics and astronomy , and department of physics and astronomy , texas a&m university , college station , tx 77843 , usa + @xmath65 department of physics , iit hyderabad , kandi , telangana 502285 , india + @xmath66 jet propulsion laboratory , california institute of technology , 4800 oak grove dr . , pasadena , ca 91109 , usa + @xmath67 department of astronomy , university of michigan , ann arbor , mi 48109 , usa + @xmath68 department of physics , university of michigan , ann arbor , mi 48109 , usa + @xmath69 kavli institute for cosmological physics , university of chicago , chicago , il 60637 , usa + @xmath70 instituto de fica teica ift - uam / csic , universidad automa de madrid , cantoblanco 28049 , madrid , spain + @xmath71 department of astronomy , university of california , berkeley , 501 campbell hall , berkeley , ca 94720 , usa + @xmath72 lawrence berkeley national laboratory , 1 cyclotron road , berkeley , ca 94720 , usa + @xmath73 slac national accelerator laboratory , menlo park , ca 94025 , usa + @xmath74 department of physics , the ohio state university , columbus , oh 43210 , usa + @xmath75 astronomy department , university of washington , box 351580 , seattle , wa 98195 , usa + @xmath76 departamento de fsica matemtica , instituto de fsica , universidade de so paulo , cp 66318 , cep 05314 - 970 , so paulo , sp , brazil + @xmath77 instituci catalana de recerca i estudis avanats , e-08010 barcelona , spain + @xmath78 institute of cosmology & gravitation , university of portsmouth , portsmouth , po1 3fx , uk + @xmath79 department of physics and astronomy , pevensey building , university of sussex , brighton , bn1 9qh , uk + @xmath80 centro de investigaciones energticas , medioambientales y tecnolgicas ( ciemat ) , madrid , spain + @xmath81 instituto de fsica , ufrgs , caixa postal 15051 , porto alegre , rs - 91501 - 970 , brazil + @xmath82 universidade federal do abc , centro de cincias naturais e humanas , av .
dos estados , 5001 , santo andr , sp , brazil , 09210 - 580 + @xmath83 universidade federal do abc , centro de cincias naturais e humanas , av
. dos estados , 5001 , santo andr , sp , brazil , 09210 - 580 + @xmath84 computer science and mathematics division , oak ridge national laboratory , oak ridge , tn 37831 , b. t. , & dey , a. 1999 , in astronomical society of the pacific conference series , vol .
191 , photometric redshifts and the detection of high redshift galaxies , ed .
r. weymann , l. storrie - lombardi , m. sawicki , & r. brunner , 111 | we present a study of quasar selection using the dark energy survey ( des ) supernova fields .
we used a quasar catalog from an overlapping portion of the sdss stripe 82 region to quantify the completeness and efficiency of selection methods involving color , probabilistic modeling , variability , and combinations of color / probabilistic modeling with variability . in all cases , we only considered objects that appear as point sources in the des images .
we examine color selection methods based on the wise mid - ir @xmath0 color , a mixture of wise and des colors ( @xmath1 and @xmath2 ) and a mixture of vhs and des colors ( @xmath1 and @xmath3 ) . for probabilistic quasar selection , we used ` xdqsoz ` , an algorithm that employs an empirical multi - wavelength flux model of quasars to assign quasar probabilities .
our variability selection uses the multi - band @xmath4-probability that sources are constant in the des year 1 @xmath5-band light curves .
the completeness and efficiency are calculated relative to an underlying sample of point sources that are detected in the required selection bands and pass our data quality and photometric error cuts .
we conduct our analyses at two magnitude limits , @xmath6 @xmath7 19.8 mag and @xmath6 @xmath7 22 mag . for the subset of sources with @xmath8 and @xmath9 detections , the @xmath0 color _ or _ `
xdqsoz ` method _ combined _ with variability gives the highest completenesses of @xmath10 85% for both @xmath6-band magnitude limits and efficiencies of @xmath10 80% to the bright limit and @xmath10 60% to the faint limit ; however , the @xmath11 and @xmath11+variability methods give the highest quasar surface densities .
the ` xdqsoz ` method and combinations of @xmath12/@xmath11/`xdqsoz ` with variability are among the better selection methods when both high completeness and high efficiency are desired .
we also present the ozdes quasar catalog of 1,263 spectroscopically - confirmed quasars from three years of ozdes observation in the 30 deg@xmath13 of the des supernova fields .
the catalog includes quasars with redshifts up to z @xmath14 4 and brighter than @xmath6=22 mag , although the catalog is not complete up this magnitude limit . |
the coherent state ( cs ) introduced in 1963 @xcite has become an useful and necessary tool to treat ideal boson fields subjected to external pumping sources .
however , cs s can not describe nonclassical effects such as antibunching , squeezing , etc .
the conventional squeezed states @xcite has been generalized to different types of higher - order ones @xcite , still for ideal bosonic particles . elementary excitations in matter , on the other hand , are quasi - particles often obeying neither bose - einstein nor fermi - dirac statistics .
recently , the notion of nonlinear coherent state ( ncs ) has emerged to adequately deal with such quasi - particles with commutation relations deformed from the usual boson / fermion ones .
@xmath4-deformed oscillators @xcite were found as a particular case of the general @xmath5-oscillators @xcite which possess in themselves a kinematic nonlinearity causing orbit - dependent oscillation frequencies .
the ncs @xcite is defined as the right - hand eigenstate @xmath6 of the nonboson operator @xmath7@xmath8 with @xmath9 @xmath10 the bosonic annihilation operator , @xmath11 a complex eigenvalue and @xmath5 an arbitrary nonlinear operator - valued function of @xmath12 an island where ncs s find their life is single trapped ions driven by lasers @xcite . in this paper
we introduce the @xmath0-quantum nonlinear coherent state ( @xmath0ncs ) which reduces to the usual ncs defined by eq .
( [ 1 ] ) when @xmath13 in section ii the formulation is given for the @xmath0ncs together with their mathematical properties .
section iii is devoted to a physical scheme of generation of @xmath0ncs s .
their nonclassical properties are studied in detail in section iv .
conclusion is the final section .
the @xmath0ncs is a generalization of the ncs to the @xmath0 right - hand eigenstates @xmath14 of the non - hermitian operator @xmath15 as @xmath16 where @xmath17 and @xmath18 spanned in the fock basis @xmath19@xmath20 the expansion coefficients satisfy the recurrence equation @xmath21 equation ( [ 4 ] ) determines all the @xmath22 s if @xmath23 with @xmath24 are known , @xmath25 where @xmath26 of course , @xmath27 in eq . (
the @xmath23 in eq .
( [ 5 ] ) are themselves determined by the normalization condition @xmath28 which yields @xmath29 ^{-1/2}. \label{7}\end{aligned}\ ] ] in principle , the @xmath30 is uncertain up to a phase factor . in eq .
( [ 7 ] ) we have chosen the phase such that the usual cs results in the limit @xmath31 and @xmath13 the explicit form of the @xmath0ncs is then @xmath32 unlike the usual cs whose eigenvalues spread the whole complex plane , those of the @xmath0ncs are bounded in each particular case .
this limitation is felt from eq .
( [ 8 ] ) that requires @xmath30 not be zeros . yet , the @xmath30 given by eq .
( [ 7 ] ) would vanish if the sum over @xmath33 diverges .
the constraint for the existence of @xmath0ncs s , i.e. for @xmath30 not to be zeros , is thus imposed simultaneously on @xmath34 @xmath5 and @xmath0 so that @xmath35 ^{-1}\right\ } = 0 .
\label{9}\ ] ] the importance of this constraint will be seen in the next section when the specific nonlinear function @xmath5 is addressed . for a given @xmath0
the eigenvalue @xmath11 is @xmath0-degenerate .
the @xmath0 corresponding eigenfunctions @xmath36 with @xmath37 are orthogonal to each other , @xmath38 as easily verified from the expansion ( [ 8 ] ) .
however , with distinct eigenvalues @xmath11 and @xmath39@xmath40 because of the non - orthogonality ( [ 11 ] ) , the @xmath0ncs s constitute an overcomplete set with the following resolution of unity @xmath41 this section is ended by an interesting observation that any state @xmath42 can be decomposed into a linear superposition of @xmath0 states @xmath43 namely , @xmath44 with @xmath45 and @xmath46 the verification is straightforward by substituting eqs .
( [ 8 ] ) , ( [ 14 ] ) and ( [ 15 ] ) into the r.h.s . of eq .
( [ 13 ] ) with subsequent use of the identities @xmath47 = p\delta _ { ll^{\prime } } \ ] ] and @xmath48 for arbitrary coefficients @xmath49 and integer @xmath50
motivated by the physical scheme proposed in @xcite we consider a single two - level ion trapped by a harmonic potential @xmath51 with @xmath52 the trapping force and @xmath53 the ion s center - of - mass position . in the quantized regime
the ion vibrates around @xmath54 with frequency @xmath55 @xmath56 the ion mass . as a whole ,
the ion hamiltonian is ( @xmath57 throughout ) @xmath58 where @xmath59 is the energy gap between the two electronic levels of the ion , @xmath10 is the bosonic annihilation operator of a quantum of the quantized vibration of the ion and , @xmath60 together with @xmath61 are the pseudospin-@xmath62 operators satisfying the commutation relations
@xmath63 = \pm \sigma _ { \pm } , \text { } \left [ \sigma _ { + } , \sigma _ { -}\right ] = 2\sigma _ { 3}. \label{17}\ ] ] by controlling the trapping potential the vibration frequency @xmath64 can be made large enough for the sidebands due to the ion quantized motion to be well - resolved .
next , the trapped ion is manipulated by laser beams which couple the ion s electronic to its vibrational degrees of freedom via the light - matter interaction hamiltonian @xmath65 in eq .
( [ 18 ] ) @xmath66 is the number of driving lasers , @xmath67 the pure electronic transition rabi frequencies , @xmath68 ( @xmath69 ) the laser frequencies ( phases ) and @xmath70 the laser spatial profile acting on the ion through its position operator @xmath71 for traveling waves @xmath72
\label{20}\ ] ] with @xmath73 the laser wavelengths and @xmath74 the lamb - dicke parameter . in the resolved sideband limit the lasers can be tuned so as @xmath75 where @xmath76 ( @xmath77 ) imply blue ( red ) detuning while resonant tuning corresponds to @xmath78 in the interaction representation associated with @xmath79 the laser - driven trapped ion is described by the total hamiltonian @xmath80 with @xmath81 putting eqs .
( [ 16 ] ) and ( [ 18 ] ) into the r.h.s . of eq .
( [ 22 ] ) with use of the baker - hausdorff identity and the relations ( [ 17 ] ) yields @xmath82 with @xmath83 given by @xmath84 a^{(|n_{l}|+n_{l})/2}\right\ } .
\label{24}\end{aligned}\ ] ] in deriving eq .
( [ 24 ] ) terms oscillating with frequencies @xmath85 ( @xmath86 ) were omitted since those average to zero in the resolved sideband limit we are interested in .
the system state @xmath87 evolves in time following the equation @xmath88 we look for long - term nontrivial steady states @xmath89 in which the internal and external degrees of freedom of the ion become decoupled , i.e. @xmath90 such nontrivial solutions would be either @xmath91 or @xmath92 where @xmath93 ( @xmath94 ) is the electronic excited ( ground ) state and @xmath95 describes the vibrational state .
the state @xmath96 is unstable due to spontaneous emission which was not treated explicitly here .
the remaining `` dark''state @xmath97 is stable with @xmath98 satisfying the condition @xmath99 similar `` dark '' states have been obtained as an ansatz of master equations with spontaneous emission included @xcite . from eqs .
( [ 25 ] ) and ( [ 27 ] ) it follows that the nontrivial state @xmath100 exists iff there are at least two driving lasers . for generality , let there be @xmath101 ( @xmath102 ) red - detuned ( blue - detuned ) laser beams and a single resonant one .
then , the coefficients of the expansion of @xmath100 in the fock space , @xmath103 are recurrent as @xmath104 @xmath105 @xmath106 where @xmath107 is the step function and @xmath108 the @xmath109-th generalized laguerre polynomial in @xmath53 for parameter @xmath33 . for the purpose of generating the @xmath0ncs defined in the preceding section we restrict to @xmath110 @xmath111 and @xmath112 and get from eq .
( [ 29 ] ) @xmath113 with @xmath114 comparing eqs .
( [ 30 ] ) and ( [ 4 ] ) indicates that @xmath100 obeying eq .
( [ 27 ] ) is a @xmath0ncs with the physical controllable eigenvalue @xmath115 and the specific , also controllable , nonlinear function @xmath116 in the next section the number distribution , squeezing and antibunching of the @xmath0ncs with the specific @xmath11 and @xmath5 given above will be studied in detail . to that aim
let us define the function @xmath117 in terms of which @xmath118 it is worth to emphasize that the @xmath0ncs does not exist for arbitrary control parameters @xmath119 and @xmath120 for a fixed trapping potential characterized by @xmath119 the amplitudes ( phases are not sensitive in this scheme with @xmath112 and @xmath121 ) of the driving fields must be chosen such that the constraint ( [ 9 ] ) is met , i.e. @xmath122 figure 1 , log(@xmath123 ) versus @xmath124 shows opposite limiting behaviors for the same value of @xmath119 but different latexmath:[$% which the @xmath0ncs vanishes .
this critical @xmath126 depends strongly on both @xmath119 and @xmath0 but not on @xmath127 phase diagrams for the existence domain of the @xmath0ncs are drawn in fig .
2 in the ( @xmath119,@xmath128)-plane for @xmath129 and @xmath130 these diagrams guide the appropriate experimental choice of the control parameters to observe the @xmath0ncs with a concrete @xmath131
the probability of finding @xmath132 quanta in the @xmath0ncs , i.e. its number distribution , is given by @xmath133 where @xmath134 equals @xmath53 if @xmath53 is a non - negative integer and zero otherwise .
the tortuous shape in fig .
1 gives rise to a multi - peaked structure of @xmath135 as depicted in fig .
3 in the case @xmath136 @xmath137 the multi - peaked distribution is very peculiar compared to the poisson s in the usual cs .
however , multiple peaking disappears in the lamb - dicke limit @xmath138 in which @xmath139 varies monotonically ( not tortuously ) for increasing @xmath132 having only one peak at some value @xmath140 and then decreasing quickly .
the developing from single- to multi - peaked structure as @xmath119 increases can be called self - splitting @xcite which would bring about abundant nonclassical phenomena including quantum interferences if the peaks are located nearby and comparable in heights .
as for higher orders @xmath141 unexpectedly , the self - splitting tends to be less pronounced as seen from fig .
4 . besides getting less tortuous , the more important fact is that the curve decreases much faster for higher @xmath131 hence , for @xmath142 the @xmath119-governed self - splitting is negligible physically . instead , for @xmath143 another kind of splitting , the @xmath0-governed one , appears independent of the lamb - dicke parameter .
this is dictated by the presence of @xmath144 in eq .
( [ 35 ] ) .
namely , for @xmath142 the number distribution `` oscillates '' with `` period '' @xmath0 and its `` phase '' is @xmath1-dependent .
the @xmath145 @xmath146 and @xmath147 plotted in fig .
5 illustrate that , in the state @xmath148 ( @xmath149 @xmath150 ) , finite is only the probability of finding @xmath151 ( @xmath152 ; @xmath153 ) quanta . generally speaking
, only a number of quanta equal to a multiple @xmath0 plus @xmath1 ( @xmath154 @xmath155 @xmath156 ) can be found in the state @xmath157 given @xmath0 each @xmath158-state occupies its own subspace and all the different @xmath1-states fill the entire fock space .
therefore , the notion `` number - parity '' might be introduced .
a state is said to have a number - parity @xmath2 if it contains only @xmath159 ( @xmath160 @xmath161 @xmath162 ) quanta . for @xmath163
the parity gets its intuitive meaning : the state @xmath164 ( @xmath165 ) contains only even ( odd ) numbers of quanta and it can be referred to as even ( odd ) ncs @xcite .
this evenness ( oddness ) is also clear from the view point of the decomposition ( [ 13 ] ) according to which @xmath166 @xmath167 by virtue of eqs . ( [ 14 ] ) and ( [ 15 ] ) , @xmath168 and @xmath169 these mean that the combination in eq .
( [ e ] ) is symmetric leading to evenness , while that in eq .
( [ o ] ) is anti - symmetric leading to oddness . because the @xmath170 determined in eqs .
( [ 23 ] ) and ( [ 24 ] ) for @xmath110 @xmath171 @xmath112 creates / annihilates @xmath0 quanta at a time the number - parity is conserved .
if initially the vibration quanta are in a fock state @xmath172 then the steady state @xmath89 will have number - parity @xmath2 with @xmath1 the remainder of @xmath33 divided by @xmath131 by sideband cooling modern techniques the fundamental limit has been reached in which the ion vibrational motion can be kept to zero - point energy 98% ( 92% ) of the time in 1d ( 3d ) @xcite .
if such a ground state is further manipulated by lasers as proposed above , then the state @xmath173 will be realized .
if , the initial quanta are prepared in a cs @xmath174 then the eventual state will be a mixed state with weighted @xmath1-components @xmath175 assuming the weights @xmath176 be as in the cs , @xmath177 the number distribution @xmath178 of the mixed state @xmath179 will differ noticeably under two situations : @xmath180 and latexmath:[$|\alpha are scattered whereas for latexmath:[$% for all @xmath1 reducing @xmath178 simply to @xmath182 with @xmath183 given in eq .
( [ 35 ] ) , independent of @xmath184 figure 6 addresses the above - said issue for @xmath185 and several values of @xmath186 the @xmath187 spans the whole fock space with different @xmath188-dependent profiles .
this is of course subject to experimental checking to which extent the assumption ( [ 37 ] ) is valid . in the small @xmath119 limit the @xmath3
ncs exhibits a single - peaked number distribution that may be super- or sub - poissonian .
so happens as well for the envelope of the @xmath142 ncs distribution .
sub - poisson distribution featuring antibunching occurs whenever the mandel parameter , @xmath189 is less than @xmath190 .
we shall also examine amplitude - quadrature squeezing of the ion center - of - mass position operator ( in units of @xmath191 , for convenience ) @xmath192 by calculating its variance @xmath193 with @xmath194 given by @xmath195 the @xmath0ncs gets squeezed in @xmath196 if @xmath197 in terms of the functions @xmath198 defined in eq .
( [ 33 ] ) @xmath199 } { % \sum_{m=0}^{\infty } \left| g_{kj}(m,0)\right| ^{2 } } , \label{41}\ ] ] @xmath200 } { % \sum_{m=0}^{\infty } \left| g_{kj}(m,0)\right| ^{2}}. \label{42}\ ] ] the analytic expressions of @xmath56 and @xmath194 are readily obtained using eqs .
( [ 41 ] ) and ( [ 42 ] ) in ( [ 39 ] ) and ( [ 40 ] ) .
for @xmath3 it is found that @xmath201 i.e. no squeezing , but @xmath56 decreases from @xmath190 for increasing @xmath202 and @xmath203 for all @xmath204 as was well - known , for @xmath31 the state @xmath205 possesses squeezing but no antibunching whereas the state @xmath206 does the inverse : antibunching but no squeezing . the @xmath0ncs with the specific nonlinear function @xmath5 given by eq . ( [ 32 ] ) causes curious changes .
the state @xmath164 is squeezed in a rather narrow range of small values of @xmath128 ( see fig .
8a ) and turns from super- to sub - poisson s as @xmath128 increases ( see fig .
8b ) . as for the state
@xmath207 it remains unsqueezed but antibunched for all @xmath208 behaves qualitatively like the state @xmath209 does .
the state @xmath210 is non - squeezed but , most curiously , its number distribution crosses @xmath190 twice ( see fig .
9 ) . at small values of @xmath128 ,
the mandel parameter @xmath211 is less than @xmath212 it crosses @xmath190 becoming super - poissonian but , after reaching a maximum , drops back below @xmath190 and remains there being sub - poissonian in the whole high - value side of @xmath204 finally , no squeezing and permanent antibunching are found in the state @xmath213
we have introduced the @xmath0-quantum nonlinear coherent state . for a given @xmath0
there are @xmath0 mutually orthogonal substates belonging to the same eigenvalue .
each of such substates can be decomposed into a linear combination of @xmath0 usual nonlinear coherent states with correlated eigenvalues and weights determined respectively by eqs .
( [ 14 ] ) and ( [ 15 ] ) .
states with the same @xmath0 and @xmath1 but belonging to two different eigenvalues @xmath11 and @xmath214 are however non - orthogonal to each other and , the states with a fixed @xmath0 and all possible @xmath11 and @xmath158 are overcomplete . a harmonically trapped two - level ion driven properly by two laser beams , one in resonance with the ion s electronic transition and the other detuned to the @xmath215 lower sideband ,
is shown to realize the @xmath0-quantum nonlinear coherent state in the steady regime for suitably chosen control parameters .
these states show up two kinds of multiple self - peaking in their number distribution .
one kind is associated with large values of the lamb - dicke parameter @xmath119 but is not pronounced for @xmath216 the other kind originates from @xmath142 independent of @xmath119 and is manifested in the number distribution that fills the fock space `` periodically '' with `` period '' @xmath141 forming a multi - peaked structure . squeezing and antibunching have been investigated in detail revealing curious features corresponding to the specific nonlinear function @xmath217 which is here well - determined in the context of a trapped ion driven by lasers with the certain setup .
another setup of driving lasers , say , with @xmath218 and @xmath219 , or whatever else , would lead to another class of states .
fig . 1 : : : log(@xmath123 ) as a function of @xmath132 for @xmath136 @xmath220 @xmath221 while a ) @xmath222 a converging case , and b ) @xmath223 a diverging case .
fig . 2 : : : phase diagrams in the ( @xmath119,@xmath128)-plane for various @xmath224 whose values are indicated near the curve .
the curves , themselves , are the @xmath126 as a function of @xmath225 the corresponding @xmath0ncs exists ( does not exist ) below ( above ) the curve .
fig . 3 : : : number distributions for the states @xmath226 with @xmath221 while a ) @xmath227 and b ) @xmath228 featuring two - peaked structures .
fig . 4 : : : log(@xmath229 ) versus @xmath132 for @xmath221 and different orders @xmath0 indicated near the curve .
the curve gets less tortuous and drops much faster for a higher @xmath131 fig . 5 : : : number distributions @xmath230 versus @xmath132 for
@xmath231 @xmath232 @xmath233 while a ) @xmath234 , b ) @xmath235 and c ) @xmath236 featuring multi - peaked structures . fig . 6 : : : number distributions of the mixed state resulting from an initial coherent state @xmath174 for @xmath237 @xmath238
@xmath239 and different values of @xmath240 as indicated . the curve with @xmath241 remains unchanged for @xmath242 fig .
7 : : : mandel parameter against @xmath128 for @xmath136 @xmath234 and @xmath243 showing antibunching as opposed to the case with @xmath244 antibunching occurs as well for higher values of @xmath225 fig . 8 : : : \a ) squeezing to non - squeezing and b ) super- to sub - poisson s transition as @xmath128 increases for the state @xmath245 with @xmath246 ( similarly for higher @xmath119 )
: : : mandel parameter as a function of @xmath128 for @xmath247 @xmath235 and @xmath246 ( similarly for higher @xmath119 ) : the transition from sub - poisson s to super - poisson s and back to sub - poisson s in the course of increasing @xmath204 | we introduce a generalized class of states called @xmath0-quantum nonlinear coherent states .
each @xmath0-state has @xmath0 @xmath1-components corresponding to one and the same eigenvalue .
each @xmath2-component can be composed of @xmath0 @xmath3-states in a correlated manner .
the introduced states are shown to be realized in the long - term behavior of the vibrational motion of an ion properly trapped and laser - driven .
nonclassical properties of the states are studied in detail .
pacs numbers : 42.50.dv , 42.50.vk , 32.80.pj |
the shortest path metric of planar graphs is one of the most popular and well - studied metrics in computer science .
countless papers , surveys , and textbooks address the computational challenges that arise when dealing with it . in this paper
, we address a core problem about this metric that has remained poorly understood .
we ask : _ how compressible is it ? _ that is
, how many bits do we need , information theoretically , in order to describe a set of distances in a planar graph ? as we discuss shortly , a better understanding of this core question is crucial to making progress on some of the biggest open problems in other well - studied subjects such as sparsification , labeling schemes , dynamic algorithms , and compact routing schemes . first , let us define our problem more formally . in the _ metric compression _
problem , we are given a set @xmath12 of @xmath0 points in some metric space with distance function @xmath13 , such as the metric of distances in an @xmath1 node planar graph , and the goal is to find an encoding @xmath14 that is as short as possible , yet still allows us to compute @xmath15 for any two points @xmath16 .
given an unweighted , undirected planar graph @xmath17 on @xmath1 nodes , and a subset @xmath12 of @xmath0 distinguished nodes in @xmath17 , compute a bit string @xmath14 that encodes the distances between all pairs of nodes in @xmath12 .
that is , there is a decoding function @xmath18 that given the encoding @xmath14 and any two nodes @xmath16 returns the @xmath19-to-@xmath20 distance in @xmath17 ( i.e. , @xmath21 ) .
there are two nave ways to solve this problem .
first , we can store all the distances explicitly as a @xmath22 matrix in the encoding @xmath14 .
the distance in a graph on @xmath1 nodes is some number in @xmath23 , and so this matrix can be encoded using @xmath3 bits .
the second option , which is better whenever @xmath24 , is to let the encoding be the graph @xmath17 itself .
navely , this is @xmath25 bits , and more sophisticated encodings give @xmath2 @xcite . using the @xmath26 notation to hide polylog factors , we get a nave upper bound of @xmath27 for our problem .
is this the best possible , or could there be a much better compression into @xmath28 or even @xmath29 bits ? for context , let us look at other metrics .
one example of a metric that admits an ultra - efficient compression into @xmath29 bits is the metric of trees or bounded treewidth graphs @xcite . for most metrics of interest , however , the exact or lossless version of the compression problem is too difficult and no non - trivial upper bounds , beyond log - factor improvements , are possible .
for example , in general ( non - planar ) graphs there is a simple @xmath7 lower bound : in any compression , each of the @xmath30 possible graphs on @xmath0 nodes must be encoded differently .
instead , it is popular to seek the optimal _ lossy _ compression from which the metric can be recovered _ approximately _
, e.g. up to a multiplicative @xmath31 error .
for example , the classical johnson - lindenstrauss @xcite embedding allows one to compress a set of @xmath0 points in euclidean @xmath13-dimensional space into roughly @xmath32 bits , so that the distances between the points can be recovered up to a @xmath31 factor , and a recent breakthrough of indyk and wagner @xcite reduced the bound to roughly @xmath33 which is tight up to a @xmath34 factor . indeed ,
if we are willing to pay a @xmath31 error , there are ingenious compressions of the planar graph metric into @xmath29 bits @xcite . but
do we have to pay this error , or are planar graphs restricted enough to allow for non - trivial compression ? [ oq1 ] can we beat @xmath35 bits for planar graph metric compression ? there are some lower bounds in our way . from the seminal work of gavoille , peleg , prennes , and raz @xcite
we know that the metric of _ weighted _ planar graphs , where the edge weights are polynomially bounded , does not admit any non - trivial compression .
the authors show that any boolean @xmath22 matrix can be `` encoded '' using the distances among a set of @xmath36 nodes in a weighted planar graph on @xmath37 nodes , where the edge weights are in @xmath38 $ ] . since we can not compress an arbitrary @xmath22 matrix into @xmath39 bits , we get a nearly - tight lower bound of @xmath40 for _ weighted _ planar graphs . for unweighted planar graphs ,
gavoille et al .
simply subdivide the edges in their construction and the number of nodes in the encoding grows to @xmath41 , which leads to a much weaker lower bound of @xmath42 ( see section [ sec : lowerbound ] for more details ) .
for example ,
when @xmath4 , the upper bound is @xmath43 and the lower bound is @xmath5 .
this subdivision of edges is rather nave , and the overall lower bound construction does not seem to capture the full power of the planar graph metric .
in fact , it can be simulated by a grid graph @xcite .
this naturally suggests the following intriguing challenge of finding a more clever encoding of matrices into planar graphs , which would lead to a negative resolution to open question [ oq1 ] .
[ ch1 ] can we encode an arbitrary @xmath22 boolean matrix @xmath44 using the distances among a subset of @xmath36 nodes @xmath45 in an _ unweighted planar graph _ with @xmath46 vertices , so that we can determine @xmath47 $ ] by only looking at the distance between @xmath19 and @xmath48 in our graph ? before presenting our results , let us discuss the state of the art on questions that are closely related to ours , in which we are interested in data structures that are not only as succinct as possible , but also have other desirable features .
along the way , we give further reasons to be pessimistic about the possibility of a non - trivial compression . [ [ sparsification . ] ] sparsification .
+ + + + + + + + + + + + + + + a natural way to compress a graph is by deleting or contracting some of its edges and nodes . finding small subgraphs or minors that preserve or approximate the distances among a given subset of @xmath0 nodes
have been studied for planar graphs @xcite and for general graphs @xcite . such compressions are appealing algorithmically , since we can readily feed them into our usual graph algorithms , and recent research suggests that , in many settings , near - optimal compression bounds can be achieved using such sparsifiers ( e.g. when compressing general graphs with additive error @xcite ) .
a discouraging lower bound of krauthgamer , nguyen , and zondiner @xcite shows that even in the case of unweighted grid graphs , it is impossible to beat the nave bound using a ( possibly weighted ) minor . thus , a positive answer to open question [ oq1 ] will have to involve a more complicated data structure .
[ [ labelling - schemes . ] ] labelling schemes .
+ + + + + + + + + + + + + + + + + + an appealing way to represent graphs is to assign a label @xmath49 to each node @xmath50 , so that by looking at the labels of two nodes @xmath51 we can infer certain properties such as the distance between them @xmath52 .
finding so - called _ distance labelling schemes _ in which the labels are as short as possible is a classical subject of study @xcite .
such labels are used for efficient algorithms both in theory @xcite and practice @xcite .
an inspiring lecture by stephen alstrup at halg 2016 surveys breakthroughs @xcite achieved in this field in the last few years , all of which involve shaving constants or logarithmic factors .
a famous open question is to close the rare _ polynomial _ gap in the bounds for planar graphs that has been embarrassingly open since the work of gavoile et al .
@xcite : the upper bound is @xmath10 bits per label ( due to @xcite who shaved a log factor over @xcite ) , and the lower bound is @xmath11 . the only known technique to prove polynomial lower bounds lower bound for nearest common ancestors in trees @xcite and @xmath53 lower bound for distance in trees @xcite .
the gist of both of them is being able to argue about how much information can be shared by labels of two nodes .
if the graph is not a tree , this seems very challenging .
] is to argue that labelling schemes are one way to compress graphs , and then use facts about the limits of graph compression .
for example , the lower bound for distance labelling of planar graphs @xcite follows because labels of size @xmath54 can be used to solve the metric compression problem using @xmath55 bits , which contradicts the lower bounds above .
in fact , the tight lower bound for metric compression of _ weighted _ planar graphs leads to a tight lower bound for labelling schemes @xcite .
thus , to prove a tight lower bound of @xmath56 for labelling schemes in unweighted planar graphs , the _ only _ approach we have with current techniques is to negatively resolve open question [ oq1 ] , e.g. by accomplishing challenge [ ch1 ] .
[ [ routing - and - dynamic - algorithms . ] ] routing and dynamic algorithms .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + a compact routing scheme assigns names and tables to the nodes of a graph , so that each node @xmath57 can find out the first edge on the shortest path ( or some approximate path ) to any target node @xmath58 only using the name of @xmath58 and the local table stored at @xmath57 .
there is a vast literature on the topic , seeking the best possible tradeoff between sizes of the tables and the stretch in many different graph families ( we refer the reader to peleg s book @xcite and the extensive surveys @xcite ) .
for planar graphs , abraham , gavoille , and malkhi @xcite write : `` _ _ surprisingly , for stretch 1 , the complexity of the size of the routing tables is not known .
_ _ '' a simple upper bound is @xmath59 total table size , and an adaptation of the same gavoille et al .
construction gives a lower bound of @xmath60 @xcite .
it is likely that accomplishing challenge [ ch1 ] would resolve this gap as well .
yet another problem with similar state - of - the - art is the all pairs shortest paths problem in _ dynamic _ planar graphs . here , the goal is to have a data structure that supports efficient updates to the graph ( edge additions or removals ) , and can answer shortest path queries efficiently .
the breakthrough algorithm of fakcharoenphol and rao @xcite , and the later optimizations @xcite , achieve @xmath61 time for updates and queries .
the only framework for showing polynomial lower bound was recently proposed by abboud and dahlgaard @xcite who proved a lower bound of @xmath62 under the popular apsp conjecture @xcite .
using their framework , accomplishing challenge [ ch1 ] directly leads to a higher lower bound of @xmath63 , as is known in the weighted case .
history suggests that weighted planar graph metrics might be harder to work with , but they are never _ truly _ harder . in so many cases , a new algorithm for the unweighted case
is followed by an almost - as - good algorithm for the weighted case , a few years later .
for example , a ptas for the travelling salesman problem in the unweighted planar metric was found in 1995 @xcite , and then for the weighted case in 1998 @xcite .
perhaps it is only a matter of time until our lower bounds for the unweighted metric match the weighted .
our first result is a new compression scheme for the planar graph metric , which achieves the information theoretically best possible bit complexity , up to log - factors .
we give a _ positive _ resolution to open question [ oq1 ] , deem challenge [ ch1 ] to be infeasible , and show that unweighted planar graphs are inherently less complex than weighted ones ; in fact , they admit a polynomially more efficient metric compression .
given an unweighted undirected planar graph on @xmath1 nodes and a subset @xmath12 of @xmath0 nodes , we can return a binary encoding of length @xmath64 from which all pairwise distances in @xmath12 can be recovered exactly .
this shrinks the gap in our understanding of the planar metric compression problem from polynomial to polylogarithmic ( removing this polylogarithmic gap remains an open question ) . for comparison ,
when @xmath65 , we show that @xmath66 bits are necessary and sufficient , while in the weighted case the bound is @xmath67 .
our encoding breaks the lower bound of krauthgamer et al .
@xcite for compressions using minors , and raises the question whether it can be matched via other forms of sparsification or _
graphical _ compressions .
it is unclear whether our new compression scheme will lead to improved upper bounds for labeling , routing , or dynamic algorithms . in section [ sec : unit ]
, we discuss the difficulty in turning it into a labelling scheme .
still , it certainly shakes our beliefs about the right bounds for those problems .
even if better upper bounds are not possible , it is no longer a mere puzzle as in challenge [ ch1 ] that is standing in the way of higher lower bounds substantially new techniques and frameworks must be developed .
[ [ distance - oracles . ] ] distance oracles .
+ + + + + + + + + + + + + + + + + our first result was a mathematical advance in the understanding of the planar graph metric .
next , we use it algorithmically to achieve a new _ subset distance oracle _ that could be an appealing choice in many applications .
a distance oracle is an encoding of a graph from which a pairwise distance can be queried efficiently . since the seminal paper of thorup and
zwick @xcite , a central subject of study in graph algorithms has been to understand the inherent tradeoff between the parameters of these distance oracles ( see the survey by sommer @xcite ) : the _ size _ of the compression , the _ query time _ for returning a distance , the _ error _ in the answers , the _ preprocessing time _ to construct the compression , and so on .
many _ exact _ distance oracles for planar graphs have been proposed @xcite , mostly focusing on the tradeoff between space and query time , and giving @xmath68 space and @xmath69 query time @xcite .
cohen - addad , dahlgaard , and wulff - nilsen @xcite show that the technique of abstract voronoi diagrams recently introduced into the field of planar graphs by cabello @xcite leads to an oracle with @xmath70 space and @xmath71 query time , suggesting that a better tradeoff is possible . to get even better tradeoffs
we might allow @xmath31 error @xcite : we can achieve very small @xmath72 space and fast @xmath71 query time .
note that @xmath73 space is impossible in this setting , no matter what query time we allow .
however , another natural way to get better tradeoffs is to restrict our attention to a subset of the nodes .
a _ subset distance oracle _ is a small space data structure that can efficiently return the distance between any pair of nodes from a set @xmath12 of @xmath0 nodes . here , for any @xmath74 ,
e.g. @xmath75 , our new compression scheme suggests that a distance oracle might have @xmath73 space .
subset distance oracles arise naturally . in typical applications of distance oracles
, one can predict that all queries will be among a subset of @xmath74 nodes .
space efficiency is often a high priority .
for example , if our graph is the national road network , one might be interested in a mobile app that can return the distance between any pair of bus stops .
our second result is the first subset distance oracle with non - trivial space bounds .
notably , all previous distance oracles in the literature work equally well for weighted and unweighted graphs , while ours uses new techniques that are provably impossible for weighted graphs . given an unweighted undirected planar graph on @xmath1 nodes and a subset @xmath12 of @xmath0 nodes
, there is a polynomial - time algorithm that returns a data structure @xmath76 of size @xmath64 such that given @xmath76 and any pair of nodes in @xmath12 we can return the exact distance in @xmath77 time .
the main open question left by our work is whether our query time can be improved , perhaps all the way down to @xmath71 .
this would be an essentially optimal distance oracle .
but even as it is , our query time is sublinear , and our space is sublinear for any @xmath74 , making it an appealing choice in applications with strict space constraints .
finally , an intriguing and wide open question is to extend ( any of ) our upper bounds to _ directed _ planar graphs .
can we accomplish challenge [ ch1 ] if we allow directed edges ?
our tools heavily rely on the graph being undirected , yet it remains unclear if a higher lower bound can be proven for directed graphs .
we exhibit the first use of the _ unit - monge _ property to the algorithmic study of planar graphs .
it is well known that distances in a planar graph enjoy this property , due to the non - crossing nature of shortest paths in the plane , but prior to our work , only the _ ( non - unit ) monge _ property , was known to be algorithmically exploitable for planar graphs . for the past few decades , it has been heavily utilized in numerous algorithms for problems related to shortest paths or minimum cuts in planar graphs ( e.g. @xcite ) , and beyond , in dozens of papers on computational geometry ( e.g. @xcite ) and pattern matching ( e.g. @xcite ) . meanwhile , the stronger _ unit - monge _ property has only been exploited for algorithms on sequences where it has already led to several breakthroughs .
we refer the reader to the 159-page monograph of tiskin @xcite for an exposition of these applications .
recall that we want to encode the distances among @xmath0 nodes in a planar graph .
let us assume that we are lucky and all the nodes lie on a single face of the graph .
denote the nodes appearing on the face in order @xmath78 , and for simplicity assume that we only want to encode @xmath79-to-@xmath80 distances .
let @xmath44 be the @xmath81 matrix of distances so that @xmath47=d(s_i , t_j)$ ] .
this matrix has the monge property : for any @xmath82 we have that @xmath83 - m[i , j ] \le m[i+1,j+1 ] - m[i , j+1]$ ] .
this is because the @xmath79-to-@xmath80 shortest path and the @xmath84-to-@xmath85 shortest path must cross . moreover , it is _ unit - monge _ , that is , @xmath83-m[i , j ] \in \{-1,0,1\}$ ] .
this is because there is an edge between @xmath79 and @xmath84 and so the distances involving these nodes are always at most @xmath86 apart .
our gains come from the fact that unit - monge matrices are compressible into @xmath87 bits ! for non - unit monge matrices , the construction of gavoille et al .
implies an @xmath7 lower bound .
another striking example for the extra power of the unit - monge property is the fact that ( a compact representation of ) the distance product of two @xmath88 such matrices can be computed in @xmath89 time @xcite , while for non - unit monge matrices only @xmath90 algorithms are possible .
the main issue for us , and in general when exploiting monge properties , is that the nodes we care about do not necessarily lie on a cycle .
the simple solution is to _ add _ a cycle connecting our @xmath0 nodes and assign weight @xmath91 to the new edges so that they do not change the distances , or more formally , to _ triangulate _ the graph .
after we do this , we have the monge property , but because of the infinite weight edges , we do not have the unit - monge property .
this solution is common to all the algorithms cited above that use the monge property , and is quite reasonable when the graph is weighted to begin with . for unweighted graphs , on the other hand
, our work proves that it is too lossy and a more involved solution leads to much better results . at a very high - level ,
our approach is to use a baker - like decomposition into slices ( vertices at consecutive levels of some specific bfs tree ) whose boundaries are cycles , and to store distances to the slice boundaries . observe that when we argued above that the unit monge property holds because there is an edge between @xmath79 and @xmath84 , we did not require that there is also an edge between @xmath80 and @xmath85 . in our solution
there is an edge between consecutive vertices on the boundary cycle of each slice .
therefore , even if we triangulate each slice using infinite weight edges , we can still exploit the unit monge property when storing distances between certain vertices in a slice and the slice boundary .
the decomposition into slices is such that , after triangulation , the slices have small cycle separators .
we recursively separate the vertices of the set @xmath12 within each slice using small cycle separators .
we store distances between separators and the slice boundary ( using the unit monge property ) and between vertices of @xmath12 and separators ( using the fact that separators are small ) .
significant technical issues arise with the nesting structure of slices .
this gives rise to so - called _ holes _ in a slice .
dealing with multiple holes requires a detailed study of additional structural properties , and a more complicated recursive solution based on these properties . in essence , we show that whenever a nave solution does not work in the presence of multiple holes , there is one hole that can be handled efficiently using a different approach .
we believe it is very likely that other problems in unweighted planar graphs can be solved by exploiting the unit - monge property .
our near - optimal metric compression serves as a proof of concept that this is possible .
however , technical challenges might have to be overcome in each specific application .
in particular , the fast distance product algorithm for unit - monge matrices @xcite appears to be a strong and relevant technique that we are so far unable to exploit for solving problems in planar graphs .
we assume basic familiarity with planar graphs and planar graph duality . we denote the primal graph by @xmath17 and the dual graph by @xmath92 . for a spanning tree @xmath93 of @xmath17
, we use @xmath94 to denote the spanning tree of @xmath92 .
it is well known @xcite that the set of edges of @xmath17 not in @xmath93 form a spanning tree @xmath94 of @xmath92 .
we often refer to @xmath94 as the _ cotree _ of @xmath93 @xcite . for a spanning tree @xmath93 of @xmath17 and an edge @xmath95 of @xmath17 not in @xmath93 ,
the _ fundamental cycle _ of @xmath95 with respect to @xmath93 in @xmath17 is the simple cycle consisting of @xmath95 and the unique simple path in @xmath93 between the endpoints of @xmath95 .
given an assignment of nonnegative weights to the faces of @xmath17 , we say that a simple cycle @xmath96 is a _ balanced separator _ if the total weight of faces strictly enclosed by @xmath96 and the total weight of faces not enclosed by @xmath96 are each at most @xmath97 of the total weight . or @xmath98
. however , in our particular application @xmath97 turns out to be more convenient .
] we often assign weights to vertices rather than to faces . finding a balanced separator with respect to vertex weights
reduces to the case of face weights ( for each vertex , simply remove its weight and add it to an incident face ) .
it is well known ( see , e.g. , @xcite ) that in triangulated planar graphs there exists a balanced separator that is a fundamental cycle assuming that no face has more than @xmath99 of the total weight ( in fact , this is true for any planar graph such that @xmath94 has maximum degree 3 ) . for vertex - weights , if no vertex has more than 1/2 of the total weight and the graph is triangulated and there are no self loops then by evenly transferring the weights to faces we obtain that no face receives more than 1/2 of the total weight ( because every node is incident to at least two faces ) and we can invoke the face - weights version of the balanced separator .
many planar graph algorithms triangulate the graph by adding edges to ensure that short balanced cycle separators exists .
the lengths of the added edges is set to be sufficiently large so as not to change distances in the graph .
this is clearly not possible in unweighted planar graphs , and is one of the obstacles we will need to overcome .
one of the main tools we use for succinct representation of distances in unweighted undirected planar graphs is the _ unit monge property_. we next prove a sequence of lemmas that utilize this property to efficiently store distances between vertices on cycles .
we begin with encoding distances between disjoint sets of vertices on a single face ( lemma [ lem : unit ] ) , then encoding all - pair distances between the vertices on a single face ( lemma [ lem : unit2 ] ) , then encoding all - pair distances between the vertices on a single simple cycle ( lemma [ lem : cycle ] ) , and finally , encoding the distances between the vertices of two faces ( lemma [ lem : hole ] ) .
[ lem : unit ] let @xmath100 be the cyclic walk of a face of a planar graph partitioned into two parts @xmath101 and @xmath102 . then ,
for any subset @xmath103 of @xmath104 , all distances between vertices of @xmath105 and vertices of @xmath103 can be encoded in @xmath106 bits .
let @xmath107 .
we define an @xmath108 matrix @xmath44 such that @xmath47 $ ] equals the distance in @xmath17 between @xmath19 and @xmath109 .
the matrix @xmath44 is monge , that is @xmath110 - m[i , j ] \le m[i+1,j+1 ] - m[i , j+1]\ ] ] for any @xmath111 $ ] and @xmath112 $ ] .
this is because the shortest @xmath19-to-@xmath113 and @xmath114-to-@xmath115 paths must necessarily cross .
furthermore , the matrix @xmath44 is unit - monge , that is @xmath110-m[i , j ] \in \{-1,0,1\}\ ] ] for any @xmath111 $ ] and @xmath116 $ ] , because there is an edge @xmath117 .
consequently , for any @xmath111 $ ] , the sequence of differences @xmath83-m[i , j]$ ] is nondecreasing and contains only values from @xmath118 , so can be encoded by storing the positions of the first 0 and the first 1 . storing these positions for every @xmath119 $ ] takes @xmath120 bits . to encode the whole @xmath44 , we additionally store @xmath121 $ ] for every @xmath116 $ ] using @xmath122 bits .
[ lem : unit2 ] let @xmath100 be the cyclic walk of a face of a planar graph .
then , all distances between vertices of @xmath96 can be encoded in @xmath123 bits .
we recursively encode all distances between vertices from a contiguous fragment of @xmath96 using lemma [ lem : unit ] .
we start with the whole @xmath124 . to encode the distances between all vertices @xmath125 , where @xmath126 , we set @xmath127 and proceed as follows : 1 .
recursively encode the distances between all vertices @xmath128 .
recursively encode the distances between all vertices @xmath129 .
3 . apply lemma [ lem : unit ] with @xmath130 and @xmath131 .
the total size of the encoding is described by the recurrence @xmath132 , hence solves to @xmath123 .
[ lem : cycle ] let @xmath100 be a simple cycle in a planar graph .
then , all distances between vertices of @xmath96 can be encoded in @xmath123 bits .
consider the two planar graphs @xmath133 ( @xmath134 ) obtained by removing all vertices enclosed ( not enclosed ) by @xmath96 .
@xmath96 is the cyclic walk of a face in @xmath134 and @xmath133 , hence we can apply lemma [ lem : unit2 ] to store the distances in @xmath133 and in @xmath134 between vertices of @xmath96 .
this is enough to encode the distances in @xmath17 between vertices of @xmath96 , as any such shortest path can be partitioned into shortest paths between two vertices of @xmath96 such that each of these paths exists in either @xmath134 or @xmath133 .
[ lem : hole ] let @xmath135 and @xmath136 be the cyclic walk of two faces of a planar graph .
then , all distances between a prefix @xmath137 of @xmath138 and any subset @xmath139 of @xmath140 can be encoded in @xmath141 bits .
we first choose a shortest path @xmath142 between @xmath137 and @xmath140 and let @xmath19 and @xmath143 be its endpoints .
we make an incision along @xmath142 and apply lemma [ lem : unit ] to encode the distances between @xmath137 and @xmath139 corresponding to shortest paths that do not cross @xmath142 using @xmath141 bits of space .
it remains to encode distances corresponding to shortest paths that do cross @xmath142 . without loss of generality
@xmath142 connects @xmath144 and @xmath145 .
we orient @xmath138 and @xmath140 so that after making an incision along @xmath142 the vertices @xmath146 and @xmath147 are adjacent to the endpoints of @xmath142 .
.[fig : cross],scaledwidth=40.0% ] consider a shortest path @xmath148 from @xmath19 to @xmath143 crossing @xmath142 , see figure [ fig : cross ] .
because both @xmath142 and @xmath148 are shortest paths , @xmath148 can be assumed to cross @xmath142 exactly once .
similarly , a shortest path @xmath149 from @xmath114 to @xmath150 crossing @xmath142 can be assumed to cross @xmath142 exactly once .
we claim that @xmath148 must cross @xmath149 .
otherwise , by considering an incision along @xmath148 we can conclude that @xmath149 crosses @xmath142 an even number of times but this is a contradiction .
therefore , any such @xmath148 and @xmath149 must cross .
this means that the matrix @xmath44 , where @xmath47 $ ] is set to be the length of a shortest path between @xmath19 and @xmath143 crossing @xmath142 once , is monge .
that is , @xmath151 + m[i , j ] \geq m[i+1,j]+m[i , j+1].\ ] ] additionally , @xmath47-m[i+1,j ] \in \{-1,0,1\}$ ] because @xmath117 is an edge .
we can hence apply the reasoning from lemma [ lem : unit ] to encode @xmath44 using @xmath141 bits .
our encoding is based on decomposing the input graph @xmath17 into _ slices_. to define the slices , recall the _
face - vertex incidence graph
_ @xmath152 of a planar graph @xmath17 : it has a vertex for every vertex @xmath50 of @xmath17 and a vertex for every face @xmath18 of @xmath17 , and if a vertex @xmath50 of @xmath17 is incident to a face @xmath18 of @xmath17 then there is an edge between their corresponding vertices in @xmath152 .
we run a breadth - first search in @xmath152 , starting from the node representing the infinite face of @xmath17 . after every even number of steps
, the yet unexplored part of the graph can be decomposed into a number of connected components , the boundary of each being a simple cycle .
more formally , we assume that the infinite face of @xmath17 is a triangle by enclosing the whole graph in a triangle , which is connected to one of the original vertices with a single edge .
we define the level of a face @xmath18 or a vertex @xmath50 of @xmath17 to be its depth in the bfs tree of @xmath152 .
thus , e.g. , the level of the infinite face of @xmath17 is zero , and the level of the vertices incident to the infinite face of @xmath17 is 1 . for each even integer @xmath153 , we define @xmath154 to be the set of connected components of the subgraph of @xmath17 induced by the faces with level at least @xmath155 .
we call each component @xmath156 a level-@xmath155 component .
we use a tree @xmath157 called the _ component tree _ of @xmath17 to capture the nesting of level components .
the nodes of @xmath157 are the level components of @xmath17 .
a level component @xmath158 is an ancestor of a level component @xmath159 in @xmath157 if the set of faces in @xmath158 contains the set of faces in @xmath159 .
since we assume that the infinite face of @xmath17 is a simple triangle , @xmath160 is indeed a tree whose root is the component corresponding to the set of all faces of @xmath17 except for the infinite face .
the boundary of a component @xmath158 is the set of edges that are incident to a face in @xmath158 and to a face not in @xmath158 .
it is not difficult to see that the boundary of each component @xmath158 is a simple cycle in @xmath17 , and that the boundaries of different components are edge - disjoint .
see @xcite for these and other properties of components and the component tree . for a node @xmath161 ,
we associate @xmath0 with the boundary cycle @xmath162 of the level component represented by @xmath0 , and define the cost of @xmath0 denoted @xmath163 to be the length of @xmath162 . for example
, for the root @xmath164 of @xmath157 we have that @xmath165 is a triangle and that @xmath166 .
[ lem : slices ] for any @xmath167 , there exists @xmath168 such that the total cost of all nodes of @xmath157 at depth @xmath169 is @xmath170 .
@xmath171 because cycles corresponding to the nodes of @xmath172 are pairwise edge disjoint .
let @xmath173 consist of all nodes of @xmath172 at depth @xmath169 .
then @xmath174 for @xmath175 and @xmath176 , so there exists @xmath168 such that @xmath177 as claimed . to define the slices we apply lemma [ lem : slices ] and call the nodes of @xmath172 at depth @xmath169 _ marked_. the root of @xmath172 is also marked .
then , for every marked node @xmath178 , the slice of @xmath50 is the subgraph of @xmath17 enclosed by @xmath179 and not strictly enclosed by @xmath180 for any marked descendent @xmath181 of @xmath50 .
the embedding of slices is inherited from the embedding of @xmath17 .
thus , the boundary of the infinite face of the slice @xmath57 of @xmath50 is @xmath179 .
the cycle @xmath179 is also called _ the boundary of the slice @xmath57_. each cycle @xmath180 corresponding to a marked descendant @xmath181 of @xmath50 such that there are no other marked nodes on the @xmath50-to-@xmath181 path becomes a face in the slice @xmath57 .
such a face is called a _ hole _ of @xmath57 , and @xmath180 is called the _ boundary of the hole_. note that , by definition , @xmath180 is the boundary of the level component that is embedded in the hole @xmath181 .
because the total cost of all marked nodes is @xmath170 and the cost of the root is 3 , the total size of all boundaries in all slices is @xmath170 .
additionally , by construction , for any slice @xmath57 , a breadth - first search of @xmath182 , the face - vertex incidence graph of @xmath57 , starting at the infinite face of @xmath57 , terminates after @xmath183 iterations and every hole is a leaf in the obtained breadth - first search tree
. by definition of slices , each slice contains faces and vertices at @xmath183 consecutive levels .
we would like to use in our solution short ( i.e. , @xmath183 ) fundamental cycle separators within each slice .
however , the diameter of a slice is not necessarily @xmath183 because face sizes may be large . to deal with this issue
we triangulate the faces so that a bfs tree of a slice will have depth @xmath183 , and will be consistent with the bfs tree of @xmath182 .
[ lem : triangulation ] we can triangulate all faces of a slice @xmath57 so that a bfs starting from the external face produces a spanning tree @xmath184 with the property that vertex @xmath50 is the parent of vertex @xmath181 in @xmath184 if and only if @xmath50 is the grandparent of @xmath181 in the bfs tree of @xmath182 .
let @xmath185 be the bfs tree of @xmath182 .
if @xmath50 and @xmath181 are incident to the same face in @xmath57 , and @xmath50 is a grandparent of @xmath181 , and @xmath186 is not an edge in @xmath57 , we add @xmath186 as an artificial triangulation edge to @xmath57 . adding these edges can be done consistently with the embedding of @xmath57 because the path in @xmath185 can be embedded on the same plane as @xmath57 such that @xmath57 and @xmath185 only intersect at vertices of @xmath57 .
see figure [ fig : slices ] .
we introduce an artificial vertex @xmath187 embedded in the infinite face of @xmath57 and triangulate the infinite face of @xmath57 by adding edges between @xmath187 and every vertex of the infinite face of @xmath57 .
similarly , we triangulate each hole @xmath188 of @xmath57 by introducing an artificial vertex @xmath189 , embedded in @xmath188 , and adding edges between @xmath189 and every vertex on the boundary of @xmath188 .
any remaining non - triangulated faces can be triangulated arbitrarily . since for every grandparent to grandchild path in @xmath185
there is a corresponding edge in the triangulation of @xmath57 , there exists a bfs tree @xmath184 rooted at the artificial vertex @xmath187 that satisfies the statement of the lemma .
note that all the artificial vertices embedded in holes of @xmath57 are leaves of @xmath184 , and hence satisfy the statement of the lemma vacuously . with two holes
are shown in solid black . in this example @xmath190 .
edges of @xmath185 , the bfs tree of @xmath191 ) are shown in double red lines .
only some of the edges of @xmath185 are shown to avoid clutter .
artificial triangulation edges are shown in dashed gray .
the bfs tree @xmath184 of the triangulation of @xmath57 is shown in blue .
[ fig : slices ] ] with two holes are shown in solid black . in this example @xmath190 .
edges of @xmath185 , the bfs tree of @xmath191 ) are shown in double red lines .
only some of the edges of @xmath185 are shown to avoid clutter .
artificial triangulation edges are shown in dashed gray . the bfs tree @xmath184 of the triangulation of @xmath57
is shown in blue .
[ fig : slices ] ] with two holes are shown in solid black . in this example @xmath190 .
edges of @xmath185 , the bfs tree of @xmath191 ) are shown in double red lines . only some of the edges of @xmath185 are shown to avoid clutter .
artificial triangulation edges are shown in dashed gray .
the bfs tree @xmath184 of the triangulation of @xmath57 is shown in blue . [
fig : slices],scaledwidth=80.0% ] with two holes are shown in solid black . in this example @xmath190 .
edges of @xmath185 , the bfs tree of @xmath191 ) are shown in double red lines .
only some of the edges of @xmath185 are shown to avoid clutter .
artificial triangulation edges are shown in dashed gray .
the bfs tree @xmath184 of the triangulation of @xmath57 is shown in blue .
[ fig : slices],scaledwidth=80.0% ] let @xmath192 be the graph obtained from the slice @xmath57 after applying lemma [ lem : triangulation ] .
let @xmath184 be the bfs tree of @xmath192 .
note that , any fundamental cycle @xmath96 w.r.t .
@xmath184 consists of two paths in @xmath184 , each consisting of @xmath183 vertices due to the triangulation .
however , @xmath96 may use edges that are not original edges of @xmath57 ( i.e. , artificial triangulation edges ) .
we do not want to consider such edges when dealing with distances , because distances in @xmath192 differ from distances in @xmath57 . to this end
we use the notion of a jordan curve . a jordan curve in @xmath57 is an embedded curve that intersects the embedding of @xmath57 only at vertices of @xmath57 .
since the embedding of the triangulation @xmath192 is consistent with that of @xmath57 , each path in @xmath184 is a jordan curve in @xmath57 .
we say that @xmath184 is a _ jordan tree _ in @xmath57 .
in particular , any fundamental cycle w.r.t .
@xmath184 is a jordan cycle ( closed jordan curve ) in @xmath57 .
we next describe how the tree @xmath184 can be used to recursively decompose @xmath57 into subgraphs called _
regions_. a region @xmath193 is a subgraph of @xmath57 .
the boundary of @xmath193 is defined as the set of vertices of @xmath193 that are incident ( in @xmath17 ) to both an edge in @xmath193 and to an edge not in @xmath193 .
thus , for example , the boundary of the region consisting of the entire slice @xmath57 consists of the external boundary @xmath138 of @xmath57 and of the boundaries of all the holes of @xmath57 .
let @xmath193 be a region .
let @xmath96 be a fundamental cycle w.r.t .
the tree @xmath184 may contain edges that are not edges of @xmath193 ( either because they are triangulation edges , or because they are edges of @xmath57 that do not belong to the region @xmath193 ) . since the embedding of @xmath184 is consistent with the embedding of any subgraph of @xmath57 , @xmath96 is a jordan cycle in @xmath193 .
the operation of separating @xmath193 using @xmath96 yields two subgraphs .
one is the subgraph induced by the faces of @xmath193 strictly enclosed by @xmath96 and the other is the subgraph induced by the faces of @xmath193 not strictly enclosed by @xmath96 .
this view of @xmath184 as a jordan tree in any region allows us to reuse the same tree @xmath184 throughout the recursive decomposition .
this recursive process can be described by a binary tree @xmath194 .
each node @xmath50 of @xmath194 corresponds to a region ( subgraph ) @xmath195 of @xmath57 .
the root of @xmath194 is the entire slice @xmath57 .
each non - leaf node @xmath50 of @xmath194 is associated with a ( jordan ) fundamental cycle separator of @xmath184 , which we denote @xmath196 .
the regions of the two children of @xmath50 are the regions obtained by separating @xmath195 with the jordan cycle @xmath196 .
we begin with the simplified case , in which we assume that each slice has a single hole .
this is the case , for example , when the input planar graph is a grid ( with possibly subdivided edges ) .
first we use lemmas [ lem : cycle ] and [ lem : hole ] to store , for each slice @xmath57 with external boundary @xmath138 , and a single hole @xmath188 with external boundary @xmath197 the following distances .
the * boundary - to - boundary * distances : the distances ( in @xmath57 ) among the vertices of @xmath138 , and the * hole - to - boundary * distances : the distances ( in @xmath57 ) between the vertices of @xmath138 and the vertices of @xmath197 .
boundary - to - boundary and hole - to - boundary distances encode distances `` between slices '' .
we also need to encode distances `` within slices '' .
we will use the fact that @xmath57 has a spanning tree of depth @xmath183 to decompose @xmath57 into regions , each containing a single distinguished node ( i.e. , node of @xmath12 ) , and having a boundary that consists of @xmath183 vertices .
then we can afford to store , for each distinguished node , its distance to the boundary of its region , and , using the unit - monge property , to also store the distances between the @xmath183 vertices on the boundary of each region @xmath193 to the vertices of @xmath138 and @xmath197 ( i.e. , the boundary of @xmath57 ) that belong to @xmath193 .
these distances will suffice for reconstructing the distance between any pair of distinguished nodes .
let @xmath198 denote the set of distinguished vertices in slice @xmath57 .
we use fundamental ( jordan ) cycle separators w.r.t . the tree @xmath184 to recursively divide @xmath57 into regions , until each region contains a single distinguished vertex . at each recursive
step we separate a region @xmath193 into two subregions by choosing a fundamental cycle separator w.r.t . @xmath184
that balances the number of distinguished vertices in @xmath193 ( i.e. , assigning unit weight to each distinguished vertex in @xmath193 and zero weight to all other vertices ) .
note that , since we use balanced separators , the depth of the recursion tree @xmath194 is @xmath199 .
recall that the fundamental cycle separators w.r.t .
@xmath184 do not cross each other , and , by construction of @xmath184 in lemma [ lem : triangulation ] , each fundamental cycle separator crosses each of the external boundary of @xmath57 and the hole of @xmath57 at most twice .
therefore , the boundary of each region @xmath195 corresponding to a node @xmath50 in the recursive decomposition tree @xmath194 contains @xmath200 vertex disjoint maximal subpaths of @xmath138 , and @xmath200 vertex disjoint maximal subpaths of @xmath197 . at the step of the recursive decomposition corresponding to node @xmath201 with separator @xmath196 and two children @xmath202 ,
we store * @xmath203-to - separator * distances : explicitly store the distances ( in @xmath195 ) between every vertex of @xmath12 in @xmath195 and every vertex of @xmath196 , * separator - to - boundary * distances and * separator - to - hole * distances : for @xmath204 , the distances ( in @xmath205 ) between every vertex of @xmath196 and every maximal subpath of @xmath138 or @xmath197 on the boundary of @xmath205 , using lemma [ lem : unit ] or lemma [ lem : hole ] ( depending whether they lie on a single or two faces of @xmath205 ) . finally , for every leaf @xmath206 , we store * @xmath12-to - boundary * distances and * @xmath12-to - hole * distances : the distance between the unique distinguished vertex in @xmath195 to every vertex of @xmath138 or @xmath197 on the boundary of @xmath195 . [
[ analysis . ] ] analysis .
+ + + + + + + + + we first show that the total space is @xmath207 , and then show that the distances between any pair of vertices in @xmath12 can be recovered using just the information we stored . since the total size of all slice boundaries is @xmath170 , storing the boundary - to - boundary distances and the hole - to - boundary distances takes @xmath208 using lemmas [ lem : cycle ] and [ lem : hole ] .
since the depth of @xmath194 is @xmath200 , each vertex of @xmath198 belongs to @xmath200 regions in the decomposition of @xmath57 .
since , in addition , @xmath209 for every @xmath206 , the total space required for storing the @xmath12-to - separator distances is @xmath210 .
consider a region @xmath193 of a slice @xmath57 .
recall that the vertices of @xmath138 ( @xmath197 ) that belong to @xmath193 lie on @xmath211 vertex disjoint maximal subpaths of @xmath138 ( @xmath197 ) .
the endpoints of each such maximal subpath may belong to another region @xmath212 at the same depth in @xmath194 .
therefore , @xmath193 shares @xmath211 vertices of @xmath138 ( @xmath197 ) with other regions at the same depth in @xmath194 . finally , recall that the number of regions of @xmath57 is @xmath213 .
therefore , using lemma [ lem : unit ] or lemma [ lem : hole ] , the total space for storing the separator - to - boundary and separator - to - hole distances is @xmath214 . in more detail , let @xmath215 be the total number of slice / hole boundary vertices in the @xmath155-th slice .
then , in every slice every boundary / hole vertex that is not an endpoint of a maximal subpath contributes at most once at each level of recursion . at each level
, we have at most @xmath0 recursive calls , so at most @xmath216 maximal subpaths and at most @xmath0 fundamental cycle separators .
therefore , the total space is @xmath217 .
storing @xmath12-to - boundary distances and @xmath12-to - hole distances at the leaves of the recursion tree requires total @xmath218 bits since each boundary or hole vertex belongs to exactly one leaf region , except for @xmath219 vertices ( endpoints of maximal subpaths ) . choosing @xmath220 proves the space bound . finally , we prove that the distances between any pair of vertices in @xmath12 can be recovered using just the information we stored . for any @xmath221
, if a shortest @xmath222-to-@xmath223 path does not leave @xmath57 then @xmath224 , and the distance can be obtained using the @xmath12-to - separator distances stored in the lowest common ancestor of the regions of @xmath222 and @xmath223 in @xmath194 .
otherwise , let @xmath142 be a shortest path between vertices @xmath225 and @xmath226 ( where @xmath192 is either @xmath57 or , wlog , enclosed by the hole of @xmath57 ) .
let @xmath227 $ ] denote the subpath of @xmath142 between vertices @xmath155 and @xmath228 .
let @xmath50 be the first vertex of @xmath142 that belongs to the boundary of @xmath192 or to the boundary of a hole of @xmath192 .
if @xmath229 $ ] contains some vertex of a fundamental cycle separator used in processing @xmath192 , let @xmath181 be the last vertex of @xmath142 that belongs to the earliest such separator .
if @xmath181 does not exist , then the length of @xmath229 $ ] is stored as an @xmath12-to - boundary or an @xmath12-to - hole distance .
if @xmath181 exists then the length of @xmath230 $ ] is stored as a @xmath12-to - separator distance , and the length of @xmath231 $ ] is stored as a separator - to - boundary or separator - to - hole distance .
let @xmath232 be the last vertex of @xmath142 that belongs to the boundary of @xmath57 .
the length of @xmath233 $ ] can be computed from boundary - to - boundary and hole - to - boundary distances since @xmath233 $ ] can be decomposed into subpaths between boundary vertices of slices .
the length of the suffix @xmath234 $ ] can be computed in a similar manner to that of the prefix @xmath229 $ ] .
a difficulty that arises in the presence of multiple holes is that since the number of holes is not bounded by a constant , we can not afford to store distances involving holes .
for example , storing hole - to - boundary distances between the external boundary @xmath138 of a slice @xmath57 and the boundary of each hole of @xmath57 requires @xmath235 bits per hole .
since the number of holes can be @xmath236 , the total space could be @xmath237 .
the role of storing distances involving boundaries of holes was to allow the recovery of distances to distinguished vertices enclosed in these holes .
we modify our approach for processing a slice @xmath57 to take into account the distinguished vertices enclosed in holes of @xmath57 as well as the distinguished vertices in @xmath57 itself . as in the single hole case
, the slice @xmath57 will be recursively divided using fundamental cycle separators . for any region @xmath193 encountered along the recursive process , let @xmath238 denote the subset of the distinguished vertices in @xmath193 , as well as those enclosed by any hole in @xmath193 .
thus , for example , @xmath239 is the set of all vertices in @xmath12 that are enclosed ( in @xmath17 ) by the external boundary of slice @xmath57 .
we say that a jordan cycle separator @xmath96 of a region @xmath193 is _ good _ if it is balanced w.r.t .
@xmath238 and does not go through any hole of @xmath193 .
the problem with jordan separators that go through some hole @xmath188 is that they partition the distinguished vertices enclosed by @xmath188 in an unspecified way since these distinguished vertices are not represented in @xmath193 .
it is not hard to see that if a good separator always exists then we do not need to store any distances involving holes . in reality
we can not always find a good separator .
consider , for example , the case where some hole @xmath188 of a region @xmath193 encloses most of the vertices of @xmath238 . clearly , a separator that is balanced w.r.t .
@xmath238 must go through @xmath188 .
thus , there is no good separator in such a case .
we show , however , that we can always either find a good separator , or there exists some hole ( which we call a _ disposable _ hole ) that can be dealt with in a special way .
this is reminiscent of recursive procedures based on heavy path decomposition , where heavy nodes ( disposable holes in our case ) are treated differently than light ones .
we guarantee that , in either case , each resulting subregion contains only a constant fraction of @xmath238 , so the depth of the recursion is @xmath211 .
we next explain the details .
[ [ good - separators - and - disposable - holes . ] ] good separators and disposable holes .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + let @xmath193 be a region .
we define the weight of each vertex @xmath50 of @xmath193 to be @xmath86 if @xmath50 is a distinguished vertex . for each hole @xmath188 of @xmath193
, we define the weight of the artificial vertex @xmath189 embedded in @xmath188 to be the number of distinguished vertices strictly enclosed ( in the whole graph @xmath17 ) by the boundary of @xmath188 .
all other vertices are assigned weight zero .
recall that a cycle separator is _ good _ if it does not go through any hole .
we would like to separate @xmath193 using a good fundamental cycle separator @xmath240 of some edge @xmath95 w.r.t .
if we can find such separator @xmath240 where @xmath95 is not incident to @xmath189 for some hole @xmath188 , then @xmath240 is a good separator ( since the vertices @xmath189 are leaves of the spanning tree @xmath184 ) . otherwise , we must separate @xmath193 with a fundamental cycle separator that goes through holes .
we next define _ disposable _
holes , and then show that we can allow the fundamental cycles to go through such holes .
let @xmath0 be a node ( level component ) in @xmath157 .
let @xmath162 be the boundary cycle of @xmath0 .
let @xmath95 be an edge of @xmath162 .
note that @xmath241 .
this is because both endpoints @xmath95 have the same level , so , by lemma [ lem : triangulation ] , neither can be the parent of the other in @xmath184 .
let @xmath242 be the endpoints of @xmath95 in the dual graph , such that @xmath18 is a face in @xmath0 and @xmath243 a face not in @xmath0 .
since @xmath241 , @xmath95 is in the cotree @xmath244 .
consider breaking @xmath244 into two subtrees by deleting @xmath95 .
we say that the edge @xmath95 is _ light _ if the subtree of @xmath244 that contains @xmath243 has weight at most @xmath245 where @xmath246 is the total weight of the vertices of @xmath193 .
note that we defined weights of primal vertices , whereas the vertices of @xmath244 are primal faces . to define face weights ,
evenly redistribute the weight of each vertex among all of its incident faces .
there is an equivalent , primal view of light edges : the jordan cycle @xmath240 partitions @xmath193 into two subgraphs , exactly one of which contains the faces of the level component corresponding to @xmath0 .
we say @xmath95 is _ light _ if the weight of the subgraph that does not contain the level component @xmath0 is at most half the weight of @xmath193 .
we say that a level component @xmath0 is _ disposable _ in region @xmath193 if there are boundary edges of @xmath0 in @xmath193 , and if every edge @xmath95 of the boundary of @xmath0 that is also in @xmath193 is light .
note that , in particular , this definition applies to holes ( since holes are level components ) .
see figure [ fig : disposable - hole ] . are shown in double - line grey .
since the boundary of a hole is a level cycle , none of the edges of the boundary of @xmath188 belongs to the spanning tree @xmath184 ( blue ) .
the artificial triangulation vertex @xmath189 of @xmath188 and the triangulation edges ( grey dashed ) are shown .
the cotree @xmath247 is shown in thin red .
suppose that the number of distinguished nodes enclosed by @xmath188 is at least @xmath245 ( so the weight of @xmath189 is at least @xmath245 ) .
then , for any edge @xmath95 of the boundary of @xmath188 , the part of @xmath248 that does not contain faces of @xmath188 weighs at most @xmath245 , so @xmath95 is a light edge and @xmath188 is a disposable hole . [
fig : disposable - hole],scaledwidth=40.0% ] before showing why disposable holes exist and that they are useful , we first mention a simple property of @xmath247 and then use it to prove the existence of disposable holes .
[ dualtree ] the cotree @xmath247 enters each level component exactly once .
the spanning tree @xmath184 is monotone with respect to node levels .
thus , if @xmath95 is an edge of the boundary of a level component @xmath0 , then one of the components of @xmath249 contains no other faces , vertices or edges of @xmath0 .
see figure [ fig : tree - cotree ] for an illustration .
( blue ) , the cotree @xmath244 ( red ) , and boundary of level components ( black cycles ) . since @xmath184 is monotone with respect to levels
, @xmath244 enters each level component exactly once .
[ fig : tree - cotree],scaledwidth=50.0% ] [ lem : recursion ] if a region @xmath193 contains more than one vertex with non - zero weight , then there exists either a good balanced fundamental cycle separator or a disposable hole in @xmath193 .
let @xmath246 be the total weight of vertices in @xmath193 .
consider the component tree @xmath157 .
let @xmath181 be a deepest disposable component in @xmath157 such that @xmath180 has an edge in @xmath193 . if @xmath181 is a hole of @xmath193 then we found a disposable hole , and we are done .
otherwise , we next show that there exists a good separator .
let @xmath250 be the children of @xmath181 in @xmath157 ( if there is no disposable component in @xmath193 , then define @xmath181 to be @xmath193 , @xmath180 to be the external boundary of @xmath193 , and let @xmath251 be the set of rootmost components in @xmath157 such that @xmath252 has an edge in @xmath193 ) .
since none of the @xmath253 s is disposable , for each @xmath253 there exists exactly one boundary edge @xmath254 ( here we wrote @xmath255 as a dual edge , and @xmath256 is the endpoint of @xmath255 that belongs to @xmath253 ) , such that the subtree of @xmath257 that contains @xmath258 has weight at least @xmath245 .
consider the following two phase process ( see figure [ fig : structural - lemma ] for an illustration ) : let @xmath259 .
if @xmath260 contains more than a single face of some @xmath143 ( in which case it must contain all faces of @xmath143 by property [ dualtree ] ) , then @xmath261 is obtained from @xmath260 by rooting @xmath260 at @xmath262 and deleting all the strict descendants of @xmath263 in @xmath260 , so that @xmath263 becomes a leaf .
the weight assigned to @xmath263 in @xmath261 is the total weight of all the vertices in the deleted subtree .
thus , the weight of @xmath261 remains @xmath246 , and , by definition of @xmath264 , the weight of @xmath263 is at most @xmath245 .
the first phase terminates when @xmath260 contains at most one face ( @xmath263 ) from each @xmath143 .
in the second phase , while @xmath260 contains an edge @xmath95 of @xmath180 that is not a leaf edge of @xmath260 , then @xmath261 is obtained from @xmath260 by rooting @xmath260 at the endpoint @xmath243 of @xmath95 that belongs to @xmath181 , and deleting all the strict descendants of the other endpoint @xmath18 of @xmath95 in @xmath260 , so that @xmath18 becomes a leaf .
similarly to the first phase , the weight of @xmath18 in @xmath261 is set to the total weight of all the vertices in the deleted subtree . since @xmath181 is disposable , the weight of @xmath18 is at most @xmath245 .
let @xmath265 be the resulting tree .
since @xmath265 contains at most one face from each @xmath253 , @xmath265 contains no triangulation edges of a hole ( both endpoints of a triangulation edge of a hole belong to the hole ) .
furthermore , the total weight of @xmath265 is @xmath246 , and every leaf of @xmath265 created during the two phase process has weight at most @xmath245 ( by definition ) . for the remaining nodes of @xmath265 ,
the degree is at most 3 and the weight is also at most @xmath245 , because the original weights in @xmath247 are at most @xmath245 ( otherwise , the node corresponds to a hole of weight at least @xmath245 that is , by definition , disposable , and we are done ) .
therefore , there exists an edge @xmath95 whose deletion from @xmath265 results in two trees , none of which weighs more than @xmath266 . by construction of the weights of @xmath265 , the balance of the fundamental cycle of @xmath95 w.r.t .
@xmath184 is exactly the ratio of the weights of the subtrees obtained by deleting @xmath95 from @xmath265 .
therefore , the fundamental cycle @xmath240 of @xmath95 w.r.t .
@xmath184 is a balanced jordan cycle separator . since no edge of @xmath265 is a triangulation edge of a hole
, @xmath240 is a good separator . according to which a good separator is found .
three level boundaries are shown ( black cycles ) .
the external one is @xmath180 , the two cycles enclosed by @xmath180 are @xmath267 and @xmath268 .
the subfigures show the process of splitting the cotree @xmath247 , first at @xmath269 , then at @xmath270 . in the second phase the cotree is further split at the two remaining edges of @xmath180 .
the resulting tree @xmath265 is an induced subtree of @xmath247 in which a balanced edge - separator can be found .
since none of the edges of @xmath265 has both endpoints in any @xmath253 , none of the edges of @xmath265 are triangulation edges of a hole .
[ fig : structural - lemma],title="fig:",scaledwidth=30.0% ] according to which a good separator is found .
three level boundaries are shown ( black cycles ) .
the external one is @xmath180 , the two cycles enclosed by @xmath180 are @xmath267 and @xmath268 .
the subfigures show the process of splitting the cotree @xmath247 , first at @xmath269 , then at @xmath270 . in the second phase the cotree is further split at the two remaining edges of @xmath180 .
the resulting tree @xmath265 is an induced subtree of @xmath247 in which a balanced edge - separator can be found .
since none of the edges of @xmath265 has both endpoints in any @xmath253 , none of the edges of @xmath265 are triangulation edges of a hole .
[ fig : structural - lemma],title="fig:",scaledwidth=30.0% ] + according to which a good separator is found .
three level boundaries are shown ( black cycles ) .
the external one is @xmath180 , the two cycles enclosed by @xmath180 are @xmath267 and @xmath268 .
the subfigures show the process of splitting the cotree @xmath247 , first at @xmath269 , then at @xmath270 . in the second phase the cotree is further split at the two remaining edges of @xmath180 .
the resulting tree @xmath265 is an induced subtree of @xmath247 in which a balanced edge - separator can be found .
since none of the edges of @xmath265 has both endpoints in any @xmath253 , none of the edges of @xmath265 are triangulation edges of a hole .
[ fig : structural - lemma],title="fig:",scaledwidth=30.0% ] according to which a good separator is found .
three level boundaries are shown ( black cycles ) .
the external one is @xmath180 , the two cycles enclosed by @xmath180 are @xmath267 and @xmath268 .
the subfigures show the process of splitting the cotree @xmath247 , first at @xmath269 , then at @xmath270 . in the second phase the cotree is further split at the two remaining edges of @xmath180 .
the resulting tree @xmath265 is an induced subtree of @xmath247 in which a balanced edge - separator can be found .
since none of the edges of @xmath265 has both endpoints in any @xmath253 , none of the edges of @xmath265 are triangulation edges of a hole .
[ fig : structural - lemma],title="fig:",scaledwidth=30.0% ] according to which a good separator is found .
three level boundaries are shown ( black cycles ) .
the external one is @xmath180 , the two cycles enclosed by @xmath180 are @xmath267 and @xmath268 .
the subfigures show the process of splitting the cotree @xmath247 , first at @xmath269 , then at @xmath270 . in the second phase the cotree is further split at the two remaining edges of @xmath180 .
the resulting tree @xmath265 is an induced subtree of @xmath247 in which a balanced edge - separator can be found . since none of the edges of @xmath265 has both endpoints in any @xmath253 ,
none of the edges of @xmath265 are triangulation edges of a hole .
[ fig : structural - lemma],title="fig:",scaledwidth=30.0% ] with this structural lemma we can now describe our oracle . consider a slice @xmath57 and let @xmath271 be the subgraph of @xmath17 enclosed by the boundary of @xmath57 .
the goal of processing slice @xmath57 is to store information ( distances ) so that the following distances ( in @xmath271 ) can be recovered from the information stored for all slices contained in @xmath271 . 1 . the distance between any two distinguished nodes in @xmath271 , 2 .
the distance between any distinguished node in @xmath271 and any vertex on the boundary of @xmath57 .
3 . the distance between any two vertices on the boundary of @xmath57 . encoding this information for all slices guarantees that distances between the distinguished vertices in the whole graph are captured .
[ [ the - encoding . ] ] the encoding .
+ + + + + + + + + + + + + to process a slice @xmath57 , we first encode * boundary - to - boundary * distances : the distances ( in @xmath271 ) between vertices on the boundary of @xmath57 using lemma [ lem : cycle ] . we then triangulate @xmath57 and define its spanning tree @xmath184 using lemma [ lem : triangulation ] .
next , we recursively separate @xmath57 using fundamental cycle separators .
the initial region @xmath193 is the entire slice @xmath57 .
its boundary is the external boundary @xmath138 of @xmath57 .
a region @xmath193 is separated into subregions obtained by cutting @xmath193 along some fundamental cycle separator @xmath96 w.r.t .
@xmath184 . since we only use fundamental cycle separators w.r.t .
the same tree @xmath184 , the separators never cross .
hence , the boundary of each new region @xmath212 consists of the contiguous portion of @xmath96 that belongs to @xmath193 , and possibly portions of the boundary of @xmath193 . since @xmath96 crosses @xmath138 at most twice ( at most once for each of the two paths in the fundamental cycle @xmath96 ) , the number of contiguous maximal fragments of @xmath138 in the boundary of @xmath212 is at most one plus the number of such fragments in the boundary of @xmath193 .
consequently , the number of contiguous maximal fragments of @xmath138 in the boundary of any region is bounded by the depth of the recursion , which we will show is @xmath272 .
we now explain how to choose the fundamental cycle separator @xmath96 with which we separate @xmath193 .
this is achieved using two interleaving recursive processes .
we refer to the first one as the outer recursion , and to the second one as the hole elimination recursion . in a step of the outer recursion
we apply lemma [ lem : recursion ] .
if we find a good balanced fundamental cycle separator @xmath96 , then we use it to separate the region @xmath193 .
every vertex in @xmath238 explicitly stores * @xmath12-to - separator * distances : its distance ( in @xmath271 ) to every vertex of @xmath96 .
in addition , for each subregion @xmath212 , for each contiguous maximal fragment @xmath273 of @xmath138 in @xmath212 , we encode * separator - to - boundary * distances : the distances ( in @xmath212 ) between @xmath273 and @xmath96 using lemma [ lem : unit ] or lemma [ lem : hole ] ( depending on whether the vertices of the separator @xmath96 and the vertices of @xmath273 lie on a single or two faces of @xmath212 ) .
then , we call the outer recursion recursively for each subregion @xmath212 .
the outer recursion terminates when there is at most one vertex with positive weight in the current region @xmath193 .
if the only remaining object is an artificial vertex @xmath189 , we apply lemma [ lem : hole ] to encode * hole - to - boundary * distances : the distances ( in @xmath193 ) between the boundary @xmath197 of @xmath188 and @xmath273 , for each contiguous maximal fragment @xmath273 of @xmath138 in @xmath193 . if the only remaining object is a distinguished vertex @xmath181 , we store * @xmath12-to - boundary * distances : the distances ( in @xmath271 ) from @xmath181 to every vertex of every @xmath273 . if the current region @xmath193 contains no vertices with positive weight , the outer recursion terminates .
if , on the other hand , we found a disposable hole @xmath188 , we store * hole - to - boundary * distances : distances between the boundary @xmath197 of @xmath188 and every contiguous maximal fragment @xmath273 of @xmath138 in @xmath193 .
the weight of the artificial vertex @xmath189 is set to zero .
this reflects the fact that for the rest of the processing of @xmath57 , distinguished vertices enclosed by the hole @xmath188 will not be treated individually and directly , but rather by encoding distances involving the vertices of @xmath197 . from this point on
, vertices of @xmath12 inside @xmath188 are no longer considered vertices of @xmath238 .
we then call the hole elimination process for the hole @xmath188 in region @xmath193 ( see figure [ fig : disposable ] ) . in a single step of the hole elimination recursion
, a region @xmath193 is separated using a fundamental cycle separator @xmath96 w.r.t . @xmath184
that is balanced w.r.t .
the number of vertices of @xmath197 in @xmath193 ( i.e. , a weight 1 is assigned to each vertex of @xmath197 and 0 to all other vertices ) . note that @xmath96 is necessarily a fundamental cycle w.r.t .
@xmath184 of some triangulation edge that is incident to @xmath189 .
the boundary of each of the two resulting regions contains a single contiguous portion of @xmath197 consisting of roughly half the vertices of @xmath197 in @xmath193 .
similarly to the single hole case , we store * @xmath12-to - separator * distances : distances ( in @xmath271 ) from every vertex of @xmath238 to every vertex of @xmath96 . for each subregion @xmath212 obtained by separating @xmath193 along @xmath96 , for each contiguous fragment @xmath273 of @xmath138 in @xmath212 , we encode * separator - to - boundary * distances : the distances ( in @xmath212 ) between @xmath273 and @xmath96 using lemma [ lem : unit ] or lemma [ lem : hole ] , and * separator - to - hole * distances : the distances ( in @xmath212 ) between @xmath96 and the single contiguous fragment of @xmath197 that belongs to @xmath212 using lemma [ lem : unit ] .
we then apply the hole elimination process recursively to each subregion @xmath212 .
it terminates when the current region @xmath193 contains at most two consecutive vertices of @xmath197 , or when it contains at most one distinguished vertices .
when this happens , we continue with the outer recursion on @xmath193 . with artificial vertex @xmath189 in a region @xmath274
is shown in top - left .
the spanning tree @xmath184 is indicated in blue .
@xmath274 is recursively separated using fundamental cycle separators that are balanced w.r.t . the number of nodes of the boundary @xmath197 of @xmath188 .
the elimination process finishes ( bottom - left ) when the current region @xmath193 contains only two consecutive vertices of @xmath197 , which are the endpoints of some edge @xmath95 .
this region @xmath193 differs from the the region @xmath212 obtained by separating @xmath274 using the fundamental cycle of @xmath95 w.r.t .
@xmath184 by a single vertex ( @xmath189).[fig : disposable],title="fig:",scaledwidth=25.0% ] with artificial vertex @xmath189 in a region @xmath274 is shown in top - left .
the spanning tree @xmath184 is indicated in blue .
@xmath274 is recursively separated using fundamental cycle separators that are balanced w.r.t . the number of nodes of the boundary @xmath197 of @xmath188 .
the elimination process finishes ( bottom - left ) when the current region @xmath193 contains only two consecutive vertices of @xmath197 , which are the endpoints of some edge @xmath95 .
this region @xmath193 differs from the the region @xmath212 obtained by separating @xmath274 using the fundamental cycle of @xmath95 w.r.t .
@xmath184 by a single vertex ( @xmath189).[fig : disposable],title="fig:",scaledwidth=25.0% ] with artificial vertex @xmath189 in a region @xmath274 is shown in top - left .
the spanning tree @xmath184 is indicated in blue .
@xmath274 is recursively separated using fundamental cycle separators that are balanced w.r.t . the number of nodes of the boundary @xmath197 of @xmath188 .
the elimination process finishes ( bottom - left ) when the current region @xmath193 contains only two consecutive vertices of @xmath197 , which are the endpoints of some edge @xmath95 .
this region @xmath193 differs from the the region @xmath212 obtained by separating @xmath274 using the fundamental cycle of @xmath95 w.r.t .
@xmath184 by a single vertex ( @xmath189).[fig : disposable],title="fig:",scaledwidth=25.0% ] with artificial vertex @xmath189 in a region @xmath274 is shown in top - left .
the spanning tree @xmath184 is indicated in blue .
@xmath274 is recursively separated using fundamental cycle separators that are balanced w.r.t . the number of nodes of the boundary @xmath197 of @xmath188 .
the elimination process finishes ( bottom - left ) when the current region @xmath193 contains only two consecutive vertices of @xmath197 , which are the endpoints of some edge @xmath95 .
this region @xmath193 differs from the the region @xmath212 obtained by separating @xmath274 using the fundamental cycle of @xmath95 w.r.t .
@xmath184 by a single vertex ( @xmath189).[fig : disposable],title="fig:",scaledwidth=25.0% ] with artificial vertex @xmath189 in a region @xmath274 is shown in top - left .
the spanning tree @xmath184 is indicated in blue .
@xmath274 is recursively separated using fundamental cycle separators that are balanced w.r.t . the number of nodes of the boundary @xmath197 of @xmath188 .
the elimination process finishes ( bottom - left ) when the current region @xmath193 contains only two consecutive vertices of @xmath197 , which are the endpoints of some edge @xmath95 .
this region @xmath193 differs from the the region @xmath212 obtained by separating @xmath274 using the fundamental cycle of @xmath95 w.r.t .
@xmath184 by a single vertex ( @xmath189).[fig : disposable],title="fig:",scaledwidth=25.0% ] we next prove that the total depth of the entire recursive procedure is @xmath275 . [ [ analyzing - the - recursion - depth . ] ] analyzing the recursion depth .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we begin with the initial region @xmath274 being the entire slice . in a single step of the outer recursion ,
if we find a good separator then we use it to separate the current region @xmath274 thus decreasing the weight of each resulting region @xmath193 by a constant factor .
if however we do not find a good separator , then we apply the hole elimination process on a disposable hole @xmath188 of the current region @xmath274 . since @xmath276 , and since every recursive call to the hole elimination process decreases the number of nodes of @xmath197 by half
, we get that after @xmath200 recursive calls the hole elimination process terminates , with each resulting region @xmath193 containing only two nodes of @xmath197 .
observe that these two nodes must be adjacent on @xmath197 ( see figure [ fig : disposable ] ) .
let @xmath95 be the edge between them and let @xmath240 be the fundamental cycle of @xmath95 w.r.t . @xmath184 .
since @xmath188 is disposable , the weight of the region @xmath212 obtained by separating @xmath274 using @xmath240 is at most half the weight of region @xmath274 . since @xmath277 and
since the weight of @xmath189 is zero this means that the weight of @xmath193 is at most half the weight of @xmath274 .
we conclude that every @xmath200 consecutive recursive calls the total weight of a region decreases by a constant factor .
this shows that the depth of the recursion is @xmath278 .
[ [ correctness . ] ] correctness .
+ + + + + + + + + + + + we next prove that the distance between any two distinguished vertices in @xmath12 can be recovered from our encoding .
[ lem : x_s - to - b ] the length of a shortest path @xmath142 in @xmath271 from any @xmath279 to any @xmath280 can be recovered from the encoding .
if @xmath142 contains some vertex of a fundamental cycle separator used in processing @xmath57 , let @xmath50 be the last vertex of @xmath142 that belongs to the earliest such separator .
by choice of the earliest separator , the @xmath222-to-@xmath50 distance ( in @xmath271 ) is stored ( @xmath12-to - separator distance ) . by choice of the last vertex on @xmath142 that belongs to that separator , the @xmath50-to-@xmath223 distance ( in the region of @xmath57 that contains @xmath281 $ ] )
is stored ( separator - to - boundary distance ) .
thus , the length of @xmath142 can be recovered .
if @xmath142 contains no such vertex , then @xmath222 and @xmath223 are in the same region when the recursion terminates , so the @xmath222-to-@xmath223 distance ( in @xmath271 ) is stored as a @xmath12-to - boundary distance .
we extend the previous lemma and show that it applies also to distinguished vertices enclosed by holes of @xmath57 ( i.e. , for @xmath239 instead of @xmath198 ) .
[ lem : x^s - to - b ] the length of a shortest path @xmath142 in @xmath271 from any @xmath282 to any @xmath280 can be recovered from the encoding .
the proof is by induction on the nesting depth of slice @xmath57 .
the base case follows from lemma [ lem : x_s - to - b ] . for the inductive step ,
if @xmath279 we are done by lemma [ lem : x_s - to - b ] , so assume @xmath222 is enclosed by some hole @xmath188 of @xmath57 .
if @xmath142 contains some vertex of a fundamental cycle separator used in processing @xmath57 before hole @xmath188 is eliminated , let @xmath50 be the last vertex of @xmath142 that belongs to the earliest such separator .
by choice of the earliest separator , the @xmath222-to-@xmath50 distance ( in @xmath271 ) is stored ( @xmath12-to - separator distance ) , and by choice of the last vertex of that separator on @xmath142 , the @xmath50-to-@xmath223 distance ( in a region of @xmath57 that contains @xmath281 $ ] ) is stored ( separator - to - boundary distance ) .
thus , the length of @xmath142 can be recovered .
if @xmath142 contains no such vertex , then the artificial vertex @xmath189 and @xmath223 are in the same region @xmath193 when either the recursion terminates , or the hole @xmath188 is eliminated . in either case , the @xmath197-to-@xmath223 distances are stored ( hole - to - boundary distance ) .
decompose @xmath142 into a maximal prefix @xmath229 $ ] enclosed by the slice @xmath192 whose boundary is @xmath197 , a maximal suffix @xmath234 $ ] enclosed by @xmath193 , and an infix @xmath233 $ ] .
the length of the prefix is stored by the inductive hypothesis for @xmath192 .
the length of the infix is represented by the boundary - to - boundary distances for @xmath192 .
the length of the suffix is stored ( hole - to - boundary distance ) .
finally , we extend the previous lemma and show that it applies to any two distinguished vertices . .
a slice @xmath57 with two holes is shown .
the boundary @xmath138 is double - lined . a @xmath222-to-@xmath223 shortest path @xmath142
is shown in solid black .
the vertices @xmath222 and @xmath223 belong to different holes of @xmath57 ( black cycles ) .
the path @xmath142 crosses a fundamental cycle separator ( blue , parts that do not belong to @xmath57 are dashed ) used in eliminating the hole to which @xmath223 belongs .
[ fig : correct],scaledwidth=50.0% ] [ lem : x - to - x ] the length of a shortest path @xmath142 in @xmath271 from any @xmath283 to any @xmath284 can be recovered from the encoding .
assume , wlog , that both @xmath222 and @xmath223 are enclosed in holes of @xmath57 ( the other cases are similar and less general ) . if @xmath142 contains some vertex of a fundamental cycle separator used in processing @xmath57
before either hole is eliminated , then let @xmath50 be a vertex of @xmath142 that belongs to the earliest such separator .
by choice of the earliest separator , both the @xmath222-to-@xmath50 and the @xmath223-to-@xmath50 distances ( in @xmath271 ) are stored ( @xmath12-to - separator distance ) .
thus , the length of @xmath142 can be recovered .
otherwise , the hole of @xmath222 and the hole of @xmath223 are in the same region @xmath193 when one of them , say the hole @xmath188 of @xmath222 , is eliminated .
if @xmath142 intersects one of the fundamental cycle separators used during the elimination process of hole @xmath188 , then let @xmath50 be the last vertex on the earliest such separator ( see figure [ fig : correct ] ) .
by choice of earliest separator , the @xmath222-to-@xmath50 distance is stored ( @xmath12-to - separator distance ) . by lemma [ lem : x^s - to - b ] ,
the length of the maximal suffix @xmath234 $ ] enclosed in @xmath188 is also stored .
let @xmath181 be the first vertex of @xmath233 $ ] that belongs to either @xmath138 or @xmath197 ( @xmath181 exists because @xmath223 is enclosed by @xmath197 and @xmath50 is not ) .
* if @xmath181 belongs to @xmath138 then the length of @xmath285 $ ] is stored ( separator - to - boundary distance ) . in this case , let @xmath286 be the last vertex of @xmath142 that belongs to @xmath138 .
the length of @xmath287 $ ] is represented by boundary - to - boundary distances for @xmath138 .
let @xmath58 be the first vertex of @xmath288 $ ] that belongs to @xmath197 .
the length of @xmath289 $ ] is represented as a hole - to - boundary distance ( when @xmath188 is eliminated ) .
let @xmath164 be the last vertex of @xmath290 $ ] that belongs to @xmath197 .
the length of @xmath291 $ ] is represented by boundary - to - boundary distances for @xmath197 , and the length of @xmath292 $ ] is represented by lemma [ lem : x^s - to - b ] .
see figure [ fig : correct ] for an illustration .
* if @xmath181 belongs to @xmath197 then the length of @xmath285 $ ] is represented as a separator - to - hole distance .
the representation of the suffix @xmath293 $ ] is then similar to the previous case . finally , we need to treat the case where @xmath142 does not intersect any fundamental cycle used in eliminating the hole @xmath188 . in this case
@xmath142 can be decomposed into a @xmath222-to-@xmath138 prefix , a @xmath197-to-@xmath223 suffix , and subpaths of @xmath142 between vertices of @xmath294 .
the prefix and suffix are represented by lemma [ lem : x^s - to - b ] .
the other subpaths are represented as hole - to - boundary or boundary - to - boundary distances as in the two cases above .
finally , we now show that the entire encoding requires only @xmath295 bits .
[ [ the - encoding - size . ] ] the encoding size .
+ + + + + + + + + + + + + + + + + + the space required for the boundary - to - boundary distances for all slices is @xmath296 since the total boundary size is @xmath170 , and by lemma [ lem : cycle ] .
we next bound the total space required for @xmath12-to - separator distances for all slices .
whenever a distinguished vertex stores its distances to a path @xmath142 explicitly , the total weight of its region decreases by a constant factor within @xmath272 recursive steps ( either immediately , if this happens in the outer recursion , or otherwise by the time the hole - elimination process ends ) .
so this can happen @xmath272 times per distinguished vertex . because @xmath297 ( by the height of @xmath184 ) , this sums up to a total of @xmath298 bits . the analysis of the remaining distances is done for each slice separately .
we have already argued that the depth of the recursive process to handle a slice @xmath57 is @xmath275 .
similarly to the analysis in section [ singleholesection ] of the single hole case , the total space required for storing separator - to - boundary distances using lemma [ lem : unit ] or lemma [ lem : hole ] at all calls at the same recursive level is @xmath299 . for exactly the same reasons ,
the total space required for storing separator - to - hole distances using lemma [ lem : unit ] at all calls at the same recursive level ( this only happens in the hole - elimination recursion ) is @xmath300 .
hole - to - boundary distances are stored using lemma [ lem : hole ] for at most one hole in each region along the recursion .
each invocation of lemma [ lem : hole ] for hole @xmath197 and boundary fragment @xmath273 requires @xmath301 bits . for a single level of the recursion ,
this sums up to @xmath218 because the total size of all boundaries is @xmath170 and there are @xmath302 vertices that contribute in more than one region ( endpoints of @xmath273 s ) .
the bound for @xmath12-to - boundary distances is @xmath218 for the same reasons .
to conclude , we showed that the total size of the entire encoding is bounded by @xmath303 , which is @xmath295 by choosing @xmath220 .
recall that gavoille et al .
@xcite show how to construct , given a boolean @xmath304 matrix @xmath305 , a planar grid @xmath306 containing @xmath307 vertices , such that @xmath305 can be recovered from the distances between @xmath0 distinguished vertices of @xmath306 .
this shows that , for @xmath308 , encoding all distances between @xmath0 vertices of a planar graph requires @xmath309 bits . for @xmath310 ,
we consider @xmath58 boolean @xmath311 matrices @xmath312 .
for each of these matrices , we construct a planar grid containing @xmath313 vertices .
the disjoint union of all these grids is a planar graph on @xmath314 vertices , such that all boolean matrices can be recovered from the distances between the @xmath0 distinguished vertices .
hence , encoding all such distances requires @xmath315 bits . setting @xmath316
we obtain that encoding all distances between the @xmath0 distinguished vertices of a planar graph on @xmath1 vertices requires @xmath42 bits .
the goal of section [ sec : upper ] was to guarantee that all distances between distinguished vertices are captured , but we were not concerned with the complexity of retrieving such a distance . in this section
we explain how to augment the encoding to allow efficient extraction of the stored distances .
we start with reformulating our encoding using the notion of _ dense distance graphs_. vertices of a dense distance graph are listed explicitly , but its edges are described implicitly with unit monge matrices .
each such matrix describes lengths of the edges between every @xmath317 and @xmath318 , for some subsets of nodes @xmath319 and @xmath320 .
the matrix is represented using @xmath321 bits as described in lemma [ lem : unit ] . in particular , we may have @xmath322 and then the matrix simply stores the length of a single edge explicitly .
the size of a dense distance graph is the total number of vertices plus the sum of @xmath323 over all matrices describing length of the edges . by construction ,
our encoding described in section [ sec : upper ] is based on defining a dense distance graph of size @xmath295 .
every distinguished node of the original graph is a vertex of the dense distance graph , and the distance between two distinguished nodes of the original ( unweighted ) graph is the same as the distance between their corresponding vertices in the ( weighted ) dene distance graph .
fakcharoenphol and rao designed an efficient algorithm for computing the shortest paths in such a graph , nicknames the fr - dijkstra : -division of a planar graph .
the vertices are the boundary nodes and distances between boundary nodes in the same region are represented with multiple monge matrices . however , it is easy to see that their algorithm work for any dense distance graph as defined above . ]
[ lem : fr ] distance between any two vertices of a dense distance graph of size @xmath57 can be found in @xmath122 time .
applying lemma [ lem : fr ] gives us an oracle of size @xmath295 answering queries in @xmath295 time .
for very large @xmath0 , say @xmath324 , the query time is clearly not optimal , as there exists an oracle of size @xmath325 answering queries in @xmath326 @xcite time . in the remaining part of this section
we will describe how to construct an oracle of size @xmath295 answering queries in @xmath327 time . to improve the query time , we apply the vanilla planar separator lemma . [
lem : separator ] for any planar graph @xmath17 on @xmath1 nodes , there exists a partition of the nodes of @xmath17 into sets @xmath328 , @xmath305 , and @xmath12 , such that @xmath329 , @xmath330 , and there are no edges between the nodes of @xmath328 and @xmath305 .
we recursively apply lemma [ lem : separator ] to construct a hierarchical decomposition of the whole graph .
the recursion is described by a binary tree @xmath172 , where every node @xmath331 corresponds to an induced subgraph @xmath332 of the original graph .
we let @xmath333 and @xmath334 denote the number of nodes and distinguished nodes in @xmath332 , respectively . we terminate the recursion as soon as @xmath335 .
if @xmath331 is a leaf , we define its set of distinguished nodes @xmath336 to consists of all the distinguished nodes of @xmath332 .
otherwise , @xmath336 consists of the following nodes : 1 . the separator of @xmath332 , 2 . for every child @xmath50 of @xmath181 that is a leaf , all the distinguished nodes of @xmath337 , 3
for every child @xmath50 of @xmath181 that is not a leaf , the separator of @xmath337 . then , we construct a dense distance graph of size @xmath338 capturing distances between any two nodes from @xmath336 in @xmath332 .
to calculate the distance @xmath339 between two distinguished nodes @xmath181 and @xmath50 in @xmath17 , we locate the deepest nodes @xmath340 and @xmath341 of @xmath172 , such that @xmath342 and @xmath343 .
then , we consider the union of all dense distance graphs constructed for the nodes of @xmath172 on the paths from @xmath340 and @xmath341 to the root .
note that the same node of @xmath17 might appear in more than one of these dense distance graphs , and we identify all of its copies . by construction
, the obtained dense distance graph captures the sought distance .
furthermore , its size is bounded by @xmath344 therefore , by lemma [ lem : fr ] we can answer a query in @xmath327 time . it remains to bound the size of the resulting oracle .
[ lem : nodesize ] the dense distance graph constructed for node @xmath331 is of size @xmath345 . to prove the lemma it is enough to bound @xmath346 by @xmath347 .
if @xmath181 is a leaf , this is clear .
otherwise , @xmath348 and @xmath336 consists of the following nodes : 1 .
the separator of @xmath332 of size @xmath349 .
2 . for every child
@xmath50 of @xmath181 that is a leaf , all @xmath350 distinguished nodes of @xmath337 , 3 . for every child @xmath50 of @xmath181 that is not a leaf , the separator of @xmath337 of size @xmath351 .
node @xmath181 has at most two children , so indeed @xmath352 .
to upper bound the size of the oracle , we need to upper bound the sum @xmath353 . to this end
, we separately consider all nodes @xmath331 such that @xmath354 , for every @xmath355 .
fix @xmath356 and call these nodes @xmath357 .
then , no @xmath358 is a descendant of another @xmath359 , so every node of the original graph appears in at most one @xmath360 .
therefore , @xmath361 and @xmath362 . from the latter inequality and the lower bound on @xmath333
we obtain that @xmath363 .
now we want to upper bound the following sum : @xmath364 from the concavity of @xmath365 , the above sum is maximized when @xmath366 , so we obtain : @xmath367 to obtain an upper bound on @xmath353 , we only need to multiply the above bound by @xmath368 because for every @xmath331 there exists @xmath356 such that @xmath333 belongs to the appropriate interval , so the total size of the oracle is @xmath369 .
a distance labeling scheme is a way to compress graphs that allows for distributed decoding .
the goal is to assign a label @xmath49 for each node @xmath50 , so that by looking at the labels of two nodes @xmath51 ( without access to the original graph ) we can infer the distance between them @xmath52 .
the main question one asks about such schemes is _ how small can the labels be _ ?
a famous open question is to close the gap between the @xmath370 upper bound @xcite and the @xmath11 @xcite lower bound for planar graphs . the only known technique capable of proving a tight lower bound is via a lower bound for the metric compression problem :
if you show that the metric can not be compressed into @xmath371 bits , then you show that no labels of size @xmath372 are possible .
our work deems this approach impassable , since such compressions are indeed possible .
optimistically , it is natural to ask if our upper bound for compression could lead to a better upper bound for labeling .
our encoding assigns @xmath373 bits per node , but can we distribute these bits to the nodes while allowing any pair of nodes to deduce the distance from their local information ? in this section , we discuss why this seems difficult .
the heart of our encoding is lemma [ lem : unit ] , which is repeatedly used to capture pairwise distances between a large subset of nodes of the graph using space proportional to the size of the subset .
a key part in the proof of the lemma is an efficient encoding of an @xmath88 matrix into @xmath43 bits , as long as it has the _ unit - monge _ property , that is : @xmath374 + m[i , j]- m[i , j+1 ] - m[i+1,j ] & \geq & 0 & \qquad\text { for any } i , j \in [ 1,n-1 ] , \\ and every
@xmath47 $ ] is an non - negative integer not exceeding @xmath1 .
the corresponding labeling problem would be to assign a label to every row and every column of @xmath44 , such that @xmath47 $ ] can be computed from the label of the @xmath155-th row and the @xmath228-th column .
we will show that the @xmath43 bits of the encoding _ can not _
be distributed into @xmath71 bits per label . in any such labeling scheme
, some labels must be of length @xmath375 . for completeness
, we will also provide a matching upper bound of @xmath326 .
we start with recalling the following connection between unit - monge matrices and permutation matrices .
@xmath142 is a permutation matrix if every row and every column contains at most one 1 and 0s elsewhere .
then , it is straightforward to verify that , for any permutation matrix @xmath142 the matrix @xmath44 defined as @xmath47=\sum_{i ' \geq i , j'\geq j } p[i',j']$ ] is a unit - monge matrix .
in fact , essentially any unit - monge matrix can be obtained through such transformation .
this is known , see e.g. section 2 in @xcite , but we provide a proof for completeness .
[ lem : unitperm ] for any unit - monge matrix @xmath44 , there exists a permutation matrix @xmath142 , such that @xmath376 = h[i]+v[j]+\sum_{i ' \geq i , j'\geq j } p[2i',2j'].\ ] ] where @xmath377 and @xmath320 are vectors of length @xmath1 with non - negative entries bounded by @xmath1 .
we define an @xmath378 matrix @xmath148 as follows : @xmath379 = m[i+1,j+1 ] + m[i , j ] - m[i , j+1 ] - m[i+1,j].\ ] ] by monge , clearly @xmath380 \geq 0 $ ] , and by unit @xmath380\leq 2 $ ] .
in fact , unit also implies that the sum in every row and every column of @xmath148 is at most 2 . to see this for rows ,
consider @xmath381+p'[i,2]+\ldots+p'[i , n]$ ] .
after telescoping , this is @xmath381-p[i+1,1]+p[i+1,n]-p[i , n]$ ] , so by unit at most 2 as claimed
. then , consider @xmath382 $ ] .
after substituting the definition of @xmath148 and telescoping , this becomes @xmath47+m[n , n]-m[i , n]-m[n , j]$ ] .
hence , if we define @xmath383=m[i , n]-m[n , n]/2 $ ] and @xmath384=m[n , j]-m[n , n]/2 $ ] it holds that @xmath47 = h[i]+v[j ] + \sum_{i ' \geq i , j ' \geq j } p'[i',j']$ ] .
finally , we create an @xmath385 matrix @xmath142 , where every @xmath386 block corresponds to a single @xmath380 $ ] , that is , the sum of values in the block is equal to @xmath380 $ ] .
it is always possible to define @xmath142 so that it is a permutation matrix . to see this ,
consider a row of @xmath148 .
the values there sum up to at most 2 , say @xmath380=p'[i , j']=1 $ ] for some @xmath387
. then , @xmath380 $ ] should correspond to a 1 in the first row of its block and @xmath388 $ ] to a 1 in the second row of its block .
if @xmath389 then in the corresponding block we create two 1s , one per row .
columns are chosen with a symmetric reasoning .
due to the above lemma , we can focus on assigning a label to every row and column of @xmath142 , such that given the label of the @xmath155-th row and the @xmath228-th column we can compute @xmath390 $ ] .
we call this problem labeling @xmath391 unit - monge matrices for dominance sum queries .
we can assume that there is exactly one 1 in every row and column of @xmath142 .
therefore , the input is fully described by a permutation @xmath393 .
any permutation on @xmath1 elements can be decomposed by up to @xmath394 increasing subsequences @xmath395 and up to @xmath394 decreasing subsequences @xmath396 .
the label of every row and every column consists of @xmath200 bits stored for every such subsequence , thus @xmath392 bits in total .
we think of every subsequence as a set of points @xmath397 and the @xmath200 bits corresponding to this subsequence in the label of the @xmath155-th row and the @xmath228-th column should be enough to determine the number of points @xmath398 such that @xmath399 and @xmath400 .
we separately describe what should be stored for an increasing subsequence and then for a decreasing subsequence .
consider an increasing subsequence consisting of points @xmath397 , such that @xmath401 and @xmath402 for every @xmath403 .
then , the label of the @xmath155-th row stores the smallest @xmath0 such that @xmath399 , and similarly the label of the @xmath228-th row stores the smallest @xmath0 such that @xmath400 . by taking the maximum of these two numbers we can determine the number of points @xmath398 such that @xmath399 and @xmath400 . now consider a decreasing subsequence consisting of points @xmath397 , such that @xmath401 and @xmath404 for every @xmath403 .
then , the label of the @xmath155-th row stores the smallest @xmath0 such that @xmath399 .
the label of the @xmath228-th row stores the largest @xmath0 such that @xmath400 . denoting the number stored for the @xmath155-th row and the @xmath228-th row by @xmath356 and @xmath164 , respectively , the number of points @xmath398 such that @xmath399 and @xmath400 can be calculated as @xmath405 .
we conceptually divide an @xmath378 matrix @xmath142 into blocks of size @xmath407 , thus creating an @xmath407 matrix @xmath305 , where every entry @xmath408 $ ] corresponds to a block of @xmath142 . for every block
@xmath408 $ ] we choose one bit @xmath409 $ ] .
we will show that then it is always possible to construct the matrix @xmath142 , such that all bits @xmath409 $ ] can be retrieved from the labels of rows of the form @xmath410 and columns of the form @xmath410 .
then it follows that we can encode @xmath1 bits of information in @xmath411 labels , hence one of these labels must consist of @xmath412 bits .
it remains to construct @xmath142 .
we construct @xmath142 incrementally .
we call a row or a column of @xmath142 active if there is no 1 there .
we start with an empty @xmath142 and keep adding 1s there while making ensure that there is at most one 1 in every row and column . given the labels of all rows @xmath410 and all columns of the form @xmath410 we can count 1s in every block of @xmath142 .
the goal is to ensure that this count is equal to @xmath409 $ ] .
assume that this is already the case for every @xmath409 $ ] such that @xmath413 or @xmath414 and @xmath387 and consider @xmath415 $ ] .
if @xmath415=0 $ ] we continue . otherwise , we have to choose exactly one active row @xmath164 in the range @xmath416 $ ] and exactly one active column @xmath417 in the range @xmath418 $ ] , and set @xmath419=1 $ ] , thus making both @xmath164 and @xmath417 inactive .
this clearly guarantees that there is exactly one 1 in the corresponding block of @xmath142 .
the only problem is to guarantee that there is at least one active row and column in the appropriate ranges .
however , we have deactivated less than @xmath420 rows in the range @xmath421 $ ] so far , and similarly less than @xmath422 columns in the range @xmath423 $ ] , so indeed there is at least one active row and column that we can use .
a. aggarwal and j. k. park
. notes on searching in multidimensional monotone arrays ( preliminary version ) . in _ 29th annual symposium on foundations of computer science , white plains , new york , usa , 24 - 26 october 1988 _ , pages 497512 , 1988 .
a. aggarwal and s. suri .
fast algorithms for computing the largest empty rectangle . in _ proceedings of the third annual symposium on computational geometry , waterloo , ontario , canada , june 8 - 10 , 1987 _ , pages 278290 , 1987 .
s. arikati , d. z. chen , l. p. chew , g. das , m. smid , and c. d. zaroliagis .
planar spanners and approximate shortest path queries among obstacles in the plane . in _
4th esa _ ,
pages 514528 .
springer , 1996 . | the planar graph metric compression problem is to compactly encode the distances among @xmath0 nodes in a planar graph of size @xmath1 .
two nave solutions are to store the graph using @xmath2 bits , or to explicitly store the distance matrix with @xmath3 bits . the only lower bounds are from the seminal work of gavoille , peleg , prennes , and raz [ soda01 ] , who rule out compressions into a polynomially smaller number of bits , for _ weighted _ planar graphs , but leave a large gap for unweighted planar graphs .
for example , when @xmath4 , the upper bound is @xmath2 and their constructions imply an @xmath5 lower bound .
this gap is directly related to other major open questions in labelling schemes , dynamic algorithms , and compact routing .
our main result is a new compression of the planar graph metric into @xmath6 bits , which is optimal up to log factors .
our data structure breaks an @xmath7 lower bound of krauthgamer , nguyen , and zondiner [ sicomp14 ] for compression using minors , and the lower bound of gavoille et al . for compression of weighted planar graphs .
this is an unexpected and decisive proof that weights can make planar graphs inherently more complex .
moreover , we design a new _ subset distance oracle _ for planar graphs with @xmath8 space , and @xmath9 query time .
our work carries strong messages to related fields .
in particular , the famous @xmath10 vs. @xmath11 gap for distance labelling schemes in planar graphs _ can not _ be resolved with the current lower bound techniques . |
relativistic jets are observed from actively - accreting black holes on all scales , from stellar - mass black holes in low - mass x - ray binary ( lmxb ) systems to supermassive black holes at the centres of galaxies .
there is a clear but poorly understood relationship between the morphology and spectrum of the jets and the structure of the accretion flow ( e.g. * ? ? ?
* ; * ? ? ?
lmxbs provide ideal laboratories to study jets and their connection to the accretion flow as they evolve through their full duty cycles on humanly - observable timescales .
spending the majority of their lifetimes in a low - luminosity quiescent state , lmxbs occasionally go through a much brighter outburst phase ( the result of increased accretion onto the central compact object ) where the jet properties change significantly as the system transitions through a number of canonical x - ray states ( see @xcite for a full review of x - ray spectral states ) . in a typical outburst , lmxbs
are initially observed in the hard state ( hs ) , which is associated with persistent radio emission from a steady , partially self - absorbed , compact jet that has a flat or inverted , optically - thick radio spectrum ( @xmath7 , where @xmath8 ; @xcite ) .
as the outburst progresses , the accretion rate increases and the system may transit through a number of intermediate states as it approaches the soft state . during the transition to the full soft state the jet emission transforms into discrete , bright , relativistically - moving knots with an optically - thin radio spectrum @xcite .
the compact jet is quenched by at least 2.5 orders of magnitude @xcite as the jet spectral break moves to lower frequencies ( @xcite ; @xcite ) and particle acceleration seemingly occurs at different locations within the jet @xcite . at the end of the outburst ,
the x - ray luminosity decreases and the source transitions back to the hard state . in this reverse state transition
the compact jet is gradually re - established ( first switching on in the radio band and then the infrared / optical band ; @xcite ) , indicating the potential build up of particle accelerating structures increasingly closer to the black hole .
the processes that are responsible for launching the jet outflow are poorly understood .
however , there exists a well - studied relationship between the radio and x - ray luminosities in hard state lmxbs , which has been observed for both individual sources ( e.g. gx 339@xmath04 ; ) and the entire sample of bh x - ray binaries ( e.g. @xcite ) .
although recent work has cast some doubt over the universality of the relation @xcite , it is generally described by two distinct power - law tracks in the radio / x - ray luminosity plane , an upper track and a lower , steeper , ` radio - quiet ' track @xcite .
the track followed by a given source does not appear to be connected to any property of the object ( such as spin ; @xcite ) and there is evidence for some individual sources transitioning between the two tracks as their x - ray luminosity fades @xcite .
it has also been suggested that this bimodal relationship may also exist in agn , supporting the theory that the mechanisms responsible for extracting energy from the accretion flow are similar across the mass scale @xcite .
high resolution , very long baseline interferometry ( vlbi ) observations can reveal important information about the changing properties of the jet during an outburst .
directly imaging the jet can determine the orientation of the jet axis , the bulk motion and structure of the jet , and how these change with the evolving accretion flow . while discrete ejecta have been observed in a number of systems in outburst ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the hard state compact jet has only been spatially resolved in two systems , grs 1915@xmath9105 @xcite and cygnus x@xmath01 @xcite , both of which are not standard lmxbs .
high - cadence observations of the discrete ejecta can determine their proper motions .
extrapolating backwards in time , the moment of launching may then be coupled to changes within the accretion flow to identify the mechanisms responsible for launching @xcite .
astrometric observations of a system over time may also allow for the proper motion of the system to be calculated , which can then be used to constrain the formation mechanism of black holes ( for full discussion , see * ? ? ? * and references therein ) .
maxij1836@xmath0194 was discovered at the onset of its outburst in august 2011 , triggering an intensive multiwavelength observing campaign .
classified as a black hole candidate due to its radio and x - ray properties @xcite , maxij1836@xmath0194 transitioned from the hard state ( spectral states are defined by x - ray spectral and timing properties ; * ? ? ?
* ) to the hard intermediate state ( hims ) on 2011 september 11 .
the outburst ` failed ' to enter the full soft state , remaining in the hims ( see @xcite for further discussion on failed outbursts ) , reaching its softest point on september 16 , following which it transitioned back to the hard state on september 28 and faded towards quiescence . in 2012
march , the system underwent a period of renewed activity @xcite before returning to quiescence by 2012 july 05 @xcite . this low - inclination system ( between 4@xmath10 and 15@xmath10 ; @xcite )
is between 4 and 10 kpc away @xcite and has proven itself to be an ideal system in which to study the evolution of the radio jet due to its bright radio emission ( @xcite ; @xcite ) . in this paper we present results from an intensive vla , atca and vlba radio monitoring campaign of maxij1836@xmath0194 , which began during the early phase of the outburst ( 2011 september 3 ) and continued until the source became sun constrained ( our last radio observation occurred on 2011 december 3 ) .
we also present atca and quasi - simultaneous _ swift _ monitoring of the source during its 2012 re - flare .
in particular , we will discuss the evolution of the polarised radio jet , the jet morphology , and the vlbi astrometry during the 2011 outburst , as well as the hard state radio / x - ray correlation as the system faded towards quiescence .
from 2011 september 3 through december 3 , we conducted an intensive monitoring campaign on maxij1836@xmath0194 with the karl g. jansky very large array ( vla ) , under project code vla/10b-218 .
we have previously presented the six epochs of vla observations with simultaneous millimetre , mid-/near - infrared , optical and x - ray data ( @xcite ; @xcite ) . here , we report the full 18 epochs of vla radio observations covering the rise , transition to and from the hims , and subsequent decay of the 2011 outburst of maxij1836@xmath0194 . the vla observations were taken every 25 days during the brighter phase of the outburst ( september 3
september 30 ) and every 1 2 weeks as the outburst decayed , until the final observation on 2011 december 3 .
the array was in its most - extended a configuration until september 12 , in the a@xmath11d move - configuration until september 18 , and thereafter in the most - compact d configuration .
the observations were taken over a wide range of frequencies ( between 1 and 43ghz ) with an integration time of 3seconds for all observations prior to november 11 , following which the data were taken with an integration time of 5 seconds .
the lowest - frequency 12ghz observations were recorded in two 512-mhz basebands , each containing eight 64-mhz sub - bands made up of sixty - four 1-mhz channels .
all other frequencies were recorded in two 1024-mhz basebands comprised of eight 128-mhz sub - bands made up of sixty - four 2-mhz channels . following standard data reduction procedures within the common astronomy software application ( casa ;
@xcite ) the data were corrected for shadowing , instrumental issues , and radio frequency interference .
3c286 was used for bandpass and amplitude calibration , j1407@xmath92827 was used as a polarisation leakage calibrator , and phase calibration was performed using the nearby ( @xmath12 away ) compact source j1820@xmath02528 .
calibrated data were then imaged on a per - baseband basis using the multi - frequency synthesis algorithm within casa and subjected to multiple rounds of self - calibration . to determine the flux density of the source , we fitted a point source to the target in the image plane ( stokes @xmath13 ) and standard vla systematic errors were added .
stokes @xmath14 and @xmath15 were also measured at the position of the peak source flux in the lower - frequency ( @xmath1610ghz ) data ( table [ tab : vla_flux ] ) . at no epoch
was the source observed to be significantly extended .
a nearby source is located to the south - west of maxij1836@xmath0194 ( @xmath130@xmath17away ; see figure [ fig : source ] , inset ) .
this unrelated , steep - spectrum source was not aligned with the jet axis and was not variable over the duration of the outburst . at times
when the vla was in its most compact d - configuration ( for all observations after 2011 september 18 ) , our low - frequency ( @xmath164ghz ) vla observations could not resolve the target from the confusing source .
therefore , where unresolved , our observations of maxij1836@xmath0194 are contaminated by the nearby source .
we must determine the spectrum of the unrelated object to estimate its contribution and then subtract from necessary observations . using an epoch when maxij1836@xmath0194 was faint ( 2011 november 01 ) , we measured the brightness of the confusing source ( figure [ fig : source ] ) . at frequencies where we were unable to resolve the two sources from each other , we subtracted the expected maxij1836@xmath0194 brightness ( determined by fitting a power law to the higher frequencies ) from the total spectrum to give an approximation of the unrelated source .
fitting the measured radio flux densities , @xmath18 , against frequency , @xmath19 , with a power law , we then estimate the spectrum of the nearby source and maxij1836@xmath0194 .
the spectrum of maxij1836@xmath0194 was well fit by a single power law at all epochs .
[ tab : vla_flux ] [ cols="^,^,^,^,^,^,^ " , ] the compact jet was not directly resolved in the individual images , except on september 19 at 2.3 ghz .
model fitting in the _
uv_-plane shows the source to be slightly extended at all frequencies , and in all but the last epoch ( table [ tab : xband_fits ] ) , as confirmed during the deconvolution process while imaging . as a check on the quality of the fits ,
the fitted gaussian was subtracted from the _ uv_-data , and the data were re - imaged .
the resulting residual images were noise - like , with the exception of the 2.3 ghz september 19 observation , for which some structure remained at the level of 1.0mjybeam@xmath20 , suggesting that a simple elliptical gaussian was not a good representation and that the source structure was more complex ( figure [ fig : vlba_images ] ) .
however , since it can account for @xmath2196% of the emission at that epoch , we take the fitted parameters as a good approximation to the source structure in this epoch , together with a faint residual jet component separated by @xmath115mas from the core , along the position angle @xmath22 .
excluding the september 04 2.27ghz observation , which was taken early in the outburst with relatively low resolution , model fitting in the
_ uv_-plane gave a consistent position angle , with a weighted mean of @xmath2 .
hereafter , we take the weighted mean as indicative of the observed jet axis . the fitted major axis of the source ( which was smaller at higher - frequency , corresponding to smaller scales within the jet ; see also @xcite ) did not change significantly during the outburst ( table [ tab : xband_fits ] ) . to constrain the proper motion of the system
, we measured the source position in each of the 8.4ghz images prior to performing any self - calibration ( table [ tab : vlba_positions ] ) . in all cases , we fitted the source in the image plane with an elliptical gaussian , for consistency with the known source structure .
we estimated systematics from the scatter on the two check source positions over the five epochs , weighted by their relative distance to the phase calibrator .
over the two months of our observing campaign , the fitted source position changed by [email protected] in right ascension ( r.a . ) and [email protected] in declination ( dec . ) . to increase the size of our data set , we included the 8.3-ghz position derived by @xcite from their chinese vlbi network ( cvn ) observations .
although the cvn observations used the same phase reference source , j1832 - 2039 , they used a less recent global astrometric solution to determine the assumed position of that calibrator , so we corrected for the shift in calibrator position before adding the cvn point to our astrometric sample .
we also attempt to correct for the motion of the optical depth @xmath24 surface along the jet axis , by calculating the expected core shift as a fraction of the component of proper motion along the vlbi jet axis .
@xcite present a model for the expected core shift in a compact jet .
assuming equipartition between the particles and the magnetic field and following their assumptions , we calculate the expected core shift at each epoch . we find a cumulative shift that is @xmath2515% of the total positional shift along the jet axis , which we include ( on a point by point basis ) in the uncertainties of our proper motion fit . using jackknife resampling with a linear function to fit the motion in both co - ordinates , we derived proper motions in r.a . and dec .
of @xmath26 with the reference position @xmath27 at the reference date of mjd 55800.0 ( table [ tab : xband_fits ] and figure [ fig : pm ] ) .
194 as a function of time ( positions given in table [ tab : xband_fits ] ) . here ,
the motion in r.a . and dec .
are shown in the upper and lower panels , respectively .
vlba measurements are shown in red , and the 8.3-ghz cvn measurement of @xcite is shown in blue .
we include the systematic uncertainty determined from the scatter in the check source positions , as well as the uncertainty due to the core shift of the jet in our error bars ( applied on a point by point basis ) .
these high - resolution observations show the proper motion of the system . ]
during the initial radio observations of maxij1836@xmath0194 , the system was in the hard x - ray spectral state . in this state ,
the radio spectrum was very inverted ( figure [ fig : lightcurves ] , bottom panel ) and linear polarisation was observed at levels of a few percent ( figure [ fig : polarisation ] ) , significantly lower than the maximum expected theoretical levels for a perfectly ordered magnetic field .
synchrotron theory predicts linear polarisation at levels of up to @xmath110% for optically - thick synchrotron emission and up to @xmath172% for optically - thin synchrotron emission in the presence of an ordered magnetic field .
however , these levels have rarely been observed in lmxbs .
for example , like maxij1836@xmath0194 , linear polarisation at levels of a few percent was observed in the hard states of v404 cyg @xcite and gx 339@xmath04 .
the low polarisation implies either that the magnetic field was intrinsically disordered on large scales ( e.g. * ? ? ?
* ) , possibly due to multiple polarised components ( which are polarised differently ; @xcite ) , a rotation of the polarisation angle within the jet @xcite or faraday rotation within the jet ( see @xcite for further discussion ) .
as the system transitioned to the hims , the radio emission brightened , the radio spectrum flattened , and the linear polarisation remained at a few percent ( although it did increase marginally ) . during this time , the polarisation spectrum steepened slightly ( figure [ fig : polarisation ] , second panel ) , which the evolution of _ fp _ shows is not just due to the change in the spectral index of stokes _ i_. while our observations do not allow for a clear interpretation of this result , we speculate that this steepening may have been due to the magnetic field becoming intrinsically more ordered further along the jet ( away from the compact object ) , or less ordered closer to the compact object .
the outburst then failed in the hims and the radio spectrum became more inverted as the source transitioned back to the hard state . during the initial hard state observations of maxij1836@xmath0194 , the radio spectrum was very inverted , with a spectral index of @xmath28 ( determined from the stokes _ i _ ) . while x - ray binaries in the hard state typically exhibit flat or slightly inverted radio spectra , where @xmath29 ( e.g. * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , inversions of @xmath30 were observed during the decay phase of the 2010 - 2011 outburst of gx 339@xmath04 @xcite .
the radio spectrum of maxij1836@xmath0194 then flattened ( to @xmath31 ) as the system transitioned into the hims .
our results show that during the transition phases of the outburst , the radio spectrum appeared to track the x - ray hardness ( figure [ fig : alpha_gamma ] ) .
however , during the hard state decay the spectrum of the radio jet continued to evolve while the x - ray hardness remained steady . during this time , the jet spectral break was observed to change by an order of magnitude indicating that the jet properties are still changing significantly ( @xcite ; @xcite ; @xcite ) .
a flattening of the radio spectrum as the source softens has been observed in some ( but not all ) other systems .
the shape of the spectrum is thought to be dependent on the geometry of the compact jet and the flow of material along the jet ( e.g. the bulk flow velocity along the jet and shocks and turbulence within the compact jet ; ) . at the beginning of its 2003 outburst , the radio spectrum of h 1743@xmath0322 showed a similar evolution to maxij1836@xmath0194 , where the inverted spectrum of the compact jet flattened as the system softened during its transition to the intermediate state @xcite .
maxij1836@xmath0194 reached its softest point in the hims , following which the jet spectrum became more inverted as the system transitioned back to the hard state .
gx 339 - 4 @xcite showed a similar evolution where the radio spectrum became more inverted following the transition back to the hard state ( after the compact jet was re - established ) .
however , a radio spectrum that steepens as the source softens is not observed in all lmxbs ( e.g. * ? ? ?
* ; * ? ? ?
* ) . from this small sample , the difference in the evolution of the jet spectrum between systems does not appear to be related to any physical parameter of the system such as inclination angle and black hole mass . while it appears that the evolution of the radio spectrum must be driven by some property of the accretion flow ( that may also be driving x - ray hardness ) , the surrounding environment , or the magnetic field ( that may affect the jet flow and shape ) , our observations
do not allow us to identify what processes may be driving the evolution . ) and _ rxte_/pca hardness ratio ( 6 - 16kev/2 - 6kev ) of maxij1836@xmath0194 during its 2011 outburst .
@xmath32 ( left hand axis ) is represented by the black points and x - ray hardness ( right axis ) is shown as the red squares .
the grey shaded region represents the hims . during the hims and transition
back to the hs @xmath32 appeared to track the x - ray hardness .
following the transition back to the hard state the x - ray hardness remained steady while the jet continued to evolve . ]
interestingly , while the flux density of the target changed , the fitted size of the compact jet at 8ghz remained relatively constant ( table [ tab : xband_fits ] ) . according to @xcite ,
the distance of the radio core along the jet axis , @xmath33 , is expected to scale with the observed flux density , @xmath18 , where @xmath34 .
the size of the compact jet , @xmath35 , is expected to scale linearly with the distance of the radio core from the origin @xcite .
therefore , the fitted size of the compact jet should relate to the flux density of the system , where @xmath36 . in our vlba observations
we therefore expect to see the jet change size by a factor of @xmath12.5 at 8ghz .
however , our results show that the jet size remained steady during the outburst ( even as the jet faded significantly ) , which could be due to physical properties of the jet changing as it faded ( such as changes to the opening angle , jet confinement region etc . ) .
the vla observations show that the _ rm _ remained approximately constant throughout the observations ( figure [ fig : polarisation ] ) , where @xmath37_rm_@xmath3832 radm@xmath39 ( to 1@xmath40 ) .
this result is consistent with values found by @xcite for radio sources along a similar line of sight ( where @xmath41 _ rm _ @xmath25 100 radm@xmath39 ) .
faraday rotation of the linearly polarised synchrotron emission is related to the thermal electron density and magnetic field along the line of sight .
therefore , the stable _ rm _ suggests that the rotation is not dominated by local material ( to the source ) .
the low value of the rm suggests that faraday rotation is most likely not responsible for the low levels of observed linear polarisation .
we find an agreement between the calculated evpa ( @xmath42 ; section [ sec : results_polarisation ] ) and the observed vlbi jet axis ( @xmath43 ; section [ sec : vlbi_axis ] ) .
the total jet spectrum is composed of contributions from both the optically - thick and optically - thin regions of each of the individual synchrotron components @xcite .
fractional polarisation is considerably higher in the optically - thin region of each synchrotron component than the optically - thick region ; hence , it is expected that the observed polarisation is dominated by contributions from the optically - thin region of each synchrotron component @xcite and therefore that the evpa will be aligned perpendicular to the magnetic field .
this implies that the magnetic field is perpendicular to the jet axis in maxij1836@xmath0194 .
a perpendicular magnetic field may arise from shock compression of the magnetic field @xcite , knots or hotspots in the jet @xcite , or a helical magnetic field structure dominating the observed emission @xcite , a consequence of the rotation of the accretion disk from which the jets may be launched @xcite .
a similar alignment between jet axis and evpa has also been observed in the black hole x - ray binaries gx 339@xmath04 , gro j1655 - 40 @xcite and cygnus x@xmath01 ( * ? ? ? * although this measurement is model - dependent ) .
polarisation studies of bl lac sources have shown a bimodal distribution of the evpa relative to the jet axis , where the majority of sources show an alignment between the evpa and the jet axis ( e.g. @xcite and references therein ) .
if an alignment between the evpa and jet axis is common to black hole x - ray binaries , high resolution images would not be required to determine the orientation of the jets . ) of maxij1836@xmath0194 as a function of source distance ( in kpc ) .
the black solid line is @xmath44 , the dotted lines are the error bars from uncertainties in proper motion and systemic velocity , and the unshaded regions denote the source distance limit of between 4 and 10 kpc @xcite .
the area above the horizontal dashed line represents the requirement of an asymmetric supernova kick .
our results suggest that an asymmetric supernova kick is required to explain the peculiar velocity of the system . ] from this single outburst we calculate the proper motion of the system by linearly fitting its vlbi positions in both co - ordinates over time ( see section [ sec : pm ] and figure [ fig : pm ] ) , which can then be used to constrain the formation mechanism of the black hole .
black holes in lmxb systems are thought to form in two ways @xcite : either a massive star collapsing directly into a black hole ( e.g. * ? ? ?
* ; * ? ? ?
* ) , or delayed formation in a supernova , as fallback on to the neutron star of material ejected during the explosion creates the black hole ( e.g. * ? ? ?
following a supernova , the centre of mass of the ejected material continues moving with the velocity of the progenitor immediately prior to the explosion .
the centre of mass of the binary system then recoils in the opposite direction and is constrained to lie in the orbital plane ( blaauw kick ; @xcite ) . in the presence of asymmetries in the supernova explosion , a further asymmetric kick , which need not be in the orbital plane , may be imparted to the binary . for further discussion on formation see @xcite and references therein . with our calculated proper motion , source distance ( 410 kpc ; @xcite ) , and systemic radial velocity ( 61@xmath2315 km s@xmath20 ; @xcite )
, we can determine the peculiar velocity of this system .
we calculate the heliocentric galactic space velocity components using the formalism of @xcite , assuming the standard solar motion measured by @xcite .
we assume circular rotation about the galactic centre in the plane of the disk , a flat rotation curve with a tangential velocity of 240kms@xmath20 , and a galactocentric distance for the sun of 8.34kpc @xcite . defining the peculiar velocity as the difference between the measured three - dimensional space velocity and the expected motion of the system due to galactic rotation (
see , e.g. * ? ? ?
* ) , we find that the peculiar velocity is @xmath45kms@xmath20 ( figure [ fig : vpec ] ) .
since stellar velocity dispersion in the disk gives a typical peculiar velocity of @xmath46kms@xmath20 even for the oldest m - type stars , and typical blaauw kicks in lmxbs ( while system parameter dependent ) have a maximum recoil velocity of @xmath47kms@xmath20 , an asymmetric natal kick was most likely required in maxij1836@xmath0194 .
this is just the sixth system in which constraints have been placed on the formation mechanism of the black hole ( see * ? ? ?
* and references therein ) .
while measuring the distribution of black hole kick velocities can constrain the formation mechanism of black holes , these observations must be carried out in a quiescent or hard state , which are faint ( high resolution observations of lmxbs are generally taken during full outburst ) .
therefore , with so few systems having accurate constraints , there is relatively little observational data on black hole kicks and observations such as these are required to make further progress .
194 . the white triangles and squares are representative data from gx 339@xmath04 @xcite and h1743@xmath0322 @xcite , respectively .
maxij1836@xmath0194 data from the 2011 outburst decay are shown by the filled green and red circles for the allowable distance limits of 4 and 10 kpc , respectively , while the open red and green circles show the monitoring from its minor re - brightening in 2012 .
the observed radio / x - ray correlation for this system is significantly steeper than either of the standard tracks shown by the representative data . ] .
top panel : the hard - state radio / x - ray correlation of maxij1836@xmath0194 ( red points ) , with the de - boosted , modelled intrinsic data ( blue points ) .
lower panel : @xmath48 required to shift @xmath49 to @xmath50 , as a function of x - ray luminosity assuming a distance of 10kpc and an inclination of 10@xmath10 @xcite . here , @xmath51 , where @xmath52 and @xmath53 100120 ( depending on the normalisation ) . here ,
the solid blue and dashed blue lines show results from the minimum and maximum normalisations used to account for the observed scatter about the lower track .
the results show that variable doppler boosting of the radio emission could account for the unusually steep radio / x - ray correlation in this system . ]
the decay phase of the outburst allows us to probe the hard - state radio / x - ray correlation of maxij1836@xmath0194 . in figure
[ fig : lrlx_representative ] we plot the observed hard state radio / x - ray correlation of maxij1836@xmath0194 with representative data from the upper branch ( gx 339@xmath04 ; @xcite ) and the lower branch ( h1743@xmath0322 ; @xcite ) . for comparison
, we calculate the luminosities of maxij1836@xmath0194 over the same energies as @xcite and @xcite . during the decay
, we find a steeper than usual radio / x - ray correlation , with an index of 1.8 @xmath23 0.2 .
this correlation index does not appear to be dependent on frequency ( e.g. taking any band of our radio data produces a similar index ) and therefore can not be explained by the changing radio spectral index . comparing this result with expected relations from @xcite
also shows that the steep correlation can not be explained by models where the x - ray emission originates either as inverse compton from a corona , or is dominated by optically thin synchrotron radiation from the jet .
we note that our 2011 x - ray observations only span @xmath11 order of magnitude in x - ray luminosity ( @xmath54 ) and deviations from the expected relationship have been found when the x - ray luminosity range is roughly one order of magnitude ( thought to be due to changes in the spectral index , electron cooling , additional x - ray emitting regions , etc ; @xcite ) . however , if we include atca and _ swift _ monitoring of the source during ( and following ) a re - brightening phase in 2012 we extend the x - ray luminosity range .
our 2012 data show that the source appears to lie on same steep track and at 5@xmath5510@xmath56ergs@xmath20 the source is still radio under - luminous compared to other x - ray binaries at comparable @xmath6 ( figure [ fig : lrlx_representative ] ) .
the relation we observe also holds over @xmath12.5 orders of magnitude in radio luminosity and does not appear to be random scatter about the standard correlation .
given the low inclination of this system ( between 4@xmath57 and 15@xmath57 ; @xcite ) it is plausible that variable doppler boosting affects the correlation .
we now investigate this possibility . assuming random inclination angles and highly variable boosting , @xcite tested a model to determine if variable boosting could account for the scatter of the radio / x - ray correlation for the available sample of black holes . with their model
, they were able to qualitatively reproduce both the scatter of the correlation , as well as the ` radio - quiet ' lower branch but noted that the model had its limitations . here
, we use a similar approach to test if the steep radio / x - ray correlation could be explained by variable relativistic beaming . while recent work has cast doubt over the universality of the two well - defined tracks for the radio / x - ray correlation @xcite
, we assume that the intrinsic radio luminosity of maxij1836@xmath0194 lies on the lower canonical hard - state tracks .
we then determine what boosting would be required to account for the observed radio luminosity , testing whether a variable @xmath48 can plausibly account for the steeper than usual correlation .
for the system to lie on the upper track , the radio emission at low x - ray luminosities would need to be significantly deboosted ( while being boosted at higher x - ray luminosities ; figure [ fig : lrlx_representative ] ) .
deboosted radio emission has been observed in lmxbs ( e.g. xte j1752@xmath0223 ; @xcite ) but is unlikely to have affected our observations due to the low inclination angle of the system .
we assume that the x - ray emission remains unbeamed ( but see @xcite ) and relate the observed beamed radio luminosity , @xmath58 , to the intrinsic radio luminosity , @xmath59 , by @xmath60 where @xmath61 is the doppler factor , _ n_=2 ( for a continuously replenished jet ; @xcite ) , and @xmath32 is the spectral index of the radio emission .
the doppler factor is a function of @xmath48 where @xmath62 , @xmath63 is the velocity of the jet emission as a fraction of the speed of light , and @xmath64 is the inclination of the system ; because @xmath65 , we are then able to calculate the lorentz factor required to shift the modelled intrinsic radio luminosity to that observed .
if we assume an inclination angle of 10@xmath57 @xcite , a source distance of 10 kpc @xcite and a jet axis perpendicular to the plane of the binary , then for @xmath59 to lie on the lower track of the radio / x - ray correlation @xmath48 must change by a factor of @xmath134 during the outburst ( from @xmath661 at the lowest x - ray luminosity to @xmath6734 at the highest ; figure [ fig : lrlx ] ) .
this result is sensitive to the normalisation of the assumed intrinsic radio luminosity , therefore we have tested a range of normalisations that account for the observed scatter about the lower track .
the maximum values of @xmath48 are higher than previously thought for both a compact jet ( where @xmath682 ; @xcite ) and the transient jet of grs 1915 + 105 ( where [email protected] ; @xcite ) .
however , due to the low inclination of maxij1836@xmath0194 , the values we estimate are not so high that beaming can be ruled out as an explanation for the steep radio / x - ray correlation .
fitting @xmath48 against x - ray luminosity ( as a fraction of the eddington luminosity for a 10m@xmath70 black hole ) gives @xmath51 , where @xmath52 and @xmath53 100120 . with this result , the bulk lorentz factor becomes larger than @xmath11 above 0.05 percent @xmath71 . for comparison ,
@xcite found @xmath48 becomes @xmath72 above 0.1 percent @xmath71 .
this level of boosting would be expected to have consequences on the observed radio spectrum , which coupled with the radio spectral index may then provide some insight into physical parameters of the jet ( such as the opening angle ) . however , a full modelling analysis would be required to determine the jet parameters which is beyond the scope of this paper . at a source distance of @xmath168 kpc ,
the radio emission at low x - ray luminosities falls well below the canonical relation ( figure [ fig : lrlx_representative ] ) and could not be explained unless it was deboosted . due to the low inclination angle of the system
, we do not expect deboosting . therefore ,
if we assume that the steep radio / x - ray correlation is due to a variable lorentz factor , maxij1836@xmath0194 must be at a distance greater than 8kpc .
our results show that relativistic beaming can account for the observed radio / x - ray correlation in maxij1836@xmath0194 .
however , similar to the results of @xcite , if we applied our boosting argument to the full sample of lmxbs we would see the correlation break down at x - ray luminosities @xmath73 percent eddington .
following an outburst , black hole lmxbs generally transition back to the hard state at luminosities between 0.3 and 3 percent eddington @xcite , with a mean value of @xmath12 percent , demonstrating that x - ray luminosities of @xmath73 percent eddington are expected during the hard state decay .
@xcite analysed the radio / x - ray relation of gx 339@xmath04 and v404 cyg and argued that similar lorentz factors are expected between sources .
we therefore conclude that while a variable lorentz factor can explain the unusually steep radio / x - ray correlation in maxij1836@xmath0194 , it can not be universally applied to the broad sample of black hole binaries .
it is possible that the steep correlation is a temporary deviation from the standard track @xcite .
however , when we include data from the 2012 re - flare , we see the source appears to lie on the same steep radio / x - ray track and the steep correlation and does not immediately rejoin the standard track .
this system could simply produce an intrinsically steep correlation ( @xmath12@xmath40 from the steeper track of @xcite , or 4@xmath40 from that of @xcite , but see recent results from @xcite ) , however , this index can not be explained by models of x - ray emission originating as inverse compton from a corona , or being dominated by optically - thin synchrotron emission from the compact jet .
we have presented the full vla , atca and vlba monitoring of maxij1836@xmath0194 during its 2011 outburst . with our intensive radio coverage of the system
we have observed the evolution of the polarised compact jet .
we find that the compact jet was linearly polarised at a few percent during the early , bright phase of the outburst . at later times
we were unable to place any constraint on the linear polarisation from the source .
we determined the vlbi jet axis and found an alignment with the _ evpa _ ( which was calculated from the measured stokes parameters ) .
this implies that the magnetic field was perpendicular to the compact jet axis , possibly due to a helical structure from the rotation of the accretion disk or compression due to internal shocks within the compact jet .
if this alignment is common for all lmxbs then high resolution vlbi observations may not be required to determine the orientation of the jet .
however , further studies are required to gauge whether the _ evpa _ is always aligned with the jet axis . calculating the peculiar velocity from our astrometric observations
, we find that the black hole likely required an asymmetric natal kick during formation to account for its high peculiar velocity . during the decay phase of the outburst we observed an unusually steep radio / x - ray correlation where @xmath74 , much steeper than other black hole lmxb systems .
we test whether the steep correlation may be due to a variable jet lorentz factor .
if we assume that the correlation agrees with the radio quiet radio / x - ray correlation track , the lorentz factor needs to vary from @xmath751 at low x - ray luminosities to @xmath7634 at high x - ray luminosities to produce the observed correlation .
however , this relation is unlikely to explain to the full sample of black hole lmxbs , and it is possible that we are observing a temporary deviation to the standard correlation .
interestingly , data from a re - flare event in 2012 show that the source appears to once again lie on the unusually steep track followed during the 2011 outburst .
if the steep radio / x - ray correlation is due only to a variable lorentz factor , our data require that the source lies at a distance greater than 8kpc .
we would like to thank the anonymous referee for their helpful comments and suggestions .
we also thank tom maccarone for useful discussions .
this research was supported under the australian research council s discovery projects funding scheme ( project number dp 120102393 ) .
da acknowledges support from the royal society .
sc acknowledges funding support from the french research national agency : chaos project anr-12-bs05 - 0009 ( http://www.chaos-project.fr ) and financial support from the univearths labex program of sorbonne paris cit ( anr-10-labx-0023 and anr-11-idex-0005 - 02 ) .
grs and ajt are supported by an nserc discovery grant .
tmb acknowledges support from inaf - prin 2012 - 6 .
sm acknowledges support by the spanish ministerio de economa y competitividad and european social funds through a ramn y cajal fellowship and the spanish ministerio de ciencia e innovacin ( sm ; grant aya2013 - 47447-c03 - 1-p ) .
this research has made use of nasa s astrophysics data system .
the international centre for radio astronomy research is a joint venture between curtin university and the university of western australia , funded by the state government of western australia and the joint venture partners .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
the australia telescope compact array is part of the australia telescope national facility which is funded by the commonwealth of australia for operation as a national facility managed by csiro . , k. , swank , j. h. , giles , a. b. , stark , m. j. , strohmayer , t. , zhang , w. , & morgan , e. h. 1996 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 2808 , euv , x - ray , and gamma - ray instrumentation for astronomy vii , ed . o. h. siegmund & m. a. gummin , 59 , j. p. , waters , b. , schiebel , d. , young , w. , & golap , k. 2007 , in astronomical society of the pacific conference series , vol .
376 , astronomical data analysis software and systems xvi , ed .
r. a. shaw , f. hill , & d. j. bell , 127 , r. j. , teuben , p. j. , & wright , m. c. h. 1995 , in astronomical society of the pacific conference series , vol .
77 , astronomical data analysis software and systems iv , ed .
r. a. shaw , h. e. payne , & j. j. e. hayes , 433 2012 sep 03 & 55807.1 & 4.60 & [email protected] & [email protected] & [email protected] & [email protected] + & & 7.90 & [email protected] & [email protected] & [email protected] & + 2011 sep 05 & 55809.1 & 5.00 & [email protected] & [email protected] & [email protected] & [email protected] + & & 7.45 & [email protected] & [email protected] & [email protected] & + 2011 sep 07 & 55811.2 & 1.50 & [email protected] & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & 68@xmath232 & & & + & & 25.90 & 79@xmath232 & & & + & & 32.02 & 88@xmath235 & & & + & & 41.00 & 87@xmath238 & & & + 2011 sep 09 & 55813.2 & 1.60 & [email protected] & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & 66@xmath232 & & & + & & 25.90 & 73@xmath232 & & & + & & 32.02 & 95@xmath235 & & & + & & 41.00 & 108@xmath236 & & & + 2011 sep 12 & 55816.1 & 1.60 & [email protected] & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & 47@xmath232 & & & + & & 25.90 & 48@xmath232 & & & + & & 32.02 & 51@xmath233 & & & + & & 41.00 & 55@xmath233 & & & + 2011 sep 13 & 55817.0 & 1.60 & [email protected] & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & 40.7@xmath231 & & & + & & 25.90 & 44@xmath232 & & & + & & 32.02 & 46@xmath233 & & & + & & 41.00 & 46@xmath233 & & & + 2011 sep 18 & 55822.0 & 1.50 & 37@xmath231 & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & 44@xmath231 & & & + & & 25.90 & 46@xmath232 & & & + & & 32.02 & 50@xmath232 & & & + & & 41.00 & 441@xmath233 & & & + 2011 sep 18 & 55822.9 & 1.56 & 33@xmath232 & & & [email protected] + & & 5.00 & [email protected] & [email protected] & [email protected] & + & & 7.45 & [email protected] & [email protected] & [email protected] & + & & 20.80 & [email protected] & & & + & & 25.90 & 38@xmath232 & & & + & & 32.02 & 44@xmath232 & & & + & & 41.00 & 37@xmath232 & & & + 2011 sep 23 & 55827.061 & 5.00 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & 51@xmath232 & & & + & & 25.90 & 54@xmath232 & & & + & & 32.02 & 62@xmath233 & & & + & & 41.00 & 66@xmath234 & & & + 2011 sep 26 & 55830.9 & 5.00 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + 2011 sep 26 & 55830.9 & 20.80 & 27@xmath231 & & & + & & 25.90 & 30@xmath231 & & & [email protected] + & & 31.52 & 36@xmath232 & & & + & & 37.52 & 40@xmath232 & & & + 2011 sep 30 & 55834.0 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & 26@xmath231 & & & + & & 31.52 & 30@xmath232 & & & + & & 37.52 & 32@xmath232 & & & + 2011 oct 07 & 55841.9 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + & & 31.52 & 22@xmath232 & & & + & & 37.52 & 25@xmath232 & & & + 2011 oct 12 & 55846.0 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + & & 31.52 & 14@xmath231 & & & + & & 37.52 & 14@xmath231 & & & + 2011 oct 22 & 55856.9 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + & & 31.52 & [email protected] & & & + & & 37.52 & [email protected] & & & + 2011 nov 01&55866.9 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + & & 31.52 & [email protected] & & & + & & 37.52 & [email protected] & & & + 2011 nov 11 & 55876.8 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + 2011 nov 18 & 55883.9 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + 2011 dec 03 & 55898.8 & 5.26 & [email protected] & & & [email protected] + & & 7.45 & [email protected] & & & + & & 20.80 & [email protected] & & & + & & 25.90 & [email protected] & & & + | maxij1836@xmath0194 is a galactic black hole candidate x - ray binary that was discovered in 2011 when it went into outburst . in this paper , we present the full radio monitoring of this system during its ` failed ' outburst , in which the source did not complete a full set of state changes , only transitioning as far as the hard intermediate state .
observations with the karl g. jansky very large array ( vla ) and australia telescope compact array ( atca ) show that the jet properties changed significantly during the outburst . the vla observations detected linearly polarised emission at a level of @xmath11% early in the outburst , increasing to @xmath13% as the outburst peaked .
high - resolution images with the very long baseline array ( vlba ) show a @xmath115mas jet along the position angle @xmath2 , in agreement with the electric vector position angle found from our polarisation results ( @xmath3 ) , implying that the magnetic field is perpendicular to the jet .
astrometric observations suggest that the system required an asymmetric natal kick to explain its observed space velocity .
comparing quasi - simultaneous x - ray monitoring with the 5ghz vla observations from the 2011 outburst shows an unusually steep hard - state radio / x - ray correlation of @xmath4 , where @xmath5 and @xmath6 denote the radio and x - ray luminosities , respectively . with atca and
_ swift _ monitoring of the source during a period of re - brightening in 2012 , we show that the system lay on the same steep correlation . due to the low inclination of this system
, we then investigate the possibility that the observed correlation may have been steepened by variable doppler boosting .
[ firstpage ] x - rays : binaries radio continuum : stars stars : individual ( maxi j1836 - 194 ) ism : jets and outflows |
for all terms and definitions , not defined specifically in this paper , we refer to @xcite , @xcite and @xcite .
unless mentioned otherwise , all graphs considered here are simple , finite and have no isolated vertices .
the sum set of two sets @xmath11 and @xmath12 , denoted @xmath13 , is the set defined by @xmath14 .
if either @xmath11 or @xmath12 is countably infinite , then their sum set will also be countably infinite .
hence , all sets we consider in this study are finite sets .
the cardinality of a set @xmath11 is denoted by @xmath15 .
the power set of a set @xmath11 is denoted by @xmath16 .
let @xmath0 denote the set of all non - negative integers .
an _ integer additive set - indexer _ ( iasi , in short ) of a graph @xmath2 is defined in @xcite as an injective function @xmath3 such that the induced function @xmath4 defined by @xmath5 is also injective .
a _ weak iasi _
@xmath9 is ( see @xcite ) an iasi such that @xmath17 for all @xmath18 .
a weak iasi @xmath9 is said to be _ weakly uniform iasi _ if @xmath19 , for all @xmath18 and for some positive integer @xmath20 .
a graph which admits a weak iasi may be called a _ weak iasi graph_. the following result is a necessary and sufficient condition for a given graph to admit a weak iasi .
@xcite a graph @xmath2 admits a weak integer additive set - indexer if and only if every edge of @xmath2 has at least one mono - indexed end vertex .
the following definitions are made in @xcite .
the cardinality of the labeling set of an element ( vertex or edge ) of a graph @xmath2 is called the _ set - indexing number _ of that element .
an element ( a vertex or an edge ) of graph which has the set - indexing number @xmath21 is called a _
mono - indexed element _ of that graph .
the _ sparing number _ of a graph @xmath2 is defined to be the minimum number of mono - indexed edges required for @xmath2 to admit a weak iasi and is denoted by @xmath22 .
certain studies about weak iasis have been done already and the following are some major results about weak iasi graphs relevant in this study .
[ t - wsg ] @xcite a subgraph of weak iasi graph is also a weak iasi graph .
[ t - wuc ] @xcite a graph @xmath2 admits a weak iasi if and only if @xmath2 is bipartite or it has at least one mono - indexed edge .
also , the sparing number of a bipartite graph @xmath2 is @xmath23 .
[ t - wuoc ] @xcite let @xmath24 be a cycle of length @xmath25 which admits a weak iasi , for a positive integer @xmath25 .
then , @xmath24 has an odd number of mono - indexed edges when it is an odd cycle and has even number of mono - indexed edges , when it is an even cycle .
an odd cycle @xmath24 has a weak iasi if and only if it has at least one mono - indexed edge .
[ t - wug ] @xcite the graph @xmath26 admits a weak iasi if and only if both @xmath27 and @xmath28 are weak iasi graphs .
more over , @xmath29 .
[ t - wkn ] @xcite a complete graph can have at most one vertex that is not mono - indexed . also , the sparing number of a complete graph @xmath30 is @xmath31 .
the admissibility of weak iasi by certain graph products and their sparing numbers have been studied in @xcite , @xcite and @xcite . in this paper ,
our intention is to study about the admissibility of weak iasi by a particular product , called edge corona , of two given graphs and estimate the corresponding sparing number .
let us first recall the definition of the edge corona of two graphs .
[ d-5.1]@xcite let @xmath27 be a graph with @xmath32 vertices and @xmath33 edges and @xmath28 be a graph with @xmath34 vertices and @xmath35 edges .
then , the _ edge corona _ of @xmath27 and @xmath28 , denoted by @xmath36 , is the graph obtained by taking @xmath33 copies of @xmath28 and then joining the end vertices of @xmath37-th edge of @xmath27 to every vertex in the @xmath37-th copy of @xmath28 .
figure [ fig : g - ecor ] is an example for the graph which is the edge corona of the cycles @xmath38 and @xmath39 . .
] the weak iasis of @xmath27 and @xmath28 may not induce a weak iasi for @xmath36 .
hence , we need to define an iasi independently for a graph product .
we say that a graph @xmath2 is said to be a _
@xmath21-uniform graph _ if the set - labels of all elements ( vertices and edges ) of @xmath2 are singleton sets . by the term an _ integral multiple of a set @xmath11 _ , we mean the set obtained by multiplying every element of @xmath11 by a same integer .
the following theorem establishes a necessary condition for the edge corona of two weak iasi graphs to admit a weak iasi .
[ t - snecg1 ] for two given graphs @xmath27 and @xmath28 , if @xmath36 admits a weak iasi , then either @xmath27 is @xmath21-uniform or @xmath40 copies of @xmath28 are @xmath21-uniform , where @xmath41 is the number of mono - indexed edges in @xmath27 .
let @xmath27 be a graph on @xmath32 vertices and and @xmath33 edges and @xmath28 be a graph on @xmath34 vertices and and @xmath35 edges .
let @xmath42 and @xmath43 be the vertex sets and @xmath44 and @xmath45 be the edge sets of @xmath27 and @xmath28 respectively .
let @xmath46 be the @xmath47-th copy of @xmath28 corresponding to the @xmath47-th edge @xmath48 of @xmath27 in @xmath36 and @xmath49 .
then , the subgraph of @xmath36 induced by the vertices @xmath50 is the complete graph @xmath51 , for any two adjacent vertices @xmath52 and @xmath53 in @xmath46 .
that is , all edges of @xmath54 are the edges of different complete graphs @xmath51 in @xmath36 , all of these complete graphs have the common edge @xmath48 .
first assume that @xmath36 admits a weak iasi .
then , we have to consider the following two cases .
assume that @xmath27 is not @xmath21-uniform .
then , @xmath27 will have some elements which are not mono - indexed .
without loss of generality , assume that the edge @xmath55 is not mono - indexed .
then either @xmath56 or @xmath57 must have a non - singleton set - label .
let @xmath56 be the vertex that is not mono - indexed .
then , by theorem [ t - wkn ] , no other vertex @xmath58 can have a non - singleton set - label .
therefore , the copy @xmath46 is @xmath21-uniform .
this argument is valid for the copies of @xmath28 corresponding to all edges of @xmath2 that are not mono - indexed .
therefore , at least @xmath40 copies of @xmath28 must be @xmath21-uniform , where @xmath41 is the number of mono - indexed edges in @xmath27 .
assume that no copy of @xmath28 is @xmath21-uniform .
then , each copy @xmath46 of @xmath28 has at least one edge that is not mono - indexed .
let the edge @xmath59 of @xmath46 has the non - singleton set - label . then
, by theorem [ t - wkn ] , the end vertices @xmath56 and @xmath57 of the the corresponding edge @xmath55 of @xmath27 can not have non - singleton set - label . hence , as no copy of @xmath28 are @xmath21-uniform
, no vertex of @xmath27 can have a non - singleton set - label .
that is , @xmath27 is @xmath21-uniform .
the converse of the theorem is also valid for with respect to the weak iasis defined on @xmath27 and @xmath28 .
let @xmath60 and @xmath61 the weak iasis defined on @xmath27 and @xmath28 , which need not be @xmath21-uniform .
the vertices of the copies of @xmath28 corresponding to the non - mono - indexed edges of @xmath27 need to be re - labeled using distinct singleton sets and the vertices of the copies of @xmath28 corresponding to the mono - indexed edges of @xmath27 can be labeled by distinct integral multiples of the set - labels of the corresponding vertices of @xmath28 .
clearly , this new labeling will be a weak iasi of @xmath36 .
hence , we have the following necessary and sufficient condition for the edge corona of two weak iasi graphs to admit a weak iasi .
[ t - snecg2 ] for given weak iasi graphs @xmath27 and @xmath28 , @xmath36 admits a weak iasi if and only if @xmath40 copies of @xmath28 are @xmath21-uniform , where @xmath41 is the number of mono - indexed edges in @xmath27 . in view of theorem [ t - snecg2 ] , we can estimate the number of mono - indexed edges in the edge corona of two given graphs . for given graphs @xmath27 and @xmath28 , the number of mono - indexed edges in @xmath36 is @xmath62 , where @xmath63 is the number of edges and @xmath64 is the the number of vertices of @xmath65 for @xmath66 and @xmath67 is the number of mono - indexed edges and @xmath68 is the the number of mono - indexed vertices of @xmath65 with respect to a weak iasi defined on @xmath65 for @xmath66 .
let @xmath27 be a graph on @xmath32 vertices and @xmath33 edges and @xmath28 be a graph on @xmath34 vertices and @xmath35 edges .
let @xmath60 and @xmath61 be the weak iasis defined on @xmath27 and @xmath28 respectively .
let @xmath68 and @xmath67 be the number of vertices and edges of @xmath65 that are mono - indexed under the weak iasi @xmath69 for @xmath66 .
let @xmath70 be a weak iasi graph .
assume that @xmath27 is not @xmath21-uniform .
then , @xmath27 has some elements having non - singleton set - labels .
then , by theorem [ t - snecg1 ] , @xmath40 copies of @xmath28 must be @xmath21-uniform .
let @xmath71 be the set of all @xmath21-uniform copies of @xmath28 in @xmath36 .
therefore , the members of @xmath71 contributes a total of @xmath72 mono - indexed edges to @xmath36 . in the remaining @xmath41 copies of @xmath28
, we can label the vertices by the distinct integral multiples of the set - labels of the corresponding vertices of @xmath28 with respect to @xmath61 .
let @xmath73 be the collection of these copies of @xmath28 .
then , each element in @xmath73 has @xmath74 mono - indexed edges . therefore
, the elements of @xmath73 contributes a total of @xmath75 mono - indexed edges .
it remains to determine the number of mono - indexed edges between @xmath27 and different copies of @xmath28 .
the mono - indexed vertex of every non - mono - indexed edge of @xmath27 is adjacent to all vertices of the corresponding copy of @xmath28 , which is also @xmath37-uniform .
the number of such mono - indexed edges is @xmath76 .
both end vertices of each mono - indexed edge of @xmath2 are adjacent to @xmath77 mono - indexed vertices of the corresponding copies of @xmath28 .
the number of such mono - indexed edges is @xmath78 .
therefore , the total number of mono - indexed edges in @xmath36 is @xmath79 . in view of theorem [ t - snecg1 ] ,
let us now proceed to discuss the sparing number of the edge corona of certain graphs .
we shall first consider the edge corona of two path graphs .
[ t - ec2p ] let @xmath80 and @xmath81 be two paths on @xmath82 and @xmath25 vertices respectively , for @xmath83 .
then , the sparing number of the edge corona of @xmath80 and @xmath81 is @xmath84 let @xmath85 .
assume that an internal vertex @xmath86 of @xmath80 has a non - singleton set - label .
then , the @xmath87 edges incident on @xmath86 become non - mono - indexed . but , two copies of @xmath81 whose vertices are adjacent to @xmath86 become @xmath21-uniform and @xmath88 edges of each of these copies of @xmath81 become mono - indexed .
more over , @xmath89 edges between @xmath80 and each of these two copies of @xmath90 become mono - indexed if @xmath25 odd and @xmath89 edges between @xmath80 and each of these two copies of @xmath90 become mono - indexed if @xmath25 even .
therefore , in both cases , we have more mono - indexed edges than when @xmath27 is @xmath21-uniform .
therefore , @xmath2 has minimum number of mono - indexed edges when @xmath27 is @xmath21-uniform .
if @xmath80 is @xmath21-uniform , each copy of @xmath81 can be labeled in an injective manner alternately by non - singleton sets and singleton sets .
therefore , no edges in these copies need to be mono - indexed .
then , each vertex of @xmath80 @xmath91 mono - indexed edges together with the mono - indexed vertices of the corresponding copy of @xmath81 . therefore , if @xmath25 is even , @xmath2 has @xmath92 mono - indexed edges and if @xmath25 is odd , @xmath2 has @xmath93 mono - indexed edges .
the sparing number of the edge corona of two graphs in which one is a path and the other is a cycle has been determined in the following theorems .
[ t - ecpc ] let @xmath80 be a path on @xmath82 vertices and and @xmath24 be a cycle on @xmath25 vertices , for @xmath94 .
then , the sparing number of the edge corona of @xmath80 and @xmath24 is @xmath95 let @xmath96 . as proved in theorem [ t - ec2p ] , @xmath2 has minimum number of mono - indexed edges when @xmath80 is @xmath21-uniform .
then , each copy of @xmath24 can be labeled in an injective manner alternately by singleton sets and non - singleton sets and hence no edges in these copies are mono - indexed . with respect to this labeling , each copy of @xmath24 contains @xmath97 mono - indexed vertices and makes @xmath97 mono - indexed edges with each vertex of @xmath80 .
therefore , if @xmath25 is even , no copy of @xmath24 need to have a mono - indexed edge and hence @xmath2 has @xmath92 mono - indexed edges .
if @xmath25 is odd , then each copy of @xmath24 must have a mono - indexed edge and hence @xmath2 has @xmath98 mono - indexed edges .
[ t - eccp ] let @xmath99 be a cycle on @xmath82 vertices and and @xmath81 be a path on @xmath25 vertices , for @xmath100 .
then , the sparing number of the edge corona of @xmath99 and @xmath81 is @xmath101 let @xmath96 . as we have already proved in theorem [ t - ec2p ]
, @xmath2 has minimum number of mono - indexed edges when @xmath99 is @xmath21-uniform .
then , each copy of @xmath81 can be labeled alternately by non - singleton sets and singleton sets and no edges in them are mono - indexed .
also , each copy of @xmath81 contains @xmath102 mono - indexed vertices and makes the same number of mono - indexed edges with each vertex of @xmath99 .
therefore , if @xmath25 is even , @xmath2 has @xmath103 mono - indexed edges and if @xmath25 is odd , @xmath2 has @xmath104 mono - indexed edges . in the following result , we study the sparing number of the edge corona of two cycle graphs .
[ t - ec2c ] let @xmath99 and @xmath24 be two cycles on @xmath82 and @xmath25 vertices respectively .
then , the sparing number of the edge corona of @xmath99 and @xmath24 is @xmath105 let @xmath96 . then , as we have stated in above theorems , @xmath2 has minimum number of mono - indexed edges when @xmath99 is @xmath21-uniform and we can label each copy of @xmath24 alternately by singleton sets and non - singleton sets . hence , no edges in these copies will be mono - indexed . with respect to this labeling , each copy of @xmath24 contains @xmath97 mono - indexed vertices and makes @xmath97 mono - indexed edges with each vertex of @xmath80 .
therefore , if @xmath25 is even , @xmath2 has @xmath103 mono - indexed edges . if @xmath25 is odd , then each copy of @xmath24 has one mono - indexed edge and hence @xmath2 has @xmath106 mono - indexed edges .
so far , we have discussed about the edge corona of certain regular graphs having same vertex degree @xmath107 .
can we generalise this result to all regular graphs having same vertex degree ?
the following result provide a solution to this problem .
[ t - ec2 g ] let @xmath27 and @xmath28 be two @xmath108-regular graphs on @xmath82 and @xmath25 vertices respectively , for @xmath83 .
then , the sparing number of the edge corona of @xmath27 and @xmath28 is @xmath109 $ ] , where @xmath110 is the minimum number of mono - indexed vertices required in @xmath28 .
let @xmath70 .
assume that an internal vertex @xmath86 of @xmath27 has a non - singleton set - label .
then , the @xmath111 edges incident on @xmath86 become non - mono - indexed .
but , @xmath108 copies of @xmath28 whose vertices are adjacent to @xmath86 become @xmath21-uniform and @xmath112 more edges of each of these copies of @xmath28 become mono - indexed , where @xmath113 is the mono - indexed edges in @xmath28 .
let @xmath32 vertices having non - singleton set - labels in @xmath28 must be relabeled by singleton sets in these @xmath108 copies of @xmath28 . therefore , @xmath114 edges between @xmath27 and each of these @xmath108 copies of @xmath28 become mono - indexed .
the total number of new mono - indexed edges in @xmath2 is @xmath115 $ ] .
therefore , in this case , we have more mono - indexed edges than when @xmath27 is @xmath21-uniform .
therefore , @xmath2 has minimum number of mono - indexed edges when @xmath27 is @xmath21-uniform . if @xmath27 is @xmath21-uniform , vertices of each copy of @xmath28 can be labeled in an injective manner alternately by distinct integral multiples of the set - labels of the corresponding vertices of @xmath28 .
then , the number of mono - indexed edges in each copy of @xmath28 is @xmath116 .
then , the total number of mono - indexed edges in @xmath36 is @xmath117 $ ] , where @xmath110 is the minimum number of mono - indexed vertices required in @xmath28 .
we can extend the above theorem for the edge corona of an @xmath108-regular graph @xmath27 and an @xmath118-regular graph @xmath28 , where @xmath119 as follows .
[ t - ec2g1 ] let @xmath27 be an @xmath108-regular graph on @xmath82 and @xmath25 vertices and @xmath28 be an @xmath118-regular graph on @xmath25 vertices , for @xmath83 and @xmath120 .
then , the sparing number of the edge corona of @xmath27 and @xmath28 is @xmath121 , where @xmath110 is the minimum number of mono - indexed vertices required in @xmath28 .
let @xmath70 .
assume that an internal vertex @xmath86 of @xmath27 has a non - singleton set - label .
then , as mentioned in the previous theorem , @xmath111 edges incident on @xmath86 become non - mono - indexed .
but , @xmath108 copies of @xmath28 corresponding to the edges incident on @xmath86 become @xmath21-uniform and hence @xmath122 more edges of each of these copies of @xmath28 become mono - indexed , where @xmath113 is the mono - indexed edges in @xmath28
. moreover , @xmath114 edges between @xmath27 and each of these @xmath108 copies of @xmath28 also become mono - indexed , where @xmath32 is the number of vertices having non - singleton set - labels in @xmath28 . hence , the number of new mono - indexed edges in @xmath2 is greater than the new non - mono - indexed edges in @xmath2 . therefore , in this case also , we have the minimum number of mono - indexed edges when @xmath27 is @xmath21-uniform . if @xmath27 is @xmath21-uniform , vertices of each copy of @xmath28 can be labeled in an injective manner alternately by distinct integral multiples of the set - labels of the corresponding vertices of @xmath28 .
then , the number of mono - indexed edges in each copy of @xmath28 is @xmath116 .
hence , the total number of mono - indexed edges in @xmath2 is @xmath123 $ ] , where @xmath110 is the minimum number of mono - indexed vertices required in @xmath28 .
another important problem in this area is about the edge corona of two graphs in which one graph is a complete graph .
first consider the edge corona of a path @xmath80 and a complete graph @xmath30 .
let @xmath80 be a path on @xmath82 vertices and @xmath30 be a complete graph on @xmath25 vertices .
then , the sparing number of @xmath124 is @xmath125 .
let @xmath126 .
then , @xmath2 can be considered as the one point union of @xmath127 complete graphs on @xmath128 vertices .
therefore , by theorem [ t - wkn ] , each @xmath129 has @xmath130 mono - indexed edges . since each @xmath129
are edge disjoint , by theorem [ t - wug ] , the total number of mono - indexed edges is @xmath125 .
next , let us consider the edge corona of a cycle @xmath99 and a complete graph @xmath30 .
let @xmath99 be a cycle on @xmath82 vertices and @xmath30 be a complete graph on @xmath25 vertices .
then , the sparing number of @xmath131 is @xmath132 .
let @xmath133 .
then , @xmath2 can be considered as the one point union of @xmath82 complete graphs @xmath129 and by theorem [ t - wkn ] , each of these @xmath129 has @xmath130 mono - indexed edges . since each @xmath129
are edge disjoint , by theorem [ t - wug ] , the total number of mono - indexed edges is @xmath132 .
the above two results can be generalised for the edge corona of an @xmath108-regular graph @xmath2 on @xmath82 vertices and a complete graph @xmath30 as follows .
let @xmath2 be an @xmath108-regular graph on @xmath82 vertices and @xmath30 be a complete graph on @xmath25 vertices , where @xmath134 .
then , the sparing number of @xmath135 is @xmath136 . since @xmath2 is an @xmath108-regular graph on @xmath82 vertices , then the number of edges in @xmath2is @xmath137 .
then , @xmath138 can be considered as a one point union of @xmath137 complete graphs on @xmath128 vertices .
then , the total number of mono - indexed edges in @xmath138 is @xmath139 .
in this paper , we have discussed about the sparing number of the edge corona of certain graphs . some problems in this area are still open . for an @xmath108-regular graph @xmath27 and an @xmath118-regular graph
@xmath28 , with @xmath140 , estimation of the sparing number of their edge corona is very complex . for some values @xmath108 and @xmath118
, we get minimum number of mono - indexed edges when @xmath27 is @xmath21-uniform and for some other values of @xmath108 and @xmath118 , we have minimum number of mono - indexed edges when @xmath27 is not @xmath21-uniform .
hence , determining the sparing number of @xmath36 in such a situation still remains as an open problem . in this paper
, we have not addressed the problem of determining the sparing number of the edge corona of two graphs in which the first graph is a complete graph .
the case when both the graphs are complete graphs are also not attempted . for two arbitrary graphs , determining the sparing number of their edge corona is more complicated .
the uncertainty in the adjacency and incidence pattern of arbitrary graphs makes this study complex .
hence , determining the sparing number of @xmath36 for arbitrary graphs @xmath27 and @xmath28 is also an open problem . | let @xmath0 be the set of all non - negative integers and @xmath1 be its the power set .
an integer additive set - indexer ( iasi ) of a graph @xmath2 is an injective function @xmath3 such that the induced function @xmath4 defined by @xmath5 is also injective , where @xmath6 is the sum set of @xmath7 and @xmath8 .
an integer additive set - indexer @xmath9 is said to be a weak integer additive set - indexer ( weak iasi ) if @xmath10 .
the minimum number of singleton set - labeled edges required for the graph @xmath2 to admit an iasi is called the sparing number of the graph . in this paper
, we discuss the admissibility of weak iasi by a particular type of graph product called the edge corona of two given graphs and determine the sparing number of the edge corona of certain graphs .
* key words : * integer additive set - indexers , mono - indexed elements of a graphs , weak integer additive set - indexers , sparing number of a graph , edge corona of a graph . *
ams subject classification : 05c78 * |
photonic materials possess structure ( crystalline order ) at the scale of the wavelength of light which they therefore diffract strongly , just like x - rays are diffracted from atomic crystals . by adjusting the geometric and dielectric parametres of a photonic crystal
, we can alter the _ optical _ properties at will@xcite , in the same way as a semiconductor crystal can be built to support desirable _ electronic _ properties .
for example , the dispersion relation ( band structure ) of light inside a photonic crystal is very different from the linear @xmath1 behaviour we have in homogeneous media .
another startling effect is the possibility for no allowed optical modes to exist at certain frequency ranges ( band gaps ) which , coupled with the ability to inject optical modes at specific frequencies by introducing defects in the crystalline structure , has fuelled prospects for novel optical phenomena such as photon localisation , inhibition of spontaneous emission and enhanced nonlinear effects .
novel devices ( single - mode leds , thresholdless lasers , etc . ) are also expected . where do we find photonic crystals ?
natural gemstone opals owe their beautiful iridescent colours to their structure which strongly diffracts visible light .
however their relatively large amount of disorder as well as the poor refractive index contrast they display renders them impractical for the detailed tailoring of optical properties one needs to attain if the above mentioned novel effects are to be sought after .
the first man - made photonic crystals were made ten years ago by yablonovitch and they operated at the scale of microwaves . intensified efforts for the past years have born fruit and we are now starting to have artificial opals at visible wavelengths , fabricated by a wealth of experimental techniques which bears witness to the ingenuity with which the problem has been tackled . some methods aim at building crystals from colloidal suspensions utilising the influence of gravitational , electromagnetic or biochemical interactions , while others concentrate on selectively removing material from a matrix ( etching ) and thus forming a desired structure . `
mixed ' approaches are also used ( removing material followed by ordering or stacking of crystal layers)@xcite .
the theoretical understanding of photonic structures has centred at solving maxwell s equations for the em wavefield in the presence of matter .
thus the classical theory of vector fields is applied .
this methodology , appropriate for scales large enough compared to the underlying atomic structure , has been successfully applied to predict both the dispersion relation ( band structure ) as well as the scattering properties of such systems .
nature favours phenomena that are energetically profitable .
for example , the van der waals force between two macroscopic objects can be found by calculating the change in the self - energy of the vacuum - fluctuation wavefields which multiply scatter off the objects . in electronic band structure ,
total energy calculations combined with variational arguments have been successfully used to predict the equilibrium structure of crystals and to realistically estimate a plethora of physical quantities@xcite .
but while the former example involves virtual light waves and the latter employs real electronic waves , there is an increasingly widening range of systems where real photons apply ; of those , photonic crystals form a subset whose potential for new physics and new applications is only beginning to be realised .
in such a crystal , the photonic band structure contains the energy levels accessible to photons .
but unlike atomic crystals , where the sea of electrons occupies a range of modes , it is possible with an external source of monochromatic light to selectively populate a particular mode in a photonic crystal .
the light source injects many photons but owing to the bosonic and non - interacting nature of light , each photon would remain in the same state .
thus photonic crystals are suitable for observing single - mode band - structure effects , a property implying rich force spectra .
to give an example , consider what happens as we traverse with a laser of tunable frequency the first band gap in a photonic crystal made of dielectric spheres in air .
on the low energy side , the em wavefield is concentrated outside the spheres ( _ i.e. _ at regions of low dielectric constant ) , hence the energy of this mode is lower . inside the gap ,
the fields are everywhere zero since all light is expelled from the crystal . and
on the high energy side , the field is concentrated inside the spheres .
this shifting of the wavefield from one region to another translates into changes in the spatial distribution of em energy inside the crystal , and would therefore strongly affect the forces . apart from performing total energy calculations , em forces can be also computed in an alternative , more formal manner : namely within the maxwell stress tensor ( mst ) formalism
. in this methodology
, the various components of the mst are constructed from the electric and magnetic fields in real space .
integration of the tensor around a closed surface surrounding the body of interest yields the total force and torque acting on it .
there is in principle no restriction as to the size and shape of the body , nor as to its dielectric properties .
the generality of this method makes its applicability wide , ranging from purely electric to purely magnetic to fully electromagnetic effects ; from virtual to real photons ; and from travelling to evanescent wavefields .
our approach in this paper is as follows : first we explain in some detail how the concept of a stress tensor arises in the calculation of forces .
we define the tensor for the case of electromagnetism and show how it can be incorporated in numerical calculations of photonic effects .
we then put our methodology at work . as a demonstration of some principal features of em forces
we treat first the case of a beam of laser light scattering from an homogeneous medium ( metallic or dielectric ) of variable thickness .
this is followed by the main body of our results which regards light - induced forces on dielectric crystals .
we are interested in calculating em forces on macroscopic bodies in the nanometre scales .
it is therefore natural to seek expressions for the forces in terms of macroscopic quantities ( such as dielectric functions ) .
we therefore make at the outset an implicit adoption of the continuum picture of matter .
we follow the treatment of landau and lifshitz@xcite .
a force acting on a macroscopic body is a sum of the forces @xmath2 on each of the volume elements that constitute the body .
thus the total force @xmath3 is @xmath4 if @xmath5 can be expressed as the divergence of a tensor of rank two , then we can transform the volume integral into a surface integral and the total force becomes @xmath6 where a sum over repeated indices is implied from here onwards .
the _ stress tensor _ @xmath7 contains all the information we need .
note it has dimensions of energy density .
this is not accidental , since forces can be also thought to arise due to spatial variations in the energy ( energy - gradient picture ) .
they act in order to lower the system s energy _
i.e. _ to push the system into regions where the energy is lowest .
therefore only the spatially varying part of @xmath7 contributes to forces , a fact which can be seen in eq .
( [ eq : tdv ] ) .
( [ eq : tds ] ) is of great practical benefit in that it is often easier to calculate the stress tensor on the surface of a solid body rather than in its interior .
so far the discussion is general . to concentrate on em forces
we need the stress tensor for the case of electromagnetism . when a solid body is immersed in a linear , isotropic medium of permittivity @xmath8 and permeability @xmath9 , then the spatially non - constant part of the stress tensor in the region external to the body
is given in terms of the electric ( @xmath10 ) and magnetic ( @xmath11 ) fields by @xcite @xmath12 provided the medium remains in mechanical and thermal equilibrium under the influence of the em fields ( _ i.e. _ electrostriction and magnetostriction effects are ignored ) .
thus by integrating @xmath7 of eq .
( [ eq : mstdef ] ) over a closed surface @xmath13 enclosing the body , the _ full _ em force on the body is obtained . no further approximations or limiting assumptions
need be made , as has often been done with other methodologies employed for em effects ( e.g. dipole approximation , vanishing particle size compared to the wavelength of light in scattering phenomena , etc . ) .
our methodology can be applied to dielectric or metallic objects of any shape and at any frequency range in principle .
the only requirement for meaningful results is that the integration is carried out over scales which are large with respect to atomic sizes , so that the continuum picture of matter is justified .
for nanostructure scales of interest to us ( @xmath14 nm ) this requirement is surely satisfied : the dynamics of the em wavefield is more properly described by the equations of maxwell rather than by the schrdinger equation for all the electronic charges which make up the dielectric response of the system .
finally , our choice of macroscopic scales dictates a relevant timescale also ( no macroscopic object can respond synchronously to thz oscillations ) : we must time - average the forces ( _ i.e. _ the stress tensor ) over many cycles when dealing with harmonic fields at optical frequencies .
we discretise maxwell s equations on a simple cubic ( sc ) mesh . for calculations of photonic band structure ,
periodic boundary conditions are used in all directions . for transmission results ,
we consider a crystal finite in the @xmath15 direction , having a thickness of @xmath16 unit cells ( layers ) , @xmath16 being allowed to vary .
we shall discuss here the case where a monochromatic light beam of frequency @xmath17 , polarisation @xmath18 and wavevector @xmath19 is incident on the crystal from the @xmath20 side . using the transfer matrix formalism we calculate reflection coefficients @xmath21 for scattering into plane waves of wavevector @xmath22 and polarisation @xmath23 , where @xmath24 is the component of @xmath19 parallel to the @xmath25 plane and @xmath26 is a _ two - dimensional _ reciprocal lattice vector also lying on the @xmath25 plane .
our computer codes have been published and issues of stability and accuracy have been discussed elsewhere @xcite . the new ingredient is the calculation of forces . from the reflection coefficients we construct the total fields on the @xmath20 side of the crystal , [ eq : field_reconstruction ] @xmath27 }
\ , , \label{eq : fielde}\ ] ] @xmath28 } \label{eq : fieldh } \,.\ ] ] in eq .
( [ eq : field_reconstruction ] ) @xmath29 are the amplitudes of the incident fields and @xmath30 are the unit vectors for each @xmath31 and each polarisation @xmath23 .
the kronecker delta functions serve as to add the contribution from the incident field , which is present on the @xmath20 side , to the reflected field . for the total fields on the @xmath32 side the only contribution comes from the transmitted fields .
having constructed the total fields outside the crystal , we invoke once again the matrix transfer equations to find the total field at each mesh point inside the structure . when this is accomplished , the stress tensor components are calculated and time - averaged .
the latter procedure can be made in a standard way since the fields are expressed as sums of plane waves at the same frequency @xmath17 .
finally , the forces are found by integrating the stress tensor over a surface enclosing the body of interest , which can be of arbitrary shape and complexity . in this work we present force calculations on isolated spheres , crystal layers and the entire crystal sample .
probably the simplest system for studying macroscopic radiation forces on materials is a homogeneous half - space ( dielectric or metallic ) upon which a single beam of light is incident .
the simplicity of this model system will allow us to put forward in a transparent manner some important features of our methodology .
furthermore , as this problem is analytically solvable it provides a nice test - ground for our numerical results .
we model the semi - infinite medium by considering a layer of material of sufficient thickness in the @xmath15 direction so that all light entering the layer is absorbed before reaching the far end . for a metal
this would only take a thickness equal to the skin depth ( a few nm typically ) when @xmath17 lies below the plasma frequency @xmath33 ; for a real dielectric , the tiny absorption present makes it necessary to consider much thicker samples , perhaps as much as several cm . in either case ,
only the incident and specularly reflected wavefields are present outside the system .
the force on the layer is found by integrating the stress tensor for the total field ( incident plus reflected ) on a surface @xmath34 which lies on the @xmath20 side immediately above the medium and extends to infinity in the @xmath25 plane ( the contribution from the integration on the far end is zero , since the fields vanish there ) .
a more appropriate quantity for this system is the em pressure @xmath35 on the layer . as a dielectric system we chose gap which is transparent ( _ i.e. _ has low absorption and the dielectric function @xmath36 is mostly real and positive ) over near infrared and visible frequencies but behaves as a poor metal ( _ i.e. _ there is a plasma frequency at @xmath37 13.5 ev but the absorptive part of @xmath36 becomes significant with respect to the real part ) in the ultraviolet spectrum . as a metallic system we use al , for which the dielectric function can be modelled as @xmath38 where @xmath39 ev and the damping coefficient for the intraband transitions is @xmath40 ev . our data for the dielectric properties of these materials come from @xcite .
we summarise our main findings for the forces on this system : + ( a ) a homogeneous system can only scatter light into directions belonging to the scattering plane ( the plane defined by @xmath19 and @xmath41 ) so em forces must lie on this plane . also the polarisation of light is conserved upon scattering .
+ ( b ) the response to light incident on a homogeneous half - space will range from full reflection ( e.g. for a perfect metal at frequencies below @xmath33 ) to full absorption ( for a dielectric at the brewster angle and @xmath42-polarised light ) . for normal incidence , the pressure on the half - space will range respectively from @xmath43 ( full reflection ) to @xmath44 ( full absorption ) , where @xmath45 is the radiation pressure of the incident beam which has intensity @xmath46 and travels at the free - space speed @xmath47 . if we replace the half - space with a layer of finite thickness such that some light is transmitted through the far end , taking care to include the contribution from the far end in the stress tensor integration , then the lower bound for the force becomes zero , corresponding to full transmission ( it i.e. zero reflection and zero absorption ) .
these effects are demonstrated for gap and al samples in figs .
[ fig.1 ] , [ fig.2 ] .
it is clear from these plots that the force on the system equals minus the net momentum change of the light , just as one would expect from the action - reaction principle .
thus the explanation of forces in terms of the momentum - balance picture is valid .
the three values ( 0,@xmath44 and @xmath43 ) constitute the natural limits for _ total _ ( _ i.e. _ external )
em forces on a macroscopic body ( homogeneous or not , finite or not ) of arbitrary size , due to incidence of a single beam of light .
the only way for these limits to be exceeded is when a resonance condition is fulfilled .
we shall come to this interesting case when dealing with forces on dielectric crystals .
( c ) since forces can just as well be attributed to spatial variations of energy as to temporal variations of momentum , we can use the energy - gradient picture to study em forces on bodies . in this picture , the em force on an infinitesimal volume element located at * r * is @xmath48 where @xmath49 is the energy of the em wavefield at @xmath50 .
the force on a macroscopic body is found by adding the contributions from all elements comprising the body .
for example , when light is incident at normal angle on a layer of material extending from @xmath51 to @xmath52 in the @xmath15 direction , the total ( _ i.e. _ average ) force on the layer is @xmath53 we show in fig .
[ fig.2 ] the total pressure on a layer of al , 400 nm in thickness .
the pressure is calculated both with the stress - tensor method ( solid grey line ) and with the energy - gradient method ( dotted grey line ) .
the agreement between the two is excellent , demonstrating the validity and usefulness of the energy - gradient picture .
after all , the physical basis for forces in nature is the lowering of the energy , and electromagnetism is no exception to this .
thus the sharpness of the spatial variations of the stress tensor ( or the energy ) is a measure of how strong and localised forces are .
smooth , continuous variations of @xmath7 imply forces that vanish when the volume element they act on shrinks to zero size .
surface forces , in contrast , are localised at the interface between media ; they are associated with abrupt , discontinuous changes in the stress tensor and can be explained physically as spatial discontinuities in the energy across the interface .
the next level of complexity involves homogeneous systems which are exposed to evanescent wavefields . in such fields
the light wavevector * k * has got an imaginary component so that the dispersion relation is always satisfied : @xmath54 where @xmath55 is the dielectric function of the medium of propagation .
our system is a dielectric layer ( denoted as medium ii ) of thickness @xmath52 , separating two dielectric half - spaces i , iii ( see fig .
[ fig.homog.layer ] ) , with the planar interfaces lying in the @xmath25 plane . for simplicity
we choose all media to have zero absorption .
a beam of light of wavelength @xmath56 is incident from medium i at an angle @xmath57 and a wavevector @xmath58 .
upon penetration into region ii , the wavevector may pick up an imaginary component along @xmath59 : @xmath60 by choosing @xmath61 and adjusting @xmath57 ( since @xmath62 ) appropriately , we can create an evanescent wave in region ii with the desirable imaginary momentum @xmath63 . here
we concentrate on the case of medium ii being vacuum ( @xmath64 ) .
we discuss _ total _ forces on media i and iii , not interface forces .
when only travelling waves exist in all three media , the two half - spaces repel each other . in this case
, the photons can be thought of as tennis balls : as they scatter off the walls of i and ii , they exert a positive pressure on each wall . however , when @xmath65 has a sufficiently large imaginary component along @xmath59 , the force changes sign and becomes attractive . in this case
the tennis - ball analogy breaks down .
our results indicate that at the electrostatic limit @xmath66 , the total force on iii per unit area varies as @xmath67 thus the force becomes zero when @xmath68 . curiously , this is the same condition for the angles of refraction @xmath69 to be equal , since @xmath70 when @xmath71 the force is repelling , and @xmath72 , just as in the case of a travelling wave scattering off a denser medium . on the other hand , when @xmath73 , we have @xmath74 , which is reminiscent of a travelling wave scattering off a dense / rare interface ; but at the same time the total force is attractive , a fact for which there is no analogue in waves of real @xmath19 .
the magnitude of the attractive force is sensitive to the ratio of the wavelength @xmath56 to the layer thickness @xmath52 over which the evanescent fields decay before entering medium iii .
we find that for @xmath75 the force has already dropped by more than three orders of magnitude compared to its value at the electrostatic limit @xmath66 . in fig .
[ fig.evan ] we plot the normal pressure @xmath76 versus @xmath77 for three cases , ( i ) @xmath78 , ( ii ) @xmath79 and ( iii ) @xmath80 . for ( i ) and ( ii ) the attractive part of @xmath76 displays a smooth maximum , whereas for ( iii ) it peaks sharply . in all cases
the highest value of @xmath76 at the electrostatic limit is of the same order of magnitude as the incident radiation pressure . for ( i ) and
( ii ) the field in region iii is that of a travelling wave for all angles and therefore @xmath81 .
however for ( iii ) the field in iii is evanescent when @xmath82 , @xmath83 being the angle for total internal reflection at an interface between media i and iii is present always at @xmath83 , provided @xmath80 .
its position in the @xmath84 axis is independent of the dielectric properties of layer ii . ] . at @xmath83
we have @xmath85 . from eq .
( [ eq : p_z ] ) it becomes clear why a vanishing @xmath86 results into a sharp peak in @xmath76 .
for moderate absorption in iii , @xmath87 does not quite reach zero at @xmath83 and the strength of the attraction is reduced although the peak is still clearly seen : for example , the maximum drops by 40 @xmath88 when @xmath89 . since the em force is proportional to the light intensity @xmath46
, we can adjust the latter so that the force induced by the evanescent ( or the travelling ) wavefield supersedes the dispersion forces between the half - spaces .
we have indicated the strength of the van der waals attraction in the figure , calculated in the macroscopic lifshitz theory with retardation and non - additivity effects considered .
owing to the @xmath90 behaviour of the van der waals force@xcite , only large enough separations will allow the photo - induced force to dominate .
we now turn our attention to a crystal of dielectric ( gap ) spheres arranged in a sc lattice .
gap has an essentially real , constant dielectric function ( modelled here as @xmath91 ) over the frequency range of relevance to this work ( 0.07 - 0.9 ev ) .
the actual absorption of gap is too small to have any sizeable effect on our force calculations , as we later show .
we consider spheres with a radius @xmath92 nm , and choose a lattice spacing @xmath93 nm , so the spherical surfaces are separated by @xmath94 nm . we often consider the reference medium in which the spheres are embedded to be air ( @xmath95 ) , although , for reasons of experimental feasibility , it is really a liquid medium we have in mind .
such a medium does not alter qualitatively the effects we describe provided its absorption is low enough .
for example , we show that water is suitable .
most of our results are for a discretisation mesh @xmath96 .
the crystal is infinite in the @xmath97 and @xmath98 directions , and has thickness @xmath16 unit cells in the @xmath15 direction . for the em properties of this crystal we expect the following behaviour : \(a ) propagation of em waves in an `` effective medium '' at large wavelengths @xmath99 where light senses the crystal as an homogeneous dielectric medium .
\(b ) multiple scattering effects : _
i ) _ appearance of forbidden ranges of frequencies as @xmath100 where the multiply scattered waves interfere destructively and annihilate each other inside the crystal ; _ ii ) _ scattering by specific crystal planes into non - specular directions as @xmath101 , @xmath102 being a reciprocal lattice vector .
\(c ) resonant features at frequencies corresponding to the em eigenmodes of each sphere .
we observe these effects in fig .
[ fig.pbs.t.force ] , where the band structure ( dispersion relation ) is shown alongside transmittance and force for one layer of thickness , normal incidence . at low energies @xmath17 is proportional to the wavevector @xmath103 , just as in free - wave propagation , the constant of proportionality being the wave velocity of a homogeneous effective medium replacing the crystal . for a system of unconnected spheres like ours ,
the dielectric function @xmath104 of the effective medium is well approximated within the garnett theory@xcite , @xmath105 @xmath106 being the volume fraction of the spheres in the crystal . for our system eq .
( [ eq : e_eff ] ) yields @xmath107 .
we find that at low energies ( or wavelengths @xmath108 ) the force on the crystal is well approximated by the analytical expression for the force on a homogeneous film of the same thickness and dielectric function @xmath104 . at higher energies
band gaps open up and the propagation of all em modes is forbidden inside an infinite crystal : all incident light will be reflected back when the frequency lies within a band gap .
it is impressive that even for one layer of thickness the major band gap centred at around 0.7 ev causes a low transmission .
consequently the pressure on the layer approaches the upper natural limit @xmath43 for frequencies inside this gap .
the lowest band gap requires a thicker sample before it can clearly show its signature in transmission and force spectra , but otherwise it affects the system s behaviour in the same way .
another consequence of multiple scattering is that under certain conditions , the scattered fields can interfere constructively and produce outgoing waves in additional directions to those of the specularly reflected and directly transmitted beams .
we call this scattering _ off - diagonal _ , since when it occurs the wavevector of the scattered field picks up a 2d reciprocal lattice vector component , @xmath109 most often the energetics of the problem are such that @xmath110 is imaginary and all the off - diagonal fields are evanescent .
however at certain ranges of frequencies a real @xmath110 is allowed and the off - diagonally scattered beams are travelling waves . for a sc lattice the minimum energy for this to occur is when @xmath24 lies at the edge of the first brillouin zone along the @xmath111 direction . considering an incident beam with @xmath112 ,
the condition becomes @xmath113 eq .
( [ eq : ods ] ) gives the lowest frequency for off - diagonal scattering for a crystal of lattice constant @xmath114 and a wave incident at angle @xmath84 .
as one imagines , the effects on forces are rich as the possible combinations of momentum - carrying outgoing waves are endless .
to give one example , consider the incidence at @xmath115 of a laser beam on our crystal of one layer thickness .
for this system we plot the intensity of reflection @xmath116 and transmission @xmath117 of each outgoing wave in fig .
[ fig.ods.r ] . off - diagonal scattering sets in for @xmath118 at frequency 0.71 ev in good agreement with eq .
( [ eq : ods ] ) . at slightly higher frequencies
the extra energy is given to @xmath110 while @xmath119 is fixed to the value @xmath120 , provided the angle of incidence is not changed .
thus the outgoing waves include two reflected and two transmitted beams , all forming an angle @xmath84 with the surface normal ( fig . [ fig.ods.beams ] ) .
in addition , notice that at 0.76 ev the intensities of the leftwards and the rightwards travelling waves are equal : @xmath121 .
this means that the total momentum of the scattered em fields along @xmath122 is zero . in other words ,
the crystal absorbs all incident momentum along @xmath122 , and the @xmath97-force on a surface @xmath123 equals @xmath124 our force calculations confirm that @xmath125 approaches this value at @xmath126 ev .
off - diagonal scattering survives a moderate amount of absorption in the crystal .
we find that for @xmath127 the reflectivities of the off - diagonal beams decrease by less than 40% ( depending on the frequency ) , which is a little larger than the decrease the diagonal beams suffer ( @xmath128 ) .
thirdly , we note that at specific frequencies there appear highly flat modes in the band structure combined with sharp features in the transmittance spectrum ( for example at 0.624,0.768,0.776,0.786,0.85,0.88 ev on the @xmath129 axis , fig .
[ fig.pbs.t.force ] ) . due to their low dispersion
they possess a low group velocity @xmath130 and for this reason they have been conveniently termed ` heavy photons ' by ohtaka and tanabe@xcite .
as has been argued elsewhere@xcite , these are resonance peaks corresponding to the em eigenmodes of isolated spheres ( _ mie resonances _ ) .
their presence strongly affects the position of the band gaps on the frequency axis@xcite .
the eigenmodes are spherical waves labelled by three quantum numbers , just as in the case of electronic orbitals in atoms , and are generally electric or magnetic oscillating multipoles . due to the spherical symmetry
they possess a @xmath131 degeneracy , @xmath132 being the azimuthal quantum number .
however the presence of a plane wave incident along the direction of @xmath19 breaks this symmetry and modifies the boundary conditions so that _ only one _ eigenmode for each @xmath132 can be excited ; this , at any rate , is the standard mie theory result and strictly applies to the case of an isolated sphere .
the notation for the excitable electric ( magnetic ) modes is @xmath133 ( @xmath134 ) , @xmath135 being the sequential index of the resonance in the frequency axis . for dielectric spheres the electric modes @xmath136
are broader whereas the magnetic modes @xmath134 are sharper , the sharpness of both types of modes increasing with frequency . for metallic spheres ,
only the electric modes survive , being better known to some as ` surface plasmons ' .
the first few mie resonances for one gap sphere subjected to a plane wave are shown in table [ table1 ] .
when many spheres approach each other to form the lattice , interactions occur which shift the resonances with respect to the frequencies for an isolated sphere ( fig .
[ fig.w.r ] ) .
unless the coupling between neighbouring spheres is strong , each resonance in the crystal will retain its characteristic em field distributions on the surface and the interior of each sphere .
thus the lowest - in - frequency heavy - photon resonance in our gap crystal is the @xmath137 magnetic dipole mode , followed by the electric dipole @xmath138 mode ( fig .
[ fig.modes ] ) .
the em fields for these two resonances clearly match the expected patterns for mie scattering from a single sphere . for the next two sharp peaks in transmittance the em fields are similar but not identical to those for the magnetic quadrupole mode @xmath139 ; and likewise for the following two peaks , whereby the fields display partial similarity with those for the electric quadrupole mode @xmath140 .
these are cases of mixing from more than one mie modes owing to the organisation of the spheres into a lattice . to quantify the contribution of each mie mode into the heavy - photon resonance one has to decompose the scattered field into its constitutive spherical waves , as ohtaka and tanabe have done@xcite .
however the mode that follows in frequency is a distinct electric quadrupole , with field distributions identical to the @xmath140 mie mode .
depending on @xmath17 , the incident field produces polarisation charges and/or induced currents in the spheres and orchestrates their response according to which mode is excited each time .
it is interesting that spheres which are initially non - magnetic ( @xmath141 ) can acquire magnetic properties at the occurrence of any @xmath142 resonance .
an effective homogeneous medium replacing the crystal structure would have @xmath143 at this instance . what happens at a resonance ?
each sphere acts as a cavity that traps the em waves .
the energy density internal to the spheres is 2 - 3 orders of magnitude larger than for the incident light beam@xcite . as van albada _
et al _ have shown@xcite , the transport of em energy can be slowed down considerably at the conditions for mie scattering .
they derive a rigorous microscopic formula for the velocity of energy transport @xmath144 , and subsequently propose a heuristic but physically plausible model for @xmath144 , which they show to agree with the exact model and experimental results within reasonable accuracy , and which reads @xmath145 where @xmath47 the speed of light in vacuum , @xmath106 the volume fraction and @xmath146 the energy density internal to the spheres relative to the surrounding medium ( vacuum in our case ) .
our computations suggest that @xmath147 for the first few heavy - photon modes , resulting in a velocity @xmath144 that is much lower than one expects from a straightforward estimate for the average velocity away from resonances , @xmath148 thus for @xmath149 the equations above give @xmath150 , whereas @xmath151 . for comparison ,
the group velocity @xmath152 varies with frequency and ranges from @xmath153 to @xmath154 for these resonances .
the accumulation of energy in the system at the resonance condition means that the overall em fields are enhanced , and this results in large forces when the stress - tensor integration is carried out .
the magnitude of the photo - induced force is proportional to @xmath46 , the intensity of the incident beam .
provided the spheres are in a liquid medium which facilitates heat evacuation ( so that melting is avoided ) , the photo - induced force can dominate gravitational , thermal - fluctuation and van der waals forces . assuming a moderate laser of @xmath155 w / m@xmath156 , the incident radiation pressure is @xmath157 pn per unit cell ( @xmath158 nm@xmath156 ) . for a crystal of one layer thickness , and when @xmath17 lies inside the second band gap ( centred at 0.7 ev in fig .
[ fig.pbs.t.force ] ) the force on a unit cell is @xmath159 pn ( the ` natural limit ' for full reflection ) , whereas at resonance conditions it exceeds this limit and can reach much higher values ( e.g. 5 pn for the @xmath137 resonance ) .
by contrast , the gravitational force on a gap sphere is @xmath160 pn .
thermal - fluctuation forces are comparable to gravity .
finally , the van der waals attraction between two gap spheres limit and ignoring retardation effects and is therefore an overestimate . ] is less than 0.2 pn@xcite .
the sharp force features shift slightly with the angle of incidence .
more importantly , they are sensitive to the amount of absorption present in the spheres or the surrounding medium : a moderate @xmath127 is enough to suppress them . for gap spheres in water , however
, we have found that the absorption is too small to blur the sharp features in forces , within the frequency range of interest ( fig .
[ fig.f_vs_abs ] ) .
the effect of increasing thickness is of particular importance with reference to resonances . when the number of layers is doubled , the number of sharp features doubles also@xcite .
this effect is directly equivalent to degeneracy splitting in atoms .
two oscillating multipoles ( gap spheres in a heavy photon mode ) brought close together interact , their fields hybridise and the two new energy levels of the combined system lie slightly below ( at @xmath161 ) and slightly above ( at @xmath162 ) the frequency of the original level ( at @xmath163 ) .
this atomic physics analogy is far reaching : the mode at @xmath161 is bonding ( attractive forces ) whereas at @xmath162 it is anti - bonding ( repulsive forces ) .
we see a series of alternating signs in the force on a sphere as different resonances are encountered ( fig .
[ fig.f_j=1 ] ) for a crystal two layers thick .
for the magnetic dipole ( @xmath137 ) mode in particular , we have shown @xcite that at @xmath161 all spheres have their dipoles parallel , whereas at @xmath162 the dipoles of the top layer are anti - parallel to those of the bottom layer .
the same is true for the electric dipole ( @xmath138 ) mode : all dipoles are parallel at @xmath164 ev , anti - parallel at @xmath165 for the system of fig .
[ fig.f_j=1 ] .
similarly , the multipoles of the higher - energy modes can be oriented constructively , in which case an attraction occurs , or destructively , leading to repulsion between neighbouring spheres .
this argument holds also for the mixed hybrid states that are made out of more than one mie modes , for example at 0.86 ev and 0.875 ev in fig .
[ fig.f_j=1 ] .
as expected , the splitting of the energy levels and the force strength increase as the spheres of two vertically adjacent layers approach each other ( fig .
[ fig.f_split.with.dist ] ) .
the nature of the sharp forces between adjacent layers therefore lies in the sphere - sphere interactions _ induced _ by the light beam , when the latter populates the various mie resonances . the analogy between resonant , light - induced , sphere - sphere interactions and the bonding of atoms that we just described is not the only one to be made .
an analogy exists also with the case of cutting a metal in two along a plane , whereby the bulk plasmons reorganise into surface plasmons of attractive and repelling nature at frequencies below and above the original frequency@xcite .
the similarities between these three very different systems prompts us to acknowledge a general tendency in nature for the formation of energetically favourable em bonds between two systems in the same state that come into proximity .
the rest of the force spectrum displays mostly smooth , attractive interactions due to the band gaps between the heavy - photon bands .
this off - resonance attraction is produced as the band - gap condition for high reflectivity causes the top layer to absorb almost all the shock from the reflection of the incident light and be therefore pushed towards the bottom layer which remains relatively inert .
next we explore the possibility that the induced bonding interactions may lead to the formation of stable crystalline structures with full 3d periodicity .
we start with two crystal layers on top of each other in a sc arrangement and gradually displace the top layer along the surface diagonal @xmath166 , such that the relative distances between spheres of the _ same _ layer do not change , as shown in fig .
[ fig.diag.disp ] .
we monitor the normal ( @xmath167 ) and shear ( @xmath168 ) forces for the bonding mode of the magnetic - dipole @xmath137 resonance .
we find that relative to a given sphere of the bottom layer ( @xmath169 ) , the shear force on the sphere above ( @xmath170 ) alternates in sign as the top layer is displaced along the diagonal . for small displacement @xmath171 , the sphere @xmath170 is attracted back to its initial position @xmath172 _ i.e. _ directly above @xmath169 of the bottom layer . as @xmath173 increases ( @xmath174 ) the top sphere
is attracted towards the void region in the centre of the square @xmath175 ( @xmath176 ) , with a shear - force strength 5 times as large as the attraction to @xmath177 .
finally for @xmath178 it is attracted towards @xmath179 , _
i.e. _ directly above the sphere @xmath180 of the bottom layer , with a force strength similar to that for the attraction towards @xmath177 .
the normal force between the two layers remains attractive ( since we are monitoring the bonding mode ) but its magnitude displays a global maximum at @xmath181 and two local maxima at @xmath182 .
consequently , the spheres of the top layer experience two potential wells at @xmath182 , whereby a stacking sequence @xmath183 is obtained .
however these wells are relatively shallow and when the potential barrier is crossed , the spheres find themselves in a much deeper well at @xmath181 where a sequence @xmath184 is energetically favourable ( fig . [ fig.pot.wells ] ) . as for the anti - bonding mode
, there are two high potential barriers at @xmath177 and @xmath185 and a barrier of lower height at @xmath181 .
the occurrence of resonant bonding effects between spheres as well as the energetic preference for the arrangement of crystal layers into the sequence @xmath184 , may add to the current methods for engineering 3d photonic crystals out of assemblies of colloidal spheres in liquid solutions .
experiments on gold nanospheres ( radius @xmath186 nm ) by kimura have shown fast formation of aggregates upon illumination at the surface - plasmon frequency ( _ i.e. _ the electric - dipole mode @xmath138 ) , in agreement with calculations of the photo - induced van der waals dipole - dipole interaction@xcite .
burns , fournier and golovchenko on the other hand have concentrated on dielectric microspheres ( @xmath187 nm)@xcite .
they have experimentally demonstrated the presence of long - range attractive forces , induced by laser light , between two polystyrene spheres in water .
they argue that these are magnetic - dipole interactions , owing their existence to induced currents on the spheres .
they calculate the interaction energy between two magnetic dipoles based on a simplified model that works in the long - distance limit @xmath188 , where @xmath189 the distance between the dipoles and @xmath56 the wavelength of light . from the perspective of mie resonances
this approach is justified because although the spheres burns _ et al _ use possess many different resonances at or near the frequency of their laser light , the large inter - sphere distance ensures that the dipole approximation works to leading order .
also the electric - dipole solutions are of secondary importance since they are damped compared to the magnetic dipoles .
in contrast to the work of kimura ( metallic spheres ) and that of burns _ et al _ ( @xmath188 ) , we are dealing with _ dielectric _
spheres at _ close proximity _ ( @xmath190 ) . in general , the sphere - sphere interactions in our system display a _ mixing _ of em - multipole behaviour , unless a specific condition for an isolated mie resonance is met .
for example two spheres of @xmath191 nm interact at @xmath37 0.65 ev as oscillating magnetic dipoles , at @xmath37 0.80 ev as oscillating electric dipoles and so on .
although the model of burns , fournier and golovchenko does not apply to our system ( due to the close proximity between the spheres ) , we believe that their physical explanation for the spatially oscillatory shear forces is basically correct : the oscillation is due to the changes in the phase shifts associated with retardation between spheres of different layers .
in conclusion , we have presented a general methodology for accurate and efficient computation of em fields and forces in matter .
we have applied this first to homogeneous systems ( dielectric and metallic ) and studied the effects of travelling and evanescent waves on the forces exerted on a slab of material . under specific conditions
, evanescent fields can produce attractive total forces which are even stronger than the van der waals attraction between two half - spaces .
next , we have concentrated on a dielectric crystal of gap spheres arranged in a sc lattice .
we observe the connection between lattice effects ( band gaps , bragg scattering ) and em forces .
more interestingly , however , we find large resonant forces at frequencies corresponding to the em eigenmodes of isolated spheres .
an analogy to atomic physics proves fruitful as the em fields on the spheres arrange into bonding and anti - bonding states , causing highly attractive and repelling interactions which may become dominant over gravitational , thermal and van der waals forces .
we believe these effects to be experimentally observable and relevant to the formation of stable crystals of colloidal nanospheres . many thanks to f .- j .
garcia - vidal for valuable discussions and comments .
i would like to acknowledge the trustees for the beit fellowship for scientific research , as well as my parents , yannis and chryssoyla , for financial assistance .
antonoyiannakis .
beit scientific research fellow . + e - mail : [email protected] + + edited by c. m. soukoulis ( plenum press , 1993 ) . edited by c. m. soukoulis * e 315 * ( kluwer academic publishers , 1996 ) .
joannopoulos , r.d .
meade , j.n .
winn , _ photonic crystals _ ( princeton university press , 1995 ) . , no.2 ( 1993 ) .
( 1993 ) .
g.feiertag , w. ehrfeld , h. freimuth , h. kolle , h. lehr , m. schmidt , m.m .
sigalas , c.m .
soukoulis , g. kiriakidis , t.pedersen , j. kuhl , w.koenig , appl .
. lett . * 71 * , no.11 , 1441 ( 1997 ) ; r.d .
pradhan , j.a .
bloodgood and g.h .
watson , phys .
b * 55 * , no.15 , 9503 ( 1997 ) ; k.a . mcintosh , l.j .
mahoney , k.m .
molvar , o.b .
mcmahon , s. verghese , m. rothschild and e.r .
brown , appl .
. lett . * 70 * , no .
22 , 2937 ( 1997 ) ; s. noda , n. yamamoto and a. sasaki , jpn . j. appl . phys .
* 35 * , l909 ( 1996 ) ; g. decher , science * 277 * , 1232 ( 1997 ) and references therein ; s - r chen , j.a .
kornfield , s.d .
smith , j.t .
grothaus and m.m .
satkowski , science * 277 * , 1248 ( 1997 ) and references therein ; w. shenton , d. pum , u.b .
sleytr and s. mann , nature * 389 * , 585 ( 1997 ) and references therein ; also see references in @xcite .
payne , m.p .
teter , d.c .
allan , t.a .
arias and j.d .
joannopoulos , rev .
phys . * 64 * , 1045 ( 1992 ) .
l. d. landau and e. m. lifshitz , _ theory of elasticity _
( pergamon press , 1975 ) @xmath192 2 . l. d. landau and e. m. lifshitz , _ electrodynamics of continuous media _
( pergamon press , 1960 ) @xmath192 15,16 .
pendry , in ref.@xcite pg .
209 ; p.m. bell , j.b .
pendry , l. martn - moreno and a. ward , comp .
. comm . * 85 * , 306 ( 1995 ) ; j.b .
pendry and p.m. bell , in ref.@xcite pg .
e. d. palik ( ed . ) , _ handbook of optical constants of solids _ ( academic press , 1985 ) part i p.369 and part ii p.445 . ,
progr . surf .
* 7 * , 1 , ( 1973 ) .
london a * 203 * , 385 ( 1904 ) ; philos .
london a * 205 * , 237 ( 1906 ) .
, j. phys .
japan * 65 * , no . 7 , 2265 and 2276 ( 1996 ) .
, europhys . lett .
* 40 * , no.6 , 613 ( 1997 ) .
b * 40 * , no.2 , 1334 ( 1989 ) ; s. datta , c.t .
chan , k.m . ho , c.m .
soukoulis and e.n .
economou , in ref.@xcite pg .
, j. phys .
japan * 65 * , no . 8 , 2670 ( 1996 ) . , j. opt .
am . a * 4 * , no.8 , 1361 ( 1987 ) .
* 270 * , 143 ( 1996 ) .
lett . * 60 * , no .
11 , 1085 ( 1988 ) . ,
* 1 * , 127 ( 1973 ) . ,
rev . and lett .
* 3 * , no.1 , 1219 ( 1996 ) . , phys .
* 63 * , no.12 , 1233 ( 1989 ) ; science * 249 * , 749 ( 1990 ) .
.mie modes of an isolated gap sphere ( @xmath91 ) and their corresponding _
order parametre _ ( mode frequency times sphere radius ) in units of ev @xmath193 nm .
@xmath142 @xmath194 modes represent magnetic ( electric ) multipoles oscillating at the light frequency .
[ table1 ] [ cols="^,^,^,^,^,^,^,^,^,^,^,^ " , ] | we have developed a general methodology for numerical computations of electromagnetic ( em ) fields and forces in matter , based on solving the macroscopic maxwell s equations in real space and adopting the maxwell stress tensor formalism .
our approach can be applied to both dielectric and metallic systems of frequency - dependent dielectric function ; as well as to objects of any size and geometrical properties in principle .
we are particularly interested in calculating forces on nanostructures .
our findings confirm that a body reacts to the em field by minimising its energy , _
i.e. _ it is attracted ( repelled ) by regions of lower ( higher ) em energy . when travelling waves ( of real wavevector ) are involved , forces can be additionally understood in terms of momentum exchange between the body and its environment .
however when evanescent waves ( of complex wavevector ) dominate , the forces are complicated , often become attractive and can not be explained by means of real momentum being exchanged .
we have studied the em forces induced by a laser beam on a crystal of dielectric spheres of gap .
we observe effects due to the lattice structure , as well as due to the single scattering from each sphere . in the former case
the two main features are a maximum momentum exchange ( and largest forces ) when the frequency lies within a band gap ; and a multitude of force orientations when the bragg conditions for multiple outgoing waves are met . in the latter case the radiation couples to the em eigenmodes of isolated spheres ( mie resonances ) and very sharp attractive and repulsive forces occur . depending on the intensity of the incident radiation these forces
can overcome all other interactions present ( gravitational , thermal and van der waals ) and may provide the main mechanism for formation of stable structures in colloidal systems .
# 1@xmath0#1 |
in the first paper of this series , @xcite ( paper i henceforth ) , we have described extensions of the antares code ( @xcite ) which enable the investigation of the pulsation - convection interaction by numerically solving the equations of radiation - hydrodynamics with realistic microphysics and opacities .
we now turn to the description and analysis of a 2d nonlinear cepheid model .
while the first paper centered around technical issues , the present one concentrates on the physical properties of the convection zone caused by the ionisation of he ii which also hosts much of the pulsational driving exhibited by the @xmath2-mechanism . to motivate this work we first discuss here a few items related to the physics of cepheid pulsations which explains the interest in the modelling of the pulsation - convection interaction .
rewards from such work are expected in the course of time , both in the immediate and in the later future .
the success of the effort put into modelling the pulsation - convection interaction has , to a large degree , been measured by the accuracy up to which properties of _ individual cepheids _ can be reproduced ( light curve , in particular also special features such as humps and bumps , etc . ) .
there is now a large body of 1d simulations which are utilized in various astrophysical contexts .
some problems and trends in cepheid research have been described in two reviews , @xcite and @xcite .
the status and progress in research on the related rr lyr stars has recently been reviewed by @xcite .
it is remarkable that both authors stress open problems in the treatment of pulsation - convection interaction . instead of repeating the problems discussed in those reviews we want to shortly describe a few problem areas which are either different from those discussed there or are considered here from a different point of view , and where better insights to the pulsation - convection interaction can expected to be crucial for progress .
let us , first of all , point out that in the usual 1d simulations for those stars convection is modelled by simplified non - local , time - dependent models , specifically adapted to an environment varying due to the radial pulsations . generally , such models need some equation for the turbulent kinetic energy , and due to the variable environment at least a time evolution equation for the turbulent kinetic energy is necessary .
this is in contrast to the possibly static equation for the turbulent energy which is applicable in stars which do not exhibit pulsation .
see the discussion in @xcite where the most widely used models are described .
these models require , however , to know a number of parameters , the so - called @xmath3 s , which are unknown a priori .
typically , these parameters are calibrated so that models reproduce properties which those objects exhibit , for example , the existence and the location of the red edge of the pulsation instability strip .
sometimes they are simply guessed .
since the number of these parameters is , in addition , considerable , and even the form of the added equations or closure approximations may well be ambiguous concern must be expressed as to the reliability of this approach . as a consequence ,
these models must be tested and , if feasible , the parameters calibrated in different , independent ways .
presently , it is not known to what extent the results obtained from conventional models are affected , at least in the case of cooler pulsating stars where effects due to convection are largest .
it is therefore of interest to consider a specific problem where what had once been considered progress has again been challenged just on grounds of inappropriate modelling of convection : the ( 1d ) simulation of double mode cepheids . in parameter regions where double mode behaviour is observed , nonlinear , radially symmetric models for pulsating stars have failed to reproduce such a behaviour for a long time .
it has therefore been considered to be an important breakthrough when , ultimately , models did exhibit double mode pulsation indeed .
these results have , however , been criticized by @xcite , where also references to other papers are given , on grounds of inappropriate modelling of convection , in particular because of much too large overshoot .
indeed , the question of how to explain double mode pulsation in cepheids is now considered open again .
quite generally , these closure parameters of the convection models are known to depend on effective temperature , chemical composition , etc .
, insofar as they are can be determined at all with some reasonable reliability .
this holds already true for the `` classical '' mixing - length parameter @xmath4 .
consider for example the discussion in @xcite .
increasingly , cepheids and related variables are used to investigate stellar populations , see , for example , @xcite .
hence , a proper assessment of the validity of conclusions is becoming both more important and , because of the possible influence of differing chemical abundances on the mixing length and other parameters , more difficult .
another emerging area of research is the detection of multiple periodicities present in some of these objects due to the advent of long , uninterrupted observations of high quality with the corot and kepler missions .
the detection of nonradial modes , e.g. @xcite , amounts to new tasks for nonlinear modelling and in particular leads to the need for multidimensional simulations .
let us also point out the usefulness of multidimensional cepheid and rr lyr studies for investigation of the atmospheres of these objects and the implications of such work .
whereas in solar and in general stellar physics use is increasingly made of multidimensional model atmospheres for abundance analysis and other types of research , e.g. @xcite , static model atmospheres are predominantly used in work about cepheids with only few exceptions , e.g. @xcite .
an appropriate treatment of the so - called projection factor which relates actual pulsation velocities to those observed turns out to be of importance in establishing the rotation law of the galaxy and issues regarding cepheid distance scale calibration ( cf .
quite generally , the increasingly sophisticated analysis of light curves of cepheids and rr lyr stars , e.g. @xcite , should be backed up by a comparable sophistication of atmospheric models , therefore possibly hydrodynamical and , if convection is important , multidimensional models . with these long - term goals in mind
we present in this paper 2d hydrodynamical models of a cepheid .
as discussed already in paper i , the vastly different spatial scales and therefore stringent resolution requirements make this a computationally quite intensive task even in 2d . in the following we investigate properties of the he ii partial ionisation zone , in particular issues regarding the energetics of the convection and the overshoot zone and implications for modelling of convection done in the traditional way in this area ( models of such a sort that they fit into the framework of the geometrically one dimensional hydrodynamic modelling of radially pulsating stars ) , and
we discuss issues of the atmospheric structure and numerical complications which are not so severe in low resolution and show their full relevance only in high resolution simulations .
the remainder of this paper is organised as follows . in sect .
[ sec : models ] we describe the physical parameters of the cepheid which we model with our simulations .
the analysis of the enthalpy flux in and around the partial ionisation zone of he ii and the mechanical work performed by the pulsation over a complete cycle as well as the comparison of these quantities with predictions from time - dependent convection models is presented in sect .
[ sec : pulsation ] . in sect .
[ sec : discussion ] we provide a discussion of our results and present our conclusions .
the model we discuss in this paper corresponds to a fundamental tone cepheid .
its parameters have already been given in paper i. we repeat them here for convenience .
our star has an effective temperature of @xmath5 and a gravitational acceleration at its surface of @xmath6 .
its luminosity is @xmath7 , the initial radius is @xmath8 ( which is about @xmath9 ) , and the total stellar mass amounts to @xmath10 .
the chemical composition is characterized by @xmath11 and @xmath12 , the mix for @xmath13 being taken from @xcite . the outer 11.3 gm are modelled . as a result the pulsation period @xmath14 measured from our simulations is about 3.85 days which is slightly shorter than if one included the entire star in the computational domain .
the present work is based on a model whose sector has an opening angle of @xmath15 .
the vertical cell size ranges from @xmath16 at the top to 124 mm at the bottom .
further details on the numerical parameters for the model and on the numerical methods used to solve the dynamical equations as well as a description of the dynamical equations themselves are given in paper i.
in this section we discuss properties of the he ii of the model just mentioned .
this work is based on data covering ten useful pulsation periods after model relaxation .
averaging over ten periods allows for a sufficient removal of special properties of individual periods .
after all , a problem typically experienced when running 2d convection simulations the small number of plumes present at any given time unless the domain is very broad .
this makes for poor statistics compared to 3d runs with similar horizontal extent .
the he ii zone is well resolved in each of the models we consider in the following .
given the importance of convection modelling in traditional ( 1d ) cepheid calculations naturally great efforts have been undertaken to derive such models .
the most frequently used models are those of @xcite and of @xcite as well as a number of variants based on them .
convection modelling in the sense of kuhfu has been adopted with modifications in the code described by @xcite .
variants of the stellingwerf model and the kuhfu model have been used in codes such as those of @xcite and the florida - budapest cooperation ( e.g @xcite ) and also in the more recent code described in @xcite .
these convection models introduce one additional time - dependent differential equation for the turbulent kinetic energy density , @xmath17 , plus algebraic recipes for various physical quantities .
more elaborate convection models using the reynolds stress approach have been developed , in particular those of @xcite and @xcite , but they have not yet been integrated into @xmath18 nonlinear stellar pulsation codes .
the comparably complex model of @xcite has been used to study the pulsational stability of red giants , but only at the level of providing the background model for linear , nonadiabatic stability analyses ( see @xcite and @xcite ) .
we therefore concentrate on examining the prescriptions of the kuhfu and stellingwerf models .
one of the differences between these models is the way in which the convective flux is expressed from the basic quantities , i.e. , those derived from the navier - stokes equations plus @xmath17 .
errors in the prediction of the convective flux may stem from two sources : from errors in @xmath17 itself ( and therefore the basic equation for this quantity ) and from an inadequate algebraic expression used to calculate the convective flux @xmath19 . in the tests performed here
we compute @xmath17 from our @xmath20 simulations and thus we test the algebraic form for @xmath19 . for two reasons
this is the proper way to proceed in the present context .
the first is a physical one and the other is of a primarily numerical nature .
the _ physical _ reason is conected with the formulation of the time dependent convection ( tdc ) equation . in the form given by @xcite
it reads @xmath21 where @xmath22 is the turbulent pressure and @xmath23 is the viscous eddy pressure .
@xmath24 is the flux of turbulent kinetic energy and @xmath25 is a coupling term of the form @xmath26 where @xmath27 denotes a source term and @xmath28 is the dissipation rate of the ( turbulent ) kinetic energy .
both the viscous eddy pressure @xmath29 and the dissipation rate @xmath30 concern actions on a very small scale and in actual practice can not be computed without model assumptions concerning just those small scale processes . from the standpoint of @xmath20 or @xmath31 modelling that means , however , that these quantities can not readily or unambiguously be recovered from the numerical simulations .
after all , the latter contain information about the resolved scales only .
the hope and expectation ( which to a considerable extent was supported e.g. by simulations of solar granulation ) in connection with such models is that they deliver useful and reliable answers even if the parameters ( reynolds number , rayleigh number , etc . ) in the numerical models differ widely from those found in actual objects ( cf .
@xcite ) . at the same time
it should be kept in mind that eq .
( [ eq : eturb ] ) in itself is phenomenological only .
it is in at least one way connected with simplified expressions for @xmath28 and connected to the kolmogorov dissipation picture as described in @xcite . however , eq . ( [ eq : eturb ] ) is not built into the multidimensional simulations . considered from a more basic point of view , eq .
( [ eq : eturb ] ) is not conservative .
moreover it can not be brought into that form with the approximate associated expressions for @xmath28 and @xmath27 .
the physical reason for this is that turbulent kinetic energy considered on its own is simply not a conserved quantity : there has to be a source of turbulent kinetic energy to prevent it from decaying . in the models used in practice the source and dissipation terms
are , from a quite fundamental point of view , not provided in conservation form . on the other hand , our numerical methodology in antares rests on conservation form .
this is for the very good reason that in hydrodynamics only conservation form can be expected to deliver correct answers for shock speeds and the like ( cf .
@xcite ) . nevertheless , even if we chose to solve eq .
( [ eq : eturb ] ) alongside , it would not have the same , or any , backreaction on our simulations because it is not one of the dynamic equations we solve .
it would therefore concern a quantity different from that one appearing in the tdc models despite of a formal , superficial analogy . therefore , similar to @xcite
, we undertake a direct comparison of the convective flux between the numerical hydrodynamical simulations and its approximation provided by the tdc models .
each of the tdc based fluxes has the form @xmath32 here , @xmath33 is @xmath34 for kuhfu and @xmath35 for stellingwerf .
@xmath36 is a correction factor whose purpose is to account for radiative energy losses of the convective elements . for the kuhfu and the stellingwerf case a slightly different functional form
is used ( see below ) .
@xmath36 contains , among others , the peclet number @xmath37 , which , again , is calculated somewhat differently in the kuhfu and in the stellingwerf setting .
our use of @xmath38 complies with @xcite .
note that e.g. @xcite use a somewhat different way to write the factor .
the functions @xmath39 , @xmath40 and @xmath36 depend on the depth . in the _ stellingwerf _ model @xmath41 is computed from @xmath42 where @xmath43 .
it furthermore adopts @xmath44 with a peclet number computed from the following expression : @xmath45 this form was proposed in @xcite ( see sect .
5.3 therein ) originally with the model of kuhfu in mind , although at this level the expressions for the convective flux have the same general structure ( [ eq : flux ] ) for both the stellingwerf and the kuhfu models . in ( [ eq :
eta_sw ] ) , @xmath46 is a parameter , unknown from the outset , @xmath47 is the rosseland mean opacity , @xmath48 the stefan boltzmann constant and @xmath49 the temperature .
moreover , @xmath50 is the usual heat capacity at constant pressure , @xmath51 denotes the density , @xmath17 the turbulent kinetic energy density , and @xmath52 stands for the pressure scale height . finally , @xmath53 is the superadiabatic temperature gradient , @xmath54 and @xmath55 denotes the actual temperature gradient @xmath56 while @xmath57 is the adiabatic temperature gradient .
for the _ kuhfu _ model we have @xmath58 to demonstrate the effect of radiative losses on the kuhfu model if approximated as in @xcite we introduce here a factor @xmath59 which contains some parameter @xmath60 . again
, it influences the efficiency with which the model diffuses surplus energy of convective elements and it is assumed to be of the form @xmath61 the peclet number for the kuhfu model is computed from @xmath62 where @xmath63 is the radiative conductivity .
this is the expression for @xmath64 derived by @xcite for the case that the dissipation rate @xmath28 is computed by means of a mixing length @xmath65 in the special case @xmath66 .
note that eq .
( [ eq : eta_kf ] ) can be transformed into eq .
( [ eq : eta_sw ] ) : the role of @xmath60 is reversed between the two , i.e.@xmath67 while @xmath68 .
consequently , all the differences between eq .
( [ eq : eta_sw ] ) and eq .
( [ eq : eta_kf ] ) can be absorbed into the numerical value for @xmath60 and the two forms of computing @xmath69 are equivalent .
( [ eq : eta_kf ] ) has the advantage of a more straightforward comparison with the model of @xcite and extensions thereof in future work .
we note that more detailed prescriptions to compute radiative losses within the kuhfu model itself have been suggested by @xcite , which , however , quite closely couple the computation of the convective flux with the dynamical equation for @xmath17 . in this study
we limit ourselves to the prescription given by eq .
( [ eq : flux ] ) which can readily be applied to both tdc models , as intended by @xcite .
last but not least we also compute the convective flux from our simulations through @xmath70 here @xmath71 is the vertical convective velocity ( the radial velocity after subtraction of the radial pulsational velocity ) and @xmath72 the enthalpy . in order to compare the fluxes used in traditional cepheid modelling with those from our simulations and relating them to the actual convective fluxes as calculated form the simulations we have averaged the latter ones over ten useful pulsation periods . in this way we improve the statistics .
after all , it is known that 2d simulations have disadvantages regarding statistics compared to 3d simulations due to the small number of convective plumes present at any given instance of time unless the domain is very wide . for each of the results we discuss in this section
the fluxes have been averaged horizontally .
[ fig : convflux ] displays the convective flux plotted against phase ( averaged over ten periods ) .
the variation of convection in intensity as a function of the phase is clearly visible . as fig .
[ pic : fc3 ] shows , convective activity ( as witnessed by convective flux ) commences its activity high up in the convectively unstable layer .
subsequently , plumes descend downwards . while their effect is still recognizable deep down ( in the overshoot zone , in fact ) , new convective activity again sets in at the top of the he ii partial ionisation zone .
the two subfigures in fig .
[ fig : convflux ] display two variants of the convective flux , namely @xmath73 , based on entropy perturbation against the mean , in fig .
[ pic : fc3 ] , and @xmath74 in fig .
[ pic : fc2 ] , where the total enthalpy enters ( cf .
[ eq : flux ] ) . in the upper picture
there is a noticeable orange - coloured streak , corresponding to plumes directed downwards , in particular in the overshoot region .
they have a large negative value of velocity , and @xmath75 because they are hotter than the ambience in the overshoot region .
this leads to a large positive value of @xmath73 , but to a negative value for @xmath19 .
in fig . [
fig : flux tot domain ] the mean convective flux @xmath19 is compared to its approximation provided by the tdc models over the total computational domain .
the left panel refers to the kuhfu and stellingwerf fluxes without the correction factor , which amounts to @xmath76 . in the panel to the right the correction factor is included .
similiar to the findings in @xcite , but less pronounced , the kuhfu formulation overestimates the convective flux in the overshooting region also in our case .
partial remedy can be achieved by including the peclet - factor @xmath77 as is evident from fig .
[ pic : phieta ] .
the results based on the stellingwerf model turn out to be only weakly dependent on the peclet correction .
the approximations also predict a slight negative overshooting above the convection zone .
the transition from convection to overshooting occurs closer to the surface than for the convective flux proper . in all figures and tables hereafter the peclet - factor will therefore be included .
[ fig : alpha_stripes ] shows @xmath78 as a function of phase for one full period ( actually even from phase -0.2 to 1.2 for clearer representation ) .
@xmath78 is obtained from @xmath79 .
the reddish vertical stripes represent the upper and lower limits , respectively , of the he ii convection zone proper quite well .
a systematic variation of @xmath40 with both the depth in the convection zone and with pulsation phase is immediately apparent .
furthermore , comparing corresponding values from the convection zone and the overshoot zone , one can easily spot phases where the @xmath40-values differ by a factor 5 - 15 and hence by an order of magnitude .
[ pic : kf - stripe ] [ pic : sw - stripe ] we therefore need to investigate the coefficients @xmath78 as a function of phase separately averaged over the convection zone and the overshooting zone . to this end
the following procedure has proved reliable and useful to determine whether a depth point belongs , at a given instant of time , either to the convection zone or the overshooting zone or to neither of them .
a layer belongs to the _ convection zone _ if , at this depth , more than @xmath80 percent of the horizontal points have a positive superadiabatic gradient , @xmath81 .
the gradients are computed from local values .
points below the convection zone belong to the _ overshooting region _ if they satisfy @xmath82 and fulfill , in addition , @xmath83 .
this choice guarantees both for the convection zone and the overshooting zone that the upper and lower boundary of the convection zone are excluded . near these boundaries the convective flux changes sign and hence is practically @xmath84 so that the ratio @xmath85 requested for the evaluation of @xmath0 ( cf . fig .
[ eq : flux ] ) leads to meaningless results . moreover
, in this region the enthalpy flux is small and the ratio @xmath85 is physically unimportant .
inspection of the simulation data shows that the overshooting region defined in this way nicely contains the downwards penetrating plumes .
[ pic : kf - conv ] [ pic : sw - conv ] [ pic : kf - over ] [ pic : sw - over ] [ pic : be - kf - conv ] [ pic : be - sw - conv ]
[ pic : be - kf - over ] [ pic : be - sw - over ] [ cols="^,^,>,>,>,^,>,>,^",options="header " , ] the cycle averages over the horizontal average @xmath86 of these work densities ( @xcite ) are the work integrands and express the amount of energy converted into pulsational energy . in fig .
[ fig : workintegrand ] the work integrands are plotted for the fluctuating velocity . when comparing to fig .
[ fig : flux tot domain ] it can be seen that for the thermodynamic pressure dampening occurs at the top of the convection zone and that the transition from excitation to dampening is within the overshooting zone .
the turbulent pressure has a dampening effect throughout where it is not small .
these work densities can then be integrated , at any time , over the convection zone ( @xmath87 ) , the overshooting zone ( @xmath88 ) .
for the present purpose the definition of the convection zone is the same as given in sect .
[ sec : flux_properties ] .
the overshooting zone is now defined a little bit differently from the one provided there .
namely , now there is no gap between it and the convection zone : points below the convection zone belong to the _ overshooting region _ if they satisfy @xmath82 and fulfill , @xmath89 ( instead of @xmath83 in sect .
[ sec : flux_properties ] ) .
this results in terms such as , for example , @xmath90 referring to the overshooting zone . in this way
the rather different physical nature of the zones is taken into account .
within such an integral positive and negative contributions may cancel out very significantly .
this can be seen by considering the integrated _ modulus _ of the pdv - terms such as @xmath91 alongside with the original values . for the thermodynamic pressure @xmath92 the variation with phase
is shown in fig .
[ fig : w_press ] . the work @xmath93 can be seen to mainly follow the pulsation pattern for each zone , more clearly so in the overshooting zone .
the average values @xmath94 ( see tab . [
tab : pw - values ] ) are positive everywhere , since excitation work can be observed at the top of the overshooting zone ( see fig .
[ fig : w_press ] ( a ) ) . in the convection zone
the pdv - terms reach their smallest values during contraction while the average values are always positive and the modulus shows little variation indicating roughly constant activity .
the panels on the right - hand side of fig .
[ fig : w_press ] show that in the overshooting zone the modulus of the pdv - terms derived from fluctuation quantities reach their highest values right after maximal contraction at the beginning of the expansion .
cancellation effects in the work integral are most dramatic for the @xmath95 terms , reaching up to about three orders of magnitude ( compare fig [ fig : w_press ] ( c ) and ( d ) ) . for the turbulent pressure the pdv - terms follow the pulsation pattern if at least one component includes pulsation . in the case of @xmath96 one
sees ( fig .
[ fig : w_pturb ] ( g ) and ( h ) ) that the dampening effects of the turbulent pressure perturbation are at full contraction at their maximum in the overshooting zone but at their lowest in the convection zone .
the modulus @xmath97 in the convection varies little .
[ pic : pu ] [ pic : mpu ] [ pic : p1u ] [ pic : mp1u ] [ pic : pu1 ] [ pic : mpu1 ] [ pic : p1u1 ] [ pic : mp1u1 ] [ pic : ptu ] [ pic : mptu ] [ pic : pt1u ] [ pic : mpt1u ] [ pic : ptu1 ] [ pic : mptu1 ] [ pic : pt1u1 ] [ pic : mpt1u1 ]
while simulations of pulsating stars ( cepheids , rr lyr ) have been computationally one - dimensional for decades , assuming radial symmetry , multidimensional calculations start to be feasible only recently .
they allow in particular the investigation of the pulsation - convection interaction which is considered a main open problem in the simulation of such stars ( @xcite ) .
we have constructed a two - dimensional model regarding convection and pulsation in a cepheid type star .
the justification for adopting two dimensions at the present state of the matter can be taken from a comparison of 2d and 3d results in @xcite .
these results show that ( for models off rr lyr stars ) many quantities such as peak convective flux as a function of phase , growth rates or amplitudes are quite similar in 2d and 3d , differences being mainly visible in the slope of the light curve at various times in the descending branch .
the aim of our investigation is to figure out basic properties of convection in the presence of pulsation and to draw inferences on existing ways of modelling these issues via time dependent convection models ( tdcs ) in a computationally 1d setting .
the tdcs just mentioned entail a number of coefficients not priorly known .
with respect to the coefficient @xmath98 , the calibration factor for obtaining the physically correct convective flux values from the tdc models of kuhfu and stellingwerf , the following conclusions can be drawn from our work . in the he ii partial ionisation zone and its surrounding overshooting zone convection is subject to large radiative losses , as is witnessed by a low peclet number ( @xmath99 predominantly ) .
it is thus difficult to calibrate @xmath0 separately from a coefficient @xmath60 , connected to radiative losses , at least from our present model .
after all , as explained in sect . 3.1 ,
these coefficients enter essentially only in the combination @xmath100 ( or @xmath101 in case of the stellingwerf model ) for our case of low peclet numbers .
our models show that using one and the same coefficient @xmath0 for the convection zone ( c.z . ) and the overshoot zone ( o.z . ) would result in much too large overshooting predicted by the tdcs .
however , @xmath0 and @xmath60 enter only in the combination described in sec .
[ sec : flux_properties ] .
consequently , including radiative losses by any choice of @xmath60 in the prescription for the convective flux can not correct the discrepancy of too large fluxes predicted by the model expressions for the overshooting zone in comparison with the interior of the convection zone . at present , the main effect of accounting for radiative losses , i.e. introducing the factor @xmath69 in eq .
( 3 ) , is to at least limit a more excessive overshooting which would otherwise result from applying the prescription of kuhfu ( see fig .
this model is more affected by that problem than the model of stellingwerf .
accounting for the change of anisotropy of the contributions to turbulent kinetic energy , however , is much more important . to avoid excessive
overshooting in the model requires introducing a functional dependence in eq .
( 3 ) which can take care of the fact that convective zones show stronger vertical fluxes than horizontal ones while the opposite holds for overshooting zones . however , as we have shown , even after such corrections a large discrepancy remains .
this requires to at least vary @xmath102 by a factor of 1.5 for the stellingwerf model and a factor of 4 for the kuhfu model to obtain convective fluxes from the models which are consistent with our numerical simulations .
in addition , there is a clear systematic trend of @xmath98 as a function of depth within the convection zone , see fig . [
fig : alpha_stripes ] .
the effect of this trend on modelling the pulsation in the traditional way is unknown .
some contribution to this might actually be due to the coupling between the mean structure and the turbulent kinetic energy which is certainly more complex than expressed through eq .
we should mention that the present comparison is already the most favorable one for the tdc models in the sense that we only take the model expressions to be meant to predict statistical averages : variations for particular instants of phase for a specific point in time may be larger and any attempt to interpret the expressions in a point - wise sense is falsified immediately by unacceptably large discrepancies ( at least two orders of magnitudes larger than the already large differences observed for the averaged quantities considered here ) . in this sense
the present simulations unveil major weaknesses of tdc models used to explain the pulsational properties of cepheids .
the present simulations allow to quantify the discrepancies expected from the tdc models for a certain group of cepheids .
since idealized simulations such as those of @xcite come to similar conclusions on a qualitative level , the discrepancies found are most likely of a fundamental nature and thus have to be taken care of .
otherwise one may fall into the trap of obtaining the right results for what might be the wrong reasons .
our results suggest that in 1d simulations one gets artificially large overshooting below the convection zone when applying identical values of @xmath0 in the c.z . and the o.z .. this reinforces the doubts which have been expressed by @xcite in respect to the achievements on double mode pulsation in 1d cepheid models .
these doubts are based on the suspicion of too large overshooting in the models which have shown double mode pulsation in a broader parameter range than previously .
one additional point is that 2d simulations tend to lead to larger overshooting than similar 3d simulations produce ( @xcite ) . considering this aspect as well
, the problem of depth - dependency of @xmath0 may well be even more severe than is borne out by the present investigation .
our results regarding the contribution of fluctuating quantities to the work integral reveal an intricate change of correlations as a function of phase . in table [
tab : pw - values ] we compare values of the contributions to the work integral , integrated over the convection zone , and the overshoot zone .
considering contributions to the work integral where both entering quantities are fluctuating ones ( instead of corresponding to the radial symmetric component ) it turns out that no such combination can be disregarded : at least for one domain of integration every combination leads to a contribution of the same size as the others .
that means that in principle each tdc model should faithfully represent the terms corresponding to all of the possible combinations ( plus those with just one fluctuating quantity of course ) .
it can be expected that work integral contributions with two fluctuating quantities will be more important in cases of stronger convection zones , since due to the more pronounced nonlinearity , their higher order terms will likely be larger .
a comparison of our simulations to tdc models regarding the energetics of the pulsation - convection interaction can not be undertaken in the same manner as has been possible for the @xmath3-factors discussed above .
there , the relevant flux as delivered by the tdc models can , by a simple algebraic formula , directly be evaluated and the scaling factor be determined ; the constancy or variability of this factor then immediately leads to an assessment of the quality .
this is to be contrasted to the case of the work integral .
there , the relevant quantities in the tdc ( turbulent energy or pressure plus furthermore others , not really provided by the usual tdcs ) are not algebraic in nature . instead , the equation for turbulent energy is an evolution equation where @xmath103 attains a value deemed physically correct only after it has been evolving for some time .
hence , comparison to the quantities from the multidimensional simulations can not be performed on the basis of a temporal snapshot . at the same time it is not really possible to include an equation for the turbulent energy @xmath103 in the multidimensional simulations : the simulations explicitly _ calculate _ @xmath103 while the equation for turbulent energy in the tdcs _ models _ @xmath103 .
consequently , there is a branching quite at the roots of the modelling processes between one- and multidimensional simulations which prevents a meaningful comparison of the two values of @xmath103 by solving the tdc equation alongside in multidimensional work .
after all , there is no room to allow for backreaction of the tdc variant of @xmath103 in multidimensions .
our calculations however show that , for modelling these values in the sense of tdcs , phase dependent properties of hydrodynamic quantities ought to be built into the models . after all
, the contributions to the work integral ( which are related to correlations of physical quantities ) have a distinctive phase dependence which , moreover , is not in all cases similar in the convection and the overshoot zone .
this can not really come as a surprise given that convection in a pulsating star is not in a statistically steady state .
rather , it continually evolves in a specific way : at first convection cells form at the top of the convection zone at a certain phase , plumes then descend , overshoot into the stable zone below at some later phase and ultimately the remnants die off deep below the convection zone . explicitly or implicitly ,
each tdc should account for such a sequence of events .
the simulations reported above suggest some guidelines for the development of revised tdcs to eventually be used in 1d simulations .
radiative losses have to be included in the model , but in addition the convective flux will have to evolve in an explicitly phase dependent manner .
therefore an additional evolution equation for @xmath104 seems inevitable , similar to the model of @xcite , further developed in @xcite , or the rarely used 3-equation variant of the kuhfu model .
moreover , the distribution of kinetic energy clearly changes as a function of depth in a non - trivial manner and most likely also features phase dependencies .
thus , a fully non - local reynolds stress model as given by @xcite , @xcite and @xcite may be required as a minimum level of complexity to substantially improve tdc models in comparison to those presently used .
when finally turning our attention to further work along these lines , several items come to mind .
one obvious goal consists in the construction of cepheid models with a stronger he ii convection zone in order to more fully sample the parameter space .
such material may be crucial for improved tdcs for use by the 1d pulsation modelling community .
one further goal is an investigation of the upper , h+hei , convection zone and , hence , the atmosphere .
we have already performed some initial runs with the necessary resolution which is considerably higher in the upper layers than in the models which we have reported here . due to the fact that , with higher resolution , the h+hei convection zone is much stronger than in our present models
we have had some difficulty with the higher resolution models . in particular
, throughout the atmosphere ( and also below ) there is vivid non - radial shock activity .
this has lead to numerical problems which required the development of numerically suitable boundary conditions .
such conditions were devised in the context of solar granulation ( @xcite ) .
a modification of these conditions has recently been found to work for the cepheid case .
shock activity with shocks up to a mach number of about 4 , triggered by convection , has also been found in preliminary calculations we have done with these new conditions . this highly dynamical situation in the atmosphere of the cepheid model also leads to a highly corrugated optical surface .
to describe such a situation by a static or even a time - dependent model one - dimensional seems extremely challenging , and the implications for an analysis of photometric or spectroscopic data will need to be figured out .
a different avenue of research could center on investing pulsation in a ring ( or , in 3d , a spherical shell ) as initiated in preliminary work ( @xcite ) .
such simulations could shed light on the excitation of non - radial modes if they indeed show up , connected or not connected with convection , whatever the outcome of the simulations would be .
this work has been supported by the austrian science foundation , fwf grant p18224 .
we are thankful to g. houdek for supplying us with starting models .
calculations have been performed at the vsc cluster of the vienna universities .
fk and em gratefully acknowledge support through the austrian science foundation fwf , grant p21742 . the work of fk has also been supported by fwf grant 25229 . | non - local , time - dependent convection models have been used to explain the location of double - mode pulsations in cepheids in the hr diagram as well as the existence and location of the red edge of the instability strip .
these properties are highly sensitive to model parameters .
we use 2d radiation hydrodynamical simulations with realistic microphysics and grey radiative - transfer to model a short period cepheid .
+ the simulations show that the strength of the convection zone varies significantly over the pulsation period and exhibits a phase shift relative to the variations in radius .
we evaluate the convective flux and the work integral as predicted by the most common convection models .
it turns out that over one pulsation cycle the model parameter @xmath0 , has to be varied by up to a factor of beyond 2 to match the convective flux obtained from the simulations . to bring convective fluxes integrated over the he ii convection zone and the overshoot zone below into agreement
, this parameter has to be varied by a factor of up to @xmath1 ( kuhfu ) .
+ we then present results on the energetics of the convection and overshoot zone by radially symmetric and fluctuating quantities . to successfully model this scenario by a static
, one dimensional or even by a simple time - dependent model appears extremely challenging .
we conclude that significant improvements are needed to make predictions based on 1d models more robust and to improve the reliability of conclusions on the convection - pulsation coupling drawn from them .
multidimensional simulations can provide guidelines for developing descriptions of convection then applied in traditional 1d modelling .
[ firstpage ] stars : variables : cepheids hydrodynamics convection |
evacuation through a doorway or a narrowing is a long - standing issue . in homer
s _ odyssey _ , odysseus and his men need to escape from a cave without being noticed by its blinded guardian , polyphemus the cyclops , who relies on his sense of touch to mount guard ; they manage to do so by tying themselves to the undersides of sheep . nowadays
, the possibility to evacuate quickly and safely from a public facility or building in emergency conditions should be ensured by the compliance with building codes @xcite .
these codes may be prescription - based or performance - based @xcite , depending on whether they constrain certain design elements , e.g. , door widths , or set quantitative goals , such as permitting the evacuation of @xmath6 persons within a given time .
a central issue is then the assessment of exit capacities @xmath1 , that is , the number of people who can walk through a given exit per unit time .
often , @xmath1 is evaluated on the basis of tabulated values resulting from empirical observations or controlled experiments .
the required amount of tabulated data can be reduced by defining a specific capacity @xmath7 , i.e. , a capacity per metre width , so that @xmath8 for a narrowing of width @xmath9 .
this simplification relies on the assumption of a roughly linear growth of @xmath1 with @xmath9 , which is supported by some experimental evidence@xcite , at least at moderate pedestrian densities and for @xmath10 . when the exit width @xmath9 is varied , kretz et al .
@xcite reported a moderate decrease of @xmath7 with @xmath9 , whereas seyfried et al .
@xcite found a slight increase .
reference @xcite stands out in the literature , in that it suggests a stepwise increase of the capacity with @xmath9 , due to the formation of equally spaced lanes , but this hypothesis stems from very limited evidence ( only two widths were studied ) . ] unfortunately , as often emphasised in the literature @xcite , the tabulated capacities vary widely between handbooks or papers : for instance , for constrictions of width @xmath11 , we find @xmath12 in the sfpe handbook @xcite , @xmath13 in the `` planning for foot traffic flow in buildings''@xcite , @xmath14 in kretz et al.@xcite , and @xmath15 in seyfried et al . @xcite .
these discrepancies are ascribed to various factors . to start with , the use of distinct measurement methods , e.g.
, the recourse to either spatial or temporal averages to define the flow or the density , has been incriminated and can indeed lead to different results , especially at high densities , even in a simple single - file flow @xcite ; but zhang et al .
came to the conclusion that this point only has a relatively minor impact at lower densities @xcite . besides this technical aspect
, cultural differences between populations may play a role ; their importance is however debated , notably owing to the similarity in the measurements performed in the london underground ( uk ) and in osaka business district ( japan ) , as mentioned by seyfried et al.@xcite .
the composition of the population and the associated morphological differences are arguably of greater importance , with higher flow rates for crowds consisting of children @xcite ( @xmath16 ) , due to their smaller size .
furthermore , it was found that short bottlenecks ( mimicking the passage through a doorway ) allowed higher flow rates than longer ones , which simulate the entrance to a narrow corridor @xcite ; in the former case , it is possible to slither around the door jamb . perhaps for similar reasons , trying to predict an exit capacity on the basis of the peak value of the flow rate _ vs. _ density curve for a simple uni - directional flow ( the so called fundamental diagram ) tends to underestimate the actual capacity @xcite .
even more notably , the density of the pedestrian crowd at the beginning of the evacuation strongly impacts the flow rate @xmath1 , with a faster flow for initially denser crowds @xcite . not unrelated to this dependence on the initial configuration
is the question of the stationarity of the pedestrian flow @xcite : often the measurements are not performed in the stationary regime , which further complicates the interpretation of the results . to study the evacuation dynamics in simpler , or experimentally more tractable , systems ,
researchers have investigated constricted flows of less complex entities , such as vibrated grains@xcite , ants@xcite , and sheep @xcite , and brought to light similarities with pedestrians ( for ants , the analogy is however debated @xcite ) .
but the complex psychology of humans adds a level of complexity to their response .
in particular , on account of their individual characters , pedestrians exhibit diverse reactions to external stimuli and variable behaviours . in a series of repetitive evacuation experiments , habituation and boredom may lead to considerable variations between repetitions @xcite . besides , most of the aforementioned research focuses on pedestrian flows in normal , cooperative conditions , and competitive evacuation experiments are rare , probably because of the risks they present ,
although in reality impatient , competitive , aggressive or vying behaviours are also observed in some situations @xcite . among but a few other examples , muir et al.@xcite simulated the evacuation from an aircraft , in which the participants competitiveness was whetted by the possibility of a reward ; nagai et al .
@xcite investigated the evacuation of relatively competitive crawlers and walkers ; helbing et al . conducted panic - like evacuation experiments , but with only about 20 participants .
( also see ref .
@xcite for a study of uni - directional flows , with some prescribed participants behaviours ) .
more recently , in a series of papers@xcite , garcimartn , zuriguel , and co - workers carried out a detailed analysis of controlled evacuations in which distinct levels of competitiveness were prescribed to the crowd .
( also see @xcite for experiments in a virtual environment . )
it is noteworthy that , compared to the values reported above for normal conditions , the specific capacities measured in those ( diversely ) competitive settings are considerably larger : @xmath17 in the work of nagai et al.@xcite and helbing et al.@xcite , @xmath18 in garcimartn et al.s paper@xcite for door widths of 70 to 80 cm .
however , these evacuations were performed with crowds of homogeneous competitiveness . to get greater insight into the effect of the individual pedestrians behaviours , we conduct and analyse controlled evacuations through a narrow doorway with participants that are prescribed
_ distinct _
behaviours , namely , either a polite or a selfish behaviour .
we also vary their eagerness to egress .
although these behavioural aspects add yet another component to a complex problem , the global picture of the evacuation process is clarified by considering a ( well - nigh ) stationary regime and seeking robust `` microscopic '' characteristics of the flow .
let us mention that a concise account of the _ global _ flow properties observed in these experiments was proposed in a preliminary report@xcite .
the evacuation experiments were performed in the gymnasium of centro atmico bariloche ( cab ) , argentina , and involved more than 80 voluntary participants ( students and researchers ) , aged 20 to 55 for the greatest part , with a woman / man ratio of about 1:3 .
the participants were asked to evacuate a delimited area through a @xmath19-wide door ; the geometry is sketched in fig .
[ fig : sketch ] .
this doorway was created by moving @xmath20-thick sliding walls .
safety was a central concern .
accordingly , the door jambs were padded with training mats , which reduced the aperture to 72 cm . in addition , because of the possible ethics and safety issues , the protocol was validated beforehand by a local ethics committee and the experiments were prepared in collaboration with the safety and hygiene group of cab .
the evacuation process was supervised by three staff members who could stop it at any moment by blowing a whistle .
the experimental design is directly inspired by the experiments of zuriguel s group@xcite , but we introduced some important changes . in each evacuation experiment , a fraction @xmath21 of the participants was told to behave selfishly while the rest should behave politely .
the selfish agents were selected randomly and varied between the experiments ; they were asked to wear a red headscarf to be recognisable on the videos .
they were allowed to `` elbow their way through the crowd , with mild contacts but no violence whatsoever '' .
their polite counterparts , on the other hand , were to `` avoid any contact and try to keep their distance '' . before each experiment , all participants were asked to walk as close as possible to the door , without crossing it .
only then did we announce who would behave selfishly in the upcoming run .
the session started with a mock evacuation .
then , in a first series of experiments , here referred to as experiments with _ placid _ walkers , all participants were instructed to `` head for the door '' , without further specification . in the second series , they were told , in addition , to `` walk purposefully ( _ con ganas _ in spanish ) , but without running , pushing or hitting others '' ; in this case , the walkers are said to be _
overall , the experiments lasted for somewhat more than one hour . finally , the often reported lack of stationarity in the flow , along with the word of caution in ref .
@xcite about the risk of faulty statistics with too small crowds , raised strong concerns in our minds about finite - size effects .
to curb these effects , we decided to innovate by imposing `` periodic boundary conditions '' on the crowd : after egressing , participants followed a circuit that led to their re - injection into the room .
in fact , to optimise the randomisation and limit the clustering of , say , fast participants , two re - injection circuits were set up , a short one and a longer one , and the evacuees were directed to either one alternatively . with this contrivance , about 250 passages through the door ( between 177 and 352 , to be precise ) were obtained in each experiment .
, @xmath22 , and the door is @xmath19-wide .
the short and long re - injection circuits are approximately @xmath23 and @xmath24 long , respectively . ] to complement our controlled experiments , we filmed the egresses from a conference room at the various breaks of the trefemac conference , a three - day physics workshop held at cab .
the participants were informed of the recording once and for all . some tendency to align before passing through the 82-cm - wide exit door
was observed ; the door was opened at more than @xmath25 .
the data , presented in fig .
3e of the main paper , consist of about 260 individual passages .
we also filmed two collective egresses from the auditorium of cab at the end of the weekly seminars .
the 75-cm - wide door was fully open and the 50 to 100 participants were unaware of the recording
. some were carrying a chair when passing through the doorway .
the data , presented in fig . 3f , consist of about 150 individual passages .
in these uncontrolled settings , large time lapses between egresses are irrelevant here , insofar as they are due to the foot - dragging of participants within the conference room , rather than to the dynamics at the exit .
the controlled evacuations were recorded with two 60hz , large - angle go pro cameras and one standard 60hz camera .
two cameras were placed above the door , and one filmed the crowd in the room from a more distant point of view .
some of the videos are available as supplemental materials . with respect to the image analysis ,
a standard method to get the egress times is to build a time line of passages ( see supplementary fig .
s1 ) by extracting a few lines of pixels just past the door from all video frames and then stitching them together . but to get more comprehensive information , we resort to a more exhaustive semi - manual analysis .
we define the exit zone as the @xmath26 semi - circular area in front of the door shown in supplementary fig .
s2 and we flick through all video frames , down - rated to 30hz , clicking on the participants heads when they enter the exit zone and pressing a dedicated key when they cross a line beyond the door .
both methods agree very well regarding the exit times .
thus , the density @xmath2 is assessed by detecting the pedestrians first entrance into the exit zone .
this method creates a small artefact in the most competitive evacuations , in which a few participants were shortly pushed out of the zone , due to a crowd movement , before moving in again and passing the door , a possibility that is discarded in our routine , hence a slight overestimation of the density in the zone in these cases .
moreover , no correction for optical distortion was performed , because we used the central part of the images . despite this central position
, the parallax resulting from the camera not being strictly at the zenith ( owing to the impossibility to place it high enough ) implies that the moment of detection of a given participant s entrance into the exit zone somewhat depends on their height .
this results in a slight inaccuracy with respect to the effective area of the zone in front of the door , hence the use of a pre - factor @xmath27 in the definition of the density unit ( @xmath28 ) . wherever relevant ,
we show error bars on the graphs to give an idea of the uncertainty due to the finite number of measurements we performed . in these cases ,
the data points are obtained as averages over a sub - sample of our measurements ; the error bars are systematically @xmath29 on each side of the plotted data point , where @xmath30 is the standard deviation of the sub - sample and @xmath31 is the sub - sample size . under the reasonable assumptions that the sample points are approximately uncorrelated and that the central limit theorem applies , the error bar then represents a 95% confidence interval .
note however that this definition of the confidence interval is valid only in the limit of large @xmath31 , whereas in our case @xmath31 is generally of order 100 - 300 ( in fig . [
fig : t_w](right ) , @xmath32 ) .
the experiments went as planned , without incident .
the re - injection process operated properly : the flow was limited by the passage through the doorway , and not by travel times in the re - injection circuits .
some selfish agents tried to overtake the rest of the crowd by walking near the lateral boundaries of the delimited area , instead of going into the thick of the crowd .
more generally , we observed the formation of files of a few selfish agents following each other and taking advantage of the ` voids ' opened by their predecessors .
we believe that this follows from a general mechanism of clustering via a coupling between the structure of the medium and the motion of the particles ( for instance , the pairing of electrons in a superconductor @xcite ) , although it is true that , here , the formation of files may have been facilitated by the conspicuousness of the headscarf - wearing selfish agents .
importantly , given that selfish agents made their way faster , they were re - injected into the room more frequently .
accordingly , the effective fraction @xmath33 of selfish evacuees is generally larger than the prescribed ( nominal ) fraction @xmath0
. the results below will be presented as a function of @xmath33 , where relevant , rather than @xmath0 . for polite crowds ,
i.e. , at low @xmath33 , the evacuation looks orderly and contacts between pedestrians are avoided .
some polite participants yield deliberately and may even wave to their neighbours to go ahead .
very generally , there is no more than one pedestrian crossing the door at each time . for larger @xmath33 ,
the crowd in front of the door becomes more compact , contacts surge , and the flow looks more chaotic as several pedestrians try to walk through the door simultaneously .
such endeavours sometimes lead to clogs due to the formation of ` arches ' of three to four participants across the door , particularly at @xmath34 . but these clogs are rapidly resolved , in about one second or less , and do not _ substantially _ delay the flow . in the most competitive evacuations ( @xmath35(purposeful ) and
@xmath36(purposeful ) ) , some agents grasp the door jamb and use it to pull themselves past the door ; a few pedestrians spin around upon egressing , because of the contacts with their neighbours .
to go beyond these general observations , we extract the exit times @xmath37 of each pedestrian @xmath38 , numbered according to their order of egress , regardless of their identity ( see methods for details ) .
this allows us to compute the time - dependent flow rate @xmath39 , which is a moving average over a time window @xmath40 , for each evacuation , viz .
, @xmath41 where @xmath42 is the heaviside function , viz .
, @xmath43 if @xmath44 , 0 otherwise . for @xmath45 ,
the curve @xmath39 is extremely jagged , as can be seen in fig .
[ fig : flowrates_7](left ) .
it consists of somewhat irregular spikes that are all the larger as the evacuation is competitive ( at large @xmath33 ) . but
averaging over a larger time window , @xmath46 , yields a smoother curve , whose relative flatness proves that , macroscopically , the pedestrian flow is quasi - stationary .
interestingly , this quasi - stationary is reached almost immediately after the beginning of the experiments and persists until at least a dozen seconds before their ends(purposeful ) is a slight exception , insofar as the flow rate wanes moderately in the last 30 or 40 seconds . ] .
the density of pedestrians in the exit zone is also roughly stationary in general ( see supplementary fig .
our observation of macroscopically quasi - stationary dynamics contrasts with the non - stationary flows found in a number of related works@xcite .
we see two main reasons for this contrast : a 3-m - large gap ( 1-m - large in ref .
@xcite ) in front of the exit was initially left free of participants in ref .
@xcite , which explains the existence of a transient regime , and no re - injection of the participants was enforced in those earlier experiments .
no matter how limited the transient effects are in our case , to further reduce them , we henceforth discard the first three and the last ten seconds of all experiments .
@xmath2 ( @xmath47 ) & flow rate @xmath1 ( @xmath48 ) & number of full turns + + 0% & 0% & 2.69 & 1.01 & 0 + 30% & 45% & 4.09 & 1.35 & 0 + 30% & 47% & 4.94 & 1.41 & 0 + 60% & 71% & 6.04 & 1.71 & 0 + + 0% & 0% & 3.70 & 1.26 & 0 + 10% & 18% & 4.49 & 1.39 & 0 + 60% & 71% & 7.63 & 2.20 & 3 + 90% & 92% & 8.26 & 2.36 & 3 + 100% & 100% & 8.98 & 2.41 & 5 + in the evacuation with @xmath36 ( purposeful ) , for @xmath49 ( thin lines ) and @xmath50 ( thick lines ) .
( right ) global flow rate @xmath1 as a function of the local density @xmath2 at the door .
the dots represent the experiments with placid participants and the triangles , the purposeful ones .
the dashed line is the prediction of the naive model presented in the text .
the geometrical prefactor @xmath51 in the density unit is due to experimental uncertainty . ]
we now compute the global flow rates @xmath1 for each evacuation and list them in table [ tab : global ] .
overall , @xmath1 grows monotonically with the concentration @xmath33 of selfish agents , and a peak value of 2.4 persons per second is reached at @xmath36 . but
@xmath1 also increases when the placid crowd is asked to walk purposefully ( irrespective of the individual behaviours ) : for the same value of @xmath33 , namely 71% , @xmath1 surges by almost 30% following this change .
this sensitivity to the participants attitudes , while the geometry and the crowd are kept fixed , renders the divergence of published values for the exit capacity quite understandable .
turning to a comparison with these published values , we see that the specific capacity corresponding to the placid polite crowd ( @xmath52 ) is comparable to the rather conservative specification of the sfpe handbook@xcite , namely , @xmath53 , while the specific capacity for the purposeful polite crowd ( @xmath54 ) is similar to kretz et al.s measurement @xcite ( @xmath55 ) for a flow through @xmath56-wide , @xmath57-long bottleneck in normal conditions , and comparable to the value @xmath58 reported by seyfried et al.@xcite for a flow through @xmath59-wide , @xmath60-long bottleneck in normal conditions . at the other extreme , the highest flow rate that we measured , @xmath61 for @xmath36(purposeful ) ,
is slightly below , but similar to , the range of flow rates obtained in the more or less competitive evacuations of pastor et al.@xcite ( @xmath62 ) for a 69-cm - wide ( instead of 72 here ) door .
the increase of @xmath1 with @xmath33 and with the agents purposefulness in our experiments points to the absence of ` faster - is - slower ' effect , or , better said , a ` more - competitive - is - slower ' effect .
this may seem to be at odds with the conclusions of pastor et al .
@xcite , but we claim that there is no contradiction : unlike @xcite , our experiments are not ( all ) above the threshold of competitiveness required to observe this effect . indeed , in @xcite , competitive participants were allowed to push each other with stretched arms , unlike in ours ; the larger competitiveness is also reflected by the larger flow rates that they measured .
the granular analogy then helps clarify the consequence of this difference @xcite : in a granular hopper flow , as the panel supporting the grains is tilted and the effective gravity increases accordingly , in a first regime the flow becomes faster , because grains fall faster , but then it slows down under higher gravity .
this slowdown is due to the stabilisation of clog - inducing arches by pressure .
models for pedestrian dynamics premised on newton s equations with `` social forces '' predict a similar non - monotonic effect of the pedestrians desired velocity @xcite . as this velocity
is increased , in a first regime the evacuation gets faster , but a second regime then emerges , where the repulsive forces between clustered pedestrians at the exit exceed the driving force associated with the desired velocity ; the ensuing congestion reduces the flow rate .
we presume that the evacuations of @xcite are all within the second regime , whereas ours are not .
we have already noted that the flow rate is not uniquely determined by @xmath33 in our experiments .
following seyfried et al.s @xcite idea that the discrepancies between the exit capacities in the literature are largely due to different initial crowd densities , we investigate the dependence of @xmath1 on the density @xmath2 in the exit zone ( a @xmath63 semi - circular area in front of the door ) . some experimental uncertainty regarding our density measurements , detailed in the subsection _ video analysis _ of _ methods _ , notably leads us to introduce a pre - factor @xmath64 in the density unit ( @xmath65 ) .
the highest densities that we report , namely @xmath66 , may appear intriguingly large ; it should however be emphasised that here @xmath2 is not the average density in the crowd , but its _ local _ value near the exit , where pedestrians are most densely packed ; previous works have already shed light on the considerable deviations between the average density and its peak local value @xcite . the graph @xmath67 plotted in fig .
[ fig : flowrates_7](right ) confirms the relevance of the density parameter : the flow rates measured for different behavioural prescriptions collapse onto a smooth master curve .
the monotonic increase of the flow rate @xmath1 with @xmath2 is remarkable . at low densities ,
the order of magnitude of @xmath1 and its growth with @xmath2 can be rationalised rather straightforwardly .
first , notice that pedestrians in the exit zone are separated by @xmath68 on average .
besides , they generally egress one by one and lower their velocity when the door is blocked by somebody else s passage . should one crudely assume that this produces a halt , after which they gather speed as @xmath69 with a final velocity @xmath70 and a response time @xmath71 , one will easily obtain the time interval required to walk a distance @xmath72 , @xmath73 ( provided that @xmath74 ) , hence a flow rate @xmath75 .
this is undoubtedly a very crude model , but its prediction , represented as dashed lines in fig .
[ fig : flowrates_7](right ) ) , is in broad agreement with the low - density data .
much more surprisingly , @xmath1 keeps increasing up to high densities , @xmath76 , contrary to most fundamental diagrams in the literature , which predict its decline above much smaller densities@xcite ( see fig .
11 of ref .
the slight uncertainty in our assessment of the density can hardly be the cause of discrepancy : in the highest - density experiment , at @xmath77 , the crowd is tightly packed near the door , which is consistent with the reported density value .
instead , it is likely that at such high densities the flow rate is not uniquely determined by @xmath2 , but strongly depends on other parameters , such as the pressure in the crowd .
indeed , the tightly - packed crowd is nearly incompressible ; therefore , the mechanical pressure can increase vastly , hence stabilising clogs , while @xmath2 only changes little . crowd turbulence
@xcite may also arise , thereby affecting the density and flow . in our settings , excessive mechanical pressures and turbulence
were warded off thanks to the strict instructions given to the participants .
importantly , the dependence of the flow rate on the density holds not only for the mean values of these quantities , but also for their temporal fluctuations . indeed , as exposed in the supplementary information ( supplementary fig .
s4 ) , with 95% confidence , flow rate fluctuations are positively correlated with density fluctuations in the vast majority of the experiments , provided that one considers moving averages of these fluctuations over time windows @xmath78 . on average , the flow rate is related to the mean time lapse @xmath79 between successive egresses via @xmath80 .
but , since fluctuations are large , it is actually worth considering the full distribution @xmath81 of time lapses . figure [ fig : time_lapses ] reveals substantial differences in these distributions when the crowd behaviour is varied .
evacuations with a mostly polite crowd display a relatively sharp peak at a characteristic time @xmath82 slightly below one second , and a depleted region at low @xmath83 .
interestingly , these features are also observed in uncontrolled collective egresses from conference rooms through doorways of similar widths , as shown in fig .
[ fig : time_lapses]e - f ( these uncontrolled egresses are described in the supplementary information ) . on the other hand , competitive egresses at large @xmath33 ( in our controlled experiments ) do not present a very well defined peak and the most frequent time lapses are shifted to smaller values .
most conspicuously , they include a non - negligible fraction of very short time lapses @xmath84 , which correspond to quasi - simultaneous egresses .
focusing on the large values of @xmath83 , caused by congestion , the data corresponding to our most competitive evacuation ( @xmath36 , purposeful ) seem to support garcimartin et al.s@xcite claim of a slow decay of @xmath81 at large @xmath83 , characterised by a power - law - like tail : we find a higher likelihood for the tail to be power - law - like than exponential ( @xmath85 for a continuous data set , according to the method of clauset et al.@xcite implemented in citealstott2014powerlaw ) .
in other situations , our data are insufficient to validate this claim , and even tend to contradict it in the most polite evacuations . between egresses for different evacuations .
the first four panels correspond to experiments for ( a ) @xmath86(placid ) , ( b ) @xmath87(placid ) , ( c ) @xmath35(purposeful ) , ( d ) @xmath36(purposeful ) .
the last two panels correspond to uncontrolled egresses ( see the text ) from ( e ) a conference room during a three - day congress and ( f ) the auditorium of cab at the end of the weekly seminars . ]
the existence of short time lapses @xmath83 hints at bursty dynamics . here
, passages through the door will belong to a burst of escapes if they are separated by less than 60% of the mean time lapse of the considered experiment , viz .
, @xmath88 ( the results are obviously robust to small variations of this criterion ) .
thus defined , bursts do not correspond to intervals of fast flow separated by long clogs ( only short - lived clogs were observed here ) , but to almost simultaneous escapes .
the distribution @xmath89 of burst sizes is plotted in fig .
[ fig : burst_sizes ] . as expected , evacuations of polite crowds consist of isolated escapes , with a small proportion of bursts of @xmath90 pedestrians or more ( @xmath91 , with probability @xmath92 ) , whereas bursts of up to @xmath93 pedestrians are observed with more selfish and competitive crowds .
our data are perfectly compatible with the ( fast ) exponential decay of @xmath89 reported in @xcite , but are insufficient to be assertive in this regard .
incidentally , note that , in all but one experiment ( @xmath94(purposeful ) ) , the average fraction of selfish agents in a burst increases with the size of the burst .
this is particularly noticeable in the `` placid '' experiments .
however , collecting statistics over the entire evacuation , we have not detected substantial temporal correlations in the behaviours of successively egressing participants .
( i.e. , egresses coming in rapid succession ) in the evacuations with purposeful walkers . ] although successions of @xmath91 short time lapses exist ( bursts of size @xmath91 ) , statistically anti - correlations between successive @xmath83 prevail by far , in all experiments but two ( @xmath95(placid ) and , to a lesser extent , @xmath96(placid ) ) .
this is apparent in the negative dip in the correlation functions of successive time lapses @xmath97 ( ordered by the rank of egress ) , @xmath98 where the brackets denote an average over all pedestrians @xmath99 in the experiment , an example of which is shown in fig .
[ fig : delta_t_correl ] .
this means that , in general , there is an alternation between shorter time lapses and longer ones .
such an alternation has already been observed in cooperative pedestrian flows at the entrance of ( long ) bottlenecks @xcite .
it was then ascribed to a `` zipper effect '' , whereby pedestrian lanes form in the bottleneck and need to be intercalated , because the bottleneck width does not allow pedestrians from different lanes to stand shoulder to shoulder . within each lane ,
the headway that pedestrians maintain with respect to the walker just in front of them ( of order one second in @xcite ) imposes a finite minimal distance between the pedestrians .
but this does not apply across lanes , so that a pedestrian may come close to contact with a walker from a neighbouring lane , without being able overtake this neighbour because of the constrained lateral space .
however , in our experiments , strong anti - correlations are also seen in very competitive and disorderly evacuations , for instance at @xmath36(purposeful ) , in which no lanes whatsoever can form . to explain this , we believe that the `` zipper '' mechanism can be generalised : if some free space is available right in front of them , pedestrians will step forward and try to cross the door at ( almost ) the same time as their predecessor who comes from another direction . but they will not risk this manoeuvre if there is not sufficient headway just in front of them , in particular if a competition for the exit already blocks the door . to bolster our claim , let us study the pedestrians angles of incidence @xmath100 into the exit zone , the angle @xmath101 corresponding to normal incidence ( `` at the centre '' ) .
the distributions of @xmath100 are approximately flat over an interval of the form @xmath102 $ ] , where the maximal angle @xmath103 seems to increase with the competitiveness of the evacuation .
more relevantly , we remark that the directions @xmath100 exhibit strong anti - correlations in time , pointing to the prevalence of alternations between small and large @xmath100 .
this is striking , for instance , in the autocorrelation function , @xmath104 plotted in fig .
[ fig : delta_t_correl](right ) for @xmath36(purposeful ) , where the negative dip for successively egressing pedestrians ( @xmath105 ) lies more than @xmath106 below 0 , where @xmath30 is the standard deviation of the autocorrelations of @xmath100 ( whose average gives @xmath107 ) and @xmath31 is the number of points used to compute the average .
moreover , if pedestrians @xmath38 and @xmath108 egress in fast succession ( @xmath109 ) , their angles of incidence differ more on average than if there is more delay between their egresses ( @xmath110 ) in strictly _ all _ experiments , with the most marked contrasts for polite crowds ( e.g. , @xmath111 vs. @xmath112 for @xmath86(purposeful ) ) .
this confirms that the short time lapses correspond to a participant from _ another _ direction passing through the door very shortly after the previous one .
conversely , the headway left behind a participant coming from the same direction is reflected by the fact that one systematically has to wait longer ( on average and in median ) to see another participant egress from the same direction ( within @xmath113 ) than from a randomly selected direction in the empirical distribution of @xmath100 .
the foregoing discussion implies that time lapses @xmath83 longer than the median value are not due to slower , lazier , or more patient pedestrians , but result from the organisation of the constricted flow .
we should mention that the alternations between short and long @xmath83 are absent from our data sets for uncontrolled egresses , where people generally passed through the door one by one , in an orderly and non - competitive way , sometimes even starting to align beforehand . and
@xmath114 , averaged over @xmath99 ( left ) and correlations between the angles of incidence @xmath100 of a pedestrian and his or her @xmath115-th successor ( right ) .
the error bars represent 95% confidence intervals . ]
so far we have focused on the time lapses between egresses .
but , from the individual pedestrian s perspective , a possibly more relevant duration is the time they have to wait in the above - defined exit zone , of area @xmath116 , before they can egress .
let us call this dwell time ( or waiting time ) @xmath117 .
its mean value @xmath118 is related to the mean time lapse via @xmath119 ; the derivation is presented as supplementary information . in the exit zone in evacuations with purposeful walkers .
normalised histograms of @xmath117 for @xmath86(purposeful ) ( left ) and @xmath36(purposeful ) ( middle ) .
the right panel shows the average dwell time for selfish , purposeful pedestrians as a function of the number @xmath120 of selfish participants in the zone , for distinct numbers of polite participants .
the error bars are 95% confidence intervals . ] sharply peaked distributions @xmath121 correspond to orderly evacuations , with approximately equal waiting times for all .
this case is exemplified in fig .
[ fig : t_w](left ) , for @xmath86(purposeful ) . on the contrary , broad @xmath121
reveal the presence of heterogeneity and/or disorder . for @xmath36(purposeful ) , since the crowd is behaviourally homogeneous , the breadth of @xmath121 ( fig .
[ fig : t_w](right ) ) has to be ascribed to disorder . on the other hand , for inhomogeneous crowds , the overall distribution @xmath121 mingles the two distinct types of pedestrians . to refine the analysis , we study the histograms of @xmath117 as a function of the pedestrian s behaviour , namely , whether they are polite agents ( pola ) or selfish moving agents ( selma ) , and their direction of incidence into the exit zone : from the left ( @xmath122 ) , from the centre ( @xmath123 ) , or from the right ( @xmath124 ) .
an example of such a histogram is shown in fig .
[ fig : t_w_dist_direc ] . beyond the statistical noise
, we see that , in all cases , the distributions are peaked at a value around @xmath125 and , as expected , pola s distributions are shifted to larger @xmath117 , and stretched , compared to selma s . for either type of behaviour , the distributions look relatively similar for all incident directions,(purposeful ) , where the average waiting time differs by more than @xmath126
. this could be due to statistical noise , individual preferences for one side , or a slight , unnoticed asymmetry in the setup . ] with a noticeable exception : polas display more frequent outliers at large @xmath117 if they come from the sides than if they come from the centre . indeed
, pola tends to be swept along by the pedestrian stream if she comes from the centre , whereas she can be blocked at the sides of the door otherwise .
there is no such a marked effect for selma , probably because the latter tries her hardest , and manages , to egress rapidly in any event . .
the stars indicate the mean value in each case . ]
how do the @xmath117 depend on the population in the exit zone ( averaged over the pedestrian s dwell time ) ? figure [ fig : t_w](right ) evinces that , in the purposeful experiments , selma s mean waiting time very generally grows with the number of selfish agents in the zone , for any fixed number of polite ones , and also increases when there are more polite participants , for a fixed number of selfish ones .
these features hold true in the placid experiments ( supplementary fig .
s6 ) , with the exception that @xmath127 seems to shorten in the competitive situations where 3 participants are simultaneously in the zone , perhaps due to a sampling bias . on the whole
, these observations also apply for pola s waiting times , but the picture is less clear . the positive correlation between the waiting time and the occupancy of the zone is intuitive , but not entirely trivial , because escapes come in faster succession if the zone is more crowded .
the aforementioned correlations are only valid in a statistical sense .
we have not found any law that can accurately predict @xmath117 as a function of the local parameters .
nevertheless , can some generally valid `` priority rules '' for the flow be unveiled ? making a list of the pedestrians that an agent of a given type yields to , that is to say , the pedestrians that entered the zone later than this agent but egressed before , we find that overall selma yields to fewer pedestrians than pola and that , for both types , it is more common to yield to selfish pedestrians than to polite ones , which is not so surprising .
it should however be stressed that this tendency is not systematic : polite agents ( possibly carried along by the pedestrian flow ) also overtake their selfish counterparts from time to time . on the other hand , in virtually no instance
does selma yield to a polite pedestrian coming from the same direction ( left , centre or right ) .
this rule only suffers two exceptions over the whole set of experiments . clearly , these `` priority rules '' are limited in scope and can not tell us how orderly the flow is . to quantify disorder ,
we choose to compute the average difference @xmath128 between the order of entrance @xmath129 of pedestrian _ i _ in the zone and their order of egress @xmath130 , viz .
, @xmath131 .
the disorder estimator @xmath128 increases with @xmath2 ( see supplementary fig .
s7 ) , but also with @xmath33 ; which of these two parameters is best correlated with @xmath128 is not obvious .
but it is noteworthy that the crowd at @xmath35 has a larger @xmath128 value than that at @xmath36 , which points to the enhancement of disorder due to behavioural heterogeneity . besides disorder ,
another practically relevant feature of the evacuation is the pressure exerted applied on the participants , insofar as suffocation and compressive asphyxia , originating in the compression of the lungs , have been reported as a cause of death in major crowd disasters @xcite .
accordingly , after each experiment , we asked a dozen randomly picked participants to rate the level of mechanical pressure that they experienced during the evacuation from 1 to 10 , for want of more objective measurements of pressure .
the perceived pressure follows a clear trend , more precisely , a monotonic increase with the density @xmath2 in the exit zone ( and thus also with @xmath33 ) , as shown in supplementary fig .
s8 ) . therefore ,
competitive evacuations , associated with larger densities at the door , involve more disorder and higher pressure than their more cooperative counterparts .
the most competitive ones even feature signs of incipient turbulence , in the form of `` vortices '' at the exit , i.e. , participants spinning around a full 360@xmath132 when egressing ( see table [ tab : global ] and videos in the supplemental materials ) .
in summary , we have performed evacuation experiments through a 72-cm - wide door , in which a ( variable ) fraction @xmath33 of the participants were asked to behave selfishly , while the rest behaved politely . irrespective of these behaviours , a first series of experiments was conducted with placid walkers and a second one with purposeful walkers . by re - injecting
egressing participants into the room , we managed to improve the statistics . furthermore , despite instantaneous fluctuations , the pedestrian flow was found to be quasi - stationary at the macroscopic scale , unlike in other experiments @xcite .
the flow rate @xmath1 gets higher for larger @xmath33 and with more purposeful walkers .
the absence of a ` faster - is - slower ' effect does not contradict its possible occurrence effect@xcite when considering crowds above a competitiveness threshold .
irrespective of the behaviours , the flow rate and other flow properties such as the disorder in the passages and the pressure perceived by the participants exhibit a simple dependence on the density @xmath2 in the exit zone .
thus , our results suggest that in a coarse macroscopic approach , and for a given composition of the crowd , the behavioural aspects can be left aside in favour of the density @xmath2 ; this confirms the key role played by the latter in determining the flow rate @xcite .
somewhat surprisingly , @xmath1 was observed to grow monotonically with @xmath2 up to close - packing ( @xmath3 ) , contrary to most fundamental diagrams in the literature @xcite .
in fact , we believe that at high densities the global flow rate will strongly depend on other parameters , in particular , the pressure in the crowd , thus , to what extent people are pushing their neighbours . here , the participants were not allowed to push . yet
, the evacuation dynamics clearly became more disorderly with increasing @xmath33 and increasing purposefulness .
this disorder was notably reflected in the distribution @xmath81 of time lapses @xmath83 between successive escapes : @xmath81 displays a relatively sharp peak at @xmath133 of order @xmath134 for polite crowds . in more competitive evacuations the low-@xmath83 region
gets populated ; bursts of quasi - simultaneous escapes occur and seem to be exponentially distributed in size . in the most competitive evacuations , some signs of incipient flow turbulence
were even detected , e.g. , pedestrian `` vortices '' .
shorter and longer time lapses @xmath83 were found to alternate between successive escapes in almost all experiments .
we explained this in terms of a generalised zipper effect , whereby pedestrians strive to keep a finite headway behind an agent walking in the same direction , while coming close to contact with those from another direction .
this idea is supported by the marked anti - correlations in the angles of incidence at the door . finally , we investigated the pedestrians dwell time @xmath117 in the exit zone .
the mean dwell time clearly increases with the occupancy of the zone , especially for selfish agents .
not surprisingly , @xmath117 is shorter and more narrowly distributed for selfish agents . also , polite pedestrians coming from the sides display more outliers at large @xmath117 . several of the effects observed in our `` microscopic '' study of the evacuation dynamics are rather intuitive .
but their quantitative characterisation opens the door ( so to speak ) to more thorough tests of evacuation models in constrained settings , as in ref .
@xcite , and , accordingly , more precise understanding of the process .
our study may also prompt the idea that the effect of the pedestrians behaviours on the evacuation dynamics is amenable to a simple description .
the authors are grateful to the grupo de higiene y seguridad of cab for their help in devising safe evacuation experiments , to all members of the fiestin group , in particular to guillermo abramson and damin zanette , to alejandro kolton and pablo gleiser for lending us their cameras , and to all the voluntary participants .
thanks iker zuriguel for a useful discussion .
this work received partial funding from conicet ( under grant no .
pip 11220110100310 ) and cnea , both argentinian agencies .
a.n . , s.b . , and m.k .
conceived and conducted the experiments .
analysed the results and wrote the paper .
all authors discussed the analysis and reviewed the manuscript .
( s.b . and m.k .
contributed equally to this work . )
the authors declare no competing financial interests .
t. kretz , a. grnebohm , and m. schreckenberg , `` experimental study of pedestrian flow through a bottleneck , '' _ journal of statistical mechanics : theory and experiment _ , vol .
2006 , no . 10 , p. p10014 , 2006 .
j. zhang , w. klingsch , a. schadschneider , and a. seyfried , `` transitions in pedestrian fundamental diagrams of straight corridors and t - junctions , '' _ journal of statistical mechanics : theory and experiment _ , vol .
. 06 , p. p06004 , 2011 . v. predtechenskii and a. i. milinski , _
planning for foot traffic flow in buildings_. national bureau of standards , us department of commerce , and the national science foundation , washington , dc , 1978 .
a. seyfried , m. boltes , j. khler , w. klingsch , a. portz , t. rupprecht , a. schadschneider , b. steffen , and a. winkens , `` enhanced empirical data for the fundamental diagram and the flow through bottlenecks , '' in _ pedestrian and evacuation dynamics 2008 _ , pp . 145156 , springer , 2010 .
j. liddle , a. seyfried , w. klingsch , t. rupprecht , a. schadschneider , and a. winkens , `` an experimental study of pedestrian congestions : influence of bottleneck width and length , '' _ arxiv preprint arxiv:0911.4350 _ , 2009 .
j. liddle , a. seyfried , b. steffen , w. klingsch , t. rupprecht , a. winkens , and m. boltes , `` microscopic insights into pedestrian motion through a bottleneck , resolving spatial and temporal variations , '' _ arxiv preprint arxiv:1105.1532 _ , 2011 .
a. seyfried , b. steffen , a. winkens , t. rupprecht , m. boltes , and w. klingsch , `` empirical data for pedestrian flow through bottlenecks , '' in _ traffic and granular flow07 _ , pp . 189199 , springer , 2009 .
a. garcimartn , d. parisi , j. pastor , c. martn - gmez , and i. zuriguel , `` flow of pedestrians through narrow doors with different competitiveness , '' _ journal of statistical mechanics : theory and experiment _ , vol .
2016 , no . 4 , p. 043402 , 2016 .
i. zuriguel , d. r. parisi , r. c. hidalgo , c. lozano , a. janda , p. a. gago , j. p. peralta , l. m. ferrer , l. a. pugnaloni , e. clment , _
_ , `` clogging transition of many - particle systems flowing through bottlenecks , '' _ scientific reports _ , vol
. 4 , 2014 . n. shiwakoti , m. sarvi , g. rose , and m. burd , `` animal dynamics based approach for modeling pedestrian crowd egress under panic conditions , '' _ transportation research part b : methodological _ , vol .
45 , no . 9 , pp . 14331449 , 2011 .
a. garcimartn , j. pastor , l. m. ferrer , j. ramos , c. martn - gmez , and i. zuriguel , `` flow and clogging of a sheep herd passing through a bottleneck , '' _ physical review e _ , vol .
91 , no . 2 , p. 022808
, 2015 .
i. zuriguel , j. olivares , j. m. pastor , c. martn - gmez , l. m. ferrer , j. j. ramos , and a. garcimartn , `` effect of obstacle position in the flow of sheep through a narrow door , '' _ physical review e _
94 , no . 3 , p. 032302
, 2016 .
n. kugihara , `` effects of aggressive behaviour and group size on collective escape in an emergency : a test between a social identity model and deindividuation theory , '' _ british journal of social psychology _ , vol .
40 , no . 4 , pp .
575598 , 2001 .
h. c. muir , d. m. bottomley , and c. marrison , `` effects of motivation and cabin configuration on emergency aircraft evacuation behavior and rates of egress , '' _ the international journal of aviation psychology _ , vol . 6 , no . 1 ,
pp . 5777 , 1996 . j. m. pastor , a. garcimartn , p. a. gago , j. p. peralta , c. martn - gmez , l. m. ferrer , d. maza
, d. r. parisi , l. a. pugnaloni , and i. zuriguel , `` experimental proof of faster - is - slower in systems of frictional particles flowing through constrictions , '' _ physical review e _
92 , no . 6 , p. 062817
, 2015 .
m. moussad , m. kapadia , t. thrash , r. w. sumner , m. gross , d. helbing , and c. hlscher , `` crowd behaviour during high - stress evacuations in an immersive virtual environment , '' _ journal of the royal society interface _ , vol . 13 , no . 122 , p. 20160414
, 2016 .
d. helbing , l. buzna , a. johansson , and t. werner , `` self - organized pedestrian crowd dynamics : experiments , simulations , and design solutions , '' _ transportation science _ , vol .
39 , no . 1 ,
pp . 124 , 2005 .
a. nicolas , s. bouzat , and m. kuperman , `` influence of selfish and polite behaviours on a pedestrian evacuation through a narrow exit : a quantitative characterisation , '' _ arxiv preprint arxiv:1608.04863 _ ,
2016 .
t. robin , g. antonini , m. bierlaire , and j. cruz , `` specification , estimation and validation of a pedestrian walking behavior model , '' _ transportation research part b : methodological _ , vol .
43 , no . 1 ,
pp . 3656 , 2009 . | we study the dynamics of pedestrian evacuations through a narrow doorway by means of controlled experiments .
the influence of the pedestrians behaviours is investigated by prescribing a selfish attitude to a fraction @xmath0 of the participants , while the others behave politely .
thanks to an original setup enabling the re - injection of egressed participants into the room , the analysis is conducted in a ( macroscopically ) quasi - stationary regime . we find that , as @xmath0 is increased , the flow rate @xmath1 rises , interpolating between published values for egresses in normal conditions and measurements for competitive evacuations .
the dependence of several flow properties on the pedestrian density @xmath2 at the door , independently of @xmath0 , suggests that macroscopically the behavioural aspects could be subsumed under the density , at least in our specific settings with limited crowd pressure . in particular , under these conditions
, @xmath1 grows monotonically with @xmath2 up to `` close - packing '' ( @xmath3 ) .
the flow is then characterised microscopically . among other quantities ,
the time lapses between successive escapes , the pedestrians waiting times in front of the door , and their angles of incidence are analysed statistically . in a nutshell , our main results show that the flow is orderly for polite crowds , with narrowly distributed time lapses between egresses , while for larger @xmath0 the flow gets disorderly and vanishing time lapses emerge . for all @xmath0
, we find an alternation between short and long time lapses , which we ascribe to a generalised zipper effect .
the average waiting time in the exit zone increases with its occupancy . the disorder in the flow and the pressure felt by participants
are also assessed . _
keywords : _ pedestrian evacuation ; behaviours ; exit capacity ; bottleneck flow @xmath4 lptms , cnrs & univ .
paris - sud , f-91405 orsay , france @xmath5 conicet & centro atmico bariloche , bariloche , 8400 , argentina _ _ |
active materials composed of motile bodies define a quickly - growing field of statistical and soft - matter physics @xcite . over the last decades
, much attention has been devoted to the individual dynamics of self - propelled particles ( e.g. effective diffusion and migration in an external field ) , and to the collective properties of large populations ( e.g. transition to collective motion , and emergence of coherent spatial patterns ) @xcite . from a theoretical perspective , the large - scale properties of active populations have been investigated for several interaction schemes at the single - particle level , see e.g. @xcite . in most of the models , dynamical rules such as velocity - alignment or hard - core repulsion were included without refereeing to the microscopic physics responsible for these couplings . a priori a number of alternative phenomenological rules could be considered , yet no global framework exists to elaborate and classify such interactions with overlooked symmetries . from an experimental perspective , significant progress has been made over the last years , and a number of artificial active systems are now available , including self - propelled colloids @xcite , vibrating grains @xcite , biofilaments @xcite .
now that the fabrication of motile microscopic systems is a problem that has been partly solved , a natural next step is to consider the self - assembly of these autonomous units into new materials .
the design of such active phases requires a deeper understanding of the interaction symmetries between their elementary units .
surprisingly , until now , the two - body interactions between self - propelled colloids and filaments have been scarcely characterized . in this paper , we classify the symmetries of the mutual interactions between self - propelled bodies moving at a constant speed .
we first decompose systematically the two - body force fields in a generalized multipole expansion which solely requires that the particles live in a homogeneous space ( translational invariance ) .
we then show how little additional information about the interaction process further simplify the form of the interactions .
we consider explicitly three relevant cases : ( i ) isotropic particles interacting only via potential interactions , ( ii ) isotropic particles interacting via two - body inelastic collisions , and ( iii ) particles of arbitrary shape that interact via non - reciprocal interactions ( _ viz .
_ interactions that are non invariant upon galilean transformations ) .
we systematically exemplify our results with classical active - matter models , which we rationalize within a unified framework .
we close this paper by suggesting guidelines for the design of new active materials .
by definition , self - propelled particles convert stored internal energy to propel themselves in the absence of any external force @xcite .
we consider here the simplest framework supporting this definition : point particles moving at a _ constant _ speed @xmath0 in a homogeneous space .
more precisely , any force acting on the particle @xmath1 located at @xmath2 and moving at a velocity @xmath3 alters only its direction of motion , defined by the unit vector @xmath4 .
we also assume that the particles interact via pairwise - additive couplings .
the equations of motion for the particle @xmath1 are then the newton s equations completed by the constant velocity constraint : @xmath5 where @xmath6 is the force exerted by particle @xmath7 on particle @xmath1 , and where we have set @xmath8 .
the projection operator @xmath9 ensures that the norm of the velocity vector @xmath4 remains constant .
implies that particles having a constant speed reorient and align along the net force that they experience .
these equations , that have been extensively studied @xcite , can also be viewed as the asymptotic limit of a broader class of active - particle dynamics , see e.g. @xcite .
our predictions do not strictly require the propulsion speed to be a constant , but only that some internal dissipation mechanisms causes @xmath0 to relax towards its stationary value in a time much shorter than the changes in the velocity direction .
we further clarify this point and discuss the robustness of eqs . and in appendix a , by studying a specific energy - exchange model @xcite .
we now establish a generic expression which classifies the interactions according to their angular symmetries . for sake of clarity , we restrict the discussion to two spatial dimensions ; the generalization to a 3d system is provided in appendix b. the force exerted by particle @xmath7 on particle @xmath1 is a priori a function of the two positions @xmath2 , @xmath10 and orientations @xmath4 , @xmath11 . however , by definition , in a homogenous medium @xmath12 is translationally invariant , and therefore only depend on the relative positions of the particles : @xmath13 .
we now identify 2d vectors to complex numbers , and note @xmath14 , where @xmath15 is the interparticle distance and @xmath16 the relative angular position . without any additional assumption , we fourier transform @xmath6 with respect to @xmath16 and obtain : @xmath17}. \end{aligned}\ ] ] transforming the real and imaginary parts into polynomial series in @xmath18 and @xmath19 , and after some elementary algebra , eq .
is recast into a more intuitive expansion akin to a multipolar series : @xmath20 @xmath21 and @xmath22 are scalars , @xmath23 is the completely antisymmetric levy - civita symbol , and all the other coefficients are vectors : @xmath24 , @xmath25 and @xmath26 .
we recall that all the @xmath27 depend in principles on the relative position and on the orientations of particles @xmath1 and @xmath7 .
the terms in eq . are classified according to their angular periodicity , the index @xmath28 corresponds to the topological charge of the force field induced by the particle @xmath7 : the larger @xmath28 , the faster the angular variations of the field .
is already a pivotal result of this paper as it describes all the possible symmetries of the field in which a self - propelled particle reorients . in order to gain more physical insight into this formal expansion , the first modes are plotted in fig .
[ fields2d ] , and discussed below . *
the @xmath29 component is a field having a constant orientation .
the particle @xmath1 reorients along @xmath30 , which is aligned with the vector @xmath31 regardless of the relative positions of the particles . * the @xmath32 component in eq
. generically corresponds to a spiral force field . to better understand its contribution , it is convenient to distinguish between its divergence and rotational components as done in eq .
[ tensor ] and fig .
1 . the first term , @xmath33 , has the symmetry of a monopolar field .
this term therefore gives rise to repulsive ( resp .
attractive ) couplings .
for example , particle @xmath1 reorients along @xmath34 if @xmath35 . the second term , the rotational part @xmath36 , breaks the bottom - top symmetry in 2d , which implies that @xmath37 is a pseudoscalar quantity .
consequently , @xmath37 is non - zero only for particles having some chiral features . in such an interaction field
, a chiral particle @xmath7 could be forced to circle around its neighbor . *
the @xmath38 component has the form of an hyperbolic elongation field with a negative topological charge .
the @xmath39 contribution has the dipolar symmetry , the direction of the dipole being set by the vector @xmath40 .
similarly , the higher - order terms of the fourier expansion correspond to force fields having positive or negative topological charges . note that the contributions from the @xmath41 components do not correspond to conventional multipoles associated with a laplacian field .
we have classified all the possible symmetries of the fields that cause the particle reorientation . in order to fully prescribe the orientational dynamics , we now have to specify how the @xmath42s relate to the particle orientations and relative distance . in order to do so
, we focus on three types of interactions which encompass most of the numerical and experimental systems .
we first consider the simplest possible setup : isotropic particles interacting via potential interactions .
@xmath43 derives from a potential @xmath44 which only depends on the interparticle distance .
consequently , @xmath30 readily reduces to the sole curl - free part of the mode @xmath32 in eq . .
the force field has the symmetry of a monopole , with a strength @xmath45 , it results in attractive or repulsive couplings .
such potential interactions have been studied in a number of numerical models , see e.g. @xcite .
for instance , dorsogna _ et al .
_ demonstrated that forces deriving from a morse potential can lead to the emergence of a number of patterns all having an a rotational symmetry , such as vortices , rings and circular clumps @xcite .
we stress that potential forces can not explicitly couple the velocities of isotropic particles .
however we will show in sect .
[ non - reciprocal ] , and in appendix c , that potential forces can yield net alignment interactions between slender bodies .
we now turn to a more general situation and assume that the particles undergo physical two - body collisions , where @xmath46 could also be associated to a dissipative process .
however , we still restrain here to the situation where @xmath46 is invariant upon galilean transformations . in other words
, @xmath46 is assumed to only depend on the relative position , @xmath47 , and on the relative orientation / velocity , @xmath48 .
the coefficients of the fourier series hence depend only on @xmath48 , and the vector coefficients @xmath49 in eq .
are all oriented along @xmath48 : @xmath50 it follows from eq .
that @xmath51 .
even though the self - propulsion mechanism does not conserve momentum , see eq .
, the invariance of the forces upon galilean transformations implies that they obey the newton s third law .
this situation typically corresponds to the model for vibrated polar disks introduced in @xcite .
et al . _ studied numerically a system of polar grains set in motion by the vibration of a substrate and interacting via short - range inelastic collisions @xcite .
they model the interactions by the spring - and - dash - pot model and assume that @xmath52\hat{{\mathbf{r}}}_{ab}+ \kappa ( d - r_{ab } ) \hat{{\mathbf{r}}}_{ab}$ ] .
the first term is associated with viscous friction , and the last term corresponds to elastic repulsion for @xmath53 .
they observed numerically that such inelastic collisions can lead to velocity alignment interactions , thereby giving rise to a macroscopic polar order . in the classification given by eq .
, the force corresponds to a combination of the three components @xmath29 , 1 and 2 , where @xmath54 and @xmath55 . beyond this specify model which beautifully accounts for experimental results ,
the present framework makes it possible to provide a clear design rule for collision - induced velocity alignment at the two - particle level .
the lowest - order mode of @xmath46 , @xmath56 , is the only one that promotes a net polar alignment regardless of the two particles conformation provided that @xmath57 ( @xmath58 would result in anti - alignement interactions ) .
particles interacting via @xmath59 evolve according to @xmath60 consequently , any other contribution to the force expansion , eq . , would compete with this velocity - alignment rule and , in principle , could yield macroscopic states with more complex symmetries that a mere polar phase .
a an asymmetric dumbbell reorients in the field . b an isotropic particle carrying a polar internal structure can also align its velocity with the field . ]
we now consider forces that are not necessarily invariant upon galilean transformations , and therefore relax the reciprocal condition .
@xmath46 can now be different from @xmath61 .
prominent instances are : ( i ) hydrodynamic interactions between simmers @xcite : when swimming in a liquid , particle @xmath7 creates a flow field that causes particle @xmath7 to reorient in this flow , thereby yielding effective long - range interactions .
( ii ) polar phoretic particles respond to the local variations of a scalar quantity ( chemical potential @xcite , temperature @xcite ) .
they are also prone to align in the phoretic field induced by their neighbors @xcite .
more generally , we here describe particles that reorient in a field originated from its neighbors .
note that this broader class encompasses the potential interactions that we first considered .
the field in which the particle aligns would precisely correspond to the gradient of the potential . in all that follows , we denote by @xmath62 the field created by particle @xmath7 at position @xmath63 .
we assume that @xmath64 depends on time only via the particles conformation ( this hypothesis corresponds to a zero - peclet number approximation in the context of phoretic particles ) and that it does not dependent on @xmath65 .
this approximation is well - suited in a far - field description , since at long distance the orientation of particle @xmath1 does not modify the field induced by particle @xmath7 .
we now have to specify how the particle reorients in the external field @xmath64 to fully define the expression of @xmath46 . in view of gaining more physical intuition ,
we first introduce a prototypical dumbbell model that can be simply solved .
we consider dumbbells made of two rigidly connected disks separated by a distance @xmath66 , fig
looking at the motion of the dumbbell @xmath1 , we denote by @xmath67 and @xmath68 the positions of the two disks .
we assume that the particle propels along its main axis : @xmath69 .
the dynamics of the dumbbell is then modeled as follows .
both disks respond linearly to the field @xmath70 ) exerted by particle @xmath7 , where @xmath71 , @xmath72 .
as thoroughly demonstrated in appendix c , the equation of motion of the dumbbell orientation then reduces to : @xmath73,\end{aligned}\ ] ] where @xmath74 and @xmath75 are two constant scalar quantities .
this equation correspond to eq .
with @xmath76 given by : @xmath77 we also show in appendix c that @xmath74 can either be a positive or a negative quantity depending of the relative mobility coefficients of the two disks .
hence two different behaviors are obtained .
if the two disks are _ not _ identical ( polar dumbbell ) , @xmath78 and the force is @xmath79 .
the dumbbell aligns with , or opposite to @xmath80 .
conversely , if the two disks are identical , then @xmath81 and the first term in eq . vanishes .
the force then reduces to @xmath82 , and the dumbbell aligns nematically in a direction set by the field gradient .
this minimalist setup already shows that multiple field alignment rules can exist .
we now go beyond this specific picture and discuss more generally the polar and nematic cases .
we first assume polar alignment in the direction of the field , which , regardless of the shape of the particle , translates into : @xmath83 where @xmath74 is here a phenomenological coefficient .
the force is independent of @xmath4 . as a consequence , the coefficients of the expansion in eq .
only depend on @xmath15 and @xmath11 , and the scalar coefficients @xmath84 and @xmath37 solely depend on the interparticle distance .
in addition , the vector coefficients @xmath49 are necessarily oriented along @xmath11 : @xmath85 we illustrate this result with three concrete examples .
( i ) this situation has been considered in the context of interacting polar swimmers in confinement @xcite , and of the motion of biofilament in plant cytoskeleton @xcite . for confined swimmers , eq .
reduces to the sole dipolar term ( @xmath86 ) which reflects the potential flow induced by any type of self - propelled object in a rigidly confined liquid film . in a different context , kumar _
et al . _ have demonstrated experimentally and numerically that pointy rods lying on a shaken bed of isotropic grains experience self - propulsion and velocity alignment interactions @xcite .
the rods interact effectively as they align locally with the polar displacement field induced by the motion of their neighbors on top of the shaken bead layer .
( ii ) this type of coupling is also responsible for the emergence of flocking patterns and of spontaneously flowing phases in the ensembles of colloidal rollers introduced in @xcite .
the velocity of a roller aligns in the direction of the flow field induced by the surrounding motile colloids . as show in @xcite ( supplementary materials ) , within a far - field approximation , the roller - roller interaction combines the first terms of eq . , @xmath29 , 1 and 2 .
we emphasize that the colloidal rollers have a perfectly isotropic shape , however their velocity very quickly relaxes in the direction of their dipolar electric - charge distribution .
this polar internal degree of freedom is responsible for the polar - alignement rule in the external field , see fig .
( iii ) finally this type of coupling also encompasses the interactions used in agent - based models for collective motion , provided that they involve pairwise - additive interactions .
in all these sequels of the seminal vicsek model @xcite , see e.g. @xcite , the interactions exactly correspond to the first fourier mode , @xmath87 . on top of this velocity - alignment rule , short - range repulsion and long - range attraction have also been considered to reproduce the morphology of cohesive flocks akin to animal populations @xcite .
they are associated with the @xmath32 component of the multipole expansion . in line with the previous discussion , and with eq .
, some motile particle , say @xmath1 , can also align nematically with the field @xmath88 induced by particle @xmath7 .
this condition translates into the following generic expression for the induced force : @xmath89 were @xmath90 is a tensor that can be constructed from the field @xmath64 and the @xmath91-operator , depending on the type of induced field and particle shape we are talking about .
for instance , for the symmetric dumbbells @xmath90 takes the simple form : @xmath92 , see eq . .
more generally , when the nematic - alignment rule arises from the slenderness of the particle , the expression of the tensor @xmath90 is given by the so - called jeffery s equation first introduced in the context of fluid mechanics @xcite .
. implies that the force @xmath6 linearly depends on @xmath4 .
the amplitudes of the fourier coefficients , @xmath93 in eq .
, therefore include a factor @xmath94 or @xmath95 , and the vectors @xmath49 in eq .
are oriented along @xmath11 .
the k - coefficient of the classification in eq . takes the generic form : @xmath96\hat{{\mathbf{p}}}_b .\ ] ] an important example concerns the zeroth - order term : both the @xmath97 and the @xmath98 contributions promote nematic alignment between the two interacting particles @xmath1 and @xmath7 for this first lowest - order term .
we are now aware of any experimental realization , where self - propelled particles would experience nematic interactions mediated by an induced field .
for instance in the seminal pusher / puller model of saintillan et al
. active swimmer align nematically in the far - field flow induced by their motion .
however these hydrodynamic interactions only correspond to the @xmath99 mode in eq .
, thereby yielding complex spatiotemporal fluctuations in swimmer suspensions which can not support nematic order @xcite .
for particles having more complex shapes , the reorientation in the field in principle involves higher symmetries than the polar or nematic modes .
for instance , the force can include higher - order terms such as @xmath100 , where @xmath101 is a third - rank tensor build from the field @xmath64 and its derivatives .
following the same procedure , classifying the interactions associated with these higher - order couplings is straightforward .
building only on symmetry arguments , we have introduced a formal classification of all the possible interactions between self - propelled particles .
this classification does not only rationalize all the previously introduced models within a unique formal framework , it also brings two unanticipated perspectives on the physics of interacting motile bodies .
firstly , from a technical perspective , the classification defined by eq . makes it possible to quickly identify the salient features of the two - body dynamics , even if the interactions take a complex form .
for instance , it offers a simple mean to check wether the interactions promote velocity alignment ( polar or nematic ) . in practical terms , eqs . and imply that that @xmath102 supports a net alignment of the velocities only if its zero mode does not vanish .
equivalently , a simple criteria for two - body alignment is that the angular integral of the two - body force does not vanish . from eq . , it necessarily takes the simple form : @xmath103 where the @xmath4 term does not contribute to the orientational dynamics , see eq . . any nonzero value for @xmath104 yields alignment . to distinguish between interactions that promote polar or nematic ordering
, one should also look at the dependence on @xmath4 of @xmath104 . if @xmath104 is unchanged as the particle @xmath1 changes its orientation ( @xmath105 ) , then @xmath106 and leads to polar alignment in eq .
in contrast , if the sign of @xmath104 changes as particle @xmath1 changes its orientation ( @xmath105 ) , then @xmath46 leads to nematic alignment in eq . secondly ,
going back to the initial motivation of this work , the phase of an ensemble of interacting particles is set by the symmetries of the interactions at the microscopic level .
although this statement is obvious for equilibrium systems , it has been surprisingly overlooked when considering active matter . until now
, all the theoretical models and the ( quantitatively characterized ) experimental realizations of active matter involving physical interactions have been restricted to the @xmath107 modes of the classification .
eq . demonstrate that there exist a number of possible interactions rules between self - propelled particles that have not been considered at all , even though they should yield novel macroscopic phases of active matter
. a special attention should be devoted to the @xmath108 and @xmath109 modes of eq . .
all the kinetic theories of active matter confirmed that interactions with a low angular symmetry strongly contribute to the large - scale hydrodynamics of these systems .
these types of interactions are therefore expected to significantly contribute to the phase behavior of unanticipated active materials .
this work was partly funded by anr grant mitra and by the institut universitaire de france .
in this section , we further extend the range of validity of the equations of motion . let us consider the more general case where small fluctuations of the particle speed are allowed .
following e.g. @xcite , we model self - propulsion by a non - linear friction force and assume the following equations of motion : @xmath110 eq . is a momentum conservation equation . the first term on the r.h.s . models self - propulsion .
it describes the conversion of internal energy into translational motion , and accounts for dissipative friction forces . as a result
, the particle speed relaxes to @xmath111 .
in addition , @xmath112 is the total external force acting on particle @xmath1 .
we also include here a possible noise term @xmath113 with zero mean and variance @xmath114 .
for the sake of clarity we discard active noise , which would stem from fluctuations in the propulsion mechanism and could be easily included as well , see e.g. @xcite .
we focus on small fluctuations of the speed , i.e. we assume that @xmath115 and @xmath116 , where @xmath117 . it is worth noting that eq .
is the only possible expression for the propulsion force at first order in @xmath104 . for sake of simplicity
we restrain here to particles moving in a 2d space , as the generalization to 3d is straightforward . in order to obtain stochastic equations for the particle position and orientation
, we have to integrate out the speed fluctuations .
this is easily done by using polar coordinates .
we introduce @xmath118 , where @xmath119 , and we describe the particle orientation by the polar angle @xmath120 : @xmath121 .
similarly , we write the force as : @xmath122 . eqs . are recast into : @xmath123 where @xmath124 and @xmath125 are independent white noises with zero mean and variance @xmath126 .
the term @xmath127 in eq . is a spurious drift term that classically results from the stratonovich discretization scheme used here to define the noise terms , see e.g. @xcite .
involve two well separated time scales . in eq .
, the particle speed relaxes to @xmath111 in a time @xmath128 .
the force and noise terms only give rise to subdominant corrections .
conversely , in eq .
the particle orientation evolves on a much longer time scale @xmath129 .
indeed , self - propulsion corresponds to a spontaneous breaking of the rotational symmetry , the orientation is not constrained to relax toward a position set by any potential . only the external forces and noise dictate the orientational dynamics .
we now want to describe the particle dynamics on time scales much longer than @xmath130 .
in other words , we want to average the dynamics over the fast variations of the particle speed .
firstly , we integrate eq . and combine this result with eq . , which yields @xmath131 \hat{{\mathbf{p}}}_a(t)\ , { \mathrm{d}}t ' , \end{aligned}\ ] ] and @xmath132 where @xmath133 is a colored translational noise , with zero mean and correlations defined by @xmath134}$ ] .
the kernel @xmath135 is given by @xmath136 \ , \theta(t)$ ] , where @xmath137 is the heaviside step function .
when considering only time variations at scales much larger than @xmath130 , the time correlations of the translational noise vanish , @xmath138 and we recover a simple markovian kernel @xmath139 . coming back to vector notations for the particle orientation , the two coarse - grained equations of motion are : @xmath140 .\end{aligned}\ ] ] we recover eq . for the orientation dynamics .
the velocity of the particle , eq . , is the sum of the self - propulsion @xmath4 and additional advection terms due to the interactions and the noise .
however , these contributions correspond to small corrections as @xmath141 and @xmath127 were assumed to be of order @xmath104 . the equations of motion that we assumed in the main text therefore correspond to the limit @xmath142 of this more general model . in other words , eqs . and are valid when the relaxation of the speed towards @xmath111 is much faster than the modification of the particle s direction due to interactions and noise .
the 3d generalization of eq . is a spherical harmonics expansion , and contains a much larger number of terms .
however , most of them are discarded by the symmetries of the interactions . for potential interactions between isotropic particles ,
the force is radial . for galilean - invariant interactions ,
the force field depends on a unique vector , @xmath143 .
its expression is therefore invariant by rotation around the @xmath144-axis . for particles aligning in a field
, the force field also depends on a unique vector , @xmath145 , if the alignment rule is polar .
the case of particles aligning nematically is readily inferred from the polar case by adding a factor @xmath95 or @xmath94 in the expression of the force .
these symmetry requirements greatly simplify the spherical harmonics expansion . in an orthonormal basis @xmath146
, we find : @xmath147 , \end{aligned}\ ] ] where @xmath148 are the spherical harmonics , @xmath149 and @xmath150 . in vector notations ,
the above equation takes a much more compact form : @xmath151 the components @xmath152 correspond to a standard multipolar series , they have the symmetries of monopolar , dipolar , quadrupolar fields etc . the coefficients @xmath153 and @xmath154 are associated with elongational and rotational fields , respectively .
we consider particles composed of two rigidly connected disks separated by a fixed distance @xmath66 .
looking at the motion of particle @xmath1 , we denote by @xmath67 and @xmath68 the positions of the two disks .
we assume here that the dumbbell propels along its principal axis : @xmath69 .
the dynamics of the dumbbell is modeled as follows .
we assume that the disk @xmath155 propels itself in the direction @xmath4 and experiences the force field @xmath156 ) exerted by particle @xmath7 ( @xmath157 or 2 ) .
the corresponding equations of motion are : @xmath158 where , the first term on the r.h.s accounts for the competition between the propulsion and the drag force experienced by the two disks .
the phenomenological drag coefficients @xmath159 and @xmath160 depend on the sizes of the disks .
the internal tension forces @xmath161 and @xmath162 are lagrange multipliers that preserve the inextensibility constraint @xmath163 .
the timescales @xmath164 and @xmath165 control the relaxation of the particle speed towards @xmath4 . as we did it in appendix a
, we assume that they are much faster than the evolution of the force field @xmath64 . in this overdamped limit ,
the dynamics of the dumbbell becomes : @xmath166 in addition , the inextensibility condition imposes @xmath167 . introducing the center of drag of the dumbbell , @xmath168 , and assuming that the variations of the force field @xmath169 occur on large length scales compared to the particle size , the equations of motion are : @xmath170 , \\
\label{dumbbell2 } & \partial_t \hat{{\mathbf{p}}}_a = ( { \mathbb i } - \hat{{\mathbf{p}}}_a \hat{{\mathbf{p}}}_a ) \cdot \left [ \alpha { \mathbf{h}}({\mathbf{r}}_a - { \mathbf{r}}_b,\hat{{\mathbf{p}}}_b ) + \beta ( \hat{{\mathbf{p}}}_a \cdot \nabla ) { \mathbf{h } } \right],\end{aligned}\ ] ] where @xmath171 , @xmath172 and @xmath173 .
. describes the translational motion of the particle .
note that the advection terms are are subdominant in the small-@xmath174 limit ( as we explained it in appendix a. the orientational dynamics of the dumbbell due to interactions with particle @xmath7 is described by eq . .
this equation is identical to eq . with the effective force @xmath77 | we classify the interactions between self - propelled particles moving at a constant speed from symmetry considerations .
we establish a systematic expansion for the two - body forces in the spirit of a multipolar expansion .
this formulation makes it possible to rationalize most of the models introduced so far within a common framework .
we distinguish between three classes of physical interactions : ( i ) potential forces , ( ii ) inelastic collisions and ( iii ) non - reciprocal interactions involving polar or nematic alignment with an induced field .
this framework provides simple design rules for the modeling and the fabrication of self - propelled bodies interacting via physical interactions . a class of possible interactions that should yield new phases of active matter is highlighted . |
we consider the possibility of using of @xmath0 meson decays with missing energy , to test the models with a light spin-0 dm particle .
dark matter search in heavy meson ( and in particular @xmath0 meson ) decays may be complimentary to such experiments as dama @xcite , cdms @xcite and xenon @xcite , which rely on the measurement of kinematic recoil of nuclei in dm interactions and lose ( for cold dm particles ) sensitivity with decreasing mass of the wimp , as the recoil energy becomes small .
so far , @xmath0 meson decays into dark matter have been considered within the models , where dm particles interaction with an ordinary matter is mediated by some light degree of freedom @xcite .
apart from desire of having dm annihilation enhancement ( due to a light intermediate resonance ) and thus having no tension with the dm relic abundance condition @xcite-@xcite , it is also known that @xmath0 meson sm decay is predominantly due to strong interactions .
thus , the wimp production branching ratio , in general , is greatly suppressed compared to relevant weak b decays , and in particular to @xmath1 transition @xcite . in light of this
, it might seem natural to concentrate only on the models within which dark matter production in @xmath0 decays is enhanced due to exchange of a light particle propagator .
yet , our aim is to study @xmath2 decay into a pair of spin-0 dm particles , @xmath3 , and @xmath4 decay into a pair of spin-0 dm particles and a photon , @xmath5 , within the models where light dark matter interaction with an ordinary matter is due to exchange of _ heavy _ particles ( with masses exceeding the bottomonium mass ) .
these models may be free of tension related to satisfying the dm relic abundance constraint as well @xcite .
also , new experimental data on @xmath0 decays into invisible states have been reported by the babar collaboration @xcite . according to these data , @xmath6 and @xmath7 where the interval in the r.h.s .
( [ i3 ] ) is related to the choice of the final state missing mass .
these bounds are significantly stronger than those on invisible @xmath2 decays ( with or without a photon emission ) , reported previously by belle and cleo @xcite and quoted by particle data group @xcite .
we show that babar experimental data on @xmath0 meson invisible decays may constrain the parameter space of light scalar dark matter models , even if there is no dark matter production enhancement due to light intermediate states .
we also illustrate that the study of dark matter production in @xmath0 decays allows us to test regions of parameter space of light spin-0 dm models that are inaccessible for b meson decays with missing energy .
it is also worth mentioning that @xmath0 decays are sensitive to a wider range of wimp mass than b decays .
thus , the study of wimp production in @xmath0 decays is complementary to that for b meson decays . at the energy scales , associated with @xmath0
decays , heavy intermediate degrees of freedom may be integrated out , thus leading to a low - energy effective theory of four - particle interactions .
our strategy would be deriving first model - independent formulae for the @xmath8 and @xmath5 branching ratios within the low - energy effective theory .
then , we confront our predictions with the experimental data , deriving model - independent bounds in terms of the wilson operator expansion coefficients , as the parameters that carry the information on an underlying new physics model .
finally , within a given model , using the matching conditions for the wilson coefficients , we translate these bounds into those on the relevant parameters of the considered model .
the talk is based on the results presented in @xcite and @xcite .
we treat @xmath0 states - neglecting the sea quark and gluon distributions - as bound states of @xmath9 valence quark - antiquark pair that annihilates - with or without emission of a photon - into a pair of dark matter particles . to this approximation
, the relevant low - energy effective hamiltonian may be written as @xmath10 where @xmath11 is the heavy mass and @xmath12 with @xmath13 .
it is worth noting that with the notations used in ( [ g1 ] ) and ( [ g2 ] ) , all the operators @xmath14 , i=1, .. 4 , are hermitean , thus all the wilson coefficients @xmath15 must be real . if dm consists of particles that are their own antiparticles , then only first two operators in eq .
( [ g2 ] ) would contribute . for dm field being a complex scalar state , @xmath3 branching ratio
is given by the following expression : @xmath16^{3/2 } \label{g5}\end{aligned}\ ] ] where @xmath17 is the dm particle mass and @xmath18 , @xmath19 , @xmath20 are the @xmath2 total width , mass and decay constant respectively .
only operator @xmath21 contributes to the decay rate . for @xmath22
being a self - conjugate spin-0 state , @xmath23 , contribution of @xmath21 vanishes as well , as noted above .
thus , one has @xmath24 this result is related to the fact that the final dm particle pair state must be a p - wave , which is impossible due to the bose - einstein symmetry of identical spin-0 particles . in what follows
, @xmath25 must also vanish in higher orders in @xmath26 operator product expansion .
thus , provided that dm pair production is the dominant invisible channel ( the neutrino background may be neglected @xcite ) , the signal for @xmath27 decay would imply that the light spin-0 dm field has a complex nature . no evidence for the @xmath27 mode may lead to some constraints on the parameters of the models with light complex scalar dark matter .
yet , in order to derive such constraints , the experimental limit on the @xmath28 mode must be strong enough .
indeed , using the numerical values of @xmath18 , @xmath19 and @xmath20 @xcite , @xcite , one may rewrite eq .
( [ g5 ] ) as @xmath29^{3/2 } \label{i10}\end{aligned}\ ] ] in what follows , the relevant experiments must be sensitive ( at least ) to @xmath30
. this sensitivity has been reached by the babar experiment @xcite , as it follows from the bound on @xmath31 , given by eq .
( [ i2 ] ) . substituting ( [ i2 ] ) into ( [ i10 ] )
, one derives the following constraint on @xmath32 as a function of @xmath17 and @xmath11 : @xmath33 the behavior of the upper bound on @xmath32 with the dm particle mass and the heavy mass is presented in fig.s [ f7]a ) and [ f7]b ) respectively . for @xmath34 and @xmath35 ,
this bound may be translated into constraints on the relevant couplings of models with a light complex spin-0 dm field . for @xmath36 or for @xmath37 , further improvement of the experimental sensitivity to @xmath28 transition
is necessary .
note that even for @xmath34 and @xmath35 , bound ( [ g25 ] ) still allows @xmath38 to be of the order of unity .
it seems to be very unlikely to saturate such a ( rather weak ) bound , if within a given new physics model , @xmath8 transition is loop - induced .
thus , bound ( [ g25 ] ) on @xmath32 seems to be useful only within the models , where @xmath8 decay occurs at tree level .
further discussion of the scenarios with complex scalar dark matter goes beyond the scope of this talk . yet , in ref .
@xcite we use bound ( [ g25 ] ) to derive constraints on the parameter space of the models with mirror fermions , in particular on that of the mssm with gauge mediated susy breaking and the dm particle in the hidden sector , while mirror fermions being connectors between the hidden and the mssm sectors .
we derive for the first time bounds on the couplings of b - quark interactions with its mirror counterpart , as functions of the wimp mass and the mirror fermion mass . for the scenarios with the dm particle being its own antiparticle , the decay @xmath39 is relevant .
as the neutrino background is negligible @xcite , @xmath39 may be the dominant channel in the @xmath40 mode .
as it was mentioned above , the strongest experimental constraint on @xmath41 is that derived by the babar collaboration @xcite and given by eq .
( [ i3 ] ) .
it may also be written as @xmath42 for the final state missing mass being less than 7 gev , or approximately @xmath43 .
we have to note however that this bound has been derived assuming that @xmath40 transition is mediated by an intermediate resonant higgs state @xmath44 that exists in the mssm extensions with an additional higgs singlet @xcite .
for the models with non - resonant dm production considered here , bound ( [ i13 ] ) may be used only to get _ a preliminary estimate _ of possible constraints on parameters of light spin-0 self - conjugate dm models @xcite , the only existing constraint for the case of two - particle invisible states and the photon having non - monochromatic energy , @xmath45 @xcite , is too weak to have any use .
due to the factor @xmath46 , @xmath47 is always below this limit . ] .
more rigourously , the experimental analysis performed in ref .
@xcite should be extended to the cases , when the emitted photon energy is non - monochromatic and is in the range @xmath48 . to our knowledge , this work is in progress now . as a function of @xmath17 , for @xmath49.,width=302 ] within self - conjugate scalar dm scenarios ,
the partially integrated branching ratio for @xmath39 decay is given by @xmath50 \label{i8}\end{aligned}\ ] ] for @xmath51 , using the numerical values of the @xmath4 mass , total width and decay constant @xcite , one may rewrite eq .
( [ i8 ] ) in the following form : @xmath52 where @xmath53 and @xmath54 at first glance , it may seem that @xmath39 branching ratio is far out of reach of the babar experimental sensitivity , for a reasonable choice of @xmath55 and @xmath56 .
notice , however , that within certain models with light spin-0 dark matter , the wilson coefficients @xmath55 and/or @xmath56 may be enormously large , as they contain some enhancement factors , such as the ratio @xmath57 - due to the mass term in the numerator of a heavy fermion propagator , or the higgs vev s ratio @xmath58 ( with the latter being , say , @xmath59 ) - due to dm particle pair production via exchange of a heavy non - sm higgs degree of freedom .
these enhancement factors can make @xmath55 and/or @xmath56 to be @xmath60 and hence @xmath61 to be @xmath62 , i.e. significantly exceeding bound ( [ i13 ] ) on @xmath41 . using bound ( [ i13 ] ) yields @xmath63 the behavior of the bound on @xmath64 with the dm particle mass is presented in fig .
[ f4 ] for @xmath49 . in the next section we consider the simplest possible model where
the wilson coefficients are made enormously large due to an enhancement factor .
we transform the constraint on @xmath64 into that on the relevant parameters of the model .
as a function of @xmath17 , for @xmath65 and @xmath66 ( line 1 ) , @xmath67 ( line 2).,width=302 ] in this section we consider the two - higgs doublet model ( 2hdm ) with a gauge singlet real scalar dm particle .
the dm interaction part of lagrangian , relevant for our analysis , may be written as @xcite @xmath68 following refs .
@xcite , we consider type - ii version of 2hdm , where @xmath69 generates masses of down - type quarks and charged leptons , whereas @xmath70 generates masses of up - type quarks .
we assume for the higgs vev s ratio to be large , i.e. @xmath71 .
three coupling constants entering lagrangian ( [ t1 ] ) are unknown parameters .
constraints on the couplings @xmath72 and @xmath73 may be derived from the study of @xmath74 transition .
those , in general , are strong enough : combined with the ones coming from the dm relic abundance condition , they rule out a wide range of the wimp mass for the scenarios with @xmath72 and/or @xmath73 dominant @xcite . yet , due to cancellation effects in the relevant diagrams , wimp pair production rate in b meson decays is insensitive to the value of @xmath75 @xcite
. the scenarios with @xmath75 dominant , or at least non - negligible , are thus far unconstrained .
study of dm production in @xmath0 decays enables one to constrain these scenarios , inaccessible by b meson decays with missing energy .
the model contains two cp - even , one cp - odd and two complex charged physical higgs degrees of freedom . to the leading order in the perturbation theory ,
@xmath39 transition occurs at tree level by exchange of a single higgs boson . to this approximation ,
only the cp - even higgs states , @xmath76 and @xmath77 , are relevant for our analysis .
one of the cp - even higgs bosons ( not - necessarily the lightest one ) is standard model - like : its phenomenology is similar to that of the sm higgs boson and the experimental bound on its mass is close to the sm limit ( see @xcite and references therein ) .
contribution of the diagrams with the sm - like higgs boson exchange to the @xmath39 decay amplitude has no any enhancement factor - thus , we may further disregard the dark matter interaction with the sm - like cp - even higgs .
the other cp - even higgs boson is `` new - physics ( np ) like '' : its phenomenology differs drastically from that of the sm higgs boson @xcite .
if dark matter is produced due to exchange of this higgs particle , @xmath39 decay amplitude is enhanced by @xmath58 factor . as discussed above , in this case
the decay branching ratio is within the reach of the babar experimental sensitivity .
if the np - like higgs is the lightest one , its mass may be much below the standard model experimental limit : according the existing experimental data @xcite , @xmath78 or @xmath79 in the general type ii 2hdm . as it was mentioned above , the light higgs scenario is beyond the scope of the present paper , thus we assume here that @xmath78 .
notice , however , that this bound is derived , provided that no invisible higgs decay mode exists . on the other hand ,
if the np - like higgs invisible decay mode is dominant , it may escape detection .
no bound on @xmath80 , to our best knowledge , exists in that case . within the considered model with light scalar dark matter , analysis of the @xmath81 mode
may restrict the scenarios with an invisibly decaying lightest higgs boson by putting severe constraints on the @xmath82 interaction coupling , @xmath83 .
for @xmath84 it is given by @xmath85 the last term in the r.h.s .
of ( [ l9 ] ) , although being suppressed by a factor of @xmath86 , must be retained because of possible hierarchy in the values of @xmath75 and @xmath72 or @xmath73 .
scenarios with such a hierarchy may be of importance : using the matching conditions for the wilson coefficients , @xmath87 and bound ( [ g27 ] ) on @xmath64 , one finds that @xmath83 is constrained to be @xmath88 , if @xmath76 mass approaches to its lower limit , @xmath65 .
we illustrate this in fig .
[ f6 ] for @xmath66 and @xmath67 . for @xmath83 being
so strongly constrained , it seems to be very unlikely that @xmath76 would escape detection and its mass be below 55 gev .
more rigorously , however , detailed re - analysis of the higgs production and decay rates , including that of @xmath89 , should be performed .
one can show @xcite that when choosing @xmath58 sufficiently large , bound on @xmath83 may still be rigorous , if the @xmath76 mass is heavier than 55 gev .
thus , within the type ii 2hdm with a light spin-0 dark matter , study of @xmath81 decay channel may lead to severe constraints on the lightest cp - even higgs invisible decay coupling , if that higgs is new - physics like .
the remarkable feature of the model is that constraints on the parameter space are derived even if the np - like higgs is the heaviest cp - even one .
we illustrate this choosing @xmath90 - the existing theoretical upper bound on the sm - like higgs boson @xcite and exclusion of the sm higgs mass interval @xmath91 with 95% c. l. by the cdf and d0 data @xcite allow us to infer that above 160 gev , the cp - even higgs boson is presumably the heaviest one and np - like . the matching conditions for the wilson coefficients are now @xmath92 thus , bound ( [ g27 ] ) on @xmath64 may be transformed into that on @xmath93 as a function of the wimp mass , @xmath58 and the heaviest cp - even higgs mass . as one can see from fig .
[ f5]a ) , for @xmath94 and @xmath95 , @xmath75 is constrained to be of order of the sm weak coupling or smaller ( @xmath96 ) , if @xmath97 .
also , for the same choice of the higgs mass and @xmath58 , @xmath93 is to be less than one , if the wimp mass is less than 2 gev . bound on @xmath93
is significantly more rigorous for higher values of @xmath58 .
for instance , if choosing @xmath98 , one gets @xmath99 and @xmath100 for @xmath97 and @xmath101 respectively .
the restrictions on @xmath93 are essential also for higher values of the heaviest cp - even higgs mass : for @xmath102 , they are still of the interest up to @xmath103 , as one can see from fig .
[ f5]b ) .
thus , within the type ii 2hdm with a scalar dark matter , for large @xmath58 scenario , @xmath0 meson decay into a dark matter particles pair and a photon , @xmath104 , may be used to derive essential constraints on the parameters of the model , which otherwise can not be tested by b meson decays with invisible outcoming particles .
thus , spin-0 dark matter production in @xmath0 meson decays has been investigated .
we restricted ourselves by consideration of the models where the decays occur due to exchange of heavy non - resonant degrees of freedom .
we performed our calculations within low - energy effective theory , integrating out heavy degrees of freedom .
this way we derived model - independent formulae for the considered branching ratios .
we used these formulae to confront our theoretical predictions with existing experimental data on invisible @xmath0 decays , both in a model - independent way and within particular models .
it has been shown that within the considered class of models , dm production rate in @xmath0 decays is within the reach of the present experimental sensitivity .
thus , @xmath0 meson decays into dark matter , with or without a photon emission , may be used to constrain the models with a gev or lighter spin-0 dm . experimental constraints on the @xmath40 mode are derived assuming that dark matter is produced by exchange of a light resonant scalar state . within the scenarios with non - resonant dm production considered here ,
these constraints may be used only to make preliminary estimates of possible bounds on the parameters of the models . yet
, those estimates show that these bounds may be rigorous enough ; besides , they are derived within the least restrained presently light scalar dm scenarios .
our goal is thus to encourage the experimental groups to analyze the experimental data on @xmath105 also for the case of non - monochromatic photon emission and spin-0 invisible states .
so , from our analysis one may conclude that dark matter production in @xmath0 meson decays may serve as an interesting alternative to commonly used dm search methods , capable of providing a valuable information on dm particles , if those turn to have a mass of the order of a few gev or smaller .
author is grateful to yu .
kolomensky , a. blechman , b. mcelrath and j. cao for valuable comments and stimulating discussions . +
this work has been supported by the grants nsf phy-0547794 and doe de - fgo2 - 96er41005 .
99 r. bernabei et al . , ( dama collaboration ) eur .
j. c 56 , 333 ( 2008 ) .
r. bernabei et al . , ( dama collaboration ) int .
d 13 , 2127 ( 2004 ) .
r. bernabei et al . , ( dama collaboration ) riv .
nuovo cimento , 26 n1 , 1 ( 2003 ) , arxiv : astro - ph/0307403 .
z. ahmed et al .
, cdms collaboration , phys .
102 , 011301 , ( 2009 ) .
j. angle et al .
( xenon collaboration ) phys .
lett . 101 , 091301 ( 2008 ) .
j. angle et al .
( xenon collaboration ) phys .
100 , 021303 ( 2008 ) .
j. f. gunion , d. hooper , b. mcelrath , phys .
d73 , 015011 ( 2006 ) .
p. fayet , phys .
d 75 , 115017 ( 2007 ) .
p. fayet , phys . rev .
d 74 , 054034 ( 2006 ) . c. boehm , p. fayet , nucl .
b 683 , 219 ( 2004 ) .
y. kahn , m. schmitt , t. m. p. tait , phys .
d 78 , 115002 ( 2008 ) .
d. hooper , k. m. zurek , phys . rev .
d 77 , 087302 ( 2008 ) .
j. cao , h. e. logan , j. m. yang , phys .
d 79 , 091701 ( 2009 ) . c. bird , r. kowalewski , m. pospelov , mod .
lett . a 21 , 457 ( 2006 ) .
c. bird , p. jackson , r. kowalewski , m. pospelov , phys .
93 , 201803 ( 2004 ) .
j. l. feng , j. kumar , phys .
lett . 101 , 231301 ( 2008 ) .
j. l. feng , j. kumar , l. e. strigari , phys .
b 670 , 37 ( 2008 ) .
j. l. feng , j. kumar , i. learned , l. e. strigari , jcap 0901 , 032 ( 2009 ) , arxiv:0808.4151[hep - ph ] .
the babar collaboration , b. aubert et al .
, arxiv : 0908.2840 [ hep - ex ] .
the babar collaboration , b. aubert et al . , arxiv : 0808.0017 [ hep - ex ] ; + s. j. sekula , for the babar collaboration , arxiv : 0810.0315 [ hep - ex ] . o. tajima et al . ,
the belle collaboration , phys .
98 , 132001 ( 2007 ) .
r. balest et al .
, cleo collaboration , phys .
d 51 , 2053 ( 1995 ) . c. amsler et al .
( particle data group ) , phys .
b 667 , 1 ( 2008 ) .
g. k. yeghiyan , arxiv : 0909.4919 [ hep - ph ] .
a. badin , a. a. petrov , g. k. yeghiyan , to be published .
h. m. choi , phys .
d 75 , 073016 , ( 2007 ) .
r. dermisek , j. f. gunion , b. mcelrath , phys .
d 76 , 051105 ( 2007 ) ; r. dermisek , j. f. gunion , phys . rev .
d 75 , 075019 ( 2007 ) ; r. dermisek , j. f. gunion , phys . rev .
d73 , 111701 ( 2006 ) .
x. g. he et al .
, phys . rev .
d 79 , 023521 ( 2009 ) .
j. abdallah , et al .
, delphi collaboration , eur .
journal c32 , 475 , ( 2004 ) .
p. achard et al . , l3 collaboration , phys .
b 609 , 35 ( 2005 ) .
j. f. gunion , h. e. haber , g. kane , s. dawson , _ the higgs hunter s guide _ , addison - wesley publishing company , 1990 .
g. abbiendi et al . , the opal collaboration , eur .
journal c 40 , 317 ( 2005 ) .
g. abbiendi et al . , the opal collaboration , eur .
journal c 18 , 425 ( 2001 ) .
g. m. asatrian , a. n. ioannisian , s. g. matinyan , sov .
53 , 371 , ( 1991 ) , yad .
53 , 592 , ( 1991 ) .
d. benjamin , for the cdf , d0 collaborations , arxiv : 0906.1403 [ hep - ex ] . | we examine decays of a spin-1 bottomonium into a pair of light scalar dark matter ( dm ) particles , assuming that dark matter is produced due to exchange of heavy degrees of freedom .
we perform a model - independent analysis and derive formulae for the branching ratios of these decays .
we confront our calculation results with the experimental data .
we show that dark matter production in @xmath0 decays may lead to constraints on parameters of the models containing a light spin-0 dm particle . |
in condensed matter physics the origins of , and phenomena associated with , metal - insulator and magnetic transitions remain issues of fundamental interest , especially when they do not fit into a conventional mold . from this viewpoint , a few @xmath13-based oxides , including the osmate studied here
, that display various high formal oxidation states are attracting great interest .
they display an abundant variety of physical phenomena , including superconductivity , important effects due to large spin - orbit coupling ( soc ) , unconventional metal - insulator transitions ( mit ) ( the focus of this paper ) , and unusual magnetic behavior . within heptavalent os@xmath14 @xmath15 systems ,
the cubic double perovskite ba@xmath16naoso@xmath2 is ferromagnetic with an unusually small magnetic moment while maintaining a cubic structure,@xcite whereas the magnetic structures of the trigonal na@xmath0oso@xmath17 , li@xmath17oso@xmath2 , and the double perovskite ba@xmath16lioso@xmath2 are based on antiferromagnetic ( afm ) order.@xcite in the hexavalent @xmath18 systems , the double perovskite ba@xmath16caoso@xmath2 is antiferromagnetic.@xcite at the metallic @xmath19 oxidation level , kos@xmath16o@xmath2 in the @xmath20-pyrochlore structure shows superconductivity@xcite at @xmath21=8 k , whereas the triple perovskite ba@xmath0lios@xmath16o@xmath22 becomes afm@xcite at 13 k , but shows significant reduction in the effective moment of 3.34 @xmath8 from the theoretical spin - only value of 4.78 @xmath8 , suggesting strong soc effects . in a pentavalent @xmath11 system , the slightly distorted double perovskite sr@xmath16croso@xmath2 is a semimetallic ferromagnet in which soc plays an important part in the magnetic behavior , by reducing the spin moment by 0.27 @xmath8 and inducing an orbital moment of 0.17 @xmath8 on the os ion , decreasing the net moment by nearly a factor of two.@xcite pentavalent cd@xmath16os@xmath16o@xmath23 , with its magnetic frustration on the pyrochlore os sublattice , has been suggested to be an itinerant slater insulator,@xcite _ i.e. _ a mit induced by antiferromagnetic ordering rather than by strong correlation effects.@xcite regarding slater insulating states , @xmath24 sr@xmath16iro@xmath25 , with a weakly canted antiferromagnetic mit@xcite at 250 k , has recently been characterized as a slater insulator.@xcite the character of the non - collinear ordered @xmath11 cd@xmath16os@xmath16o@xmath23 is still under discussion.@xcite the orthorhombic perovskite naoso@xmath0 , also with pentavalent os , synthesized by shi _
et al . _ , undergoes a mit coincident with antiferromagnetic ordering at the rather high temperature of @xmath26=410 k with no evidence for any further structural distortion below 600 k.@xcite the resistivity increases , and the hall number decreases , by @xmath103 orders of magnitude from t@xmath27 to 200 k and continued decreasing to 2 k , the insulating character becoming more accelerated below 25 k toward a strongly insulating ground state .
the electronic contribution to the specific heat measurement vanished at low temperature , indicating a fully gapped fermi surface . due to this uncommon behavior , seemingly a continuously increasing gapping of the fermi surface leading to decrease and finally vanishing of carriers ,
this system was argued to be the first example of the continuous mit ( so - called the slater insulator with continuous opening of the afm gap ) .
recently , calder _ et al .
_ confirmed the continuous transition as a simultaneous change in magnetic and transport character.@xcite the effective curie moment of 2.71 @xmath8 observed by shi _
et al . _ is reasonably consistent with a @xmath28 moment , reduced ( by 30% ) by hybridization with oxygen and perhaps compensated somewhat by an orbital moment .
this picture is one of high spin @xmath11 os@xmath29 .
[ ] neutron power diffraction ( npd ) measurements indicate g - type antiferromagnetism ( g - afm ) with direction of moments alternating on neighboring os ions , but an os _ ordered moment _ of only 1 @xmath8.@xcite this reduced value is difficult to reconcile with a @xmath28 picture , since three dimensional arrays of large spins should not be subject to reduction by fluctuations . in the only previous theoretical work on naoso@xmath0 , du et al . calculated a similar ( even somewhat smaller ) value from the sort of first principles calculations we will describe and use below.@xcite the npd measurements show the spin oriented along the @xmath30-axis without canting of the spins,@xcite which provides additional information related to the impact of large soc , one aspect that we want to investigate more thoroughly .
du _ et al .
_ obtained a magnetic ground state of the observed g - afm type , and provided a comparison of this state to other types of magnetic order that they could obtain in their calculations for naoso@xmath0.@xcite there remain several unresolved questions : ( 1 ) what is the origin and the nature of the afm ordering , and what other states are energetically nearby ? ( 2 ) why is the ordered moment only 1 @xmath8 for pentavalent os @xmath11 ? is strong soc a factor ? ( 3 ) why is this mit continuous instead of the more common first - order type of mit in transition metal oxides ? ( 4 ) does soc perhaps play a more central role than is apparent from the results presented by du _
et al . _ ?
the orientation of the moments ( magnetocrystalline anisotropy ) may provide useful conditions to address this question . for another perovskite - based osmate ba@xmath16naoso@xmath2 , we observed strong interplay between correlation effects , small bandwidth , and strong soc that are necessary to consider together to build an interpretation of the unusual features of that system,@xcite which include that it is an uncommon _ ferromagnetic _ mott insulator , it displays a very small ordered moment , and it remains cubic when it seems the @xmath15 configuration should favor at least a small jahn - teller distortion . in this paper , we address most of the questions we raised above , using first principles spin - polarized calculations including both correlation effects and soc . .
atomic positions of atoms of naoso@xmath0 in the @xmath31 ( no .
62 ) structure , as experimentally observed.@xcite [ cols="^,^,^,^,^,^,^",options="header " , ] [ table2 ] at @xmath32 , the @xmath33 occupation matrix substantiates this conclusion .
the majority @xmath3 orbitals are almost equally occupied at about 0.7@xmath34 per orbital , substantially reduced from the formal value of unity .
the majority @xmath35 and all the minority @xmath33 orbitals , which are formally unoccupied , each have occupation of about 0.3@xmath34 per orbital , reflecting substantial @xmath9 hybridization and indicating a strong deviation from the conventional ionic @xmath11 viewpoint . in total
there are 4.3 electrons throughout the @xmath33 orbitals , instead of the formal value 3 .
this ` discrepancy ' is not so surprising , as it is widely recognized that an actual charge on @xmath33 cations differs from the formal charge , sometimes@xcite by a factor of 2 . for os here , the formal charge of + 5
is contrasted with an actual charge of around + 3.7 .
the reduction by a factor of three of the formal moment is a more notable feature , and unlike the case of ba@xmath16naoso@xmath2 where the moment is also vastly reduced,@xcite here it is unrelated to soc .
naoso@xmath0 is a case of an insulating transition metal oxide where the formal charge ( oxidation state ) picture is murky at best , consistent with the itineracy evident in the electronic structure and apparently in the mit as well .
figure [ gap ] shows change in the direct and indirect gaps due to varying @xmath36 . in the distorted structure and
including soc , an energy gap opens only at @xmath32=2.2 ev .
[ ] above @xmath32 , both gaps increase linearly at a rate of 0.16 ev / ev , with the difference of 0.3 ev independent of @xmath36 .
below @xmath32 , a negative indirect gap ( not shown ) results due to significant lowering of the bottom of the conduction band at the @xmath37 point , though the direct gap is already established ( the occupied and unoccupied bands are disjoint ) . this feature may be related with the proposal of a continuous transition in opening an energy gap , as discussed in another osmate cd@xmath16os@xmath16o@xmath23 by shinaoka _ et al._@xcite considering a likely @xmath380.5 ev , @xmath39 2.7 ev @xmath40 suggests for this multiband system that a mott transition is not the appropriate picture for this continuous mit . the corresponding band structure at @xmath32 is displayed in fig .
[ band2 ] .
the soc splitting is about 0.2 ev at the @xmath41-point .
an unusual outcome of the structural distortion is that the top of the valence bands is nearly flat , reflected in the step increase in the valence band dos evident in fig .
[ afmdos ] , and providing a two - dimensional character to the low energy phase space .
we also point out that the dispersion is very similar above and below @xmath42 , relevant information for yet - to - be - done optical experiments since the @xmath43-conserving joint dos may be sharply peaked at the onset of absorption . within lsda+u+soc
as in lsda , the afm state is nearly degenerate with the nm state .
the fact that neither state is energetically favored is consistent with the finite temperature transition between the two states .
the value of @xmath4 appropriate for naoso@xmath0 has already been mentioned in the literature , in addition to the conclusion @xmath442 ev of du _
_ that is nearly identical to ours .
et al . _ reported data on magnetic x - ray resonant scattering ( mxrs ) with interpretation from x - ray absorption near edge spectra ( xanes ) obtained from joly s finite difference code.@xcite they obtained a better correspondence using @xmath45 than with nonzero @xmath4 .
we think there is no direct contradiction . in their work
they are dealing with resonant transitions between deep @xmath46 and @xmath47 levels and unoccupied @xmath13 states . in our work , based on the self - consistent lda+u method , the effect of @xmath4 is to increase the separation of occupied @xmath13 bands and unoccupied @xmath13 bands , which is a different issue involving a @xmath4 in a different manner . as discussed for several decades , @xmath3 orbitals when split off from the @xmath35 orbitals by the crystal field ,
form a representation of an angular momentum of @xmath48.@xcite in the spherical hund s rule limit the total angular momentum is zero for a high spin @xmath49 system . especially in this locally - cubic geometry and half filling , the orbital moment should be small . as shown in table [ table2 ] , inclusion of soc leads to an orbital moment antialigned with the spin moment and with magnitude no larger than 0.1 @xmath8 .
thus in the insulating state the net moment remains about 1 @xmath8 , consistent with the experimentally observed value,@xcite and relatively insensitive to @xmath36 in the range studied here ( see table [ table2 ] ) .
we have calculated total energies with lda+u+soc with four spin orientations @xmath50 , @xmath51 , @xmath52 , and @xmath53 to determine the easy axis . as observed by calder _
et al._,@xcite @xmath52 is the easy axis . along @xmath51 , both spin and orbital moments increase by 10% in magnitude , but the energy cost is large , 474 mev / f.u .
, despite some additional hund s coupling energy .
along @xmath53 , the spin moment is enhanced to 1.25 @xmath8 while the energy is higher by 51 mev / f.u .
consistent with the similarities between the @xmath54 and @xmath30 axes ( the octahedron rotation is around the @xmath55 axis ) the energy difference is small , only 5 mev / f.u .
approached from high temperature , a continuous mit occurs at the point where the spin - symmetric phase becomes unstable to magnetic order , which is afm order in the case of naoso@xmath0 . in slater - type ( itinerant ) transitions , ordering conventionally is driven by fermi surface ( fs ) nesting , as in the classic case of chromium .
the common picture for a slater mit is that a spin density wave ( sdw ) gaps at least a portion of the fs , and the increasing amplitude of the sdw destroys more and more of the fs as the temperature is lowered .
the specific heat data for naoso@xmath0 indicate the fs is completely gone as t@xmath560 .
the resistivity data strongly suggests it has been effectively destroyed at intermediate temperature , since the magnitude and temperature coefficient reflect insulating behavior even as spin - fluctuation scattering has began to be frozen out by the magnetic order ( which , in itself , should _ lower _ rather than increase the resistivity ) , and the strongly decreasing hall coefficient is consistent with rapidly shrinking fss . in this section
we first discuss the electronic structure in the ( high temperature ) nonmagnetic metallic state , albeit without consideration of the thermal broadening and the lattice dynamical and spin fluctuation scattering that may be necessary for a detailed picture .
then we provide a scenario for the mit that is consistent with observations and what is understood about the electronic structure .
the lda and lda+soc band structures without magnetic order are displayed together in fig .
[ nmband ] .
we have checked that including u@xmath57 in this nonmagnetic phase leads to negligible effect , as is commonly observed . in the previous section we established
that soc has relatively little impact on the bands in the afm phase , displaying its main impact by its seeming competition with correlation effects so that the correlation strength necessary to open the gap ( within the afm phase ) is doubled to @xmath102 ev . near @xmath42 , the effect of soc on the band structure ( fig .
[ nmband ] ) in the high t phase again seems minor , but it is clearly visible at symmetry points ( y , s ) where it removes degeneracies and splits bands .
moreover , the effect on the dos is minor , as is clear from the inset of fig .
[ nmband ] . with or without soc ,
the fermi level dos n(@xmath42 ) @xmath58 3 states / ev - f.u . for both spins , corresponding to a linear specific heat coefficient @xmath59 = 7 mj / mol - k@xmath60 .
this standard metallic value is corroborated by large fs sheets , which we now discuss .
because the topology of the fss is central to a sdw transition , we display fss without and with soc in fig .
the small shifts in band energies due to soc that can be seen in fig .
[ nmband ] are translated by the small velocities into considerable changes in the complex multi - sheeted fss , with some of the clearest changes occurring around the symmetry points .
the fss retain to some extent an approximate 4-fold symmetry in the @xmath54-@xmath30 plane ( @xmath61 plane in fig .
[ fs ] ) , which of course is not a true symmetry of the @xmath31 structure but is useful to recognize .
there are both electron and hole type surfaces , making the magnitude of the hall coefficient difficult to interpret .
most theoretical work on afm transitions in metals , including very recent ones , focus either on fs nesting ( cr is the classic example ) or fs `` hot spots '' involving band and fs folding due to the finite-@xmath12 order that set in at the transition . in the first case
the nested sections of fs become gapped , while in the second case the fs also reconstructs but in a less obvious manner .
for example , sachdev _
et al._@xcite review the latter scenario and provide evidence of a breakdown of fermionic quasiparticle excitations at the transition , and a fs reconstruction from large to small surfaces that have been discussed for many years in the high temperature superconducting cuprates .
both the nesting scenario and the hot spot scenario of the continuous slater transition to an afm incurs difficulty in naoso@xmath0 , because it is a _
q=0 ordering _ for which both scenarios , which focus on a nonzero wavevector @xmath62 , become moot .
this situation can arise , and not uncommonly does so , in crystal structures with more than one magnetic ion of a given type . in naoso@xmath0 , the four symmetry - related os ions in the gdfeo@xmath0 distorted lattice devolve into two pairs , with spins in opposite directions .
interestingly , the original exposition by slater@xcite did not involve finite-@xmath12 physics in any explicit manner .
his picture , based on spin - unrestricted hartree - fock theory , was that afm ordering would introduce new fourier components in the ( spin dependent ) potential , leading to new gaps in the band structure that could well appear at the fermi level , especially in stoichiometric compounds where there is the likelihood that bands can be completely occupied , or completely unoccupied .
if the new gaps appear at a zone boundary , it will be energetically favored to double the periodicity , as in the simplest picture of sdw order . yet
if the new gaps appear only at @xmath12=0 ( no cell doubling ) , the band structure may still become gapped in spite of the lack of any extension of periodicity .
the effect of afm ordering in naoso@xmath0 on the electronic structure can be seen by comparing the metallic bands of the nonmagnetic high t phase in fig .
[ nmband ] ( solid lines ) with the afm insulating bands in fig .
[ band2 ] .
new fourier components of the potential separate the bands strongly at zone boundary symmetry points ( _ viz .
_ r , y , s ) and at the zone center the bands are already split .
slater also tied this type of ordering to bonding - antibonding band formation , which provides a band energy ( bonding energy ) gain in addition to the hund s rule ( magnetic ) gain in energy .
a rudimentary picture that is consistent with data and with this understanding of the electronic structure can be assembled . already at high t the strongly distorted , four f.u .
per primitive cell , structure is dictated by the tolerance ratio of 0.87 .
the distortion leaves flat bands at and near the fs , which is large and multi - sheeted . at t@xmath27 , magnetic order sets in and the metallic band structure begins to morph continuously toward the t=0 afm insulating band structure .
the fss , both electron and hole type , begin to shrink in volume and vanish one - by - one , until no fs remains .
the change from the metallic to the insulating band structure should be monotonic in , and nearly proportional to , the magnetic order parameter .
the magnetic order parameter measured by calder _
_ was provided only down to @xmath100.85 t@xmath27,@xcite so it is not possible to estimate when the last fs may be expected to disappear .
as mentioned earlier , thermal smearing and lattice and spin fluctuations will complicate a detailed picture .
one loose end in this picture is the apparent discontinuity a jump by roughly 50% in the paramagnetic susceptibility@xcite just below t@xmath27 . without a fermi surface or hot spot origin ,
the mechanism of g - type ordering is not clear , and was not addressed by du _
_ in their theoretical study of magnetic ordering.@xcite since g - type ordering is common in perovskite transition metal oxides , it seems likely this ordering represents an emergence of the nearest neighbor afm coupling commonly occurring in perovskites , in spite of the apparent itineracy in the electronic structure and the collapse of the @xmath11 valence state picture .
our results have clarified a number of features of naoso@xmath0 , and the questions posed in sec .
i. first , a combination of the distortion of the perovskite structure and moderate correlation effects is responsible for both antiferromagnetism and the insulating gap in the ground state .
our calculations account naturally for the factor of three reduction in the observed moment ( 1@xmath8 ) relative to the ideal @xmath49 value , due to very strong hybridization with the o @xmath63 orbitals that render the conduction states , and the magnetism , itinerant in nature . the oxygen ions , being coordinated with two os ions with antialigned spins , are unpolarized by ( approximate ) symmetry .
the calculated energetics , with small energy difference between the magnetic and non - magnetic states , is also consistent with a slater - type metal - to - insulator transition to an afm state .
moreover , the simplest specification of a mott insulator is one in which the insulating state does not depend on magnetic order . in naoso@xmath0
all the evidence indicates that magnetic order is _ necessary _ for insulating behavior .
because of the itinerant nature of the moment , a heisenberg model can not provide a representation which could then be applied to predict the nel temperature .
spin - orbit coupling , often a dominant force in @xmath13 metal oxides , has a more modest effect in this compound .
while it effects the band structure enough that the effective coupling strength u@xmath57 is increased from 1 ev to 2.2 ev , it is in itself ineffective in decreasing the local moment . the strong moment reduction ( from a localized high - spin @xmath3 picture of 3@xmath8 )
is due to strong hybridization with o @xmath63 orbitals that is , itineracy which is itself affected by the distortion .
the minor importance of soc is consistent with the finding of calder _ et al . _
that xray resonant photoelectron spectra show only minor effect of soc.@xcite though @xmath64 perovskite oxides have been studied extensively for decades , their @xmath13 counterparts are relatively new to detailed inspection of their properties and the underlying mechanisms . our results for naoso@xmath0 will help to provide a basis for understanding the distinctions in the physics of @xmath13 versus @xmath64 ( and @xmath65 ) oxides .
this research was supported by the basic science research program through nrf of korea under grant no .
2012 - 0002245 .
w.e.p was supported by doe under grant no .
de - fg02 - 04er46111 .
y. krockenberger , k. mogare , m. reehuis , m. tovar , m. jansen , g. vaitheeswaran , v. kanchana , f. bultmark , a. delin , f. wilhelm , a. rogalev , a. winkler , and l. alff , phys .
b * 75 * , 020404(r ) ( 2007 ) .
y. g. shi , y. f. guo , s. yu , m. arai , a. a. belik , a. sato , k. yamaura , e. takayama - muromachi , h. f. tian , h. x. yang , j. q. li , t. varga , j. f. mitchell , and s. okamoto , phys .
b * 80 * , 161104(r ) ( 2009 ) .
s. calder , v. o. garlea , d. f. mcmorrow , m. d. lumsden , m. b. stone , j. c. lang , j .- w .
kim , j. a. schlueter , y. g. shi , k. yamaura , y. s. sun , y. tsujimoto , and a. d. christianson , phys .
. lett . * 108 * , 257209 ( 2012 ) . in ref .
et al . _ quoted @xmath66=2 ev , without mentioning @xmath67 explicitly .
considering that they also used wien2k , this value seems to be @xmath32 which is consistent with our results .
y. joly , phys .
b * 63 * , 125120 ( 2001 ) . | the orthorhombic perovskite naoso@xmath0 undergoes a continuous metal - insulator transition ( mit ) , accompanied by antiferromagnetic ( afm ) order at @xmath1 k , suggested to be an example of the rare slater ( itinerant ) mit .
we study this system using _ ab initio _ and related methods , focusing on the origin and nature of magnetic ordering and the mit .
the rotation and tilting of oso@xmath2 octahedra in the gdfeo@xmath0 structure result in moderate narrowing the band width of the @xmath3 manifold , but sufficient to induce flattening of bands and afm order within the local spin density approximation ( lsda ) , where it remains metallic but with a deep pseudogap . including on - site
coulomb repulsion @xmath4 , at @xmath5 ev a mit occurs only in the afm state .
effects of spin - orbit coupling ( soc ) on the band structure seem minor as expected for a half - filled @xmath6 shell , but soc doubles the critical value @xmath7 necessary to open a gap and also leads to large magnetocrystalline energy differences in spite of normal orbital moments no greater than 0.1@xmath8 .
our results are consistent with a slater mit driven by magnetic order , induced by a combination of structurally - induced band narrowing and moderate coulomb repulsion , with soc necessary for a full picture .
strong @xmath9 hybridization reduces the moment , and when bootstrapped by the reduced hund s rule coupling ( proportional to the moment ) gives a calculated moment of @xmath101 @xmath8 , consistent with the observed moment and only a third of the formal @xmath11 value .
we raise and discuss one important question : since this afm ordering is at @xmath12=0 ( in the 20 atom cell ) where nesting is a moot issue , what is the microscopic driving force for ordering and the accompanying mit ? |
we thank adrian dumitru , kevin dusling and yasushi nara for help with their rcbk codes .
we thank zhongbo kang for his contributions at early stages of this work , and jean - philippe lansberg and lijuan ruan for helpful discussions .
this work was supported in part by the u. s. department of energy under contract no .
de - ac02 - 98ch10886 .
n. brambilla , s. eidelman , b. heltsley , r. vogt , g. bodwin , _ et al .
_ , _ heavy quarkonium : progress , puzzles , and opportunities _ , http://dx.doi.org/10.1140/epjc/s10052-010-1534-9[_eur.phys.j._ * c71 * ( 2011 ) 1534 ] [ http://arxiv.org/abs/1010.5827[arxiv:1010.5827 ] ] [ http://inspirehep.net/record/874793[inspire ] ]
. g. t. bodwin , e. braaten , and g. p. lepage , _ rigorous qcd analysis of inclusive annihilation and production of heavy quarkonium _ , http://dx.doi.org/10.1103/physrevd.55.5853 , 10.1103/physrevd.51.1125[_phys.rev .
_ * d51 * ( 1995 ) 11251171 ] [ http://arxiv.org/abs/hep-ph/9407339[hep-ph/9407339 ] ] [ http://inspirehep.net/record/374999[inspire ] ] .
m. butenschoen and b. a. kniehl , _
@xmath1 polarization at tevatron and lhc : nonrelativistic - qcd factorization at the crossroads _ , http://dx.doi.org/10.1103/physrevlett.108.172002[_phys.rev.lett._ * 108 * ( 2012 ) 172002 ] [ http://arxiv.org/abs/1201.1872[arxiv:1201.1872 ] ] [ http://inspirehep.net/record/1083802[inspire ] ] .
chao , y .- q .
ma , h .- s .
shao , k. wang , and y .- j .
@xmath1 polarization at hadron colliders in nonrelativistic qcd _ , http://dx.doi.org/10.1103/physrevlett.108.242004[_phys.rev.lett._ * 108 * ( 2012 ) 242004 ] [ http://arxiv.org/abs/1201.2675[arxiv:1201.2675 ] ] [ http://inspirehep.net/record/1084398[inspire ] ] . b. gong , l .- p .
wan , j .- x .
wang , and h .- f .
zhang , _ polarization for prompt @xmath1 , @xmath100 production at the tevatron and lhc _ , http://dx.doi.org/10.1103/physrevlett.110.042002[_phys.rev.lett._ * 110 * ( 2013 ) 042002 ] [ http://arxiv.org/abs/1205.6682[arxiv:1205.6682 ] ] [ http://inspirehep.net/record/1116546[inspire ] ] .
kang , y .- q .
ma , j .- w .
qiu , and g. sterman , _ heavy quarkonium production at collider energies : factorization and evolution _ , http://dx.doi.org/10.1103/physrevd.90.034006[_phys.rev._ * d90 * ( 2014 ) 034006 ] [ http://arxiv.org/abs/1401.0923 [ arxiv:1401.0923 ] ] [ http://inspirehep.net/record/1276125 [ inspire ] ]
. g. t. bodwin , h. s. chung , u .- r .
kim , and j. lee , _ fragmentation contributions to @xmath1 production at the tevatron and the lhc _ , http://dx.doi.org/10.1103/physrevlett.113.022001[_phys.rev.lett._ * 113 * ( 2014 ) 022001 ] [ http://arxiv.org/abs/1403.3612[arxiv:1403.3612 ] ] [ http://inspirehep.net/record/1285966[inspire ] ] .
ma , j .- w .
qiu , g. sterman , and h. zhang , _ factorized power expansion for high-@xmath101 heavy quarkonium production _ , [ http://arxiv.org/abs/1407.0383[arxiv:1407.0383 ] ] [ http://inspirehep.net/record/1304290[inspire ] ] .
l. gribov , e. levin , and m. ryskin , _ semihard processes in qcd _ , http://dx.doi.org/10.1016/0370-1573(83)90022-4[_phys.rept._ * 100 * ( 1983 ) 1150 ] [ http://inspirehep.net/record/200601[inspire ] ] .
a. h. mueller and j. qiu , _ gluon recombination and shadowing at small values of x _ , http://dx.doi.org/10.1016/0550-3213(86)90164-1[_nucl.phys._ * b268 * ( 1986 ) 427 ] [ http://inspirehep.net/record/17408[inspire ] ] . z .-
kang , y .- q .
ma , and r. venugopalan , _ quarkonium production in high energy proton - nucleus collisions : cgc meets nrqcd _ , http://dx.doi.org/10.1007/jhep01(2014)056[_jhep_ * 1401 * ( 2014 ) 056 ] [ http://arxiv.org/abs/1309.7337 [ arxiv:1309.7337 ] ] [ http://inspirehep.net/record/1255833 [ inspire ] ] .
f. gelis , e. iancu , j. jalilian - marian , and r. venugopalan , _ the color glass condensate _ , http://dx.doi.org/10.1146/annurev.nucl.010909.083629 [ _ ann.rev.nucl.part.sci .
_ * 60 * ( 2010 ) 463489 ] [ http://arxiv.org/abs/1002.0333[arxiv:1002.0333 ] ] [ http://inspirehep.net/record/844547[inspire ] ] .
j. p. blaizot , f. gelis , and r. venugopalan , _ high - energy pa collisions in the color glass condensate approach .
2 . quark production _ ,
http://dx.doi.org/10.1016/j.nuclphysa.2004.07.006[_nucl.phys._ * a743 * ( 2004 ) 5791 ] [ http://arxiv.org/abs/hep-ph/0402257[hep-ph/0402257 ] ] [ http://inspirehep.net/record/645059[inspire ] ] . h. fujii , f. gelis , and r. venugopalan , _ quantitative study of the violation of @xmath102-factorization in hadroproduction of quarks at collider energies _ , http://dx.doi.org/10.1103/physrevlett.95.162002[_phys.rev.lett._ * 95 * ( 2005 ) 162002 ] [ http://arxiv.org/abs/hep-ph/0504047[hep-ph/0504047 ] ] [ http://inspirehep.net/record/679806[inspire ] ] . h. fujii , f. gelis , and r. venugopalan , _ quark pair production in high energy pa collisions : general features _ , http://dx.doi.org/10.1016/j.nuclphysa.2006.09.012[_nucl.phys._ * a780 * ( 2006 ) 146174 ] [ http://arxiv.org/abs/hep-ph/0603099[hep-ph/0603099 ] ] [ http://inspirehep.net/record/712422[inspire ] ] . h. fujii and k. watanabe , _ heavy quark pair production in high energy pa collisions : quarkonium _ , http://dx.doi.org/10.1016/j.nuclphysa.2013.06.011[_nucl.phys._ * a915 * ( 2013 ) 123 ] [ http://arxiv.org/abs/1304.2221[arxiv:1304.2221 ] ] [ http://inspirehep.net/record/1227544[inspire ] ] .
d. kharzeev , e. levin , and k. tuchin , _ nuclear modification of the j/@xmath103 transverse momentum distributions in high energy pa and aa collisions _ , http://dx.doi.org/10.1016/j.nuclphysa.2014.01.006[_nucl.phys._ * a924 * ( 2014 ) 4764 ] [ http://arxiv.org/abs/1205.1554[arxiv:1205.1554 ] ] [ http://inspirehep.net/record/1113916[inspire ] ] .
f. gelis , k. kajantie , and t. lappi , _ chemical thermalization in relativistic heavy ion collisions _ , http://dx.doi.org/10.1103/physrevlett.96.032304[_phys.rev.lett._ * 96 * ( 2006 ) 032304 ] [ http://arxiv.org/abs/hep-ph/0508229[hep-ph/0508229 ] ] [ http://inspirehep.net/record/690379[inspire ] ]
. j. jalilian - marian and y. v. kovchegov , _ inclusive two - gluon and valence quark - gluon production in dis and pa _ , http://dx.doi.org/10.1103/physrevd.71.079901 , 10.1103/physrevd.70.114017[_phys.rev . _ * d70 * ( 2004 ) 114017 ] [ http://arxiv.org/abs/hep-ph/0405266[hep-ph/0405266 ] ] [ http://inspirehep.net/record/651090[inspire ] ] .
f. dominguez , c. marquet , b .- w .
xiao , and f. yuan , _ universality of unintegrated gluon distributions at small x _ , http://dx.doi.org/10.1103/physrevd.83.105005[_phys.rev._ * d83 * ( 2011 ) 105005 ] [ http://arxiv.org/abs/1101.0715 [ arxiv:1101.0715 ] ] [ http://inspirehep.net/record/883101 [ inspire ] ] .
i. balitsky , _ operator expansion for high - energy scattering _ , http://dx.doi.org/10.1016/0550-3213(95)00638-9[_nucl.phys._ * b463 * ( 1996 ) 99160 ] [ http://arxiv.org/abs/hep-ph/9509348[hep-ph/9509348 ] ] [ http://inspirehep.net/record/399690[inspire ] ] . j. jalilian - marian , a. kovner , and h. weigert , _ the wilson renormalization group for low x physics : gluon evolution at finite parton density _ , http://dx.doi.org/10.1103/physrevd.59.014015[_phys.rev._ * d59 * ( 1998 ) 014015 ] [ http://arxiv.org/abs/hep-ph/9709432 [ hep - ph/9709432 ] ] [ http://inspirehep.net/record/448750 [ inspire ] ] .
e. iancu , a. leonidov , and l. d. mclerran , _ nonlinear gluon evolution in the color glass condensate
_ , http://dx.doi.org/10.1016/s0375-9474(01)00642-x[_nucl.phys._ * a692 * ( 2001 ) 583645 ] [ http://arxiv.org/abs/hep-ph/0011241[hep-ph/0011241 ] ] [ http://inspirehep.net/record/537171[inspire ] ] . y. v. kovchegov , _ small - x @xmath60 structure function of a nucleus including multiple pomeron exchanges _ , http://dx.doi.org/10.1103/physrevd.60.034008[_phys.rev._ * d60 * ( 1999 ) 034008 ] [ http://arxiv.org/abs/hep-ph/9901281 [ hep - ph/9901281 ] ] [ http://inspirehep.net/record/493965 [ inspire ] ] .
. ma and r. venugopalan _ in preparation _ .
a. dumitru , j. jalilian - marian , t. lappi , b. schenke , and r. venugopalan , _ renormalization group evolution of multi - gluon correlators in high energy qcd _ , http://dx.doi.org/10.1016/j.physletb.2011.11.002[_phys.lett._ * b706 * ( 2011 ) 219224 ] [ http://arxiv.org/abs/1108.4764[arxiv:1108.4764 ] ] [ http://inspirehep.net/record/924854[inspire ] ] . l. d. mclerran and r. venugopalan , _ computing quark and gluon distribution functions for very large nuclei _ , http://dx.doi.org/10.1103/physrevd.49.2233[_phys.rev._ * d49 * ( 1994 ) 22332241 ] [ http://arxiv.org/abs/hep-ph/9309289[hep-ph/9309289 ] ] [ http://inspirehep.net/record/358358[inspire ] ] .
l. d. mclerran and r. venugopalan , _ gluon distribution functions for very large nuclei at small transverse momentum _ , http://dx.doi.org/10.1103/physrevd.49.3352[_phys.rev._ * d49 * ( 1994 ) 33523355 ] [ http://arxiv.org/abs/hep-ph/9311205[hep-ph/9311205 ] ] [ http://inspirehep.net/record/359612[inspire ] ] . j. l. albacete , a. dumitru , h. fujii , and y. nara , _ cgc predictions for p+pb collisions at the lhc _ , http://dx.doi.org/10.1016/j.nuclphysa.2012.09.012[_nucl.phys._ * a897 * ( 2013 ) 127 ] [ http://arxiv.org/abs/1209.2001[arxiv:1209.2001 ] ] [ http://inspirehep.net/record/1184977[inspire ] ] . j. l. albacete and c. marquet , _ gluon saturation and initial conditions for relativistic heavy ion collisions _ , http://dx.doi.org/10.1016/j.ppnp.2014.01.004[_prog.part.nucl.phys._ * 76 * ( 2014 ) 142 ] [ http://arxiv.org/abs/1401.4866[arxiv:1401.4866 ] ] [ http://inspirehep.net/record/1278217[inspire ] ] .
s. catani , m. ciafaloni , and f. hautmann , _ gluon contributions to small x heavy flavor production _ , http://dx.doi.org/10.1016/0370-2693(90)91601-7[_phys.lett._ * b242 * ( 1990 ) 97 ] [ http://inspirehep.net/record/295304[inspire ] ] .
s. catani , m. ciafaloni , and f. hautmann , _ high - energy factorization and small x heavy flavor production _ , http://dx.doi.org/10.1016/0550-3213(91)90055-3[_nucl.phys._ * b366 * ( 1991 ) 135188 ] [ http://inspirehep.net/record/302969[inspire ] ] .
ma , k. wang , and k .-
@xmath104 production at the tevatron and lhc at @xmath105 in nonrelativistic qcd _ , http://dx.doi.org/10.1103/physrevlett.106.042002[_phys.rev.lett._ * 106 * ( 2011 ) 042002 ] [ http://arxiv.org/abs/1009.3655[arxiv:1009.3655 ] ] [ http://inspirehep.net/record/868936[inspire ] ] .
m. butenschoen and b. a. kniehl , _ reconciling @xmath1 production at hera , rhic , tevatron , and lhc with nrqcd factorization at next - to - leading order _ , http://dx.doi.org/10.1103/physrevlett.106.022003[_phys.rev.lett._ * 106 * ( 2011 ) 022003 ] [ http://arxiv.org/abs/1009.5662[arxiv:1009.5662 ] ] [ http://inspirehep.net/record/871033[inspire ] ] .
j. pumplin , d. stump , j. huston , h. lai , p. m. nadolsky , _ et al .
_ , _ new generation of parton distributions with uncertainties from global qcd analysis _ , http://dx.doi.org/10.1088/1126-6708/2002/07/012[_jhep_ * 0207 * ( 2002 ) 012 ] [ http://arxiv.org/abs/hep-ph/0201195 [ hep - ph/0201195 ] ] [ http://inspirehep.net/record/581996 [ inspire ] ] .
a. caldwell and h. kowalski , _ investigating the gluonic structure of nuclei via @xmath1 scattering _ , http://dx.doi.org/10.1103/physrevc.81.025203[_phys.rev._ * c81 * ( 2010 ) 025203 ] [ http://inspirehep.net/record/850035 [ inspire ] ] .
, a. adare _ et al .
_ , _ ground and excited charmonium state production in @xmath106 collisions at @xmath107 gev _ , http://dx.doi.org/10.1103/physrevd.85.092004[_phys.rev._ * d85 * ( 2012 ) 092004 ] [ http://arxiv.org/abs/1105.1966 [ arxiv:1105.1966 ] ] [ http://inspirehep.net/record/899065 [ inspire ] ] .
, d. acosta _ et al .
_ , _ measurement of the @xmath1 meson and @xmath108hadron production cross sections in @xmath109 collisions at @xmath110 gev _ , http://dx.doi.org/10.1103/physrevd.71.032001[_phys.rev._ * d71 * ( 2005 ) 032001 ] [ http://arxiv.org/abs/hep-ex/0412071 [ hep - ex/0412071 ] ] [ http://inspirehep.net/record/668026 [ inspire ] ] .
, b. abelev _ et al .
_ , _ inclusive @xmath1 production in @xmath111 collisions at @xmath112 tev _ , http://dx.doi.org/10.1016/j.physletb.2012.10.078[_phys.lett._ * b718 * ( 2012 ) 295306 ] [ http://arxiv.org/abs/1203.3641[arxiv:1203.3641 ] ] [ http://inspirehep.net/record/1094079[inspire ] ] .
, k. aamodt _ et al .
_ , _ rapidity and transverse momentum dependence of inclusive @xmath1 production in @xmath111 collisions at @xmath113 tev _ , http://dx.doi.org/10.1016/j.physletb.2011.09.054 , 10.1016/j.physletb.2012.10.060[_phys.lett .
_ * b704 * ( 2011 ) 442455 ] [ http://arxiv.org/abs/1105.0380 [ arxiv:1105.0380 ] ] [ http://inspirehep.net/record/897764 [ inspire ] ] .
chang , _ hadronic production of @xmath1 associated with a gluon _ , http://dx.doi.org/10.1016/0550-3213(80)90175-3[_nucl.phys._ * b172 * ( 1980 ) 425434 ] [ http://inspirehep.net/record/143086[inspire ] ] .
r. baier and r. ruckl , _ hadronic production of @xmath1 and @xmath114 : transverse momentum distributions _ , http://dx.doi.org/10.1016/0370-2693(81)90636-5[_phys.lett._ * b102 * ( 1981 ) 364 ] [ http://inspirehep.net/record/164479[inspire ] ] . , b. b. abelev _
_ , _ measurement of quarkonium production at forward rapidity in pp collisions at @xmath8= 7 tev _ ,
[ http://arxiv.org/abs/1403.3648 [ arxiv:1403.3648 ] ] [ http://inspirehep.net/record/1285950 [ inspire ] ] . , v. khachatryan _ et al .
_ , _ prompt and non - prompt @xmath1 production in @xmath111 collisions at @xmath115 tev _ , http://dx.doi.org/10.1140/epjc/s10052-011-1575-8[_eur.phys.j._ * c71 * ( 2011 ) 1575 ] [ http://arxiv.org/abs/1011.4193[arxiv:1011.4193 ] ] [ http://inspirehep.net/record/878118[inspire ] ] . , r. aaij _ et al .
_ , _ measurement of @xmath1 production in @xmath111 collisions at @xmath116 _ , http://dx.doi.org/10.1140/epjc/s10052-011-1645-y[_eur.phys.j._ * c71 * ( 2011 ) 1645 ] [ http://arxiv.org/abs/1103.0423[arxiv:1103.0423 ] ] [ http://inspirehep.net/record/891233[inspire ] ] .
, g. aad _ et al .
_ , _ measurement of the differential cross - sections of inclusive , prompt and non - prompt @xmath1 production in proton - proton collisions at @xmath115 tev _ , http://dx.doi.org/10.1016/j.nuclphysb.2011.05.015[_nucl.phys._ * b850 * ( 2011 ) 387444 ] [ http://arxiv.org/abs/1104.3038[arxiv:1104.3038 ] ] [ http://inspirehep.net/record/896268[inspire ] ] .
, l. adamczyk _ et al .
_ , _ @xmath1 production at high transverse momenta in @xmath106 and au+au collisions at @xmath117 gev _ , http://dx.doi.org/10.1016/j.physletb.2013.04.010[_phys.lett._ * b722 * ( 2013 ) 5562 ] [ http://arxiv.org/abs/1208.2736[arxiv:1208.2736 ] ] [ http://inspirehep.net/record/1127499[inspire ] ] . , r. aaij _ et al .
_ , _ measurement of @xmath100 meson production in pp collisions at sqrt(s)=7 tev _ , http://dx.doi.org/10.1140/epjc/s10052-012-2100-4[_eur.phys.j._ * c72 * ( 2012 ) 2100 ] [ http://arxiv.org/abs/1204.1258[arxiv:1204.1258 ] ] [ http://inspirehep.net/record/1104754[inspire ] ] .
ma , k. wang , and k .-
chao , _ a complete nlo calculation of the @xmath1 and @xmath118 production at hadron colliders _ , http://dx.doi.org/10.1103/physrevd.84.114001[_phys.rev._ * d84 * ( 2011 ) 114001 ] [ http://arxiv.org/abs/1012.1030 [ arxiv:1012.1030 ] ] [ http://inspirehep.net/record/879577 [ inspire ] ] .
a. m. stasto , b .- w .
xiao , f. yuan , and d. zaslavsky , _ matching collinear and small - x factorization calculations for inclusive hadron production in pa collisions _ , http://dx.doi.org/10.1103/physrevd.90.014047[_phys.rev._ * d90 * ( 2014 ) 014047 ] [ http://arxiv.org/abs/1405.6311 [ arxiv:1405.6311 ] ] [ http://inspirehep.net/record/1298042 [ inspire ] ]
. f. gelis and r. venugopalan , _ large mass @xmath119 production from the color glass condensate _ , http://dx.doi.org/10.1103/physrevd.69.014019[_phys.rev._ * d69 * ( 2004 ) 014019 ] [ http://arxiv.org/abs/hep-ph/0310090 [ hep - ph/0310090 ] ] [ http://inspirehep.net/record/630106 [ inspire ] ] .
e. l. berger , j. qiu , and y. wang , _ transverse momentum distribution of @xmath114 production in hadronic collisions _ , http://dx.doi.org/10.1103/physrevd.71.034007[_phys.rev._ * d71 * ( 2005 ) 034007 ] [ http://arxiv.org/abs/hep-ph/0404158 [ hep - ph/0404158 ] ] [ http://inspirehep.net/record/648526 [ inspire ] ] .
p. sun , c .- p .
yuan , and f. yuan , _ heavy quarkonium production at low pt in nrqcd with soft gluon resummation _ , http://dx.doi.org/10.1103/physrevd.88.054008[_phys.rev._ * d88 * ( 2013 ) 054008 ] [ http://arxiv.org/abs/1210.3432 [ arxiv:1210.3432 ] ] [ http://inspirehep.net/record/1190526 [ inspire ] ] .
qiu , p. sun , b .- w .
xiao , and f. yuan , _ universal suppression of heavy quarkonium production in pa collisions at low transverse momentum _ , http://dx.doi.org/10.1103/physrevd.89.034007[_phys.rev._ * d89 * ( 2014 ) 034007 ] [ http://arxiv.org/abs/1310.2230 [ arxiv:1310.2230 ] ] [ http://inspirehep.net/record/1257636 [ inspire ] ]
. s. j. brodsky and j .- p .
lansberg , _ heavy - quarkonium production in high energy proton - proton collisions at rhic _ , http://dx.doi.org/10.1103/physrevd.81.051502[_phys.rev._ * d81 * ( 2010 ) 051502 ] [ http://arxiv.org/abs/0908.0754 [ arxiv:0908.0754 ] ] [ http://inspirehep.net/record/827958 [ inspire ] ]
. j. lansberg , _
total @xmath1 and @xmath114 production cross section at the lhc : theory vs. experiment _ , _ pos _ * ichep2010 * ( 2010 ) 206 [ http://arxiv.org/abs/1012.2815[arxiv:1012.2815 ] ] [ http://inspirehep.net/record/881402[inspire ] ] .
r. nelson , r. vogt , and a. frawley , _ narrowing the uncertainty on the total charm cross section and its effect on the @xmath1 cross section _ , http://dx.doi.org/10.1103/physrevc.87.014908[_phys.rev._ * c87 * ( 2013 ) 014908 ] [ http://arxiv.org/abs/1210.4610 [ arxiv:1210.4610 ] ] [ http://inspirehep.net/record/1191030 [ inspire ] ] . | we employ a small @xmath0 color glass condensate ( cgc)+ non - relativistic qcd ( nrqcd ) formalism to compute @xmath1 production at low @xmath2 in proton - proton collisions at collider energies .
very good agreement is obtained for total cross - sections , rapidity distributions and low momentum @xmath2 distributions .
similar agreement is obtained for @xmath3 production .
we observe an overlap region in @xmath2 where our results match smoothly to those obtained in a next - to - leading order ( nlo ) collinearly factorized nrqcd formalism .
the relative contribution of color singlet and color octet contributions can be quantified in the cgc+nrqcd framework , with the former contributing approximately @xmath4 of the total cross - section .
the production of heavy quarkonium states is an excellent laboratory to understand hadronization in qcd @xcite .
a significant development was the non - relativistic qcd ( nrqcd ) factorization formalism @xcite that provides systematic estimates of the magnitudes of universal long distance matrix elements ( ldmes ) contributing to cross - sections of heavy quarkonium states . based on this nrqcd framework ,
next - to - leading order ( nlo ) computations performed in recent years @xcite describe nearly all the high @xmath2 ( @xmath5 , where @xmath6 is the quarkonium mass ) data on heavy quarkonium production at hadron colliders , with further improvements anticipated from high @xmath2 logarithmic resummations @xcite in contrast , heavy quarkonium production in the low @xmath7 region is far from understood .
this regime dominates the total cross - section for production of heavy quarkonia at colliders . for collider center of mass ( c.m ) energies @xmath8 , the dynamics is sensitive to large logarithms in @xmath9 .
summing these small @xmath0 logs leads to the phenomenon of gluon saturation , characterized by a dynamically generated semi - hard scale @xmath10 in the hadron wavefunctions @xcite . recently in @xcite , the color class condensate ( cgc )
effective theory of small @xmath0 qcd @xcite results for the short distance heavy quark pair production cross - section @xcite were combined with the ldmes of nrqcd to provide analytic expressions for a large number of quarkonium final states . in this letter
, we will provide first quantitative results in this novel cgc+nrqcd framework for @xmath1 production in proton - proton ( p+p ) collisions at collider energies .
we show that results in this framework can be matched smoothly to nlo perturbative qcd results at high @xmath2 , thereby providing the missing link for a unified description for quarkonium production in all phase space .
this unified framework is identical for the description of both p+p and proton - nucleus ( p+a ) collisions . with some numerical effort
, it can be extended systematically @xcite to quarkonium production in the early stages of nucleus - nucleus ( a+a ) collisions , where quarkonium dissociation is a signature of quark - gluon plasma formation .
a further nice feature of this framework is that allows one to quantify the relative contribution of color singlet and color octet channels to quarkonium final states in both p+p and p+a collisions . in the nrqcd factorization formalism @xcite , the inclusive production of a heavy quarkonium state @xmath11
is expressed as @xmath12 where @xmath13}}}$ ] are the quantum numbers of the produced intermediate heavy quark pair with standard spectroscopic notation for the angular momentum , and the superscript @xmath14 denotes the color state of the pair , which can be either color singlet ( cs ) with @xmath15 or color octet ( co ) with @xmath16 . for @xmath17 production , the primary focus here ,
the most important intermediate states are @xmath18}}}}$ ] , @xmath19}}}}$ ] , @xmath20}}}}$ ] and @xmath21}}}}$ ] . in eq .
, @xmath22 are non - perturbative but universal nrqcd ldmes , which can be extracted from data .
the @xmath23 denote the short distance coefficients for the production of a heavy quark pair with quantum number @xmath24 .
based on the heavy quark pair production amplitude in @xcite , @xmath23 have been calculated for all @xmath25-wave and @xmath26-wave intermediate states in @xcite . in the following , we will give general expressions for dilute - dense scattering " .
in the cgc power counting @xcite , this could represent either p+a collisions or forward p+p collisions . at collider energies , @xmath0 in the forward going
dilute " proton may be small enough that small @xmath0 resummation is relevant , but phase space densities are still small .
conversely , phase space densities in the dense " backward going proton can reach their maximal value of @xmath27 , indicating gluon saturation . because large @xmath0 modes are probed at high @xmath28
, there is a limited range in @xmath2 where the cgc effective theory is valid . at such large @xmath2 , as we shall discuss further , perturbative computations in the collinear factorization framework should be more reliable . in this
dilute dense cgc framework , we can express @xmath23 in the color octet channel as @xcite @xmath29 when @xmath24 is co , @xmath30 , @xmath31 ( @xmath32 ) are transverse momentum ( rapidity ) of produced heavy quarkonium , @xmath33 [ @xmath34 is the rapidity of gluons coming from dilute proton [ dense proton ] .
the expression for @xmath35 can be found in @xcite .
further , @xmath36 where @xmath37\right\rangle_y / n_c$ ] is the dipole forward scattering amplitude , @xmath38 are lightlike wilson lines in the fundamental representation , and @xmath39 the number of colors . here
@xmath40 represents the renormalization group evolved expectation value of this correlator in the target background field evaluated at rapidity @xmath41 . for more details ,
we refer the reader to @xcite .
the unintegrated gluon distribution inside the proton is then expressed as @xmath42 where @xmath43 is the effective transverse area of the proton , and @xmath44 is similar to @xmath45 in eq . but with its wilson lines in the adjoint representation .
interestingly , if @xmath24 is cs , @xmath23 depends on both dipole correlators and novel quadrupole correlators @xcite .
these latter are defined as @xmath46\right\rangle_y / n_c\,$ ] . like the dipole correlators , they are universal gauge invariant quantities , and appear in a number of final states @xcite .
the energy evolution of both dipole and quadrupole correlators is described by the balitsky - jimwlk hierarchy of small @xmath0 rg equations @xcite .
for the dipole correlator , in the large @xmath39 limit , the hierarchy has a closed form expression , the well known balitsky - kovchegov ( bk ) equation @xcite .
this equation can be solved numerically and is widely used in phenomenological applications .
in contrast , no such simple form exists for the quadrupole correlator and solving the corresponding balitsky - jimwlk equation is cumbersome for phenomenological applications .
while the expressions simplify in a quasi - classical approximation in the large @xmath39 limit @xcite , the result is still too complicated for our purposes . for our study here
, we discovered an approximate factorized expression for the quadrupole correlator , @xmath47 this result is exact when any two adjacent positions coincide : @xmath48 , @xmath49 , @xmath50 or @xmath51 .
we have checked @xcite that it is a good approximation to the balitsky - jimwlk results in @xcite . making use of eq .
, the short distance coefficients for the cs channels in @xcite simplify significantly to @xmath52 where @xmath53 has a different functional form from @xmath54 in @xcite because of the above simplification .
in particular , for the @xmath18}}}}$ ] channel , @xmath55}}}}}_1=&\frac{k_{1\perp}^2\left ( k_{1\perp}^2 + 4m^2\right)}{12 m}\\ & \hspace{-0.5cm}\times\left(\frac{1}{l_{\perp}^2+k_{1\perp}^2/4+m^2}-\frac{1}{{l'}_{\perp}^2+k_{1\perp}^2/4+m^2}\right)^2 , \end{split}\end{aligned}\ ] ] with @xmath56 . the key ingredient in both eqs . ( [ eq : dsktco ] ) and ( [ eq : dsktcs ] ) is @xmath57 , which is obtained by solving the running coupling bk ( rcbk ) equation directly in momentum space .
we use mclerran - venugopalan ( mv ) initial conditions @xcite for the dipole amplitude at the initial rapidity scale @xmath58 ( with @xmath59 ) for small @xmath0 evolution .
since the structure function @xmath60 in deeply inelastic scattering ( dis ) is directly propotional ( at small @xmath0 ) to the dipole amplitude , all the parameters in the rcbk equation are fixed from fits to the hera dis data @xcite .
other initial conditions discussed in the literature @xcite give similar results at low @xmath2 for heavy quark pair production but differ significantly at high @xmath2 where the present formalism is not reliable no matter which initial condition is used . while the formalism captures the dominant higher order corrections at small @xmath0 @xcite ,
this is less assured as one goes to larger @xmath0 with increasing @xmath2 .
because nlo cgc expressions for eqs . and that extend their @xmath2 range are not available , we choose to restrict data comparisons to the more reliable small to moderate @xmath2 region .
we shall explore whether there is a @xmath2 range where our results overlap with the successful high @xmath2 nlo collinear factorization approach @xcite .
a similar issue of matching to the collinear framework occurs when @xmath61 is accessed respectively in the forward or backward going protons .
since the rcbk equation is solved only for @xmath62 , we devised a scheme to extrapolate the dipole amplitude to larger @xmath0 , using results for the collinear factorized gluon distributions ( pdfs ) . from the relation between the pdfs and the unintegrated gluon distribution @xcite , using eq .
, one obtains @xmath63 where @xmath64 and @xmath65 is an extra free function that we will now determine .
we assume @xmath66 when @xmath67 is close to @xmath68 , such that @xmath69 and @xmath70 .
these two conditions result in two equations , which allow us to simultaneously determine @xmath71 and @xmath72 .
if we use the cteq6 m pdfs @xcite for @xmath73 and the rcbk equation with the mv initial condition to calculate @xmath74 , we determine @xmath75 and @xmath76 fm .
( interestingly , this value for @xmath71 is in very good agreement with the transverse gluon radius of the proton extracted from hera diffractive data @xcite . )
having fixed @xmath71 and @xmath77 with our matching conditions , we use the extrapolation @xmath78 with @xmath65 determined to be @xmath79^{-1}\hspace{-0.2cm}.\end{aligned}\ ] ] the dilute - dense approximation employed here will break down when the phase space densities in both protons are large .
this may be the case at central rapidities , especially at the lhc . in the cgc framework , quark pair production in this dense - dense regime can not be computed analytically but may be feasible numerically .
this is however beyond the scope of the present work and is a further source of systematic uncertainty . before we confront our framework to the data , we need to fix the charm quark mass and determine the ldmes .
we set the charm quark mass to be @xmath80 , approximately one half the @xmath1 mass .
the co ldmes were extracted in the nlo collinear factorized nrqcd formalism @xcite by fitting tevatron high @xmath2 prompt @xmath17 production data ; this gives @xmath81}}}})\rangle}}=1.16/(2n_c ) { \mathrm{~gev}}^3 $ ] , @xmath82}}}})\rangle}}=0.089\pm0.0098 { \mathrm{~gev}}^3 $ ] , @xmath83}}}})\rangle}}=0.0030\pm0.0012 { \mathrm{~gev}}^3 $ ] and @xmath84}}}})\rangle}}=0.0056\pm0.0021{\mathrm{~gev}}^3 $ ] .
note that the high sensitivity of short distance cross - sections to quark mass is mitigated by the mass dependence of the ldmes .
the latter are extracted from particular data sets using eq .
( [ eq : factorization ] ) . for a different quark mass
, the ldmes extracted would be quite different @xcite , thereby ensuring only a weak quark mass uncertainty to nrqcd predictions .
we also note that while other nlo nrqcd fits exist , they are not suitable for our use .
for example , the fit in @xcite relies on a cancellation between the @xmath20}}}}$ ] channel and @xmath21}}}}$ ] channels , which occurs only at nlo where short distance coefficients can be negative .
to account for the uncertainties outlined , as well as higher order in @xmath85 corrections , we introduce a systematic uncertainty of @xmath86 on top of the statistical uncertainties from the ldmes .
total production cross section at hadron colliders in the cgc+nrqcd framework compared to data at midrapidity .
the data are from @xcite . ] in fig .
[ fig : ss ] , we compare our results for the @xmath17 total cross section at mid - rapidity for @xmath87 from @xmath88 to @xmath89 . in general ,
our predictions are consistent with the collider data . for the top rhic energy ,
our result is a little larger then the phenix data .
however , ( see fig . [
fig : pt ] ) , we are able to describe the phenix data on the @xmath2 dependence at forward rapidity . the small discrepancy with phenix data at central rapidities may reflect the fact that our formalism is most reliable for large c.m energy or for forward rapidities . as stated previously , our computation is the first that includes contributions from both the color singlet and color octet channels to the total cross - section .
in contrast to the color singlet model ( csm ) @xcite prediction , we find that in our formalism the cs contribution is only 10% of the total cross section .
our estimates show that for a large nucleus this relative contribution will rise to approximately 15%- 20% .
differential cross section as a function of rapidity at lhc .
data from @xcite . ] our results for the differential cross section as a function of rapidity compared to @xmath89 data at lhc are shown in fig .
[ fig : y ] .
all the data lie well within the error band . at the larger rapidities ,
the results become sensitive to @xmath90 in either the projectile or target ( and conversely very small @xmath0 in the target or projectile ) . at these forward ( or backward ) rapidities , as discussed previously , the matching of the dipole amplitude to the unintegrated pdfs at large @xmath0 is important .
note that this matching condition allowed us to fix the radius @xmath71 which was the only normalization parameter in eqs .
( [ eq : dsktco ] ) and ( [ eq : dsktcs ] ) the apparent @xmath85 dependence cancels out when eq .
( [ eq : unintegrated ] ) is substituted in these equations .
it is therefore striking that our results explain both the overall normalization at @xmath91 and the relative normalization at forward rapidities after matching the pdfs smoothly to the dipole amplitude at @xmath92 .
( top curve ) and @xmath17 ( other four curves ) differential cross section as a function of @xmath2 .
data from @xcite .
nlo nrqcd predictions are taken from @xcite . ] in fig .
[ fig : pt ] , we compare our results for @xmath1 differential cross section as a function of @xmath2 with experimental data at several c.m energies and rapidity regions .
it is clear that small @xmath2 data is well described by our cgc+nrqcd formalism .
further , as anticipated in the previous discussion , the results in this formalism begin to disagree with data at higher @xmath2 . from previous experience with single inclusive lhc data @xcite ,
one expects to overshoot the data with mv initial conditions this is precisely what we see . in fig .
[ fig : pt ] , we also show the nlo collinear factorized nrqcd results @xcite which show good agreement with data at large @xmath2 .
it is very interesting to observe an overlap region around @xmath93 , which can be described by both the cgc+nrqcd formalism and the nlo collinearly factorized nrqcd formalism .
a good matching of the small @xmath0 and collinear factorized formalisms at large @xmath2 is seen in single inclusive hadron production by imposing exact kinematic constraints in the small @xmath0 formalism @xcite . since leading
order collinear factorization results for heavy quark pair production are obtained as a limit of the cgc result @xcite , imposing exact kinematics may help better understand the overlap between the two formalisms . in the low @xmath2 region ,
a further refinement of our formalism will include resummation of logarithms of @xmath94 for @xmath95 @xcite .
most of the experimental data presented are for inclusive @xmath17 production .
these include @xmath1 s produced from @xmath96-meson decays as well as prompt production of @xmath1 s .
the latter includes feeddown from higher excited charmonium states as well as direct @xmath17 production .
however we only considered direct @xmath17 production contribution in our theory results .
nevertheless , the comparison is meaningful .
firstly , the @xmath96-meson decay contribution in the small @xmath2 region is small , of order less than 10% .
secondly , the ldmes in @xcite are obtained by fitting prompt @xmath17 data .
thus feeddown contributions are already roughly estimated in our results . with the expressions in @xcite for the higher charmonium states , a fully consistent treatment of prompt
@xmath1 production data is feasible in the near future . as a first step , we compare our results for the @xmath3 differential cross section as a function of @xmath2 with data in fig .
[ fig : pt ] . in this comparison
, we set the charm quark mass to be @xmath97 and used the co ldmes extracted in @xcite .
theory and data agree well . however , if we set @xmath80 for @xmath3 , the results overshoot the data . as noted
, one anticipates the sensitivity to the quark mass in the short distance cross sections to be offset by their dependence in the ldmes .
however , for @xmath3 , only two linear combinations of the three co ldmes of @xmath98 can be determined @xcite ; because it is unconstrained , we set @xmath99}}}})\rangle}}=0 $ ] here . thus @xmath3 has larger systematic uncertainties relative to @xmath17 that subsume the sensitivity of results to the quark mass .
based on our present work , all three @xmath98 co ldmes can be determined from a global fit .
a consistent treatment of all charmonium and bottomonium states is in progress @xcite . in the latter case
, we anticipate a larger contribution from the logarithmic resummation of @xcite .
comparing our results to related work , @xmath17 total cross sections were studied recently in the csm within the broad framework of collinear factorization @xcite .
while the trend of the total cross section is described , the theory uncertainties have to be as large as a factor of 10 for data to lie within errors .
likewise , the cem with collinear factorization @xcite describes the total cross section , rapidity distributions and @xmath2 distributions with much larger uncertainties than our framework .
the results presented here are timely given the wealth of data extant and anticipated from rhic to lhc for p+p and p+a collisions .
when combined with developments in collinearly factorized nrqcd , they set the groundwork for a comprehensive understanding of the mechanism of heavy quark hadronization in qcd . |
quantum entanglement of a system quantifies the correlations between the parts of the system@xcite , which serves as a resource for quantum information processing tasks . the block entanglement , viz .
the entropy of a subsystem , is a widely - used entanglement measure , that has been used to investigate critical behaviour near quantum phase transitions in spin systems@xcite .
but there are not many studies of the entanglement in interacting electron systems , which exhibit substantially richer structure than interacting spin systems as they carry additional charge degrees of freedom . in this article
, we study the strong correlation effect on the entanglement spectrum of the one - dimensional gutzwiller state@xcite , as a proto - type strongly - correlated state . [ [ the - gutzwiller - state ] ] the gutzwiller state : + + + + + + + + + + + + + + + + + + + + + + the gutzwiller state was initially suggested as a variational ground state for the hubbard model @xcite . in this state ,
the strong on - site correlation effect of the hubbard model ground state is mimicked by applying a projection operator on the non - interacting metallic state to decrease the double occupancy . at one end of the control
parameter is the metallic state with no projection , viz .
a fermi ground state constructed from occupying lowest - lying one - electron plane - wave states for both up and down spin electrons .
the metallic state maximises the double occupancy as there is no correlation between the up and down spin electron . at the other extreme
, there is an insulator phase , corresponding to the fully - projected state with no double occupancy .
the gutzwiller state for a lattice of @xmath5 sites is given by @xmath6 where @xmath7 is the metallic fermi state constructed from the vacuum state by using electron creation operators @xmath8 with a momentum @xmath9 and spin @xmath10 , and @xmath0 is a parameter taking values from @xmath11 to @xmath12 , @xmath13 being the non - interacting case , and @xmath14 being the limit of infinite interactions .
the filling factor is determined by @xmath15,the fermi momentum . while the gutzwiller state gives good agreement with hubbard model ground state for three dimensions , in 1d , it is different from the ground state of the hubbard model . at half - filling ,
the ground state of the hubbard model in 1d describes a luttinger liquid , whereas the gutzwiller state shows fermi liquid behaviour @xcite . in the thermodynamic limit , for any @xmath16
, the momentum space distribution of electrons for the 1d half - filled gutzwiller state has a discontinuity at the fermi wave vector .
therefore , in the thermodynamic limit , the system is metallic for any @xmath16 .
also , in 1d , the half - filled gutzwiller state for @xmath14 is the exact ground state of the haldane - shastry model @xcite , which is a heisenberg - like model with long range interactions .
hence the gutzwiller state is , by itself , an interesting correlated electron state . here
we will be interested in the entanglement spectrum of the half - filled gutzwiller state .
[ [ previous - results - entanglement - entropy - fluctuations - and - metal - insulator - crossover ] ] previous results : entanglement entropy , fluctuations and metal - insulator crossover : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the block entanglement entropy and bipartite fluctuations of the gutzwiller state in 1d for half bipartition has been recently studied as a function of the correlation factor @xmath0 and the number of sites @xmath5 @xcite.the block entanglement entropy @xmath17 for half - bipartition of half - filled gutzwiller state scales as : @xmath18 which has the same form as conformal field theory ( cft ) result for one dimensional systems except with an effective central charge @xmath19 which is a function of both @xmath0 and @xmath5 . at @xmath13 , @xmath20 and at @xmath14 , @xmath21 which are independent of @xmath5 and are the correct results in these two limits from cft .
the @xmath5 dependence occurs for intermediate values of @xmath0 .
the scaling of @xmath19 indicates a metal - insulator crossover .
bipartite fluctuations also show a scaling .
the scaling of both bipartite fluctuations and @xmath19 show that for @xmath22 , the relevant scaling variable is @xmath23 and metal - insulator crossover occurs at @xmath24 .
the reason behind deviation from the the cft result for intermediate values of @xmath0 is the existence of a finite correlation length .
as shown in ref @xcite , the correlation length between opposite spins is infinte at @xmath14 , but decreases with @xmath0 and becomes @xmath11 for @xmath13 , whereas the correlation length between same spins always remain infinite .
hence , for intermediate values of @xmath0 , there is still long range correlation in the system , but the system is no longer scale invariant because of the finite correlation length between opposite spins .
so , cft result for entanglement entropy with long range correlations is not valid , and also area law of entanglement entropy is not valid .
we see that , in such case , the system still retains the @xmath25 divergence of the cft result for long range correlations , but with an effective central charge which is depends on @xmath5 . for @xmath14 and @xmath13 , the system is scale invariant and the cft result for entanglement entropy with long range correlations holds with the appropriate central charge .
it is important to note that the deviation from cft result is only a finite size effect and will not be observed if the system is thermodynamically large .
this is because the two relevant length scales in the system are the finite correlation length and the system size and the physics is governed by their ratio . in the thermodynamic limit , for any value of the finite correlation length , this ratio is @xmath11 , and so , the system behaves like it is at @xmath13 .
hence , in thermodynamic limit , the cft result holds with central charge @xmath26 for @xmath16 , and with @xmath27 for @xmath14 .
this is consistent with the known description for gutzwiller state in thermodynamic limit , which says , it is metallic for all @xmath28 . for , finite size
, however , the system shows a metal - insulator crossover when the ratio of the finite correlation length and system size is of order one , which occurs for a finite value of @xmath0 . for @xmath22 , this value of @xmath0
is given by @xmath29 .
physical systems in this size range which may be described by the gutzwiller state in one dimension are nanochains of strongly correlated metals .
[ [ entanglement - spectrum ] ] entanglement spectrum : + + + + + + + + + + + + + + + + + + + + + + all the above previous results are based on the study of entanglement entropy and bipartite fluctuations . while entanglement entropy is a number characterising the correlations and entanglement in the system , the entanglement spectrum , which is the full spectrum of eigenvalues of the reduced density matrix , is more fundamental .
more information regarding correlations and entanglement is expected to be there in the entanglement spectrum .
also , in numerical techniques based on matrix product states , like the density matrix renormalization group , all observables are calculated from a truncated spectrum of the reduced density matrix .
thus study of full spectrum of the reduced density matrix also helps in determining convergence and accuracy for such numerical schemes . in this paper ,
we investigate the spectral features of the reduced density matrix of the finite size 1d half - filled gutzwiller state for half bipartition .
entanglement spectrum has been previously studied for 1d systems within the framework of cft @xcite .
it is interesting to see how these results are modified for our case , where we know cft is not valid .
the results are based on exact numerical schmidt decomposition of systems upto @xmath30 sites .
we consider a ring with @xmath5 electrons , with equal number of up and down spins , corresponding to the half - filling case . we partition the system into two equal parts , with the first @xmath31 sites belong to one subsystem . in the following ,
we keep the system size at @xmath30 and study various observables as a function of @xmath0 .
finite - size scaling is difficult to be implemented as the data is insufficient .
however , since , as described above , the correct scaling variable is a product of an increasing function of @xmath0 and an increasing function of @xmath5 , we expect that the variation of observables with @xmath0 for fixed @xmath5 is representative of variation of observables with @xmath5 for a fixed @xmath0 .
the variation of the product of the largest eigenvalue @xmath32 and rank @xmath4 of the reduced density matrix with on - site correlation factor @xmath0 for @xmath30 sites at half - filling is shown .
the inset shows the variation of @xmath32 with @xmath0 .
crossover to metallic limit is shown by the product becoming almost constant . ]
the dimension of the reduced density matrix of the subsystem with @xmath31 sites is @xmath33 , and in principle as many eigenvalues can be nonzero .
but , the entanglement entropy shows only a logarithmic divergence with @xmath5 , implying only o(@xmath5 ) number of nonzero eigenvalues .
thus the rank of the reduced density matrix is vastly reduced . since on - site
correlation factor @xmath0 reduces the probability of doubly - occupied sites , the rank is reduced as @xmath0 decreases .
reduction of the rank of the reduced density matrix is one of the reasons for a decrease of the entanglement as @xmath34 .
the largest eigenvalue is an important parameter governing the distribution of eigenvalues .
the range of the distribution of eigenvalues @xmath35 is determined by the largest eigenvalue @xmath32 . for gutzwiller state at half - filling , @xmath32 decreases as correlation factor @xmath0 increases .
this is because , as the rank of the reduced density matrix increases , the normalisation factor demands a decrease of @xmath32 . for a given rank @xmath4 ,
the minimum value of @xmath32 is @xmath36 , which corresponds to the maximally entangled state where all eigenvalues are equal .
thus minimum value of the product @xmath37 is @xmath12 .
hence as the product increases from @xmath12 , it gives a measure of how far the system is from the maximally entangled state of the corresponding rank . in fig.[fig :
emaxr ] , we see that the product @xmath38 is @xmath39 and increases rapidly from the insulating limit and saturates at the metallic limit at @xmath13 .
this means that the state is always far from the maximally entangled state .
but the strongly interacting case is closer to the maximally entangled state of the corresponding rank , than the metallic state , even though the metallic state has greater entanglement entropy ( see introduction , eq [ ee ] ) due to its greater rank .
the onset of the crossover to metallic region is seen by the product becoming almost constant .
the crossover region shown by this curve matches with that from the scaling form for the entanglement entropy@xcite .
hence distance from the maximally entangled state also characterises the crossover .
variation of ratio of number of distinct eigenvalues @xmath3 to rank @xmath4 with @xmath0 for @xmath30 sites at half - filling is shown .
it remains nearly constant , exhibiting a ` chaotic ' behaviour in the crossover region , but it decreases with @xmath0 close to @xmath14 and @xmath13 , showing an ` integrable ' behavior . ]
fraction of distinct eigenvalues can be used to track quantum chaos . in study of quantum chaos ,
a hamiltonian describing a chaotic system is one which can be modelled by a random matrix ensemble with the same degree of freedom as the hamiltonian , viz .
both have the same spacing distribution given by the wigner surmise ( the goe , gue or gse spacing distribution ) @xcite .
the wigner surmise gives a zero at zero spacing , which means there can be no degeneracy of eigenvalues .
for an integrable system , however , the spacing distribution is given by a poisson distribution , which has a peak at zero spacing , which means a large number of eigenvalues are degenerate .
if we have an integrable system and then introduce a symmetry breaking perturbation via a control parameter that makes it non - integrable , it has been recently shown @xcite that the spacing distribution changes from poisson to the wigner surmise . therefore , the fraction of distinct eigenvalues changes from less than one to one as the symmetry - breaking parameter is increased and then remains constant at one .
thus , the fraction of distinct eigenvalues remaining constant as the symmetry - breaking parameter is changed can be taken as a signature of quantum chaos . instead of the hamiltonian
, here we are concerned with the reduced density matrix of bipartition of the gutzwiller state , but , we have a similar situation . for @xmath14 and @xmath13 , because of scale invariance , the system is amenable to analytical methods via cft , and thus is ` integrable ' , whereas for intermediate values of @xmath0 , scale invariance is broken and it becomes ` non - integrable ' .
the parameter @xmath0 therefore acts like the symmetry - breaking parameter .
fig [ fig : n / r ] shows the variation of the fraction of distinct eigenvalues with @xmath0 for the reduced density matrix of half - filled gutzwiller state .
we see that for intermediate values of @xmath0 , the fraction of distinct eigenvalues remain almost constant .
this then may be a signature of quantum chaos .
however , this constant is much smaller than one .
this is because , even though scale invariance is broken , the symmetries and the block diagonal nature of the reduced density matrix preserves a lot of degeneracy of the eigenvalues . that this is indeed a signature of quantum chaos
will be made clear from the detailed analysis of the spacing distribution ( section [ spacing distribution ] . ) .
the detailed description of the various symmetries of the reduced density matrix will also be given while studying the spacing distribution .
near @xmath13 , the fraction of distinct eigenvalues decreases from the constant value because degeneracy increases , while the rank remains nearly the same . on the other hand , near @xmath14 , the fraction of distinct eigenvalues increases from the constant value because of a drastic decrease in rank .
next we look at the density distribution of non - zero eigenvalues .
the block entanglement entropy can be written , using the density of eigenvalues , as @xmath40 here , the integrated density of eigenvalues @xmath41 is given by @xmath42 calabrese and lefevre@xcite calculated this function for one - dimensional gapless systems using cft .
they started from the result obtained by scaling analysis for large subsystem size : @xmath43 where @xmath44 is the central charge of the cft , and @xmath45 is the subsystem size . taking the coefficient @xmath46 to be constant , they calcualted the density distribution of eigenvlaues of the reduced denstity matrix and from there obtained the result for the integrated density of eigenvalues as : @xmath47 where @xmath48 is modified bessel function of zeroth order
, @xmath32 is the largest eigenvalue .
various authors have shown @xcite : @xmath49 in the large @xmath45 limit .
also , following cardy and calabrese @xcite , entanglement entropy is given by : @xmath50 which is also true in the large @xmath45 limit .
above statement is just restatement of the result that entanglement entropy equals the single - copy entanglement in the large @xmath45 limit . in fig.[fig :
n(e)1](@xmath51 ) , the integrated eigenvalue density @xmath52 is plotted as a function of @xmath53 for @xmath30 sites for various values of on - site correlation factor @xmath0 , along with the corresponding plots of the cft result given above .
the cft result is seen to be correct only when @xmath14 and @xmath13 and not for intermediate values of @xmath0 .
this , of course , as described in introduction , is not surprising in light of previous work @xcite .
there it is shown that cft results are valid only at the two end points .
the reason is that it is only at the two end points that all underlying correlation lengths are infinite . for intermediate values of @xmath0
, there is a finite correlation length that comes from correlation between opposite spins .
hence the system does not remain scale invariant and finite - size corrections become important .
deviations from cft result signifies the crossover . however , the entanglement entropy retains cft - like scaling only with a different effective central charge which includes finite - size corrections .
so , we expect that upon replacing the central charge in above equations by the effective central charge will give the correct @xmath52 for all values of @xmath0 .
equivalently , we can just go back to the result @xmath54 .
so , the formula for integrated eigenvalue spectrum becomes : @xmath55 where the modulus is inserted because the term inside is no longer guaranteed to be positive .
we have shown in fig.[fig : n(e)1](@xmath56 ) the entanglement spectrum staircase plotted with the modified function given above .
it is seen that eq.[nu2 ] holds for all values of @xmath0 .
this confirms eq .
[ ralpha ] holds for intermidiate @xmath0 also but with an effective finite - size corrected central charge .
spacing distribution of eigenvalues of hamiltonian systems has been traditionally used to investigate the quantum chaos@xcite , and to track metal - insulator transition@xcite .
the uncorrelated metallic state shows a ` correlated ' single - particle energy level spacing distribution with level repulsion , whereas the correlated insulator state is associated with a ` uncorrelated ' poisson spacing distribution . for a chaotic system ,
the spacing distribution is seen to be same as that of an ensemble of random matrices having the same number of degrees of freedom as the system . from our investigation of fraction of distinct eigenvalues
we have found that for crossover values of @xmath0 , the eigenvalues of reduced density matrix behave like a ` chaotic ' system .
so , in this region , we expect that the spacing distribution of eigenvalues of @xmath57 can be understood from a random matrix ensemble . for that purpose , we first look at the symmetries of @xmath57 .
@xmath57 is block diagonal , with each block specified by the number of up spins @xmath58 and the number of down spins @xmath59 in @xmath60 .
furthermore , we note that the labels of @xmath61 and @xmath62 , as well as those of @xmath60 and @xmath63 are arbitrary and can be interchanged . hence the eigenvalues of blocks of reduced density matrix with spins flipped are same .
also , all the eigenvalues from blocks with @xmath64 particles in @xmath60 are same as those from blocks with @xmath65 particles in @xmath60 .
hence there are inherent degeneracies among the blocks .
finding eigenvalues of each block of @xmath57 is a formidable task because the size and number of blocks increase exponentially with @xmath5 .
let @xmath4 denote a block of the reduced density matrix @xmath57 with @xmath66 and @xmath67 .
the matrix element @xmath68 between two configurations labelled by the set of positions of up and down spin electrons in subsystem @xmath60 is given by , @xmath69 where @xmath70 is constant that depends on @xmath5 , @xmath58 and @xmath59 , @xmath71 is a function of @xmath0 coming from the double occupancy of the configurations @xmath51 and @xmath72 of subsystem @xmath60 , @xmath73 is a function @xmath0 coming from the double occupancy of the configuration of subsystem @xmath63 .
the label @xmath74 ( @xmath75 ) is a set of @xmath76 ( @xmath77 ) numbers giving positions of up ( down ) spins in subsystem @xmath60 in @xmath51th configuration and similarly for the @xmath72th configuration , and @xmath78 ( @xmath79 ) is the set of positions of up ( down ) spin in subsystem @xmath63 , and @xmath80 is the slater determinant with momentum labels given by the set @xmath9 and position labels given by the set @xmath81 , @xmath82 denotes complex conjugate .
the sum is over all possible configurations of @xmath63 , i.e , all possible choices of @xmath78 and @xmath79 .
writing out the slater determinants explicitly and simplifying , it can be shown that there are some terms in the sum that are independent of the configurations of @xmath63 . upon performing the sum over configurations of @xmath63 , these terms
add up coherently whereas the other terms add incoherently .
since the number of all possible configurations of @xmath63 is quite large , we can approximate @xmath68 by considering only the terms independent of configurations of @xmath63 .
these terms will only pick up a factor of some constant times a function of @xmath0 from the sum .
the result is of the form : @xmath83 \equiv q f_{ij}(g ) m_{ij},\ ] ] where all constants have been absorbed into @xmath70 , all @xmath0 dependence coming from both subsystems @xmath60 and @xmath63 has been absorbed into @xmath84 , @xmath85 ( @xmath86 ) is a set of @xmath76 ( @xmath77 ) values of @xmath9 chosen from all possible values of @xmath9 below @xmath15 and the sums are over all possible such choices , @xmath87 is the term within square brackets which is the sum of all possible terms in eq [ r1 ] which are independent of configurations of @xmath63 . since taking complex conjugate of slater determinant simply means @xmath88 , and since both these levels are allowed , the above sums are real .
thus the @xmath87 is real . also due to the same reason ,
@xmath87 depends only on the relative positions of up ( down ) spins in the @xmath51th and @xmath72th configurations .
these properties manifest in symmetries of the matrix @xmath89 . for @xmath90
, explicit calculation shows that the matrix @xmath89 is a real symmetric toeplitz matrix.for example , for @xmath90 , for the block with @xmath91 and @xmath92 , @xmath87 is given by : @xmath93 which can be cast into toeplitz form with proper ordering . for @xmath94 ,
the matrix @xmath89 becomes block real symmetric toeplitz matrix with each block also being a real symmetric toeplitz matrix .
such a matrix has the same number of degrees of freedom as a real symmetric toeplitz matrix .
however , it is difficult to prove such a result for general @xmath5 directly .
absolute value of @xmath95 of spacing distribution of the unfolded eigenvalues of @xmath57 is plotted with the adjacent level spacing @xmath2 , for @xmath30 sites at half - filling @xmath96 , and is linear fitted with slope @xmath97 .
this confirms prediction from random matrix theory .
( refer fig [ fig : p(s)1 ] . ) ] but we can prove that by noting that the matrix @xmath89 is proportional to the reduced density matrix for @xmath13 , where the system is non - interacting . in that case ,
as shown by various authors @xcite , wick decomposition allows eigenvalues of @xmath57 to be written in terms of eigenvalues of the correlation matrix @xmath98 .
the eigenvalues @xmath99 of the correlation matrix are related to eigevalues @xmath100 by the realtion : @xmath101 following f. arres et.al @xcite , the elements of the correlation matrix are given by : @xmath102 where the sum is over all occupied values of @xmath9 .
we note that , since @xmath9 and @xmath103 are both allowed in the sum , @xmath104 is real .
hence the correlation matrix has a real symmetric toeplitz structure .
therefore , each block of @xmath57 , which is proportional to the matrix @xmath89 , must have the same number of degrees of freedom as a real symmetric toeplitz matrix .
motivated by random matrix theory and the fact that the system behaves ` chaotic ' for intermediate values of @xmath0 , we claim that the adjacent level spacing distribution for intermediate values of @xmath0 will be the same as that of a random matrix ensemble of block diagonal matrices where each block is a real symmetric toeplitz matrix whose elements are drawn from some distribution with mean @xmath11 and variance @xmath12 .
ensemble of random real symmetric toeplitz matrices have been previously investigated and the spacing distribution is found to be poisson distribution , and not the usual goe spacing distribution @xcite .
this is because the number of degrees of freedom of @xmath105 real symmetric toeplitz matrix is @xmath106 and not @xmath107 . in our case , since sizes of blocks vary , we have many ensembles of random real symmetric toeplitz matrices of various sizes .
so the spacing distribution will not exactly be the poisson distribution . since size of matrix is the only parameter in a random matrix ensemble , blocks of same size will have same distribution of eigenvalues
. also eigenvalue distributions of blocks of different sizes will overlap .
hence , probability of degeneracy of the eigenvalues will be very high .
thus in the spacing distribution , there will be a pronounced peak at zero spacing .
apart from the peak , the effect of having blocks of various sizes will be to slow down the poisson like decay .
so the exponent in poisson distribution will pick up a constant factor .
the amplitude of the poisson like decay function will be much smaller than the peak at zero spacing .
thus the spacing distribution we expect from random matrix theory is highly peaked at zero and having an exponentially decaying tail , the peak being much higher than amplitude of the exponential function .
the exponential decay scale depends , as we will see below ( see fig.4 ) , on the ratio of the number of distinct eigenvalues and the rank that we have investigated ( see fig.2 ) earlier . to check the above result with exact numerical calculations ,
we need to unfold the eigenvalue spectrum to get a non - trivial distribution , viz .
the average spacing of the unfolded energy level should be independent of the system size .
we define the unfolded eigenvalues @xmath108 , using the analytical result of the spectrum given in eq.[nu2 ] , @xmath109 the adjacent level spacing distribution is then given by , @xmath110 where @xmath111 is the mean adjacent level spacing of @xmath112 . even though the distribution of @xmath113 is constrained by @xmath114 ,
the above distribution is unconstrained because @xmath112 diverges as @xmath115 . the spacing distribution for @xmath30 sites is shown in fig.[fig : p(s)1 ] for various values of @xmath0 .
the distribution has a pronounced peak at zero spacing , but it goes down very fast to a peak of a single value , i.e , there is only one value corresponding to that spacing .
after that , of course , there can be no decay .
thereafter , the peaks of single values occur at large intervals , the interval size also increasing with spacing , which means the effective probability of finding a spacing at a given interval keeps on decreasing .
this discrete tail , of course , can not be described by a continuous function .
however , our understanding from random matrix gives the decay after the peak at zero spacing till the first peak of single value for intermediate values of @xmath0 .
the constant factor in the exponent is phenomenologically found to be equal to the fraction of distinct eigenvalues @xmath97 , which would of course be equal to @xmath12 if there were no degeneracy .
[ fig : p(s)2 ] gives the semilog plot . at @xmath14 and @xmath13 ,
the spacing distributions are different because there the system is ` integrable ' . there
the spacing distribution may still be poisson - like with a different constant factor in the exponent , but no conclusion can be drawn at present . away from the half - filling case ,
none of the above measures show much variation with the correlation factor @xmath0 , which goes well with previous results from calculation of entanglement entropy @xcite .
we have investigated eigenvalue spectrum of the reduced density matrix , as a function of the correlation factor @xmath0 for gutzwiller state at half - filling .
we have found that the rank @xmath4 and the largest eigenvalue @xmath32 of the reduced density matrix vary in such a way that their product clearly demarcates the insulating and metallic regions . in the insulating region it increases rapidly with @xmath0 and becomes almost constant in the metallic region . the fraction of distinct eigenvalues @xmath97 exhibits a ` chaotic ' behaviour in the crossover region , and it shows a ` integrable ' behaviour at both insulating and metallic ends .
we have shown that the integrated entanglement spectrum deviates from the calabrese - lefevre result in the crossover region .
we have related the deviation to existence of a finite correlation length and have shown that replacing the central charge by an effective finite - size corrected central charge , or , equivalently , writing the calabrese - lefevre result in terms of entanglement entropy gives the correct behaviour .
similar work has recently been done in the context of heisenberg model @xcite .
we have investigated the spacing distribution of non - zero eigenvalues of the reduced density matrix and have provided an understanding of it in terms of random matrices .
the adjacent level spacing distribution of unfolded non - zero eigenvalues for intermediate values of gutzwiller projection parameter @xmath0 is the same as that of an ensemble of random matrices obtained by replacing each block of reduced density matrix by a random real symmetric toeplitz matrix .
it is strongly peaked at zero , with an exponential tail proportional to @xmath1 , where @xmath2 is the adjacent level spacing , the proportionality constant being much smaller than the height of the peak at zero spacing .
all our results relating to the crossover region are essentailly finite - size effects .
this is because in the thermodynamic limit , for intermediate values of @xmath0 , the ratio of the finite correlation length and the subsystem size goes to zero , and hence the system behaves exactly the non - interacting @xmath13 case for any non - zero value of @xmath0 @xcite . but as discussed in the introduction , finite - size effects remain significant at least as long as @xmath22 , which are system sizes realizable in nanochains . | entanglement spectrum of finite - size correlated electron systems are investigated using the gutzwiller projection technique .
the product of largest eigenvalue and rank of the block reduced density matrix , which is a measure of distance of the state from the maximally entangled state of the corresponding rank , is seen to characterise the insulator to metal crossover in the state .
the fraction of distinct eigenvalues exhibits a ` chaotic ' behaviour in the crossover region , and it shows a ` integrable ' behaviour at both insulating and metallic ends .
the integrated entanglement spectrum obeys conformal field theory ( cft ) prediction at the metal and insulator ends , but shows a noticeable deviation from cft prediction in the crossover regime , thus it can also track a metal - insulator crossover . a modification of the cft result for the entanglement spectrum for finite size is proposed which holds in the crossover regime also .
the adjacent level spacing distribution of unfolded non - zero eigenvalues for intermediate values of gutzwiller projection parameter @xmath0 is the same as that of an ensemble of random matrices obtained by replacing each block of reduced density matrix by a random real symmetric toeplitz matrix .
it is strongly peaked at zero , with an exponential tail proportional to @xmath1 , where @xmath2 is the adjacent level spacing , @xmath3 is number of distinct eigenvalues and @xmath4 is the rank of the reduced density matrix .
0.5 cm |
it is well - known that in any symmetric monoidal category , there are useful intrinsic notions of _ duality _ and _ trace _ ; see , for example , @xcite .
one of the original motivating examples is that traces in the stable homotopy category compute fixed - point indices , an observation which leads directly to the lefschetz fixed point theorem .
the search for generalizations and converses of the lefschetz fixed point theorem has led to various generalizations and refinements of the fixed - point index , such as the reidemeister trace @xcite . in this paper
we present an abstract framework for constructing refinements of traces in symmetric monoidal categories , which produces the reidemeister trace as a particular example .
other examples , which we will not discuss here , include equivariant @xcite , relative @xcite , and fiberwise @xcite generalizations of the lefschetz number and reidemeister trace .
these refinements arise from generalizations of duality and trace to _ indexed _
symmetric monoidal categories .
an indexed symmetric monoidal category is a family of symmetric monoidal categories @xmath0 , one for each object @xmath1 of a cartesian monoidal base category @xmath2 , equipped with base change functors induced by the morphisms of @xmath2 .
the simplest case is when @xmath2 is the category of sets and @xmath0 the category of @xmath1-indexed families of objects of some fixed symmetric monoidal category , such as abelian groups or chain complexes .
we can also allow @xmath2 to consist of groupoids or higher groupoids .
the prototypical example in homotopy theory , which will give rise to the reidemeister trace , is when @xmath2 is the category of topological spaces and @xmath0 is the homotopy category of parametrized spectra over @xmath1 , as in @xcite .
more generally , @xmath2 could be any category of spaces , or of schemes , or a topos , and @xmath0 a category of sheaves of abelian groups , modules , spaces , or parametrized spectra on @xmath1 , or a derived category thereof . in any such context
we can , of course , consider duality and trace in the individual symmetric monoidal categories @xmath0 .
following @xcite we call these notions _ fiberwise _ , since in examples they are usually equivalent to duality and trace acting on each fiber or stalk separately .
thus , if @xmath3 is fiberwise dualizable , any endomorphism @xmath4 has a fiberwise trace , which is an endomorphism of the unit object @xmath5 of @xmath0 . in examples ,
this trace essentially calculates the ordinary traces of the induced endomorphisms @xmath6 of the fibers @xmath7 , for all @xmath8 .
our first theorem gives a refinement of this fiberwise symmetric monoidal trace .
[ thm : intro - fiberwise ] if @xmath9 is fiberwise dualizable and @xmath4 is any endomorphism , then the symmetric monoidal trace of @xmath10 factors as a composite @xmath11{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{a}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$ } } } } \xto{{\ensuremath{\operatorname{tr}}}(\widehat{f } ) } i_a.\ ] ] ( in fact , we prove a stronger theorem about `` partial traces '' whose domains and codomains are `` twisted '' by an additional object . ) the first morphism in this factorization is defined purely in terms of @xmath1 , while the second morphism contains the information about @xmath3 and @xmath10 .
thus , this theorem refines @xmath12 by lifting its domain to a potentially larger one . in examples ,
the refined trace @xmath13 also includes the traces of the composites of the fiberwise endomorphisms @xmath14 with the action of `` loops '' in @xmath1 on the fibers of @xmath3 ( in cases where @xmath1 is something which contains loops , like a groupoid or a topological space ) .
often , however , we are in a somewhat different situation : we want to extract trace - like information ( such as fixed point invariants ) from an endomorphism @xmath15 in some _ cartesian _ monoidal category @xmath2 , such as sets , groupoids , or spaces .
since there are no nontrivial dualities in a cartesian monoidal category , the standard approach is to choose a non - cartesian monoidal category @xmath16 ( such as abelian groups , chain complexes , or spectra ) , apply a functor @xmath17 ( such as the free abelian group or suspension spectrum ) , and then consider the symmetric monoidal trace of @xmath18 in @xmath16 .
the trace obtained from the free abelian group functor in this way simply counts the number of fixed points of a set - endofunction , while that obtained from the suspension spectrum functor is exactly the classical fixed - point index .
now , it turns out that in most examples where this is done , there is actually an indexed symmetric monoidal category over @xmath2 , such that @xmath19 is the category indexed by the terminal object of @xmath2 .
moreover , the functor @xmath20 is definable in terms of this structure : @xmath21 is the `` pushforward '' to @xmath22 of the unit object of @xmath0 .
for example , the free abelian group functor arises from `` set - indexed families of abelian groups '' , and the suspension spectrum functor arises from spectra parametrized over spaces .
our second theorem gives a refinement of the trace of @xmath18 in this situation .
we need a slightly stronger hypothesis , however : rather than requiring @xmath21 to be dualizable in @xmath22 ( the requirement in order for @xmath18 to have a trace ) , we need to assume that @xmath5 is _ totally dualizable _ in @xmath0 .
total duality is a new type of duality , not reducible to symmetric monoidal duality , which can be defined in any good indexed symmetric monoidal category ; it was first noticed in the case of parametrized spectra by costenoble and waner @xcite , and studied further in @xcite .
total dualizability of @xmath5 implies ordinary dualizability of @xmath21 in @xmath22 , and in examples it seems to be not much stronger than this ( if at all ) . [ thm : intro - total ] if @xmath5 is totally dualizable and @xmath23 is an endomorphism in @xmath2 , then the symmetric monoidal trace of @xmath18 factors as a composite @xmath24{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{a_f}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$ } } } } { \ensuremath{\overset{}{\longrightarrow}}}i_\star.\ ] ] here , it is the _ first _ morphism in the factorization which should be regarded as a refined trace of @xmath10 , while the second morphism forgets the additional information contained therein ( although in this case , the intermediate object @xmath25{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{a_f}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$}}}}$ ] also depends on @xmath10 ) . in examples ,
the object @xmath25{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{a_f}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$}}}}$ ] is `` generated by '' the `` fixed - point classes '' of @xmath10 , and the refined trace @xmath26 separates out the contributions to the fixed - point index of @xmath10 depending on which fixed - point class they belong to . in particular ,
in the case of spectra parametrized over spaces , we obtain the reidemeister trace , and says that the reidemeister trace refines the fixed - point index . as with , is a special case of a stronger theorem about `` partial traces '' and arbitrary totally dualizable objects .
we can also exploit this generalization to describe a further connection to the classical symmetric monoidal theory .
perhaps the most important `` partial '' or `` twisted '' trace is the trace of the composite @xmath27 which is a map @xmath28 called the * transfer of @xmath10 * ( or the `` trace of @xmath10 with respect to @xmath29 '' ) .
the transfer of @xmath10 is also a refinement of @xmath30 , in that when composed with the augmentation @xmath31 it reproduces the symmetric monoidal trace of @xmath10 .
the generalized version of then implies that @xmath26 refines not only @xmath30 , but also the trace of @xmath10 with respect to @xmath29 .
[ thm : intro - total2 ] in the situation of , the transfer of @xmath10 factors as a composite @xmath24{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{a_f}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$ } } } } { \ensuremath{\overset{}{\longrightarrow}}}\sigma(a).\ ] ] although they may seem different on the surface , theorems [ thm : intro - fiberwise ] and [ thm : intro - total][thm : intro - total2 ] actually apply in _ formally dual _ situations ( although their conclusions and proofs are not dual ) .
this observation is originally due to may and sigurdsson @xcite , who constructed a _
bicategory _ out of parametrized spectra , and identified fiberwise and costenoble - waner duality as cases of the general notion of duality in this bicategory ( which comes in dual flavors , since composition in a bicategory is not symmetric ) . in @xcite ,
the second author generalized their construction to any indexed symmetric monoidal category .
part of proving the above theorems , therefore , will be to extend the results of @xcite on duality to the case of a general indexed symmetric monoidal category .
the consideration of _ traces _ for bicategorical duality , on the other hand , requires an additional structure called a _ shadow _ , introduced by the first author in @xcite and studied further in @xcite .
thus , another necessary preliminary will be to prove that the bicategory arising from any indexed symmetric monoidal category has a canonical shadow .
this done , we will be able to identify the morphisms @xmath13 and @xmath26 as traces relative to the bicategorical incarnations of fiberwise and total duality , respectively , and deduce theorems [ thm : intro - fiberwise ] , [ thm : intro - total ] , and [ thm : intro - total2 ] .
finally , there is one further ingredient in this paper : the use of a _ string diagram _ calculus for indexed monoidal categories . in general ,
string diagram calculus is a `` poincar dual '' way of drawing composition in categorical structures , which tends to make complicated computations much more visually evident . in particular , many basic equalities , such as the naturality of tensor products ,
are realized by simple isotopies .
string diagrams for monoidal categories and bicategories are described in @xcite , and a generalization for bicategories with shadows was given in @xcite .
several of the results of this paper require fairly involved computations which are made much more tractable by an appropriate string diagram calculus .
however , for clarity of exposition , we postpone the introduction of string diagrams ( along with the proofs which depend on them , of course ) to the very end of the paper .
thus , the reader can avoid string diagrams altogether by stopping after [ sec : total ] .
the organization of this paper is as follows . in [ sec : shad - from - mono][sec : indexed - coproducts ] we introduce indexed symmetric monoidal categories . then in [ sec : smc - traces ] we recall the classical notion of trace in symmetric monoidal categories , and construct the functor @xmath32 referred to above . in [
sec : shadows ] we describe the construction of a bicategory with a shadow from an indexed symmetric monoidal category . in [ sec : fiberwise ] we state the `` fiberwise '' duality and trace theorems and its generalizations although the proofs are postponed until
[ sec : fiberwise - proofs ] .
next , [ sec : base - change ] is devoted to the consideration of `` base change objects '' , an additional bit of structure that exists in bicategories arising from indexed monoidal categories .
this is then used in
[ sec : total ] , where we state and prove the `` total '' duality and trace theorems , [ thm : intro - total ] and [ thm : intro - total2 ] and their generalizations .
the proofs in this case are actually easier than the fiberwise case , although some calculations still have to be postponed to
[ sec : tdproofs ] . finally , in
[ sec : string - diagrams][sec : colored - strings ] we introduce our string diagram calculus , and in [ sec : fiberwise - proofs][sec : tdproofs ] we apply it to the postponed calculations . throughout the paper , we make use of three primary examples : 1 . @xmath33 sets , @xmath34 @xmath1-indexed families of abelian groups
@xmath33 topological spaces , @xmath34 spectra parametrized over @xmath1 .
@xmath33 groupoids , @xmath34 @xmath1-indexed diagrams of chain complexes .
the first is a `` toy '' example , which is easy to compute with , but which is so degenerate that it fails to display all of the interesting phenomena .
the second is the motivating example which matters most in applications ( together with various generalizations ) , but it is technically fairly complicated , so we omit all proofs relating to it ; many of them can be found in @xcite .
( in particular , no knowledge of spectra is necessary to read this paper . )
the third is an intermediate example which is easier to understand , but which still displays most of the interesting phenomena .
we would like to thank david corfield , todd trimble , and daniel schppi for helpful suggestions about string diagrams and monoidal bicategories .
the second - named author gratefully acknowledges the hospitality of the university of kentucky .
we begin with the following standard definition ( cf . for instance @xcite or ( * ? ? ? * b1.2 ) ) .
let @xmath2 be a category .
an * @xmath2-indexed category * @xmath35 is a pseudofunctor from @xmath36 to @xmath37 . in elementary terms , this means we have a category for each object @xmath38 , which we write as @xmath0 , and a functor for each morphism @xmath39 in @xmath2 , which we write as @xmath40 , such that composition and identities are preserved up to coherent natural isomorphism .
we sometimes refer to @xmath2 as the _ base category _ and the @xmath0 as the _ fiber categories _ , and we refer to the functors @xmath41 as _ reindexing functors_. [ eg : ring - ind - mod ] if @xmath42 is the category of rings , there is an @xmath2-indexed category sending a ring @xmath1 to @xmath43 , with @xmath41 given by restriction of scalars .
there are similar examples using chain complexes , dgas , or ring spectra .
[ eg : families ] for any category @xmath16 , there is a @xmath44-indexed category sending each set @xmath1 to the category @xmath45 of @xmath1-indexed families of objects of @xmath16 .
[ eg : spaces - over - spaces ] if @xmath2 has finite limits , then there is an @xmath2-indexed category sending each object @xmath1 to its slice category @xmath46 , with the functors @xmath41 given by pullback . in the important special case
when @xmath2 is the category @xmath47 of topological spaces , we also have a `` homotopy '' version @xmath48 , in which we invert the weak homotopy equivalences in each fiber category .
see @xcite for an extensive formal development of this example .
similar examples can be constructed whenever @xmath2 has a `` homotopy theory '' ( e.g. when it is a quillen model category ) .
`` derived '' examples such as the last one are of particular interest to us . here
are two more such examples .
[ eg : spectra - over - spaces ] if @xmath47 denotes a suitably nice category of topological spaces , there is a @xmath47-indexed category defined by @xmath49 , where @xmath50 denotes a point - set level category of spectra parametrized over @xmath1 , such as the parametrized orthogonal spectra used in @xcite .
we can similarly invert the weak equivalences of spectra to obtain a derived @xmath47-indexed category @xmath51 .
[ eg : derived - families ] if the category @xmath16 in has a homotopy theory , then we can pass to homotopy categories in that example to obtain another @xmath44-indexed category @xmath52 .
however , since @xmath53 for any set @xmath1 , we gain no extra generality thereby .
we will see in later sections that qualitatively different phenomena arise in `` derived '' situations , when the base category @xmath2 also has a `` homotopy theory '' which is taken into account somehow in the fibers .
this is the case in examples [ eg : spaces - over - spaces ] and [ eg : spectra - over - spaces ] , but not in . since examples [ eg : spaces - over - spaces ] and [ eg : spectra - over - spaces ] are somewhat technically complicated , it will be useful to have a simpler example which exhibits some of the same derived phenomena .
we can obtain such examples by replacing spaces by groupoids ( which , up to homotopy , can be identified with homotopy 1-types ) .
[ eg : gpds ] if @xmath54 is the category of groupoids and @xmath16 is any category , we can define @xmath55 to be the category of functors from the groupoid @xmath1 to @xmath16 . for simplicity , we usually take @xmath16 to be abelian groups .
this is an enlargement of , if we consider sets as discrete groupoids . on the other hand ,
if @xmath1 is a one - object groupoid , i.e. a group @xmath56 , then @xmath57 can be identified with the category of modules over the group ring @xmath58 $ ] .
[ eg : hogpds ] we have `` derived phenomena '' in even when @xmath16 does not have its own homotopy theory , because the construction automatically `` sees '' the homotopy theory of groupoids .
( for instance , we have @xmath59 whenever @xmath60 are _ equivalent _ groupoids , not only when they are isomorphic )
. however , we can also `` derive it further '' if @xmath16 _ does _ have a homotopy theory , obtaining a @xmath61-indexed category with @xmath62 , where we invert the objectwise weak equivalences in @xmath45 . to maximize concreteness and familiarity , we may take @xmath63 to be the category of ( unbounded ) chain complexes , with its usual homotopy theory . note
that in contrast to , when @xmath1 is not discrete , @xmath52 will generally be different from @xmath64 .
thus , this example is genuinely different from . now in most of these examples , each category @xmath0 has a monoidal structure , which is respected by the transition functors .
namely : * if @xmath16 is a monoidal category ( such as @xmath65 or @xmath66 with tensor product ) , then so is @xmath45 for any set or groupoid @xmath1 ; * when @xmath2 has finite limits , each slice category @xmath46 is cartesian monoidal ; * each category @xmath50 has a parametrized smash product ; and * all of these monoidal structures descend to homotopy categories .
this `` fiberwise '' monoidal structure is simply described by the following definition . for a category @xmath2 ,
an * @xmath2-indexed symmetric monoidal category * is a pseudofunctor from @xmath36 to the 2-category @xmath67 of symmetric monoidal categories , strong symmetric monoidal functors , and monoidal natural transformations . in elementary terms
, this means an indexed category such that each @xmath0 is a symmetric monoidal category , each reindexing functor is strong symmetric monoidal , and each constraint isomorphism is a monoidal transformation .
we write @xmath68 and @xmath5 for the tensor product and unit object of the monoidal category @xmath0 , and refer to them as the _ fiberwise _ monoidal structure . however , for many purposes it is more effective to express the monoidal structure differently .
if @xmath2 has finite products , including a terminal object @xmath69 , then in any @xmath2-indexed monoidal category @xmath35we can define an _ external product _
functor @xmath70 ( here @xmath71 and @xmath72 are the projections from the cartesian product . ) this external product is coherently associative and unital , in a suitable sense ( see @xcite ) , with unit object @xmath73 .
moreover , we can recover the fiberwise monoidal structures ( or _ internal products _ ) from the external products , via the isomorphisms @xmath74 ( here @xmath75 is the diagonal of the cartesian product . )
these hold for any @xmath38 and @xmath76 , by monoidality of the functors @xmath41 . in the same way we can construct natural isomorphisms @xmath77 for any @xmath78 for which this makes sense .
if @xmath16 is a symmetric monoidal category , then in the @xmath44-indexed symmetric monoidal category @xmath79 of , the fiberwise product of @xmath80 and @xmath81 in @xmath45 is defined by @xmath82 whereas the external product of @xmath83 and @xmath84 is defined by @xmath85 the case of is similar . note that if @xmath1 is one - object groupoid , hence a group @xmath56 , and we identify @xmath57 with the category of @xmath58$]-modules , then its fiberwise monoidal structure is the tensor product over @xmath86 , made into a @xmath58$]-module using the fact that @xmath58 $ ] is a bialgebra .
it is _ not _ a tensor product over @xmath58 $ ] , which would nt make sense anyway since @xmath58 $ ] is not commutative .
if @xmath2 has finite limits , then in the @xmath2-indexed symmetric monoidal category @xmath87 of , the fiberwise product of @xmath88 is their fiber product ( pullback ) : @xmath89 whereas the external product of @xmath90 and @xmath91 is just their cartesian product , with the induced projection to @xmath92 : @xmath93 the case of is similar , with cartesian products replaced by smash products .
both the fiberwise and external products of parametrized spectra are used in @xcite .
when the base category @xmath2 is monoidal but not cartesian monoidal , we can still define `` an external product '' for @xmath35 to consist of functors @xmath94 which are coherently associative and unital .
for instance , has an external product , but not an internal one .
however , our interest here is solely in the case when @xmath2 is cartesian monoidal .
in order to construct a bicategory from an indexed symmetric monoidal category , and thereby prove our refinements of the symmetric monoidal trace , we will need one further piece of structure .
[ def : indcoprod ] let @xmath2 be a cartesian monoidal category , and @xmath35 an @xmath2-indexed category .
we say that @xmath35 has * @xmath2-indexed coproducts * if 1 .
each reindexing functor @xmath41 has a left adjoint @xmath95 , and 2 . for any pullback square @xmath96^f\ar[d]_h & b \ar[d]^g\\ c\ar[r]_k & d}\ ] ] in @xmath2 , the composite
@xmath97 is an isomorphism ( the _ beck - chevalley condition _ ) .
if @xmath35 is symmetric monoidal , we say that * @xmath98 preserves indexed coproducts * ( in each variable separately ) , or that the * projection formula holds * , if 1 . for any @xmath39 in @xmath2 and any @xmath99 , @xmath100 , the canonical map @xmath101 is an isomorphism . is not symmetric
, we must assert also the analogous condition on the other side.][item : prescoprod ] we will only use the beck - chevalley condition for a few types of pullback squares .
the basic such squares are shown in figure [ fig : pullbacks ] .
( these are all actually pullback squares in any cartesian monoidal category , whether or not it has pullbacks in general .
see also @xcite . )
we will also consider the transposes of figures [ fig : frobpb ] and [ fig : slidpb ] , as well as squares obtained from one of these by taking a cartesian product with a fixed object . for ease of reference
, we have assigned a name to each of these beck - chevalley conditions .
we call figure [ fig : reindpb ] `` commutativity with reindexing '' because the condition @xmath102 says that the indexed - coproduct functor @xmath103 commutes with the reindexing functor @xmath41 .
we call figure [ fig : frobpb ] the `` frobenius axiom '' because the condition @xmath104 has the same form as one of the axioms of a frobenius algebra
. the name of figure [ fig : slidpb ] will make more sense once we introduce string diagrams ; see [ sec : string - diagrams ] . finally , we call figure [ fig : mdpb ] `` monic diagonals '' because the fact that that square is a pullback says precisely that @xmath105 is a monomorphism .
[ cols="^,^ " , ] thus , it remains only to verify that @xmath106 and @xmath107 satisfy the appropriate triangle identities if and only if @xmath108 and @xmath109 satisfy _ their _ triangle identities .
we will verify this for one triangle identity ; the other is similar and left to the reader .
refer to figures [ fig : fib - duals ] and [ fig : fib - duals-2 ] . in figure
[ fig : fib - duals - step1 ] we show the composite which the bicategorical triangle identity asserts to be equal to an identity .
the morphisms occuring at the top prior to the coevaluation @xmath106 comprise the isomorphism from figure [ fig : bicatunit ] ( spelled out in more detail ) , using the definition of the beck - chevalley morphism from figure [ fig : mdbc ] . in figure
[ fig : fib - duals - step2 ] we have replaced @xmath106 and @xmath107 with their equivalents in terms of the symmetric monoidal @xmath108 and @xmath109 .
then in figure [ fig : fib - duals - step3 ] , we slide a number of morphisms past each other , in the way which string diagram notation makes easy to see .
to get to figure [ fig : fib - duals - step4 ] , we apply a triangle identity for the adjunction @xmath110 , as in figure [ fig : surface - adj - diag ] . in figure
[ fig : fib - duals - step5 ] we do some more sliding , and then finally in figure [ fig : fib - duals - step6 ] we apply ( the dual of ) the curlycue identity from figure [ fig : curlycue ] . but figure [ fig : fib - duals - step6 ] is exactly the composite which the symmetric monoidal triangle identity asserts equal to an identity . thus , one triangle identity holds if and only if the other does .
the definition of @xmath111 is shown in figure [ fig : fibtrace1-gbar ] .
we must show that the bicategorical trace @xmath112{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{q}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$}}}}{\ensuremath{\overset{}{\longrightarrow}}}{{\ensuremath{\hspace{1mm}\makebox[-1mm]{$\langle$ } \makebox[0mm]{$\langle$}\hspace{1mm}{p}\makebox[1mm]{$\rangle$ } \makebox[0mm]{$\rangle$ } } } } = p\ ] ] is equal to the composite @xmath113 in figure [ fig : fibtrace1-bicat ] we show @xmath114 , while in figure [ fig : fibtrace1-symmon ] we show with the definition of @xmath111 substituted .
we can first of all consider the parts of these diagrams occurring above @xmath115 separately from those occurring below @xmath115 ; it will clearly suffice to show that each of these pairs are equal . in figure
[ fig : fibtrace1-lower ] we treat the lower parts .
figure [ fig :
fibtrace1-lower - a ] is the lower part of figure [ fig : fibtrace1-symmon ] . to obtain figure [ fig : fibtrace1-lower - b ] , we change the isomorphism @xmath116 from one which contracts @xmath29 with the second @xmath117 to one which contracts it with the first .
the two are equal by pseudofunctor coherence , since both are induced by the equality @xmath118 .
then in figure [ fig : fibtrace1-lower - c ] we slide @xmath109 and the symmetry down to the bottom .
this yields the lower part of figure [ fig : fibtrace1-bicat ] ( recalling the relationship of the symmetric monoidal and bicategorical evaluations ) together with a composite of pseudofunctoriality isomorphisms on top .
but by pseudofunctor coherence , this composite is equal to the identity .
next , in figure [ fig : fibtrace1-upperi ] we begin considering the upper parts . in figure
[ fig : fibtrace1-upper - a ] we show the upper part of figure [ fig : fibtrace1-symmon ] , with the symmetric monoidal coevaluation written in terms of the bicategorical one .
then in figure [ fig : fibtrace1-upper - b ] we slide the bicategorical @xmath108 all the way to the bottom ( directly above @xmath115 , which divides the upper from the lower parts ) .
since @xmath108 also occurs directly above @xmath115 in figure [ fig : fibtrace1-bicat ] , it suffices to compare the parts of figures [ fig : fibtrace1-bicat ] and [ fig : fibtrace1-upper - b ] occurring above @xmath108 .
moreover , we may also ignore the @xmath119 node and the @xmath120 node on the far left , since neither plays any role in either of these diagrams .
now , in figure [ fig : fibtrace1-usub-1 ] we have copied the part of figure [ fig : fibtrace1-bicat ] above @xmath108 , with @xmath119 and @xmath120 omitted and the definition of the frobenius morphism substituted . in figure
[ fig : fibtrace1-usub-2 ] we simply slide , and then to get to figure [ fig : fibtrace1-usub-3 ] we apply the broken zigzag identity ( figure [ fig : curlycue - mate ] ) . in figure [ fig : fibtrace1-usub-4 ] we replace the isomorphism at the very top by a composite of two other isomorphisms , which is equal to it by pseudofunctor coherence .
figure [ fig : fibtrace1-usub-5 ] is a simple slide , and then in figure [ fig : fibtrace1-usub-6 ] we replace the composite of two associativity isomorphisms by the composite of three ( the `` pentagon identity , '' which also holds by pseudofunctor coherence ) . in figure [
fig : fibtrace1-usub-7 ] we slide the lower of these associativities past a unit , and finally in figure [ fig : fibtrace1-usub-8 ] we compose two isomorphisms at the bottom to obtain a different one ( again by pseudofunctor coherence ) .
the result is exactly the part of figure [ fig : fibtrace1-upper - b ] above @xmath108 , with @xmath119 and @xmath120 omitted .
we define @xmath121 to be the composite @xmath122 this is pictured in figure [ fig : ftilde ] .
we now claim that the following square commutes .
@xmath123 \ar[d]_f & ( \delta_a)^ * ( \delta_a ) _ !
q \otimes m\ar[d]^{\overline{\widetilde{f}}}\\ m\otimes p \ar[r ] & m\otimes ( \pi_a)^ * ( \pi_a ) _ ! p}\label{eq : ftilde - square}\ ] ] to show this , we substitute @xmath121 from figure [ fig : ftilde ] as @xmath115 in figure [ fig : fibtrace1-gbar ] ( replacing @xmath119 by @xmath124 and @xmath125 by @xmath126 ) . canceling the isomorphism @xmath127 with its inverse
, we see that what is left at the bottom of the resulting diagram is exactly the left - bottom composite of .
therefore , it suffices to show that what remains above this becomes the identity when composed with the top arrow in . by inverting the first two transitions in figure [ fig : ftilde ] , this is equivalent to the equality shown in figure [ fig : ftilde2 ] .
however , this follows by a simple sliding and the broken zigzag identity from figure [ fig : curlycue - mate ] .
\(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 4,,2 ) ; ( d4 ) at ( 5,,1 ) ; ( d5 ) at ( 6 , , 1 ) ; ( m ) ( d4 ) ( d5 ) ; ( q)(d3 ) ( d4 ) ; ( d3)(7 , , 3 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 4,,1.5 ) ; ( d4 ) at ( 5,,1.5 ) ; ( d5 ) at ( 6 , , 1 ) ; ( m ) ( d3 ) ; ( d4 ) ( d5 ) ; ( q)(d3 ) ( d4 ) (7 , , 3 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d4 ) at ( 4,,1 ) ; ( m )
( d4 ) (7 , , 1 ) ; ( q )
( d4 ) ; \(p ) at ( -.5,,0 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( d4 ) at ( 4,,1 ) ; ( m ) ( d4 )
(7 , , 1 ) ; ( p ) ( d4 ) ; \(f ) at ( -.25,2.1,1 ) @xmath10 ; ( m1 )
( f ) ; ( q1 )
( f ) ; ( f ) ( p3 ) ; ( f ) ( m3 ) ; \(p ) at ( -.5,,0 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( d2 ) at ( 2,,0 ) ; ( d3 ) at ( 3,,0 ) ; ( d4 ) at ( 4,,1 ) ; ( m ) ( d4 )
(7 , , 1 ) ; ( p ) ( d2 ) ; ( d3)(d4 ) ; ( q-1 ) ( q0 ) ( q1 ) ; \(p ) at ( -.5,,0 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( d2 ) at ( 2,,0 ) ; \(m ) ( 7 , , 2 ) ; ( p ) ( d2 ) ; ( d30 ) ( d41) ( f ) ( d43)(d44 ) to[out=-90,in=-90,looseness=2 ] node[inneriso ] ( d34 ) ; ( d5 - 1 ) ( d50 ) to[out=-90,in=-90,looseness=2 ] node[inneriso ] ( d40 ) ; ( d40 ) ( d3 - 1 ) ; ( d4 - 1 ) ( d30 ) ; at ( intersection of d3 - 1d40 and d4 - 1d30 ) ; ( d24)(d25 ) ; ( m-1 ) ( m0 ) ( m1 ) ; ( m3 ) ( m4)(m5 ) ; ( p3)(p4)(p5 ) ; ( d34 ) to[out=90,in=90,looseness=1.5 ] node[innerunit ] ( d24 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 3,,1 ) ; ( m ) ( d3 ) ( 6,,1 ) ; ( q ) ( d3 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 3,,1 ) ; ( d4 ) at ( 4,,1 ) ; ( d5 ) at ( 5,,0 ) ; ( m ) ( d3 ) ( d4 ) ( 6,,1 ) ; ( q ) ( d3 ) ; ( d4 ) ( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,2 ) ; ( d3 ) at ( 3,,1 ) ; ( d4 ) at ( 4,,1 ) ; ( d5 ) at ( 5,,0 ) ; ( m ) ( d3 ) ( d4 ) ( 6,,1 ) ; ( q ) ( d1 ) ; ( d2 ) ( d3 ) ; ( d1 ) to[bend right ] ( d2 ) ; ( d1 ) to[bend left ] ( d2 ) ; ( d4 ) ( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,0 ) ; ( d3 ) at ( 3,,1 ) ; ( d4 ) at ( 4,,1 ) ; ( d5 ) at ( 5,,0 ) ; ( m ) ( d2 ) ; ( q ) ( d1 ) ; ( d2 ) ( d3 ) ( d4 )
( 6,,1 ) ; ( d1 ) ( d2 ) ; ( d1 ) ( d3 ) ; ( d4 )
( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,0 ) ; ( d5 ) at ( 5,,0 ) ; ( m ) ( d2 ) ( d5 ) ; ( q ) ( d1 ) ( 6,,2 ) ; ( d30 ) ( d31 ) ( d32 ) ( d23 ) ( d24 ) ; ( d33 ) ( d22 ) ; ( d22 ) to[out=90,in=90,looseness=3 ] node[innerunit ] ( d12 ) ; ( d12 ) ( d13 ) ( d14 ) ; ( d33 ) to[out=-90,in=-90,looseness=2 ] node[innercounit ] ( d43 ) ; ( d43 ) ( d42 ) ( d41 ) ; ( d41 ) to[out=90,in=90,looseness=3 ] node[inneriso ] ( d51 ) ; ( d51 ) ( d52 ) ( d53 ) ( d54 ) ; at ( intersection of d22d33 and d32d23 ) ; ( m0 ) ( m1 ) ( m2 ) ( m3 ) ( m4 ) ; ( q0 ) ( q1 ) ( q2 ) ( q3 ) ( q4 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d5 ) at ( 5,,1 ) ; ( m ) ( d5 ) ( 6,,1 ) ; ( q ) ( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,2 ) ; ( d5 ) at ( 5,,1 ) ; ( m ) ( d5 ) ( 6,,1 ) ; ( d1 ) to[bend right ] ( d2 ) ; ( d1 ) to[bend left ] ( d2 ) ; ( q ) ( d1 ) ; ( d2 ) ( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,0 ) ; ( d5 ) at ( 5,,1 ) ; ( m ) ( d2 ) ( d5 ) ( 6,,1 ) ; ( d1 ) ( d2 ) ; ( q ) ( d1 ) ; ( d1 ) ( d5 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,0 ) ; ( d3 ) at ( 3,,0 ) ; ( d4 ) at ( 4,,0 ) ; ( d5 ) at ( 5,,1 ) ; ( m )
( d2 ) ( d3 ) ; ( d4 ) ( d5 ) ; ( d1 ) ( d2 ) ; ( q ) ( d1 ) ; ( d1 ) ( d5 ) ( 6,,1 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,0 ) ; ( d3 ) at ( 3,,0 ) ; ( m ) ( d2 ) ( d3 ) ; ( d1 ) ( d2 ) ; ( q ) ( d1 ) ; ( d1 ) ( 6,,1 ) ; ( d50 ) ( d51 ) ( d22 ) ; ( d21 ) ( d52 ) ( d53 ) ; at ( intersection of d22d51 and d21d52 ) ; ( d53 ) to[out=-90,in=-90,looseness=2 ] node[inneriso ] ( d43 ) ; ( d43 ) to[out=90,in=90,looseness=2 ] node[innerunit ] ( d33 ) ; ( d33 ) ( d43 ) ; ( d21 ) to[out=90,in=90,looseness=3 ] node[innerunit ] ( d11 ) ; ( d11 ) ( d12 ) ( d13 ) ( d14 ) ; ( d22 ) ( d23 ) ( d24 ) ; ( d33 ) ( d34 ) ; ( m0 ) ( m1 ) ( m2 ) ( m3 ) ( m4 ) ; ( q0 ) ( q1 ) ( q2 ) ( q3 ) ( q4 ) ; now , from the conclusion of [ item : fibtrace1 ] and the naturality of symmetric monoidal traces , we can conclude that the following diagram commutes .
@xmath128 \ar[d]_{(\pi_a ) _ ! { \ensuremath{\operatorname{tr}}}(f ) } & ( \pi_a)_!(\delta_a)^ * ( \delta_a ) _ ! q \ar[d]_{(\pi_a)_!{\ensuremath{\operatorname{tr}}}(\overline{\widetilde{f } } ) } \ar[dr]^{{\ensuremath{\operatorname{tr}}}(\widetilde{f } ) } \\ ( \pi_a ) _ ! p \ar[r ] & ( \pi_a ) _ ! ( \pi_a)^ * ( \pi_a ) _ ! p \ar[r ] & ( \pi_a ) _ ! p } \ ] ] but the composite along the bottom of this diagram is the identity , by a triangle law for the adjunction @xmath129
. this yields the desired result .
the bijection between the two types of morphism is immediate , since the respective domains and codomains are isomorphic . for a morphism @xmath130
, the corresponding @xmath131 is shown on the left side of figure [ fig : fhat ] .
moreover , given @xmath130 , we can also construct @xmath121 as in part [ item : fibtrace2 ] ( replacing @xmath125 by @xmath132 ) , and we claim that @xmath133 is the composite @xmath134 ^ -{\widetilde{f } } & \widehat{m } \odot ( \pi_a)_!(\pi_a)^ * p \ar[r]^-{{\ensuremath{\operatorname{id}}}\odot { \ensuremath{\varepsilon } } } & \widehat{m } \odot p.}\label{eq : fhat - ftilde}\ ] ] this composite is shown on the right side of figure [ fig : fhat ] ; a slide and a triangle identity suffice to prove the equality .
now the desired result follows from the conclusion of part [ item : fibtrace2 ] together with the naturality of bicategorical traces ( * ? ? ?
\(m ) at ( -1,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,2 ) ; ( d2 ) at ( 2,,1 ) ; ( d3 ) at ( 3,,1 ) ; ( m ) ( d2 ) ( d3 ) ; ( q ) ( d1 ) ( d2 ) ; ( d1 ) ( 4,,2 ) ; \(m ) at ( -1,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,1 ) ; ( d2 ) at ( 2,,1 ) ; ( d3 ) at ( 3,,0 ) ; ( m )
( d1 ) ( d2 ) ( d3 ) ; ( q ) ( d1 ) ; ( d2 ) ( 4,,1 ) ; \(m ) at ( -1,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d1 ) at ( 1,,1 ) ; ( m ) ( d1 ) ( 4,,1 ) ; ( q ) ( d1 ) ; \(p ) at ( -1,,0 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( d1 ) at ( 1,,0 ) ; ( d2 ) at ( 2,,1 ) ; ( d1 ) ( d2 ) ( 4,,1 ) ; ( m ) ( d2 ) ; \(p ) at ( -1,,0 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( m ) ( 4,,2 ) ; \(f ) at ( -.3,3,1 ) @xmath10 ; ( m0 ) ( m1 )
( m2 ) ( f ) ( m4 ) ( m5 ) ; ( q0 ) ( q1 ) ( q2 ) ( f ) ( p4 ) ( p5 ) ; ( d20 ) ( d11 ) ( d12 )
( f ) ; ( d10 ) ( d21 ) ; at ( intersection of d10d21 and d20d11 ) ; ( d21 ) to[out=-90,in=-90,looseness=3 ] node[inneriso ] ( d31 ) ; ( d30 ) ( d31 ) ; ( f ) ( d14 ) ; ( f ) ( d24 ) ; ( d14 ) to[out=-90,in=-90,looseness=3 ] node[inneriso ] ( d24 ) ; at ( 0,7,0 ) @xmath3 ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 4,,2 ) ; ( d4 ) at ( 5,,1 ) ; ( d5 ) at ( 6 , , 1 ) ; ( m ) ( d4 ) ( d5 ) ; ( q)(d3 ) ( d4 ) ; ( d3)(7 , , 3 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d3 ) at ( 4,,1.5 ) ; ( d4 ) at ( 5,,1.5 ) ; ( d5 ) at ( 6 , , 1 ) ; ( m ) ( d3 ) ; ( d4 ) ( d5 ) ; ( q)(d3 ) ( d4 ) (7 , , 3 ) ; \(m ) at ( -.5,,0 ) @xmath3 ; ( q ) at ( 0,,2 ) @xmath119 ; ( d4 ) at ( 4,,1 ) ; ( m ) ( d4 ) (7 , , 1 ) ; ( q ) ( d4 ) ; \(p ) at ( -.5,,0 ) @xmath125 ; ( d1 ) at ( 1,,0 ) ; ( m ) at ( 0,,2 ) @xmath3 ; ( d4 ) at ( 4,,1 ) ; ( m ) ( d4 ) (7 , , 1 ) ; ( d1 ) ( d4 ) ; \(f ) at ( -.25,2.1,1 ) @xmath10 ; ( m1 )
( f ) ; ( q1 )
( f ) ; ( f ) ( d13 ) ; ( f ) ( m3 ) ; \(p ) at ( -.5,,0 ) @xmath125 ; ( d1 ) at ( 1,,0 ) ; ( m ) at ( 0,,1.5 ) @xmath3 ; ( d2 ) at ( 2,,0 ) ; ( d3 ) at ( 3,,0 ) ; ( d4 ) at ( 4,,1 ) ; ( m ) ( d4 ) (7 , , 1 ) ; ( d1 ) ( d2 ) ; ( d3)(d4 ) ; ( q-1 ) ( q0 ) ( q1 ) ; \(p ) at ( -.5,,0.5 ) @xmath125 ; ( d1 ) at ( 1,,0.5 ) ; ( m ) at ( 0,,2 ) @xmath3 ; ( d2 ) at ( 2,,0.5 ) ; \(m ) ( 7 , , 2 ) ; ( d1 ) ( d2 ) ; \(p ) at ( -.5,,0.5 ) @xmath125 ; ( m ) at ( 0,,2 ) @xmath3 ; ( m )
( 7 , , 2 ) ; ( d30 ) ( d41) ( f ) ( d43)(d44 ) to[out=-90,in=-90,looseness=2 ] node[inneriso ] ( d34 ) ; ( d5 - 1 ) ( d50 ) to[out=-90,in=-90,looseness=2 ] node[inneriso ] ( d40 ) ; ( d40 ) ( d3 - 1 ) ; ( d4 - 1 ) ( d30 ) ; at ( intersection of d3 - 1d40 and d4 - 1d30 ) ; ( d24)(d25 ) ; ( m-1 ) ( m0 ) ( m1 ) ; ( m3 ) ( m4)(m5 ) ( m6 ) ; ( d13)(d14)(d15 ) ; ( f ) ( p3)(p4)(p5 ) ( p6 ) ; ( d15 ) to[out=-90,in=-90,looseness=4 ] node[innercounit ] ( d25 ) ; ( d34 ) to[out=90,in=90,looseness=1.5 ] node[innerunit ] ( d24 ) ;
in contrast to the fiberwise case , our general comparison theorems for total duality traces were completely formal , not requiring any string diagram calculations . however , the two most interesting applications , namely corollaries [ cor : totaltr3 ] and [ thm : totaltr - transfer ] , required identifying the composites @xmath135 and @xmath136 as @xmath18 and @xmath137 , respectively , and for this we do need some calculation .
recall that we have an endomorphism @xmath138 of @xmath139 and that we defined @xmath10 to be the isomorphism @xmath140 regarded as a morphism @xmath141 .
we defined @xmath142 be the morphism @xmath143 by definition , then , the composite @xmath135 becomes @xmath144 to give a simpler description of this composite we start by considering the composite of only the third and fourth morphisms ( the two on the second line of ) .
these are instances of .
tracing through their definitions yields a fairly long composite of isomorphisms , but fortunately it is equal to something simpler , as an instance of the following lemma . using , and unfolding the definition of the `` sliding and splitting '' beck - chevalley morphism , becomes the morphism shown in figure [ fig : xif1 ] . in figure
[ fig : xif2 ] we simply rearrange the order of all the morphisms , sliding the counit @xmath151 on the far left down to the bottom , the isomorphism @xmath152 on the far right up to the top , and rearranging things in the middle so that the other unit and counit meet . in figure
[ fig : xif3 ] we cancel the internal unit and counit using a triangle identity .
now we have the isomorphisms @xmath153 and @xmath154 at the top , a counit @xmath155 at the bottom , and in the middle a composite of pseudofunctoriality isomorphisms for reindexing functors , which starts and ends at @xmath156 .
this middle composite is extracted in figure [ fig : xifm ] , after peeling off the @xmath157 and @xmath158 ( which are unchanged throughout ) . by coherence for pseudofunctors ,
this composite is equal to the identity , so that figure [ fig : xif3 ] is equal to figure [ fig : xif4 ] .
but this is exactly @xmath150 ( see ) . since @xmath159 factors through @xmath142 , @xmath136 factors through @xmath135 . applying
, we see that @xmath136 is equal to the composite @xmath162 thus , it will suffice to show that the composite @xmath163 is equal to @xmath164 .
now , by definition , @xmath164 is equal to the composite in figure [ fig : zeta1 ] .
we can rewrite the initial pseudofunctoriality isomorphisms to obtain figure [ fig : zeta2 ] , and then apply the broken zigzag identity ( figure [ fig : curlycue - mate ] ) to obtain figure [ fig : zeta3 ] . in figure
[ fig : zeta4 ] we isotope sideways , and then in figure [ fig : zeta5 ] we slide the adjunction unit to the top .
finally , we can cancel the isomorphisms in the middle by pseudofunctor coherence , obtaining figure [ fig : zeta6 ] , which is exactly .
g. m. kelly and ross street .
review of the elements of @xmath165-categories . in _
category seminar ( proc .
sydney , 1972/1973 ) _ , volume 420 of _ lecture notes in mathematics _ , pages 75103 .
springer , berlin , 1974 .
f. william lawvere .
equality in hyperdoctrines and comprehension schema as an adjoint functor . in _ applications of categorical algebra ( proc .
pure math .
xvii , new york , 1968 ) _ , pages 114 .
soc . , providence , r.i . , 1970 .
robert par and dietmar schumacher .
abstract families and the adjoint functor theorems . in _ indexed categories and their applications _ , volume 661 of _ lecture notes in math .
_ , pages 1125 .
springer , berlin , 1978 .
peter selinger .
a survey of graphical languages for monoidal categories . in bob
coeke , editor , _ new structures for physics _ , chapter 4 .
springer , 2011 .
available at http://www.mscs.dal.ca/~selinger/papers.html#graphical and arxiv:0908.3347 .
r.f.c . walters .
categorical algebras of relations .
http://rfcwalters.blogspot.com/2009/10/categorical - algebras - of - relation% s.html[http://rfcwalters.blogspot.com/2009/10/categorical - algebras - of - relation% s.html ] , october 2009 . | by the lefschetz fixed point theorem , if an endomorphism of a topological space is fixed - point - free , then its lefschetz number vanishes .
this necessary condition is not usually sufficient , however ; for that we need a refinement of the lefschetz number called the reidemeister trace .
abstractly , the lefschetz number is a trace in a symmetric monoidal category , while the reidemeister trace is a trace in a bicategory ; in this paper we relate these contexts using indexed symmetric monoidal categories .
in particular , we will show that for any symmetric monoidal category with an associated indexed symmetric monoidal category , there is an associated bicategory which produces refinements of trace analogous to the reidemeister trace .
this bicategory also produces a new notion of trace for parametrized spaces with dualizable fibers , which refines the obvious `` fiberwise '' traces by incorporating the action of the fundamental group of the base space .
we also advance the basic theory of indexed monoidal categories , including introducing a string diagram calculus which makes calculations much more tractable . this abstract framework lays the foundation for generalizations of these ideas to other contexts . |
understanding the quark structure of the lightest meson , the pion , the prototype for a meson bound state is arguably one of the most basic , albeit challenging , questions qcd is still facing even after decades of intense investigations . integrating over transverse momenta up to some resolution scale @xmath2
, one gets from the pion s light - cone wave function the pion s leading twist-2 distribution amplitude ( da ) , @xmath3 , defined in terms of a nonlocal axial current : @xmath4 where is the longitudinal momentum fraction carried by the valence quark ( antiquark ) in the pion and the path - ordered exponential @xmath5 $ ] ensures gauge invariance . appealing to its renormalization - group properties
@xcite , we can expand @xmath6 in terms of its one - loop eigenfunctions , alias the gegenbauer polynomials , to obtain @xmath7 + \ldots \label{eq : pion - da}\end{aligned}\ ] ] here @xmath8 is the asymptotic pion da and by virtue of the leptonic decay @xmath9 one has the normalization @xmath10 . relying only upon the first two gegenbauer coefficients , eq.([eq : pion - da ] )
can yield distinct profiles , as shown in fig.[fig : pion - dilemma ] .
we will see in the next section how two - photon processes can be used to resolve pion s dilemma and reveal its parton substructure in agreement with the experimental data and the latest lattice calculations .
gev@xmath11 . ]
the entire nonperturbative content of the pion da is encoded in the expansion coefficients @xmath12 which in turn can be derived from the moments @xmath13 where @xmath14 , that decrease with increasing polynomial order @xmath15 to 0 : @xmath16 . $ ] the evolution behavior of @xmath12 is controlled by the erbl equation @xcite ( for a pedagogical exposition , see @xcite ) . in our approach
@xcite we have determined @xmath17 at a normalization scale @xmath18 gev@xmath19 with the help of qcd sum rules with nonlocal condensates @xcite .
there are alternative approaches , based , for instance , on local qcd sum rules @xcite , instantons @xcite , etc .
the predictive power of these theoretical approaches was challenged by the high - precision data of the cleo collaboration on the pion - photon transition @xcite , in which one photon has a large virtuality @xmath20 , while the other is nearly on shell .
kroll and raulfs @xcite were the first to show that the popular cz pion da was overshooting these data considerably . other analyses , having recourse to light - cone sum rules with a spectral density obtained in the standard factorization scheme of perturbative qcd in leading order ( lo ) @xcite and next - to - leading - order ( nlo ) @xcite , followed , which established the following facts ( consult fig . [ fig : collage - cleo - lattice ] , drawn at the main scale , @xmath21 , probed in the cleo experiment ) : ( i ) the cz pion da is outside the @xmath22 error ellipse of the cleo data .
( ii ) the asymptotic pion da is outside the @xmath23 ellipse , while ( iii ) other proposed models close to that are at least @xmath24 s off .
( iv ) the `` bunch '' of pion das derived from nonlocal qcd sum rules mostly overlaps with the @xmath25 error ellipse , with the bms model da being entirely inside . of various pion da models with the cleo data in terms of the @xmath25 ellipse ( solid line ) and recent lattice simulations , denoted by vertical dashed lines @xcite and solid ones @xcite .
the symbols mark the models discussed in tables [ tab : second_moment ] and [ tab : fourth_moment ] .
the slanted shaded rectangle represents the bms da `` bunch '' @xcite .
the dashed @xmath25 ellipse corresponds to the inclusion of the twist-4 contribution to the pion da via renormalons @xcite .
the red bullet denotes the prediction derived from the holographic model based on ads / cft duality ( see last section ) . ]
( slanted broken lines ) intersecting with the @xmath25 error ellipse ( inner solid line ) of the cleo data and the latest lattice results on @xmath26 of the ukqcd - rbc collaboration @xcite ( vertical shaded band ) . ]
the most recent lattice calculations @xcite ( larger band bounded by dashed lines ) and @xcite ( narrower band within solid lines ) in fig .
[ fig : collage - cleo - lattice ] support and enhance these findings as regards the range of values of the first gegenbauer coefficient @xmath26 .
first , as one sees from this figure , they rather disfavor a relatively large twist-4 contribution to the pion da , estimated with the help of renormalons @xcite ( dashed @xmath25 error ellipse ) .
second , the latest calculation @xcite , with even smaller uncertainties than @xcite , indicates a trend further away from the asymptotic da , but still in compliance with the bms results .
tables [ tab : second_moment ] and [ tab : fourth_moment ] show the values of the second and the fourth moment of various pion das at the lattice reference scale @xmath27 gev@xmath28 . .predictions for @xmath29 from various models ( methods ) after nlo evolution from their proprietary normalization point to the scale @xmath27 gev@xmath28 .
[ tab : second_moment ] [ cols="<,<,<,<",options="header " , ] in fig .
[ fig : xi-4 ] , we determine the range of values of the fourth moment @xmath30 ( denoted by the slanted broken lines ) that simultaneously fulfill the cleo data and the lattice constraints on @xmath29 from @xcite . for each error ellipse
, there is some overlap with the vertical @xmath26-band estimated in @xcite . for the maximum and minimum of @xmath31 with respect to the @xmath25 error ellipse
, we find @xmath32 corresponding to @xmath33 and @xmath34 [ recall that @xmath35 it happens that these values almost coincide with the intersection points ( open circles ) of the @xmath25 error ellipse and the boundaries of the vertical band of the @xmath26 lattice constraints .
figure [ fig : pi - gamma - ff ] displays various theoretical predictions for @xmath36 vs. @xmath20 in comparison with the cello ( diamonds , @xcite ) and the cleo ( triangles , @xcite ) experimental data , evaluated with the twist-4 parameter value @xmath37 gev@xmath38 @xcite .
the other curves shown correspond to selected pion das : the asymptotic da @xmath39 ( lower dashed line ) , @xmath40 ( upper dashed line ) @xcite , and two instanton - based models , viz .
, @xcite ( dotted line ) and @xcite ( dash - dotted line ) . an important observation from this figure
is that the shaded strip , which corresponds to the bms `` bunch '' @xcite , becomes narrower at lower scales around 1 gev@xmath11 .
the reason is that at such low scales , the form factor is dominated by its twist-4 contribution , while the leading twist-2 part dies out .
moreover , fig .
[ fig : pi - gamma - ff ] makes it clear that the low-@xmath20 cello data @xcite exclude @xmath41 and clones , while the high-@xmath20 cleo data @xcite rule out @xmath42 .
figure [ fig : pi - gamma - pie ] gives an illustration of the partial contributions to @xmath43 , originating from different sources at @xmath44 .
a comprehensive account of these effects can be found in @xcite . vs. @xmath20 in comparison with available experimental data ( see text ) . ] at the typical scale @xmath45 gev@xmath11 of the cleo data @xcite .
the next - to - next - to - leading order ( nnlo ) estimate is based on @xcite , whereas the uncertainty owing to the ( res)onance model used , was discussed in @xcite . ]
the analysis of the pion s electromagnetic form factor involves on the nonperturbative side the bms `` bunch '' of pion das in comparison with @xmath41 and @xmath42 . on the perturbative side ,
a theoretical scheme is used , which consists of expanding the form factor in terms of analytic images of the strong running coupling and its powers up to the nlo .
the basis of this approach develops from @xcite and can account for more than one hard scale by incorporating into the `` analytization '' procedure all terms that contribute to the spectral density of the amplitude @xcite .
the bedrock of the approach was developed in @xcite and its application to @xmath46 was considered in @xcite ( see also @xcite ) .
the results are displayed in fig .
[ fig : pion - ff ] in comparison with experimental data .
note that the quantity shown comprises a soft non - factorizing contribution which dominates at currently accessible momentum transfers @xcite .
the main characteristics of this approach are ( i ) a strongly reduced sensitivity on the renormalization and the factorization scale , ( ii ) an undiminished quality of precision in adopting different choices of renormalization schemes and scale settings , virtually eliminating the dependence variations from scheme to scheme and scale to scale @xcite .
( iii ) another important finding is that , within such an analytic approach , the form - factor predictions ( shaded strip in fig .
[ fig : pion - ff ] ) turn out to be very close to that computed with @xmath41 , albeit the underlying pion da profiles are very different .
this proved that what really matters is the behavior of the pion da at the endpoints @xmath47 .
we have shown that the cleo data on @xmath48 pose a veto to a variety of proposed models for the pion da ( see fig . [
fig : collage - cleo - lattice ] ) and favor an endpoint - suppressed
b-shaped pion da a bactrian `` camelino '' , the endpoint suppression being provided by the vacuum nonlocality @xmath49 gev@xmath11 . the crucial ` missing link ' is the determination of @xmath31 on the lattice . a value within the range @xmath50$](2 gev ) , extracted from the cleo data in conjunction with the lattice constraints on @xmath29 from @xcite ( cf .
fig.[fig : xi-4 ] ) , would validate the claim that the pion da is bms - like .
the drell - yan process @xmath51 for lepton - pair production with a large invariant mass @xmath20 provides an additional useful tool to probe and test different pion das in terms of azimuthal asymmetries , as fig .
[ fig : dy ] illustrates for the case of the kinematic variable @xmath1 . as a function of @xmath52 for @xmath53 .
the shaded strip contains the results for the bms `` bunch '' of das @xcite .
the solid line corresponds to the bms model , the dotted solid denotes the result for @xmath41 , and the ( red ) dashed line is the prediction for the _ endpoint dominated _ cz da @xcite .
one - loop evolution of the pion das to each measured @xmath20 value is included ( data taken from @xcite ) . ]
overall , a rather good agreement of the bms `` bunch '' , derived from nonlocal qcd sum rules @xcite , with available data was found @xcite , though the existing data can not single out a particular model . in this respect ,
the planned compass experiment may be of significant relevance .
in this extra section , contained only in the arxiv version of this paper , we provide some predictions from the ads / qcd approach . in the holographic model of brodsky and de tramond ( see , e.g. , @xcite ) ,
the pion da has the form @xmath54 with the normalization @xmath55 and moments given by @xmath56 where @xmath57 is the euler beta function . taking the first derivative of @xmath58 , @xmath59
, one realizes that this da has its endpoints strongly enhanced , even relative to the asymptotic da , let alone to the bms da ( cf .
[ fig : pion - dilemma ] ) .
note that also the twist-4 da contribution , extracted from the holographic model by agaev and gomshi nobary @xcite , shows endpoint enhancement , although this effect is milder compared to the twist-2 da case .
it is interesting to recall at this point that eq .
( [ eq : bt - da ] ) was considered long ago by mikhailov and radyushkin in an attempt to reconstruct the pion da from its first few moments within the context of qcd sum rules with nonlocal condensates @xcite .
the values of the moments @xmath29 and @xmath31 of the holographic model have been inserted in tables [ tab : second_moment ] and [ tab : fourth_moment ] , respectively , ( last entry in each of these tables ) after nlo evolution from their initial values @xmath60 and @xmath61 to the lattice scale @xmath62 ( with associated gegenbauer coefficients @xmath63 and @xmath64 ) . the holographic model is also included in fig.[fig : collage - cleo - lattice ] in comparison with other theoretical models and confronted with the cleo data @xcite , in parallel with the two most recent and precise lattice calculations @xcite . as for the moments above , we have assumed that the holographic da , given by eq .
( [ eq : bt - da ] ) is normalized at the scale @xmath65 gev@xmath11 with the gegenbauer coefficients @xmath66 and @xmath67 .
we then performed a two - loop evolution to the scale @xmath44 and found @xmath68 the position of the holographic model is marked in fig.[fig : collage - cleo - lattice ] by a red bullet and turns out to be ( cf .
[ fig : xi-4 ] ) just inside the border of the @xmath24 ellipse of the cleo data , being also inside the predictions range of the qcdsf / ukqcd collaboration , but outside the limits on @xmath26 determined by the ukqcd - rbc07 collaboration .
predictions for the spacelike ( and timelike ) pion form factor , derived from ads / qcd , can be found , for example , in @xcite ; they are not addressed here .
i wish to thank a.p .
bakulev for collaboration and s.v .
mikhailov for collaboration and technical help in preparing this contribution .
i am indebted to the organizers of the workshop for financial support and the deutsche forschungsgemeinschaft ( dfg ) for a travel grant .
the works presented here were supported in part by the dfg under grant 436 rus 113/881/0 and the heisenberg - landau programme ( grant 2008 ) .
finally , i would like to thank s. brodsky and g. de tramond for attracting my attention to holographic qcd . | this talk presents issues pertaining to the quark structure of the pion within qcd , both from the theoretical and from the experimental point of view .
we review and discuss the pion - photon transition form factor and the pion s electromagnetic form factor vs. corresponding experimental data from the cleo collaboration and the jlab .
we also examine the extent to which recent high - precision lattice computations of the second moment of the pion s distribution amplitude conform with theoretical models . finally , we include predictions for the azimuthal asymmetry of the @xmath0 distribution in the polarized @xmath1-pair - induced dy production employing various pion distribution amplitudes . |
conventional time series analysis is heavily dependent on the assumption of stationarity . but this assumption is unsatisfactory for modelling many real - life phenomena that exhibit seasonal behaviour .
seasonal variations in the mean of time series data can be easily removed by a variety of methods , but when the variance varies with the season , the use of periodic time series models is suggested . in order to model periodicity in autocorrelations ,
a class of periodic autoregressive moving average ( parma ) models was introduced .
the parma@xmath2 system is given by : @xmath3 where @xmath4 and the coefficients @xmath5 and @xmath6 are nonzero sequences periodic in @xmath7 with the same period @xmath8 while the innovations @xmath9 are independent gaussian random variables .
for such parma models , the covariance function is a tool for describing the dependence structure of the time series and we can recall that the sequence given by ( [ parma ] ) is periodically correlated , more precisely the covariance function @xmath10 is @xmath11periodic in @xmath12 for every @xmath13 ( see @xcite ) . as the coefficients in ( [ parma ] ) are periodic , it is obvious that the class of parma models is an extension of the class of commonly used arma models . due to their interesting properties ,
parma systems have received much attention in the literature and turned out to be an alternative to the conventional stationary time series as they allow for modelling many phenomena in various areas , e.g. , in hydrology ( @xcite ) , meteorology ( @xcite ) , economics ( @xcite ) and electrical engineering ( @xcite ) .
the assumption of normality for the observations seems not to be reasonable in the number of applications , such as signal processing , telecommunications , finance , physics and chemistry , and heavy - tailed distributions seem to be more appropriate , see e.g. @xcite .
an important class of distributions in this context is the class of @xmath1-stable ( stable ) distributions because it is flexible for data modelling and includes the gaussian distribution as a special case .
the importance of this class of distributions is strongly supported by the limit theorems which indicate that the stable distribution is the only possible limiting distribution for the normed sum of independent and identically distributed random variables .
stable random variables have found many practical applications , for instance in finance ( @xcite ) , physics ( @xcite ) , electrical engineering ( @xcite ) .
parma models with symmetric stable innovations combine the advantages of classical parma models and stable distributions they offer an alternative for modelling periodic time series with heavy tails .
however , in this case the covariance function is not defined and thus other measures of dependence have to be used .
the most popular are the covariation and codifference presented in @xcite . in this paper
we consider a special case of stable parma models , i.e. parma(1,1 ) systems with symmetric @xmath1-stable innovations . in this case
we rewrite the equation ( [ parma ] ) in a simple way , i.e. : @xmath14 where @xmath4 and the coefficients @xmath15 and @xmath16 are nonzero periodic sequences with the same period @xmath8 while the innovations are independent symmetric @xmath1-stable ( s@xmath1s for short ) random variables given by the following characteristic function : @xmath17 where @xmath18 denotes the scale parameter .
let us define @xmath19 and @xmath20 , with the convention @xmath21 if @xmath22 .
as the covariance function is not defined for stable random vectors , in section [ s_mofdep ] we present two other measures of dependence that can be used for symmetric stable time series the covariation and the codiffrence . in section [ s_bsol ] we discuss the necessary conditions for existence of the bounded solution of parma(1,1 ) systems for @xmath23 and we note that results obtained in @xcite for parma systems with gaussian innovations can be extended to the case of stable innovations .
the covariation and the codifference for parma(1,1 ) models with stable innovations are studied in section [ s_depstr ] and the asymptotic relation between these two measures of dependence for the considered models is examined there .
we find it interesting to illustrate theoretical results and thus in section [ s_ex ] we give an example of parma(1,1 ) systems with s@xmath1s innovations and illustrate the periodicity of the considered measures of dependence and the asymptotic relation between them .
let @xmath24 and @xmath25 be jointly s@xmath1s and let @xmath26 be the spectral measure of the random vector @xmath27 ( see for instance @xcite ) . if @xmath28 then the covariance is not defined and thus other measures of dependence have to be used .
the most popular measures are : the * covariation @xmath29 * of @xmath24 on @xmath25 defined in the following way : @xmath30 where @xmath31 and the signed power @xmath32 is given by @xmath33 , and the * codifference @xmath34 * of @xmath24 on @xmath25 defined for @xmath35 : @xmath36 properties of the considered measures of dependence one can find in @xcite .
let us only mention here that , in contrast to the codifference , the covariation is not symmetric in its arguments .
moreover , when @xmath37 both measures reduce to the covariance , namely @xmath38 the covariation induces a norm on the linear space of jointly s@xmath1s random variables @xmath39 and this norm is equal to the scale parameter , see @xcite . hence throughout this paper the norm ( so - called covariation norm ) @xmath40 is defined by @xmath41 , for a s@xmath1s random variable @xmath24 and @xmath42 .
the sequence @xmath43 , is bounded in a space @xmath39 with norm @xmath44 if @xmath45 . moreover , in this paper we write @xmath46 in @xmath39 if and only if @xmath47 . if it is possible to transform the sequence @xmath48 to the moving average representation @xmath49 , where the innovations @xmath9 are independent s@xmath1s random variables with parameters @xmath18 and @xmath50 , then both the covariation and the codifference can be expressed in terms of the coefficients @xmath51 ( see @xcite ) : @xmath52 @xmath53
in this section we consider parma(1,1 ) system given by ( [ parma12 ] ) with the s@xmath1s innovations for @xmath54 and we investigate when the bounded solution exists .
let us assume that @xmath55 .
in this case we show that the considered parma(1,1 ) system has a bounded solution given by the formula : @xmath56 provided that every @xmath57 can be represented as @xmath58 for some @xmath59 and @xmath60 , we have @xmath61 now it sufficies to notice that by periodicity of coefficients and @xmath62 and that @xmath63 .
thus @xmath64 where @xmath65 is a real constant , @xmath66 .
this implies that @xmath67 and hence @xmath48 given by ( [ forma1 ] ) is bounded .
moreover , it is easy to check that @xmath48 given by ( [ forma1 ] ) satisfies equation ( [ parma12 ] ) . to prove the converse ,
let us note that iterating the equation ( [ parma12 ] ) yields : @xmath68 for each @xmath4 and @xmath69 by periodicity of the coefficients we have @xmath70 .
therefore , if @xmath55 and @xmath48 is a bounded solution of ( [ parma12 ] ) then @xmath71 this means that the bounded solution of the system considered in this section is given by @xmath72 in a similar manner , it can be shown , that if @xmath73 , then the bounded solution of the system under study is given by @xmath74 it is worth pointing out here that the solution for the parma(1,1 ) system with s@xmath1s innovations takes the same form of the moving average as in the case of gaussian innovations obtained in @xcite and it reduces to well - known formula for arma models in case of constant coefficients .
let us consider parma(1,1 ) system given by ( [ parma12 ] ) with s@xmath1s innovations for @xmath23 .
moreover , we will restrict our attention to the case @xmath55 because this case is more important for applications as shown in section [ s_bsol ] , the considered system has a bounded solution @xmath48 in the linear space of jointly s@xmath1s random variables with norm @xmath44 given by the causal moving average representation ( [ forma1 ] ) . in order to investigate the dependence structure of stable parma(1,1 ) models we will first rewrite @xmath75 as @xmath76 , where @xmath77 and then using results presented in section [ s_mofdep ] we will find formulas for the covariation and the codifference .
we will also study periodicity of these measures and their asymptotic relation . *
the covariation . *
if @xmath78 , then for @xmath50 we have @xmath79 it is easy to notice that @xmath80 and @xmath81 .
therefore , for @xmath82 we obtain @xmath83 as every @xmath57 can be represented as @xmath58 for some @xmath59 and @xmath60 , we have @xmath84 now it suffices to notice that by periodicity of coefficients and @xmath85 and that @xmath63 .
thus @xmath86 if @xmath87 , then the covariation is given by @xmath88 in this case @xmath89 which results in formula @xmath90 the simple calculation similar like in the previous case leads us to the result @xmath91 as the sequences @xmath16 and @xmath15 are periodic in @xmath7 with period @xmath8 , the covariation is periodic in @xmath7 and @xmath92 with the same period , indeed @xmath93 . * the codifference . * for @xmath50 and
@xmath78 we have to calculate @xmath94 observe that there are @xmath95 and @xmath96 ( as @xmath97 ) in each part of the above formula .
therefore we can write @xmath98 that can be transformed to @xmath99 and now it is sufficient to notice that the sum @xmath100 has been already calculated for the covariation .
so we obtain @xmath101 it is not necessary to calculate the codifference for @xmath87 as this measure is symmetric in its arguments , i.e. for every @xmath102 we have @xmath103 . by the assumption , the coefficients @xmath16 and @xmath15 are periodic in @xmath7 with period @xmath8 .
thus it is not difficult to notice that @xmath104 , which means that the codifference is periodic in @xmath7 and @xmath92 with the period @xmath8 . *
asymptotic relation between cv and cd .
* using formulas ( [ cv1 ] ) , ( [ cv2 ] ) and ( [ cd1 ] ) and two simple facts that hold for @xmath54 , @xmath105 and any real constant @xmath106 : @xmath107 @xmath108 it can be proved that for the considered parma(1,1 ) systems with s@xmath1s innovations the following relations hold for each @xmath4 and @xmath54 : @xmath109 @xmath110 the interesting point is that , because of asymmetry of the covariance , we obtained quite different asymptotic results . moreover , ( [ glowne11 ] )
extends the results obtained in @xcite for arma models and it reduces to ( [ covcvcd ] ) for @xmath37 .
in order to illustrate our theoretical results let us consider parma(1,1 ) model with s@xmath1s innovations with @xmath111 , where the coefficients are given by @xmath112 it is clear that the coefficients are periodic with @xmath113 and in this case @xmath114 .
therefore we are allowed to use formulas obtained in sections [ s_bsol ] and [ s_depstr ] .
we first want to demonstrate how the parameter @xmath1 influences the behaviour of the time series , so we plot 1000 realizations of the considered model for @xmath37 , @xmath115 and @xmath116 , see figure [ fig_real ] .
it is easy to notice that the smaller @xmath1 we take , the greater values of the time series can appear ( property of heavy - tailed distributions ) .
next , we want to show the dependence structure of the considered model , especially periodicity of measures of dependence . in order to do this
we plot the codifference @xmath117 and the covariation @xmath118 for @xmath115 and @xmath116 in case of @xmath119 , see figure [ fig_miary ] .
although the behaviour of the measures of dependence depends on the parameter @xmath1 , one can observe that both measures are periodic with the same period @xmath113 .
finally , let us illustrate the asymptotic relation between the covariation and the codifference that is studied in section [ s_depstr ] .
figure [ fig_asym ] contains plots of the functions @xmath120 , @xmath121 , @xmath122 and @xmath123 for @xmath124 , @xmath125 and @xmath115 and @xmath116 . according to the theoretical results , the first two quotients tend to 1 and the next two tend to 0 as @xmath13 increases .
e.g. gladyshev : _ sov .
_ * 2 * , 4 ( 1961 ) .
a. makagon , a.g .
miamee , h. salehi : in _ nonstationary stochastic processes and their applications _ , edited by a. g. miamee ( word scientific 1991 , p. 147 ) a. makagon , a. weron , a. wyomaska : _ appl
_ * 31 * , 273 ( 2004 ) .
vecchia : _ water resources bulletin _ * 21 * , 721 ( 1985 ) .
p. bloomfield , h.l .
hurd , r.b .
j. time ser .
* 15 * , 127 ( 1994 ) .
e. broszkiewicz - suwaj , a. makagon , r. weron , a. wyomaska : _ physica a _ * 336 * , 196 ( 2004 ) .
e. parzen , m. pagano : _ journal of econometrics _ * 9 * , 137 ( 1979 )
. w. gardner , l.e .
franks : _ ieee transactions of information theory _ * 21 * , 4 ( 1975 ) .
s. mittnik , s.t .
ratchev : _ stable paretian models in finance _ , wiley , new york , 2000 .
a. janicki , a. weron : _ statist .
* 9 * , 109 ( 1994 ) .
b. stuck , b. kleiner : _ bell system technical journal _ * 53 * , 1263 ( 1974 ) .
g. samorodnitsky , m.s .
taqqu : _ stable non - gaussian random processes _ , chapman & hall , new york , 1994 .
j. nowicka , a. weron : _ annales _ * li 1 ( 14 * ) , 133 ( 1997 )
. j. nowicka : _ stochastic models _ * 13 * , 673 ( 1997 )
. j. nowicka - zagrajek , a. wyomaska : `` asymptotic behaviour of measures of dependence for arma(1,2 ) models with stable innovations .
stationary and non - stationary coefficients '' , research report hsc/06/7
( 2006 ) , wroclaw university of technology . | in this paper we investigate the dependence structure for parma models ( i.e. arma models with periodic coefficients ) with symmetric @xmath0stable innovations . in this case the covariance function is not defined and therefore other measures of dependence have to be used . after obtaining the form of the bounded solution of the parma system with symmetric @xmath1-stable innovations , we study the codifference and the covariation the most popular measures of dependence defined for symmetric stable time series .
we show that both considered measures are periodic .
moreover we determine the cases when the codifference and the covariation are asymptotically proportional with the coefficient of proportionality equal to @xmath1 .
= by -1 pacs : 89.65.gh , 05.45.tp , 02.50.cw |
[ fig1 ] = 0.99
[ fig2 ] = 0.83
the density matrix @xmath7 representing the state of an @xmath4-level quantum system is a positive , unit - trace @xmath8 matrix .
we denote by @xmath101 the set of all @xmath102 complex matrices , and by @xmath103 the set of all density matrices .
a map @xmath104 is positive if @xmath105 for any @xmath106 in @xmath107 .
a map @xmath108 is completely positive ( cp ) if for any @xmath109 the map @xmath110 is positive ( @xmath111 is the identity map in @xmath112 ) .
a cp map is trace - preserving if @xmath113 = \mathrm{tr } ( \rho)$ ] for any @xmath114 .
we denote by @xmath11 the set of all cp , trace - preserving maps acting in @xmath107 , referred to as _
kraus maps _ or _ quantum operations _ @xcite .
r. fanciulli , a. m. weiner , m. m. dignam , d. meinhold , and k. leo , phys .
b * 71 * , 153304 ( 2005 ) ; n. dudovich , t. polack , a. peer , and y. silberberg , phys . rev . lett .
* 94 * , 083002 ( 2005 ) ; z. amitay , a. gandman , l. chuntonov , and l. rybak , _ ibid .
_ * 100 * , 193002 ( 2008 ) ; n. timoney , v. elman , s. glaser , c. weiss , m. johanning , w. neuhauser , and c. wunderlich , phys . rev .
a * 77 * , 052334 ( 2008 ) ; m. j. biercuk , h. uys , a. p. vandevender , n. shiga , w. m. itano , and j. j. bollinger , _ ibid .
_ * 79 * , 062324 ( 2009 ) .
h. stapelfeldt and t. seideman , rev .
phys . * 75 * , 543 ( 2003 ) ; m. dantus and v. v. lozovoy , chem . rev . * 104 * , 1813 ( 2004 ) ; y. silberberg , ann . rev
. phys . chem . * 60 * , 277 ( 2009 ) ; f. krausz and m. ivanov , rev .
phys . * 81 * , 163 ( 2009 ) .
a. assion _ et al .
_ , science * 282 * , 919 ( 1998 ) ; d. meshulach and y. silberberg , nature * 396 * , 239 ( 1998 ) ; t. c. weinacht , j. ahn , and p. h. bucksbaum , _ ibid . _ * 397 * , 233 ( 1999 ) ; r. bartels _ et al .
_ , _ ibid .
_ * 406 * , 164 ( 2000 ) ; t. brixner , n. h. damrauer , p. niklaus , and g. gerber , _ ibid . _
* 414 * , 57 ( 2001 ) ; j. l. herek , w. wohlleben , r. j. cogdell , d. zeidler , and m. motzkus , _ ibid . _ * 417 * , 533 ( 2002 ) ; r. j. levis , g. m. menkir , and h. rabitz , science * 292 * , 709 ( 2001 ) ; t. brixner _
lett . * 92 * , 208301 ( 2004 ) ; a. lindinger , c. lupulescu , m. plewicki , f. vetter , a. merli , s. m. weber , and l. wste , _ ibid . _ * 93 * , 033001 ( 2004 ) ; v. i. prokhorenko , a. m. nagy , s. a. waschuk , l. s. brown , r. r. birge , and r. j. d. miller , science * 313 * , 1257 ( 2006 ) ; t. laarmann _ et al .
. lett . * 98 * , 058302 ( 2007 ) ; m. p.
a. branderhorst , p. londero , p. wasylczyk , c. brif , r. l. kosut , h. rabitz , and i. a. walmsley , science * 320 * , 638 ( 2008 ) ; m. j. biercuk , h. uys , a. p. vandevender , n. shiga , w. m. itano , and j. j. bollinger , nature * 458 * , 996 ( 2009 ) ; m. roth , l. guyon , j. roslund , v. boutou , f. courvoisier , j .-
wolf , and h. rabitz , phys .
* 102 * , 253001 ( 2009 ) .
h. rabitz , r. de vivie - riedle , m. motzkus , and k. kompa , science * 288 * , 824 ( 2000 ) ; r. j. levis and h. a. rabitz , j. phys .
a * 106 * , 6427 ( 2002 ) ; d. goswami , phys . rep . * 374 * , 385 ( 2003 ) ; t. brixner and g. gerber , chemphyschem * 4 * , 418 ( 2003 ) ; m. wollenhaupt , v. engel , and t. baumert , ann .
phys . chem . *
56 * , 25 ( 2005 ) ; w. wohlleben , t. buckup , j. l. herek , and m. motzkus , chemphyschem * 6 * , 850 ( 2005 ) ; t. pfeifer , c. spielmann , and g. gerber , rep . prog
. phys . * 69 * , 443 ( 2006 ) ; p. nuernberger , g. vogt , t. brixner , and g. gerber , phys .
* 9 * , 2470 ( 2007 ) ; c. winterfeldt , c. spielmann , and g. gerber , rev .
phys . * 80 * , 117 ( 2008 ) .
a. c. doherty , s. habib , k. jacobs , h. mabuchi , and s. m. tan , phys .
a * 62 * , 012105 ( 2000 ) ; a. doherty , j. doyle , h. mabuchi , k. jacobs , and s. habib , in _ proceedings of the 39th ieee conference on decision and control _
( ieee , 2000 ) , vol . 1 , p. 949 .
r. l. kosut , r. j. levis , h. mabuchi , h. rabitz , i. a. walmsley , and e. yablonovich , `` managing uncertainties in the control of quantum systems , '' _ summary of 2003 princeton - oxford workshop on control of quantum systems _
( unpublished ) . | many proposals have been put forth for controlling quantum phenomena , including open - loop , adaptive feedback , and real - time feedback control .
each of these approaches has been viewed as operationally , and even physically , distinct from the others .
this work shows that all such scenarios inherently share the same fundamental control features residing in the topology of the landscape relating the target physical observable to the applied controls . this unified foundation may provide a basis for development of hybrid control schemes that would combine the advantages of the existing approaches to achieve the best overall performance
.
steering the dynamics of quantum systems by means of external controls is a central goal in many areas of modern science , ranging from the creation of selective molecular transformations to quantum information processing @xcite .
the control may be either coherent ( e.g. , a shaped laser pulse @xcite ) or incoherent ( e.g. , a tailored environment @xcite or a sequence of quantum measurements @xcite ) .
quantum control is proving to be successful in the laboratory for manipulating a broad variety of physical , chemical , and biologically relevant processes @xcite .
many seemingly distinct approaches have been developed for controlling quantum phenomena @xcite , including open - loop control ( olc ) in which model - based designs are directly applied in experiments @xcite , adaptive feedback control ( afc ) which is a measurement - guided closed - loop laboratory procedure including resetting the system to the initial state on each loop iteration @xcite , and real - time feedback control ( rtfc ) which involves quantum measurement back - action on the system upon traversing the loop @xcite .
each of these control schemes has been argued to be operationally , and even physically , distinct from the others .
this rapid communication shows that all such scenarios are unified on a fundamental level by the common character of the control landscape @xcite which relates the physical objective @xmath0 $ ] to the control variables @xmath1 .
this general foundation makes it possible to extend the powerful landscape - based methods of optimality analysis , originally developed for olc and afc @xcite , to rtfc .
furthermore , one can envision combinations of such schemes @xcite , or even the prospect of new unanticipated ones , possibly arising in the future , all of which will share the same fundamental control landscape features .
the structure of the control landscape ( i.e. , its topology characterized by the nature of the local and global extrema ) determines the efficacy of a search for optimal solutions to the posed control problem @xcite .
searching for the global maximum of the objective function @xmath0 $ ] in the laboratory would be significantly hindered by the existence of local maxima that act as traps during the optimization procedure .
even stochastic optimization algorithms could be impeded by a high density of traps @xcite .
if the objective function has no traps , then , in principle , nothing lies in the way of reaching the global maximum ( i.e. , the best possible yield ) , except limited controls .
this work establishes that a common landscape topology is shared by all quantum control schemes , leading to two general conclusions about controlling quantum phenomena .
first , there are no local optima to act as traps on the landscape for a wide class of control problems , thereby enabling highly efficient searches for globally optimal controls .
second , for each particular objective @xmath2 , there exists a single special control @xmath3 which is universally optimal for all initial states of a system ; this result implies that @xmath3 provides inherent robustness to variations of the initial system s state . for practical applications
, these generic fundamental properties can facilitate more flexible laboratory implementations of quantum control as well as guide the design of suitable algorithms for laboratory optimization @xcite .
the control landscapes for olc and afc were studied in recent works @xcite . for closed quantum systems ,
a control objective @xmath2 can be cast as a function of unitary evolution operators .
assuming evolution - operator controllability ( i.e. , that any unitary evolution can be produced by the available controls ) , the landscapes were shown to have no traps for typical objectives , including the expectation value of an observable @xcite and the fidelity of a unitary quantum gate @xcite .
the absence of landscape traps was also analyzed for closed systems in the space of actual controls @xcite .
the landscapes for olc and afc of open quantum systems can be studied by casting an objective as a function of completely positive , trace - preserving maps ( i.e. , kraus maps @xcite ) which describe the system evolution . assuming kraus - map controllability ( i.e. , that any kraus map can be produced by the available controls )
, the trap - free landscape topology was established for open - system observable control @xcite .
the analysis of the control landscapes for rtfc is performed below to significantly extend the prior analysis @xcite and provide a single complete framework unifying the landscapes for all quantum control schemes , including olc , afc , and rtfc . two distinct approaches to rtfc of quantum systems
are considered . the first one is based on feedback employing measurements of some quantum system output channel to guide a classical controller @xcite .
the other feedback approach is a quantum analog of watt s flyball governor a self - regulating quantum machine . in this scenario ( called coherent rtfc ) , no measurements with a classical output signal are involved , and instead another quantum subsystem is employed to facilitate the control action , so that the evolution of the composite quantum system , consisting of the `` plant '' ( the subsystem of interest ) and `` controller '' ( the auxiliary subsystem ) , is coherent @xcite .
the performance of practical rtfc controllers ( classical , quantum , or both ) can be , in principle , optimized using afc @xcite , possibly guided by an initial olc design , further enhancing the significance of a universal landscape topology common to all quantum control approaches . in rtfc based on selective measurements of a single quantum system ,
the outcome of each measurement is random . after averaging over the random processes corresponding to all possible measurement outcomes , feedback control produces kraus - type evolution of the controlled system . in rtfc based on measurements of an ensemble of quantum systems ,
the ensemble average produces a kraus map as well . in coherent rtfc , if the plant and controller are initially prepared in a product state , the evolution of the plant is once again represented by a kraus map .
therefore , all of these incarnations of rtfc generate kraus - type evolution .
arbitrary kraus maps can be engineered , for example , by using a simple quantum measurement combined with coherent feedback actions @xcite . adopting a previous analysis @xcite , we show , under the kraus - map controllability assumption , that quantum control landscapes have no traps for all considered types of rtfc .
we will start with basic definitions , then prove that each considered type of rtfc produces kraus - type evolution , and , finally , use these results to determine the quantum control landscape topology .
the conclusion of the rapid communication will then draw the analysis together for a unified control landscape formulation including olc , afc , and rtfc . in practically important situations
the evolution of an open quantum system can be represented by a kraus map @xcite . for an @xmath4-level system
, any such map @xmath5 can be cast as the operator - sum representation ( osr ) @xcite : @xmath6 , where @xmath7 is the density matrix ( a positive , unit - trace @xmath8 matrix ) and @xmath9 is a set of kraus operators ( complex @xmath8 matrices satisfying the trace - preserving condition @xmath10 ) .
the osr is not unique : any map @xmath5 can be represented using infinitely many sets of kraus operators .
we denote the set of all kraus maps by @xmath11 .
the kraus - map description of open - system dynamics is very general and includes both markovian and non - markovian regimes @xcite .
a generalized quantum measurement with @xmath12 possible outcomes @xmath13 ( @xmath14 ) is characterized by a family of kraus operators @xmath15 .
we denote the set of outcomes and corresponding kraus operators for a given quantum measurement as @xmath16 . in particular , the projective measurement of a hermitian observable @xmath17 with eigenvalues @xmath18 and spectral projectors @xmath19 corresponds to the case @xmath20 .
if the measurement starts at time @xmath21 and the system density matrix before the measurement is @xmath22 , then the probability of the measurement outcome @xmath18 will be @xmath23 $ ] .
when a selective measurement of duration @xmath24 is performed and the outcome @xmath18 is observed , the density matrix evolves as @xmath25 , where @xmath26 if the measurement is nonselective ( i.e. , the measurement outcome is not observed ) , the corresponding evolution @xmath27 will be the average over all possible measurement outcomes : @xmath28 the evolution in eq .
( [ eq : non - selective ] ) defines a kraus map @xmath29=\rho(t+\tau_{\text{m}})$ ] , which is completely determined by the measurement @xmath30 .
in rtfc , the measurement ( or , in another variation , the interaction with an auxiliary quantum `` controller '' ) alters the evolution of the quantum system at each feedback iteration .
thus , the same quantum system is followed in real time in the feedback loop .
implementing rtfc on the atomic or molecular scale appears to be a very challenging technical problem , but its practical realization promises to significantly improve the ability to stabilize and control quantum systems .
consider now rtfc of a single quantum system , where a discrete series of selective measurements is performed and each measurement is followed by a feedback action ( continuous feedback can be treated as the limit of the discrete case , resulting in the same control landscape topology ) . for the discrete case ,
the @xmath31th iteration of the feedback process consists of a measurement @xmath32 with possible outcomes @xmath33 at time @xmath34 , followed by a feedback action dependent on the measurement outcome ( the measured observables and feedback actions can be distinct at different iterations ) .
the feedback action may be generally represented by a kraus map that depends on the measurement outcome ( a special case of the feedback action is a unitary transformation corresponding to a coherent control ) .
if at the @xmath31th iteration the measurement outcome @xmath35 is observed , then the feedback action ( of duration @xmath36 ) described by the kraus map @xmath37 will be applied to the system , so that the density matrix will evolve as @xmath38 $ ] , where @xmath39 is of the form ( [ eq : selective ] ) , with the corresponding probability @xmath40 .
the map @xmath37 also includes the free evolution and influence of the environment .
the system evolution after one feedback iteration , averaged over all possible measurement outcomes , @xmath41 , is therefore given by the transformation @xmath42 $ ] , where @xmath43 = \sum_{\alpha } \lambda^i_{\alpha } \bigg [ \sum_{\beta } k^i_{\alpha,\beta } \rho(t_i ) ( k^i_{\alpha,\beta})^{\dagger } \bigg]\ ] ] and @xmath44 are the kraus operators characterizing the measurement @xmath32 .
let the osr of the feedback - action kraus map @xmath37 be @xmath45 , where @xmath46 . then eq . ( [ eq : one - iteration ] ) can be rewritten as @xmath47 = \sum_{\nu,\alpha,\beta } z^i_{\nu,\alpha,\beta } \rho(t_i ) ( z^i_{\nu,\alpha,\beta})^{\dagger } ,
\ ] ] where @xmath48 . since @xmath49 , the transformation @xmath50 of eqs .
( [ eq : one - iteration ] ) and ( [ eq : phi - one - iteration ] ) is a kraus map .
thus , eq . ( [ eq : phi - one - iteration ] ) shows that the average evolution for one feedback iteration is of the kraus type for arbitrary ( generalized ) quantum measurement and any ( coherent or incoherent ) feedback action .
for a special case of a pure , least - disturbing measurement and a coherent feedback action , this result was obtained in ref .
@xcite .
the entire feedback process for controlling the system from the initial to final time is characterized by a sequence of measurements and feedback actions : @xmath51 , where @xmath32 ( @xmath52 ) is the measurement for the @xmath31th iteration , @xmath53 is the set of all feedback actions ( kraus maps ) for the @xmath31th iteration , and @xmath54 is the number of iterations .
different trials of the feedback process will , in general , produce distinct evolutions , resulting in different system states at the final time @xmath55 .
the average output of the feedback process is given by averaging over all possible evolutions , which produces the transformation @xmath56 $ ] , where @xmath57 is a kraus map given by the composition of one - iteration kraus maps of the form ( [ eq : phi - one - iteration ] ) .
consider now rtfc of an ensemble of identical quantum systems , where measurements record the expectation value @xmath58 $ ] of an observable @xmath59 .
the density matrix representing the state of the ensemble undergoes a transformation characteristic for a nonselective measurement ( i.e. , with averaging over all possible measurement outcomes ) : @xmath60 $ ] , where the kraus map @xmath61 is defined by eq .
( [ eq : non - selective ] ) . the feedback action conditioned upon the measured value @xmath62
is generally represented by a kraus map @xmath63 , so that the ensemble evolution for one feedback iteration is @xmath64 $ ] , where the kraus map @xmath65 is the composition of the kraus maps representing the ensemble measurement and feedback action for the @xmath31th iteration .
similar to the single - system case , the overall transformation for the entire feedback process is @xmath56 $ ] , where the kraus map @xmath66 is again the composition of one - iteration kraus maps .
consider now coherent rtfc , where the quantum subsystem of interest ( the plant ) interacts with an auxiliary quantum subsystem ( the controller ) , and the evolution of the composite system is coherent : @xmath67 .
here , @xmath68 and @xmath69 are the density matrix and the unitary evolution operator , respectively , for the composite system .
an external coherent control field ( which is generally time - dependent ) can act on the plant , controller , or both .
the state of the plant at any time @xmath21 is represented by the reduced density matrix @xmath70 $ ] , where @xmath71 denotes the trace over the controller s degrees of freedom .
if the initial state of the composite system is in tensor product form , @xmath72 , then the evolution of the plant is represented by a kraus map : @xmath73 = \sum_{\nu } k^u_{\nu } \rho(0 ) ( k^u_{\nu})^{\dagger}$ ] , where the kraus operators @xmath74 depend on the evolution operator @xmath69 of the composite system @xcite . with the analysis above
, we can now assess the topology of quantum control landscapes for all considered types of rtfc . to be specific ,
a prevalent problem in quantum control is to maximize the expectation value of some target observable of the controlled system .
as we established above , in measurement - based rtfc for both a single quantum system and an ensemble , the average density matrix at the final time @xmath55 is given by @xmath75 $ ] , where the kraus map @xmath76 is determined by the set @xmath77 of measurements and feedback actions at all iterations of the feedback process .
the goal is to find an optimal feedback process @xmath78 that maximizes the expectation value of the target observable @xmath79 at the final time , @xmath80 $ ] .
the feedback process @xmath77 plays the role of a set of controls , and the objective function has the form : @xmath81 = \text{tr } \big\{\hat{a } \phi_f[\rho(0 ) ] \big\}$ ] .
an osr of the kraus map @xmath76 is given by @xmath82 = \sum_{\nu } k^f_{\nu } \rho(0 ) ( k_{\nu}^f)^{\dagger}$ ] , where @xmath83 is a set of kraus operators depending on the feedback process @xmath77 .
the objective @xmath81 $ ] can be cast as a function of kraus operators : @xmath84 = \text{tr } \left [ \hat{a } \textstyle{\sum_{\nu } } k^f_{\nu } \rho(0 ) ( k_{\nu}^f)^{\dagger } \right ] .\ ] ] we assume that the set @xmath85 of all available feedback processes ( i.e. , all available measurements and feedback actions ) is rich enough to produce all possible kraus maps , i.e. , that the system is kraus - map controllable . for coherent rtfc , the plant also undergoes a kraus - type evolution , @xmath73 $ ] , and the control objective @xmath86 can also be cast as a function of the kraus operators : @xmath87 = \text{tr } \big [ \hat{a } \sum_{\nu } k^u_{\nu } \rho(0 ) ( k_{\nu}^u)^{\dagger } \big]$ ] , which has the same form as eq .
( [ eq : of - k ] ) . for a sufficiently large controller prepared initially in a pure state ,
evolution - operator controllability of the composite system is a sufficient condition for kraus - map controllability of the plant @xcite .
building on recent olc and afc - motivated analysis of control landscape topology @xcite for a system that is kraus - map controllable , the objective function of the form ( [ eq : of - k ] ) has no local maxima for any initial state @xmath88 and any hermitian operator @xmath79 .
this result implies that _ quantum control landscapes for olc , afc , measurement - based rtfc , coherent rtfc , and combinations thereof , all have the same trap - free topology _ , provided that the available controls are sufficient to produce any kraus map .
furthermore , all local extrema of the objective function of the form ( [ eq : of - k ] ) are saddles @xcite that can be easily evaded by a suitable algorithm guiding an ascent over the landscape .
the trap - free control landscape topology is established above in the space of kraus maps , @xmath11 .
the control landscape also will be trap - free in the space of actual controls , @xmath89 , if the tangent map from @xmath89 to @xmath11 is surjective everywhere in @xmath89 .
although , in general , there exist so - called singular controls @xcite at which surjectivity does not hold , numerical results show that their impact on the search for optimal controls should be negligible @xcite . another important consequence of controllable kraus - type evolution for various types of quantum olc , afc and rtfc is the existence of a special control ( e.g. , for measurement - based rtfc , a special feedback process @xmath90 ) that is optimal for _ all _ initial states of the system .
let @xmath91 be the state maximizing the target expectation value : @xmath92 = \max_{\rho } \text{tr}(\hat{a } \rho)$ ] , and let the spectral decomposition of this final state be @xmath93 is the probability to find the system in the state @xmath94 .
for an arbitrary orthonormal basis @xmath95 in the system s hilbert space , define operators @xmath96 generates evolution @xmath97 = \rho(t)$ ] for all initial states @xmath88 @xcite .
therefore , the control that produces the kraus map @xmath98 ( e.g. , for measurement - based rtfc , the feedback process @xmath99 such that @xmath100 ) will be optimal for all initial states and this control will be robust to variations of the initial system state .
this work shows that the operationally and technologically distinct quantum control approaches of olc , afc , and rtfc share , under the condition of kraus - map controllability , a unified control landscape structure implying two common fundamental properties .
first , all such control schemes are characterized by the absence of landscape traps , with all local extrema being saddles .
second , special controls exist which are universally optimal for all initial states . these findings establish that ( 1 ) there are no inherent landscape features hindering attainment of the highest possible control yield and ( 2 ) suitable controls can provide broad scale robustness to variations of the initial conditions .
moreover , these properties are valid for any quantum control scheme which produces kraus - type evolution of the system , including feedback - based and open - loop approaches , as well as their combinations .
the unification of these seemingly different approaches at a conceptual level ( see figures in supplementary material @xcite ) should , in turn , provide a basis to ultimately unite the currently distinct laboratory realizations of quantum feedback control and open - loop control to attain the best performance under all possible conditions .
this work was supported by nsf and aro .
99 c. brif , r. chakrabarti , and h. rabitz , http://dx.doi.org/10.1088/1367-2630/12/7/075008[new j. phys .
* 12 * , 075008 ( 2010 ) . ]
a. p. peirce , m. a. dahleh , and h. rabitz , http://dx.doi.org/10.1103/physreva.37.4950[phys .
rev .
a * 37 * , 4950 ( 1988 ) ; ] s. shi and h. rabitz , http://dx.doi.org/10.1063/1.458438[j .
chem .
phys . * 92 * , 364 ( 1990 ) . ] n. timoney _ et al . _ , http://dx.doi.org/10.1103/physreva.77.052334[phys .
rev .
a * 77 * , 052334 ( 2008 ) ; ] m. j. biercuk _ et al . _ , http://dx.doi.org/10.1103/physreva.79.062324[_ibid._ * 79 * , 062324 ( 2009 ) ; ] y. silberberg , http://dx.doi.org/10.1146/annurev.physchem.040808.090427[annu
rev .
phys .
chem . * 60 * , 277 ( 2009 ) ; ] f. krausz and m. ivanov , http://dx.doi.org/10.1103/revmodphys.81.163[rev .
mod .
phys . * 81 * , 163 ( 2009 ) .
] r. s. judson and h. rabitz , http://dx.doi.org/10.1103/physrevlett.68.1500[phys .
rev .
lett . * 68 * , 1500 ( 1992 ) . ] h. rabitz _ et al . _ ,
http://dx.doi.org/10.1126/science.288.5467.824[science * 288 * , 824 ( 2000 ) . ] a. assion _ et al . _ , http://dx.doi.org/10.1126/science.282.5390.919[science * 282 * , 919 ( 1998 ) ; ] d. meshulach and y. silberberg , http://dx.doi.org/10.1038/24329[nature ( london ) * 396 * , 239 ( 1998 ) ; ] r. bartels _ et al . _ , http://dx.doi.org/10.1038/35018029[_ibid._ * 406 * , 164 ( 2000 ) ; ] t.
brixner _
et al . _ , http://dx.doi.org/10.1038/35102037[_ibid._ * 414 * , 57 ( 2001 ) ; ] j. l. herek _ et al
_ , http://dx.doi.org/10.1038/417533a[_ibid._ * 417 * , 533 ( 2002 ) ; ] r. j. levis , g. m. menkir , and h. rabitz , http://dx.doi.org/10.1126/science.1059133[science * 292 * , 709 ( 2001 ) ; ] v. i. prokhorenko
_ et al . _ , http://dx.doi.org/10.1126/science.1130747[_ibid._ * 313 * , 1257 ( 2006 ) ; ] m. p.
a. branderhorst _
et al . _ , http://dx.doi.org/10.1126/science.1154576[_ibid._ * 320 * , 638 ( 2008 ) ; ] t.
brixner _
et al . _ , http://dx.doi.org/10.1103/physrevlett.92.208301[phys
rev .
lett . * 92 * , 208301 ( 2004 ) ; ] a. lindinger _
et al . _ , http://dx.doi.org/10.1103/physrevlett.93.033001[_ibid._ * 93 * , 033001 ( 2004 ) ; ] t. laarmann _ et al . _ , http://dx.doi.org/10.1103/physrevlett.98.058302[_ibid._ * 98 * , 058302 ( 2007 ) ; ] m. roth _
et al . _ , http://dx.doi.org/10.1103/physrevlett.102.253001[_ibid._ * 102 * , 253001 ( 2009 ) ; ] p. nuernberger , g. vogt , t. brixner , and g. gerber , http://dx.doi.org/10.1039/b618760a[phys
chem .
chem .
phys .
* 9 * , 2470 ( 2007 ) . ]
a. pechen and h. rabitz , http://dx.doi.org/10.1103/physreva.73.062102[phys .
rev .
a * 73 * , 062102 ( 2006 ) ; ] r. romano and d. dalessandro , http://dx.doi.org/10.1103/physreva.73.022323[_ibid._ * 73 * , 022323 ( 2006 ) . ]
r. vilela mendes and v. i. manko , http://dx.doi.org/10.1103/physreva.67.053404[phys .
rev .
a * 67 * , 053404 ( 2003 ) ; ] a. pechen , n. ilin , f. shuang , and h. rabitz , http://dx.doi.org/10.1103/physreva.74.052102[_ibid._ * 74 * , 052102 ( 2006 ) ; ] f. shuang _ et al . _ , http://dx.doi.org/10.1103/physreva.78.063422[_ibid._ * 78 * , 063422 ( 2008 )
] see supplementary material at [ http://link.aps.org/supplemental/10.1103/physreva.82.030101[http://link.aps.org/supplemen tal/10.1103/physreva.82.030101 ] ] , for more information on quantum control approaches , a hybrid quantum control scheme , and kraus maps .
v. p. belavkin , autom .
remote control * 44 * , 178 ( 1983 ) .
h. m. wiseman , http://dx.doi.org/10.1103/physreva.49.2133[phys .
rev .
a * 49 * , 2133 ( 1994 ) . ]
h. rabitz , m. hsieh , and c. rosenthal , http://dx.doi.org/10.1126/science.1093649[science * 303 * , 1998 ( 2004 ) ; ] h. rabitz , t .- s
. ho , m. hsieh , r. kosut , and m. demiralp , http://dx.doi.org/10.1103/physreva.74.012721[phys . rev .
a * 74 * , 012721 ( 2006 ) ; ] h. rabitz , m. hsieh , and c. rosenthal , http://dx.doi.org/10.1063/1.2198837[j .
chem .
phys . * 124 * , 204107 ( 2006 ) ; ] m. hsieh , r. b. wu , and h. rabitz , http://dx.doi.org/10.1063/1.2981796[_ibid._ * 130 * , 104109 ( 2009 ) . ] m. hsieh and h. rabitz , http://dx.doi.org/10.1103/physreva.77.042306[phys .
rev .
a * 77 * , 042306 ( 2008 ) ; ] t .- s .
ho , j. dominy , and h. rabitz , http://dx.doi.org/10.1103/physreva.79.013422[_ibid._ * 79 * , 013422 ( 2009 ) . ]
t .- s . ho and h. rabitz , http://dx.doi.org/10.1016/j.jphotochem.2006.03.038[j .
photochem .
photobiol .
a * 180 * , 226 ( 2006 ) . ]
k. moore , m. hsieh , and h. rabitz , http://dx.doi.org/10.1063/1.2907740[j .
chem .
phys . * 128 * , 154117 ( 2008 ) ; ] a. oza _ et al . _ , http://dx.doi.org/10.1088/1751-8113/42/20/205305[j .
phys .
a * 42 * , 205305 ( 2009 ) . ]
j. g. digalakis and k. g. margaritis , http://dx.doi.org/10.1080/00207160108805080[int .
j. comp . math . * 77 * , 481 ( 2001 ) . ] j. roslund and h. rabitz , http://dx.doi.org/10.1103/physreva.79.053417[phys .
rev .
a * 79 * , 053417 ( 2009 ) ; ] j. roslund , o. m. shir , t. back , and h. rabitz , http://dx.doi.org/10.1103/physreva.80.043415[_ibid._ * 80 * , 043415 ( 2009 ) . ]
a. pechen , d. prokhorenko , r. b. wu , and h. rabitz , http://dx.doi.org/10.1088/1751-8113/41/4/045205[j .
phys .
a * 41 * , 045205 ( 2008 ) ; ] r. b. wu , a. pechen , h. rabitz , m. hsieh , and b. tsou , http://dx.doi.org/10.1063/1.2883738[j .
math .
phys .
* 49 * , 022108 ( 2008 ) . ]
k. kraus , _ states , effects and operations : fundamental notions of quantum theory _
( springer , berlin , 1983 ) .
a. c. doherty , s. habib , k. jacobs , h. mabuchi , and s. m. tan , http://dx.doi.org/10.1103/physreva.62.012105[phys .
rev .
a * 62 * , 012105 ( 2000 ) . ]
p. bushev _ et al . _ , http://dx.doi.org/10.1103/physrevlett.96.043003[phys
rev .
lett . * 96 * , 043003 ( 2006 ) ; ] g. g. gillett _ et al . _ , http://dx.doi.org/10.1103/physrevlett.104.080503[_ibid._ * 104 * , 080503 ( 2010 ) . ]
s. lloyd , http://dx.doi.org/10.1103/physreva.62.022108[phys .
rev .
a * 62 * , 022108 ( 2000 ) . ]
r. j. nelson , y. weinstein , d. cory , and s. lloyd , http://dx.doi.org/10.1103/physrevlett.85.3045[phys .
rev .
lett . * 85 * , 3045 ( 2000 ) ; ] h. mabuchi , http://dx.doi.org/10.1103/physreva.78.032323[phys .
rev . a * 78 * , 032323 ( 2008 ) . ] s. lloyd and l. viola , http://dx.doi.org/10.1103/physreva.65.010101[phys .
rev .
a * 65 * , 010101 ( 2001 ) . ]
d. a. lidar , z. bihary , and k. b. whaley , http://dx.doi.org/10.1016/s0301-0104(01)00330-5[chem . phys . * 268 * , 35 ( 2001 ) . ]
r. wu , a. pechen , c. brif , and h. rabitz , http://dx.doi.org/10.1088/1751-8113/40/21/015[j .
phys .
a * 40 * , 5681 ( 2007 ) . ]
b. bonnard and m. chyba , _ singular trajectories and their role in control theory _ ( springer , berlin , 2003 ) .
r. b. wu , j. dominy , t .- s . ho , and h. rabitz , e - print http://arxiv.org/abs/arxiv:0907.2354[arxiv:0907.2354 .
] * supplementary material for the manuscript * + alexander pechen , constantin brif , rebing wu , raj chakrabarti , and herschel rabitz + _ department of chemistry , princeton university , princeton , new jersey 08544 , usa _ |
the development of meta materials has led to renewed interest in the original work by veselago on negative refraction @xcite and the field of meta materials is in rapid growth . for more details
we refer to recent reviews @xcite as well as to recent special issues @xcite . in this work
we consider a photonic crystal composed of high - index dielectric rods arranged in a two - dimensional lattice , see fig .
[ fig1](a ) . for temporal harmonic modes the @xmath4 field
is governed by @xcite @xmath5 where @xmath6 is the relative dielectric permittivity , @xmath1 is the relative magnetic permeability , @xmath7 is the speed of light in vacuum , and @xmath8 is the angular frequency with @xmath9 being the free - space wave vector . for nature s
own dielectric materials we locally have , almost by definition , that @xmath10 so that everywhere in space @xmath11 @xcite .
however , for artificial composite dielectric structures in the meta material limit , where the free - space wavelength @xmath12 is much longer than the characteristic length scale ( periodicity ) @xmath13 of the underlying ( quasi ) periodic structure , one may for the slowly varying homogenized fields ( envelopes ) have an effective permeability @xmath14 which is constant in space and which at certain frequencies deviates remarkably from the microscopic permeability , i.e. @xmath15 and in particular it may become negative . the same apparent paradox is also possible for the dielectric response . as an example fig .
[ fig2 ] shows a plane - wave simulation @xcite of eq .
( [ eq : waveequation ] ) for the structure in fig .
[ fig1]a with @xmath16 and @xmath17 for the rods and @xmath18 for the background air region . for these inverse
photonic crystal structures , the band structure and the photonic band - gap formation can be understood in terms of mie resonances @xcite .
clearly , the band structure reveals a strong hybridization at the frequency corresponding to the first mie resonance of the rods as indicated by the dashed line .
it is this resonance behaviour which make the rods behave as artificial magnetic atoms with the possibility for a negative-@xmath1 response . in this work
we extract the effective permeability by applying the recently derived mesoscopic homogenization theory for p - polarized fields by felbacq and bouchitt @xcite .
we in detail study the effective permeability @xmath14 and the effect of the topology of the rods as well as the influence of spatial disorder / randomness and dispersion in the rod topology , compare panels ( b ) and ( c ) in fig .
[ fig1 ] . for rods with a more arbitrary cross section
@xmath19 we define a topological radius @xmath0 by @xmath20 where @xmath21 is the area of @xmath19 and @xmath22 is the perimeter of its surface @xmath23 , see fig . [
fig1](b ) .
it is easily verified that a circular cross section of radius @xmath24 gives @xmath25 .
more generally , the topological radius @xmath0 captures the qualitative resonance behaviour of the effective permeability as we shall see in the following .
our results for more arbitrary topologies rely on approximative results for the spectral problem of the helmholtz equation as employed recently by one of us in the context of diffusion dynamics , fluid flow , and ion transport in microfluidic problems @xcite .
we consider the problem of p - polarized fields corresponding to a magnetic field aligned parallel to the rods , see fig .
[ fig1](a ) . in the limit where @xmath26
the problem posed by maxwell s equations can be treated by a multi - scale expansion approach leading to an effective magnetic permeability given by @xcite @xmath27 where @xmath28 is the area of @xmath19 , @xmath29 is the air - filling fraction , and @xmath30 is governed by [ eq : mueff ] @xmath31 above , we have used the dirac _ bra - ket _ notation and inner products are defined as @xmath32 where @xmath33 and @xmath34 are functions on @xmath19 .
similarly , @xmath30 , or @xmath35 , in eqs .
( 3 ) and ( 4 ) corresponds to the magnetic field component along the rods and @xmath36 denotes the unity function . in the following we assume a quadratic unit cell with area @xmath37 , see fig . [ fig1](b ) , so that @xmath38 . as in refs .
@xcite we solve the problem in a complete basis defined by the helmholtz eigenvalue equation @xmath39 with the basis functions being orthonormal , i.e. @xmath40 .
obviously , the ansatz @xmath41 fulfills the boundary condition and the expansion coefficients are readily determined by standard _ bra - ket _ manipulations . in this way we arrive at the following general expression for the effective permeability @xcite @xmath42 the strength of this result is obviously the connection of @xmath14 to the spectrum and eigenstates of the helmholtz equation which have been studied in great detail in various contexts of physics , including membrane dynamics , the acoustics of drums , the single - particle eigenstates of 2d quantum dots , not to forget quantized conductance of quantum wires .
finally , the results of course in turn provides the direct link between @xmath14 and the mie resonances .
in the following we utilize approximate results for the spectral problem and its dependence on geometry .
we start by introducing the effective area @xmath43 so that eq . (
[ eq : muexpansion ] ) is rewritten as @xmath44 where we in the second line have truncated the rapidly converging sum making the approximation valid in the vicinity of the main resonance . in table
[ tab1 ] we have listed the first eigenvalue @xmath45 and the corresponding effective area @xmath46 for a number of different geometries . for the numerical results we have employed a finite - element method with an adaptive mesh algorithm ( comsol multiphysics ) ensuring fast convergence of both eigenvalues and eigenstates . inspecting the column for the eigenvalue @xmath47 in table [ tab1 ] we note that the numbers do not in general change much from geometry to geometry , despite the variation in topology .
the small spread originates in the normalization by the topological radius @xmath0 , see eq .
( [ eq : rtop ] ) . for the effective area
@xmath46 we likewise observe a modest variation .
it is worthwhile emphasizing the results for a circular cross section which become good approximations for also other relatively compact topologies . in the following we thus use the approximation that @xmath48 and @xmath49 where @xmath50 is the first zero of the bessel function , i.e. @xmath51 . with these approximations
we finally arrive at @xmath52 where all the dependence on geometry is captured by the topological radius @xmath0 .
we may now estimate the resonance wavelength .
suppose that @xmath3 , then for weakly absorbing materials , @xmath53 , we have the resonance around @xmath54 with a line width of the order @xmath55 as expected the line width reflects the damping due to @xmath56 in the high - index material .
furthermore , we may also estimate the magnitude of the effective magnetic response .
just below the resonance the real part is bounded by @xmath57 while for the imaginary part @xmath58 the above results illustrate how @xmath6 , @xmath29 , and @xmath0 influence the magnetic response .
obviously , a negative permeability requires that @xmath59 . on the other hand ,
some absorption is of course needed to have a finite line width instead of a delta - function - like response which would suffer tremendously from the inevitable effects of inhomogeneous broadening .
we shall return to this in sec .
[ sec : inhomogeneous ] .
the validity of the assumptions underlying eq .
( [ eq : mumultiscale ] ) has been tested in refs .
@xcite by comparing simulated transmission spectra of the real photonic crystal to spectra based on the corresponding homogenized meta material . in this work
we thus assume eq .
( [ eq : mumultiscale ] ) valid and focus on checking the quality of the approximate result in eq .
( [ eq : mueffapprox ] ) .
we do this by comparing the outcome of eq .
( [ eq : mueffapprox ] ) to numerical solutions of the original problem posed by eq .
( [ eq : mumultiscale ] ) .
we use a finite - element method with an adaptive mesh algorithm ( comsol multiphysics ) for this purpose . for the dielectric function
we use @xmath60 as in ref .
[ fig2 ] illustrates the corresponding photonic band structure .
clearly , there is a strong hybridization at the frequency corresponding to the isolated mie resonance of the rod , see eq .
( [ eq : lambda_resonance ] ) .
it is this avoided crossing which causes strong resonance in the effective permeability that is observed in fig .
[ fig3 ] . the approximate result in eq .
( [ eq : mueffapprox ] ) works excellently near the main resonance for rods with circular cross section as seen in fig .
[ fig3 ] while the approximation of course does nt capture high - order modes due to the truncation of the summation in eq .
( [ eq : muexpansion_truncation ] ) . in the following
we go a step further and test the result for highly non - circular geometries .
[ fig4 ] illustrates results for one such example .
as seen there is a good qualitative agreement and since the geometry is highly non - circular , see inset , the quantitative agreement is perhaps better than expected .
note how the change in topological radius @xmath0 scales the resonance wavelength @xmath61 and the line width @xmath62 compared to the circular case , fig .
[ fig3 ] , in accordance with eq .
( [ eq : lambda_resonance ] ) while the conservation of the air - filling fraction @xmath29 leads to the same magnitude of the response in accordance with eqs .
( [ eq : remu ] ) and ( [ eq : immu ] ) which depend only on @xmath29 , but not on @xmath0 .
the present results suggest that shape plays only a minor role and mainly affects the particular value of the resonance frequency ( which is proportional to the topological radius @xmath0 ) while the air - filling fraction @xmath29 and the dielectric function @xmath3 tend to govern the magnitude of the effective magnetic response .
this suggests that homogenized materials will not be strongly influenced by spatial disorder / randomness conserving the average air - filling fraction while shape dispersion ( dispersion in the topological radius , possibly also influencing the average air - filling fraction ) will give rise to a smearing of the resonance phenomena .
suppose that for a particular topological radius @xmath63 the resonance wavelength is @xmath64 , see eq .
( [ eq : lambda_resonance ] ) , and that the width of the resonance is @xmath65 . assuming a gaussian distribution of @xmath0 , @xmath66,\ ] ] and neglecting the dependence of the intrinsic line width @xmath65 on @xmath0 we get an inhomogeneous suppression of the resonance magnitude by a factor @xmath67 , @xmath68 and an inhomogeneous broadening of the resonance by a factor @xmath69 , @xmath70 thus , inhomogeneous broadening will be negligible if @xmath71 corresponding to @xmath72 fig . [ fig5 ] illustrates these results in the case of rods of circular cross section as in fig .
the results are obtained from a numerical evaluation of @xmath73 with @xmath14 given by eq .
( [ eq : mueffapprox ] ) . as also suggest by eq .
( [ eq : criterion ] ) a size dispersion with a relative variation comparable to @xmath74 preserves the negative-@xmath1 phenomena while larger variations will result in a positive @xmath14 .
we have studied the magnetic response of rod - type dielectric meta materials within the recently developed homogenization theory @xcite and we have in particular investigated the effects of topology and inhomogeneous broadening .
we predict that topology itself mainly affects the mie resonance frequency while the resonance shape mainly depends on the air - filling fraction and the ratio @xmath75 with @xmath3 being the relative dielectric function of the rods .
furthermore , we argue that randomness / disorder conserving the air - filling fraction will have a limited influence on the negative-@xmath1 phenomena while dispersion in the topological radius @xmath0 of the rods may lead to significant inhomogeneous broadening and suppression of the phenomena .
this work is financially supported by the _ danish council for strategic research _ through the _ strategic program for young researchers _ ( grant no : 2117 - 05 - 0037 ) .
e. lidorikis , m. m. sigalas , e. n. economou , and c. m. soukoulis , `` gap deformation and classical wave localization in disordered two - dimensional photonic - band - gap materials '' , phys .
b * 61 * 13458 13464 ( 2000 ) .
.central dimensionless parameters for different geometries .
@xmath76see e.g. @xcite for the eigenmodes and eigenspectrum .
@xmath77data obtained by finite - element simulations .
e.g. @xcite for the eigenmodes and eigenspectrum .
[ cols="<,^,^,^,^",options="header " , ] | we consider meta materials made from a two - dimensional dielectric rod - type photonic crystal .
the magnetic response is studied within the recently developed homogenization theory and we in particular study the effects of topology and inhomogeneous broadening . while topology itself mainly affects the mie resonance frequency we find that the dispersion in the topological radius @xmath0 of the dielectric rods may lead to significant inhomogeneous broadening and suppression of the negative-@xmath1 phenomena for @xmath2 , with @xmath3 being the relative dielectric function of the rods . |
the highly mismatched semiconductor alloy gabi@xmath11as@xmath1 , comprised of dilute concentrations of bismuth ( bi ) incorporated in gaas , is an attractive candidate for the design of highly efficient mid- and far - infrared optical @xcite and spintronic @xcite devices .
it has been shown , both experimentally @xcite and theoretically @xcite , that isovalent substitution of a small fraction of bi atoms in gaas strongly reduces the band gap energy ( e@xmath12 ) by @xmath1390 mev/% bi replacing as .
a giant bowing of the spin - orbit - splitting energy ( @xmath3 ) has also been observed in gabi@xmath11as@xmath1 , leading to the onset of an e@xmath4 regime .
this characteristic is of fundamental importance as it opens the possibility to suppress the dominant chsh auger recombination losses suffered by conventional iii - v telecomm lasers operating at high temperatures @xcite .
therefore , an understanding of the electronic structure of gabi@xmath0as@xmath1 is crucial both from a fundamental perspective , as well as for potential device applications .
co - alloying of bi and n in gaas further opens up the possibility of precise strain control in gabi@xmath0n@xmath6as@xmath7 thereby offering a large and flexible parameter space for band structure engineering . this work analyses the combined impact of bi and of n on the electronic structure of gabi@xmath0n@xmath6as@xmath7 , showing that n predominantly affects the conduction band and bi the valence band , with both effects largely being independent of each other . a detailed chart of the variations in the e@xmath2 and e@xmath12-@xmath3 energies as a function of bi and n composition is plotted as a guide for material and device design and analysis .
we have developed a nearest neighbor sp@xmath8s@xmath9 tight binding ( tb ) hamiltonian to investigate the electronic structure of the dilute bismide and the bismide - nitride alloys of gaas , including the effects of atomic relaxation determined using a valence force field model @xcite . by performing large , ordered gabi@xmath0y@xmath1 ( y = as , p ) supercell calculations using our models , we demonstrate that isovalent bi substitution introduces bi - related defect states which interact with the host gay ( y = as , p ) matrix valence band edge via a bi composition dependent band anti - crossing ( bac ) interaction .
the band anti - crossing parameters are then introduced into 12- and 14-band * * k**@xmath10*p * models for gabi@xmath0as@xmath1 and for gabi@xmath11n@xmath14as@xmath7 to investigate the alloy electronic structure .
the details of the derivation procedures are reported elsewhere @xcite .
figures 1 ( a ) , ( b ) , and ( c ) compare band dispersions computed from the tight binding ( tb ) and the * * k**@xmath10*p * models in the vicinity of the @xmath15-point for ga@xmath16as@xmath16 , ga@xmath16bi@xmath17as@xmath18 , and ga@xmath16bi@xmath17n@xmath17as@xmath19 alloy supercells .
the good agreement between the tb and * * k**@xmath10*p * models confirms the presence of the band anti - crossing interaction between the impurity states and the host matrix states .
furthermore , by carrying out 4096-atom ordered supercell calculations , we verify that the effects of bi and of n on the alloy electronic structure are largely independent of each other , with n - related defect states introducing a bac interaction in the conduction band and the bi - related impurity states introducing bac interactions in the valence band .
our calculations show that the observed strong variation in e@xmath12 for gabi@xmath0as@xmath1 can be well explained in terms of a large upward shift in the valence band edge energy due to the bac interaction and a significant contribution coming from a conventional alloy reduction in the conduction band edge energy @xcite .
the large band gap reduction with n composition in ganas is further enhanced by the presence of bi in gabi@xmath0n@xmath6as@xmath7 , leading to a giant e@xmath12 bowing @xcite . since n only weakly perturbs the gaas valence band structure , the bowing of @xmath3 in gabi@xmath0n@xmath6as@xmath7 is similar to that in the gabi@xmath0as@xmath1 alloy .
figure 1 ( d ) shows the variation of e@xmath12 ( red lines ) and of e@xmath12-@xmath3 ( blue lines ) for gabi@xmath0n@xmath6as@xmath7 grown pseudomorphically on gaas with @xmath20% and @xmath21% , for which the layers can be under compressive strain ( @xmath22 ) , strain - free ( @xmath23 ) , or under tensile strain ( @xmath24 ) as the composition is varied .
a large tuning of the band gap wavelength ( @xmath25 ) is demonstrated , to @xmath26 3.0 @xmath27 m .
the e@xmath12-@xmath3 = 0 line indicates the cut - off bi and n compositions beyond which an e@xmath12 energy lower than the @xmath3 energy can be achieved , which is key to the suppression of the chsh auger loss mechanism @xcite . in summary ,
our models provide a clear understanding of the electronic structure of gabi@xmath0as@xmath1 and of gabi@xmath0n@xmath6as@xmath7 .
we conclude that these alloys have significant potential for use in highly efficient optoelectronic devices operating with reduced temperature sensitivity both at telecommunication wavelengths and in the mid - ir , critical to the development of future green commercial telecommunication networks and for a wide range of sensing applications .
this work was supported by european union project biancho ( fp7 - 257974 ; mu and eor ) , by science foundation ireland ( 10/in.1/i2994 ; eor ) and by the irish research council under the embark initiative ( rs/2010/2766 ; cab ) . | the addition of dilute concentrations of bismuth ( bi ) into gaas to form gabi@xmath0as@xmath1 alloys results in a large reduction of the band gap energy ( e@xmath2 ) accompanied by a significant increase of the spin - orbit - splitting energy ( @xmath3 ) , leading to an e@xmath4 regime for @xmath5% which is technologically relevant for the design of highly efficient photonic devices .
the quaternary alloy gabi@xmath0n@xmath6as@xmath7 offers further flexibility for band gap tuning , because both nitrogen and bismuth can independently induce band gap reduction .
this work reports sp@xmath8s@xmath9 tight binding and 14-band * * k**@xmath10*p * models for the study of the electronic structure of gabi@xmath0as@xmath1 and gabi@xmath0n@xmath6as@xmath7 alloys .
our results are in good agreement with the available experimental data .
address = tyndall national institute , lee maltings , dyke parade , cork , ireland .
address = tyndall national institute , lee maltings , dyke parade , cork , ireland .
, altaddress = department of physics , university college cork , cork , ireland . address = tyndall national institute , lee maltings , dyke parade , cork , ireland .
, altaddress = department of physics , university college cork , cork , ireland . |
the universe appears to consist of approximately 30% dark matter , which clusters and drives the formation of large - scale structure in the universe , and 70% dark energy , which drives the late - time acceleration of the universe ( see ref .
@xcite for a recent review , and references therein ) .
numerous models for the dark energy have been proposed ; in one class of models , the dark energy is simply taken to be barotropic fluid , in which the pressure @xmath10 and energy density @xmath11 are related by @xmath12 one of the first cases to be investigated in detail was the equation of state @xmath13 where @xmath14 is a constant @xcite .
for an equation of state of this form , the density of the dark energy depends on the scale factor , @xmath15 , as @xmath16 a more complex equation of state arises in the chaplygin gas model @xcite , for which the equation of state has the form @xmath17 where @xmath18 is a constant , leading to a density evolution of the form @xmath19 in this model , the density interpolates between dark - matter - like evolution at early times and dark - energy - like ( i.e. , constant density ) evolution at late times .
thus , the chaplygin gas model can serve as a unified model of dark matter and dark energy . (
the dark matter sector of this model may have problems with structure formation @xcite , although see the discussion in refs .
@xcite for an alternate viewpoint ) .
the chaplygin gas model was generalized by bento , et al .
, @xcite to an equation of state of the form @xmath20 where @xmath18 and @xmath21 are constants .
this equation of state leads to the density evolution @xmath22^{1/(1+\alpha)}.\ ] ] again , the density evolution in the generalized chaplygin gas models changes from @xmath23 at early times to @xmath24 at late times .
note that such an equation of state can also be modelled as a dissipative matter fluid where the dissipative pressure is proportional to the energy density ; a number of exact solutions for this model have been discussed in refs .
@xcite .
in equation ( [ genc1 ] ) , one normally takes @xmath1 , while @xmath18 is chosen to be sufficiently small that @xmath2 for the generalized chaplygin gas . in this paper
, we relax these constraints and examine chaplygin gas models described by equation ( [ genc1 ] ) for which @xmath21 and @xmath14 can lie outside this range . in this case
, the chaplygin gas no longer serves as a unified model for dark matter and dark energy , but it can be taken to be a model for dark energy alone .
this sort of generalization for the special case of @xmath25 was previously undertaken by khalatnikov @xcite .
the models presented here can be derived as special cases of the more generic models examined in refs .
@xcite , and can also arise in the context of @xmath26-essence models @xcite .
these possible generalizations are mentioned in ref .
@xcite , although only the case @xmath27 , @xmath4 is explored in detail .
in addition to noting some interesting features of these models , our main new result is a set of observational constraints on these models .
the equation of state for the generalized chaplygin gas ( gcg ) is given by equation ( [ genc1 ] ) . in order to produce a negative pressure and give the currently observed acceleration of the universe , equation ( [ genc1 ] ) must have @xmath28 , so we will confine our attention to this case .
we assume further that @xmath29 , since for @xmath30 , the gcg is equivalent to the dark energy fluid described by equation ( [ constantw ] ) .
note that a particular choice of @xmath18 and @xmath21 does not uniquely determine the gcg model ; one also needs to specify , for example , the value of @xmath11 at a particular redshift .
integration of the energy conservation equation @xmath31 for @xmath32 results in @xmath33^{1/(1+\alpha)},\ ] ] where @xmath34 with @xmath35 being the present value of @xmath36 .
the choice of @xmath37 and @xmath21 _ does _ uniquely specify the gcg model . 3.2 in
since we are considering the gcg as a dark energy candidate , the expression for the hubble parameter near the present becomes @xmath38^{1/(1+\alpha)}\biggr],\end{aligned}\ ] ] where we have assumed a normal dust - like dark matter component with present density parameter @xmath39 , and @xmath40 is the present value of the density parameter for the gcg component .
we have neglected the radiation , which makes a negligible contribution to the total density around the present epoch .
we assume a flat universe : @xmath41 . the equation of state for the gcg fluid is @xmath42
taking @xmath43 in this equation , it is clear that @xmath44 where @xmath45 is the present - day value of @xmath14 for the chaplygin gas .
equation ( [ w0 ] ) gives the physical significance of @xmath37 .
the behavior of the gcg can vary significantly , depending on the values of @xmath37 and @xmath21 .
our aim in this paper is to explore all of these possible behaviours .
we consider first two trivial cases . for @xmath46 ,
the gcg behaves exactly as a cosmological constant at all times for all values of @xmath21 . for @xmath47 ,
the gcg behaves as a standard pressureless ( @xmath48 ) dust component for at all times for all values of @xmath21 .
3.2 in now consider the four non - trivial cases corresponding to the pair of choices @xmath49 , @xmath50 , and @xmath3 , @xmath1 .
case 1 ) @xmath51 ( _ standard gcg _ ) .
as previously discussed , in this region the gcg behaves as a pressureless dust component at early times , evolving asymptotically to a de - sitter regime ( @xmath52 ) at late times .
hence the gcg in this parameter region can act as a unified model for dark energy and dark matter ( udm ) .
this is the standard model of the generalized chaplygin gas , previously explored in great detail @xcite .
the remaining three cases correspond to new " versions of the generalized chaplygin gas .
case 2 ) @xmath53 , @xmath54 ( _ early phantom gcg _ ) . in this case , the gcg acts as a phantom component with @xmath4 at all times , but @xmath14 asymptotically approaches @xmath55 at late times .
hence , one has an early phantom behaviour in this region of parameter space . in this case
, @xmath56 at @xmath57^{1\over{3(1+\alpha)}}$ ] and then @xmath36 grows to become a constant at late times .
the behavior of this model at early times depends on the value of @xmath21 .
\(a ) for @xmath58 , we have @xmath59 when @xmath60 at @xmath61 .
hence , there is a pressure singularity at @xmath61 , with @xmath62 , while the scale factor @xmath15 and the expansion rate @xmath63 remain finite .
this is the so - called sudden singularity " previously discussed by barrow @xcite . in the classification scheme of nojiri , odintsov , and tsujikawa @xcite ,
this is a type ii singularity .
the gcg density @xmath11 , the equation of state parameter @xmath14 , and the deceleration parameter @xmath64 for this case are shown in figs .
1 - 3 , respectively , as solid curves .
\(b ) for @xmath65 , both @xmath36 and @xmath66 go to zero at @xmath61 .
although @xmath8 at @xmath61 , there is no singularity in this case .
afterwards the gcg behaves as a growing phantom component and then asymptotically approaches the de - sitter regime ( @xmath52 ) at late times .
the values of @xmath11 , @xmath14 , and @xmath64 are shown in this case in figs .
1 - 3 as dashed curves . 3.2 in this model has also been examined in some detail in ref .
their main point of emphasis is the interesting fact that this model allows for a phantom equation of state @xmath4 , while avoiding a future singularity .
we note further that for @xmath65 , this model is free of either past or future singularities , while still allowing for a phantom equation of state .
case 3 ) @xmath67 ( _ transient gcg _ ) . in this case , one has a de - sitter regime ( @xmath52 ) at early times , while @xmath36 asymptotically approaches pressureless dust ( @xmath9 ) at late times .
the gcg density @xmath11 , the gcg equation of state parameter @xmath14 , and the deceleration parameter @xmath64 for this case are shown in figs .
1 - 3 , respectively , as dot - dash curves .
if the gcg serves as dark energy in this case , the acceleration is a transient phenomenon .
models with transient acceleration are desirable from the point of view of string theory , as the existence of future horizons in an eternally - accelerating universe leads to a well - known problem in constructing the s - matrix in such models @xcite .
hence , a fair amount of effort has gone into constructing models in which the currently - observed acceleration is a transient phenomenon @xcite .
this case for the gcg represents another such model .
case 4 ) @xmath68 ( _ late phantom gcg _ ) in this case @xmath36 starts as a constant and then grows and eventually hits a singularity in the future .
the equation of state resembles that of a cosmological constant ( @xmath7 ) at early times , and then becomes phantom - like in the future .
the future singularity occurs at @xmath69^{1\over{3(\alpha+1)}}$ ] .
note that this singularity occurs at a finite value of the scale factor , which is different from the standard phantom scenario in which the scale factor blows up simultaneously with the energy density of the dark energy .
this singularity also differs from that in case 2 , as in this case , we have @xmath70 and @xmath71 as @xmath72 ; in the classification scheme of ref .
@xcite , this is a type iii singularity .
the gcg density @xmath11 , the equation of state parameter @xmath14 , and the deceleration parameter @xmath64 for this case are shown in figs .
1 - 3 , respectively , as dotted curves .
it is well known that , under very general assumptions , dark energy arising from a scalar field can not evolve from @xmath4 to @xmath2 or vice - versa @xcite ; this has been dubbed the phantom divide " @xcite .
the gcg model also displays a phantom divide , characterized by whether or not @xmath50 or @xmath49 . for all models with @xmath73
, we have @xmath2 at all times and for all values of @xmath21 , while @xmath50 gives phantom behavior ( @xmath4 ) at all times and for all @xmath21 . note also that all of the gcg models display asymptotic de sitter behavior ( @xmath52 ) . for @xmath3 ,
the de sitter behavior occurs asymptotically in the past , while @xmath1 gives de sitter behavior in the asymptotic future .
these results are independent of @xmath37 , which simply determines whether @xmath14 approaches @xmath55 from above or below .
many other barotropic models have been discussed in the literature , e.g. , the van der waals model @xcite and the wet fluid model @xcite . however , the barotropic model which most closely resembles the gcg models examined here is the model of ref .
@xcite ( see also @xcite ) , which has the equation of state @xmath74 while this model is qualitatively different from the gcg models , it displays similar behavior in certain limits , particularly with regard to singularities in the evolution . in particular , for @xmath75 and @xmath76 ,
the density @xmath11 is an increasing function of the scale factor , so that the second term in equation ( [ stef ] ) is dominant at late times .
thus , this model approaches the behavior of our case 4 model at late times , with a similar singularity @xcite .
in this section , we will examine constraints on these various versions of the gcg from the supernova ia data .
in deriving these limits , we assume that the gcg acts purely as dark energy , and we take the dark matter to be a separate component . the observations of supernovae measure essentially the apparent magnitude @xmath77 , which is related to the luminosity distance @xmath78 by @xmath79 where @xmath80 is the dimensionless luminosity distance and @xmath81 with @xmath82 being the comoving distance given by @xmath83 also , @xmath84 where @xmath85 is the absolute magnitude . for our analysis , we consider the data set compiled by riess _
the total data set contains the previously published 230 data points from tonry _ et al . _
@xcite , along with the 23 points from barris _ et al .
_ @xcite . but riess _ et al .
_ have discarded various points where the classification of the supernovae was not certain or the photometry was incomplete , increasing the reliablity of the sample .
ultimately the final set contains 143 points plus the 14 points discovered recently using the hubble space telescope ( hst ) , and this set of 157 points is named the gold " sample by riess _
et al . _
the data points in these samples are given in terms of the distance modulus @xmath86 where @xmath87 is measured in mpc .
the @xmath88 is calculated from @xmath89 ^ 2~ , \label{chisq2}\end{aligned}\ ] ] where present day hubble parameter , @xmath90 , is a nuisance parameter and @xmath91 are the model parameters .
marginalizing our likelihood functions over the nuisance parameter @xmath90 yields the confidence intervals in the @xmath91 parameter space . 3.5 in .
for our purposes , we have three model parameters , @xmath39 , @xmath21 and @xmath92 . for our best fit analysis
, we vary the parameters in the following ranges : @xmath39 from @xmath93 to @xmath94 , @xmath21 from @xmath95 to @xmath96 and @xmath92 from @xmath97 to @xmath98 .
the best fit values for the parameters in this case are : @xmath99 , @xmath100 and @xmath101 together with @xmath102 . in figure 4
we show the confidence contours in the @xmath103 , and @xmath104 parameter space by marginalizing over @xmath21 and @xmath39 respectively . in both cases ,
@xmath105 ( @xmath106cdm ) is rejected at the @xmath107 confidence level , while the data favor @xmath108 ( @xmath109 ) .
moreover , from the allowed @xmath104 space , one can see that a large portion of the allowed region corresponds to case 4 ( late phantom gcg ) .
this is consistent with previous analyses that suggest that the supernova data slightly favor @xmath4 at present @xcite .
there is also a small region at the @xmath107 confidence level corresponding to case 2(b ) ( early phantom gcg without an initial singularity ) . to see how the model parameters depend on @xmath39 , we plot in fig .
5 the @xmath110 confidence contours in the @xmath111 space for different values of @xmath39 .
this figure shows that the allowed parameters for the gcg depend very sensitively on the assumed value for @xmath39 ; for this set of choices for @xmath39 , any of the four cases is possible . for @xmath112 ,
the allowed region falls under the case 1 , where gcg behaves as a dust - like dark energy component , asymptotically approaching a cosmological constant . for @xmath113 , the data allow case 1 as well as case 3 where the acceleration of the universe is only a transient phenomena as it asymptotically approaches dust - like behavior .
for @xmath114 the data allow case 2 ( both a and b ) where the gcg behaves as phantom dark energy at early times but approaches a cosmological constant at late times , as well as case 4 where the gcg evolves from cosmological - constant - type behaviour to phantom behaviour . for @xmath115 only case 4 is allowed .
our results show that @xmath116 is necessary in order to have a phantom - like equation of state .
this conclusion is consistent with the recent results of jassal et al .
@xcite , although they used a different parametrization for the dark energy equation of state .
3.0 in these results show that the value of @xmath39 is crucial in determining which types of gcg dark energy are consistent with the supernova data .
the constraint on @xmath39 in @xmath106cdm models from wmap is @xmath117 @xcite , which is also consistent with recent results from sdss observations , while the more recent 2dfgrs analysis gives @xmath118 @xcite .
of course , when @xmath14 is allowed to vary from @xmath119 , these limits become functions of @xmath14 , but it is clear that the current observational constraints on @xmath39 are insufficient to rule out any of our four possible gcg models using supernova data alone .
in this section , we study the growth of density perturbations for the mixture of a matter fluid and a gcg dark energy fluid in the linear regime on subhorizon scales . in performing this calculation ,
it is necessary to assume a particular clustering behavior for the dark energy .
however , this behavior will depend on the physical model that gives rise to the chaplygin gas equation of state .
for example , if the chaplygin gas is taken to be a perfect fluid satisfying equation ( [ genc1 ] ) , then the gcg component will cluster gravitationally with a sound speed given by @xmath120 @xcite . on the other hand
, it is also possible to generate minimally - coupled scalar field models with the equation of state given by equation ( [ genc1 ] ) @xcite ( see also ref .
such models always have @xmath121 on subhorizon scales and therefore do not cluster on small scales .
the evolution of density perturbations for a gcg dark energy component that can cluster gravitationally was examined in ref .
@xcite , while the case of a smooth , unclustered gcg dark energy component was examined in ref .
we take the latter approach here . in this case , the only effect of the gcg evolution is to alter the growth of dark matter perturbations through the effect of the gcg energy density on the expansion of the universe . however , it is important to remember that the results will be very different for the case in which the gcg component is allowed to cluster .
3.0 in assuming the gcg to be a smooth component , the growth equation for the linear matter density perturbation , @xmath122 , is given by @xmath123 where `` prime '' denotes the derivative with respect to @xmath124 , `` dot '' denotes the derivative with respect to @xmath125 , and @xmath126 is the hubble parameter for the background expansion given in equation ( [ h ] ) . in equation ( [ delta ] ) , @xmath122 is the linear matter density contrast , @xmath127 , and @xmath128 is given by @xmath129^{1/(1+\alpha)}}}.\ ] ] one can easily check that for @xmath130 , the equation reduces to that for the @xmath106cdm model .
we have integrated equation ( [ delta ] ) numerically from @xmath131 to @xmath132 ( taken to be the present ) .
the initial conditions are choosen such that at @xmath131 , the standard solution @xmath133 for einstein - desitter universe is reached .
also we have assumed the matter density parameter @xmath134 throughout .
we have studied the solutions for the standard gcg ( case 1 ) , transient gcg ( case 3 ) and late phantom gcg ( case 4 ) models .
6 , 7 , and 8 show the behaviour of @xmath122 as a function of the scale factor for these three cases .
the standard gcg case has been studied previously by multamaki et al .
@xcite , who showed that for parameters slightly deviating from the @xmath106cdm universe ( @xmath46 ) , @xmath122 deviates grossly from the standard @xmath106cdm result , making this model hard to reconcile with the present observational data .
the behaviour of @xmath122 in fig .
6 agrees with this result . 3.0 in on the other hand
, one can see from figs .
7 and 8 ( the transient and late phantom cases respectively ) , with values of @xmath21 and @xmath37 deviating significantly from the @xmath106cdm case , that the behaviour of @xmath122 is practically indistinguishable from the @xmath106cdm case .
this is an interesting result ; it shows that models with gcg dark energy can have quite different equations of state from @xmath106cdm at either early or late times , yet still give similar results for the growth of linear density perturbations .
the reason for these results is clear from the behavior of @xmath11 for these models ( fig .
1 ) . both the transient and late phantom gcg models , like the cosmological constant , contribute negligibly to the density of the universe at early times ; this density is dominated by the matter component . at low redshift ,
the gcg in both models begins to dominate the expansion ( just as the cosmological constant does in @xmath106cdm ) , with the gcg density decreasing in time ( for the transient model ) or increasing in time ( for the late phantom model ) .
however , these deviations from the behavior of the cosmological constant occur over a very short range in redshift , and by forcing @xmath135 for the dark energy to be the same in all three cases , the results for the evolution of density perturbations are almost exactly the same .
3.0 in for the standard gcg , in contrast , the dust - like behavior results in a significant contribution to the density of the universe as long as dark matter is the dominant component ; if the gcg is assumed not to cluster , the result is a significant decrease in the perturbation growth .
this is illustrated more clearly in fig .
9 , where we show @xmath122 as function of the scale factor for both the transient and standard gcg models with the same @xmath92 ( @xmath136 ) , along with the @xmath106cdm model .
since the observations related to density perturbations are consistent with the @xmath106cdm model , our results suggest that , unlike the standard gcg model , both the transient and late phantom gcg models are consistent with the linear growth of density perturbations inferred from observations .
we have not shown the results for the early phantom model , but for the case where there is no early singularity , the results are also nearly identical to the @xmath106cdm model .
again , we emphasize that these results assume a non - clustering gcg . for the case
where the gcg clusters as a perfect fluid , the growth of density perturbations would be quite different .
our exploration of the the full @xmath137 parameter space for the generalized chaplygin gas yields three additional models beyond the standard gcg .
although these can not serve as unified models for dark matter and dark energy , they can have interesting consequences when treated as models for dark energy alone .
the early phantom model ( case 2 ) has the interesting property of serving as phantom dark energy ( @xmath4 ) without a late big rip singularity , and for some choices of parameters it is also free of an initial singularity . the transient cgc model ( case 3 )
provides a mechanism for accelerated behavior at the present , but it asymptotically approaches dust - like behavior at late times ( i.e. , its time evolution is exactly opposite to the standard gcg model ) .
thus , it provides a mechanism to allow for present - day acceleration without a future horizon .
the late phantom gcg model gives @xmath4 with @xmath14 decreasing with time , and it results in a future singularity at a finite value of the scale factor .
all of these models can be made consistent with the type ia supernovae observations , for an appropriate choice of @xmath39 ; the question of which models are allowed is extremely sensitive to the value of @xmath39 . if the gcg is assumed not to cluster , then all of these models ( with the exception of the subset of early phantom models with an initial singularity ) are also consistent with the growth of linear density perturbations .
of course , if the gcg is assumed to cluster , then these results will be significantly altered .
note that all three of our new " gcg models have a de sitter phase , and all three models can be tuned arbitrarily close to a cosmological constant at the present , either by pushing the phantom - like behavior arbitrarily far into the past ( early phantom model ) , or by pushing the dust - like or phantom - like behavior arbitrarily far into the future ( transient gcg and late phantom model , respectively ) .
these limits may seem uninteresting , as they reduce to the @xmath106cdm model over all observable ranges , but the one exception is the transient gcg model .
a dust - like phase for this model , even in the far future , eliminates the problem of future horizons .
thus , the transient gcg model can be made arbitrarily similar to the @xmath106cdm model but at the same time can resolve the possible conflict between the accelerating universe and string theory .
fabris , s.b.v .
gonalves and p.e .
de souza , astro - ph/0207430 ; a. dev , j.s . alcaniz and d. jain , , 023515 ( 2003 ) ; v. gorini , a. kamenshchik and u. moschella , , 063509 ( 2003 ) ; m. makler , s.q .
de oliveira and i. waga , , 1 ( 2003 ) ; j.s .
alcaniz , d. jain and a. dev , , 043514 ( 2003 ) ; m.c .
bento , o. bertolami , n.m.c santos and a.a .
sen , , 063501 ( 2005 ) . | the generalized chaplygin gas is characterized by the equation of state @xmath0 , with @xmath1 and @xmath2 .
we generalize this model to allow for the cases where @xmath3 or @xmath4 .
this generalization leads to three new versions of the generalized chaplygin gas : an early phantom model in which @xmath5 at early times and asymptotically approaches @xmath6 at late times , a late phantom model with @xmath7 at early times and @xmath8 at late times , and a transient model with @xmath7 at early times and @xmath9 at late times .
we consider these three cases as models for dark energy alone and examine constraints from type ia supernovae and from the subhorizon growth of density perturbations .
the transient chaplygin gas model provides a possible mechanism to allow for a currently accelerating universe without a future horizon , while some of the early phantom models produce @xmath4 without either past or future singularities . |
the number of known white dwarfs ( wds ) has increased dramatically since the beginning of the sloan digital sky survey ( sdss ; * ? ? ? * ) .
the most recent sdss wd catalog , collected from its data release 4 ( dr4 ; * ? ? ? * ) , yielded @xmath7 spectroscopically confirmed wds , of which @xmath8 are new discoveries @xcite .
model fits to the sdss spectra find that the majority of these objects are hydrogen ( da ) wds for which the mass distribution peaks at the expected value of @xmath9 m@xmath2 years ; see @xcite . ] .
but among the new sdss wds there are also a handful of candidate very low - mass wds ( lmwds ) with m @xmath1 m@xmath2 .
such objects are noteworthy not only because of their rarity but because the galaxy is not thought to be old enough to produce lmwds through single star evolution .
the youngest wds in the oldest milky way globular clusters have masses of @xmath10 m@xmath2 @xcite .
lower mass wds must therefore undergo significant mass loss , presumably because they form in close binaries whose evolution includes a phase of mass transfer . during this phase ,
much of the wd progenitor s envelope is removed , stunting the wd s evolution by preventing a helium flash in its core and resulting in the observed low - mass , helium - core wd m@xmath2 is @xmath11 and rises to @xmath12 for m @xmath13 m@xmath2 .
thus , even in this scenario , the likelihood of finding companions to the sdss lmwds remains high . ] . as @xcite put it ,
`` low - mass wds need friends . '' in examining seven wds with m @xmath14 m@xmath2 , they found that four are in double - degenerate close binary systems , with orbital periods on the order of a few hours to a few days ; for a fifth they identified a companion but were unable to confirm its nature or obtain an orbital period .
subsequent studies of large samples of das found that roughly @xmath15 have masses @xmath16 m@xmath17 , and that in the majority of these binaries the companion is likely to be a degenerate star @xcite .
the @xcite systems are wd / wd binaries , but the lmwd companions can also be neutron stars ( nss ; cf . discussion in * ? ? ?
* ) . in these systems ,
binary evolution produces a lmwd in a near - circular orbit around the ns `` reborn '' as a millisecond pulsar ( msp ; for discussion of different evolutionary scenarios involving nss , see * ? ? ?
observationally , roughly @xmath18 of field radio pulsars reside in binary systems .
most of these pulsars are msps , and in most cases their companions are thought to be lmwds with @xmath19 m @xmath16 m@xmath2 . of the @xmath20 msp companions identified outside of globular clusters , @xcite list @xmath21 with masses @xmath22 m@xmath2 , assuming the systems have a median inclination of @xmath23
. however , the msp companions are frequently too faint for optical spectroscopy to confirm that they are lmwds . despite the observational challenges ,
there are several well known msp / lmwd systems , and in fact the majority of known he - core wds are msp companions @xcite ; see @xcite for a review of these systems .
that said , the birthrate of msps in the galactic disk is thought to be @xmath24yr@xmath25 @xcite . estimating the number of lmwds in the galaxy is difficult ( cf
. discussion in * ? ? ?
* ) , and the sdss sample is best considered as providing a lower limit ( because wds are generally not explicitly targeted for sdss spectroscopy , it is unclear how best to correct for e.g. , incompleteness ) .
however , @xcite used the palomar green survey to estimate the formation rate of lmwds to be @xmath26 pc@xmath27 yr@xmath25 , implying that these make up about @xmath15 of all da white dwarfs larger than indicated by the relative space densities of low - mass and ordinary ( @xmath9 m@xmath17 ) wds as it accounts for the presence of degenerate companions to the lmwds . ] .
naively , this suggests that the lmwd birthrate in the volume sampled by the sdss dr4 ( @xmath28 kpc@xmath29 ) is @xmath30 yr@xmath25 , or nearly two orders of magnitude higher than that of msps .
this discrepancy in the ( admittedly very rough ) birthrates suggests that there is a low likelihood of detecting msp companions to sdss lmwds . still , because of the tantalizing theoretical and observational connections between lmwds and msps , we searched for radio pulsations from putative pulsar companions to @xmath0 spectroscopically confirmed sdss lmwds using the nrao green bank telescope ( gbt ) .
we describe the sample of sdss lmwds in [ wds ] and our @xmath3 mhz observations in [ obs ] .
we discuss the significance of our non - detections in [ discussion ] and
conclude in [ concl ] .
our sample of candidate lmwds was selected from the two sdss wd catalogs ( for a full description of the sdss wd selection , wd models , and the autofit program used to obtain estimates for wd temperatures and masses , see * ? ? ?
* ; * ? ? ?
two , sdss [email protected] and [email protected] , are described in detail in @xcite .
an additional @xmath33 candidate lmwds were later proposed by @xcite .
these wds are selected to have temperatures t@xmath34 k , surface gravities ( log @xmath35 ) more than @xmath36 below @xmath37 , and @xmath38 mag .
none of these wds show the photometric excess or spectroscopic signature associated with a companion in their discovery data ( see , for example , figures 21 and 22 in * ? ? ?
* ) , as expected if the companion is a compact object .
re - analysis of the sdss spectra by @xcite confirmed that the @xcite candidates are lmwds in all but one case : they find that sdss [email protected] is an a2 star .
this leaves us with a sample of @xmath39 lmwds from the sdss catalogs . to these
we add three lmwd candidates identified by @xcite that are present in the @xcite wd catalog but are not classified by the latter authors as lmwds .
sdss [email protected] and [email protected] are not included in the @xcite lmwd list because they have t@xmath40 and @xmath41 k , respectively .
the model fits for sdss [email protected] return t@xmath42 k and a log @xmath35 that is only @xmath43 below @xmath37 .
model fits to very cool and very hot wds are somewhat unreliable at @xmath44 k @xcite . ] , but we wish to be as inclusive as possible in assembling our sample . adding these three gives us just @xmath0 lmwd candidates spectroscopically identified in a survey volume of @xmath28 kpc@xmath29 .
lcccccccc [email protected] & @xmath45 & @xmath46 & @xmath47 & @xmath48 & @xmath49 & @xmath50 & @xmath51 & @xmath52 + [email protected] & @xmath53 & @xmath54 & @xmath55 & @xmath56 & @xmath57 & @xmath58 & @xmath59 & @xmath60 + [email protected] & @xmath61 & @xmath62 & @xmath63 & @xmath56 & @xmath64 & @xmath65 & @xmath59 & @xmath66 + [email protected] & @xmath67 & @xmath68 & @xmath69 & @xmath70 & @xmath71 & @xmath72 & @xmath73 & @xmath74 + [email protected] & @xmath75 & @xmath76 & @xmath69 & @xmath70 & @xmath71 & @xmath77 & @xmath59 & @xmath78 & @xmath74 + [email protected] & @xmath79 & @xmath80 & @xmath69 & @xmath70 & @xmath49 & @xmath81 & @xmath59 & @xmath82 + [email protected] & @xmath83 & @xmath84 & @xmath85 & @xmath86 & @xmath87 & @xmath88 & @xmath59 & @xmath66 + [email protected] & @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath93 & @xmath94 & @xmath59 & @xmath66 + [email protected] & @xmath95 & @xmath96 & @xmath91 & @xmath92 & @xmath97 & @xmath98 & @xmath59 & @xmath99 + [email protected] & @xmath100 & @xmath101 & @xmath91 & @xmath92 & @xmath49 & @xmath102 & @xmath59 & @xmath66 + [email protected] & @xmath103 & @xmath104 & @xmath91 & @xmath92 & @xmath105 & @xmath106 & @xmath107 & @xmath74 + [email protected] & @xmath108 & @xmath109 & @xmath91 & @xmath92 & @xmath110 & @xmath111 & @xmath107 & @xmath66 + [email protected] & @xmath112 & @xmath113 & @xmath114 & @xmath115 & @xmath93 & @xmath116 & @xmath107 & @xmath117 + [email protected] & @xmath118 & @xmath119 & @xmath120 & @xmath121 & @xmath71 & @xmath122 & @xmath59 & @xmath123 + [email protected] & @xmath124 & @xmath125 & @xmath126 & @xmath127 & @xmath64 & @xmath128 & @xmath51 & @xmath74 + [email protected] & @xmath129 & @xmath130 & @xmath131 & @xmath48 & @xmath132 & @xmath133 & @xmath73 & @xmath134 figure [ properties ] gives log @xmath35 vs. t@xmath135 for all of the candidate lmwds as reported by @xcite .
we use the models in figure [ properties ] to estimate m@xmath136 for each of our lmwds ; their properties are presented in table [ props ] . while the masses of these wds are comparable to those of lmwds detected as msp companions , their temperatures are very different .
for example , the companion to msp j1012@xmath315307 , whose mass of @xmath137 m@xmath2 makes it the lowest - mass confirmed wd in a msp / wd system , has t@xmath138 k @xcite , which falls in the range for previously cataloged pulsar wd companions , @xmath139 k
@xmath140 t@xmath141 k , regardless of wd type and mass @xcite . indeed , none of the known he - core wd companions whose temperature has been measured is hotter than t@xmath142 k. our lmwds all have temperatures above @xmath143 k , and most have t@xmath144 k , indicating that we are looking at a younger population of objects than that found to date as msp companions k are unlikely to be targeted for / discovered via sdss spectroscopy @xcite . ] .
a comparison to the evolutionary models produced by @xcite and @xcite suggests that all but the lowest mass wd in our sample are younger than roughly @xmath145 gyr , and that all are younger than @xmath146 gyr ( see figure [ properties ] ) .
these ages are calculated from wd cooling models , and correspond to the time since the end of mass transfer onto the wd companions . by contrast , typical msps have ages of @xmath146 to @xmath147 gyr ; these are estimated from the pulsar characteristic ( spin - down ) age and , in a few cases , also from the companion wd s cooling age ( e.g. , j1012@xmath315307 ; * ? ? ?
the lmwds were observed with the gbt over the course of five nights between 2006 january 12 and march 9 .
each session began with a 1 min observation of a test pulsar that was immediately analyzed to confirm a correct observing setup ; the integration times for each target are listed in table [ props ] . at the central observing frequency of 820mhz , the berkeley - caltech pulsar machine analog / digital filter bank ( the design is described in * ? ? ?
* ) provided 48mhz of bandwidth split into 96 spectral channels , for each of which we recorded total power samples every @xmath148s .
the resulting data were analyzed using standard pulsar search techniques implemented in the presto package @xcite installed in our computer cluster .
this analysis was similar to that described in @xcite , with appropriate ranges of trial dispersion measure ( dm ) to account for dispersive interstellar propagation , and with substantial acceleration correction to partially account for the changing pulsar period caused by putative orbital motion .
we calculated the maximum dm expected in the direction of each target using the @xcite model for the distribution of free electrons in the galaxy .
a typical value for our high - galactic - latitude targets was @xmath149cm@xmath27pc . in order to account for possible uncertainties in the model , we de - dispersed the data up to a dm limit twice that obtained from the model .
these values are given for each target in table [ props ] .
the orbital motions of the putative pulsars in these systems could significantly affect the apparent pulsar spin period , particularly for the companions to the lowest - mass wds in our sample ( known msp / wd binaries have orbital periods that range from a few hours to many months ) . in table
[ props ] , we include rough estimates for binary periods based on comparisons to the models of @xcite . @xcite calculated orbital properties for @xmath145@xmath151 m@xmath2 companions to @xmath105 m@xmath2 accreting nss , and their models are consistent with the orbital periods found for systems with less massive companions @xcite . while for @xmath152 of known pulsars the maximum orbital acceleration is @xmath153 m s@xmath154 @xcite , for @xmath155 of our targets the estimated accelerations are @xmath156 m s@xmath154 .
this significantly increases the parameter space that needs to be searched for pulsed signals and hence the computing time on the cluster , and also impacts our sensitivity limits ( see discussion in [ acc_sens ] ) .
no convincing pulsar signal was detected in our data . below we discuss the limitations of our search .
models for wds in the mass range of interest here predict absolute magnitudes of m@xmath157 @xcite .
we use sdss data release 6 photometry ( dr6 ; * ? ? ?
* ) to obtain rough distance estimates to our lmwds ( see table [ props ] ) . luminosity - wise , our searches were extremely constraining . in order to calculate the minimum detectable period - averaged flux density , we use the standard modifications to the radiometer equation .
we consider a pulsar duty cycle of 20% ( fairly typical of msps ) and compute sensitivity , conservatively , for @xmath158cm@xmath27pc . at 820mhz ,
the gbt gain is 2kjy@xmath25 and the system temperature is 25k .
the sky temperature varies with location , but at the relevant high latitudes at this frequency it only adds a few kelvin to the overall temperature .
we consider an effective threshold signal - to - noise ratio of 10 .
for an integration time of 2400s ( see table [ props ] ) , the sensitivity limit for long period pulsars is 0.16mjy .
pulsar luminosities are often measured at 1400mhz ; using a typical spectral index of @xmath159 , the limiting sensitivity at that frequency is about @xmath160mjy . for an msp period of 3ms ,
our sensitivity at 1400mhz was roughly 50% worse , or 0.1mjy ( and it quickly degrades for shorter periods , being three times worse for 1ms ) .
( twelve of our targets fall within the first footprint ; none are detected in these data , for which the sensitivity limit is roughly @xmath145 mjy @xcite . )
about 95% of msps known in the galactic disk have luminosities @xmath161mjykpc@xmath162 , and some @xmath152 have @xmath163mjykpc@xmath162 .
since our targets have typical estimated distances of @xmath164 kpc , our @xmath165 limits for 3ms periods ranged over @xmath166 mjykpc@xmath162 . as far as luminosity alone is concerned , therefore , we should have been able to detect approximately @xmath152 of known msps had they been present in our targets and beaming radio waves toward the earth .
as mentioned earlier , if the lowest - mass wds in our sample are in binaries with nss , these should be compact systems with orbital periods on the order of a few hours to a few days @xcite . as a result ,
pulse smearing due to orbital motions could occur , severely complicating our search . to gauge the extent of this effect
, we have calculated orbital accelerations for a number of systems with lmwds in orbit around a @xmath97 m@xmath2 ns .
when conducting an accelerated pulsar search with presto , the free parameter related to the maximum acceleration searched actually specifies the maximum drift of fourier components that is corrected . relating this to a physical acceleration depends on the integration time and on the putative pulsar period .
we consider a period of 3ms . for our integration times and computing resources , it was impracticable to fully correct for putative acceleration such as
might be produced in systems with orbital periods of about 8 hours or less .
for example , for j0849@xmath310445 , a m @xmath47 m@xmath2 wd that we observed for @xmath52 s , the search sensitivity to a @xmath167 ms pulsar was limited to a maximum acceleration of @xmath168 m s@xmath154 . because this search alone required several days of computing time on our @xmath169-node cluster , it simply was not feasible to extend it to the maximum accelerations predicted for a wd of its mass ( e.g. , @xmath170 m s@xmath154 if @xmath171 h and @xmath172 m s@xmath154 if @xmath173 h ) .
furthermore , for the six wds with m @xmath174 m@xmath2 , our integration times may represent a substantial fraction of the binary orbital period ( e.g. , as much as @xmath11 in the case of sdss j1234 ; see table [ props ] ) . in these cases ,
the assumption of a constant apparent acceleration used in our search breaks down , further limiting our ability to detect a pulsed signal from these targets . nevertheless ,
we were still very sensitive to most kinds of _ known _ msp / wd systems : essentially none are known in the disk of the galaxy with @xmath175h . in summary , while we were not very sensitive to new kinds of exotic msps ( e.g. , with @xmath176ms , or with @xmath177h ) , we should have detected some ordinary msps if they were orbiting a substantial fraction of our targets .
@xcite conducted a similar search for radio pulsations at @xmath4 mhz from unseen companions to eight sdss lmwds using the gbt .
none of these observations contained signals warranting follow - up .
the eight wds targeted by @xcite , described in @xcite , were identified as candidate lmwds based on initial model fits to sdss spectra done by @xcite . however , of the eight candidates , three are no longer considered lmwds based on more recent analysis of their spectra by @xcite461156.4 , j102228.02@xmath31020035 , and [email protected] . ] .
we therefore updated the analysis performed by @xcite in order to place more stringent constraints on the likelihood @xmath5 that a given lmwd has an msp companion .
@xcite assumed that a non - detection was the most likely outcome of their observations and estimated @xmath5 by using @xmath178 where @xmath179 is the msp beaming fraction , @xmath180 is the luminosity completeness of the search , @xmath181 the sensitivity to the systems accelerations , @xmath182 the search - algorithm success rate , and @xmath183 is the number of observed lmwds .
@xcite find that @xmath184 ; this limit rises to @xmath185 if one removes the three objects no longer classified as lmwds .
we use several of the same assumptions as @xcite in estimating @xmath5 .
we set @xmath186 and chose @xmath187 , a conservative estimate of our msp - detection algorithm s success rate .
our luminosity sensitivity is @xmath188 and our acceleration sensitivity is @xmath189 ; these two values reflect our confidence that our observations were sensitive enough to detect most msps with luminosities , orbital characteristics , and rotation periods comparable to those of the known population .
we then find that @xmath6 .
we conducted a search for pulsar companions to @xmath0 low - mass white dwarfs at @xmath3 mhz with the gbt .
no convincing pulsar signal was detected in our data .
this is consistent with the findings of @xcite , who conducted a search for radio pulsations at @xmath4 mhz from unseen companions to eight sdss lmwds ( three of which are no longer considered lmwds ) .
however , our non - detections place stronger constraints on the probability that the companion to a given lmwd is a radio millisecond pulsar .
our observations allow us to lower this likelihood from @xmath190 to @xmath191 .
any search for pulsars is inherently biased , and even a completely unbiased search does not preclude the existence of msps in these systems .
msp beaming fractions are thought to be about @xmath192 , and therefore in some sensitive searches msps have not been detected even though there is strong evidence for their presence ( e.g. , the companion to the young pulsar j1906@xmath310746 ; * ? ? ?
given that ns blackbody emission is gravitationally bent , allowing us to view @xmath193 of the ns surfaces in x rays @xcite , x - ray observations could yet detect msps that might exist in these systems .
none of these lmwds have been observed since the _ rosat _ all - sky survey @xcite , and none were detected in that survey .
if lmwd companions are not nss , the likelihood is that their currently unseen companions are also ( very faint ) wds .
two dozen wd / wd binaries are known ; in @xmath147 systems both wd masses have been measured ( cf . * ? ? ? * and references therein ) .
wd0957@xmath32666 has m@xmath194 and m@xmath195 m@xmath2 ; wd1101@xmath31364 has m@xmath196 and m@xmath197 m@xmath2 . the other wds in these systems
have m @xmath198 m@xmath2 . in the @xmath199 other binaries discussed by @xcite ,
only one of the wd masses is measured , with the companion wd mass given as a lower limit .
four of these systems harbor a wd with @xmath200 m @xmath201 m@xmath2 .
if the sdss lmwds are confirmed as wd / wd systems , they will represent a significant addition to this population and in a mass range that is still poorly sampled .
we thank scott ransom and the observing specialists at the gbt for their assistance and kurtis williams and craig heinke for helpful discussions .
m. ageros is supported by an nsf astronomy and astrophysics postdoctoral fellowship under award ast-0602099 .
the robert c. byrd green bank telescope is operated by the national radio astronomy observatory , which is a facility of the u.s .
national science foundation operated under cooperative agreement by associated universities , inc .
we gratefully acknowledge gbt grant gssp06 - 0008 for support of portions of our program .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington . | we have conducted a search for pulsar companions to @xmath0 low - mass white dwarfs ( lmwds ; m @xmath1 m@xmath2 ) at @xmath3 mhz with the nrao green bank telescope ( gbt ) .
these lmwds were spectroscopically identified in the sloan digital sky survey ( sdss ) , and do not show the photometric excess or spectroscopic signature associated with a companion in their discovery data .
however , lmwds are believed to evolve in binary systems and to have either a more massive wd or a neutron star as a companion .
indeed , evolutionary models of low - mass x - ray binaries , the precursors of millisecond pulsars ( msps ) , produce significant numbers of lmwds ( e.g. , * ? ? ?
* ) , suggesting that the sdss lmwds may have neutron star companions .
no convincing pulsar signal is detected in our data .
this is consistent with the findings of @xcite , who conducted a gbt search for radio pulsations at @xmath4 mhz from unseen companions to eight sdss wds ( five are still considered lmwds ; the three others are now classified as `` ordinary '' wds ) .
we discuss the constraints our non - detections place on the probability @xmath5 that the companion to a given lmwd is a radio pulsar in the context of the luminosity and acceleration limits of our search ; we find that @xmath6 . |
the spectral resolving power @xmath3 is perhaps the most important single property of a spectrograph .
the wavelength increment @xmath4 is the minimum separation for two spectral lines to be considered as just resolved .
the problem is that the definition of @xmath4 is arbitrary , and inconsistent between various usages .
classically the rayleigh criterion was used , while in recent years by far the most common practice has been to use the full - width at half maximum , i.e. @xmath5 .
it is clear that there can be no fundamental definition of the minimum resolvable wavelength difference @xmath4 , because with arbitrarily high signal / noise ratio , sufficiently fine sampling and a perfectly known instrumental response function ( here abbreviated as the line spread function , lsf ) an observed spectrum could be deconvolved to any desired spectral resolution .
what spectroscopists understand by the ` resolution ' of an instrument is the smallest @xmath4 which does not require ( significant ) deconvolution to obtain spectral line strengths and locations ( wavelengths ) .
lines of this separation can be distinguished at moderate signal / noise levels .
this arbitrariness in the definition of @xmath4 has always been recognised , from the early use of the rayleigh criterion .
there is in principle no problem with an arbitrary definition of @xmath4 and hence @xmath6 , provided it is consistent between various systems that are to be compared .
thus meaningful comparisons could be made using @xmath5 _ provided _ that the lsf has the same form in each case .
but the problem arises because this is not true : a diffraction - limited slit spectrograph gives a @xmath7 profile , a projected multi - mode circular fibre feed gives a boxy profile ( a half ellipse ) , a fabry - perot etalon with high finesse gives a lorentzian profile , a single - mode fibre or waveguide will give a gaussian profile , and a lsf with significant aberrations may resemble a gaussian but in general will have its own unique form .
it is when comparing resolving power between instruments with different forms of lsf that inconsistency arises , and as shown below the inconsistency can exceed a factor of two in resolving power .
this is a significant error in the context of scientific requirements for resolution , e.g. in stellar abundance studies .
moreover , resolving power is typically one of the formal specifications of a spectrograph , yet without a description of the lsf and the way @xmath4 is to be measured , any requirement on @xmath6 is necessarily imprecise in its meaning .
likewise , the concept of signal / noise per resolution element is vague because the ` resolution element ' is not well defined
. inconsistencies also occur between the well - known formulas for theoretical resolving power : \a ) @xmath8 for a diffraction - limited slit spectrograph with uniform illumination of all grating lines ( @xmath9 = diffraction order , @xmath10 = number of illuminated lines ) assumes a @xmath7 lsf and the rayleigh criterion , i.e. the maximum of a spectral line of wavelength @xmath11 occurs at the same position on the detector as the first zero of the line at @xmath12 .
\b ) @xmath13 for a fabry - perot instrument ( @xmath9 = order of interference , @xmath14 = etalon finesse ) assumes separation of the two lorentzian lsfs by their fwhm .
\c ) @xmath15 for a slit - limited spectrograph used in littrow configuration ( @xmath16 = collimated beam diameter , @xmath17 = grating incidence angle , @xmath18 = telescope diameter , @xmath19 = slit width in angular measure on the sky ) assumes rectangular lsfs ( i.e. perfect images of a uniformly illuminated slit ) and two lines are regarded as just resolved when the two slit images just touch .
there is thus a need to provide a more consistent definition of resolving power , so that comparisons can be made with better precision . in this paper
i first illustrate the problem by comparing various lsf forms with two lines separated according to the various criteria that have been proposed .
i then attempt to provide a consistent definition of resolution across different lsf forms .
the influence of sampling of spectra into discrete pixels is important in practice , but will be considered separately in a later work . for the present paper , sampling issues will be avoided by using a sufficiently large number of pixels so that profiles are effectively continuous .
this will keep the discussion focused on the issue of resolution itself .
the discussion here will be confined to 1-dimensional spectra , e.g. after processing to integrate over the spatial direction of a raw 2-dimensional data set .
figure [ fig : bigplot ] compares the different lsf profiles used in this work and the various resolution criteria .
there are a number of points to note from this figure . taking the rows in order : \1 ) the top row shows a single spectral emission line of each lsf form .
the @xmath7 , rectangular and lorentizian lsfs were introduced above .
the gaussian is often used as a general form of smooth profile , perhaps caused by many small errors and aberrations smoothing the ideal profile and combining via the central limit theorem to give a gaussian distribution .
the projected circle profile in column d applies to the case of a multimode fibre , which presents a uniformly illuminated circular image at the spectrograph entrance . when integrated over the spatial direction and presented as a profile along the wavelength axis , it has the form of a half - ellipse .
( this is an abel transform ; see e.g. bracewell 1995 p 367 . )
\2 ) this and the subsequent rows show a pair of identical lines separated according to various criteria .
the three numbers towards the right hand side of each panel show the separation / fwhm , the local minimum and the value of the autocorrelation at the separation shown .
panel 2a shows the classical rayleigh criterion separation of two @xmath7 profiles .
the local minimum between the peaks is 81.1% of the peak height . to many spectroscopists
this does indeed represent what is meant by two lines being just resolved .
but the separation is 1.129 @xmath20 fwhm , illustrating the inconsistency of the two criteria .
the rayleigh criterion , where one peak is placed over the zero of the other profile , can not be used for the gaussian or lorentzian profiles which do not have a zero . for the projected circle the boxy profile , with slope increasing as the edge of the profile is approached , produces the central spike in the sum as seen in all of panels 2d to 6d . in practice aberrations and pixelisation
will remove this to some extent , but its effects must still be considered .
\3 ) the rayleigh criterion can be generalised by taking its local minimum of 81.1% as the defining criterion .
this can be applied to all except the projected circle , due to its central spike .
\4 ) the fwhm is the most - used criterion nowadays .
but as panels 4a and 4b show , for the @xmath7 and gaussian profiles the resulting blended profile is not well resolved . for the @xmath7 profile ( 4a )
the local minimum is 97% of the peak , which does not accord with the common understanding of resolution .
a gaussian profile ( 4b ) is only a little better .
the projected circle ( 4d ) has an overall flux deficit between the peaks but a central spike at the midpoint . in practice
the result will depend on the degree of smoothing and pixelisation .
for the lorentzian profile ( 4e ) the relative minimum is well seen but only with good signal / noise , due to the substantial overlap of the line wings ( note the high autocorrelation of 0.498 ) .
\5 ) again using the @xmath7 profiles separated at the rayleigh criterion as a standard , this row takes the resulting autocorrelation value of 0.151 and uses it as a criterion for two lines to be just resolved . due to the high wings of the lorentzian
, it requires a separation of 2.366 @xmath20 fwhm to meet this criterion ( panel 5e ) .
\6 ) the equivalent width ( area / height ) has been proposed to meet some of the above objections ( e.g. jones et al .
1995 , 2002 ) . for the @xmath7 profiles ,
a separation of 1.0 equivalent width is extremely close to the rayleigh criterion . for other profiles
it also gives reasonable results .
the conclusion from figure [ fig : bigplot ] is that none of the separation criteria shown is clearly superior for all lsf forms , and in particular the fwhm is a poor indicator of resolution for the important cases of smooth sinc@xmath2 or gaussian profiles . for reference ,
the main properties of the lsf functional forms discussed in this paper are given in table [ tab : props ] .
lccccc & formula & peak & fwhm & ew & z + + @xmath7 & @xmath21 & @xmath22 & @xmath23 & @xmath24 & @xmath25 + gaussian & @xmath26 & @xmath27 & @xmath28 & @xmath29 & @xmath30 + projected circle & @xmath31 [ @xmath32 & @xmath33 & @xmath34 & @xmath35 & - + lorentzian & @xmath36 & @xmath37 & @xmath23 & @xmath38 & @xmath39 + + @xmath40all formulas are normalised to unit area under the profile .
this paper aims to present criteria by which resolving power can be more meaningfully compared across lsfs of different functional forms .
two approaches have been taken . in section [ sec : wavelength ]
a criterion based on wavelength accuracy will be given .
however the first criterion , to be discussed in this section , is developed by recognising that what an astronomer means by two close spectral lines being resolved is that the two can be seen separately and can have their strengths and positions ( wavelengths ) measured without undue influence of one on the other .
there will still be an arbitrary definition of what constitutes ` undue ' influence , but the aim is to ensure that there is only _
one _ arbitrary definition and that all other measures are consistent with it .
the influence of one spectral line on another is measured by its effect in increasing the noise in measurement of the flux of the line .
the procedure to use this method was to generate lsfs of various functional forms , with two equal strength peaks at separations varying from 0.8 to 2.0 @xmath20 fwhm , add noise to them and then perform least squares fits to extract the positions and strengths of the two peaks .
importantly , the width of each peak was treated as known rather than as a further variable to fit .
this was done for two reasons : ( 1 ) at the ultimate closest resolvable approach of two spectral lines it is recognised that the issue is to separate two unresolved lines .
it is well known that if the line width is itself resolved then lines would have to be further apart to be properly resolved .
this is not what ` spectral resolving power ' is taken to mean .
( 2 ) once two lines begin to blend , in the presence of noise the fitting process would be likely to result in one broadened line rather than two partly blended lines .
figure [ fig : example_dual_peak ] shows an example of two lines , with added noise , and the least squares fits .
the simulations were performed using the same noise power within the fwhm for each lsf form .
this is an unavoidably arbitrary choice of noise power normalisation , but it does not influence the results to be derived from these simulations .
the different lsfs were normalised to the same total area i.e. flux ( not peak ) .
this reflects the fact that total signal power in the spectral line is the quantity of importance to astronomers .
fwhm , with 62.6 samples over a fwhm , and subject to independent gaussian distributed noise with standard deviation 1.0 in each sample .
red curve : the least squares fit to two gaussians .
this plot shows one of 4000 realisations at one of 25 peak spacings . ]
figure [ fig : sig_flux_sepn ] shows the results of this process . for each of the five lsfs shown ( @xmath7 , gaussian , lorentzian ,
projected circle and projected circle convolved with a gaussian ) , a large number of trials ( 4000 ) was done at each of 25 separations from 0.8 to 2.0 @xmath20 the fwhm . from each set of 4000 trials the standard deviation of the least squares
fitted flux was found .
the smooth curves shown are semi - empirical model fits to the data of standard deviation versus peak separation and are used to smooth out irregularities due to random fluctuations .
the functional form fitted was : @xmath41 where @xmath42 is the lsf function , @xmath43 being the independent variable along the dispersion axis .
two free parameters , @xmath44 and @xmath45 , were adjusted to fit the simulation results for each lsf and in all cases gave a very good fit , within the residual fluctuations .
the values of @xmath46 were obtained using equation [ eqn : sig_flux ] below . at large separations
the standard deviations approach the value obtained for an isolated peak , i.e. by this criterion the lines are not influencing each other , and are fully resolved .
however the lorentz profile has such broad wings that it has not yet reached a constant level at the separation of 2.0 fwhm .
the lorentz profile shows the effect of one peak disturbing another ( i.e. increasing its noise ) at substantially larger separations than the other lsfs , when measured in multiples of the fwhm .
the @xmath7 and gaussian lsfs show very similar curves in figure [ fig : sig_flux_sepn ] , consistent with the fact that both are peaked functions which drop smoothly and rapidly towards zero .
the projected circle lsf has a very different curve of @xmath47 vs separation .
there is no effect at all of one peak on the other until they begin to touch , at @xmath48 fwhm = 1.1547 @xmath20 fwhm . at smaller separations
there is some interaction but it is very small because the profiles are convex with such steep sides .
figure [ fig : sig_flux_sepn ] also includes a curve for a projected circle lsf convolved with a gaussian of width such that the final fwhm is a minimum ( see section [ sec : proj_circ ] ) . figure [ fig : sig_flux_sepn ] makes clear that different lsf functional forms do indeed have very different properties as regards the mutual effects of two lines , and to simply use the fwhm as a resolution criterion is a poor indicator of spectral resolution as it affects line finding and fitting .
it is also clear that the lorentzian profile will give poor resolution at a given separation in fwhms , while the projected circle is exceptionally good .
the data in figure [ fig : sig_flux_sepn ] can be used to derive scaling factors to quantitatively compare different lsfs .
the method used here was to take a @xmath7 profile separated according to the rayleigh criterion as the standard of ` just resolved ' spectral lines .
this leads to a @xmath47 value increased by a factor of 1.0514 compared with its limiting value at large separations ( i.e. for isolated peaks ) .
other lsf forms will thus be considered to be just resolved when their @xmath47 values are likewise increased by @xmath49 over the value at large separations . defining a resolving power according to this criterion : @xmath50 where @xmath51 is the separation / fwhm required to achieve the above criterion , and
@xmath52 is the standard deviation of a flux measurement for an isolated peak ( equation [ eqn : sig_flux ] ) , the values given in table [ tab : scaling_factors ] are obtained .
although the @xmath7 profile was used as the standard for resolution , its value of @xmath51 is not unity because the rayleigh criterion corresponds to a peak separation of 1.129 @xmath20 fwhm .
the @xmath51 values show how much the resolving powers determined by the present criterion of equal disturbance in peak fitting due to an adjacent line differ from those based simply on the fwhm . as expected , the lorentzian is the worst , with an @xmath53 only 59% of its @xmath54 while the projected circle is the best , with @xmath55 exceeding @xmath54 by 20% .
the convolved projected circle is a more realistic case ( to be discussed in section [ sec : proj_circ ] ) and its resolving power , while less than the exact projected circle , is still substantially greater than a gaussian or @xmath7 .
vs separation of two peaks , for five different lsf forms .
from highest to lowest at peak separation = 1.0 the curves are : black - lorentzian ; green - gaussian ; blue - sinc@xmath2 ; magenta - projected circle convolved with a gaussian ( see section [ sec : proj_circ ] ) ; red - projected circle .
the blue square on the sinc@xmath2 curve indicates the rayleigh criterion separation . ]
lcc lsf form & @xmath51 & @xmath56 + + [ -0.2 cm ] @xmath7 & 1.129 & 1.129 + gaussian & 1.21 & 1.127 + lorentzian & 1.70 & 1.605 + projected circle & 0.83 + projected circle ( conv ) & 0.95 & 0.943 + figure [ fig : medplot ] shows profiles presented in the same style as figure [ fig : bigplot ] but with row 2 showing various lsf types with two peaks separated according to the criterion @xmath57 . these show the separations which are regarded as ` just resolved ' according to the criterion introduced here .
before introducing a second method of quantifying resolving power , it is necessary to review the formulas for uncertainties in the flux and position ( wavelength ) of a single spectral line peak .
when the width of the peak is known and only the amplitude ( flux ) and position ( wavelength ) are fitted by least squares , and assuming a symmetrical lsf form , clarke et al ( 1969 ) give the formulas : @xmath58 @xmath59 in these formulas @xmath60 is the rms noise in each wavelength channel and is assumed to be the same for all channels .
the summation is over all wavelength channels contributing to the profile .
the lsf function is @xmath61 , and @xmath62 denotes its derivative with respect to wavelength .
note that in these equations @xmath61 is normalised to a peak of 1.00 , and the ` pk ' in eqn [ eqn : sig_x ] is the peak flux of the response whose @xmath63 is to be found .
these formulas have been verified by monte carlo tests and show that the precision in finding the strength of a peak depends most on the values where the intensity is greatest , while the precision in location of the peak depends on the regions of greatest slope .
it is not appropriate in this paper to consider a detailed noise model where one would take into account shot noise from both the object and the background sky , as well as read - out noise and dark noise .
instead , it will suffice to use the above assumption of constant noise in all channels .
the results are thus most directly applicable to spectra that are background ( or read - out noise ) limited but serve as a guide for other noise models as well .
they can also be applied to absorption lines , especially those that do not depress the continuum by a large fraction . in the present work
a large number of channels ( pixels ) have been used , e.g. 62.5 or 100 across the fwhm , to avoid the issue of sampling effects . however , for the projected circle the gradient @xmath62 becomes infinite as the intensity drops to zero , and the sum in equation [ eqn : sig_x ] would always be dominated by the edge pixels ( see section [ sec : proj_circ ] ) .
hence this case is omitted here .
the more realistic convolved projected circle avoids this problem .
the second method to be considered originates from a somewhat independent property of high resolving power , namely the ability to measure accurate positions ( wavelengths ) of unresolved spectral lines .
thus two spectrographs can be considered as having equal resolving power if they give the same wavelength accuracy despite their different lsf forms , assuming the noise power per wavelength interval remains constant and equal total fluxes are received in both cases . to compare resolving powers using this criterion , there is no need to perform noise simulations as in section [ sec : consistent ] but instead equation [ eqn : sig_x ] can be used as follows .
define @xmath64 @xmath65 is a type of width measure of a lsf , which will be referred to as the ` noise width ' , given its role in calculating @xmath63 . for empirically determined lsfs @xmath65
will generally be calculated numerically as @xmath66 where @xmath67 is the channel width in the summation .
values of @xmath65 for the lsf types discussed here are included in table [ tab : props ] .
equation [ eqn : sig_x ] can now be written as @xmath68 where again @xmath60 is the rms noise in the channel of width @xmath69 and the subscript ` iso ' has been omitted because all profiles considered in this section are single .
the basis of this second resolution criterion is that @xmath70 of any lsf will be equated to @xmath71 , with the condition that the two profiles have equal total fluxes ( not equal peak values ) .
the condition for equal total fluxes is simply @xmath72 where @xmath73 stands for the equivalent width .
equating the @xmath63 s for the given lsf and for sinc@xmath2 and using the values of @xmath65 and @xmath73 for sinc@xmath2 from table [ tab : props ] there follows @xmath74 this is the fwhm of a sinc@xmath2 profile which would have the same wavelength noise error as the actual lsf being examined .
if the value is large it means that a wide sinc@xmath2 could give accuracy equal to the lsf i.e. the lsf is poor ( e.g. a lorentzian ) .
if the fwhm@xmath75 is narrow it means that a high resolution sinc@xmath2 is needed to equal the accuracy of a good lsf , e.g. the convolved projected circle .
the final step is to form the ratio of this calculated fwhm@xmath75 with that of the actual lsf and scale it by a factor 1.129 which will make the final scaled resolving powers consistent with the rayleigh criterion for sinc@xmath2 profiles .
this gives @xmath76 values of @xmath56 for the standard lsf forms are included in table [ tab : scaling_factors ] , except for the projected circle where the infinite gradient limit makes the calculation invalid .
values are quite similar to the @xmath51 scaling factors derived in section [ sec : consistent ] .
the interpretation of @xmath56 is that @xmath77 is the effective @xmath1 which should be used in place of the fwhm in order to calculate resolution on a scale consistent with the rayleigh criterion for a sinc@xmath2 profile .
thus @xmath78 is the resolving power on this consistent scale
. this criterion will be easier to use in practice than the @xmath79-based criterion of section [ sec : consistent ] .
for an empirically determined lsf , for example resulting from ray tracing of a spectrograph design , one would need to interpolate the lsf to a fine sampling interval , and smooth out any fine structure artefacts from the lsf calculation ( e.g. from a finite number of traced rays ) , then use equation [ eqn : z_sum ] to find the noise width and also find the fwhm and equivalent width ( area / peak ) .
then equation [ eqn : beta ] can be used to find the scaling factor which is finally applied in equation [ eqn : r ] . in the case of an asymmetric lsf the more general form of equation [ eqn : sig_x ] given by clarke et al .
( 1969 ) eqn ( a7 ) should be replaced by @xmath80 should be used , although the corrections for asymmetry are small .
the projected circle lsf is important in practice and has very different properties compared with other forms , and so warrants further discussion .
the use of multi - mode fibres to feed images to a pseudo - slit in a spectrograph is increasingly common . taking the fibre exit face as a uniformly illuminated circle ( a good approximation given the spatial scrambling produced by transmission along the fibre )
, when its image has been integrated over the spectrograph s spatial direction , the result will be the projected circle as illustrated in panel 1d of figure [ fig : bigplot ] .
it differs markedly from the sinc@xmath2 , gaussian and lorentzian forms in that the projected circle lsf approaches the x - axis with infinite slope .
this convex - outwards form results in the formation of a central spike when two such lsfs overlap , as in figure [ fig : bigplot ] .
interestingly , the projected circle line profile also results from doppler broadening of an intrinsically narrow line in a rapidly rotating star .
this is because the radial velocity is constant along strips parallel to the rotation axis , and the flux at any one wavelength is due to an integration along such a strip , i.e. a projection .
the effects of the very steep sides of such a profile have been noted , and dravins ( 1992 ) drew attention to the sharp spectral features which could appear at wavelengths where no spectral line is present , i.e. the central spikes as seen in figure [ fig : bigplot ] .
he also noted that information about the true stellar spectrum could be obtained regarding features considerably narrower than the fwhm of the full broadened profile - this is again due to the steep sides , which lead to the central spike being narrow and easily smoothed out ( in this case by intrinsic line width in a stellar spectrum ) .
as shown in section [ sec : consistent ] the lack of wings of the projected circle lsf result in minimal noise interaction of two close lines , i.e. its effective resolving power is substantially higher than its fwhm would suggest .
the pure projected circle lsf can not be directly compared with other lsfs as regards wavelength uncertainties , because of the infinite slopes .
this means that however fine the sampling may be , the @xmath56 value will still depend on the sampling interval .
this is illustrated in figure [ fig : proj_circle_beta ] which shows @xmath56 dropping approximately logarithmically with increasing sampling frequency .
the values of @xmath56 shown are all substantially less than any of those in table [ tab : scaling_factors ] .
even with some blurring due to aberrations a well - sampled lsf resembling the projected circle will have much higher wavelength accuracy than a gaussian - like peak of the same fwhm . for a projected circle lsf as a function of the number of samples across the full width to zero intensity ( fwzi ) .
] one of the peculiarities produced by the convex boxy shape of this lsf is that the fwhm is _ reduced _ by convolution with a gaussian of moderate width .
this effect was noted in the design of the aaomega spectrograph ( saunders et al .
this is another illustration of the inadequacy of fwhm as a measure of resolution , since one would not claim that convolution of the lsf by spectrograph aberrations increases the resolving power .
figure [ fig : conv_width ] illustrates this behaviour , using a projected circle lsf of fwhm = 1.00 convolved with gaussians of various fwhms up to 0.7 .
the resulting fwhm drops by as much as 5% , when the gaussian fwhm = 0.3259 , before rising again as the gaussian convolving function is further broadened .
figure [ fig : three_conv ] shows three of the profiles : the pure projected circle ; the case of the minimum final fwhm , and the case of gaussian fwhm = 0.595 which restores the final fwhm to 1.00 , albeit with a very different lsf form compared with the initial projected circle .
the case of the minimum final fwhm was used as the example of a convolved projected circle in table [ tab : scaling_factors ] and figures [ fig : sig_flux_sepn ] and [ fig : medplot ] . .
the curves from highest to lowest at the peak are : black : pure unconvolved projected circle ; blue : gaussian fwhm = 0.3259 gives the minimum final fwhm of 0.9494 ; red : gaussian fwhm = 0.595 results in a final fwhm of 1.00 . ]
the analysis above has shown that characterising the resolution @xmath81 of a spectrograph by its instrumental fwhm is a poor measure because it fails to take fully into account the variation among different line spread function forms of the quantities which matter most in spectroscopy - namely the disturbance which a spectral line causes to a near neighbour , or the accuracy with which a single line s wavelength can be measured .
using these two criteria , a very different picture emerges , as shown by the @xmath51 and @xmath56 scaling factors in table [ tab : scaling_factors ] . there is more than a factor of two difference in resolving power between the best and worst lsfs ( with identical fwhm ) when resolving power is measured on a consistent scale . comparing the various resolution criteria
shown in figure [ fig : bigplot ] with the @xmath47-based criterion of figure [ fig : medplot ] shows that the equivalent width is the one that comes closest to matching the consistent resolution scale introduced here .
but the match is not exact , with a significant difference in the case of the gaussian lsf .
the lorentzian lsf s broad wings greatly increase its effective @xmath81 and hence reduce the resolving power of an instrument with this lsf well below the value given by the fwhm .
it is well known by users of imaging fabry - perot instruments , for example , that this lsf makes the instrument unsuitable for absorption line studies , because a line core is influenced by convolution with continuum fluctuations over a substantial wavelength range . here
, this influence has been quantified and the lorentzian s low relative resolving power explicitly demonstrated .
conversely , the projected circle , even after smoothing by significant aberrations , has a steep - sided form which gives substantially higher resolving power than its fwhm would suggest .
gaussian and sinc@xmath2 profiles have properties intermediate between these two extremes .
but even they have ambiguities at the 10 - 15% level , with a pair of gaussian profiles requiring a separation of 1.129 @xmath20 fwhm to achieve the 81% relative minimum of a generalised rayleigh criterion .
either of the two resolution element scaling factors can serve as a quality indicator for any given lsf profile .
it is notable that the @xmath51 and @xmath56 scaling factors in table [ tab : scaling_factors ] are quite similar for a given lsf type , despite the former being based on the additional error in fitting the flux of a line caused by a near neighbour , while the latter is based on accuracy of wavelength determination for isolated lines .
this agreement strengthens the case for using one of these resulting scaling factors to bring resolving power of any spectrograph on to a consistent scale . in principle , the ` @xmath51 ' factor , based on mutual disturbance in fitting a line is the more appropriate in low to moderate signal / noise spectra , while the ` @xmath56 ' factor , based on wavelength accuracy , is the more appropriate for high - resolution , high signal / noise work . but given the similarity of the two factors and that the @xmath56 factor is much easier to calculate for a general empirically - determined instrumental profile , the @xmath56 factor is recommended as a suitable standard measure for comparison of resolving power between different spectrographs . + * references * + bracewell , r.n .
_ two dimensional imaging , _ prentice hall 1995 .
+ clarke , t.w . ,
frater , r.h . , large , m.i . , munro , r.e.b . and
murdoch , h.s . aust .
10 , 3 , 1969 .
+ dravins , d. _ high resolution spectroscopy with the vlt , _ ed .
ulrich , m .- h . ,
eso workshop no .
40 , 55 , 1992 .
+ jones , a.w . , bland - hawthorn , j. and shopbell , p.l .
77 , 503 , 1995 .
+ jones , d.h . , shopbell , p.l . and bland - hawthorn , j. mnras 329 , 759 , 2002 .
+ saunders , w. aao newsletter no . 108 , 8 , august 2005 .
+ spronck , j.f.p . ,
fischer , d.a . ,
kaplan , z.a . , schwab , c. and szymkowiak , a. arxiv 1303.5792 2013 . | the spectral resolving power @xmath0 is a key property of any spectrograph , but its definition is vague because the ` smallest resolvable wavelength difference ' @xmath1 does not have a consistent definition
. often the fwhm is used , but this is not consistent when comparing the resolution of instruments with different forms of spectral line spread function . here two methods for calculating resolving power on a consistent scale are given .
the first is based on the principle that two spectral lines are just resolved when the mutual disturbance in fitting the fluxes of the lines reaches a threshold ( here equal to that of sinc@xmath2 profiles at the rayleigh criterion ) .
the second criterion assumes that two spectrographs have equal resolving powers if the wavelength error in fitting a narrow spectral line is the same in each case ( given equal signal flux and noise power ) .
the two criteria give similar results , and give rise to scaling factors which can be applied to bring resolving power calculated using the fwhm on to a consistent scale .
the differences among commonly encountered line spread functions are substantial , with a lorentzian profile ( as produced by an imaging fabry - perot interferometer ) being a factor of two worse than the boxy profile from a projected circle ( as produced by integration across the spatial dimension of a multi - mode fibre ) when both have the same fwhm .
the projected circle has a larger fwhm in comparison with its true resolution , so using fwhm to characterise the resolution of a spectrograph which is fed by multi - mode fibres significantly underestimates its true resolving power if it has small aberrations and a well - sampled profile .
astronomical instrumentation spectrographs spectral resolving power data analysis and techniques spectral resolution |
a magnetic skyrmion , a nano - scale vortex - like spin texture in the real space , has attracted much attention in condensed matter physics.@xcite figure 1(a ) illustrates a magnetic skyrmion .
magnetic moments point downward at the center of the skyrmion and upward along the periphery ; intermediate magnetic moments vary their directions continuously such that the moments wrap the whole solid angle at the center .
the skyrmion s diameter is typically @xmath1 to @xmath2 nm and the magnetic skyrmion has been usually observed in a crystalline form ( the skyrmion crystal).@xcite since the skyrmion s spin configuration possesses the scalar spin chirality,@xcite topological phenomena appear in the hall effect,@xcite which is formulated by introducing an emergent electromagnetic field.@xcite besides its rich physics , the skyrmion has potential for applications especially in spintronics .
unique properties of the skyrmion , i.e. , particle - like nanostructure and topological stability , are highly promising for novel spintronic devices such as information storage or logic devices.@xcite the skyrmion is also appealing as an efficient information carrier ; in fact , recent experiments showed that magnetic skyrmions can be moved by electric current at low energy costs.@xcite in spite of the skyrmion s potential for spintronic applications , few studies have investigated skyrmions in terms of the pure spin current physics .
a pure spin current refers to a flow of spin angular momentum with no charge currents ; its understanding and control are central issues of spintronics.@xcite if the pure spin current can be generated from magnetic skyrmions , the generation should be interesting for basic research and would be a step forward in skyrmionic devices as well . here , we experimentally demonstrate spin - current generation from a resonantly driven skyrmion motion in a helimagnetic insulator @xmath0 . using a @xmath0/pt two - layer structure ,
we show that spin angular momentum transfers from the skyrmion crystal in @xmath0 to conduction electrons in pt via the spin pumping effect.@xcite the injected spin current is electrically detected in pt via the inverse spin hall effect ( ishe),@xcite which converts a spin current into a transverse electromotive force @xmath3 via large spin - orbit interaction of pt [ fig .
@xmath3 is given by@xcite @xmath4 when a spin current carries the spin polarization @xmath5 in the spatial direction @xmath6 .
we used @xmath0 for this work because the high - insulating property of @xmath0 enables electric detection of spin currents free from galvanomagnetic contamination.@xcite the skyrmion crystal in @xmath0 has been observed by lorentz transmission electron microscopy@xcite and neutron diffraction measurements.@xcite the helimagnetic ordering temperature is approximately @xmath7 k and a helical spin structure is stabilized by the dzyaloshinskii - moriya interaction at the zero magnetic field .
the skyrmion - crystal phase appears in the vicinity of the ordering temperature in a restricted window of magnetic fields for bulk @xmath0 crystals.@xcite skyrmions form a two - dimensional triangular lattice within the plane perpendicular to an applied magnetic field.@xcite the magnetic phase diagram of @xmath0 is almost the same for any direction of an applied magnetic field.@xcite
we grew single crystals of @xmath0 by a chemical vapor transport method.@xcite the volume of the crystals was typically @xmath8 mm@xmath9 .
the magnetic moment @xmath10 for the prepared crystals was investigated with a vibrating sample magnetometer ( vsm ) in a physical properties measurement system ( quantum design , inc . ) . to make a @xmath0/pt sample , we polished the surface of @xmath0 and sputtered a @xmath11-nm - thick pt film on the polished surface in an argon atmosphere .
figure 1(c ) illustrates an experimental set - up for detecting the spin pumping effect . to conduct microwave experiments
, we used a reflection - type coplanar waveguide and a network analyzer . since the skyrmion phase is restricted in a narrow window of magnetic fields , we fixed a magnetic field and swept a microwave frequency to measure magnetic resonance .
a static in - plane magnetic field was applied perpendicularly to a microwave magnetic field . when a microwave frequency fulfills resonance conditions of @xmath0 , magnetic moments are forced to precess ( magnetic resonance ) .
if this precession transfers spin angular momentum to conduction electrons in pt , the spin current is converted into @xmath3 in pt via the ishe . to measure @xmath3 , we attached two electrodes to the ends of the pt film [ fig .
applied microwaves were set at 10 mw .
the used @xmath0 sample was of column shape and 1.5 mm across and 0.5 mm high .
the spin - pumping measurement was performed in a cryogenic probe station , where the sample was cooled by attaching the bottom of a sample holder to a cold head .
a system temperature was measured with a thermometer near the cold head ; the sample temperature appeared higher than the system temperature by about 5 k in the present study .
the prepared @xmath0 single crystals were characterized by measuring @xmath10 .
figure 2 shows temperature dependence of @xmath10 at 300 mt . with decreasing temperature
, @xmath10 sharply increases at approximately 60 k and saturates at @xmath12 @xmath13 at 2 k. the ordering temperature and the saturation moment are consistent with those of preceding studies.@xcite the inset to fig .
2 shows magnetic - field dependence of @xmath10 at 5 k. @xmath10 saturates above 100 mt , which suggests transition into a ferrimagnetic phase . in the magnetization curve below 100 mt , a step - like change appears near 30 mt .
this change corresponds to a phase transition between multi - domain and single - domain helimagnetic states.@xcite hence , the magnetic properties of the prepared crystals are well consistent with the reported ones .
figure 3 shows microwave responses of a @xmath0/pt sample at 45 k and 50 k. hereafter , temperatures refer to a system temperature @xmath14 of the probe station .
figure 3(a ) shows frequency dependence of microwave absorption @xmath15 at 45 k , which is well lower than the helimagnetic ordering temperature . at 114 mt
, @xmath15 shows a dip around 3.5 ghz .
since the resonance frequency decreases with decreasing a magnetic field , the microwave absorption is attributed to a ferrimagnetic resonance of @xmath0 . below 54 mt ,
two dips appear and their resonance frequencies increase with decreasing a magnetic field .
the field dependence of the resonance frequencies is attributed to a helimagnetic phase induced by the dzyaloshinskii - moriya interaction.@xcite note that the two dips in the helimagnetic phase correspond to two spin - wave modes.@xcite figure 3(c ) summarizes field dependence of resonance frequencies at 45 k. we observed magnetic resonances of the skyrmion crystal between 47 k and 51 k. figure 3(b ) shows frequency dependence of @xmath15 at 50 k. the resonance frequency decreases with decreasing a magnetic field from @xmath16 mt , and then increases below @xmath17 mt .
this behavior is similar to that of @xmath18 k. from 30 mt through 6 mt , however , an additional excitation appears around 1.4 ghz while the excitation around 2 ghz is suppressed . below 3 mt
, the additional dip disappears and the helimagnetic one revives around 2 ghz .
figure 3(d ) summarizes field dependence of resonance frequencies at 50 k. while helimagnetic and ferrimagnetic modes appear both at 45 k and at 50 k , the new resonance mode appears at @xmath19 k only , as highlighted by the shading of fig .
its resonance frequency is lower than those of the other modes and monotonically increases with increasing a magnetic field .
this magnetic - field dependence obviously differs from that of helimagnetic modes .
moreover , the new magnetic excitation occurs in a restricted window of magnetic fields .
these results strongly indicate that magnetic excitations near 1.5 ghz originate from a collective motion of the skyrmion crystal . according to preceding studies on magnetic resonance of skyrmion crystals,@xcite
the newly observed mode is attributed to a counterclockwise rotation of magnetic skyrmions , in which skyrmion s core rotates counterclockwise .
note that the excitation observed around 2 ghz in the skyrmion - crystal phase is due to either a clockwise rotation of skyrmions@xcite or magnetic resonance of remnant helimagnetic states@xcite ; the origin of this excitation was unidentified in this experiment .
accordingly , we confine ourselves to the counterclockwise skyrmion motion to investigate the spin pumping effect .
figure 4(a ) shows a skyrmion phase diagram determined by magnetic resonance in the @xmath14 range slightly below the ordering temperature . here , critical magnetic fields @xmath20 , @xmath21 , and @xmath22 are defined as shown in fig .
3(d ) ; from @xmath20 through @xmath21 the skyrmion excitation continues while across @xmath22 a phase transition occurs from the helimagnetic state to the ferrimagnetic state . @xmath14 dependence of the critical magnetic fields reveals that magnetic excitations of the skyrmion crystal are restricted in a narrow window of @xmath14 ( approximately 4 k ) .
this observation is consistent with a phase diagram determined by measurements of ac magnetic susceptibility,@xcite although error bars for @xmath20 and @xmath21 are rather large because of small skyrmion excitations around @xmath20 and @xmath21 .
note again that the skyrmion - crystal phase was observed at @xmath14 lower than those reported elsewhere@xcite because in the present set - up we could not attach the thermometer directly to the sample as mentioned in the method part .
we demonstrate spin - current generation by measuring ishe - induced electromotive forces in pt at @xmath19 k. we first investigated the ferrimagnetic spin pumping at 66 mt [ point a in fig .
4(a ) ] . figures 4(b ) and 4(c ) represent a microwave - absorption spectrum and an electromotive - force spectrum at point a , respectively . in the microwave - absorption spectrum [ fig .
4(b ) ] , a single dip of ferrimagnetic resonance appears around 2 ghz .
correlating with the resonance , a single dip of electromotive forces appear around the resonance frequency , as shown in fig .
moreover , the electromotive forces show opposite polarity with an opposite magnetic field .
this reversal is consistent with the ishe [ eq .
( [ ishe ] ) ] and provides evidence for spin - current generation from the ferrimagnetic resonance .
next , we conducted the spin - pumping measurement in the skyrmion - crystal phase [ point b in fig .
a magnetic field was set at 15 mt to maximize microwave - absorption intensity within the skyrmion - crystal phase [ see also fig .
3(b ) ] . figures 4(d ) and 4(e ) are a microwave - absorption spectrum and an electromotive - force spectrum at point b , respectively . in the microwave - absorption spectrum [ fig .
4(d ) ] , two absorption dips appear ; the dip at the lower frequency ( @xmath231.3 ghz ) corresponds to the counterclockwise rotation of skyrmions [ see also figs .
3(b ) and 3(d ) ] . in the electromotive - force spectrum
4(e ) ] , electromotive forces appear around 1.3 ghz in correlation with the skyrmion excitation . again
, the electromotive forces show opposite polarity with an opposite magnetic field .
moreover , the electromotive force is of the same sign for skyrmion and ferrimagnetic phases for a given field direction . the results are consistent with eq .
( [ ishe ] ) because the direction of spin polarization should be determined by a static magnetic field also in the skyrmion - crystal phase .
these observations show that the resonantly excited skyrmion motion can give rise to the spin pumping effect and that the skyrmion dynamics can be detected electrically via the ishe . dividing the voltage by the intensity of microwave absorption@xcite yields 16 @xmath24 for the ferrimagnetic phase and 8 @xmath24 for the skyrmion - crystal phase , which implies the lower efficiency of the spin pumping in the skyrmion - crystal phase .
we numerically calculated spin currents related with skyrmion dynamics to demonstrate theoretically that the counterclockwise skyrmion motion can contribute to the spin pumping effect .
the dynamics of local magnetizations @xmath25 were simulated with the landau - lifshitz - gilbert ( llg ) equation@xcite for a single skyrmion on a two - dimensional square lattice in the @xmath26-@xmath27 plane .
the used hamiltonian is @xmath28 where @xmath29 mev and @xmath30 are coupling constants of ferromagnetic interaction and dzyaloshinskii - moriya interaction , respectively , and @xmath31 is a static magnetic field in the @xmath32 direction , i.e. , perpendicular to the square lattice .
the system size was @xmath33 sites .
the parameters used for the llg equation were : the gyromagnetic ratio @xmath34 hz / t ; the gilbert damping constant @xmath35 ; a static magnetic field @xmath36 mt in the @xmath32 direction ; a microwave magnetic field @xmath37 mt in the @xmath27 direction ; the microwave frequency @xmath38 ghz . with these parameters ,
the skyrmion s core resonantly rotates counterclockwise as shown in figs .
5(a)-(d ) .
note that values of the static magnetic field and the resonance frequency are not equal to those of our measurement because the simplest model of a single skyrmion was adopted to capture the essence of the spin pumping from the skyrmion dynamics . generated spin currents were calculated using the time average of @xmath39@xcite at the bottom edge of the system @xmath40 , namely @xmath41 , where @xmath42 represents the time average of @xmath43 .
as shown in figs .
5(e)-(h ) , both the magnitude and the sign of @xmath43 depend heavily on space and time .
the spatial dependence reflects a swirling spin structure due to the dzyaloshinskii - moriya interaction ; such behavior has not been observed for centrosymmetric systems such as @xmath44 , where the dzyaloshinskii - moriya interaction is absent .
the spatial and temporal dependence of @xmath43 at the edge gives rise to the complicated time evolution of generated spin currents @xmath45 as indicated by a blue dashed line in fig .
5(i ) . the temporal dependence of @xmath45 can not be described by a single sine wave , which implies that nonlinear magnetization dynamics were also driven for @xmath46 mt .
nevertheless , @xmath47 is found to be finite as highlighted by the shading in fig .
5(i ) , which theoretically substantiates our experimental results . as illustrated in figs .
5(a)-(d ) , magnetic moments even at the edge form a noncollinear configuration owing to the dzyaloshinskii - moriya interaction .
although a similar noncollinear configuration at the edge was found also for the ferrimagnetic phase in the calculation with @xmath48 t , the temporal dependence of @xmath45 for @xmath46 mt was simply sinusoidal .
we note that @xmath47 for the skyrmion phase was approximately 30 % smaller than for the ferrimagnetic phase in the present calculation .
this result indicates the lower efficiency of the spin pumping in the skyrmion phase , which is consistent with our experimental result .
the numerical calculation presented here demonstrates that non - uniform spin configurations such as skyrmions can give rise to the spin pumping effect while the observation of the spin pumping has been confined to uniform spin configurations such as ferromagnetic and ferrimagnetic states so far .
to summarize , we have observed spin - current generation from skyrmion dynamics in a @xmath0/pt two - layer structure .
we resonantly excited magnetic skyrmions in @xmath0 by a microwave and electrically detected spin currents via the ishe in pt .
a numerical calculation showed that a rotational motion of skyrmions can give rise to a finite dc spin current via the spin pumping effect .
spin - current generation via skyrmion dynamics demonstrated here can be a new functionality operating at microwave frequencies in spintronic applications .
we thank s. seki for fruitful discussions and r. iguchi for technical assistance .
this work was supported by crest - jst creation of nanosystems with novel functions through process integration , " strategic international cooperative program aspimatt from jst , and grants - in - aid for scientific research ( a ) from jsps ( no .
24244051 ) and for scientific research on innovative areas from mext ( `` topological quantum phenomena '' ( no . 25103702 ) ) as well as by mext kakenhi grant number 23108004 .
/pt two - layer structure .
a microwave magnetic field @xmath50 is applied perpendicularly to a static magnetic field @xmath51 . @xmath52 and @xmath53 denote the spin polarization and the spatial direction of a spin current , respectively .
@xmath54 denotes an electromotive force induced by the inverse spin hall effect .
( c ) experimental set - up for the spin - pumping measurement.,width=453 ] .
an applied magnetic field was set at @xmath55 mt . the inset shows magnetic - field dependence of the magnetization of @xmath0 at 5 k. the shading at @xmath56 mt highlights transitions between multi - domain and single - domain helimagnetic phases.,width=226 ] k and ( b ) at @xmath19 k at various magnetic fields .
black and blue curves represent data in ferrimangetic and helimagnetic phases , respectively .
red triangles in ( b ) indicate the magnetic resonance of the skyrmion crystal .
microwave - absorption spectra are shifted vertically for clarity .
( c),(d ) magnetic - field dependence of resonance frequencies ( c ) at @xmath18 k and ( d ) at @xmath19 k. the skyrmion resonance persists from the critical field @xmath20 through the critical field @xmath21 . across the critical field @xmath22 , a transition occurs between helimagnetic and ferrimagnetic phases.,width=453 ] , @xmath21 , and @xmath22 . for definitions of @xmath20 , @xmath21 , and @xmath22 ,
see the main text and fig .
the light pink background shows the skyrmion - crystal phase .
spin - pumping measurements were conducted at points a ( ferrimagnetic phase ) and b ( skyrmion - crystal phase ) .
( b)-(e ) frequency dependence of ( b),(d ) microwave absorption and ( c),(e ) electromotive forces at 50 k. ( b ) and ( c ) are for the ferrimagnetic phase at 66 mt ; ( d ) and ( e ) are for the skyrmion - crystal phase at 15 mt .
symbols are experimental data and solid curves are ( multi-)lorentzian fits to the data . a red dashed curve in ( d ) represents skyrmion - crystal resonance . ,
width=453 ] and @xmath27 components of local magnetizataions while the contour plot represents the @xmath32 component .
magnified views are shown in the panels below ( a)-(d ) .
( e)-(h ) representative snapshots of the spatial distribution of the damping torque , @xmath39 for the skyrmion phase .
spin currents generated at the edge , @xmath57 , were calculated .
( i ) time evolution of generated spin currents @xmath58 for the skyrmion phase ( blue dashed curve ) and a microwave magnetic field divided by a static magnetic field @xmath59 ( black solid curve ) .
the red line represents the magnitude of a time - averaged spin current @xmath47.,width=453 ] | we report spin - current generation related with skyrmion dynamics resonantly excited by a microwave in a helimagnetic insulator @xmath0 .
a pt layer was fabricated on @xmath0 and voltage in the pt layer was measured upon magnetic resonance of @xmath0 to electrically detect injected spin currents via the inverse spin hall effect ( ishe ) in pt .
we found that ishe - induced electromotive forces appear in the skyrmion phase of @xmath0 as well as in the ferrimagnetic phase , which shows that magnetic skyrmions can contribute to the spin pumping effect . |
optical antennas or nanoantennas have been the focus of research efforts due to their intriguing applications in wireless nano - links @xcite , photodetection @xcite , photo emission @xcite , and nonlinear plasmonics @xcite , to name a few .
antennas play a vital role as the mediator between free space propagation and localized energy in communication systems @xcite .
antennas may offer less loss and improved performance as compared to their plasmonic waveguide interconnect counterparts @xcite , specifically in optical wireless nano - links with long propagation distances of more than several wavelengths .
however , modifications to the conventional antenna theory and design techniques are needed in the optical frequencies due to the changes in the properties of metals @xcite .
designing individual antennas to resonate at optical frequencies is not the only problem to be solved .
impedance matched integration of antennas with waveguides , efficient energy extraction from ( or coupling to ) the antennas , and controlling the directivity of the antennas are other challenges that need to be tackled @xcite . recently , a widely used rf horn antenna @xcite design has been extended to operate at the optical frequency regime .
the horn geometry offers an inherently waveguide integrated design , a high directivity , and a large efficiency @xcite .
ramaccia et al .
first studied horn nanoantennas with a gradual exponentially tapered structure where metal - insulator - metal waveguides with cylindrical metallic pillars were incorporated @xcite .
a similar structure was adopted by yang et al .
@xcite where they designed a broadband and highly directive e - plane horn @xcite nanoantenna with straight flares fully incorporated in a homogeneous medium . in @xcite ,
yang et al . extended the same horn antenna concept , where a plasmonic channel waveguide , along with the flared section , was carved in a silver film .
both @xcite and @xcite presented a systematic study of the horn nanoantenna for different flare parameters , where the antennas were embedded in a homogeneous medium , and radiated in end - fire direction@xcite for wireless nano - link applications . on the other hand , when a nanoantenna is placed at the interface between two different media with different refractive indices
, it radiates a significant amount of power into the medium with the higher index @xcite .
if horn nanoantennas presented in @xcite , which were designed for end - fire radiation , are placed at the interface of two heterogeneous media , they will radiate mostly into the substrate . as a result ,
only a fraction of the total power can be coupled or extracted from the end - fire direction , as the antenna radiation will be directed elsewhere .
thus , there is a need to engineer the radiation of these nanoantennas , so that they can radiate in any given direction , for applications in wireless nano - links @xcite , and laser detection and ranging @xcite .
previously , plasmonic substrates ( i.e. metallic ground - planes ) were used to manipulate the nanoantennas radiation pattern @xcite .
if a reflector ground plane is employed , directly emitted light from the nanoantenna and the reflected light from the ground plane can interfere with different phases based on the thickness and type of the substrate material , enabling one to control and engineer the far - field radiation pattern @xcite . in @xcite , ghadarghadr et al .
reported that the dipole nanoantenna far - field radiation can be engineered by changing the distance between the dipole and the reflecting surface , and showed beam steering can be achieved in any arbitrary direction , ranging from end - fire to broadside .
similarly , reflector ground planes were also used by @xcite in order to direct the scattered radiation from nanoparticles or dipole nanoantennas out of the substrate plane . in this paper , we analyze and extend the plasmonic horn antenna design in @xcite to the case where the antenna is at the interface of air and a glass substrate with or without a metal mirror at its back .
analysis of the proposed design shows that beam steering can be achieved by changing the thickness of the substrate . in order to minimize the reflection from the antenna into the waveguide we use the impedance matching technique in @xcite .
rest of the paper is organized as follows . in section [ sec : sec2 ] , we introduce the horn nanoantenna .
section [ sec : sec3 ] gives an in - depth analysis of the horn nanoantenna on an infinitely thick as well as on a finite thickness , metal - backed substrate , along with the illustration of impedance matching technique .
finally , we conclude in section [ sec : sec4 ] .
the antenna structure with the ground plane and impedance matching section is shown in fig .
figure [ fig1](a ) shows top view of the antenna ( xy - plane ) , and fig .
[ fig1](b ) shows side view of the antenna ( xz - plane ) .
the horn antenna , along with the waveguide , is placed at the interface between air ( n=1 ) and glass ( n=1.44 ) .
the waveguide is in the form of a two - wire transmission line ( twtl ) with fixed dimensions @xmath1 given by @xmath2 nm , respectively , while the gap @xmath3 between these two wires is equal to 30 nm .
the horn length @xmath4 is 650 nm , the opening angle @xmath5 equal to 20@xmath6 , and width of the horn strip @xmath7 equal to 300 nm .
it should be noted that lengths @xmath8 , @xmath9 , and @xmath10 make a total length @xmath11 , where @xmath8 is position of the impedance matching ( i m ) section from flare and twtl interface , and @xmath9 is length of the i m section .
the substrate has a finite thickness @xmath12 in the case of a metal - backed antenna .
the nanoantenna was simulated and analyzed in comsol multiphysics .
the whole structure was excited by the fundamental mode that is supported by the twtl .
a rectangular _ numeric port _ was defined at the end cross - section of the twtl , and _ boundary mode analysis _ was used to calculate the mode at the specified port .
the _ boundary mode analysis _ performs an eigenfunction study to find the desired mode ( fundamental mode in this case ) supported by the twtl , and its associated propagation constant .
consequently , the same mode , with input power ( @xmath13 ) was fed to the antenna system .
the whole structure was bounded by a box which served for the purpose of calculating the far - field pattern of the antenna , and a pml layer of finite thickness was assigned at each face of the bounding box .
permittivities for a range of wavelengths of the silver @xmath14 metal were obtained by interpolating the data from @xcite .
in radio frequency ( rf ) and microwave engineering , an antenna s performance is evaluated by its near - field electric field ( e - field ) intensity , directivity , and how well the antenna is matched to the connecting waveguide by measuring the reflection coefficient @xmath15 .
directivity is defined as @xmath16 , where @xmath5 and @xmath17 are elevation and azimuth angle , respectively , @xmath18 is the angular power density , and @xmath19 is total radiated power .
figure [ fig2 ] shows the results for the nanoantenna where the substrate is an infinitely thick glass .
figure [ fig2](a ) and [ fig2](b ) show the near - field @xmath20 intensity in the xy and xz planes , respectively .
it is seen in fig .
[ fig2](a ) that the nanoantenna radiates , converting the bounded mode of twtl to planar free space propagating waves by providing a gradual impedance match between twtl and free space , while from fig .
[ fig2](b ) , one can observe that most of the radiation is directed into the substrate .
directivity of horn antennas is given in terms of e - plane and h - plane pattern @xcite .
e - plane is defined as the plane in which the primary e - field of the electromagnetic wave radiated by an antenna lies , which in our case is xy - plane , and h - plane is defined as the plane in which the magnetic field ( h - field ) lies , which is xz - plane in this case .
figure [ fig2](c ) shows the far - field h - plane directivity pattern of the antenna which has a maximum value of 14.5 at @xmath21 .
the amplitude reflection coefficient @xmath15 and the radiation efficiency @xmath19/@xmath22 are given in fig .
[ fig2](d ) as function of wavelength . at the central wavelength of 1550 nm ,
@xmath15 ( black solid line ) is less than 0.09 ( in linear scale ) meaning that @xmath23 power is delivered to the nanoantenna . during the calculation of the radiation efficiency ,
@xmath22 is calculated as @xmath24 @xcite , where @xmath25 is the waveguide s propagation length , and @xmath11 is length of the twtl .
the total radiation efficiency is around 0.61 , at 1550 nm , while the efficiency in substrate is greater than efficiency in air , which also confirms that most of the power is radiated into the substrate .
near field in the xy plane .
( b ) @xmath20 near field in the xz plane .
( c ) h - plane directivity at 1550 nm .
( d ) amplitude reflection coefficient @xmath15 , and total radiation efficiency , radiation efficiency in the substrate , and in air.,scaledwidth=90.0% ] from our analysis of the nanoantenna with an infinite substrate , we see that most of the radiation is directed into the substrate .
therefore , only a fraction of the total power can be extracted from the plane parallel to the nanoantenna , as required in the case of a nano - link @xcite . in this section
we investigate the effects of a reflective ground plane beneath the substrate .
we examine the radiation properties by changing the thickness @xmath12 of the substrate .
figure [ fig3 ] shows the results for a horn nanoantenna with a ground plane .
figure [ fig3](a ) and [ fig3](b ) show side view of near field @xmath20 of the nanoantenna for @xmath12 = 300 nm and 500 nm , respectively , where fig .
[ fig3](c ) and [ fig3](d ) show nanoantenna s directivity plot . for @xmath12 = 300 nm [ fig .
[ fig3](a ) and [ fig3](c ) ] , antenna radiates most of the power into the air with a directivity of 15 at @xmath26 . for @xmath12 = 500 nm [ fig .
[ fig3](b ) and [ fig3](d ) ] , the antenna radiates end - fire , with a directivity of 7 at @xmath27 .
similarly , the whole structure was simulated for a range of @xmath12 values between 200 and 900 nm , and we observed that for @xmath12 ranging from 200 nm to 450 nm , the antenna radiation was directed in air with the maximum directivity at @xmath5 between @xmath28 to @xmath29 . when @xmath12 was further increased above 450 nm until 650 nm , antenna radiated nearly end - fire with maximum directivity at @xmath5 ranging from @xmath30 to @xmath31 . by further increasing @xmath12 above 650 nm again
antenna radiated in air with almost same @xmath5 range as for below 450 nm .
near field in the xz plane for @xmath12 = 300 nm .
( b ) @xmath20 near field in the xz plane for @xmath12 = 500 nm .
( c ) & ( d ) are corresponding h - plane directivities at 1550 nm.,scaledwidth=90.0% ] the periodicity in antenna directivity can be explained by the antenna factor ( af ) @xcite , which is a sinusoidal function of thickness @xmath12 of the substrate , and the wave propagation constant .
in addition , the periodicity of the radiation pattern is similar to the results reported in figure 12(b ) of @xcite .
the modal reflection coefficient @xmath32 plot for the antenna with @xmath33 nm is given fig .
[ fig5](a ) with a black solid line . as seen in the figure
, @xmath15 increased to @xmath340.13 from a value of 0.09 in the case of an infinitely thick substrate . in order to reduce the back reflection from flares into twtl , in this section we employ impedance matching ( i m ) scheme based on conjugate matching @xcite .
for conjugate matching , a piece of dielectric ( glass ) was introduced in the gap of twtl of the horn nanoantenna , shown in fig .
[ fig1 ] and [ fig5](b ) , making a cascade matching network of total length @xmath11 equal to 1000 nm , comprised of air - core twtl with length @xmath10 , dielectric - core twtl with length @xmath9 , and again air - core twtl with length @xmath8 , from left to right as shown in fig .
[ fig5](b ) . for @xmath35 , according to conjugate matching , impedance of the horn flares should be equal to the complex conjugate impedance of the matching network . since impedance matching network in our case consists of the cascaded twtls with air or dielectric cores , input impedance at each interface [ fig .
[ fig5](b ) ] can be obtained by transmission line equation as follows @xmath36/[{1-\gamma(n-1)\exp(-2\gamma_nl_n)}].\ ] ] here @xmath37 is the characteristic impedance of the twtl @xcite at the end of which interface n lies , @xmath38 and @xmath39 are input impedance and reflection coefficient at interface n , and @xmath40 and @xmath41 are the complex propagation constant and length of twtl portion , respectively .
reflection coefficient @xmath15 is given by following equation @xmath42 equation [ eq:1 ] allows us to manipulate input impedance at each interface by changing @xmath41 , and @xmath43 , and by input impedance manipulation one can minimize reflection coefficient @xmath15 according to eq .
[ eq:2 ] . in order to achieve impedance matching
, we propose an analytical model , which predicts optimal lengths @xmath8 , and @xmath9 beforehand , rather than simulating for the range of lengths which is time and memory consuming . the analytical model is based on calculating normalized impedances at each interface of the horn nanoantenna with impedance matching network .
complete schematics of analytical model is given in fig .
it should be noted that all the structures used for modeling are placed on metal - backed substrate .
we start our modeling by simulating the horn nanoantenna with air - core twtl as shown in fig .
[ fig5](c ) . from the simulation
, we extract the reflection coefficient @xmath44 which is in turn used to calculate @xmath45 by transforming it by @xmath46 , according to equation @xmath45 = @xmath44@xmath47 .
the normalized impedance @xmath48 can be calculated by using eq .
similarly , as the second step of modeling , a cascade of air - core and dielectric - core twtls is simulated as shown in fig .
[ fig5](d ) , and @xmath49 was calculated as @xmath49=@xmath50@xmath51 , where @xmath52 , and @xmath53 are length and propagation constant of air - core twtl . similarly , normalized impedance @xmath54 can be calculated again from eq .
[ eq:2 ] . finally , rest of the modeling includes calculation , where fig .
[ fig5](c ) is cascaded to fig .
[ fig5](d ) , and forms fig .
[ fig5](b ) .
normalized input impedance at each interface can be calculated using eq .
[ eq:1 ] and previous modeling steps .
thus , one can obtain impedance matching by changing the lengths @xmath8 and @xmath9 .
= 300 nm ) and i m .
( a ) @xmath15 of : analytical model ( case 1 ) , simulated result ( case 2 ) with i m , and simulated antenna without i m ( case 3 ) .
( b ) analytic model for i m . ( c ) air - core twtl with antenna .
( d ) air - core twtl cascaded with glass - core twtl .
( e ) h - plane directivity at 1550 nm.,scaledwidth=60.0% ] figure [ fig5](a ) and [ fig5](e ) shows performance of antenna with ground plane and i m .
figure [ fig5](a ) gives reflection coefficient comparison of theoretical model ( case 1 ) with i m section , simulated result ( case 2 ) with i m section , and simulated result without i m section ( case 3 ) .
optimal lengths , @xmath55 and @xmath56 , equals to 350 nm and 125 nm respectively .
theoretical model and simulated results are in agreement , thus allowing one to predict optimal lengths , and reflection coefficient of antenna system beforehand .
furthermore , i m reduced reflection of antenna from @xmath57 to @xmath58 ( simulated case ) , which means that more than 61% of improvement has been achieved .
in addition , it can be observed from fig . [
fig5](e ) that i m does not have any significant effect on radiation pattern as the directivity of antenna remains the same .
in conclusion , the plasmonic horn nanoantenna with metal - backed substrate and i m has been investigated .
we studied the nanoantenna placed on semi - infinite substrate and results demonstrate that most of the power is radiated into the substrate .
we employ metal - backed substrate to achieve beam steering by varying substrate thickness .
results reveal that metal - backed substrate introduced impedance mismatch between the antenna and twtl .
we use impedance matching technique which allows us to accurately predict optimal parameters to minimize reflections .
impedance matching reduced reflections by more than 61% , and 99.75% of the input power is delivered to the nanoantenna .
we use the nanoantenna in the transmission mode , converting surface plasmons to free space propagating waves .
conversely , the angular selectivity and the enhanced directivity of the horn nanoantenna can also be utilized in the receiving mode , and can be used for potential applications in wireless nano - links , sensing , and free space to chip communications .
this work is supported by the scientific and technological research council of turkey - tubitak ( project no .
112e247 ) , tubitak - bideb 2215 - graduate scholarship program for international students , and ko university . | in this paper , we study a plasmonic horn nanoantenna on a metal - backed substrate .
the horn nanoantenna structure consists of a two - wire transmission line ( twtl ) flared at the end .
we analyze the effect of the substrate thickness on the nanoantenna s radiation pattern , and demonstrate beam steering in a broad range of elevation angles .
furthermore , we analyze the effect of the ground plane on the impedance matching between the antenna and the twtl , and observe that the ground plane increases the back reflection into the waveguide . to reduce the reflection , we develop a transmission line model to design an impedance matching section which leads to 99.75% power transmission to the nanoantenna . 10 y. yang , q. li , and m. qiu , `` broadband nanophotonic wireless links and networks using on - chip integrated plasmonic antennas , '' scientific reports * 6 * ( 2016 ) .
y. yang , d. zhao , h. gong , q. li , and m. qiu , `` plasmonic sectoral horn nanoantennas , '' opt . lett . * 39 * , 32043207 ( 2014 ) . j. m. merlo , n. t. nesbitt , y. m. calm , a. h. rose , l. dimperio , c. yang , j. r. naughton , m. j. burns , k. kempa , and m. j. naughton , `` wireless communication system via nanoscale plasmonic antennas , '' scientific reports * 6 * ( 2016 ) .
j .- s .
huang , t. feichtner , p. biagioni , and b. hecht , `` impedance matching and emission properties of nanoantennas in an optical nanocircuit , '' nano letters * 9 * , 18971902 ( 2009 ) .
a. al and n. engheta , `` wireless at the nanoscale : optical interconnects using matched nanoantennas , '' phys .
rev .
lett .
* 104 * , 213902 ( 2010 ) .
d. m. sols , j. m. taboada , f. obelleiro , and l. landesa , `` optimization of an optical wireless nanolink using directive nanoantennas , '' opt .
express * 21 * , 23692377 ( 2013 ) .
m. w. knight , h. sobhani , p. nordlander , and n. j. halas , `` photodetection with active optical antennas , '' science * 332 * , 702704 ( 2011 ) .
a. g. curto , g. volpe , t. h. taminiau , m. p. kreuzer , r. quidant , and n. f. van hulst , `` unidirectional emission of a quantum dot coupled to a nanoantenna , '' science * 329 * , 930933 ( 2010 ) .
k. lee , x. chen , h. eghlidi , p. kukura , r. lettow , a. renn , v. sandoghdar , and s. gtzinger , `` a planar dielectric antenna for directional single - photon emission and near - unity collection efficiency , '' nature photonics * 5 * , 166169 ( 2011 ) .
m. kauranen and a. v. zayats , `` nonlinear plasmonics , '' nat photon * 6 * , 737748 ( 2012 ) . l. novotny , `` effective wavelength scaling for optical antennas , '' phys . rev
. lett . * 98 * , 266802 ( 2007 ) .
d. ramaccia , f. bilotti , a. toscano , and a. massaro , `` efficient and wideband horn nanoantenna , '' opt . lett . * 36 * , 17431745 ( 2011 ) . y. xu , a. e. miroshnichenko , s. lan , q. guo , and l .- j .
wu , `` impedance matching induce high transmission and flat response band - pass plasmonic waveguides , '' plasmonics * 6 * , 337343 ( 2011 ) .
y. sachkou , a. andryieuski , and a. v. lavrinenko , `` impedance conjugate matching of plasmonic nanoantenna in optical nanocircuits , '' in `` elmar , 2011 proceedings , '' ( 2011 ) , pp . 389391 .
p.
ginzburg and m. orenstein , `` plasmonic transmission lines : from micro to nano scale with @xmath0/4 impedance matching , '' opt . express * 15 * , 67626767 ( 2007 ) . c. a. balanis , _ antenna theory : analysis and design _
( john wiley & sons , 2016 ) .
j. sun , e. timurdogan , a. yaacobi , e. s. hosseini , and m. r. watts , `` large - scale nanophotonic phased array , '' nature * 493 * , 195199 ( 2013 ) .
z. xi , y. lu , p. yao , w. yu , p. wang , and h. ming , `` controllable directive radiation of a circularly polarized dipole above planar metal surface , '' opt .
express * 21 * , 3032730335 ( 2013 ) .
q. min , y. pang , d. j. collins , n. a. kuklev , k. gottselig , d. w. steuerman , and r. gordon , `` substrate - based platform for boosting the surface - enhanced raman of plasmonic nanoparticles , '' opt .
express * 19 * , 16481655 ( 2011 ) .
a. ahmed and r. gordon , `` directivity enhanced raman spectroscopy using nanoantennas , '' nano letters * 11 * , 18001803 ( 2011 ) .
t. j. seok , a. jamshidi , m. kim , s. dhuey , a. lakhani , h. choo , p. j. schuck , s. cabrini , a. m. schwartzberg , j. bokor , e. yablonovitch , and m. c. wu , `` radiation engineering of optical antennas for maximum field enhancement , '' nano letters * 11 * , 26062610 ( 2011 ) .
s. ghadarghadr , z. hao , and h. mosallaei , `` plasmonic array nanoantennas on layered substrates : modeling and radiation characteristics , '' opt .
express * 17 * , 1855618570 ( 2009 ) .
d .- s .
ly - gagnon , s. e. kocabas , and d. a. b. miller , `` characteristic impedance model for plasmonic metal slot waveguides , '' ieee journal of selected topics in quantum electronics * 14 * , 14731478 ( 2008 ) .
p.
b. johnson and r. w. christy , `` optical constants of the noble metals , '' phys .
rev .
b * 6 * , 43704379 ( 1972 ) .
g. veronis and s. fan , `` bends and splitters in metal - dielectric - metal subwavelength plasmonic waveguides , '' applied physics letters * 87 * , 131102 ( 2005 ) . |
fluctuations and correlations are well known signatures of phase transitions . in particular , the quark / gluon to hadronic phase transition may lead to significant fluctuations @xcite . in 2010 , the relativistic heavy ion collider ( rhic ) began a program to search for the qcd critical point .
this involves an `` energy scan '' of au+au collisions from top collision energy ( @xmath3 = 200 gev ) down to energies as low as @xmath3 = 7.7 gev @xcite .
this critical point search will make use of the study of correlations and fluctuations , particularly those that could be enhanced during a phase transition that passes close to the critical point @xcite .
@xmath4 quantifies deviations in the particle ratios from those expected for an ideal statistical poissonian distribution @xcite .
the definition of @xmath5 ( describing fluctuations in the @xmath0 ratio ) is , @xmath6 a formula similar to ( [ nudyn ] ) can be constructed for @xmath1 and @xmath2 ratio fluctuations .
additional information about @xmath4 can be found in @xcite .
earlier measurements of particle ratio fluctuations utilized the variable @xmath7 @xcite .
the two variables are related as @xmath8 .
the data presented here for @xmath1 , @xmath2 , and @xmath0 dynamical fluctuations was acquired by the star experiment at rhic from minimum bias ( mb ) au+au collisions at @xmath3 = 7.7 , 11.5 , 39 , 62.4 , and 200 gev @xcite .
particle identification cuts can be found in @xcite .
data was also analyzed using the recently completed time of flight ( tof ) detector @xcite .
tof identified particles extended the @xmath9 range of pions and kaons from , @xmath10 gev/@xmath11 and protons from @xmath12 gev/@xmath11 .
we estimate that pion contamination of kaons is less than 3% using a combination of the star tpc and tof .
new results for dynamical particle ratio fluctuations from the rhic energy scan are shown in figures [ ppi_excitation ] , [ kp_excitation ] , and [ kpi_excitation ] .
figure [ ppi_excitation ] shows the measured dynamical @xmath1 fluctuations as a function of incident energy , expressed as @xmath13 .
the measured fluctuations from the star experiment are plotted as the black stars , from the na49 experiment @xcite at the sps as blue squares , and two transport model predictions from urqmd and hsd ( red and black lines , respectively ) , using the star experimental acceptance . all model predictions in this proceeding
are processed through the same acceptance .
the published na49 results have been converted to @xmath4 using the relation @xmath8 .
star measures a general trend from larger , negative values of @xmath1 fluctuations at lower energies , which then increases towards zero at @xmath3 = 200 gev .
these fluctuations are negative due to the dominance of the third term ( correlated production , e.g. @xmath14 ) in @xmath13 .
there is good agreement between the two different experiments at the lowest energies measured by star ( @xmath3 = 7.7 and 11.5 gev ) .
figure [ kp_excitation ] shows the measured dynamical @xmath2 fluctuations as a function of incident energy , expressed as @xmath15 .
results from the na49 experiment are from @xcite .
dynamical @xmath2 fluctuations show an increase from larger , negative values toward zero as a function of increasing incident energy .
this is qualitatively and quantitatively similar to the trend measured for dynamical @xmath1 fluctuations .
the experimental trend observed by star at the lowest incident energies ( @xmath3 = 7.7 and 11.5 gev ) is dramatically different than that measured by na49 .
the star result is consistent with negative dynamical @xmath2 fluctuations down to @xmath3 = 7.7 gev , while that from na49 is consistent with zero and increases to a positive value at @xmath3 = 6.3 gev .
figure [ kpi_excitation ] shows the measured dynamical @xmath0 fluctuations as a function of incident energy , expressed as @xmath5 .
results from the na49 experiment are from @xcite .
star measures dynamical @xmath0 fluctuations that are approximately independent of collision energy from @xmath3 = 7.7 - 200 gev .
this result is different than the observation of increasing fluctuations with decreasing incident energy observed by na49 .
it is worth noting that the differences between the two experiments occur for dynamical ratio fluctuations involving kaons .
there is good agreement between star and na49 for @xmath1 dynamical fluctuations .
figure [ netp ] shows the experimental measurements for the products of the higher moments of the net - proton distribution .
the higher moments of conserved quantities are predicted to be more sensitive to the correlation length of the system and may show a large change if the system passes through a critical point .
the products of the skewness*@xmath16 ( @xmath17 ) and kurtosis*@xmath18 ( @xmath19 ) are constructed to cancel volume effects @xcite . for the net - proton distribution
, it is observed that the products of the moments agree with a hadron resonance gas ( hrg ) prediction @xcite above @xmath3 = 39 gev and are below the prediction from @xmath3 = 7.7 - 39 gev .
these measurements are corrected for statistical effects related to the width of the centrality bin .
new results from dynamical particle ratio fluctuations ( @xmath1 , @xmath2 , and @xmath0 ) and the higher moments of net - proton distributions have been presented .
dynamical @xmath1 and @xmath2 fluctuations gradually increase from a larger negative value at @xmath3 = 7.7 gev toward zero at @xmath3 = 200 gev .
dynamical @xmath0 are positive and approximately energy independent from @xmath3 = 7.7 - 200 gev .
there are differences between the measured values of @xmath2 and @xmath0 fluctuations at @xmath3 = 7.7 gev from the star and na49 experiments .
results for the products of the higher moments of the net - proton distributions are in good agreement with the prediction of a hadron resonance gas model at high energies , but deviations are seen at energies below @xmath3 = 62.4 gev .
additional data at @xmath3 = 19.6 and 27 gev will provide precision measurements of all fluctuation observables .
99 v. koch , arxiv:0810.2520 [ nucl - th ] ( 2008 ) . | dynamical fluctuations in global conserved quantities such as baryon number , strangeness , or charge may be observed near a qcd critical point .
results from new measurements of dynamical @xmath0 , @xmath1 , and @xmath2 ratio fluctuations are presented .
the commencing of a qcd critical point search at rhic has extended the reach of possible measurements of dynamical @xmath0 , @xmath1 , and @xmath2 ratio fluctuations from au+au collisions to lower energies .
the star experiment has performed a comprehensive study of the energy dependence of these dynamical fluctuations in au+au collisions at the energies @xmath3 = 7.7 , 11.5 , 39 , 62.4 , and 200 gev .
new results are compared to previous measurements and to theoretical predictions from several models .
the measured dynamical @xmath0 fluctuations are found to be independent of collision energy , while dynamical @xmath1 and @xmath2 fluctuations have a negative value that increases toward zero at top rhic energy .
fluctuations of the higher moments of conserved quantities ( net - proton and net - charge ) distributions , which are predicted to be sensitive to the presence of a critical point , are also presented . |
while the spectra of late - m dwarfs are by definition very similar , their underlying nature can vary widely from youthful brown dwarfs to old low - mass stars .
the first field dwarf with a spectral type later than m6 was discovered by @xcite , yet even 50 years later the nature of such objects was open to debate , with @xcite suggesting that the latest type m dwarfs may all be young enough to be substellar .
@xcite proposed a method for discriminating between brown dwarfs and low - mass stars by using the doublet at 6708 , as models predict that stars and the highest mass brown dwarfs ( @xmath8 ) should deplete their initial lithium rapidly .
such model predictions of lithium depletion have also been used to age - date open clusters ( e.g. , * ? ? ?
* ; * ? ? ?
however , the underlying theory still remains unconstrained by direct mass measurements for ultracool dwarfs with lithium measurements near the predicted depletion boundary . in recent years
, spectroscopic signatures of low surface gravity have been used to identify very young ( @xmath910100 myr ) ultracool dwarfs among the field population of late - m and l dwarfs @xcite , thus enabling a means of discriminating between stars and brown dwarfs if the objects are young enough .
dynamical mass measurements provide a direct means for studying the diversity in the physical properties of late - m dwarfs .
we present here masses for the late - m binaries , , and based on keck natural guide star ( ngs ) and laser guide star ( lgs ) adaptive optics ( ao ) imaging from our ongoing orbital monitoring program targeting ultracool binaries .
we combine our new measurements with published results to perform a systematic study of late - m dwarfs . to that end
, we also present medium - resolution ( @xmath10 ) near - infrared ( nir ) spectra for this sample of late - m binaries .
the resulting combination of high - quality dynamical masses and spectra enables strong tests of model atmospheres over a wide range of masses , ages , and activity levels .
lp 349 - 25 was first identified as a nearby m8 dwarf by @xcite , and discovered its binary nature after being initially unresolved by @xcite .
@xcite measured a trigonometric parallax of @xmath11 mas for , corresponding to a distance of @xmath12 pc , which is somewhat more distant than the photometric estimate of @xmath13 pc .
@xcite obtained high - resolution integrated - light spectroscopy using keck / hires and found no evidence for lithium ( equivalent width @xmath14 ) .
@xcite recently published a dynamical mass of @xmath15 based on keck lgs ao data .
we derive here a dynamical mass of @xmath16 , based on an independent set of astrometry .
thus , this binary is a pair of brown dwarfs .
lhs 1901 was first identified as a nearby m6.5 dwarf by @xcite , and @xcite have independently classified it as m7.0 .
discovered to be a @xmath17 binary , which was also later identified by @xcite .
@xcite measured the parallax of to be @xmath18 mas , corresponding to a distance of @xmath19 pc .
our dynamical mass of @xmath20 indicates that this is a pair of very low - mass stars in a highly eccentric orbit ( @xmath21 ) .
gl 569b was discovered as a companion to the chromospherically active m2 star gl 569 by @xcite , and @xcite resolved it as an m8.5+m9 binary using keck ao .
there are three published orbit determinations for this binary @xcite that all give consistent values of the total mass ( @xmath22 ) . however , these orbits are all based largely on astrometry from 1999 to 2001 obtained with the keck nir cameras kcam and scam ( the slit - viewing camera for nirspec ) .
since these instruments are not astrometrically well - calibrated , we have derived a new orbit for based solely on data from keck / nirc2 and the _ hubble space telescope _ ( _ hst _ ) space telescope imaging spectrograph ( stis ) .
in addition , we use the revised _ hipparcos _
parallax for gl 569a of @xmath23 mas @xcite rather than the original value of @xmath24 mas @xcite used in all previous orbit determinations .
our improved measurement of the total dynamical mass for is @xmath25 , indicating that the binary is more massive than previously thought but still with at least one unambiguously substellar component .
in addition to these three binaries , we include the m8+m8 binary 2mass j22062280@xmath62047058ab ( hereinafter 2mass j2206@xmath62047ab ) in our study as it also has a well - determined mass ( @xmath26 ; * ? ? ? * ) and integrated - light spectroscopy . from the bright primary lhs 1070a ( m5.5 ) making it difficult to obtain spectroscopy of the late - m dwarfs .
we also do not consider four other late - m dwarf binaries from @xcite , as their dynamical masses are poorly constrained with 60190% errors . ]
we have monitored , , and with the ao system at the keck ii telescope on mauna kea , hawaii , using the facility nir camera nirc2 in its narrow field - of - view mode . at each epoch
, we obtained data in one or more filters covering the standard atmospheric windows from the mauna kea observatories ( mko ) filter consortium @xcite . for
, we used lgs ao @xcite when observing conditions allowed , and ngs ao otherwise ( see table [ tbl : obs ] ) .
the lgs brightness , as measured by the flux incident on the ao wavefront sensor , was equivalent to a @xmath27 @xmath7 9.510.4 mag star .
the tip - tilt correction and quasi - static changes in the image of the lgs as seen by the wavefront sensor were measured contemporaneously by a second , lower bandwidth wavefront sensor monitoring , which saw the equivalent of an @xmath28 @xmath7 14.314.6 mag star . for , we used ngs ao @xcite , and the incident flux on the wavefront sensor from was equivalent to a @xmath27 @xmath7 13.013.4 mag star .
for , the primary star gl 569a ( @xmath29 mag ) located 5@xmath300 from the science target provided the ngs ao correction .
our procedure for obtaining , reducing , and analyzing our images is described in detail in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
table [ tbl : obs ] summarizes our observations of , , and . the binary separation , position angle ( pa ) , and flux ratio were determined using the same three - component gaussian representation of the point - spread function ( psf ) as described in @xcite . at epochs
where the binary was sufficiently well - separated , we were also able to use the starfinder software package to simultaneously solve for the psf and binary parameters .
as described in @xcite , we assessed systematic errors in both psf - fitting procedures by applying them to artificial binary images constructed from images of psf reference stars with similar fwhm and strehl .
we found good agreement between both methods , with starfinder typically giving equivalent or smaller errors .
we adopted the astrometric calibration of @xcite , with a pixel scale of [email protected] mas and an orientation for the detector s @xmath32-axis of @xmath330@xmath3413@xmath310@xmath3402 east of north .
we applied the distortion correction developed by b. cameron ( 2007 , private communication ) , which changed our astrometry below the 1@xmath35 level .
the resulting relative astrometry and flux ratios for , , and are given in table [ tbl : astrom ] .
for , we also used our keck images to measure relative photometry between the primary star gl 569a and at @xmath36 , as there is no photometry published in the mko system .
we summed the flux in circular apertures with radii of 0@xmath303 to 0@xmath305 .
we used nirc2 in subarray mode to reduce the minimum allowed exposure time and thus avoid saturating the primary , which limited the size of the largest aperture we could use . however , in experimenting with even smaller apertures we found the same relative photometry to within 1@xmath35 .
we used a portion of the array minimally affected by light from either the bright primary star or the binary to measure the median sky level , which was subtracted from our summed aperture fluxes .
the relative flux of the binary in integrated light compared to gl 569a was @xmath37 mag , @xmath38 mag , and @xmath39 mag . to compute absolute photometry for
we applied these flux ratios to mko photometry for gl 569a , which was derived by converting its two micron all sky survey ( 2mass ) photometry to the mko system using offsets computed from synthetic photometry of a spex prism spectrum of gl 569a ( a. burgasser 2010 , private communication .
the conversions from 2mass to mko for @xmath40 , @xmath41 , and @xmath42 were @xmath60.03 mag , @xmath330.02 mag , and @xmath60.01 mag , respectively ( all @xmath43 compared to the 2mass errors ) .
thus , the final mko integrated - light photometry for was @xmath44 mag , @xmath45 mag , @xmath46 mag .
gl 569bab was observed by _ hst_/stis on 2002 june 26 ut as part of a program to obtain resolved optical spectroscopy of ultracool binaries ( go-9499 ; pi martn ) . during acquisition ,
the stis ccd imager took two images with the long - pass filter f28x50lp in which the binary is well - detected and the primary star gl 569a is saturated .
we analyzed these images using our psf - fitting routine based on the tinytim model of the _ hst _ psf @xcite , as described in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
we adopted the pixel scale and distortion solution from the stis instrument science report 2001 - 02 ( * ? ? ? * @xmath47 mas ) .
we tested for systematic errors in our best - fit binary parameters by applying our psf - fitting routine to artificial binary images created from stis ccd images of single stars from _ hst _ calibration programs in cycles 8 and 10 ( cal / stis-8422 and 8924 ) . to preserve the undersampled nature of the psf
, we created only artificial binaries separated by an integer number of pixels .
the separation and instrumental pa of in the stis images were 1.92 pixels and 3@xmath341 , so artificial binaries with @xmath48 were an excellent representation of the science data .
the single stars were @xmath710@xmath49 brighter in the stis images than , so we scaled them to match the science data and then injected them into the actual science image to accurately simulate the noise .
the binary images were injected at a distance of 5@xmath50 from gl 569a , chosen to be comparable to gl 569b s separation . for the final adopted errors we used the larger of the rms from the artificial binaries or the rms of the best - fit parameters to the science data . from our artificial binary tests , we found systematic offsets in separation ( 3.0 mas , 2.3@xmath35 ) , pa ( 0@xmath342 , 0.2@xmath35 ) and flux ratio ( 0.15 mag , 5@xmath35 ) , which we applied to our best - fit binary parameters for gl 569bab .
the final separation , pa , and flux ratio we derived were @xmath51 mas , @xmath52 , and @xmath53 mag ( table [ tbl : astrom ] ) .
we retrieved archival images of obtained with the very large telescope ( vlt ) at paranal observatory on five epochs spanning six months in 2006 .
these data were taken with the naco adaptive optics system @xcite using the n90c10 dichroic and s13 camera .
we adopt the pixel scale of @xmath54 mas derived by b. sicardy ( 2008 , private communication ) from observations of pluto s motion compared to its known ephemeris , and an orientation of the detector of @xmath55 .
we followed the same procedure for analyzing these data as in our previous work with vlt images @xcite .
we registered , sky - subtracted , and performed cosmic ray rejection on the raw archival images .
we used the same analytic psf - fitting routine as was used for the keck data to fit the vlt images , with uncertainties determined from the standard deviation of measurements from individual dithers .
the resulting best - fit parameters are summarized in table [ tbl : astrom ]
. published two epochs of astrometry for , and we investigated the possibility of improving the precision of their measurements by applying our psf - fitting procedure to their data , which is in public archives .
the cfht observations of on 2004 july 3 ut were not available in the cfht archive because they were obtained during engineering time ( t. forveille 2010 , private communication ) , but the vlt / naco observations from 2004 september 24 ut are in the vlt archive .
we measured a separation of @xmath56 mas , a pa of @xmath57 , and a flux ratio at @xmath41 band of @xmath58 mag .
our flux ratio and separation are in good agreement with those measured by , @xmath59 mag and @xmath60 mas , while our pa of @xmath57 is somewhat inconsistent with their @xmath61 ( at 2.8@xmath35 ) .
we tried using each of these measurements to fit the orbit and found that they gave consistent dynamical masses with a very similar @xmath62 for the best - fit .
our higher precision measurement would enable the uncertainty in some of the orbital parameters to be improved by up to a factor of @xmath72 .
however , we conservatively chose to use the published astrometry of as their larger errors and pa offset may result from analysis of astrometric calibration data unavailable to us that was taken along with their observations . published three epochs of astrometry for , and the first two data sets are available in the cfht archive .
unfortunately , not all of the astrometric calibration data was available in the archive , so independent measurements from these data can not be used in the orbit fit because , e.g. , the orientation of the detector at the time of those observations is not accurately known .
found uncertainties of 5 mas and 0.5 for all three data sets .
the rms scatter of our psf - fitting measurements is comparable or slightly lower ( 34 mas and 0.40.6 ) , and thus even with the needed calibration data a drastic improvement in precision seems unlikely .
thus , we use the published astrometry of when fitting the orbit of ( table [ tbl : astrom ] ) . for , we do not use any of the previously published astrometry .
we performed our own analysis on the original keck / nirc2 images from @xcite following the method described in section [ sec : keck ] .
@xcite published one epoch of astrometry for which is contemporaneous with our 2009 may data ; we conservatively choose to exclude their data from our orbit fit , although it is consistent with our best - fit orbit .
we do not include any of the keck / kcam or scam astrometry published by @xcite or @xcite , as those cameras have not been rigorously astrometrically calibrated , unlike nirc2 and the stis ccd .
in addition to astrometry , which we use to determine orbits and thus dynamical masses , we have also obtained nir spectra to perform tests of model atmospheres , as discussed in sections [ sec : results ] and [ sec : tests ] . using the nasa infrared telescope facility ( irtf ) spectrograph spex @xcite , we obtained spectra of , , and on 2008 july 6 , 2010 january 30 , and 2010 march 3 ut , respectively .
we used spex in cross - dispersed ( sxd ) mode , with five orders spanning 0.812.42 . for
we used a @xmath63 slit ( @xmath64 ) , and for and we used a @xmath65 slit ( @xmath66 ) .
we calibrated , extracted , and telluric - corrected the data using the spextool software package @xcite .
we used integrated - light photometry from 2mass to flux - calibrate our spectra of and . for
, we used our own integrated - light photometry derived from our keck images ( section [ sec : keck ] ) to flux calibrate its spectrum .
we used the mean of the @xmath36 scaling factors computed for each bandpass to place our spectra on an absolute flux scale .
combining our keck ao monitoring with published discovery data , we have observations spanning 5.9 years for , 5.9 years for , and 7.9 years for .
following our previous studies , we have determined the orbital parameters and their uncertainties using both a markov chain monte carlo ( mcmc ) approach , described in detail by @xcite , as well as the least - squares minimization routine .
in addition , we have developed our own visual binary orbit fitting routine that utilizes the levenberg - marquardt algorithm as implemented in the mpfit routine for idl @xcite . to determine uncertainties on the best - fit parameters from this routine , we have performed monte carlo simulations in which gaussian noise is added to the data corresponding to their measurement uncertainties .
we used the resulting distributions to determine the medians and standard deviations of the best - fit parameters .
when multiple astrometric measurements were available from different bandpasses at a single epoch , we used the highest precision measurement in the orbit fit ( see table [ tbl : astrom ] ) .
the best - fit orbits are shown in figures [ fig : orbit - skyplot ] and [ fig : orbit - sep - pa ] .
we found essentially symmetric uncertainties in the mcmc - derived orbital parameters for .
these parameters agreed to within @xmath67 of those derived by our custom fitting routine , and the errors agreed to within @xmath68 .
as an additional test of our levenberg - marquardt fitting routine , we found that the best - fit parameters from also agreed to within @xmath69 , and the errors derived from our monte carlo approach agreed to within @xmath70 of the errors computed from the full covariance matrix in .
the orbital parameters and errors derived by our mcmc analysis and custom fitting routine are given in table [ tbl : orbit ] .
we experimented with including the five epochs of relative astrometry from @xcite when fitting the orbit of .
if all of their data are included , the @xmath62 becomes unrealistically large : @xmath62 of 58.2 for 31 degrees of freedom ( dof ) .
this large @xmath62 value is due to the fact that two of their measurements from 2007 december and 2008 may are extreme outliers from the best - fit orbit .
( indeed , their own orbit fit is inconsistent with these data as the @xcite orbit has a reduced @xmath62 of 2.15 , which is unrealistically large . ) by excluding their most discrepant epoch ( 2008 may 30 ut ) , we were able to achieve a reasonable @xmath62 ( 33.1 for 29 dof ) .
neither the orbital parameters nor their uncertainties change significantly after including the @xcite data .
this is because their astrometry does not offer additional unique constraints on the orbit , as it was obtained contemporaneously with ours and is of lower precision by a factor of @xmath74 to 10 .
thus , we conservatively choose to exclude the @xcite astrometry from our analysis . our best - fit total mass ( ) for is @xmath71 ( 0.6% ) , and after including the 2.1% error in the parallax this becomes @xmath16 ( 6% ) . from our mcmc analysis of
, we found significant asymmetries in the posterior probability density distributions of the orbital period , semimajor axis , and eccentricity .
the confidence limits on all orbital parameters are given in table [ tbl : orbit ] , and the distributions of these three parameters are shown in figure [ fig : pdist ] .
the orbital period and semimajor axis display a high degree of covariance as expected ( e.g. , see section 3.2 of * ? ? ?
there appear to be two slightly degenerate solutions , with one having a shorter period , lower eccentricity , and higher mass than the other .
our mcmc analysis enables us to treat this degeneracy properly by accounting for it in the confidence limits of these orbital parameters and the resulting dynamical mass .
the difference in mass between these two solutions is very small ( about 1.6% ) , insignificant compared to the 12% error contributed by the distance uncertainty .
our best - fit total mass for is @xmath72 ( 1.4% ) , and after including the @xmath73% error in the parallax this becomes @xmath20 ( 12% ) .
the orbital parameters we find from are all consistent with previously published orbits , with a singular exception . the semimajor axis we find ( @xmath74 mas ) is 4.3@xmath35 higher than the value of @xmath75 mas derived by @xcite , and the @xcite and @xcite values are similarly discrepant ( by 1.6@xmath35 and 3.7@xmath35 , respectively ) .
since all orbit determinations to date have been dominated by the nine epochs of astrometry from keck kcam and scam , the semimajor axis discrepancy implies that the pixel scale of one or both of these instruments was underestimated by @xmath75% , resulting in systematically low semimajor axes .
this is supported by the fact that the previously published nirc2 astrometry from @xcite and @xcite is in good agreement with our best - fit orbit .
also , the _ hst_/stis measurement would be highly inconsistent with a value for the semimajor axis as low as previously found , showing that our semimajor axis is not simply due to a problem with the nirc2 calibration . since we find the same orbital period as in previous work but a semimajor axis that is 5.4% larger , our dynamical total mass is significantly higher than previously found .
this would be a 16% effect , but it is moderated somewhat by the revised _
hipparcos_parallax that is 1.6% ( about 1@xmath35 ) higher than the original value , which brings the mass down by 4.8% for a net change of 11% .
our best - fit total mass for is @xmath76 , and after accounting for the 1.6% parallax error this becomes @xmath77 .
our mcmc analysis also reveals a slightly degenerate orbit solution with double - peaked distributions for the argument of periastron ( @xmath78 ) and pa of the ascending node ( @xmath79 ) as shown in figure [ fig : pdist ] .
the degeneracy in these values are apparently not correlated with any other orbital parameters .
@xcite used the integrated - light optical spectrum of to derive a spectral type ( spt ) of m@xmath80 . comparing our near - infrared spectrum of to spectra of m dwarf standards from the irtf
spectral library @xcite , we find an infrared spectral type of m8 . without resolved spectroscopy of the binary , we can not directly determine the spectral types of the individual components .
however , we have used our photometry and the empirical spt absolute magnitude relations of ( * ? ? ?
* @xmath40 band only ) and ( * ? ?
* @xmath81 band only ) to estimate spectral types .
we used the method described by @xcite , which accounts for both the measurement uncertainties and the intrinsic scatter in the empirical absolute magnitude relations .
we found that the range of @xmath40-band absolute magnitudes allowed by our measurement errors exceeded the bounds of the @xcite relation , and thus we base our estimates only on the -band relation .
we estimate spectral types of m@xmath82 for the primary and m@xmath83 for the secondary , which are consistent with the published integrated - light spectral type .
@xcite measured the integrated - light spectral type of to be m6.5 , and @xcite measured m7.0 .
our near - infrared spectrum is better matched by the m dwarf standard vb 8 ( m7 ) than gj 1111 ( m6.5 ) , thus we find an infrared type of m7 for . using the method described above , the @xmath40-band absolute magnitudes of the individual components provide an estimated spectral type of m@xmath84 for both the primary and secondary , consistent with measured integrated - light spectral types .
the difference of zero subtype between the components is also consistent with our measured flux ratios and late - m dwarfs with known parallaxes .
for example , the rms scatter among @xmath42-band absolute magnitudes for m6.5 dwarfs is 0.12 mag ( cf .
@xmath85 mag for ) . for
, we adopt the spectral types of m@xmath86 and m@xmath87 derived by @xcite .
they used their resolved near - infrared spectra of the individual components to compute spectral indices and compared these values to field dwarfs of known spectral type .
they found a difference of 0.5 subtypes between the two components and assigned the integrated - light spectral type of m8.5 from @xcite to the primary , resulting in a type of m9.0 for the secondary .
these spectral types are consistent with those found by using resolved optical spectroscopy of from _ hst_/stis .
we computed the total bolometric luminosities ( ) of our sample binaries from their integrated - light spectra and photometry .
in addition to our spex spectra , we used the and band photometry from @xcite and the 24 _ spitzer _ photometry from @xcite for , and we used the band photometry from @xcite for . for , we estimated a @xmath88 color of @xmath89 mag from the empirical relation of @xcite to extend its sed to longer wavelengths . to accurately account for the sed at shorter wavelengths , we appended optical spectra , using the overlapping region of our spex spectra to determine the flux scaling .
for and we used optical spectra from @xcite and @xcite , respectively . for
we used the optical spectrum of the m7 standard vb 8 @xcite .
we numerically integrated the spectra and photometry points , interpolating between the gaps in the data , extrapolating to zero flux at zero wavelength , and assuming a rayleigh - jeans tail beyond the last photometric point .
we accounted for errors as described in @xcite , finding a total @xmath90 of @xmath91 dex for , @xmath92 dex for , and @xmath93 dex for .
we apportioned these total bolometric fluxes to the individual components by using the measured @xmath42-band flux ratio and the bolometric correction
spt relation derived by @xcite , properly accounting for uncertainties in the component spectral type estimates .
the resulting individual luminosities are given in table [ tbl : meas ] .
for the components of , our values are in good agreement with the estimates from @xcite .
as in our previous work , we correctly account for the covariance between the individual luminosities ( via the flux ratio ) in the following analysis , allowing more precise determinations of relative quantities ( e.g. , @xmath94 and the mass ratio ) .
we used a variety of solar - metallicity model atmosphere grids to fit our integrated - light irtf / spex spectra , summarized in table [ tbl : atm - ref ] .
the two publicly available grids we used are ames - dusty @xcite and phoenix - gaia @xcite .
we also used an improved version of ames - dusty that employs the same treatment of dust but with updated line lists for some important opacity sources such as feh , crh , and h@xmath95o ( see section 3.2 of @xcite ) ; this `` gaia - dusty '' model grid is the same as that employed by @xcite .
our fourth model grid drift - phoenix relies on yet another gas opacity data set and , more importantly , features improved dust opacities .
the model treats the dynamics of the dust in a presently unrivaled level of detail by calculating the rate of seed formation and solving an actual growth rate equation system for gravitational settling of composite grains .
we fit model atmospheres to our spectra using the procedure outlined by @xcite . for every model spectrum
, we found the optimal scaling factor to match the observed spectrum by minimizing the @xmath62 statistic .
the model spectrum with the lowest resulting @xmath62 value gives the best - fit parameters for and .
for each spectrum , we fit individual wavelength ranges corresponding to the standard near - infrared bandpasses ( @xmath96 ) .
we also performed fits to the entire sed , both with and without the segment of the spectrum between 0.81 and 0.95 .
this region of the spectrum is often excluded when fitting models to l and t dwarfs ( e.g. , * ? ? ?
* ) because it is strongly affected by the broad wings of the resonant doublet at 0.77 , which have long been difficult to model ( e.g. , * ? ? ?
* ) and can be impacted by dust modeling @xcite . such alkali lines are present but weaker in late - m dwarfs , and they may contribute to observed discrepancies with models between @xmath70.7 and 0.9 for objects as early as l0 .
the results of our model atmosphere fitting procedure are given in table [ tbl : atm - fit ] .
we report fits both for our full wavelength range ( 0.812.42 ) and for the nir only ( 0.952.42 ) .
we prefer the nir sed fits for the following reasons : ( 1 ) the nir sed fits are more consistent with the individual @xmath96 band fits ( typically within one model grid step of 100 k , 0.5 dex in ) ; ( 2 ) the full sed fits are systematically cooler by @xmath7100 k compared to the individual band fits , implying that the lowest @xmath97 portion of the spectrum from 0.81 to 0.95 has a disproportionate influence on the results ; and ( 3 ) by adopting a 0.95 cutoff here , our results will be directly comparable to future work on l and t dwarfs that require such a cutoff because of the problems with lines .
figure [ fig : chi2 ] shows the @xmath62 contours of the fits to our observed nir spectral energy distributions ( seds ) using all model spectra .
the elongation of the contours in the direction indicates that gravity is less well - constrained by our fits than effective temperature .
we also note that there are occasionally local @xmath62 minima in these plots that correspond to physically implausible parameters . for drift - phoenix
, the deviant @xmath62 minimum occurs at very low temperatures ( 2000 k ) and high gravities ( @xmath98 , which for masses of @xmath990.08 correspond to radii of @xmath990.045 , more than a factor of two lower than evolutionary model predicted radii ) .
for the other three model grids , @xmath62 minima extend to higher temperatures ( @xmath73300 k ) and extremely low gravities ( @xmath100 , which for masses of @xmath1010.06 correspond to radii of @xmath1010.7 ) . in the rare cases
when the global minimum is found in one of these physically implausible regions ( @xmath73% of fits , see table [ tbl : atm - fit ] ) , we select the second best fitting model instead , and this always results in best - fit parameters from the prevailing solutions of = 27003000 k. the different grids of model atmospheres generally give very similar best - fit parameters despite their significantly different input assumptions .
for example , the treatment of dust in the drift - phoenix models is much more realistic than in other models , yet for our sample it yields the same values as other grids .
thus , it appears that the effects of dust at the warm temperatures of late - m dwarfs only subtly impact the resultant model spectra for our purposes .
in fact , for the nir sed fits ( 0.952.42 ) , all four sets of models give exactly the same .
finally , we investigated how binarity impacts the atmospheric parameters derived from integrated - light spectra .
we constructed artificial binary integrated - light spectra by summing individual irtf / spex sxd spectra , both from our own spectra ( e.g. , summing the spectra of and ) and from publicly available spectra of the m7 , m8 , and m9 spectral standards vb 8 , vb 10 , and lhs 2924 @xcite .
we fit these spectra both individually and summed , after scaling down the later type spectrum to best match the measured @xmath36 flux ratios of and .2047 both have flux ratios within @xmath70.1 mag of unity across @xmath102 bandpasses and thus have components that are likely to be spectrally identical .
for example , has a bolometric flux ratio of @xmath103 dex , and thus the difference in between the individual components is expected to be 75 k ( i.e. , less than one model grid step ) for an assumed radius ratio of unity . ] comparing the summed and input `` primary '' spectrum fits , we found that the best - fit values from individual bands were nearly identical ( at most different by one grid step of 100 k or 0.5 dex ) . in some cases of we found that the summed spectra gave systematically lower temperatures of 100 k as compared to the input `` primary '' spectrum .
however , this was not the case when the input spectra had @xmath36 colors comparable to the components of and .
we note that within each binary the components have very similar @xmath36 colors , so we might expect that simulated component and summed spectra would give the same results .
for example , the integrated - light @xmath104 color of is 1.02 mag , and its components are only different by @xmath70.03 mag ( @xmath105 mag and 1.05 mag ) despite the binary flux ratio of @xmath70.3 mag ( the @xmath106 and @xmath107 colors are even closer , @xmath108 mag ) . in summary ,
our simulations show that the individual band fits should be unaffected by contamination from the spectrum of the secondaries in our sample binaries , and even if the sed fits are affected they will give values systematically lower by only 100 k. given that our best - fit values of the nir seds show no such systematic offsets from the band fits , we conclude that our fits to the integrated - light spectra accurately find the best - fit parameters of the primary component s spectrum .
we have considered whether the space motion and activity of our sample binaries can provide useful constraints on their ages . for
, there are two discrepant measurements of its radial velocity ( rv ) in the literature : ( 1 ) @xcite derived @xmath109 from cross - correlation of their keck / hires spectrum with the rv standard gl 406 ( m6 ) ; and ( 2 ) @xcite found a center - of - mass velocity of @xmath110 by fitting an orbit to their multi - epoch near - infrared rvs.phase = 0.02 ) , @xcite measured a primary rv of @xmath111 , which is more consistent ( at 1.6@xmath35 ) with the @xcite value . ] combining these rvs with the parallax and proper motion measured by ( * ? ? ?
* assuming 4 mas and 4 errors in their proper motion amplitude and pa , respectively ) , we derive heliocentric velocities for of @xmath112 = ( @xmath113 , @xmath114 , @xmath115 ) and ( @xmath116 , @xmath117 , @xmath118 ) , respectively . we adopted the sign convention for @xmath119 that is positive toward the galactic center and accounted for the errors in the parallax , proper motion , and rv in a monte carlo fashion .
these space motions place only 0.7@xmath35 or 0.9@xmath35 away from the mean of the ellipsoid defined by all other ultracool dwarfs with @xmath120 measurements .
( this sample is described in detail in section 3.4 of @xcite . ) using the same method as in our previous work , we have also assessed s membership in the galactic thin and thick disk populations using the besanon model of the galaxy , finding at most a 0.3% probability of thick disk membership using the @xcite rv and @xmath121 for the @xcite rv . thus , although the derived space motions are different , the conclusion that is a normal thin disk member is unaffected . as a likely thin disk member , the age of is essentially unconstrained by its kinematics . finally , the fact that is chromospherically active ( @xmath122 = @xmath64.52 ; * ? ? ?
* ) could potentially provide an age constraint , as the activity of m dwarfs changes with age .
however , the long lifetime of activity in such late - m dwarfs found by @xcite places only a weak constraint of @xmath123 gyr for the age of the system . given the measured mass and luminosities of
, it is in fact likely to be much younger , as discussed in detail in the following sections . for
, there is no published radial velocity , precluding the kinematic analysis described above . the proper motion and parallax of @xcite provide a measurement of its tangential velocity ( @xmath124 ) , indicating that is typical of late - m dwarfs within 20 pc ( @xmath125 ; * ? ? ?
* ) and not likely to be a member of the thick disk or halo populations . both @xcite and @xcite have shown that is not active , which is uncommon for nearby , late - m dwarfs .
@xcite found that activity lifetime increases monotonically with m dwarf spectral type , and for m6 and m7 dwarfs these lifetimes are @xmath126 gyr and @xmath127 gyr , respectively .
thus , the lack of activity seen in ( m6.5 ) indicates that its age is likely to be at least as old as 6 gyr , the 2@xmath35 lower limit for the activity lifetime of m6 and m7 dwarfs .
the age of the m2.5 primary star in the gl 569 triple system has been discussed extensively in the literature .
@xcite find that the space motion of gl 569a places it close to , but not likely a member of , the ursa major moving group ( 300 myr ; * ? ? ?
* ) . however , they suggest that it may be a member of a stream of young a and f stars identified by , which have ages of 300800 myr as derived from strmgren photometry . @xcite
prefer a younger age of 100125 myr because gl 569a seems to lie on the pleiades sequence in the color magnitude diagram of @xcite .
we have considered whether the x - ray emission from gl 569a may also place an independent constraint on the age of the system .
we computed the x - ray flux of gl 569a from its _ rosat _ count rate , using the @xcite conversion factor , finding @xmath128 erg @xmath129 s@xmath130 ( @xmath131 ) .
this lies between the median values of @xmath132 for m dwarfs in the pleiades ( 125 myr ) and hyades ( 625 myr ) as determined by @xcite .
the x - ray luminosity of gl 569a is higher than @xmath775% of hyades objects and is fainter than @xmath770% of pleiades objects .
this suggests an intermediate age of @xmath9300450 myr for the gl 569 system , although its x - ray luminosity is consistent with the full range of @xmath9125625 myr at 1@xmath35 .
directly measured dynamical masses enable unique tests of theoretical models of very low - mass stars and brown dwarfs .
we consider two independent sets of evolutionary models from the tucson @xcite and lyon groups ( dusty ; * ? ? ?
following the approach developed in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , we use the most directly measured properties of our sample binaries ( i.e. , total mass and individual component luminosities ) to derive additional properties ( e.g. , ) from evolutionary models .
these model - derived properties are then compared to other available constraints ( e.g. , from atmospheric model fitting ) .
we note that currently available evolutionary models do not incorporate many of the theoretical advances used in the latest model atmospheres , such as detailed dust cloud models and more accurate molecular and dust opacities . however , our line of analysis is necessarily centered on evolutionary models because we directly measure mass and , not the model atmosphere parameters and .
in addition to this analysis , we also explore the approach recently utilized by @xcite in order to directly compare our results to their findings for the same models . as described in detail by @xcite and @xcite , the total mass of a binary along with its individual component luminosities
can be used to estimate the age of the system from evolutionary models .
this age estimate can be quite precise ( @xmath710% ) when both components are substellar , since the luminosities of brown dwarfs depend very sensitively on age .
this is the case for : the lyon and tucson models give consistent ages of @xmath133 gyr and @xmath134 gyr , respectively ( figure [ fig : mtotage ] , table [ tbl : lpmodel ] ) .
the model - derived age of the system is consistent with its kinematics ( section [ sec : age ] ) , and the implications of such a young age for the system are addressed in detail in section [ sec : lpab ] .
because is composed of two very low - mass stars whose luminosities remain essentially constant over their main - sequence lifetime , the model - derived age for this system is not very precise .
lyon and tucson models give consistent ages of @xmath135 gyr and @xmath136 gyr , respectively ( table [ tbl : lhsmodel ] ) .
the large upper limits to the age distributions are set by the oldest age for which the model grids are computed ( 10 gyr ) .
these broad age distributions are consistent with the age constraint that comes from the lack of chromospheric activity in ( @xmath3 gyr ; section [ sec : age ] ) . because the median model - derived ages are quite young , they would be inconsistent with an age as old as 6 gyr , but a more precise parallax measurement is needed to test this .
for example , if we fix the parallax at its presently measured value but improve its precision from 3.0 mas to 1.0 mas , an attainable goal with seeing - limited astrometry , this age discrepancy would rise to @xmath137 significance for the lyon models .
gl 569bab has a higher mass than but lower component luminosities , and thus the lyon and tucson models give older ages of @xmath138 gyr and @xmath139 gyr , respectively ( figure [ fig : mtotage ] , table [ tbl : glmodel ] ) .
these are consistent with the age constraints from gl 569a discussed in section [ sec : age ] , with the exception of the age of 100125 myr proposed by @xcite based on its position on the color magnitude diagram . given the unavailability of such data for m dwarfs at @xmath9500 myr , we suggest that gl 569a may well be consistent with such an older age . our model - derived age for the system is also in agreement with the age estimates from @xcite and @xcite .
all of our target binaries have flux ratios near unity , thus enabling robust individual mass estimates with only a very weak dependence on model assumptions . for ,
the lyon and tucson models give mass ratios ( @xmath140 @xmath141 @xmath142 ) of @xmath143 and @xmath144 , respectively .
thus , the model - derived individual masses ( listed in table [ tbl : lhsmodel ] ) are nearly identical , and their precision is dominated by the @xmath145 uncertainty in . because the model - derived ages for are quite young , the inferred mass ratios are somewhat further from unity than they would otherwise be for the measured @xmath146 dex luminosity ratio .
the lyon and tucson models give consistent mass ratios of @xmath147 and @xmath148 , respectively .
this results in individual masses of @xmath149 and @xmath150 for the components of ( lyon ) , placing both below the substellar boundary ( @xmath151 ; * ? ? ?
* ) and likely below the lithium depletion boundary .
@xcite reported resolved rv measurements for and , combined with their astrometric orbit , found @xmath152 .
this is nominally inconsistent with our model - derived mass ratio at 1.6@xmath35 , with their value corresponding to the secondary being _ more _ massive than the primary . in figure
[ fig : lp - rv ] , we show their measurements along with the rv curve predicted for mass ratios of 0.87 and 2.0 , assuming their best - fit center - of - mass velocity of @xmath110 . fixing the orbital elements to the values from the astrometric orbit , we find @xmath62 values for the rv measurements that are unreasonably small for both values of @xmath140 ( @xmath153 for @xmath154 and @xmath155 for @xmath156 ; 6 dof ) .
this implies that the errors on the rv measurements are overestimated , and within these large errors both values of the mass ratio ( 0.87 and 2.0 ) are acceptable .
we discuss the plausibility of these different mass ratios in detail in section [ sec : lpab ] . for ,
the lyon and tucson models give consistent mass ratios of @xmath157 and @xmath158 , respectively .
this agrees with the spectroscopically determined mass ratio of 0.71@xmath159 from @xcite .
our slightly higher model - derived mass ratio results in somewhat more similar masses ( @xmath160 and @xmath161 , lyon ) than implied by the directly measured mass ratio ( @xmath162 and @xmath163 ) . in both cases however , is expected to lie near the mass - limit for lithium burning , which can be tested directly in the future with resolved optical spectroscopy of the doublet at 6708 .
in addition , absolute astrometric monitoring of the 2.4-yr orbital period binary can readily yield a refined measurement of the mass ratio : 1 mas astrometry would give a 5% mass ratio error , @xmath94@xmath49 better than spectroscopy .
effective temperature ( ) is one of the most difficult properties to measure directly , as it requires a direct measurement of the radius and luminosity . to date ,
radius measurements remain elusive for very low - mass stars and brown dwarfs in the field .
thus , when testing model predictions we are restricted to consistency checks between different methods of estimating . in section [ sec
: atm - fit ] , we estimated the temperatures of the primary components of our sample binaries by fitting their integrated - light spectra with atmospheric models .
the best - fit drift - phoenix and gaia - dusty model spectra are shown in figure [ fig : atm - fit ] plotted over the observed spectra .
evolutionary models provide a nearly independent temperature estimate by providing a prediction of the radius , which yields when combined with the measured .10% when comparing dustless to extremely dusty model atmospheres at low temperatures ( @xmath164 k ) , but the effect is @xmath69 at the warmer temperatures of late - m dwarfs ( see figure 2 of * ? ? ? * ) . ] for and , lyon evolutionary models give effective temperatures of @xmath165 k and @xmath166 k , respectively , and the tucson evolutionary models give 100 k warmer temperatures of @xmath167 k and @xmath168 k. both sets of models give consistent surface gravities of = 5.2 ( cgs ) for both components .
all nir atmospheric model fits gave systematically higher temperatures of 30003100 k and surface gravities near the predicted value ( @xmath169 dex ) .
given the nominal uncertainty ( i.e. , the model atmosphere grid step ) of 100 k in the best - fit values , the largest discrepancy of @xmath7150 k with the lyon evolutionary models is marginally significant . to provide an additional point of comparison for the estimates , we compiled all m6.5 and m7 dwarfs with effective temperatures derived by @xcite using the nearly model - independent infrared flux method @xcite .
we found a mean and standard deviation of @xmath170 k for these objects , which is lower than estimates from both classes of models by 140290 k , agreeing somewhat better with evolutionary models . for and , lyon models give effective temperatures of @xmath171 k and @xmath172 k , respectively , and the tucson models give 120 k warmer temperatures of @xmath173 k and @xmath174 k. lyon models predict surface gravities of about 4.9 ( cgs ) for both components , and tucson models predict 5.0 ( cgs ) . in comparison ,
atmospheric model fits of gave 28003000 k and surface gravities near the predicted value ( @xmath169 dex).0.4 given the known mass of .
such an unrealistic surface gravity is also found from the ames - dusty models for .
] again , the tucson evolutionary model is consistent with that derived from atmospheric models , while the lyon evolutionary models are lower by @xmath7150 k. for and , lyon models give effective temperatures of @xmath175 k and @xmath176 k , respectively , and the tucson models give 100 k warmer temperatures of @xmath177 k and @xmath178 k. lyon models predict surface gravities of about 5.2 ( cgs ) for both components , and tucson models predict 5.3 ( cgs ) . in comparison ,
atmospheric model fits of gave 27002900 k , with a wide range of surface gravities ( @xmath179 to 6.0 ) .
these values are much higher than evolutionary models by 150250 k , and the broad range of surface gravities is in reasonable agreement . in order to understand why temperature estimates from the two classes of models typically do not agree ,
we show our observed spectra in figure [ fig : atm - evol ] plotted with drift - phoenix and gaia - dusty model spectra that have and values closest to the lyon dusty evolutionary model - derived values .
this enables us to assess what spectral features prevented our fitting procedure from selecting these model spectra , which are very similar to the spectra used for the atmospheric boundary condition in the lyon models . ) .
this enables us to identify remaining problems and not dwell on issues that have already been resolved in the latest generation of that model grid .
] the nir sed fits do not seem to be significantly affected by broadband colors ( e.g. , the better fitting models do not necessarily have @xmath104 colors closer to what is observed ) . rather , the shapes of the spectra , which are sculpted by broad molecular absorption bands , seem to determine the best - fit models .
* the disagreement in shape is most pronounced at @xmath41 band , where the observed spectrum is much flatter than the low- , low - gravity model spectra that would better correspond to evolutionary model - derived properties .
this could be due partly to missing feh opacity in the models between @xmath91.5 and 1.7 @xcite . *
the observed @xmath42-band shape is also flatter ( i.e. , the band peak at 2.2 is more suppressed ) than in the low- , low - gravity model spectra . * in @xmath40 band , the fit seems to be driven by the deep h@xmath95o absorption band at 1.33 , which is too deep in the low- , low - gravity model spectra . * in @xmath180 band , the feh 0@xmath60 bandhead at 0.99 is too strong in the spectra corresponding to evolutionary model - derived properties .
this could alternatively be interpreted as the overall continuum at @xmath180 band being too high in the atmospheric models ( e.g. , missing opacity from wings of resonant alkali lines such as ) .
in general , we have found that atmospheric models predict systematically higher effective temperatures than evolutionary models .-
band spectra of mid- to late - m dwarfs .
most of their sample are young m dwarfs ( @xmath181 myr ) , but for their three field m7 to m9 dwarfs they find higher values than previously published estimates based on measured luminosities and estimated radii for objects of similar spectral type .
thus , their findings are consistent with our results . ]
this pattern is illustrated in figure [ fig : dteff ] , which shows the estimates from each class of models plotted against each other .
a systematic shift of @xmath7250 k would bring the models into agreement , with either the atmospheric model estimates being too warm or the evolutionary model estimates too cool ( i.e. , model radii too large by 1520% ) . ) give values @xmath7100 k more consistent with evolutionary models .
this indicates that optical diagnostics utilizing tio bandhead strengths ( e.g. , see * ? ? ?
* ) may be needed to provide more accurate estimates than can be obtained from near - infrared spectroscopy alone .
] we note that the observed discrepancy can not simply be due to the fact that our method relies on fitting the integrated - light spectra of binaries . as discussed in section [ sec : atm - fit ] , if the secondary component s flux impacted the best - fit model it would have the effect of lowering the derived , whereas we find temperatures that are too warm .
this means that the discrepancy could only be larger than we observe .
in addition , even a significant nonsolar metallicity for our binaries could not explain the observed temperature offset , as @xcite showed that models of varying metallicity gave the same best - fit to within the model grid step of 100 k for the ultracool subdwarf hd 114762b ( d / sdm9 ) .
we have considered what systematic errors could contribute to this discrepancy between evolutionary and atmospheric models .
currently available evolutionary models use much older versions of model atmospheres as their boundary condition , and this could potentially result in systematic errors in their output .
@xcite calculated evolutionary models both with and without the effects of dust clouds in the atmosphere and found at worst 6% differences in the resulting radii , which is much lower than what is needed to account for our observed discrepancy .
we can also estimate how other input assumptions impact evolutionary model output simply by comparing the predictions of the lyon and tucson models , which use different helium abundances , model atmospheres , and dust treatment .
these two sets of models predict radii different by 710% and temperatures different by @xmath7100 k for the objects in our sample more than a factor of two lower than what is needed to fully account for the observed discrepancy .
presently available model atmospheres are more advanced than evolutionary models , but as shown in figure [ fig : atm - fit ] they can not completely match our observed near - infrared spectra . therefore , the resulting temperature values must harbor systematic errors at some level .
thus , we conclude that the model atmospheres must be responsible for at least part of the observed 250 k discrepancy .
in fact , we suggest that the atmospheric models are more likely than the evolutionary models to be the primary source of the discrepancy , since roughly the same offset is observed over a wide range of masses , ages , and activity levels but the same temperature range .
this proposition will be readily testable with eclipsing binary measurements given the large implied radius difference between models ( 1520% ) .
we computed resolved photometry for our sample binaries in the mko photometric system to compare to lyon model predictions . for ,
we used our integrated - light mko photometry described in section [ sec : keck ] along with the best available flux ratios from table [ tbl : astrom ] . for and , we first converted their integrated - light 2mass photometry to the mko system by deriving correction terms from our near - infrared spectra
. the resulting component magnitudes for all three binaries are listed in table [ tbl : meas ] .
near - infrared colors on the mko photometric system were derived from the lyon models in the same fashion as other properties ( e.g. , ) , and the results are listed in tables [ tbl : lpmodel ] , [ tbl : lhsmodel ] , and [ tbl : glmodel ] . in figures [ fig : lp - jhk ] and [ fig : gl - jhk ] , we show the photometry of and compared to lyon mass tracks computed for the individual masses of both components .
the observed @xmath102 photometry is significantly discrepant with model tracks at all ages , with models typically predicting @xmath70.10.2 mag bluer @xmath36 colors than observed and @xmath70.10.2 mag redder @xmath88 colors . for
, the @xmath36 colors predicted by models are in much better agreement with the observations .
the most discrepant color is @xmath106 , which is only 0.07 mag ( 2.3@xmath35 ) redder than predicted , and this may be due to missing feh opacity in the @xmath41 band .
( we do not show a corresponding figure for as its components are main - sequence stars that evolve at essentially constant color . )
@xcite have recently described an approach to test models of ultracool dwarfs with dynamical mass measurements that is somewhat different from ours .
they use the measured individual luminosities along with estimates for each component to derive any other property ( e.g. , or age ) from evolutionary models .
( their estimates are derived by fitting model atmospheres to resolved broadband photometry of the binaries . )
this approach is similar to ours in that it uses two quantities to determine a third from evolutionary models , which is necessary because brown dwarfs occupy a massage ( or age ) relation instead of the simpler mass relation for stars on the main sequence . in @xcite , we considered in detail such model tests using different combinations of these fundamental parameters , and we ultimately chose to use mass and to derive other properties as these are the most directly measured quantities .
however , any discrepancies we have found in the models should also manifest themselves in the
approach that has been adopted by @xcite . in this section ,
we examine the results of analyzing our late - m dwarf measurements using the @xcite approach .
we first interpolated the lyon and tucson model grids at individual ( , ) points in the same fashion as @xcite , with our temperatures coming from the gaia - dusty atmospheric model fitting of the nir sed as described in section [ sec : atm - fit ] .
( this is the exact same model atmosphere grid as used by @xcite . ) in order to compute uncertainties in the model - derived properties using this approach , we used @xmath182 luminosity and temperature values randomly drawn from normal distributions corresponding to the measurement errors .
we assumed 100 k errors in , equal to the model atmosphere grid step .
the resulting evolutionary model - derived masses are plotted in figure [ fig : dmass ] against those derived by @xcite .
our masses are significantly higher than theirs by about a factor of @xmath92 to 4 because our estimates are systematically higher while their luminosity measurements are very similar to ours . in order for models to match our much higher temperatures at similar luminosities
, the masses must be larger and the ages somewhat older .
the best illustration of this disagreement is the case of 2mass j2206@xmath62047a , as @xcite used essentially identical luminosities and photometry as @xcite , and the component fluxes are nearly identical , removing any ambiguity in fitting its integrated - light spectrum .
@xcite found @xmath183 k from fitting its resolved broadband photometry , while our best - fit spectrum from the very same grid of model atmospheres has a temperature of @xmath184 k. therefore , despite the other similarities in our inputs , we found a primary mass of @xmath185 using the lyon dusty evolutionary models , while they found @xmath186 .
neither of these masses agrees with the measured mass ( @xmath187 , table 7 of @xcite ) : our mass is higher , and the @xcite mass is much lower .
the differences between our derived masses for and are even larger , as @xcite found extremely low masses for these objects ( @xmath188 and @xmath189 , respectively).20 reported by @xcite imply very young ages that are inconsistent with the lack of any low - gravity signatures in these objects spectra . ] fundamentally , both the @xcite approach and our standard method to test models provide similar information : a consistency check on the temperatures from atmospheric and evolutionary models .
the comparison presented here highlights how the amplitude , and even the sign , of discrepancies found between measured and model - predicted masses depend strongly on the input temperature estimates . using the exact same atmospheric model grids , but different methods , our two groups have arrived at values of different by @xmath7650 k. this is due to the fact that the fitting method of @xcite relies solely on the broadband colors of models , whereas our method is driven by the shape of specific spectral features . as shown in figure [ fig : lp - jhk ] and in our previous work @xcite
, the broadband colors of models consistently disagree with observations of objects of known mass over a broad range of spectral types .
we also note that a significant disadvantage of the @xcite approach is that evolutionary models occupy a very thin strip in this parameter space as opposed to @xmath190 .
this resulted in many of our randomly drawn ( , ) measurements falling outside the range predicted by models ( from 1.2% to 99.8% in the worst case ) , and we simply excluded these points from the analysis .
@xcite interpreted their observed discrepancies with theory as an error in the evolutionary model cooling curves , i.e. , a mass problem . using their method
, we find an opposite mass discrepancy because our estimates are @xmath7650 k higher than theirs , resulting in higher masses .
we suggest that the strong sensitivity of this model - testing method to the input estimates means that any observed discrepancies more readily identify problems in the way is determined from atmospheric models ( i.e. , a temperature problem ) , rather than a problem with evolutionary models .
it is somewhat surprising that has turned out to be a pair of young ( @xmath9140 myr ) brown dwarfs given the lack of detectable lithium @xcite and its m@xmath80 spectral classification @xcite . with a total mass of @xmath191 , one or both of the components may even lie below the theoretically predicted lithium depletion boundary at @xmath70.0550.065 .
in fact , even objects massive enough to deplete lithium take time to achieve the internal temperature necessary to do so ( @xmath192 k ; * ? ? ?
for example , @xcite show that a 0.070 object takes 220 myr to destroy 99% of its initial lithium , while at 145 myr only @xmath193 stars are similarly depleted . using the lyon and tucson models ,
we find predicted lithium fractions of @xmath194 and @xmath195 for , respectively , and @xmath196 and @xmath197 for .
thus , both sets of models predict that the primary may be depleted in lithium , but that the secondary should retain nearly all of its initial lithium .
the presence of lithium in the fainter secondary may be masked by the primary flux in the integrated - light spectroscopy of the system , so resolved spectroscopy to detect the doublet at 6708 is needed to test this prediction of brown dwarf evolutionary models directly .
evolutionary models make very precise predictions for the age of from its measured total mass and individual luminosities .
taken together , the lyon and tucson models predict an age of @xmath1 myr .
this is consistent with the age of 125 myr derived for the pleiades by @xcite using the lithium depletion boundary .
thus , we might expect to obey roughly the same boundaries in spectral type and color as found by @xcite .
the lithium boundary in the pleiades lies between spectral types of m6.5 and m7 , with objects @xmath198 having detectable lithium .
although the spectral type determinations for the components of are hampered by the lack of resolved spectroscopy , the measured integrated - light type of m@xmath80 makes it unlikely that is earlier than m7 , yet it is clearly depleted in lithium .
@xcite used spectroscopically derived @xmath199 colors to define the lithium depletion boundary in the pleiades , with all objects bluer than 2.20 mag being depleted in lithium . from the integrated - light optical spectrum of @xcite , we derived @xmath200 mag using the same method as @xcite . to be @xmath70.10 mag .
] at this color index , all brown dwarfs in the pleiades sample of @xcite show evidence for the presence of lithium .
lp 349 - 25a thus stands out as potentially anomalous compared to determination of the lithium boundary in the pleiades .
( note that our comparison is purely empirical in nature and does not depend on any theoretical assumptions regarding lithium depletion in brown dwarfs . )
this indicates that there may be a problem with the model - derived age for . at older ages , the lithium depletion boundary moves to lower masses ( i.e. , later spectral types ) ,
so an older age for would bring it into better agreement with the pleiades .
the age derived in our analysis is based solely on the model - predicted evolution , and for its actual age to be older either : ( 1 ) the measured parallax would have to be smaller ( i.e. , more distant ) ; or ( 2 ) the models would need to under - predict the luminosities of . in the former case , a parallax that is 3.5@xmath35 smaller ( placing the binary at @xmath201 pc ) would increase the total mass to 0.152 , resulting in a model - derived age @xmath202 myr and individual components predicted to be nearly fully depleted in lithium . in the latter case , the luminosities we measure would indeed be high for their mass , not because the objects are very young but rather because model evolution is not correct .
such an error has previously been suggested by @xcite to explain the higher - than - predicted luminosities of the benchmark brown dwarf binary hd 130948bc ( l4+l4 ) .
an independent measurement of the parallax for and resolved optical spectroscopy is needed to better assess this possible discrepancy with brown dwarfs in the pleiades .
the youth of could potentially be reflected in low - gravity spectral features such as have been noted for other field ultracool dwarfs ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the integrated - light optical spectrum from @xcite and our near - infrared spectrum do not show obvious hallmarks of low surface gravity
. however , near - infrared spectra of m7 and m8 dwarfs in the pleiades recently published by @xcite also show no distinguishing characteristics relative to field dwarfs of the same spectral type , at least at low resolution , @xmath203 , and modest @xmath97 .
we have also checked for possible association of with young moving groups using our derived @xmath112 heliocentric space velocity ( section [ sec : age ] ) .
we find no such association with any of the groups listed by and @xcite .
lp 349 - 25ab is one of the relatively few ultracool dwarfs that displays radio emission ( @xmath910% occurrence rate for @xmath101m7 dwarfs ; * ? ? ?
@xcite discovered the radio emission from at 8.5 ghz but with a beam size of @xmath204 were unable to resolve the binary to determine whether both components are radio - luminous .
@xcite presented additional unresolved , multi - frequency , multi - epoch observations to better constrain the emission mechanism , finding a lack of variability in both the radio flux and spectral index on both short ( hours ) and long ( months ) timescales .
their observations spanning 10.7 hours would have easily captured multiple rotations of the emitting component , as the measured @xmath205 of @xmath206 @xcite implies a maximum rotation period of @xmath207 hours ( assuming an average model - derived radius of 0.136 ) or a period of 2.6 hours if the rotation axis is co - aligned with the binary orbit .
is the only known radio - luminous ultracool dwarf that does not display some level of variability . to explain all observed radio properties
, @xcite favored a long - lived ( @xmath208 yr ) , high - latitude ( i.e. , polar ) source emitting gyrosynchrotron radiation .
future observations at higher angular resolution using the extended very large array are needed to determine whether only one or both components are radio - luminous .
our dynamical mass measurement will be a key input in developing customized models to better understand the origin of the magnetic structures generating the unique radio emission of .
finally , we consider the possibility that is actually a higher order multiple system .
the resolved colors and magnitudes of the two components are consistent with a simple binary as the secondary is slightly fainter and redder than the primary . as discussed in section [ sec : qratio ] , we find that the resolved radial velocity measurements of @xcite are consistent with both the model - derived mass ratio of 0.87 and their best fit value of 2.0 , in which the secondary has twice the mass of the primary . in this scenario ,
the b component would have to actually comprise two fainter brown dwarfs whose combined luminosity is below that of the single a component .
that can be ruled out is that the secondary component of is more luminous than the primary .
as shown in figure 1 of @xcite , a secondary with a putative mass of 0.04 would never outshine the primary , even during its deuterium - fusing phase at @xmath72.5 myr .
] however , in this case the a component s mass would only be about 0.04 , which is well below the lithium depletion limit and thus should display detectable lithium absorption .
in addition , dividing the binary into a higher order multiple would necessarily require the age to be younger for such lower mass component objects to output the same amount of energy . at younger ages
, it would be even less likely that the most massive component ( whatever its mass ) would have had time to reach the core temperature needed to destroy its initial lithium .
thus , we conclude that given all available constraints can not be a higher order multiple .
we have determined the orbits of three late - m binaries from keck ngs and lgs ao orbital monitoring . from observations spanning 5.9 years ( 75% ) of the 7.8 yr orbit of ,
we have determined a total mass of @xmath191 . for , we have found a highly eccentric ( @xmath209 ) 16.1-yr orbit from observations spanning 5.9 years ( 36% of the orbit , including periastron passage ) , deriving a dynamical mass of @xmath20 . we have also determined a new orbit for based solely on astrometrically well - calibrated instruments ( keck / nirc2 and _ hst_/stis ) , finding a larger semimajor axis and thus larger dynamical mass ( @xmath25 ) than previous work . for all binaries , the most significant contribution to their mass uncertainties are the errors in their parallaxes . despite s integrated - light spectral type of m8
_ and _ lack of detectable lithium absorption , this binary has turned out to be pair of brown dwarfs . to match the observed total mass and component luminosities ,
evolutionary models predict that the system must be quite young ( @xmath1 myr ) and that at least the secondary component should have retained most of its initial lithium .
if the model - derived age for is correct , it is nearly the same age as the pleiades ; however , it appears to be discrepant with the boundary between lithium - depleted and lithium - bearing objects from @xcite .
if the parallax measured by @xcite is accurate , this disagreement with the pleiades indicates a possible problem with the model - derived age , which in our analysis is based solely on the predicted evolution .
an older age for would remedy this problem , as the lithium depletion boundary moves to later spectral types at older ages .
this solution would require that models under - predict the luminosities of brown dwarfs , which has previously been suggested by @xcite to explain the higher - than - predicted luminosities of the benchmark brown dwarf binary hd 130948bc ( l4+l4 ) .
we also find that the lack of lithium depletion in at least the primary component effectively rules out the possibility that is a higher order multiple . as in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , we have used measured dynamical masses to test the predictions of evolutionary and atmospheric models .
properties such as and are derived from evolutionary models using only and the resolved component luminosities .
we have also fit model atmospheres to the integrated - light 0.952.42 spectra as an independent method for estimating .
including our previous work on 2mass j2206@xmath62047ab @xcite , there is now a sample of four late - m binaries with precise dynamical masses and near - infrared spectra enabling a consistency check between evolutionary model - derived properties and model atmospheres .
all four model atmosphere grids we tested give effective temperatures @xmath7250 k warmer than predicted for our objects of known mass from lyon dusty evolutionary model radii ( including the ames - dusty model atmospheres which were used as the boundary condition for the dusty evolutionary models ) .
although model atmospheres now incorporate considerably more advanced theory than the decade - old evolutionary models , we propose that such a large offset can not be entirely explained by problems with model radii . instead ,
model atmospheres are likely to be the major source of this discrepancy given that : ( 1 ) they must harbor systematic errors at some level , since they do not completely match our observed spectra ; and ( 2 ) roughly the same offset is observed over a narrow range of but a wide range of masses , ages , and activity levels
. such a discrepancy would be difficult to explain in the context of evolutionary models , although they may also harbor systematic errors that partially contribute to this inconsistency .
directly measured radii are needed to determine which set of models correctly predicts ( if either do ) .
this can be readily tested with future discoveries of late - m eclipsing binaries , as a 250 k offset corresponds to a substantial radius difference ( 1520% ) .
in addition to our standard line of analysis , we have also explored the approach to testing models used by @xcite for the late - m binaries common to our two samples .
our model atmosphere fitting of integrated - light spectra yields higher estimates than their approach of fitting resolved broadband photometry .
consequently , we find evolutionary model - derived masses from their approach that are a factor of @xmath92 to 4 higher than theirs , resulting in an opposite discrepancy with measured dynamical masses .
this illustrates that their method sensitively relies upon the input , a property that can not be determined directly without radius measurements .
thus , this approach may have limited utility in testing evolutionary models , as observed discrepancies are more readily caused by problems with estimates from model atmospheres . to summarize how our conclusions contrast with those of @xcite : ( 1 ) they found estimates for from model atmospheres that are @xmath7400 k _ lower _ than our evolutionary model - derived values , whereas we find estimates that are @xmath7250 k _
higher_. ( 2 ) they interpreted these results in terms of mass , concluding that systematic errors in the evolutionary model cooling curves were responsible for observed discrepancies .
we propose instead that the discrepancy is largely caused by systematic errors in the model atmospheres .
as the sample of ultracool dwarfs with dynamical mass measurements grows , we are beginning to probe the diversity of the field population , which comprises both very low - mass stars and brown dwarfs . at the cool temperatures of late - m dwarfs
, field objects can span nearly a factor a two in mass depending on their age , which makes mass measurements crucial for breaking this degeneracy . at cooler temperatures
there are even fewer mass measurements presently available , and the effects of complex processes such as atmospheric dust formation become more important .
therefore , continued orbital monitoring will be imperative to bring similar tests of models as we report here into the much cooler regimes of the l and t dwarfs .
it is a pleasure to thank joel aycock , randy campbell , al conrad , heather hershley , jim lyke , jason mcilroy , gary punawai , julie riviera , hien tran , cynthia wilburn , and the keck observatory staff for assistance with the keck observations .
we are very grateful to michal simon and chad bender for providing us with their keck / nirc2 images of .
we also thank france allard for providing evolutionary models with near - infrared photometry on the mko system , travis barman for providing us with the gaia - dusty model spectra , kelle cruz for providing the optical spectrum of , adam burgasser for providing the spex prism spectrum of gl 569a , brian cameron for making available his nirc2 distortion solution , and cline reyl for customized besanon galaxy models .
we are indebted to katelyn allers for assistance in obtaining irtf / spex data of .
our research has employed the 2mass data products ; nasa s astrophysical data system ; the simbad database operated at cds , strasbourg , france ; and the m , l , and t dwarf compendium housed at ` http://www.dwarfarchives.org ` and maintained by chris gelino , davy kirkpatrick , and adam burgasser @xcite .
mcl , tjd , and bpb acknowledge support for this work from nsf grants ast-0507833 and ast-0909222 .
finally , the authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community .
we are most fortunate to have the opportunity to conduct observations from this mountain .
asplund , m. , grevesse , n. , & sauval , a. j. 2005 , in astronomical society of the pacific conference series , vol .
336 , cosmic abundances as records of stellar evolution and nucleosynthesis , ed .
t. g. barnes iii & f. n. bash , 25 bernath , p. 2006 , in american institute of physics conference series , vol .
855 , astrochemistry - from laboratory studies to astronomical observations , ed .
r. i. kaiser , p. bernath , y. osamura , s. petrie , & a. m. mebel , 143148 perryman , m. a. c. , & esa , eds .
1997 , esa special publication , vol .
1200 , the hipparcos and tycho catalogues .
astrometric and photometric star catalogues derived from the esa hipparcos space astrometry mission van leeuwen , f. 2007 , hipparcos , the new reduction of the raw data ( hipparcos , the new reduction of the raw data . by floor van leeuwen , institute of astronomy , cambridge university , cambridge , uk series : astrophysics and space science library , vol . 350 20 springer dordrecht ) )
measurements with observational error bars .
[ fig : orbit - sep - pa],title="fig:",width=163 ] 0.4 in ) measurements with observational error bars .
[ fig : orbit - sep - pa],title="fig:",width=163 ] 0.4 in ) measurements with observational error bars .
[ fig : orbit - sep - pa],title="fig:",width=163 ] saved steps from our mcmc chains of length @xmath210 . for and
the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions .
[ fig : pdist],title="fig:",width=259 ] 0.1 in saved steps from our mcmc chains of length @xmath210 . for and
the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions .
[ fig : pdist],title="fig:",width=259 ] saved steps from our mcmc chains of length @xmath210 . for and the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions . [
fig : pdist],title="fig:",width=259 ] 0.1 in saved steps from our mcmc chains of length @xmath210 . for and
the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions .
[ fig : pdist],title="fig:",width=259 ] saved steps from our mcmc chains of length @xmath210 . for and the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions .
[ fig : pdist],title="fig:",width=259 ] 0.1 in saved steps from our mcmc chains of length @xmath210 . for and
the correlation between @xmath211 and @xmath212 enables the total mass to be determined more precisely than from simple propagation of errors , as illustrated by lines drawn demarcating the 3@xmath35 range for the total mass ( prior accounting for the distance uncertainty ) .
the double - peaked nature of the mcmc parameter distributions for and also reveal slightly degenerate orbit solutions .
[ fig : pdist],title="fig:",width=259 ] values from model atmosphere fits to our sample binaries observed integrated - light 0.952.42 spectra ( drawn at 1.02 , 1.1 , 1.3 , 1.5 , 1.7 , 2.0 , 2.5 , and 3.5 times the minimum @xmath62 ) .
is generally better constrained than , as indicated by contours that are elongated in the direction .
white space shows where models are not present in our grids . for comparison ,
the gold stars show lyon dusty evolutionary model - derived properties , with in general agreement but systematically lower than the best - fit model atmospheres ( see section [ sec : teff - logg ] ) .
[ fig : chi2],height=576 ] and 2@xmath35 ranges in the model - derived masses , which increase as a function of age . by applying the independent constraint of the measured total masses
, we used lyon models ( left panel ) and tucson models ( right panel ) to derive ages for these binaries ( see tables [ tbl : lpmodel ] and [ tbl : glmodel ] ) .
the horizontal gray bars show our 1@xmath35 and 2@xmath35 constraints on the total mass , and the resulting median , 1@xmath35 , and 2@xmath35 model - inferred ages are shown by solid , dashed , and dotted lines , respectively . because has a lower mass than ( and higher luminosities ) , its derived age is significantly younger than for .
[ fig : mtotage],width=336 ] and 2@xmath35 ranges in the model - derived masses , which increase as a function of age . by applying the independent constraint of the measured total masses , we used lyon models ( left panel ) and tucson models ( right panel ) to derive ages for these binaries ( see tables [ tbl : lpmodel ] and [ tbl : glmodel ] ) .
the horizontal gray bars show our 1@xmath35 and 2@xmath35 constraints on the total mass , and the resulting median , 1@xmath35 , and 2@xmath35 model - inferred ages are shown by solid , dashed , and dotted lines , respectively .
because has a lower mass than ( and higher luminosities ) , its derived age is significantly younger than for .
[ fig : mtotage],width=336 ] ) . combining our measured luminosity ratio and total mass ,
evolutionary models predict @xmath156 , whereas @xcite found @xmath154 from the best fit to their data .
their 13 errors result in an unrealistically low @xmath62 ( 0.5 for 6 dof ) , suggesting that the errors are overestimated . within these large errors , we find that both mass ratios are allowed by their data .
[ fig : lp - rv],width=307 ] 0.003 for display only .
the most prominent deviations in the model spectra are : the exaggerated shapes of @xmath41 and @xmath42 bands ; the overpredicted depth of h@xmath95o absorption at 1.33 ; and the strength of the feh bandhead at 0.99 . without directly measured radii
we can not determine if the problem lies in the model spectra , the evolutionary model radii , or both .
[ fig : atm - evol],width=595 ] 0.003 for display only .
the most prominent deviations in the model spectra are : the exaggerated shapes of @xmath41 and @xmath42 bands ; the overpredicted depth of h@xmath95o absorption at 1.33 ; and the strength of the feh bandhead at 0.99 . without
directly measured radii we can not determine if the problem lies in the model spectra , the evolutionary model radii , or both .
[ fig : atm - evol],width=595 ] mass uncertainties , and open squares demarcate ages of 0.1 , 0.2 , and 0.3 gyr .
field dwarfs with parallax measurements are shown for comparison as filled gray circles .
lp 349 - 25a and b are somewhat redder ( 0.10.2 mag ) than evolutionary model tracks for @xmath36 colors , while their @xmath88 colors are 0.10.2 mag bluer than models .
[ fig : lp - jhk],width=624 ] 2 to 4 ( left ) because our model atmosphere fitting of integrated - light spectra yielded @xmath7650 k higher temperatures than their fitting of resolved broadband photometry ( right ) .
we suggest that the sensitivity of this model testing approach to the input estimates makes it more likely to identify problems with model atmospheres than evolutionary models .
[ fig : dmass],title="fig:",width=307 ] 2 to 4 ( left ) because our model atmosphere fitting of integrated - light spectra yielded @xmath7650 k higher temperatures than their fitting of resolved broadband photometry ( right ) .
we suggest that the sensitivity of this model testing approach to the input estimates makes it more likely to identify problems with model atmospheres than evolutionary models .
[ fig : dmass],title="fig:",width=307 ] 2008 jan 16 & 05:56 & 1.280 & & @xmath213 & @xmath214 + 2008 jun 30 & 13:20 & 1.314 & @xmath40 & @xmath215 & @xmath216 + & 13:29 & 1.274 & @xmath41 & @xmath217 & @xmath218 + & 13:12 & 1.350 & & @xmath219 & @xmath220 + 2008 aug 20 & 14:21 & 1.069 & & @xmath221 & @xmath222 + 2008 sep 9 & 13:51 & 1.169 & & @xmath223 & @xmath224 + 2009 sep 28 & 13:17 & 1.328 & & @xmath225 & @xmath226 + 2009 dec 15 & 05:22 & 1.479 & @xmath42 & @xmath227 & @xmath228 + 2010 may 22 & 14:54 & 1.721 & @xmath42 & @xmath229 & @xmath230 + 2008 jan 15 & 11:05 & 1.128 & & @xmath231 & @xmath232 + 2008 sep 9 & 15:13 & 1.411 & @xmath40 & @xmath233 & @xmath234 + & 14:59 & 1.475 & @xmath41 & @xmath235 & @xmath236 + & 15:17 & 1.393 & & @xmath237 & @xmath238 + 2009 sep 28 & 15:37 & 1.150 & & @xmath239 & @xmath240 + 2009 dec 16 & 08:45 & 1.428 & & @xmath241 & @xmath242 + 2010 mar 22 & 05:59 & 1.096 & @xmath40 & @xmath243 & @xmath244 + & 05:55 & 1.094 & @xmath41 & @xmath245 & @xmath246 + & 05:52 & 1.093 & @xmath42 & @xmath247 & @xmath248 + & 06:04 & 1.097 & & @xmath249 & @xmath250 + 2004 dec 24 & 16:22 & 1.270 & @xmath41 & @xmath251 & @xmath252 + 2005 feb 25 & 12:08 & 1.298 & & @xmath253 & @xmath246 + 2008 jan 16 & 16:33 & 1.036 & & @xmath254 & @xmath255 + 2009 apr 29 & 12:24 & 1.089 & & @xmath256 & @xmath257 + 2009 may 29 & 10:56 & 1.159 & & @xmath258 & @xmath259 + 2010 mar 22 & 15:48 & 1.235 & @xmath42 & @xmath260 & @xmath261 + 2010 may 23 & 11:10 & 1.133 & @xmath40 & @xmath262 & @xmath263 + & 11:16 & 1.148 & @xmath41 & @xmath264 & @xmath265 + & 11:23 & 1.166 & @xmath42 & @xmath266 & @xmath267 + & 11:32 & 1.193 & & @xmath268 & @xmath269 + 2004 jul 3 & cfht / pueo & & @xmath270 & @xmath271 & @xmath272 + 2004 sep 26 & vlt / naco & @xmath41 & @xmath60 & @xmath273 & @xmath59 + 2006 jul 13 & vlt / naco & & @xmath274 & @xmath275 & @xmath228 + 2006 sep 21 & vlt / naco & & @xmath276 & @xmath277 & @xmath278 + 2006 oct 9 & vlt / naco & & @xmath279 & @xmath280 & @xmath281 + 2006 nov 11 & vlt / naco & & @xmath282 & @xmath283 & @xmath284 + 2006 dec 24 & vlt / naco & & @xmath285 & @xmath286 & @xmath287 + 2008 jan 16 & keck / nirc2 & & @xmath288 & @xmath289 & @xmath290 + 2008 jun 30 & keck / nirc2 & @xmath40 & @xmath291 & @xmath292 & @xmath293 + & keck / nirc2 & @xmath41 & @xmath294 & @xmath295 & @xmath296 + & keck / nirc2 & & @xmath297 & @xmath298 & @xmath299 + 2008 aug 20 & keck / nirc2 & & @xmath300 & @xmath301 & @xmath302 + 2008 sep 9 & keck / nirc2 & & @xmath303 & @xmath304 & @xmath305 + 2009 sep 28 & keck / nirc2 & & @xmath306 & @xmath307 & @xmath308 + 2009 dec 15 & keck / nirc2 & @xmath42 & @xmath309 & @xmath310 & @xmath311 + 2010 may 22 & keck / nirc2 & @xmath42 & @xmath312 & @xmath313 & @xmath314 + 2004 jan 8 & cfht / pueo & & @xmath315 & @xmath316 & @xmath317 + 2005 apr 27 & cfht / pueo & & @xmath318 & @xmath319 & @xmath320 + 2005 oct 14 & cfht / pueo & @xmath41 & @xmath321 & @xmath322 & @xmath323 + 2008 jan 15 & keck / nirc2 & & @xmath324 & @xmath325 & @xmath326 + 2008 sep 9 & keck / nirc2 & @xmath40 & @xmath327 & @xmath328 & @xmath329 + & keck / nirc2 & @xmath41 & @xmath330 & @xmath331 & @xmath332 + & keck / nirc2 & & @xmath333 & @xmath334 & @xmath335 + 2009 sep 28 & keck / nirc2 & & @xmath336 & @xmath337 & @xmath338 + 2009 dec 16 & keck / nirc2 & & @xmath339 & @xmath340 & @xmath341 + 2010 mar 22 & keck / nirc2 & @xmath40 & @xmath342 & @xmath343 & @xmath344 + & keck / nirc2 & @xmath41 & @xmath342 & @xmath345 & @xmath346 + & keck / nirc2 & @xmath42 & @xmath347 & @xmath348 & @xmath349 + & keck / nirc2 & & @xmath350 & @xmath351 & @xmath352 + 2002 jun 26 & _ hst_/stis&f28x50lp&@xmath51 & @xmath353 & @xmath53 + 2004 dec 24 & keck / nirc2 & @xmath41 & @xmath354 & @xmath355 & @xmath356 + 2005 feb 25 & keck / nirc2 & & @xmath357 & @xmath358 & @xmath359 + 2008 jan 16 & keck / nirc2 & & @xmath360 & @xmath361 & @xmath362 + 2009 apr 29 & keck / nirc2 & & @xmath363 & @xmath364 & @xmath365 + 2009 may 29 & keck / nirc2 & & @xmath366 & @xmath367 & @xmath368 + 2010 mar 22 & keck / nirc2 & @xmath42 & @xmath369 & @xmath370 & @xmath371 + 2010 may 23 & keck / nirc2 & @xmath40 & @xmath372 & @xmath373 & @xmath374 + & keck / nirc2 & @xmath41 & @xmath375 & @xmath376 & @xmath377 + & keck / nirc2 & @xmath42 & @xmath378 & @xmath379 & @xmath380 + & keck / nirc2 & & @xmath381 & @xmath382 & @xmath383 + + semimajor axis @xmath212 ( mas ) & 146.7 & @xmath60.6 , 0.6 & @xmath61.1 , 1.2 & @xmath384 + orbital period @xmath211 ( yr ) & 7.76 & @xmath60.04 , 0.04 & @xmath60.07 , 0.08 & @xmath385 + eccentricity @xmath386 & 0.051 & @xmath60.003 , 0.003 & @xmath60.005 , 0.006 & @xmath387 + inclination @xmath388 ( ) & 117.24 & @xmath60.14 , 0.14 & @xmath60.3 , 0.3 & @xmath389 + time of periastron passage @xmath390 ( jd ) & 0 & @xmath624 , 26 & @xmath650 , 50 & @xmath391 + pa of the ascending node @xmath79 ( ) & 35.95 & @xmath60.12 , 0.12 & @xmath60.23 , 0.23 & @xmath392 + argument of periastron @xmath78 ( ) & 250 & @xmath63 , 4 & @xmath66 , 7 & @xmath393 + total mass ( ): fitted & 0.1205 & @xmath60.0007 , 0.0007 & @xmath60.0013 , 0.0013 & @xmath394 + total mass ( ): final & 0.120 & @xmath60.007 , 0.008 & @xmath60.014 , 0.017 & @xmath395 + @xmath62 ( 21 degrees of freedom ) & & & & 18.1 + semimajor axis @xmath212 ( mas ) & 288 & @xmath65 , 5 & @xmath68 , 8 & @xmath396 + orbital period @xmath211 ( yr ) & 16.1 & @xmath60.5 , 0.5 & @xmath60.8 , 0.8 & @xmath397 + eccentricity @xmath386 & 0.830 & @xmath60.005 , 0.005 & @xmath60.008 , 0.008 & @xmath398 + inclination @xmath388 ( ) & 72.1 & @xmath60.3 , 0.3 & @xmath60.5 , 0.5 & @xmath399 + time of periastron passage @xmath400 ( jd ) & 0.0 & @xmath62.5 , 2.5 & @xmath64 , 4 & @xmath401 + pa of the ascending node @xmath79 ( ) & 182.0 & @xmath60.2 , 0.2 & @xmath60.4 , 0.4 & @xmath402 + argument of periastron @xmath78 ( ) & 224.8 & @xmath60.6 , 0.6 & @xmath61.1 , 1.1 & @xmath403 + total mass ( ): fitted & 0.1944 & @xmath60.0028 , 0.0028 & @xmath60.005 , 0.005 & @xmath404 + total mass ( ): final & 0.194 & @xmath60.021 , 0.025 & @xmath60.039 , 0.053 & @xmath405 + @xmath62 ( 9 degrees of freedom ) & & & & 6.8 + semimajor axis @xmath212 ( mas ) & 95.6 & @xmath61.0 , 1.1 & @xmath61.8 , 2.0 & @xmath406 + orbital period @xmath211 ( yr ) & 2.367 & @xmath60.003 , 0.003 & @xmath60.006 , 0.006 & @xmath407 + eccentricity @xmath386 & 0.316 & @xmath60.005 , 0.005 & @xmath60.009 , 0.009 & @xmath408 + inclination @xmath388 ( ) & 35.0 & @xmath61.1 , 1.1 & @xmath62.2 , 2.0 & @xmath409 + time of periastron passage @xmath410 ( jd ) & 0.0 & @xmath61.5 , 1.5 & @xmath62.5 , 2.5 & @xmath411 + pa of the ascending node @xmath79 ( ) & 324.8 & @xmath62.0 , 2.0 & @xmath63.3 , 3.3 & @xmath412 + argument of periastron @xmath78 ( ) & 257.9 & @xmath62.0 , 2.0 & @xmath63.0 , 3.3 & @xmath413 + total mass ( ): fitted & 0.140 & @xmath60.004 , 0.005 & @xmath60.007 , 0.009 & @xmath414 + total mass ( ): final & 0.140 & @xmath60.008 , 0.009 & @xmath60.015 , 0.018 & @xmath415 + @xmath62 ( 9 degrees of freedom ) & & & & 6.3 + ( ) & & 1,2 & & & 1,3 & & & 1,4 + semimajor axis ( au ) & & 1,2 & & & 1,3 & & & 1,4 + @xmath416 ( pc ) & & 2 & & & 3 & & & 4 + spectral type & @xmath417 & @xmath418 & 1 & & @xmath419 & @xmath419 & 1 & & @xmath420 & @xmath421 & 5,6 + @xmath40 ( mag ) & @xmath422 & @xmath423 & 1,7 & & @xmath424&@xmath425 & 1,7 & & @xmath426&@xmath427 & 1 + @xmath41 ( mag ) & @xmath428 & @xmath429 & 1,7 & & @xmath430&@xmath431 & 1,7 & & @xmath432&@xmath433 & 1 + @xmath42 ( mag ) & @xmath434 & @xmath435 & 1,7 & & @xmath436&@xmath437 & 1,7 & & @xmath438&@xmath439 & 1 + ( mag ) & @xmath440 & @xmath441 & 1,8 & & & & & & @xmath442&@xmath443 & 1,9 + @xmath104 ( mag ) & @xmath444 & @xmath445 & 1,7 & & @xmath446&@xmath447 & 1,7 & & @xmath448&@xmath449 & 1 + @xmath106 ( mag ) & @xmath450 & @xmath451 & 1,7 & & @xmath452&@xmath453 & 1,7 & & @xmath454&@xmath455 & 1 + @xmath107 ( mag ) & @xmath456 & @xmath457 & 1,7 & & @xmath458&@xmath459 & 1,7 & & @xmath460&@xmath461 & 1 + @xmath88 ( mag ) & @xmath462 & @xmath463 & 1,7,8 & & & & & & @xmath464&@xmath465 & 1,9 + @xmath466 ( mag ) & @xmath467 & @xmath468 & 1,2,7 & & @xmath469&@xmath470 & 1,3,7 & & @xmath471&@xmath472 & 1,4 + @xmath473 ( mag ) & @xmath474 & @xmath475 & 1,2,7 & & @xmath476&@xmath477 & 1,3,7 & & @xmath478&@xmath479 & 1,4 + @xmath480 ( mag ) & @xmath481 & @xmath482 & 1,2,7 & & @xmath483&@xmath484 & 1,3,7 & & @xmath485&@xmath486 & 1,4 + @xmath487 ( mag ) & @xmath488 & @xmath489 & 1,2,8 & & & & & & @xmath490&@xmath491 & 1,4,9 + @xmath492(/ ) & @xmath493 & @xmath494 & 1 & & @xmath495&@xmath496 & 1 & & @xmath497&@xmath498 & 1 + @xmath499 ( ) & & 1 & & & 1 & & & 1
+ phoenix - gaia & bh05 & & ames & ames & da99 & f99 & gn93 & dust disappears after forming ( `` cond '' ) + ames - dusty & a01 & & ames & ames & pd93 & f99 & gn93 & dust stays where it forms ( `` dusty '' ) + gaia - dusty & b10 & & bt06 & ames & d03 & b02 & a05 & dust stays where it forms ( `` dusty '' ) + drift - phoenix & w09 & & ames & ames & b06 & b06 & gn93 & nonequilibrium cloud model ( `` drift '' ) + + & * 2800 * & * 4.5 * & & * 2800 * & * 5.5 * & & * 2800 * & * 5.0 * & & * 2800 * & * 5.5 * + all ( 0.812.42 ) & 2600 & 4.0 & & 2600 & 4.0 & & 2600 & 4.0 & & 2600 & 4.0 + @xmath180 ( 0.951.12 ) & 2800 & 4.5 & & 2800 & 5.5 & & 2800 & 4.5 & & 2700 & 5.0 + @xmath40 ( 1.101.34 ) & 2900 & 5.5 & & 2900 & 5.5 & & 2900 & 5.5 & & 2800 & 5.5 + @xmath41 ( 1.401.80 ) & 2900 & 5.5 & & 2900 & 5.5 & & 2900 & 5.5 & & 2800 & 6.0 + @xmath42 ( 1.902.40 ) & 3000 & 5.5 & & 3000 & 5.5 & & 2900 & 5.5 & & 2900 & 6.0 + + + & * 3000 * & * 5.0 * & & * 3000 * & * 5.0 * & & * 3000 * & * 5.0 * & & * 3000 * & * 5.0 * + all ( 0.812.42 ) & 2900 & 4.5 & & 2900 & 4.0 & & 2900 & 4.5 & & 2900 & 5.0 + @xmath180 ( 0.951.12 ) & 3100 & 4.5 & & 3000 & 5.0 & & 3000 & 4.5 & & 3000 & 5.0 + @xmath40 ( 1.101.34 ) & 3000 & 5.5 & & 3000 & 5.5 & & 3000 & 5.5 & & 3000 & 5.5 + @xmath41 ( 1.401.80 ) & 3000 & 5.5 & & 3000 & 5.5 & & 3000 & 5.5 & & 3000 & 6.0 + @xmath42 ( 1.902.40 ) & 3100 & 5.5 & & 3100 & 5.5 & & 3100 & 5.5 & & 3000 & 5.5 + + + & * 2800 * & * 4.5 * & & * 2800 * & * 5.5 * & & * 2800 * & * 5.0 * & & * 2800 * & * 5.5 * + all ( 0.812.42 ) & 2700 & 4.5 & & 2800 & 5.5 & & 2700 & 5.0 & & 2700 & 5.0 + @xmath180 ( 0.951.12 ) & 2900 & 5.0 & & 2900 & 6.0 & & 2800 & 5.0 & & 2700 & 5.5 + @xmath40 ( 1.101.34 ) & 2800 & 5.5 & & 2700 & 5.5 & & 2700 & 5.5 & & 2700 & 5.5 + @xmath41 ( 1.401.80 ) & 2900 & 5.5 & & 2800 & 6.0 & & 2800 & 6.0 & & 2800 & 6.0 + @xmath42 ( 1.902.40 ) & 2900 & 5.5 & & 2900 & 5.5 & & 2800 & 5.5 & & 2800 & 6.0 + + + & * 2900 * & * 5.0 * & & * 2900 * & * 6.0 * & & * 2900 * & * 5.5 * & & * 2900 * & * 5.5 * + all ( 0.812.42 ) & 2800 & 4.5 & & 2800 & 6.0 & & 2700 & 5.0 & & 2800 & 5.5 + @xmath180 ( 0.951.12 ) & 2900 & 5.0 & & 2900 & 6.0 & & 2800 & 5.0 & & 2700 & 5.5 + @xmath40 ( 1.101.34 ) & 2900 & 5.5 & & 2800 & 5.5 & & 2800 & 5.5 & & 3000 & 6.0 + @xmath41 ( 1.401.80 ) & 2900 & 5.5 & & 2900 & 6.0 & & 2900 & 5.5 & & 2900 & 6.0 + @xmath42 ( 1.902.40 ) & 3000 & 5.5 & & 3000 & 5.5 & & 2900 & 5.5 & & 2900 & 6.0 + + @xmath507 ( ) & 0.065 & @xmath508 & @xmath509 + @xmath510 ( k ) & 2780 & @xmath511 & @xmath512 + @xmath513 ( cgs ) & 5.02 & @xmath514 & @xmath515 + @xmath516 ( ) & 0.130 & @xmath517 & @xmath518 + li@xmath519/li@xmath520 & 0.75 & @xmath521 & @xmath522 + + @xmath523 ( ) & 0.056 & @xmath524 & @xmath525 + @xmath526 ( k ) & 2640 & @xmath511 & @xmath527 + @xmath528 ( cgs ) & 5.00 & @xmath514 & @xmath515 + @xmath529 ( ) & 0.124 & @xmath517 & @xmath530 + li@xmath531/li@xmath520 & 0.96 & @xmath532 & @xmath533 + + + + age ( gyr ) & 0.127 & @xmath534 & @xmath535 + @xmath140 ( @xmath142 ) & 0.872 & @xmath536 & @xmath503 + @xmath504 ( k ) & 134 & @xmath537 & @xmath506 + + @xmath507 ( ) & 0.064 & @xmath538 & @xmath539 + @xmath510 ( k ) & 2660 & @xmath511 & @xmath512 + @xmath513 ( cgs ) & 4.93 & @xmath514 & @xmath540 + @xmath516 ( ) & 0.144 & @xmath517 & @xmath541 + li@xmath519/li@xmath520 & 0.68 & @xmath542 & @xmath543 + @xmath544 ( mag ) & 0.795 & @xmath545 & @xmath524 + @xmath546 ( mag ) & 0.375 & @xmath547 & @xmath548 + @xmath549 ( mag ) & 0.420 & @xmath550 & @xmath509 + @xmath551 ( mag ) & 0.544 & @xmath552 & @xmath553 + + @xmath523 ( ) & 0.056 & @xmath508 & @xmath554 + @xmath526 ( k ) & 2520 & @xmath511 & @xmath527 + @xmath528 ( cgs ) & 4.91 & @xmath514 & @xmath515 + @xmath529 ( ) & 0.137 & @xmath517 & @xmath518 + li@xmath531/li@xmath520 & 0.95 & @xmath555 & @xmath556 + @xmath557 ( mag ) & 0.810 & @xmath518 & @xmath558 + @xmath559 ( mag ) & 0.396 & @xmath560 & @xmath561 + @xmath562 ( mag ) & 0.415 & @xmath538 & @xmath554 + @xmath563 ( mag ) & 0.594 & @xmath564 & @xmath565 + lccc + + age ( gyr ) & 0.37 & @xmath566 & @xmath567 + @xmath140 ( @xmath142 ) & 0.966 & @xmath568 & @xmath569 + @xmath504 ( k ) & 35 & @xmath570 & @xmath571 + + @xmath507 ( ) & 0.098 & @xmath572 & @xmath573 + @xmath510 ( k ) & 2960 & @xmath511 & @xmath574 + @xmath513 ( cgs ) & 5.22 & @xmath575 & @xmath576 + @xmath516 ( ) & 0.128 & @xmath508 & @xmath539 + + @xmath523 ( ) & 0.095 & @xmath577 & @xmath578 + @xmath526 ( k ) & 2930 & @xmath579 & @xmath580 + @xmath528 ( cgs ) & 5.23 & @xmath532 & @xmath581 + @xmath529 ( ) & 0.124 & @xmath524 & @xmath582 + + + + age ( gyr ) & 0.28 & @xmath583 & @xmath584 + @xmath140 ( @xmath142 ) & 0.958 & @xmath585 & @xmath586 + @xmath504 ( k ) & 39 & @xmath587 & @xmath588 + + @xmath507 ( ) & 0.099 & @xmath589 & @xmath590 + @xmath510 ( k ) & 2860 & @xmath512 & @xmath591 + @xmath513 ( cgs ) & 5.15 & @xmath592 & @xmath593 + @xmath516 ( ) & 0.138 & @xmath518 & @xmath594 + @xmath544 ( mag ) & 0.787 & @xmath595 & @xmath545 + @xmath546 ( mag ) & 0.336 & @xmath596 & @xmath597 + @xmath549 ( mag ) & 0.451 & @xmath552 & @xmath598 + + @xmath523 ( ) & 0.094 & @xmath599 & @xmath600 + @xmath526 ( k ) & 2820 & @xmath506 & @xmath601 + @xmath528 ( cgs ) & 5.15 & @xmath602 & @xmath603 + @xmath529 ( ) & 0.135 & @xmath518 & @xmath594 + @xmath557 ( mag ) & 0.788 & @xmath595 & @xmath545 + @xmath559 ( mag ) & 0.341 & @xmath604 & @xmath605 + @xmath562 ( mag ) & 0.447 & @xmath606 & @xmath607 + lccc + + age ( gyr ) & 0.51 & @xmath608 & @xmath609 + @xmath140 ( @xmath610 ) & 0.886 & @xmath534 & @xmath611 + @xmath504 ( k ) & 229 & @xmath612 & @xmath613 + + @xmath614 ( ) & 0.074 & @xmath508 & @xmath554 + @xmath615 ( k ) & 2530 & @xmath511 & @xmath512 + @xmath616 ( cgs ) & 5.30 & @xmath503 & @xmath617 + @xmath618 ( ) & 0.101 & @xmath545 & @xmath517 + li@xmath619/li@xmath520 & 0.00 & @xmath620 & @xmath620 + + @xmath621 ( ) & 0.066 & @xmath538 & @xmath622 + @xmath623 ( k ) & 2300 & @xmath511 & @xmath527 + @xmath624 ( cgs ) & 5.28 & @xmath625 & @xmath626 + @xmath627 ( ) & 0.097 & @xmath545 & @xmath508 + li@xmath628/li@xmath520 & 0.02 & @xmath629 & @xmath630 + + + + age ( gyr ) & 0.46 & @xmath631 & @xmath632 + @xmath140 ( @xmath610 ) & 0.866 & @xmath605 & @xmath611 + @xmath504 ( k ) & 221 & @xmath633 & @xmath634 + + @xmath614 ( ) & 0.075 & @xmath508 & @xmath509 + @xmath615 ( k ) & 2430 & @xmath511 & @xmath512 + @xmath616 ( cgs ) & 5.23 & @xmath514 & @xmath635 + @xmath618 ( ) & 0.110 & @xmath545 & @xmath636 + li@xmath619/li@xmath520 & 0.00 & @xmath620 & @xmath620 + @xmath637 ( mag ) & 0.827 & @xmath638 & @xmath639 + @xmath640 ( mag ) & 0.371 & @xmath641 & @xmath642 + @xmath643 ( mag ) & 0.456 & @xmath508 & @xmath606 + @xmath644 ( mag ) & 0.660 & @xmath645 & @xmath646 + + @xmath621 ( ) & 0.065 & @xmath538 & @xmath622 + @xmath623 ( k ) & 2210 & @xmath511 & @xmath527 + @xmath624 ( cgs ) & 5.20 & @xmath625 & @xmath647 + @xmath627 ( ) & 0.106 & @xmath648 & @xmath550 + li@xmath628/li@xmath520 & 0.00 & @xmath649 & @xmath650 + @xmath651 ( mag ) & 0.96 & @xmath646 & @xmath652 + @xmath653 ( mag ) & 0.432 & @xmath654 & @xmath535 + @xmath655 ( mag ) & 0.525 & @xmath656 & @xmath657 + @xmath658 ( mag ) & 0.786 & @xmath659 & @xmath646 + | we present a systematic study of the physical properties of late - m dwarfs based on high - quality dynamical mass measurements and near - infrared ( nir ) spectroscopy .
we use astrometry from keck natural and laser guide star adaptive optics imaging to determine orbits for the late - m binaries ( m7.5+m8 ) , ( m6.5+m6.5 ) , and ( m8.5+m9 ) .
we find that ( @xmath0 ) is a pair of young brown dwarfs for which lyon and tucson evolutionary models jointly predict an age of @xmath1 myr , consistent with the age of the pleiades .
however , at least the primary component seems to defy the empirical pleiades lithium depletion boundary , implying that the system is in fact older ( if the parallax is correct ) and that evolutionary models underpredict the component luminosities for this magnetically active binary .
we find that is a pair of very low - mass stars ( @xmath2 ) with evolutionary model - derived ages consistent with the old age ( @xmath3 gyr ) implied by its lack of activity . our improved orbit for results in a higher mass for this binary ( @xmath4 ) compared to previous work ( @xmath5 * ?
? ?
* ) .
we use these mass measurements along with our published results for 2mass j2206@xmath62047ab ( m8+m8 ) to test four sets of ultracool model atmospheres currently in use . fitting these models to our nir integrated - light spectra
provides temperature estimates warmer by @xmath7250 k than those derived independently from dusty evolutionary models given the measured masses and luminosities .
we propose that model atmospheres are more likely to be the source of this discrepancy , as it would be difficult to explain a uniform temperature offset over such a wide range of masses , ages , and activity levels in the context of evolutionary models .
this contrasts with the conclusion of @xcite that model - predicted masses ( given input and ) are at fault for differences between theory and observations .
in addition , we find an opposite ( and smaller ) mass discrepancy from what they report when we adopt their model - testing approach : masses are too high rather than too low because our estimates derived from fitting nir spectra are @xmath7650 k higher than their values from fitting broadband photometry alone . |
major progress has been made in our understanding of galaxy formation and evolution since the discovery of a significant population of submm / mm - bright , dusty galaxies .
the first handful of galaxies , discovered behind lensing clusters @xcite , and in the hubble deep field @xcite , revealed the diversity of the population .
scuba @xcite , the innovative camera with which smgs were first seen , was commissioned on the james clerk maxwell telescope during 19967 . since then , the 117-bolometer , 1200-@xmath0 m mambo camera @xcite has been installed on the iram 30-m telescope , spain .
surveys at 1200@xmath0 m are sensitive to the same dusty galaxies as scuba , yet with a selection function that potentially stretches to higher redshifts and is sensitive to lower dust temperatures @xcite . a moderate fraction of smgs can be pinpointed by their @xmath0jy - level emission at 1.4ghz ( ivison et al .
2002 ) , allowing us to identify their counterparts in other wavebands .
however , up to 30% of the population remain undetected in the radio waveband in even the deepest maps and the properties of these galaxies remain a mystery . those smgs which can be localised through their radio emission
are usually found in the optical / ir wavebands to be faint , morphologically complex systems , often comprising red galaxies with bluer companions , as expected for a distant , dust - reddened , interacting star - forming population @xcite .
recent advances in their study include the measurement of the redshift distribution of radio - detected smgs @xcite , the resolution of most of the submm background by exploiting gravitational lensing @xcite , the identification of x - ray emission from a significant fraction of the population @xcite , the detection via broad co lines of collosal quantities ( @xmath510@xmath6m@xmath7 ) of molecular gas @xcite , and the first indications of the strong clustering expected for such massive galaxies @xcite .
these discoveries underline the importance of studying the far - ir luminous smgs for our understanding the formation of massive galaxies at high redshifts . for local ir - luminous galaxies , mid - ir data from the _ infrared space observatory ( iso ) _
were used to develop a number of useful diagnostics to differentiate between those powered predominantly by starbursts and by active galactic nuclei ( agn ) @xcite .
_ iso _ was sensitive to only the most luminous smgs , so efforts to assess the preponderance of agn in smgs have so far been reliant on uv / optical spectroscopy , radio observations , and the aforementioned x - ray imaging , often with ambiguous results .
the fraction of distant , far - ir - luminous galaxies with an energetically dominant agn remains poorly determined , although it is expected to be small : even smgs with unambiguous agn characteristics @xcite are thought to have a major starburst contribution to their bolometric luminosities @xcite .
this finding has been echoed in studies of distant quasars and radio galaxies : far - ir - luminous examples are found to be gas - rich and their submm emission is often resolved , as expected if their power originates in large part from stars @xcite . in this letter , we use mid - ir continuum imaging of nine mambo - selected smgs in the lockman hole to investigate the use of mid - ir data to identify the counterparts at other wavelengths of this population and demonstrate the power of mid - ir data for determining the likely power source ( agn or starburst ) in these galaxies .
@xcite reported the results of a deep unbiased mambo survey of the lockman hole , a region mapped earlier by scuba for the ` 8-mjy survey ' @xcite .
this region was selected for reasons that included its low galactic cirrus emission and the availability of high - quality complementary data at x - ray , optical / ir and radio wavelengths .
not surprisingly , the same region was selected for some of the earliest imaging with _
spitzer _ @xcite .
the data exploited herein were obtained using irac @xcite and mips @xcite , reaching noise levels of @xmath2 = 0.8 , 0.8 , 2.8 , 1.7 and 30@xmath0jy at 3.6 , 4.5 , 5.8 , 8.0 and 24@xmath0 m with absolute calibration accurate to @xmath8% .
these compare with 5-@xmath2 confusion limits ( 20 beams per source ) of around 1.0 , 1.2 and 56@xmath0jy at 3.6 , 8.0 and 24@xmath0 m @xcite .
the images cover a @xmath9 region , centered near the smg , le850.01 ( table 1 ) .
data were reduced using standard _ spitzer _ data analysis tools , as described by @xcite and @xcite .
our analysis also makes use of deep optical imaging from suprimecam on the subaru 8-m telescope ( @xmath10 , vega , 2@xmath11-radius aperture ) , _ xmm - newton _ x - ray data ( @xmath12 , 0.510kev ) and a very large array radio image ( @xmath13 , 1.4ghz ) @xcite .
the @xcite 1200-@xmath0 m sample includes nine @xmath13@xmath2 mambo sources within the region covered by the irac / mips observations .
four are detected at @xmath14@xmath2 , all of which should be secure .
the other five are @xmath13-@xmath2 detections , of which up to two may be the result of noise fluctuations , eddington bias @xcite or confusion .
@xcite performed monte carlo simulations to determine the reliability , completeness and positional accuracy of their source catalog .
we list the flux densities , significance and positions for all nine sources in table 1 and note that the positions are accurate to @xmath144@xmath11 ( 95% confidence ) .
1 shows the seds of typical local starburst- and agn - dominated ir - luminous galaxies and several smgs with well - sampled uv radio spectra .
it is worth reviewing the emission mechanisms operating across the observable wavelength range : where our adopted sample was selected , at 1200@xmath0 m ( rest - frame far - ir ) , we are sensitive to cold dust ( @xmath540k ) created in copious quantities by supernovae ( sne ; @xcite ) , which re - radiates energy absorbed in the uv from hot , young stars . in the radio , at 1.4ghz , we are again sensitive to sne ( and hence to recent star formation ) , via synchrotron radiation from relativistic electrons .
contamination via radio - loud agn is possible and , at @xmath0jy flux levels , is virtually impossible to distinguish from pure sne - related emission .
we are also sensitive to agn via x - ray emission from accretion disks and their associated coronae , although determining the origin of the x - rays is not always trivial and heavily obscured ( @xmath15@xmath16 , compton - thick ) agn can evade detection altogether @xcite . at the shortest ir wavelengths accessible to _ spitzer _
( 3.68.0@xmath0 m ) , we are provided via photospheric emission from stars with a relatively unobscured measure of stellar luminosity , possibly even of stellar mass in more evolved systems out to @xmath17 . at 24@xmath0 m ( rest - frame @xmath57@xmath0 m for a galaxy at @xmath183 )
we are sensitive to emission from @xmath5500-k dust in the circumnuclear torus of agn , and to the warmest dust in starbursts .
do our _ spitzer _ observations provide sufficient sensitivity to detect dust - enshrouded starbursts and agn in the distant universe ? in fig .
1 we illustrate the seds of a selection of well - studied dusty , luminous galaxies and models ( both agn and starbursts ) . these demonstrate the range of rest - frame mid - ir luminosities for these systems and we select arp220 and mrk231 to represent extremes of the the starburst- and agn - dominated sub - classes .
2 shows that even with its steep mid - ir spectrum ( @xmath19 , where @xmath20 ) , arp220 would be detected at 3.6 and 24@xmath0 m out to @xmath21 ( see also @xcite ) .
mrk231 , its mid - ir emission boosted an order of magnitude higher than arp220 by its agn , would be detected out to @xmath22 at 24@xmath0 m .
the diagnostic power of the near- and mid - ir bands the ability to discriminate between starbursts with and without buried , active nuclei arises from the different physical regimes and spectral features probed across the mid - ir waveband @xcite .
a steeper slope between rest - frame @xmath2310@xmath0 m is apparent for starbursts than for agn - dominated galaxies .
the latter , typified by mrk231 and smmj02399@xmath240136 ( fig . 1 ) , have power - law spectra covering rest - frame @xmath50.210@xmath0 m @xcite . in contrast
, the sed of a starburst like arp220 has a flatter region between 1 and 4@xmath0 m .
thus for mambo galaxies at @xmath253 ( assuming the same median redshift as for scuba galaxies @xcite ) , the key spectral indices are available in the 3.624-@xmath0 m bands covered by irac and mips .
, a @xmath26 versus @xmath27 color color plot , shows the tracks of arp220 and mrk231 as they are moved from the local universe to @xmath28 , as well as the track of a synthetic @xmath29-l@xmath7 starburst from @xcite .
we can see immediately that the starburst and agn populations are well separated , and that this plot has strong diagnostic potential : we can define regions where starburst and active galaxies are likely to lie at @xmath303 and hence use this to classify our smgs .
we search for radio and mid - ir counterparts to the mambo sources within their nominal error circles ( illustrated in fig . 4 ) .
we find that all of the four most secure @xmath14@xmath2 mambo sources have robust radio counterparts at 1.4ghz ( as defined by @xcite ) .
this drops slightly to 7/9 robust radio counterparts using all @xmath13@xmath2 mambo sources , somewhat higher than the 18/42 detection rate60-mjy sources , with a noise level 23 times higher than the radio data utilized here .
using equivalent data , our fraction of radio - detected mambo sources would fall to 5/9 , consistent with @xcite . ]
reported by @xcite .
the remaining two have tentative radio identifications at @xmath13@xmath2 but these are not significant enough to be reliable . if instead we had used the 24@xmath0 m _ spitzer _ imaging to identify counterparts we would have identified 8/9 of the @xmath13@xmath2 mambo galaxies with significant 24@xmath0 m sources .
interestingly , all but one of the seven mambo sources with robust radio counterparts are detected at 24@xmath0 m , the exception being mmj105207.2 + 572558 .
combining the two identification schemes increases our confidence in the identification of the proposed counterparts in both wavebands and provides reliable identifications for _ all _ nine smgs in this region .
detections at 24@xmath0 m ( fwhm @xmath5 6@xmath11 ) halve the mm positional uncertainties ( to @xmath142@xmath11 , 95% confidence ) , although the accuracy remains short of that provided by radio imaging ( @xmath140.61.6@xmath11 , 95% confidence @xcite ) . hence the radio remains the most useful waveband for localising the far - ir emission given sufficiently long integrations .
we next investigate the power of mid - ir observations to identify agn within the smg population .
only one of our nine mambo sources has an x - ray counterpart , mmj105200.2 + 572425 (= le850.08 ) , previously identified by @xcite .
it could be argued that the complexity of this immense merger there are at least seven distinct optical components in fig . 4 is such that it would be surprising if at least one of the galaxies were not active ! in the absence of the x - ray detection ,
could we have identified which components contain buried agn using only the mid - ir data ? we could .
we list the fluxes of the three resolved mid - ir components in this system separately in table 1 to check whether they have different seds .
1 shows the individual seds , arbitrarily placed at @xmath31 .
the component near - coincident with the brightest _ xmm - newton _ and radio emission ( ` b ' in fig . 4 ; see also @xcite ) has a power - law mid - ir spectrum reminiscent of mrk231 .
component ` a ' ( previously assumed to be the source of x - rays , based on a less precise centroid from _ rosat _ ) was identified as an agn at @xmath32 by @xcite ) and shows a weak break in slope at an observed wavelength around 4@xmath0 m , suggesting a possible mix of agn and star formation power sources . component ` c ' , on the other hand , has a strong short - wavelength upturn indicative of a starburst .
this system allows us to demonstrate the diagnostic power of fig . 3 , with component ` b ' lying close to the track of mrk231 , well separated from the starburst track ; component ` a ' lies between the agn and starburst tracks and within @xmath33 of its predicted redshift based on the sed of mrk231 ; component ` c ' is classified as a pure starburst .
the main area of possible confusion with this diagram is separating high - redshift starbursts from low - redshift agn ; fortunately , as in this case , we can expect agn at @xmath34 to reveal themselves in other wavebands . on the basis of these tests
, it seems reasonable to expect that mid - ir observations with _
spitzer _ will be able to weed out active galaxies within the smg population . the mid - ir data for the other smgs shown in figs 1 and 3 display a range of characteristics between the extremes of the arp220 and mrk231 templates ,
although the dispersion in fig . 1 must be due partially to the spread in redshift .
over half of the sample display an arp220-like upturn in the mid - ir with a flat sed out to 8@xmath0 m in the observed frame and a sharp rise thereafter showing that active galaxies , although present , do not dominate the mambo population . this is consistent with conclusions drawn from uv / optical spectra and x - ray imaging @xcite . using the model / empirical tracks to classify the eight mambo galaxies with measured 3.624-@xmath0 m seds , fig . 3 would indicate six starbursts and two agn .
thus @xmath35% of the smgs in our small sample have mid - ir colors characteristic of high - redshift obscured starbursts .
their mid - ir colors suggest a median redshift of @xmath51.4 for our mambo sample , somewhat lower than the @xmath31 measured for the radio - detected smgs of chapman et al .
this may reflect differences in the selection functions at 850 and 1200@xmath0 m , with the latter selecting a somewhat colder , lower redshift and intrinsically lower - luminosity section of the population , or it may simply be the result of adopting inappropriate sed templates . if we assume the mambo sources lie at similar redshifts to the scuba population and place the mambo sources at @xmath36 then 8.0@xmath0 m corresponds to rest - frame @xmath37 .
it is therefore instructive to compare the mambo galaxies with the @xmath37-corrected spectrum of arp220 ( fig .
arp220 has an absolute @xmath38 magnitude of @xmath2424.4 , roughly 1.3l@xmath39 @xcite , and would have an 8.0-@xmath0 m flux of @xmath52@xmath0jy at @xmath40 .
the median 8.0-@xmath0 m flux of the @xmath14@xmath2 mambo sources is @xmath520@xmath0jy , implying naively that these galaxies are already @xmath515l@xmath41 .
the more plausible alternatives are that these galaxies lie at lower redshifts than we have assumed ( @xmath422 , as implied by fig .
3 ) or that their observed 8.0-@xmath0 m luminosity is dominated by massive , young stars or hot dust . determining the exact cause of the apparently immense mid - ir luminosities of these smgs will require spectoscopic observations in the optical and mid - ir to confirm their redshifts and the details of their mid - ir spectral properties .
_ spitzer _ observations between 3.6 and 24@xmath0 m of nine galaxies selected at 1200@xmath0 m have improved our confidence in the identification of secure counterparts , complementing what was learned from radio imaging .
robust identifications have been possible for @xmath590% of the mambo galaxies using 24-@xmath0 m imaging , a similar fraction to that achievable from radio imaging .
together , radio and mid - ir imaging have yielded plausible counterparts for the entire mambo sample in this region .
we conclude that 3.624-@xmath0 m observations provide a useful tool to aid in identifying smgs .
accretion onto an obscured central engine is sometimes betrayed by the shape of the mid - ir continuum , confirming _ spitzer _ s potential to weed out active galaxies .
the fraction of mm galaxies with energetically - important agn in our small sample , 25% ( or 560% at the 95% confidence level ) , is similar to that estimated from uv / optical spectroscopy and deep x - ray imaging @xcite , but the majority of mambo galaxies have rest - frame mid-/far - ir seds commensurate with obscured starbursts .
this work is based in part on observations made with the _
spitzer space telescope _ , which is operated by the jet propulsion laboratory ( jpl ) , california institute of technology ( caltech ) under nasa contract 1407 .
support for this work was provided by nasa through contract number # 960785 issued by jpl / caltech .
lccccccccl & [email protected]&[email protected] & 249@xmath1431&[email protected]&[email protected]&[email protected]&[email protected]&36@xmath145&(a ) ; le850.14 + & & & &[email protected]&[email protected]&[email protected]&[email protected]&72@xmath148&(b ) + _ mmj105201.3 + 572448 _ & [email protected] & [email protected] & 193@xmath1431&[email protected]&[email protected]&[email protected]&[email protected]&73@xmath148&le850.01 + _ mmj105155.4 + 572310 _ & [email protected] & [email protected] & 125@xmath1432&[email protected]&[email protected]&[email protected]&[email protected]&47@xmath145&le850.18 + _ mmj105200.2 + 572425 _ & [email protected] & [email protected] & 534@xmath1455&112@xmath1412&[email protected]&[email protected]&[email protected]&@xmath4315&(a ) ; @xmath32 , le850.08 + & & & 282@xmath1430&[email protected]&[email protected]&[email protected]&[email protected]&80@xmath148&(b ) + & & & 162@xmath1430&[email protected]&[email protected]&[email protected]&[email protected]&22@xmath146&(c ) + & [email protected]&@xmath4311 & 121@xmath1430&[email protected]&[email protected]&[email protected]&[email protected]&21@xmath146 & + _ mmj105207.2 + 572558_&[email protected] & @xmath4312 & @xmath4390&[email protected]&[email protected]&[email protected]&[email protected]&23@xmath146 & + * mmj105203.6 + 572612 * & [email protected]&@xmath4310 & 166@xmath1430&[email protected]&[email protected]&[email protected]&[email protected]&19@xmath146 & + _ mmj105216.0 + 572506 _ & [email protected]&[email protected] & 578@xmath1459&&&&&48@xmath145&le850.29 + _ mmj105148.5 + 572408 _ & [email protected]&@xmath437.8 & 211@xmath1430&[email protected]&[email protected]&[email protected]&[email protected]&29@xmath146 & + | we present _ spitzer _ observations in five wavebands between 3.6 and 24@xmath0 m of an unbiased sample of nine luminous , dusty galaxies selected at 1200@xmath0 m by the mambo camera on the iram 30-m telescope , a population akin to the well - known submillimeter ( submm ) or ` scuba ' galaxies ( hereafter smgs ) .
owing to the coarse resolution of submm / mm instrumentation , smgs have traditionally been difficult to identify at other wavelengths .
we compare our multi - wavelength catalogs to show that the overlap between 24 and 1200@xmath0 m must be close to complete at these flux levels .
we find that all ( 4/4 ) of the most secure @xmath14@xmath2 smgs have robust @xmath14@xmath2 counterparts at 1.4ghz , while the fraction drops to 7/9 using all @xmath13@xmath2 smgs .
we show that combining mid - infrared ( mid - ir ) and marginal ( @xmath13@xmath2 ) radio detections provides plausible identifications in the remaining cases , enabling us to identify the complete sample .
accretion onto an obscured central engine is betrayed by the shape of the mid - ir continuum emission for several sources , confirming _ spitzer _ s potential to weed out active galaxies .
we demonstrate the power of a @xmath3 versus @xmath4 color - color plot as a diagnostic for this purpose .
however , we conclude that the majority ( @xmath575% ) of smgs have rest - frame mid-/far - ir spectral energy distributions ( seds ) commensurate with obscured starbursts .
sensitive 24-@xmath0 m observations are clearly a useful route to identify and characterize reliable counterparts to high - redshift far - ir - bright galaxies , complementing what is possible via deep radio imaging . |
the search for free - floating brown dwarfs ( hereafter referred to as bds ) has finally been proved succesful .
after years of continuous efforts to detect intrinsically fainter and less massive objects both in the field and in open star clusters ( see e.g. the reviews by jameson @xcite and reid @xcite ) , the recent photometric survey of a small sky area in the pleiades ( age 70120 myr ) by zapatero osorio , rebolo & martn ( @xcite ) ( hereafter paper i ) has provided objects that according to all available evolutionary models are beyond the substellar limit ( rebolo , zapatero osorio & martn @xcite ; martn , rebolo & zapatero osorio @xcite ) .
the genuine bd nature of two of them has been confirmed by the detection of high lithium abundances in their atmospheres ( rebolo et al .
the re - appearance of lithium in objects at the bottom of the pleiades main sequence ( ms ) has helped to empirically define the borderline that separates very low - mass stars from bds . at present ,
both sides of the substellar limit are populated with several objects which may allow us to establish qualitative and quantitative differences between them and also with respect to other field objects with similar spectral types .
martn et al .
( @xcite ) have started to approach this issue using intermediate - resolution optical spectroscopy . in this paper
, we examine this problem using near - infrared photometry .
we have obtained @xmath9 , @xmath10 and @xmath8 magnitudes for some pleiades very low - mass stars selected from the literature , as well as for the most interesting objects in the optical survey presented in paper i. we will focus here on the utility of combining both optical and infrared photometry to identify true pleiades bds , disentangle binarity effects and discriminate cluster members from contaminants that appear in optical surveys .
these observations are mainly aimed at characterizing the cluster substellar borderline and studying the infrared sequence that connects pleiades least massive stars with bds .
our sample is presented in table 1 where we have avoided the term `` pleiades '' for those objects taken from paper i. all objects in the sample have measured spectral types equal or later than m6 , and they could be located around the substellar limit in the pleiades cluster according to their @xmath7 magnitudes and ( @xmath11 ) colours .
ppl 15 ( discovered by stauffer , hamilton & probst @xcite ) is the m6.5-type pleiad that at present defines the borderline which separates very low - mass stars from bds in the cluster .
this is based on the fact that ppl 15 has preserved some of its initial lithium content .
the comparison with theoretical models ( basri , marcy & graham @xcite ; rebolo et al .
@xcite ) yields that the most likely mass of ppl 15 is @xmath20.08 @xmath1 .
hhj 2 ( m6.5 ) , hhj 3 ( m6 ) and hhj 6 ( m6.5 ) are proper motion members taken from hambly , hawkins & jameson ( @xcite ) .
their optical - infrared magnitudes and spectral types ( steele & jameson @xcite ) place these objects close to the substellar limit , but still on the stellar domain .
hhj 3 has depleted its lithium ( marcy , basri & graham @xcite ) .
in fact , none of the hhj objects in our sample shows lithium in their atmospheres ( basri 1996 , private communication ) , implying that all of them have masses larger than 0.08 @xmath1 and therefore , they are among the least massive stars in the cluster
. there are no lithium observations available for ppl 1 ( m6.5 ) ( stauffer et al .
@xcite ) , but it has a photometry very similar to that of hhj 3 , and a radial velocity consistent with membership .
the remaining objects in our sample are taken from the optical survey of paper i from which two genuine bds have emerged , namely teide 1 ( m8 ) and calar 3 ( m8 ) .
both have successfully passed the lithium test ( rebolo et al .
@xcite ) . according to their radial velocities and spectral types , calar 1 ( m9 ) , calar 2 ( m6 ) , calar 5 ( m6.5 ) and roque 1 ( m7 ) seem to be non - members of the pleiades ( martn et al .
the infrared photometry will shed new light on their membership in the cluster .
the @xmath9 , @xmath10 and @xmath8 observations were performed in 1996 february 9 at the 4.2 m william herschel telescope ( wht , observatorio del roque de los muchachos on the island of la palma , spain ) , using the whircam infrared camera equipped with an insb ( 256@xmath12256 ) detector array .
this detector provided 0.24 pixels and a field of view of @xmath21@xmath13@xmath121@xmath13 .
the standard @xmath9 , @xmath10 and @xmath14 infrared filter set was used .
the total integration time per object and per filter was 50 s , where each final image actually consisted of 5 co - added 10 s exposures .
raw data were processed using standard techniques within the iraf environment .
dark frames were first substracted from the images before combining them in order to obtain good flat - fields for each filter . then
, individual frames were divided by the normalized , mean flat - field .
the photometric analysis was carried out using routines within daophot .
weather conditions during the campaign were always fairly photometric .
instrumental aperture magnitudes were corrected for atmospheric extinction and transformed into the ukirt system using observations at different airmasses of the standards fs 9 and fs 12 ( casali & hawarden @xcite ) .
the @xmath14 data were converted to the @xmath8-band magnitudes using the transformation equation given in wainscoat & cowie ( @xcite ) .
we estimate that the accuracy in the final magnitudes is 0.05 mag at each wavelength .
the final @xmath9 , @xmath10 and @xmath8 photometry of the sample is provided in table 1 .
our magnitudes are in good agreement within @xmath00.1 mag with those found in the literature for the objects in common ( hhj and ppl objects ) , except for the @xmath8-band of ppl 15 , where a difference of 0.18 mag is observed with respect to basri et al.s ( @xcite ) data in the sense that our result is brighter .
we have only a @xmath9 measurement for hhj 3 ; for completeness in table 1 , @xmath10 and @xmath8 magnitudes in the ukirt system were taken from steele , jameson & hambly ( @xcite ) .
.infrared photometry [ cols="<,^,^,^,^,^,^",options="header " , ] all isochrones reproduce qualitatively the general trend of the observational data .
however , within the age range assumed for the pleiades ( 70150 myr ) , the models of burrows et al .
( @xcite ) and dantona & mazzitelli ( @xcite ) predict rather blue ( @xmath15 ) colours and/or faint @xmath8 magnitudes . only for very young ages
, those isochrones would fit the pleiades data , implying that cluster members like the faintest hhj objects should have preserved lithium , clearly in desagreement with the lithium observations of these stars .
this discrepancy for very low - masses was also noted in terms of effective temperature and luminosity in the @xmath16 persei cluster by zapatero osorio et al .
( @xcite ) .
the 70 myr evolutionary track of burrows et al .
( @xcite ) seems to coincide with that of the ` nextgen ' 120 myr isochrone of chabrier et al .
( @xcite ) . for a given age , burrows et al.s models predict effective temperatures that are hotter by 150100 k than those predicted by chabrier et al . , and luminosities that are slightly brighter by 0.030.015 dex in the mass range in common for both theoretical models . the differences between the two sets of isochrones of baraffe et al .
( ` base ' ) and chabrier et al .
( ` nextgen ' ) are remarkable , e.g. , given an age and a mass the ( @xmath15 ) offset could be as large as 0.5 mag .
the ` base ' model atmospheres have been improved recently by including more accurate molecular line lists , yielding the ` nextgen ' models .
these are claimed to reproduce the observations better than previous sets of computations ( see allard et al .
@xcite ; baraffe & chabrier @xcite ; chabrier et al .
@xcite ) . however , the masses that are inferred from the comparison of our data with the ` nextgen ' models in fig
. 4 are lower by a factor @xmath21.25 than those obtained from the lithium test .
this is basically due to the apparent different age that is derived from the isochrones in fig .
4 ( 70 myr , the canonical age of the cluster ) , and from the combination of lithium and luminosity of objects close to the substellar limit ( 120 myr , basri et al .
@xcite ) . since the input molecular line lists for computing synthetic spectra for cool temperatures
are incomplete , colours and magnitudes measured on these spectra are hampered by uncertainties of @xmath20.5 mag ( chabrier et al .
@xcite ) . on the other hand , masses obtained from lithium are more reliable because they are based on the simpler physics of the interior of fully - convective objects ( magazz , martn & rebolo @xcite ; bildsten et al .
@xcite ) . in order to bring the masses obtained from the @xmath8 vs @xmath15 diagram into agreement with those inferred from the lithium test for an age of 120 myr
, it would be required to shift the ` nextgen ' isochrone by about 0.4 mag towards redder colour , which is within the theoretical uncertainties estimated by chabrier et al .
@xcite .
the sequence of very low - mass stars and bds in the pleiades is depicted in fig .
4 . our suggested decomposition for hhj 6 and
ppl 15 is indicated .
the location of the likely companions nicely fits the sequence of single members .
we consider that the primary of ppl 15 defines the transition region between stars and bds , i.e. masses within 0.080.075 @xmath1 , and we propose that its photometry could be adopted as indicative of the location of the substellar limit in the pleiades .
the derived absolute magnitudes are listed in table 3 , where the interval of values corresponds to the above mass range .
we have estimated the star - bd frontier for older clusters like the hyades , as well as for objects of different ages in the field using the isochrones of chabrier et al .
( @xcite ) ( nextgen model atmospheres ) normalized to the primary of ppl 15 .
we consider the theoretical magnitude differences associated to different ages in order to evolve from the pleiades to 600 , 1000 and 5000 myr .
we have also inferred the spectral types of the objects at the substellar limit shown in the last column of table 3 by comparing our @xmath17 , @xmath15 and @xmath18 colours with those of kirkpatrick & mccarthy ( @xcite ) for field objects .
we obtain fainter magnitudes and cooler spectral types than suggested by the mass - luminosity and mass - spectral type relationships proposed by henry & mccarthy ( @xcite ) and kirkpatrick & mccarthy ( @xcite ) respectively .
our results are in better agreement with those of baraffe & chabrier ( @xcite ) . at the age of the hyades ,
the substellar limit has moved to m9 spectral type ; for example , cluster objects with the mass of the primary of ppl 15 might show a spectrum similar to a field m9 dwarf , whereas those like teide 1 would have become much cooler .
field very late - m dwarfs ( m7m10 ) are likely older than 1000 myr and , therefore , most of them should be very low - mass stars rather than substellar objects .
it is possible that photometric surveys reveal nearby young bds with spectral types earlier than m9m10 in the solar neighbourhood . because of their recent formation
they will be less numerous than stars of similar spectral types in the field , but their relatively higher luminosity will favour detection in all - sky near - infrared surveys like denis and 2mass . the probability of finding these field young bds will depend on the mass function for very low - masses and on the mixture of ages ( see martn , zapatero - osorio & rebolo @xcite )
. the lithium test will be useful in order to establish their substellar nature .
we have presented near - infrared @xmath9 , @xmath10 and @xmath8 photometry for some of the least luminous objects so far known in the pleiades cluster .
they are very low - mass stars approaching the substellar limit , transition objects , and massive bds ( hhj and ppl objects , and calar 3 and teide 1 ) .
in addition , we have also included in our sample four of the bd candidates proposed in paper i ( calar 1 , 2 , 4 , 5 and roque 1 ) in order to asses their membership on the basis of the infrared data .
we have made use of all the available spectroscopic and photometric knowledge on our final list of objects to characterize the sequence that cluster members around the substellar limit and beyond do follow in the infrared colour - magnitude diagrams .
this sequence is clearly defined by the faintest hhj objects , ppl 1 , ppl 15 , calar 3 and teide 1 , lying 0.30.4 mag above the field ms for spectral types later than m6 .
objects found in future surveys with similar photometry ( optical and infrared ) should be considered as likely cluster members .
undoubtely , infrared data is highly useful to discriminate those objects that are true pleiades members from those that are not .
optical surveys could be contaminated by background objects that although fitting in the cluster sequence at these wavelengths , may suffer from interestellar reddening .
as extinction affects much less infrared magnitudes , consequently these objects do not match the @xmath9 , @xmath10 and @xmath8 photometry expected for the pleiades ( this is the case of calar 2 and calar 5 ) .
foreground late - type stars are also possible contaminants in photometric surveys , although its contamination is expected to be considerably smaller . however , in the optical survey of paper i two of these field stars were detected , roque 1 and calar 1 , both having very high radial velocities ( martn et al .
the reddest object in optical wavelengths , calar 1 ( m9 ) , appears to have rather bluer infrared colours than expected for its spectral type .
this together with its high radial velocity suggests membership to an old population .
the large dispersion observed in the infrared photometry of pleiades members is likely attributable to binarity effects .
two objects in our sample , namely hhj 6 and ppl 15 , lie @xmath20.7 mag above the single cluster sequence in infrared diagrams . adopting an age of 120 myr for the cluster and using the photometric data of very low - mass pleiades stars and bds as well as the theoretical tracks of chabrier et al .
( @xcite ) , we have reproduced the photometry of both , finding that the likely masses of the components are 0.085 and [email protected] @xmath1 for hhj 6 , 0.080 and [email protected] @xmath1 for ppl 15 .
these results are in fairly good agreement with those obtained from the lithium test .
follow - up infrared spectroscopic observations of ppl 15 should be made to try to detect the effects of the companion on the spectrum , which are expected to be detectable .
since the primary of ppl 15 would be exactly located at the transition region between stars and bds in the pleiades , we have used it to define the empirical location of the substellar limit in infrared diagrams , for which we have corrected the observed magnitudes from the contribution of the secondary .
we have estimated that the star - bd frontier on the single pleiades sequence is located at magnitudes @xmath7 = 17.818.0 , @xmath9 = 15.615.7 , @xmath10 = 14.915.0 and @xmath8 = 14.514.6 and at optical spectral types m6.5m7 .
new pleiades objects that are found fainter along the photometric sequence defined in this paper might be bds indeed . based on this result , we estimate that the hydrogen - burning mass limit occurs at spectral types around m9 in the hyades cluster .
allard , f. , hauschildt , p.h .
, 1995 , apj , 445 , 433 allard , f. , hauschildt , p.h .
, 1996 , apj , submitted allard , f. , hauschildt , p.h . , baraffe , i. , chabrier , g. , 1996 , apj , 465 , l123 baraffe , i. , chabrier , g. , 1996 , apj , 461 , l51 baraffe , i. , chabrier , g. , allard , f. , hauschildt , p.h . , 1995 ,
apj , 446 , l35 basri , g. , marcy , g.w . ,
graham , j.r .
, 1996 , apj , 458 , 600 bessell , m.s . ,
stringfellow , g.s .
, 1993 , ara&a , 31 , 433 bildsten , l. , brown , e.f . ,
matzner , c.d .
, ushomirsky , g. , 1996 , apj , submitted burrows , a. , hubbard , w.b . ,
saumon , d. , lunine , j.i . , 1993 ,
apj , 406 , 158 casali , m.m . , hawarden , t.g . , 1992 , jcmt - unkirt newsletter , 4 , 33 chabrier , g. , baraffe , i. , plez , b. , 1996 , apj , 459 , l91 crawford , d.l . , perry , c.l . , 1976 , aj , 81 , 419 dantona , f. , mazzitelli , i. , 1994 , apjs , 90 , 467 eggen , o. , iben , i. , 1989 , aj , 97 , 431 hambly , n.c .
, hawkins , m.r.s . ,
jameson , r.f .
, 1993 , a&as , 100 , 607 henry , t.j . ,
mccarthy , d.w .
jr . , 1993 ,
aj , 106 , 773 jameson , r.f . , 1995 , in the bottom of the main sequence and beyond , ed .
tinney ( berlin : springer ) , p. 183
jameson , r.f .
, skillen , i. , 1989 , mnras , 239 , 247 kirkpatrick , j.d . ,
mccarthy , d.w .
jr . , 1994 ,
aj , 107 , 333 leggett , s.k . , 1992 , apjs , 82 , 351 leggett , s.k . ,
allard , f. , berriman , g. , dahn , c.c .
, hauschildt , p.h . , 1996 ,
apjs , 104 , 117 magazz , a. , martn , e.l . , rebolo , r. , 1993 , apj , 404 , l17 marcy , g.w . ,
basri , g. , graham , j.r . , 1994 , apj , 428 , l57 martn , e.l . ,
rebolo , r. , magazz , a. , apj , 436 , 262 martn , e.l . ,
rebolo , r. , zapatero osorio , m.r .
, 1996a , apj , 469 , 706 martn , e.l . , zapatero osorio , m.r . ,
rebolo , r. , 1996b , in the 3rd denis conference , puerto de la cruz , marzo 1996 , in press mead , a.r.g . , ballard , k.r . , brand , p.w.j.l . ,
hough , j.h .
, brindle , c. , bailey , j.a . , 1990 , a&as , 83 , 183 rebolo , r. , martn , e.l . ,
magazz , a. , 1992 , apj , 389 , l83 rebolo , r. , zapatero osorio , m.r . ,
martn , e.l . , 1995 , nat , 377 , 129 rebolo , r. , martn , e.l . ,
basri , g. , marcy , g.w .
, zapatero osorio , m.r .
, 1996 , apj , 469 , l53 reid , i. n. , 1995 , in the bottom of the main sequence and beyond , ed .
tinney ( berlin : springer ) , p. 53
reid , i.n . ,
gilmore , g. , 1984 , mnras , 206 , 19 rieke , g.h . , lefobsky , m.j . , 1985 , apj , 288 , 618 stauffer , j.r . , 1982 , aj , 87 , 1507 stauffer , j.r . , 1984 , apj , 280 , 189 stauffer , j.r . , hamilton , d. , probst , r .g . , 1994 , aj , 108 , 155 stauffer , j.r . ,
hamilton , d. , probst , r .g . , rieke , g. , mateo , m. , 1989 , apj , 344 , l21 stauffer , j.r . , liebert , j. , giampapa , m. , 1995 , aj , 109 , 298 steele , i.a . ,
jameson , r.f .
, 1995 , mnras , 272 , 630 steele , i.a . , jameson , r.f . ,
hambly , n.c . , 1993 , mnras , 263 , 647 steele , i.a . ,
jameson , r.f . ,
hodgkin , s.t . , hambly , n.c .
, 1995 , mnras , 275 , 841 tinney , c.g .
, mould , j.r . ,
reid , i.n .
, 1993 , aj , 105 , 1045 wainscoat , r.j . ,
cowie , l.l . , 1992 , aj , 103 , 332 williams , d.m . ,
boyle , r.p .
, morgan , w.t . , rieke , g.h . , stauffer , j.r . ,
rieke , m.j . , 1996 , apj , 464 , 238 zapatero osorio , m.r .
, rebolo , r. , martn , e.l .
, 1996a , a&a , in the press , paper i zapatero osorio , m.r . ,
rebolo , r. , martn , e.l .
, garca lpez , r.j . , 1996b , a&a , 305 , 519 | we have obtained near - infrared observations of some of the faintest objects so far known towards the pleiades young stellar cluster , with the purpose of investigating the sequence that connects cluster very low - mass stars with substellar objects .
we find that infrared data combined with optical magnitudes are a useful tool to discriminate cluster members from foreground and background late - type field stars contaminating optical surveys .
the bottom of the pleiades sequence is clearly defined by the faint hhj objects as the very low - mass stars approaching the substellar limit , by the transition object ppl 15 , which will barely ignite its hydrogen content , and by the two brown dwarfs calar 3 and teide 1 .
binarity amongst cluster members could account for the large dispersion observed in the faint end of the infrared colour - magnitude diagrams .
two objects in our sample , namely hhj 6 and ppl 15 , are overluminous compared to other members , suggesting a probable binary nature .
we have reproduced the photometric measurements of both of them by combining the magnitudes of cluster very low - mass stars and brown dwarfs and using the most recent theoretical evolutionary tracks .
the likely masses of the components are slightly above the substellar limit for hhj 6 , while they are 0.080 and [email protected] @xmath1 for ppl 15 .
these masses are consistent with the constraints imposed by the published lithium observations of these pleiads .
we find a single object infrared sequence in the pleiades cluster connecting very low - mass stars and brown dwarfs .
we propose that the substellar mass limit ( @xmath20.075 @xmath1 ) in the pleiades ( @xmath2120 myr ) takes place at absolute magnitudes @xmath3 = 12.4 , @xmath4 = 10.1 , @xmath5 = 9.4 and @xmath6 = 9.0 ( spectral type m7 ) .
cluster members fainter by 0.2 mag in the @xmath7-band and 0.1 mag in the @xmath8-band should be proper brown dwarfs .
the star - brown dwarf frontier in the hyades cluster ( 600 myr ) would be located at @xmath3 = 15.0 , @xmath4 = 11.6 , @xmath5 = 10.8 and @xmath6 = 10.4 ( spectral type around m9 ) .
fon an age older than 1000 myr we estimate that brown dwarfs are fainter than @xmath6 = 10.9 ( spectral type later than m9.5 ) . |
nonwoven materials or fleece are webs of long flexible fibers that are used for composite materials , e.g. filters , as well as in the hygiene and textile industries .
they are produced in melt spinning operations : hundreds of individual endless fibers are obtained by the continuous extrusion of a molten polymer through narrow nozzles that are densely and equidistantly placed in a row at a spinning beam .
the viscous or viscoelastic fibers are stretched and spun until they solidify due to cooling air streams . before the elastic fibers lay down on a moving conveyor belt to form a web , they become entangled and form loops due to the highly turbulent air flows .
the homogeneity and load capacity of the fiber web are the most important textile properties for quality assessment of industrial nonwoven fabrics .
the optimization and control of the fleece quality require modeling and simulation of fiber dynamics and lay down .
available data to judge the quality , at least on the industrial scale , are usually the mass per unit area of the fleece . a stochastic model for the fiber deposition in the nonwoven production was proposed and analyzed in ref .
its core is a stochastic ornstein uhlenbeck process for the random motion of the fiber .
the aim of this paper is to determine the parameters of the ornstein uhlenbeck process from available mass per unit area data , i.e. the occupation time in mathematical terms . for the sake of simplicity , we focus on a one dimensional version of the ornstein uhlenbeck process . the paper is organized as follows : in section [ s : model ]
we introduce the ornstein uhlenbeck process as a prototypic model for the fiber deposition . section [ s : theory ] is devoted to the derivation of the expectation value for the occupation times .
an algorithm to estimate the parameters in the ornstein uhlenbeck process from available occupation times is presented along with numerical experiments in section [ s : numerics ] .
finally , we draw some conclusions and give an outlook to open questions .
as a prototypic model for the fiber deposition , we consider the general one dimensional ornstein uhlenbeck process @xmath0 the equilibrium @xmath1 and the stiffness @xmath2 govern the deterministic part whereas the diffusion parameter @xmath3 and a standard wiener process ( or brownian motion ) @xmath4 contribute the stochastic part . for the sake of simplicity ,
we consider mainly the centered process @xmath5 satisfying @xmath6 the real - valued random variable @xmath5 models the deposition point of an individual fiber on the fleece . if we follow the random variable over a time interval @xmath7 $ ] for @xmath8 , we obtain the path of an individual fiber . to introduce the mathematical analog of the mass per unit area
we need the following definition .
let @xmath8 and consider an interval @xmath9\subset{\mathbb{r}}$ ] , where @xmath10 or @xmath11 are allowed . the _ occupation time _
@xmath12}$ ] is defined as @xmath13}(x_t ) : = \int_0^t \mathbf{1}_{[a , b]}(x_t ) dt = \int_0^t\int_a^b \delta_0(x_t - x)\ , dx\ , dt\;.\ ] ] here , @xmath14}$ ] denotes the indicator function of the interval @xmath9 $ ] and @xmath15 is the donsker s delta function introduced in definition [ d : donsker ] , below . the occupation time is a random variable itself .
it models the time , the random process spends inside the spatial interval @xmath9 $ ] during the time interval @xmath7 $ ] . in terms of our physical model for the nonwoven production
, the occupation time can be interpreted as the mass of fiber material deposited inside the interval @xmath9 $ ] , i.e. the mass per unit area of the final fleece .
this quantity is easily accessible even on the scale of industrial production and hence it will serve as the input to our parameter estimation problem . in the next chapter , we will present tools from white noise analysis to derive the expectation of the occupation time for the centered ornstein uhlenbeck process @xmath5 .
although it is possible to derive the results by classical stochastic analysis methods , we use a white noise approach to generalize the concepts also to higher dimensions , where one can give a rigorous meaning to multidimensional donskers delta functions as a white noise distribution , in later research .
moreover in future work an extension to more complicated processes ( e.g. with fractional noise term ) is planed .
thereafter we show , how to estimate the parameters @xmath16 , @xmath17 of the process from available data for the occupation times .
we start by considering the gelfand triple @xmath18 , where @xmath19 denotes the schwartz space of rapidly decreasing smooth functions , @xmath20 the hilbert space of real valued square integrable ( equivalence classes of ) functions on @xmath21 w.r.t . lebesgue measure and @xmath22 the topological dual of @xmath19 , i.e. the space of tempered distributions .
this particular choice is the usual one in white noise analysis @xcite . by @xmath23
we denote the duality pairing between @xmath24 and @xmath25 , an extension of the standard inner product on @xmath20 in the sense of a gelfand triple .
next , we want to introduce a probability measure on the space @xmath22 .
therefore , we consider the @xmath17algebra @xmath26 generated by the cylinder sets @xmath27 } \newcommand{\rcb}{\right\}}\lab f , \cdot{\right\rangle}:\ f\in s({\mathbb{r}})\rcb$ ] .
the white noise measure @xmath28 on @xmath29 is given via minlos theorem @xcite by its characteristic function @xmath30 @xmath31 for @xmath32 .
the hilbert space of complex valued square
integrable functions w.r.t .
this measure @xmath28 is denoted by @xmath33 . for @xmath34
we have the isometry @xmath35 thus , this result can also be extended to @xmath36 in the sense of an @xmath37limit .
hence , within the above formalism , a version of a standard wiener process can be written as @xmath38 , for @xmath39 and @xmath40 . to treat the occupation time of the ornstein uhlenbeck process in the white noise framework , we need the space of hida distributions @xmath41 . the above introduced space @xmath37 serves as the central space of the gelfand triple @xmath42 , where @xmath43 denotes the space of hida test functions .
the dual pairing of @xmath44 with @xmath45 is denoted by @xmath46 . for a detailed description of the construction of the hida triple
we refer to ref .
@xcite . for a function @xmath25 ,
the exponential @xmath47 is an element of @xmath43 .
we will characterize hida distributions with the help of the @xmath48transform and @xmath49functionals .
@xmath50 + the @xmath48transform of a hida distribution @xmath51 is defined as @xmath52 where @xmath25 . since @xmath53 ,
the expectation of a hida distribution @xmath51 can be defined by @xmath54 @xmath50 + we call @xmath55 a @xmath49-functional , if 1 . for all @xmath34 ,
the mapping @xmath56 is analytic and hence has an entire extension to @xmath57 .
2 . there exist constants @xmath58 and a continuous norm @xmath59 on @xmath19 such that @xmath60 for all @xmath61 and all @xmath62 .
the proof of the following equivalence theorem can be found in ref .
@xcite .
[ t : char ] a mapping @xmath55 is the @xmath48transform of a unique element in @xmath63 , if and only if @xmath64 is a @xmath49functional . in the sense of a limit in @xmath63
we can define the white noise process as @xmath65 where @xmath66 denotes the dirac delta in @xmath39 .
this process can be considered as the time derivative of the wiener process @xmath67 in the sense of hida distributions .
the next result follows from theorem [ t : char ] and concerns integration of a family of hida distributions , see ref .
@xcite .
[ t : intcor ] let @xmath68 be a measure space and @xmath69 a mapping from @xmath70 to @xmath63 .
we assume that the @xmath48transform @xmath71 satisfies the following conditions : 1 .
the mapping @xmath72 is measurable for all @xmath32 .
2 . there exists a continuous norm @xmath59 on @xmath19 and functions + @xmath73 and @xmath74 integrable with respect to @xmath75 such that @xmath76 for all @xmath32 , @xmath61 .
then it holds in the sense of bochner integration in a suitable sub
hilbert space of @xmath63 , that the integral of the familiy of hida distributions is itself a hida distribution , i.e. @xmath77 and the @xmath48transform interchanges with the integration @xmath78 based on the above theorem , we introduce the following hida distribution .
[ d : donsker ] we define donsker s delta at @xmath79 corresponding to @xmath80 by @xmath81 in the sense of bochner integration@xcite .
its @xmath48transform in @xmath32 is given by @xmath82 coming back to the ornstein uhlenbeck process , we note that we can write it in the framework of white noise analysis as @xmath83 for @xmath84 .
this can be seen as follows : + clearly @xmath85 is a gaussian random variable with expectation @xmath86 and covariance @xmath87 thus , by uniqueness @xmath85 is the ornstein uhlenbeck process solving the corresponding sde ( [ e : sde_ou ] ) . in the special case @xmath88 of the centered ornstein uhlenbeck process @xmath89 , where @xmath90 , we obtain that the @xmath48transform of the corresponding donsker s delta at @xmath91 is given by @xmath92 for @xmath25 . using @xmath93 the expectation is readily available by @xmath94 let @xmath5 be a centered ornstein uhlenbeck process on the time interval @xmath7 $ ] , where @xmath8 .
let @xmath9\subset{\mathbb{r}}$ ] be an interval , where @xmath10 and @xmath11 are allowed .
the expectation of the occupation time @xmath12}(x_t)$ ] is given by @xmath95}(x_t ) \right ) & \nonumber\\ = & \frac{1}{2 } \int_0^t { \mathrm{erf}}\left ( \frac{\alpha b}{1-\exp(-2\lambda t)}\right ) - { \mathrm{erf}}\left ( \frac{\alpha a}{1-\exp(-2\lambda t)}\right ) \ , dt\;,\end{aligned}\ ] ] where @xmath96 .
the occupation time , i.e. the time the process spends in the space interval @xmath9 $ ] during the time @xmath48 is given by @xmath13}(x_t ) = \int_0^t \int_a^b \delta_0(x_t - x ) \ , dx \ , dt\;.\ ] ] interchanging integrations due to theorem [ t : intcor ] , we obtain the expectation of the occupation time of the ornstein uhlenbeck process @xmath97}(x_t ) ) & = & \int_0^t \int_a^b { \mathbb{e}}_{\mu}(\delta_0(x_t - x))\ , dx\ , dt \nonumber\\ & = & \int_0^t \frac{1}{\sqrt{2\pi k } } \int_a^b \exp { \left(}-\frac{x^2}{2k}\rb \ , dx \ , dt \nonumber \\ & = & \frac{1}{2 } \int_0^t { \mathrm{erf}}{\left(}\frac{b}{\sqrt{2k}}\rb-{\mathrm{erf}}{\left(}\frac{a}{\sqrt{2k}}\rb \ , dt \nonumber\end{aligned}\ ] ] @xmath98 with @xmath99 .
the expected occupation time of the ornstein uhlenbeck process @xmath5 defined via the sde ( [ e : sde_ou_sim ] ) can be computed using eqn .
( [ e : occtime ] ) . alternatively , one can also compute the occupation time using a monte
carlo simulation of the underlying process .
we generate @xmath100 of sample paths of the ornstein uhlenbeck process with a fixed parameter set and compute the sample occupation time @xmath101}(\cdot)$ ] for each path . as in the basic idea of the monte
carlo simulation , the sample average of the occupation time serves as an estimator for the expectation value .
if large numbers of samples are considered , the estimator yields a better approximation . in the sequel
, we shortly outline the numerical approximation of a stochastic process like the ornstein uhlenbeck process ( [ e : sde_ou_sim ] ) .
consider a general non autonomous stochastic differential equation @xmath102 defined in the time interval @xmath7 $ ] , where @xmath103\times{\mathbb{r}}\to { \mathbb{r}}$ ] and @xmath4 is a standard wiener process . under mild conditions
the solution of ( [ eqn : genouprocess ] ) has the following form @xmath104 note that the solution @xmath5 is a random variable for each @xmath105 . for details on the existence and uniqueness of solutions to ( [ eqn : genouprocess ] )
, we refer to ref .
@xcite .
this solution can be numerically estimated by using the euler
maruyama method .
we discretize the interval @xmath7 $ ] using a time step @xmath106 for some positive integer @xmath107 and introduce discrete time points @xmath108 for @xmath109 .
let @xmath110 denote the numerical approximation of @xmath111 .
further , we assume that the second integral on the right hand side of ( [ eq : intergralformofsde ] ) is integrated using the it version of stochastic integrals . then the euler
maruyama method reads as @xmath112 figure [ fig : ouprocess ] shows a sample path of the ornstein uhlenbeck process computed using the euler
maruyama method . using the discrete version @xmath110 of the process
, we can easily calculate the sample occupation time @xmath101}(\tilde{x}_j)$ ] .
the accuracy of the numerical solution to the sde can be measured in two ways , namely strong and weak convergence .
strong convergence measures the accuracy on the basis of individual realizations .
the weak convergence measures the accuracy of numerical methods to sdes in case where the goal is to ascertain the probability distribution .
for example , the euler maruyama method has strong order of convergence @xmath113 . for more details ,
see refs .
@xcite . ) with parameters @xmath114 and @xmath115 on the time interval @xmath116 $ ] . ] to compute the expectation of the occupation times given by ( [ e : occtime ] ) , we have to evaluate integrals of the type @xmath117 where @xmath118 .
note that @xmath119 and @xmath120 for @xmath121 .
hence , in the limit @xmath121 , e.g. for @xmath122 and @xmath123 , the integral gets singular .
this together with the unbounded domain of integration poses some numerical difficulties , which can be overcome by splitting the integral : the part close to the asymptotic singularity at @xmath124 , and intermediate part and the part close to infinity .
we introduce @xmath125 and @xmath126 and rewrite @xmath127 the second integral @xmath128 does not pose any numerical difficulties and can easily be computed using simpson s rule .
however , we have to take care about the first and the third part . in @xmath129
, we replace the error function by its quadratic taylor polynomial @xmath130 at @xmath131 , and get @xmath132 the choice of @xmath133 depends upon the desired accuracy of the above approximation .
[ lm : lemmaintestim ] choosing @xmath134{s_0(s_0 ^ 2-c^2)\varepsilon}$ ] for a given tolerance @xmath135 , we get an approximation error @xmath136 . @xmath137 where we introduce @xmath138 and estimate the integral using the mean value theorem .
hence @xmath139 for the given choice of @xmath133 , i.e. @xmath140{s_0(s_0 ^ 2-c^2)}$ ] we get @xmath141 in the third part of the integral , we replace the error function by its limit @xmath142 and get @xmath143 choosing @xmath144 yields an approximation of the error function of less than @xmath145 .
figure [ fig : optfordiflamsig_1 ] shows the expected occupation times obtained using either eqn .
( [ e : occtime ] ) or monte - carlo simulations . in the setting that we have shown here ,
both computational methods yield indistinguishable results .
= [ -0.1 , 0.1]$ ] , time horizon @xmath146 and parameters @xmath147 and @xmath148 .
computations are carried out using either the monte carlo method or eqn .
( [ e : occtime ] )
. ] for large values of @xmath149 , the function @xmath150 in ( [ e : erfint ] ) gets hard to evaluate numerically . for @xmath151 , we obtain @xmath152 ; the usual machine precision .
hence , we limit ourselves to @xmath153 . furthermore , in case of @xmath154 , we obtain @xmath155 as well as @xmath156 .
therefore the splitting of the integration introduced above does not resolve the problem with singularity at the bounds of the integration interval .
estimating the parameters of the ornstein uhlenbeck process given by ( [ e : sde_ou_sim ] ) based on eqn .
( [ e : occtime ] ) for the occupation time is the main objective of this paper .
therefore , we formulate an optimization problem which we use to estimate the parameters .
let @xmath157 denote the ornstein uhlenbeck process with parameters @xmath16 and @xmath17 as in eqn .
( [ e : sde_ou_sim ] ) .
let a spatial interval @xmath9 $ ] be given fixed
. then we define @xmath158 } ( \lambda , \sigma):= \mathbb{e}_\mu \left ( m_{t,[a , b]}(x_{\lambda,\sigma}(t ) ) \right)$ ] as the expected occupation time of the ornstein uhlenbeck process for time horizon @xmath48 . _
problem : given different time horizons @xmath159 and intervals @xmath160 $ ] , corresponding occupation times @xmath161 for @xmath162 and @xmath163 , determine the parameters @xmath164 such that the deviation @xmath165}(\lambda,\sigma)-g_{i , j}\right)^2\ ] ] is minimal .
_ to solve this optimization problem , we apply a standard method from numerical analysis . here , we have used the simplex search method implemented in matlab as the function _
fminsearch _ , see ref . @xcite . as a stopping exit for the optimization
, we use a tolerance of @xmath166 between successive iterations . to demonstrate the parameter estimation procedure ,
consider the following situation .
let @xmath167 and @xmath168 be the time horizons @xmath159 for @xmath169 and @xmath170 $ ] , @xmath171 $ ] , @xmath172 $ ] @xmath173 $ ] be the intervals @xmath160 $ ] for @xmath174 , which use to calculate the corresponding data @xmath161 to the above mentioned optimization problem .
using either the direct equation ( [ e : occtime ] ) or the monte
carlo method , we compute occupation times @xmath161 for the parameters @xmath175 .
now we initiate the minimization procedure providing the corresponding @xmath161s for the both situations separately .
the resulting estimated values are @xmath176 in the case @xmath161 are computed via direct equation ( [ e : occtime ] ) and @xmath177 in the case @xmath161 are computed by the monte carlo methods .
the following table lists more numerical findings that are estimated correspond to different setting . [ cols="^,^,^,^,^,^ " , ] the parameters recoverd from occupation times generated using ( [ e : occtime ] ) ( columns 3 and 4 ) agree better than those recoverd from occupations times generated with the help of monte carlo simulations ( last two columns ) .
this is not surprising , since we used the same underlying equation to generate and to recover the parameters .
however , also for the parameters covered from the monte
carlo simulations we have a difference of about 10@xmath178 between the true and the recovered data .
this is for most applications a sufficient accuracy . nevertheless , increasing the number of samples in monte
carlo simulation we can improve the accuracy of the recomputed parameters .
we derived an equation to compute directly the expected occupation time of the centered ornstein uhlenbeck process .
this allows to identify the parameters of the ornstein uhlenbeck process for available occupation times via a standard least squares minimization . to test our method
, we generated occupation times and recovered the parameters with the above mentioned procedure . within the range of our numerical experiments
, we found very good agreement .
this gives hope to be able to estimate parameters in industrial fleece production processes from measurable quantities like the mass per area .
however , to get closer to the industrial applications we have to extend our method to the @xmath179case and more involved processes than the standard ornstein uhlenbeck process .
w. bock would like to thank the department of mathematics for the financial support .
the daad ( german academic exchange service ) is gratefully acknowledged for providing a scholarship to u. p. liyanage .
berezansky , y. m. kondratiev , y. g. ( 1995 ) . _ spectral methods in infinite - dimensional analysis .
dordrecht : kluwer academic publishers .
translated from the 1988 russian original by p. v. malyshev and d. v. malyshev and revised by the authors bonilla , l.l .
gtz , t. klar , a. marheineke , n. wegener , r. ( 2008 ) .
hydrodynamic limit of a fokker
planck equation describing fiber lay - down processes .
siam j. appl .
vol.68 , no.3 p.648665 gtz , t. klar , a. marheineke , n. wegener , r. ( 2007 ) . a stochastic model for the fiber lay - down process in the nonwoven production .
siam j. appl .
vol.67 , no.6 p.17041717 hida , t. ( 1980 ) .
_ brownian motion .
_ new york : springer - verlag hida , t. kuo , h .- h .
potthoff , j. streit , l. ( 1993 ) .
_ white noise .
an infinite dimensional calculus _
dordrecht , boston , london : kluwer academic publisher higham , d.(2001 ) .
an algorithmic introduction to numerical simulation of stochastic differential equations .
siam rev . vol.43 . nr.3 .
p.525546 kloeden , p.e .
platen , e .
_ numerical solution of stochastic differential equations_. springer kondratiev , yu.g .
leukert , p. potthoff , j. streit , l. westerkamp , w.(1996 ) . generalized functionals in gaussian spaces : the characterization theorem revisited .
. anal . vol.141 .
p.301318 lascheck , a. leukert , p. streit , l. westerkamp , w. ( 1994 ) .
more about donsker s delta function .
soochow journal of mathematics .
vol.20 . nr.3 .
p.401418 mathworks ( 2007 ) .
_ matlab 7 function reference_. the math works inc .
vol.2 ksendal , b. ( 2007 ) .
_ stochastic differential equations : an introduction with applications_. springer , sixth edition .
potthoff , j. streit , l. ( 1991 ) .
a characterization of hida distributions . j. funct . anal . vol.101 .
p.212229 westerkamp , w. ( 1995 ) .
recent results in infinite dimensional analysis and applications to feynman integrals .
phd thesis .
university of bielefeld | we derive an equation to compute directly the expected occupation time of the centered ornstein uhlenbeck process .
this allows us to identify the parameters of the ornstein uhlenbeck process for available occupation times via a standard least squares minimization . to test the method
, we generate occupation times via monte carlo simulations and recover the parameters with the above mentioned procedure . parameter estimation from occupation times + w. bock + _ functional analysis and stochastic analysis group , + department of mathematics , + university of kaiserslautern , 67653 kaiserslautern , germany + _ t. gtz + _ mathematical institute , + university of koblenz , universittsstr . 1 , 56070 koblenz , germany + _ m. grothaus + _ functional analysis and stochastic analysis group , + department of mathematics , + university of kaiserslautern , 67653 kaiserslautern , germany + _
u. p. liyanage + _ technomathematics group , + department of mathematics , + university of kaiserslautern , 67653 kaiserslautern , germany + _ * keywords :* white noise analysis ; ornstein uhlenbeck process ; occupation time ; parameter estimation . + * msc 2010 :* 60k30 ; 65c20 |
a new class of organic - inorganic hybrid perovskite materials , such as ( ch@xmath0nh@xmath0)pbi@xmath0 , has recently emerged as an alternative choice for photovoltaic devices and immediately drew enormous attention @xcite .
( ch@xmath0nh@xmath0)pbi@xmath0 possesses a combination of characteristics that make it attractive for solar cell applications including a favourable band gap of 1.6 ev , efficient optical adsorption , high level of mobility for charge carries of both polarities and exceptionally long carrier lifetime @xcite .
device characteristics , such as the power conversion efficiency , can be further enhanced by using mixed halide perovskites , e.g. , ( ch@xmath0nh@xmath0)pb(cl@xmath1i@xmath2)@xmath0 @xcite .
the demonstrated energy conversion efficiency of nearly 20% @xcite combined with low - temperature solution processing technology hold a promise to lead the way towards low - cost solar energy .
the crystal structure of ( ch@xmath0nh@xmath0)pbi@xmath0 exhibits several polymorphs . at low temperatures ,
an orthorhombic @xmath3-phase prevails up to 161 k @xcite .
a tetragonal @xmath4 structure is stable in the intermediate temperature range of 161330 k. the high - temperature cubic @xmath5-phase exists above 330 k @xcite .
electronic structure calculations performed for the cubic phase reveal a direct band gap of approximately 1.3 ev at the r@xmath6 point of the brillouin zone @xcite .
the direct gap moves to @xmath7 as the cubic structure transforms into the tetragonal phase @xcite , which has a more complex unit cell with larger volume .
the transformation is accompanied by widening of the band gap , which reaches the value of 1.6 ev as determined by photoluminescence measurements @xcite and confirmed by _
ab initio _ calculations using hybrid functionals and quasiparticle gw approaches @xcite .
further analysis of the band structure evolution in the course of the transition from high- to low - symmetry structure is hindered by the brillouin zone folding resulted from the increased size of the unit cell for @xmath4-phase . here we show that a direct comparison of the electronic structure between @xmath5 and @xmath4 phases of ( ch@xmath0nh@xmath0)pbi@xmath0 can be achieved by projecting the band structure of @xmath4-phase onto the brillouin zone of the cubic structure , which represents an idealized version of low - symmetry phases .
this analysis can facilitate answering the following questions : ( i ) to which extent does the bloch character of band edges in the tetragonal phase inherit the r - character of the cubic phase , and ( ii ) what is the effect of dynamic disorder in the orientation of organic groups or substitutional alloying , e.g. , in the mixed halide ( ch@xmath0nh@xmath0)pb(cl@xmath1i@xmath2)@xmath0 perovskite , on the bloch states at the band edges ?
an uncertainty in the bloch character of states near to the band edges in semiconductors can be used to reveal deterioration of charge transport characteristics due to carrier localization @xcite , which can adversely affects the performance of solar cells .
the first - principles calculations were carried out using density functional theory ( dft ) @xcite . `
abinit ` package @xcite and projector augmented - wave method were used @xcite with pseudopotentials adopted from a ` jth ` library ( version 0.2 ) @xcite .
@xcite construction of the generalized gradient approximation ( gga - pbe ) was employed for the exchange correlation functional .
completeness of the plane wave basis set was controlled though the cut - off energy , which was set at 15 ha ( 408 ev ) as determined from convergence studies .
the lattice constants of ( ch@xmath0nh@xmath0)pbi@xmath0 in pseudocubic phase ( fig .
[ fig : a](a ) ) was optimized self - consistently by simultaneous minimization of stresses and hellmann - feynman forces acting on atoms below 1 mha / bohr .
the brillouin zone was sampled using @xmath8 @xcite k - mesh .
the calculated lattice constant of 6.47 is slightly overestimated as compared to the experimental value of 6.28 @xcite .
this result is consistent with typical trends for modelling of solids with gga - pbe exchange - correlation energy functional @xcite .
nh@xmath0)pbi@xmath0 ( a ) and the corresponding relativistic band structure ( b ) along high - symmetry path in the brillouin zone shown in the inset.,title="fig:",scaledwidth=30.0% ] + the structure of tetragonal phase was constructed by repeating the cubic cell with the multiplicity of @xmath9 ( fig . [
fig : b](a ) ) . the octahedral tilting and orientation of methylammonium molecules
were initially set based on previously reported structures of @xmath4-(ch@xmath0nh@xmath0)pbi@xmath0 resolved experimentally and theoretically @xcite .
then the supercell was fully relaxed .
calculated lattice parameters for tetragonal phase @xmath10 and @xmath11 are within 4% deviation from the corresponding experimental values at room temperature @xmath12 and @xmath13 @xcite .
it should be mentioned that our tetragonal supercell differs from a conventional tetragonal unit cell by 45@xmath14 rotation .
the construction of the supercell by translation of the parent ( cubic ) structure along its lattice vectors allows us later to unfold the band structure and present the electron wave vectors in terms of the cubic reciprocal lattice .
+ unfolding of the band structure of supercells was performed by calculating the bloch spectral weight using ` fold2bloch ` package @xcite implemented in ` abinit ` . `
vesta ` 3 package @xcite was used for atomic structure visualization .
we begin the analysis of the electronic structure of ( ch@xmath0nh@xmath0)pbi@xmath0 with the high - symmetry pseudocubic structure . the calculated relativistic band structure is shown in fig . [ fig : a](b ) .
the band gap is direct near r@xmath6 point of the brillouin zone .
the magnitude of the band gap is underestimated by approximately 1 ev in comparison to the experimental value of 1.6 ev @xcite , which is a typical result for dft calculations of ( ch@xmath0nh@xmath0)pbi@xmath0 when the bear gga - pbe exchange - correlation functional is combined with spin - orbit coupling .
the agreement can be further improved by using a modified becke - johnson potential @xcite , however it is not required here due to the comparative nature of our study .
a rashba - like breaking of the spin degeneracy @xcite is observed at the band edges ( particularly strong in the conduction band ) .
@xcite reported a different chirality of the spin texture at the top valence band and the bottom conduction bands that results in the long carrier lifetime due to constrains imposed by the spin section rules . in order to elucidate origin of the rashba splitting ,
the band structure calculation was performed for cspbi@xmath0 in the pseudocubic structure ( not shown here ) . atomic position and lattice perimeters
were kept at the equilibrium values of ( ch@xmath0nh@xmath0)pbi@xmath0 , except for the methylammonium that was substituted by cs .
the band gap and the rashba splitting near to the band edges of the pseudocubic cspbi@xmath0 was found to be identical to those in ( ch@xmath0nh@xmath0)pbi@xmath0 .
furthermore , the splitting is fully suppressed in the centrosymmetric cubic cspbi@xmath0 .
this result suggests that precursors for the rashba splitting are the strong spin - orbit interaction and distortions in the pb - centred octahedron caused by the dipolar nature of ch@xmath0nh@xmath15 ions @xcite .
a significant reduction of the rashba splitting observed in ( nh@xmath16)pbi@xmath0 perovskite structures @xcite confirms validity of this argument .
the cubic structure is attractive for modelling due to its simplicity .
however , we should keep in mind that the pseudocubic structure is an idealization ; the room temperature properties are governed by the tetragonal phase shown in fig .
[ fig : b](a ) .
distinctive features of the tetragonal structure are the presence of an octahedral tilting and ch@xmath0nh@xmath15 cation rotation .
this leads to the question as to which extend the discussion of the electronic structure of the pseudocubic phase is applicable to the tetragonal structure ? in order to enable direct comparison between two electronic structures , the band structure of low - symmetry tetragonal phase need to be mapped onto the pseudocubic brillouin zone .
this can be achieved by unfolding " the supercell band structure . since
the tetragonal structure comprises of @xmath17 pseudocubic cells , each eigenstate @xmath18 in the supercell gives rise to eight k - points in the pseudocubic brillouin zone .
for instance , @xmath7 unfolds into a set of eight k - points , three of which are obtained by permutations of @xmath19 , another three additional k - points are permutations of @xmath20 , one k - point is at @xmath6 and the @xmath21-point self .
it is evident that @xmath21-point in tetragonal structure is coupled to r - point in the pseudocubic structure .
however , it is not possible to say _ a priori _ what will be the strength of r - representation .
each unfolded k - point is assigned its own bloch spectral weight @xmath22 .
the whole set satisfies normalization @xmath23 .
the spectral weight us used as an augmentation for the band structure plot @xmath24 and allows to filter out spurious bands .
the spectral weight is computed by projecting the supercell wavefunction @xmath25 onto a set of plane waves that are compatible with the wave vector @xmath26 and reciprocal lattice vectors of the pseudocubic brillouin zone .
details of this procedure are widely discussed in the literature @xcite . the effective band structure for tetragonal phase is shown in fig . [ fig : c](a ) .
the band gap increases by approximately 0.1 ev when compared to the pseudocubic structure .
the conduction and valence band edges of tetragonal phase retain 90 and 84% of the pseudocubic r - character , respectively .
the @xmath21-character of the corresponding states is only 10% at most .
the tetragonal phase exhibits a similar magnitude of the rashba splitting as in the pseudocubic structure ( fig .
[ fig : a](b ) ) .
+ now as the basis for comparison of the electronic structure for high- and low - symmetry phases is established , we can proceed with disordered structures . a low energy barrier for rotation of the ch@xmath0nh@xmath15 ion and occurrence of multiple local minima @xcite leads to discretized reorientation dynamics of the c
n axis in ( ch@xmath0nh@xmath0)pbi@xmath0 at the room temperature @xcite .
the model for disordered orientation of methylammonium molecules was generated from @xmath4-(ch@xmath0nh@xmath0)pbi@xmath0 structure ( fig .
[ fig : b](a ) ) by flipping the molecules 180@xmath27 randomly with subsequent relaxation of atomic coordinates .
the obtained structure is shown in fig .
[ fig : b](b ) .
it should be noted that the total energy of the disordered structure was even lower in comparison to the unperturbed tetragonal phase . the band structure of a disordered tetragonal phase
is presented in fig .
[ fig : c](b ) .
the disorder does not lead to significant changes in the band gap .
the band edges retain their r - character at the level of @xmath28% , which is almost identical to that in the unperturbed tetragonal phase .
thus we can conclude that the orientational disorder of organic groups does not introduce localization effects at the band edges and does not hamper transport of charge carriers .
this interpretation is based on the relation between the ambiguity of the bloch character in reciprocal space and localization in the real space through the fundamental uncertainty principle .
the only noticeable effect of the orientational thermal disorder is the reduced rashba splitting , which can make the spin selection rule , which promotes dissociation of optical excitations , less efficient .
finally , we turn the discussion to a compositional disorder in mixed halide perovskite structures .
even though the cl incorporation in an iodide - based structure is relatively low ( below 34% ) @xcite , the recombination time and diffusion length of optical excitations are prolonged by 12 orders of magnitude @xcite .
the origin for a superior performance of ( ch@xmath0nh@xmath0)pb(cl@xmath1i@xmath2)@xmath0 perovskite thin film solar cells remains an open question .
the plausible scenarios include cl@xmath29 ions involved in the following functions : facilitation the release of excess ch@xmath0nh@xmath15 during annealing @xcite , passivation of grain boundaries and the interface with the substrate @xcite , creation of an in - homogeneous built - in electric field profile similar to p - i - n junction @xcite , and reduction of the carrier effective mass in structures with high cl content @xcite . in order to model a dilute substitutional ( ch@xmath0nh@xmath0)pb(cl@xmath1i@xmath2)@xmath0 alloy , one of 24 i
- atoms in the supercell was replaced by cl that resulted in the effective composition of @xmath30 .
the structural relaxation in the vicinity of chlorine reveals asymmetric pb - cl bonds @xmath31 ( see fig . [
fig : b](c ) ) .
the pb - cl bond lengths of 2.75 and 3.48 deviate significantly from the pb - i equilibrium bond length of 3.25 .
the shorter pb - cl bond length agrees with the sum of pb@xmath32 and cl@xmath29 ionic radii , 0.94 and 1.81 respectively @xcite .
the asymmetric configuration could be an efficient way to mediate local strain due to the large size mismatch between cations and can be a driving force for segregation of cl at interfaces .
the effective band structure of ( ch@xmath0nh@xmath0)pb(cl@xmath33i@xmath34)@xmath0 is shown in fig .
[ fig : c](c ) .
incorporation of chlorine in dilute quantities does not change the band gap , which is in accord with observations made by @xcite .
the rashba splitting and the bloch character of the valence and conduction band edges remain intact with the undopped ( ch@xmath0nh@xmath0)pbi@xmath0 .
perturbations due to cl are observes for states _ within _ the conduction and valence bands as pointed out by arrows in fig .
[ fig : c](c ) .
since the access energy required to reach perturbed states is greater than the thermal energy , those perturbations in the band structure will unlikely have an adverse effect on transport coefficients of the mixed halide perovskites .
the relativistic band structure of low - symmetry @xmath4-(ch@xmath0nh@xmath0)pbi@xmath0 tetragonal phase is mapped onto the pseudocubic brillouin zone using an unfolding technique . in spite of significant differences in the structure ( the presence of an octahedral tilting and ch@xmath0nh@xmath15 cation rotation )
, the band structures of both phases share many similarities .
a strong coupling is observed between @xmath21-point in tetragonal structure and r - point in the pseudocubic structure .
the conduction and valence band edges of tetragonal phase retain 8090% of the pseudocubic r - character .
the @xmath21-character of the corresponding states is very weak in contrast to the common expectation .
both pseudocubic and tetragonal structures exhibits a similar magnitude of the rashba splitting , which plays an important role in the long carrier lifetime due to constrains imposed by the spin section rules . at room temperature
, the rashba splitting can be reduced due to an orientational disorder of ch@xmath0nh@xmath15 cations . at the same time
, the disorder neither affects the band gap nor reduces the r - character of the band edges and thus does not hamper transport of charge carriers .
mixed halide ( ch@xmath0nh@xmath0)pb(cl@xmath1i@xmath2)@xmath0 structures with dilute concentration of chlorine reveal existence of asymmetric pb - cl bonds as a way to mediate local strain due to the large size mismatch between cations that can also be a reason for segregation of cl at interfaces .
the band gap , rashba splitting and the bloch character of the valence and conduction band edges in perovskite structures with dilute substitutional chlorine remain intact with the undopped ( ch@xmath0nh@xmath0)pbi@xmath0 .
perturbations due to the incorporated chlorine are observes for states located energetically far from the band edges .
this rules out the possibility of adverse effect of chlorine substitutional incorporation on transport coefficients of the mixed halide perovskites .
the work was supported by the natural sciences and engineering research council of canada under the discovery grant program rgpin-2015 - 04518 .
computational resources were provided by high performance computing centre at lakehead university and compute canada . | the high - symmetry pseudocubic init cell is often used for modelling the electronic structure of halide perovskites .
however , direct comparison of the band structure with more realistic low - symmetry phases is impeded by the zone folding .
we utilize a spectral density technique in order to reduce the supercell band structure to a common bloch basis .
this allows us to compare the electronic structure of high- and low - symmetry phases as well as investigate effects of structural and compositional disorder on states near to the band edges that govern transport , dissociation and recombination of optical excitations . |
the nature of the star hv 2112 , located in the same part of the sky as the small magellanic cloud ( smc ) , has recently become the center of a vivid debate .
levesque et al .
( 2014 ) showed that it exhibits anomalous spectral features suggesting enhanced abundances of molybdenum , rubidium , lithium , calcium and possibly also potassium when compared to other smc supergiants .
the high inferred bolometric luminosity and the enhanced abundances of the first three elements lead to the intriguing suggestion by levesque et al .
( 2014 ) that hv 2112 could be a massive `` thorne - ytkow object '' ( to ) , the hypothesized red - giant like stellar objects with neutron stars inside their cores .
tos , first modeled by thorne & ytkow ( 1975 , 1977 ) , may form as a result of unstable mass transfer in a massive x - ray binary after the neutron star is engulfed in the envelope of its companion star ( taam et al .
alternatively , these objects may form when the supernova kick of newly formed neutron star accidentally sends it towards and into its companion ( leonard et al .
1994 ) . whether tos actually form is a matter of debate .
fryer et al ( 1996 ) and chevallier ( 1993 ) argued that the neutron star would accrete enough material to collapse to a black hole before forming a stable to .
biehle ( 1991 , 1994 ) and cannon ( 1993 ) predict that massive tzos , if stable , are powered by nuclear burning through the exotic interrupted rapid proton ( irp)-process .
therefore they may be identified by enhanced abundances of rare proton - rich elements , including mo and rb . also enhancements of li
have been predicted ( podsiadlowski et al .
the predictions triggered several ( proposed ) observing campaigns hoping to identify massive tzos ( van paradijs et al .
1995 , kuchner et al .
2002 , levesque et al .
so far hv 2112 is the most promising candidate , even though the abundances of rb and mo are not as high as expected and the enhancement of ca and k was not predicted .
tout et al .
( 2014 ) discuss an alternative explanation where hv 2112 is a `` super asymptotic giant branch ( agb ) star '' .
this is the late evolutionary phase of stars with initial mass around 8@xmath3 , massive enough to ignite carbon , but still low enough to experience thermal pulses and dredge up episodes typical for agb stars .
super agb stars have higher core masses than typical red giants , giving them luminosities as high as red supergiants and tosin their early phases ( siess 2010 ) .
similar to agb stars they can , in principle , produce mo and rb through the slow neutron capture process ( karakas & lattanzio 2014 ) and li through the cameron & fowler ( 1971 ) mechanism .
enhancement of ca or k is , however , not predicted . in both scenarios
hv 2112 is assumed to be member of the smc .
this view is supported by its projected location in the sky and the kinematic properties : the measured radial velocity ( rv ) is similar to that of the smc ( neugent et al .
however , as levesque et al .
( 2014 ) state , the possibility that it may be a halo giant with a similar rv can not be ruled out , but it would require a novel explanation for the peculiar spectral features . in this paper , we report the detection of a very large proper motion based on data from the southern proper motion survey ( girard et al . 2011 ) .
if confirmed , the proper motion excludes membership in the smc , making hv 2112 a foreground milky way star ( likely a giant due to its pulsational properties ) .
a location in the milky way would rule out both proposed scenarios , the to - nature and sagb - nature of hv 2112 .
we propose a new scenario which involves pollution by the wind of a former massive agb companion star . in this manner ,
the enhancement of molybdenum , rubidium , and possibly lithium comes from the processes that would enrich a massive agb star . by now the former companion has lost its entire envelope .
its remnant , most likely a massive white dwarf , is expected to be still around in a wide orbit .
hv 2112 in this scenario is the initially less massive secondary star .
this scenario is similar to the scenario proposed by iben & renzini ( 1983 ) to explain extrinsic s stars as well as ba and ch stars ( mcclure 1984 ) .
one of the very attractive features of this scenario ( and any scenario were it is placed in the galactic halo ) is the natural explanation provided for the ca ( and k ) features .
these elements appear enhanced when compared to smc giants , but they are normal for the galactic halo where [ ca / fe ] enhancements are typical .
we discuss various testable predictions of this scenario .
the star hv 2112 has had its proper motion measured as part of the southern proper motion survey .
this is a systematic astrometric study with a baseline of approximately four decades covering nearly the entire southern sky down to approximately @xmath4 ( girard et al .
2011 ) . for hv 2112 ,
the first epoch was obtained in 1972 and the second in 2007 .
the star shows a proper motion of [email protected] mas / year in right ascension and [email protected] mas / year in declination .
while the star is strongly variable , there are two reasons to believe that the proper motion measurements from spm 4.0 are reliable .
first , in both epochs , it was among the brightest stars in its general vicinity within two arcminutes , only one star is brighter in the second epoch v band ccd data , and only two stars are brighter in the first epoch b band photometry
. additionally , within 2 , there are only 8 stars brighter than @xmath6 , so the probability of finding a star within 10 " which is even 10% as bright as hv 2112 is about @xmath7 .
second , with the 35 year baseline , the total astrometric shift measured is about 0.35 arcseconds .
this shift is sufficiently large that it would require extreme fine tuning to produce an astrometric shift due to the combination of blending and source variability without the second object being apparent in the images , and as shown , the probability of such a star being that nearby is of order @xmath7 . the ucac4 catalog ( zacharias et al .
2013 ) often gives an additional useful second epoch for proper motions to that from the spm , although ucac4 and spm make use of the same first epoch data . for this object , however , the second epoch from ucac caught this variable star in a relatively faint state , so that the positional uncertainty is several times as large as in the spm data .
the ucac4 proper motion estimate is @xmath8 mas / yr in right ascension and @xmath9 mas / yr in declination .
if one takes a weighted average of the two measurements , then proper motion is @xmath10 mas / year in right ascension and @xmath11 mas / year in declination .
this differs from zero at roughly 3.8@xmath12 , and from the -1.1 mas / yr in declination of the smc ( kallivayalil et al .
2013 ) by @xmath13 .
the deviation from the proper motion of the smc in right ascension , which is -0.8 mas / yr ( kallivayalil et al .
2013 ) is statistically insignificant , being about 1.7@xmath12 . averaging spm and ucac using weights determined solely by their statistical uncertainties is a conservative approach , since any systematic uncertainties should be far more serious for the fainter ucac4 epoch than for the spm data , and the spm measurement indicates a larger proper motion than does the ucac4 measurement .
we have also checked for systematics in the astrometry of the field by taking the hipparcos proper motion measurements for stars within 2 degrees of hv 2112 and cross - correlating them against the spm 4.0 measurements .
we have found that there is no systematic offset at a level of about 0.5 mas / year , and that the differences between the spm and hipparcos measurements are comparable to those expected from the stated statistical errors . we have test whether the internal errors for the spm data are comparable to what is stated in the spm catalogue . using the topcat package , we have taken all the spm stars within 1.5 degrees of hv 2112 .
we have then filtered to include only stars whose 2mass colours ( cutri et al . 2003 ) place them on one of the giant branches for the smc and which have @xmath14 between 10 and 14 ( i.e. a magnitude range for which we expect the errors to be similar to those for hv 2112 ) , and then examined only those stars with proper motions of 5 masec / year or less .
we find that for these 199 stars , the standard deviation in the values of the proper motions in both dimensions is about 90% of the mean values of the uncertainties on the proper motions from spm 4.0 .
the expectation , if all the stars had measurement uncertainties of the mean uncertainty , and there were no interloper stars , would be that the standard deviation would be about 75% of the 1@xmath12 uncertainty ( the expectation is not exactly @xmath15 because the more extreme outliers will be eliminated by the 5 mas / year cut ) .
it may then be that the measurement uncertainties are underestimated by about 20% , and that the significance of the proper motion measurement relative to the smc proper motion is at only the 3.0@xmath12 level , but this gives us additional confidence that there are no major systematic uncertainties in spm 4.0 that bias our results .
any statistically significant measurement of a proper motion in data with the precision of spm 4.0 immediately provides a strong argument against a location of the star in the small magellanic cloud , because the statistical uncertainties on the proper motions are larger than the escape velocity from the smc . the proper motion in @xmath16 of 10 mas / year at a distance of 62.1 kpc ( graczyk et al 2014 ) yields a space velocity of 3100 km / sec , well above the escape velocity of the smc .
if one instead takes this to be the velocity of a halo star in the milky way halo , and assumes a 150 km / sec speed ( a reasonable value given the star s radial velocity of 157 km / sec levesque et al . 2014 ) , then it is likely to be located at a distance of about 3 kpc , although the statistical errors on the proper motion measurement allow for a wide range of possible distances .
both the space velocity , and the height , probably a few kpc below the galactic plane , given the galactic latitude of 45 degrees , suggest that this is a halo star rather than a disk star in the milky way .
= 10 to @xmath17 and @xmath18 colors consistent with the giant branch for metal poor stars ( from @xmath18=0.78 to 1.35 ) .
hv 2112 has the black circle around it in the figure .
it can be seen to be an outlier in the proper motion diagram , with the main locus of points toward the center being the smc stars , and the outliers being foreground stars.,width=8 ]
at a distance of a few kpc , the luminosity of this object is likely to be @xmath19 , rather than the @xmath20 it would have in the smc .
furthermore , is has been classified as a mira - like variable from photometry ( samus et al .
2012 ) , and its emission lines in its bright phases are also consistent with mira phenomenology ( levesque et al . 2014 ) .
the hypothesis of a mira star also fits well with the inferred luminosity for the closer distance .
these points were already noted in levesque et al .
( 2014 ) , but rejected on the basis of the then - likely association with the smc .
levesque et al . ( 2014 ) did briefly , but presciently , note that future kinematics observations could potentially place the object in the milky way , and that if they did , this would lead to a different interpretation than the thorne - ytkow object interpretation pursued in their paper .
likewise , the super asymptotic giant branch ( sagb ) star interpretation considered by tout et al .
( 2014 ) also can not explain the object if it is located in the galactic foreground due to brightness considerations . on the other hand ,
the abundances may be explained by a scenario in which matter is transferred from a massive agb primary star to a binary companion . eventually , the secondary will evolve off the main sequence and become an agb star itself , albeit of significantly lower mass , allowing such an object still to exist in the old population in the halo .
these stars are generally referred to as extrinsic s stars ( e.g. iben & renzini 1983 ) , following an initial purely phenomenological classification of s stars ( merrill 1922 ) , which were later found , in a mnemonically fortunate coincidence to be rich in s - process elements ( smith & lambert 1986 ) .
the primary must be massive enough to produce substantial enhancements of molybdenum and rubidium .
this certainly is expected for super - agb stars ( lau et al .
2011 ; tout et al .
2014 ) , but is likely to happen down to even 4 @xmath21 ( lugaro et al .
2003 ) , and is clearly seen in some s stars ( e.g. lambert et al .
1995 show that rb abundances increase with increasing s - process elemental abundances , while allen & barbuy 2006 show an excess of molbydenum in most of the members of a large sample of barium stars . ) .
the extrinsic s stars form through the binary evolution channel above , while the intrinsic s stars are formed by dredge - up of s - process elements , and show enhancement of tc .
the tc signature indicates that the material has been placed into the stellar atmosphere on a timescale of order or less than the radioactive decay time of tc .
the placement of the star in the milky way halo also helps to explain the calcium abundance anomaly that could not be explained by the super - agb star scenario discussed in tout et al .
( 2014 ) nor by the tzo scenario of levesque et al .
levesque et al . (
2014 ) did not have a calibrated estimate of the ca abundance of the star , but did show that the ratio of strengths of ca lines to fe lines was a factor of about 3 stronger than that for red supergiants in the smc .
this is indicative of a [ ca / fe ] ratio about 3 .
kobayashi et al . ( 2006 ) shows that halo stars in the solar neighborhood with [ fe / h]@xmath22 tend to have [ ca / fe ] of about 0.4 , which indicates a calcium to iron abundance ratio about 2.5 .
the most metal rich ( and presumably youngest ) smc giants have ratios of @xmath23 element abundances to iron abundances quite similar to solar ( mucciarelli 2014 ) .
relatively little discussion exists about the potassium excess in hv 2112 , but it is of similar size to the ca excess judging from the plots in levesque et al .
( 2014 ) , and this , also , can be explained well in terms of empirical halo abundances ( kobayashi et al .
2006 ) .
lithium is a fragile element and is easily destroyed at temperatures in excess of about @xmath24k .
even a modest depth of mixing can lead to efficient depletion ( stancliffe 2009 ) . in principle , lithium can be in the agb phase of the primary star ( cameron & fowler 1971 ) and accreted onto the secondary .
however , all lithium originating from the primary has most likely been destroyed when hv 2112 evolved to become an agb star and developed a deep convective envelope .
the present day lithium abundance is more likely the result of production in situ , in hv 2112 itself .
low mass agb stars can produce lithium as a result of thermal haline mixing ( stancliffe 2010 ) .
an alternative possibility is that the white dwarf left behind accretes enough material to produce a substantial nova rate .
evidence for lithium production has been found in the classical nova v339 ( tajitsu et al .
this could pollute the envelope of hv 2112 .
the model predictions remain uncertain , but lithium is not uncommon in symbiotic mira s .
for example v407 cyg shows a strong lithium excess ( tatarnikova et al .
radial pulsations have periods which scale with the dynamical timescale of the star that is , @xmath25 , where @xmath26 is the period of oscillation , and @xmath27 is the density of the star .
mira variables have pulsation period of about 140 - 700 days .
their masses tend to be @xmath28 and their radii tend to be about 300 @xmath29 .
the typical variations in radius are a factor of about 1.5 ( e.g. mahler et al .
because the mass of the thorne - ytkow object could be quite large , it is viable for such a star to have a pulsational period of about 600 days caused by the same pulsation mechanism as the classical miras , even though its radius would be much larger than a classical mira .
on the other hand , the radial velocity amplitude of the pulsations should be quite a bit larger if the object is at smc distances than if it is a foreground galactic mira variable .
in essence , one can apply the baade - wesselink method ( baade 1926 ; wesselink 1946 ) to estimate the radius and hence distance to the object . at smc distances
, the radius of the star should be about 1500 @xmath29 , so a variation of 40% about the mean radius is a variation of 600 @xmath29 .
the pulsations of miras in general , and hv 2112 in particular are such that most of the variation takes place in a relatively small range of pulse phase .
if we assume that such a size variation takes place in 0.2 times the pulse period , or 120 days , then we expect a velocity of expansion of about 50 km / sec .
this speed is not large enough to be inherently problematic for either scenario discussed here , but does provide a means for estimating the stellar radius , and hence the distance to the star if a radial velocity monitoring campaign is undertaken .
we can make a crude estimate of the number of extrinsic s stars in the galaxy with a massive agb progenitors as part of testing the plausibility of the scenario .
s stars make about 4% of giants in the temperature range where m stars are normally found , and extrinsic s stars make about 1/3 of s stars ( van eck & jorissen 2000 ) .
extrinsic s stars are thus about 1% of the number of stars that appear to be m giants in color space .
the smc spans about @xmath30 of the solid angle of the sky .
the milky way has @xmath31 stars , of which about @xmath32 are in the halo , and about @xmath33 are halo m giants .
there should then be approximately 30 extrinsic s stars in the milky way halo projected against the smc .
if we further assume that only the extrinsic s stars whose progenitors were more massive than 4 @xmath21 can have the required abundances of rubidium and molybdenum , we must correct for the fact that only 11% of the stars between 1 and 10 @xmath21 ( i.e. the ones that produce agb stars without undergoing supernovae ) are above 4 @xmath21 ( and hence are heavy enough for production of these elements ) .
we then expect a few extrinsic s stars with massive enough progenitors in the smc field . since we additionally require that they be in an agb state themselves to explain the mira variations and the li production , the expected number falls by a factor of about 30 ( girardi et al . 2010 ) ,
so our expected number of objects is about 0.1 .
the scenario we propose thus quite plausibly places a source of the type we invoke in a location projected against the smc , while also not placing so many such objects projected against the smc that we would have expected to have found many of them already .
the scenario makes several additional testable predictions .
first and foremost , the object should be in a wide binary system .
orbital periods of extrinsic s stars are generally at least 600 days ( jorissen & mayor 1992 ) .
the white dwarf in this system should likely be a onemg white dwarf , and hence should be of considerably higher mass that the s star .
nonetheless , the motions should still be of relatively low orbital velocity . assuming a total system mass of about 2 @xmath21 and an orbital period of about 2 years
, the orbital velocity of the optically bright star should be about 20 km / sec , while the orbital velocity of the white dwarf should be about 10 km / sec .
additionally , the system may show some accretion by the white dwarf in the system . in mira itself , for example
, the white dwarf companion of the optically bright star is seen to be accreting at about @xmath34 per year ( sokoloski & bildsten 2010 ) , but the orbital period of mira is about 500 years . if , instead , this system has an orbital period of only a few years , the accretion rate can be expected to be much higher .
to first order , the mass accretion rate from a stellar wind can be estimated to be : @xmath35 since the orbital speed of the donor star and the wind speeds of agb stars are of roughly the same order , and the accretor is heavier than the donor , a substantial fraction of the total wind may be accreted ( bondi & hoyle 1944 ; struck et al .
2004 ; mohamed & podsiadlowski 2012 ) .
given that the wind speed and orbital speed are not known in this system , and the accretion rate depends on the ratio of these two speeds to the fourth power , it is difficult to make a specific prediction here .
if @xmath36 1% of the wind of a typical agb star ( i.e 1% of @xmath37/yr ) is accreted , however , then the system would be expected to be bright in the ultraviolet due to strong accretion .
the system is not detected in galex data , but does show a blue excess in the spectrum of levesque et al .
more targeted searches for an accretion signature in this object are warranted , and should be spread over the @xmath36600 day period of the optically bright star , given that the accretion rate is likely to change with the pulsation phase .
the system does not show x - ray emission , with a 90% upper limit on its luminosity of about @xmath38 in data taken on march 25 , 2013 ( v. antoniou , private communication ) .
relatively few cataclysmic variables have x - ray luminosities higher than this value , and those which do are frequent dwarf nova outbursters ( e.g. britt et al . 2015 ) .
this limit , again , is not particularly constraining , as many symbiotic stars have x - ray luminosities below @xmath39 erg / sec .
we thank the lorentz center for having hosted the workshop on `` the impact of massive binaries throughout the universe '' , where the ideas in this paper were first discussed .
we thank stephen justham , ben davies , lennart van haaften and onno pols for useful discussions , vallia antoniou for sharing the x - ray upper limit in advance of publication , and terry girard for helpful and detailed discussions of possible systematic errors in the southern proper motion survey data .
sdm acknowledges support by the eu horizon 2020 program through a marie sklodowska - curie reintegration fellowship , h2020-msca - if-2014 , project i d 661502 .
99 allen d.m . , barbuy b. , 2006 a&a , 454 , 895 angulo et al . 1999 , nuclear physics a,656 , 3 baade w. , 1926 , an , 228 , 359 biehle g.t . , 1991 , apj , 380 , 167 biehle g.t . , 1994 ,
apj , 420 , 364 bondi h. , hoyle f. , 1944 , mnras , 104 , 273 britt c.t . , et al . , 2015 , mnras , 448 , 3455 cameron a.g.w . , fowler w.a . , 1971 , apj , 161 , 111 cannon r.c . , 1993 , mnras , 263 , 817 chevalier r. , 1993 , apjl , 411 , 33 cutri r.m . , et al . , 2003 , the irsa 2mass all - sky point source catalog , nasa / ipac infrared science archive fryer c.l . , benz w. , herant m. , 1996 , apj , 460 , 801 girard t.m . , et al . , 2011 , aj , 142 , 15 girardi l. , et al . , 2010 , apj , 724 , 1030 graczyk d. , et al . , 2014 , apj , 780 , 59 iben i. , renzini a. , 1983 , ara&a , 21 , 271 jorissen a. , mayor m. , 1992 , a&a , 260 , 115 kallivayalil n. , van der marel r.p .
, besla g. , anderson j. , alcock c. , 2013 , apj , 764 , 161 karakas a.i . , lattanzio j.c .
, 2014 , pasa , 31 , 30 kobayashi c. , umeda h. , nomoto k. , tominaga n. , okhubo t. , 2006 , apj , 653 , 1145 lambert d.l . , smith v.v . , busso m. , gallino r. , straniero o. , 1995 , apj , 450 , 302 lau h.h.b . , doherty c.l . ,
gil - pons p. , lattanzio j.c .
, 2011 , aspc , 445 , 45 kuchner m.j .
, vakil d. , smith v.v .
, lambert d.l .
, plez b. , phinney e.s .
, 2002 , aspc , 263 , 131 leonard p.j.t . , hills j.g . ,
dewey r.j . , 1994 ,
apj , 423 , l19 levesque e.m . ,
massey p. , ytkow a.n .
, morrell n. , 2014 , mnras , 443l , 94 lugaro m. , davis a.m. , gallino r. , pellin m.j . , straniero o. , kppeler f. , 2003 , apj , 593 , 486 mahler t.a . , wasatonic r. , guinan e.f . , 1997 , ibvs , 4500 , 1 mcclure r.d . , 1984 , apj , 208l , 31 merrill p.w . , 1922 , apj , 56 , 457 mohamed s. , podsiadlowski p. , 2012
, baltic astronomy , 21 , 88 mucciarelli a. , 2014 , an , 335 , 79 neugent k.f . ,
massey p. , skiff b. , drout m.r .
, meynet g. , olsen k.a.g . , 2010 , apj , 719 , 1784 samus , n.n .
, et al . , 2012 , general catalog of variable stars , vizier online data catalog : b / gcvs smith v.v . ,
lambert d.l . , 1986 , apj , 311 , 843 sokoloski j. , bildsten l. , 2010 , apj , 723 , 1188 stancliffe r. j. , 2009 , mnras , 394 , 1051 stancliffe , r.j . ,
2010 , mnras , 403 , 505 struck c. , cohanim b.e . , willson l.a .
, 2004 , mnras , 347 , 173 taam r.e .
, bodenheimer p. , ostriker j.p .
, 1978 , apj , 222 , 269 tajitsu , a. , sadakane , k. , naito , h. , arai , a. , aoki , w. , 2015 , nature , 518 , 381 tatarnikova a. a. , marrese p. m. , munari u. , tomov t. , whitelock p. a. , & yudin b. f. , 2003 , mnras , 344 , 1233 thorne k.s .
, ytkow z.n . , 1975 ,
apj , 199l , 19 thorne k.s . ,
ytkow a.n . , 1977 ,
apj , 212 , 832 tout c.a . ,
ytkow a.n . ,
church r.p .
, lau h.b . ,
doherty c.l . ,
izzard r.g . , 2014 ,
mnras , 445l , 36 van eck s. , jorissen a. , 2000 , a&a , 360 , 196 van paradijs j. , spruit h.c . , van langevelde h.j . , waters l.b.f.m . , 1995 ,
a&a , 303l , 25 wesselink a.j . , 1946 ,
ast . int .
, 10 , 91 zacharias n. , finch c.t . , girard t.m . ,
henden a. , batlett j.l . ,
monet d.g . , zacharias m.l . , 2013 ,
aj , 145 , 44 | using the southern proper motion ( spm ) catalog , we show that the candidate thorne - ytkow object hv 2112 has a proper motion implying a space velocity of about 3000@xmath0if the object is located at the distance of the small magellanic cloud .
the proper motion is statistically different from that of the smc at approximately @xmath1 in spm , although the result can drop to about @xmath2 significance by including the ucac4 data and considering systematic uncertainties in addition to the statistical ones .
assuming the measurement is robust , this proper motion is sufficient to exclude its proposed membership of the small magellanic cloud and to argue instead that it is likely to be a foreground star in the milky way halo .
the smaller distance and therefore lower brightness argue against its proposed nature as a thorne - ytkow object ( the hypothesized star - like object formed when a normal star and a neutron star merge ) or a super - asymptotic giant branch ( agb ) star .
instead we propose a binary scenario where this star is the companion of a former massive agb star , which polluted the object with via its stellar wind , i.e. a special case of an extrinsic s star .
our new scenario solves two additional problems with the two existing scenarios for its nature as thorne - ytkow object or present - day super - agb star .
the puzzling high ratio of the strength of calcium to iron absorption lines is unexpected for smc supergiants , but is fully consistent with the expectations for halo abundances .
secondly , its strong variability can now be explained naturally as a manifestation of the mira phenomenon .
we discuss further observational tests that could distinguish between the foreground and smc scenarios in advance of the improved proper motion measurements likely to come from gaia .
[ firstpage ] astrometry stars : abundances stars : chemically peculiar
stars : individual : hv 2112 |
there is a still growing interest in quantifying and measuring correlations in composite quantum systems ( for a recent review see @xcite ) .
such notions as entanglement , quantum discord and other entropic measures of quantumness proved to be useful in these investigations @xcite . while entanglement from a very definition reflects by itself some specific structure of the density matrix representing the state , discord measures the amount of quantum correlations between subsystems in terms of mutual von neumann entropy . on the other hand , for many years the nonadditive generalizations of shannon and von neumann entropy functions , such as renyi entropy @xcite , tsallis entropy @xcite and others @xcite , has been used to describe some aspects of statistical and thermodynamic behaviour of systems far from equilibrium .
these families of entropy functions parametrized by @xmath1 seem to incorporate in a natural manner some correlations present in composite systems .
in what follows we focus on tsallis entropy which , except additivity , displays similar properties ( nonnegativity , concavity ) to those of von neumann entropy @xcite and therefore , in most applications in the field of quantum information , can be used as well @xcite . in particular , an observation due to abe @xcite that using tsallis entropy one can detect more separable two - qubit werner states than using shannon entropy seems very promising . in this article , we investigate the generalization of mutual entropy and quantum discord when defined in terms of tsallis entropy .
it turns out that even at the classical level this generalization can be done following two paths .
hence , at the quantum level we introduce two quantities , called @xmath0-discords @xmath2 and @xmath3 , which generalize quantum discord , and we investigate them for werner and isotropic states .
we give some arguments that the use of @xmath0-mutual entropy derived from @xmath0-expectation value results in the @xmath0-discord function which , at least for werner and isotropic states , reflects in a more appropriate way the amount of quantum correlations for all values of @xmath1 than the other one . finally , we give an example indicating that both types of @xmath0-discord can take negative values for some specific quantum states belonging to the family of two - qubit x states @xcite .
let us start we some basics of classical information theory , where , in fact , some ambiguities crucial for further analysis arise .
let us consider two classical systems @xmath4 and @xmath5 described by discrete probability spaces @xmath6 and @xmath7 , respectively .
information contained in any probability distribution @xmath8 can be quantified by the shannon entropy function @xmath9 consider now a joint probability space @xmath10 , where @xmath11 is the joint probability , and define conditional probability @xmath12 as @xmath13 hence , shannon entropy of conditional probability ( [ cond - prob ] ) reads @xmath14 averaging it over subsystem b leads to the notion of _ conditional entropy _
@xmath15=\sum_jp^{(b)}_js(a|b_j),\ ] ] which is an expectation value of entropy of conditional probability with respect to marginal probability @xmath16 . as an obvious consequence , @xmath17 with @xmath18 being the joint entropy , holds . the amount of information on the subsystem a contained in b is characterized by _ mutual entropy _
@xmath19 due to ( [ s(a|b ) ] ) it follows immediately that @xmath20 can be alternatively described by @xmath21 as long as the shannon entropy function is used . as we will argue in sect .
[ ccmte ] this equivalence is no longer valid when using more general entropy function , i.e. , tsallis entropy . using generalized logarithmic function , called _
@xmath0-logarithm _ , @xmath22 one can define _ tsallis entropy _ of rank @xmath0 as @xmath23 and related to it , _ tsallis entropy of conditional probability _
@xmath24 note that @xmath25 corresponds to shannon entropy .
now , in order to define _ tsallis conditional entropy
_ one can follow several nonequivalent ways .
* in a perfect analogy to ( [ s - cond - e ] ) one could define @xmath26=\sum_jp^{(b)}_jt_q(a|b_j),\ ] ] where @xmath27 means the expectation value with respect to the marginal probability @xmath28 .
unfortunately , this quantity has no useful properties , in particular , there is no relation similar to ( [ s(a|b ) ] ) ! * if someone wants the property ( [ s(a|b ) ] ) to hold then an expectation value with modified weights @xmath29 should be considered , i.e. , @xmath30 as a consequence of the following property of @xmath0-logarythm : @xmath31 one easily obtains that @xmath32 but @xmath33 is not a probability distribution , hence @xmath34 is not a legitimate expectation value . * to avoid this , one defines a @xmath0-_expectation value _ of any random variable @xmath35 and probability @xmath8 as @xmath36\;:=\;\frac{\sum_jp_j^qf_j}{\sum_jp_j^q},\ ] ] which obviously preserves unity , @xmath37=1 $ ] .
usually , the probability distribution @xmath38 is refered to as an _ escort probability_. hence , conditional tsallis entropy @xmath39 can be defined as the @xmath0-expectation value with respect to the marginal probability @xmath28 as @xmath40=\sum_jr^{(b)}_jt_q(a|b_j)=\frac{\sum_i(p_i^{(b)})^qt_q(a|b_i ) } { \sum_j(p^{(b)}_j)^q}\ ] ] in perfect analogy to the case of shannon entropy ( [ s - cond - e ] ) , where @xmath41\,.\ ] ] now , ( [ t(a|b ) ] ) is no longer valid , but we can rewrite ( [ cond - entropy-2 ] ) using ( [ cond - entropy ] ) and the relation @xmath42 as @xmath43 where @xmath44 using ( [ q - ln ] ) results in @xmath45 \\ & = & -\frac1n\sum_{i , j}\big[p_{ij}^q\ln_qp_{ij}-p_{ij}(p^b_j)^{q-1}\ln_qp^b_j\big ] \\ & = & \frac{t_q(a , b)-t_q(b)}{n}\,.\end{aligned}\ ] ] finally , one arrives at the modified relation ( [ t(a|b ) ] ) , i.e. , @xmath46 in principle , both ( [ t(a|b ) ] ) and ( [ t_q - cond ] ) can be used to define _ mutual tsallis entropy_. using ( [ t(a|b ) ] ) leads to @xmath47 and mimic the well - known case of shannon entropy . on the other hand , if one consequently defines the mutual information as @xmath48 and takes into account ( [ t_q - cond ] ) , then @xmath49 both ( [ trad - me ] ) and ( [ mut - inf-2 ] ) can be generalized and used to quantify some correlations in composite quantum systems but we will argue in sect .
[ qt ] that the use of ( [ mut - inf-2 ] ) has some advantages .
it is a common agreement that the most suitable quantity to describe correlations between subsystems a and b is _ mutual entropy _ ( see ( [ mutual - s - entropy ] ) ) usually defined in terms of von neumann entropy @xmath50 as @xmath51 where @xmath52 , @xmath53 , @xmath54 , @xmath55 , @xmath56 being reduced density matrices .
clearly , @xmath57 contains information about all correlations enclosed in the state @xmath58 and one can pose the question which of them are of classical origin and which one are pure quantum @xcite .
one can decompose ( in a nonunique way ) total correlations @xmath57 as @xmath59 the classical part @xmath60 can be determined by optimizing the local measurement procedure as proposed in @xcite .
the quantity @xmath61 is then called _
quantum discord_. in what follows we generalize this notion using tsallis entropy instead of von neumann ones .
here we recall some properties of von neumann entropy crucial for further analysis @xcite .
the first one is additivity with respect to the tensor product , @xmath62 next , the so - called subadditivity property ( sa ) @xmath63 ensures that von neumann mutual entropy is nonnegative @xmath64 and strong subadditivity ( ssa ) @xcite @xmath65 which is essentially a tri - partite property , implies nonnegativity of quantum discord @xmath61 @xcite . in the framework of composite _ quantum _ systems _ tsallis entropy
_ can be defined in perfect analogy to the classical case as @xmath66 tsallis entropy function @xmath67 is nonnegative , concave and , if @xmath58 is pure then @xmath68 , but @xmath67 is no longer additive with respect to the tensor product @xcite , instead we have pseudo - additivity ( pa ) , @xmath69 this property , in fact , makes tsallis entropy useful in nonextensive statistical mechanics @xcite . from ( [ pa ] ) results that sa fails for arbitrary @xmath1 .
note , however , that for @xmath70 one obtains @xmath71 and moreover @xcite @xmath72 hence sa holds .
let us define two types of _ quantum mutual tsallis entropy _
@xcite , both measuring total correlations between subsystems @xmath4 and @xmath5 , as @xmath73 note that all the objects entering both types of mutual entropy given by ( [ mut - inf-3a ] ) and ( [ mut - inf-3b ] ) , are already well - defined .
moreover , ( [ t - sub ] ) implies that @xmath74 is nonnegative for @xmath70 .
unfortunately , we are lacking this property for @xmath75 . in
what follows , we will argue that ( [ mut - inf-3b ] ) displays some advantages when describing quantum correlations in composite systems . in order to determine classical correlations @xmath76 based on tsallis entropy we follow @xcite .
for this , let us consider a perfect measurement on subsystem @xmath5 , defined by a set of one - dimensional projectors @xmath77 , yielding post - measurement states @xmath78 where @xmath79 let us define _ quantum conditional tsallis entropy _ with respect to this measurement by @xmath80\;=\ ; \frac{\dst\sum_k p_k^qt_{q}(\rho_k)}{\dst\sum_kp_k^q}\,.\ ] ] in fact , @xmath81 is fully characterized by an ensemble @xmath82 .
now , the information gained about a as a result of measurement on b is @xmath83 when optimization is taken over all perfect measurements .
one assumes that the quantity @xmath76 encodes classical correlations .
finally , we define _ @xmath0-discord _ by @xmath84 note , however , that instead of @xmath75 one could use @xmath74 given by ( [ mut - inf-3a ] ) and instead of ( [ class - cor ] ) one could define @xmath85 where @xmath86\;=\ ; \sum_k p_kt_{q}(\rho_k)\ ] ] involves only the ordinary expectation value . as a consequence ,
we obtain another type of @xmath0-discord @xmath87 in @xcite it is proven that @xmath88 but some numerical results suggest that @xmath3 can also be negative for some values of @xmath70 . in sect .
[ w - q - d ] we determine and comment on both types of @xmath0-discord for werner and isotropic states .
let us consider @xmath89 werner states @xmath90 , parametrized by @xmath91 , which are invariant with respect to local unitary transformations , i.e. , @xmath92 , defined by @xmath93\,,\ ] ] where @xmath94 is the flip operator @xmath95 and @xmath96 is @xmath97 identity matrix .
straightforward algebra yields @xmath98\,.\label{q - ent - sys}\end{aligned}\ ] ] now , using ( [ rho_k ] ) and ( [ p_k ] ) , one obtains @xmath99 and @xmath100,\ ] ] where , in fact , @xmath101 is independent on a particular measurement , as one should expect according to local unitary invariance of werner states .
hence , @xmath102\ ] ] and the discords @xmath2 and @xmath3 differ one from another by total correlation terms only .
the later ones obtained from ( [ mut - inf-3a ] ) and ( [ mut - inf-3b ] ) using ( [ q - ent - sub ] ) , ( [ q - ent - sys ] ) read @xmath103\,,\\ { \cal i}_q(a : b ) & = & \frac{1}{q-1}\big[\frac12(1-\lambda)^q+\frac32\big(\frac{1+\lambda}{3}\big)^q-2^{1-q}\big ] \,.\end{aligned}\ ] ] finally , both @xmath0-discords for werner states turn out to be @xmath104\,,\\ { d}_q(a : b ) \!\!&=&\!\ ! \frac{1}{q-1}\big[\frac12(1-\lambda)^q+ \frac{1}{2}\big(\frac{1+\lambda}{3}\big)^q-\big(\frac{2-\lambda}{3}\big)^q\big]\,.\end{aligned}\ ] ] comparison between two kinds of @xmath0-discord for werner states for @xmath105 .
the upper graph corresponds to @xmath3 whereas the lower to @xmath2.,height=226 ] the comparison between @xmath2 and @xmath3 for @xmath105 is shown in fig .
[ fid - dwa ] .
note , that for classically correlated ( diagonal ) werner state corresponding to @xmath106 , @xmath107 for every @xmath1 , as one should expect .
the quantity @xmath108 is positive for this state and for @xmath109 can be even negative which makes it less useful as a measure of quantum correlations
. we can prove ( see appendix ) that the @xmath0-discord @xmath2 for @xmath89 werner states is always nonnegative for @xmath1 .
similar calculations can be carried out also for @xmath89 isotropic states defined as @xmath110 where @xmath111 and the projector @xmath112 .
this family is invariant with respect to the following action of local unitary transformations : @xmath113 after straightforward algebra one obtains two types of @xmath0-discords for isotropic states as @xmath114\,,\\ { d}_q(a : b ) \!\!&=&\!\ ! \frac{1}{q-1}\big[\frac12\big(\frac{2 - 2\lambda}{3}\big)^q+ \frac{1}{2}(2\lambda)^q-\big(\frac{1 + 2\lambda}{3}\big)^q\big]\,.\end{aligned}\ ] ] comparison between two kinds of @xmath0-discord for isotropic states for @xmath105 .
the dashed - line graph corresponds to @xmath3 whereas the dotted - line to @xmath2.,height=226 ] the comparison between @xmath2 and @xmath3 for @xmath105 is shown in fig .
[ fig - dwa - iso ] .
now , the classically correlated isotropic state corresponds to @xmath115 and as expected , for this value of @xmath116 , @xmath107 for every @xmath1 .
we can prove that the @xmath0-discord @xmath2 for @xmath89 isotropic states is always nonnegative for every @xmath1 .
the quantity @xmath108 is positive for this state and for @xmath109 can be even negative which makes its behaviour similar to that of werner states .
as a final example , let us consider the following family of circulant two - qubit states : @xmath117,\qquad 0\leq\varepsilon , g\leq 1\,.\ ] ] the family does not belong neither to the bell diagonal states considered in @xcite nor to its generalisation investigated in @xcite . but the method of discord calculation used in @xcite can be adopted for @xmath0-discord @xmath2 as well . one can immediately check that eigenvalues of @xmath118 read @xmath119 whereas eigenvalues of reductions @xmath120 and @xmath121 are the following : @xmath122 therefore , @xmath123\ ] ] and @xmath124\,.\ ] ] then , the quantum mutual tsallis information ( [ mut - inf-3b ] ) results in @xmath125^q}{(q-1)(1+\varepsilon^q)(1+\varepsilon)^q}\\ & & \hspace*{20 mm } + \;\frac{[\frac12(1+\varepsilon^2)- \frac12\sqrt{\varepsilon^4 + 2(2g^2 - 1)\varepsilon^2 + 1}]^q}{(q-1)(1+\varepsilon^q)(1+\varepsilon)^q}\,.\end{aligned}\ ] ] now , in order to obtain classical correlations @xmath76 we parameterise the projective measurement @xmath126 by @xmath127 unitary matrix @xmath128 @xmath129 of the form @xmath130
\,,\quad t_0 ^ 2+t_1 ^ 2+t_2 ^ 2+t_3 ^ 2=1\,,\ ] ] where @xmath131 are pauli matrices .
the ensemble of post - measurement states @xmath82 , @xmath132 , can be characterized by probabilities ( see @xcite for details ) @xmath133 and the eigenvalues of @xmath134 and @xmath135 ( see ( [ rho_k ] ) ) @xmath136 where @xmath137 ^ 2 + 4kl(1+g)^2 - 16 mg } { [ ( 1+\varepsilon^{-1})k+(\varepsilon+1)l]^2}}\ , , \\
\vartheta ' & = & \sqrt{\frac{[(1-\varepsilon^{-1})l+(\varepsilon-1)k]^2 + 4kl(1+g)^2 - 16 mg } { [ ( 1+\varepsilon^{-1})l+(\varepsilon+1)k]^2}}\,,\end{aligned}\ ] ] with @xmath138 $ ] , @xmath139 , @xmath140 $ ] defined in terms of @xmath141 as @xmath142 the quantum conditional tsallis entropy with respect to the measurement yields @xmath143 and @xmath144 let us define the function @xmath145\,.\ ] ] which is decreasing in @xmath146 for @xmath1 . due to @xcite the infimum in ( [ class-1 ] ) can be obtained when : * @xmath147 and @xmath148 .
this results in @xmath149 , @xmath150 , @xmath151 and @xmath152 from ( [ spektrum ] ) and ( [ cond - t - entropy ] ) one obtains @xmath153\,.\label{eq1}\end{aligned}\ ] ] * @xmath154 and @xmath155 .
this results in the same formula ( [ eq1 ] ) * @xmath156 .
this results in @xmath149 or @xmath157 , hence @xmath158 \dst\frac{\sqrt{(\varepsilon-\varepsilon^{-1})^2 + 4(1-g)^2}}{2+\varepsilon+\varepsilon^{-1 } } & \quad \dst m=\frac{1}{4}\ , , \end{array}\right.\ ] ] and @xmath159 note that @xmath160 therefore @xmath161 and ( [ class-1 ] ) takes the form @xmath162-f_q\big(\frac{\sqrt{(\varepsilon-\varepsilon^{-1})^2 + 4(1+g)^2}}{2+\varepsilon+\varepsilon^{-1}}\big).\ ] ] as noticed in @xcite , the @xmath0-discord @xmath3 can take negative values for some states and @xmath163 .
now , we discover the same for @xmath2 for some specific @xmath164-states and @xmath165 ( see fig .
[ ujemny ] ) .
@xmath0-discord for @xmath164-states . for @xmath165 and @xmath166 @xmath0-discord @xmath2 is negative for some interval of @xmath167.,height=226 ]
the author would like to acknowledge many stimulated discussions with prof .
d. chruciski .
we have the following simple lemma note , that for werner states the inequality @xmath176 is equivalent to @xmath177 or @xmath178 evidently @xmath179 hence we can apply lemma 1 which proves nonnegativity of @xmath2 for werner states .
the same method can be used to show that also for isotropic states the same conclusion holds .
10 l. c. cleri , j. maziero , and r. m. serra , int .
* 9 * , 1837 ( 2011 ) and other articles in a special issue of int .
j. quant . inf
. 9 nos . 7 & 8 ( 2011 ) .
nielsen and i.l .
chuang , _ quantum computation and quantum information _ , cambridge university press , cambridge , 2000 .
r. horodecki , p. horodecki , m. horodecki , k. horodecki , rev .
phys . * 81 * , 865942 ( 2009 ) .
h. ollivier and w.h .
urek , phys .
88 , 017901 ( 2002 ) .
l. henderson and v. vedral , j. phys . a 34 , 6899 ( 2001 ) .
m. horodecki , p. horodecki , r. horodecki , j. oppenheim , a. sen , u. sen , and b. synak - radtke , phys .
a 71 , 062307 ( 2005 ) .
r. rossignoli , n. canosa , l. ciliberti , phys .
a * 82 * , 052342 ( 2010 ) .
a. rnyi , _ probability theory _ , north - holland , amsterdam , 1970 .
c. tsallis , j. stat .
* 52 * , 479 ( 1988 ) .
m. masi , phys .
a * 338 * , 217 ( 2005 ) .
s. furuichi , k. yanagi , k. kuriyama , j. math .
. 45 48684877 ( 2004 ) .
s. furuichi , j. math .
* 47 * , 023302 ( 2006 ) .
x. hu , z. ye , j. math . phys . * 47 * , 023502 ( 2006 ) .
s. abe and a. k. rajagopal , phys .
e 60 , 3461 ( 1999 ) .
s. abe and a.k .
rajagopal , physica a * 289 * , 157164 ( 2001 ) .
t. yamano , phys .
e * 63 * , 046105 ( 2001 ) .
o. ghne and m. lewenstein , phys .
a * 70 * , 022316 ( 2004 ) .
j. s. kim , phys .
a 81 , 062328 ( 2010 ) .
s. luo , _ quantum discord for two - qubit systems _ , phys .
a * 77 * , 042303 ( 2008 ) .
v. vedral , phys .
90 , 050401 ( 2003 ) .
j. maziero , l. c. celri , r. m. serra , and v. vedral , phys .
a 80 , 044102 ( 2009 ) .
bo li , zhi - xi wang , shao - ming fei , _ quantum discord and geometry for a class of two - qubits states _ , phys .
a * 83 * , 022321 ( 2011 ) .
m. ali , a.r.p .
rau , and g. alber , _ quantum discord for two - qubit x - states _
a 81 , 042105 ( 2010 ) . | due to some ambiguity in defining mutual tsallis entropy in the classical probability theory , its generalization to quantum theory is discussed and , as a consequence , two types of generalized quantum discord , called @xmath0-discords , are defined in terms of quantum tsallis entropy .
@xmath0-discords for two - qubit werner and isotropic states are calculated and it is shown that one of them is positive , at least for states under investigation , for all @xmath1 . finally , an analytical expression for @xmath0-discord of certain family of two - qubit x states is presented . |
spatial dissipative solitons ( sdss ) are self - trapped beams of light rosanov , mt , akhmed or plasmonic waves @xcite-@xcite propagating in planar or bulk waveguides . they result from the balance between diffraction and self - focusing ( sf ) nonlinearity , which is maintained simultaneously with the balance between the material loss and compensating gain . due to their basic nature , sdss are modes of profound significance to nonlinear photonics ( optics and plasmonics ) , as concerns the fundamental studies and potential applications alike . in particular , a straightforward possibility is to use each sufficiently narrow sds beams as signal carriers in all - optical data - processing schemes .
this application , as well as other settings in which the solitons occur , stresses the importance of the stabilization of the sdss modes , and of development of enabling techniques for the generation and steering of such planar and bulk beams . in terms of the theoretical description ,
basic models of the sds dynamics make use of complex ginzburg - landau equations ( cgles ) .
the prototypical one is the cgle with the cubic nonlinearity , which includes the conservative paraxial - diffraction and kerr terms , nonlinear ( cubic ) loss with coefficient @xmath2 , which represents two - photon absorption in the medium , and the spatially uniform linear gain , with strength @xmath3 , aiming to compensate the loss @xcite : @xmath4 here @xmath5 is the complex amplitude of the electromagnetic wave in the spatial domain , @xmath6 is the propagation distance , the paraxial - diffraction operator @xmath7 acts on transverse coordinates @xmath8 in the case of the propagation in the bulk , or on the single coordinate , @xmath9 , in the planar waveguide .
accordingly , eq . ( [ gl ] )
is considered as two- or one - dimensional ( 2d or 1d ) equation in those two cases .
the equation is normalized so that the diffraction coefficient is @xmath10 , while @xmath11 is the kerr coefficient , @xmath12 and @xmath13 corresponding to the sf and self - defocusing ( sdf ) signs of the nonlinearity , respectively . a more general version of the cgle may include an imaginary part of the diffraction coefficient @xcite , which is essential , in particular , for the use of the cgle as a model of the traveling - wave convection @xcite .
however , in optical models that coefficient , which would represent diffusivity of photons , is usually absent .
a well - known fact is that the 1d version of eq .
( [ gl ] ) gives rise to an exact solution in the form of an exact _ chirped _ sds , which is often called a _
pereira - stenflo soliton
_ @xcite:@xmath14 ^{1+i\mu } , \label{ps } \\ a^{2}=3\gamma /\left
( 2\epsilon \right ) , ~\kappa ^{2}=\gamma /\mu , ~k=\left ( \gamma /2\right ) \left ( \mu ^{-1}-\mu \right ) , \label{psconstants}\end{aligned}\]]where the chirp coefficient is@xmath15this exact solution is subject to an obvious instability , due to the action of the uniform linear gain on the zero background far from the soliton s core .
therefore , an important problem is the design of physically relevant models which may produce stable sds .
one possibility is to achieve full stabilization of the solitons in systems of linearly coupled cgles modeling dual - core waveguides , with the linear gain and loss acting in different cores @xcite-@xcite , marini .
this includes , _ inter alia _ , a @xmath1-symmetric version of the system that features the exact balance between the gain and loss @xcite .
the simplest example of such a stabilization mechanism is offered by the following coupled cgle system @xcite : @xmath16 where @xmath17 is the linear - coupling coefficient , @xmath18 and @xmath19 are the electromagnetic - wave amplitude and the linear - loss rate in the stabilizing dissipative core , and @xmath20 is a possible wavenumber mismatch between the cores . in the case of @xmath21 ,
the zero background is stable in the framework of eqs .
( [ u - c ] ) and ( [ v - c ] ) under condition@xmath22 the same ansatz ( [ ps ] ) which produced the pereira - stenflo soliton for the uncoupled cgle yields an exact solution of the coupled system ( [ u - c ] ) , ( [ v - c]):@xmath23 ^{1+i\mu } , \]]with chirp @xmath24 given by the same expression ( [ mu ] ) as above , and @xmath25 ^{-1}a .
\label{b}\]]a stable soliton is obtained if a pair of distinct solutions are found , compatible with the condition of the stability for the zero background [ which is eq .
( [ gg ] ) in the case of @xmath21 ] , instead of the single solution in the case of eq .
( [ psconstants ] ) .
then , the soliton with the larger amplitude is stable , coexisting , as an _ attractor _ , with the stable zero solution , while the additional soliton with a smaller amplitude plays the role of an unstable _ separatrix _ which delineates the boundary between attraction basins of the two coexisting stable solutions @xcite . in the case of @xmath21 ,
the aforementioned condition of the existence of two solutions reduces to @xmath27 .
\label{two}\]]in particular , it follows from eq .
( [ two ] ) and ( [ mu ] ) that a related necessary condition , @xmath28 , implies @xmath29 , i.e. , the kerr coefficient , @xmath11 , must feature the sf sign , and the cubic - loss coefficient , @xmath30 , must be sufficiently small in comparison with @xmath31 . if inequality @xmath28 holds , and zero background is close to its stability boundaries , i.e. , @xmath32 and @xmath33 , see eq .
( [ gg ] ) , parameters of the stable soliton with the larger amplitude are@xmath34while propagation constant @xmath35 is given , in the first approximation , by eq .
( [ psconstants ] ) . in the same case ,
the unstable _ _ _ _ separatrix soliton has a small amplitude and large width , while its chirp keeps the above value ( [ mu]):@xmath36 getting back to models based on the single cgle , stable solitons can also be generated by the equation with cubic gain sandwiched " between linear and quintic loss terms , which corresponds to the following generalization of eq .
( [ gl ] ) : @xmath37 with @xmath38 , @xmath39 , @xmath19 , and @xmath40 .
the linear loss , represented by coefficient @xmath41 , provides for the stability of the zero solution to eq .
( [ cqgl ] ) .
the cubic - quintic ( cq ) cgle was first proposed , in a phenomenological form , by petviashvili and sergeev @xcite .
later , it was demonstrated that the cq model may be realized in optics as a combination of linear amplification and saturable absorption @xcite-@xcite .
stable dissipative solitons supported by this model were investigated in detail by means of numerical and analytical methods @xcite-@xcite .
the subject of the present mini - review is the development of another method for creating stable localized modes , which makes use of linear gain applied at a hot spot " ( hs ) , i.e. a localized amplifying region embedded into a bulk lossy waveguide .
the experimental technique which allows one to create localized gain by means of strongly inhomogeneous distributions of dopants implanted into the lossy waveguide , which produce the gain if pumped by an external source of light , is well known @xcite .
another possibility is even more feasible and versatile : the dopant density may be uniform , while the external pump beam is focused on the location where the hs should be created . supporting dissipative solitons by the localized gain
was first proposed not in the framework of cgles , but for a gap soliton pinned to an hs in a lossy bragg grating ( bg ) @xcite . in terms of the spatial - domain dynamics ,
the respective model is based on the system of coupled - mode equations ( cmes ) for counterpropagating waves , @xmath42 and @xmath42 , coupled by the bragg reflection:@xmath43where the tilt of the light beam and the reflection coefficients are normalized to be @xmath10 , the nonlinear terms account for the self- and cross - phase modulation induced by the kerr effect , @xmath3 is the linear - loss parameter , @xmath44 represents the local gain applied at the hs [ @xmath45 , @xmath46 being the dirac s delta - function ] , and the imaginary part of the gain coefficient , @xmath47 , accounts for a possible attractive potential induced by the hs ( it approximates a local increase of the refractive index around the hs ) . as was mentioned in @xcite too , and for the first time investigated in detail in @xcite , the hs embedded into the usual planar waveguide is described by the following modification of eq .
( [ gl ] ) : @xmath48 where , as well as in eqs .
( [ o1 ] ) and ( [ o2 ] ) , @xmath44 is assumed , and the negative sign in front of @xmath49 represents the linear loss in the bulk waveguide .
another hs model , based on the 1d cgle with the cq nonlinearity , was introduced in @xcite:@xmath50where @xmath51 represents the quintic self - defocusing term , @xmath3 and @xmath19 are , as above , strengths of the bulk losses and localized _ cubic _ gain , and @xmath52 is the width of the hs ( an approximation corresponding to @xmath53 , with the hs in the form of the delta - function , may be applied here too ) .
while solitons in uniform media , supported by the cubic gain , are always unstable against the blowup in the absence of the quintic loss @xcite , the analysis reported in valery demonstrates that , quite counter - intuitively , _
stable _ dissipative localized modes in the uniform lossy medium may be supported by the _ unsaturated _ localized cubic gain in the model based on eq .
( cqhs ) .
in addition to the
direct " linear gain assumed in the above - mentioned models , losses in photonic media may be compensated by parametric amplification , which , unlike the direct gain , is sensitive to the phase of the signal @xcite .
this mechanism can be used for the creation of a hs , if the parametric gain is applied in a narrow segment of the waveguide .
as proposed in @xcite , the respective 1d model is based on the following equation , cf .
( [ cghs ] ) and ( [ cqhs]):@xmath54where @xmath55 is the complex conjugate field , @xmath20 is a real phase - mismatch parameter , and , as well as in eq .
( [ cqhs ] ) , the hs may be approximated by the delta - function in the limit of @xmath53 .
models combining the localized gain and the uniformly distributed kerr nonlinearity and linear loss have been recently developed in various directions .
in particular , 1d models with two or multiple hss spotsexact1-@xcite and periodic amplifying structures kkv , yingji , as well as extended patterns @xcite , have been studied , chiefly by means of numerical methods .
the numerical analysis has made it also possible to study 2d settings , in which , most notably , stable localized vortices are supported by the gain confined to an annular - shaped area @xcite-@xcite . the parametric amplification applied at a ring may support stable vortices too , provided that the pump 2d beam itself has an inner vortical structure @xcite .
another ramification of the topic is the development of symmetric combinations of hot " and cold " spots , which offer a realization of the concept of @xmath1-symmetric systems in optical media , that were proposed and built as settings integrating the balanced spatially separated gain and loss with a spatially symmetric profile of the local refractive index @xcite-@xcite .
the study of solitons in nonlinear @xmath56-symmetric settings has drawn a great deal of attention nonlinpt1-@xcite , @xcite . in particular , it is possible to consider the 1d model in the form of a symmetric pair of hot and cold spots described by two delta - functions embedded into a bulk conservative medium with the cubic nonlinearity @xcite . a limit case of this setting , which admits exact analytical solutions for @xmath1-symmetric solitons , corresponds to a @xmath1 _ dipole _ , which is represented by the derivative of the delta - function in the following 1d equation @xcite:@xmath57here @xmath58 and @xmath59 correspond to the sf and sdf bulk nonlinearity , respectively , @xmath60 is the strength of the attractive potential , which is a natural conservative component of the @xmath1 _ dipole _ , and @xmath61 is the strength of the @xmath1 dipole .
it is also natural to consider discrete photonic settings ( lattices ) , which appear , in the form of discrete cgles , as models of arrayed optical discrcgl1-@xcite or plasmonics plasmon - array2,plasmon - array3 waveguides . in this context , lattice counterparts of the hss amount to a single @xcite or several discr2 amplified site(s ) embedded into a 1d or 2d @xcite lossy array .
being interested in tightly localized discrete states , one can additionally simplify the model by assuming that the nonlinearity is carried only by the active cores , which gives rise to the following version of the discrete cgle , written here in the general 2d form @xcite : @xmath62 \delta _ { m,0}\delta _ { n,0}u_{m , n}\ ; , \label{eq : gl}\end{aligned}\]]where @xmath63 , @xmath64 , @xmath65 , ... are discrete coordinates on the lattice , @xmath66 and @xmath67 are the kronecker s symbols , and the coefficient of the linear coupling between adjacent cores is scaled to be @xmath10 .
as above , @xmath3 is the linear loss in the bulk lattice , @xmath68 and @xmath47 represent the linear gain and linear potential applied at the hs site ( @xmath69 ) , while @xmath70 and @xmath71 account for the sf ( @xmath72 ) or sdf ( @xmath73 ) kerr nonlinearity and nonlinear loss ( @xmath74 ) or gain ( @xmath75 ) acting at the hs ( the unsaturated cubic gain may be a meaningful feature in this setting @xcite ) .
the @xmath1 symmetry can be introduced too in the framework of the lattice system . in particular , a discrete counterpart of the 1d continuous model ( [ eq ] ) with the @xmath1-symmetric dipole was recently elaborated in @xcite : @xmath76where the @xmath1 dimer ( discrete dipole ) embedded into the hamiltonian lattice is represented by @xmath77 at @xmath78 , @xmath79 at @xmath80 , and @xmath81 at @xmath82 .
a counterpart of the delta - functional attractive potential in eq .
( [ eq ] ) corresponds to a local defect in the inter - site couplings : @xmath83 , @xmath84 at @xmath85 .
lastly , the nonlinearity is assumed to be carried solely by the dimer embedded into the lattice : @xmath86 at @xmath87 , and @xmath81 at @xmath82 , cf .
( [ eq : gl ] ) .
equation ( [ dnlse ] ) admits exact analytical solutions for all @xmath1-symmetric and antisymmetric discrete solitons pinned to the dimer .
in addition to the hs , one can naturally define a warm spot " ( ws ) , in the 2d cgle with the cq nonlinearity , where the coefficient of the linear loss is given a spatial profile with a minimum at the ws ( @xmath88 ) skarka .
the equation may be taken as the 2d version of eq .
( [ cqgl ] ) with @xmath89where @xmath90 is the radial coordinate , coefficients @xmath91 and @xmath92 being positive .
this seemingly simple model gives rise to a great variety of stable 2d modes pinned to the ws .
depending on values of parameters in eqs .
( [ cqgl ] ) and ( [ ws ] ) , these may be simple vortices , rotating elliptic , eccentric , and slanted vortices , spinning crescents , etc . @xcite .
lastly , the use of the spatial modulation of loss coefficients opens another way for the stabilization of the sds : as shown in @xcite , the solitons may be readily made stable if the _ spatially uniform _ linear gain is combined with the local strength of the cubic loss , @xmath93 , growing from the center to periphery at any rate faster than @xmath94 , where @xmath95 is the distance from the center and @xmath96 the spatial dimension .
this setting is described by the following modification of eq .
( [ gl ] ) : @xmath97 |u|^{2}u+\gamma u , \label{olga}\ ] ] where , as said above , @xmath61 and @xmath93 are positive , so that @xmath98 , for @xmath99 or @xmath100 .
this mini - review aims to present a survey of basic results obtained for sdss pinned to hss in the class of models outlined above . in view of the limited length of the article ,
stress is made on the most fundamental results that can be obtained in an analytical or semi - analytical form , in combination with the related numerical findings , thus providing a deep inside into the dynamics of the underlying photonic systems .
in fact , the possibility of obtaining many essential results in an analytical form is a certain asset of these models .
first , in section ii findings are summarized for the most fundamental 1d model based on eq .
( [ cghs ] ) , which is followed , in section iii , by the consideration of the @xmath1-symmetric system ( [ eq ] ) .
the bg model ( [ o1 ] ) , ( [ o2 ] ) is the subject of section iv , and section v deals with the 1d version of the discrete model ( [ eq : gl ] ) .
the article is concluded by section vi .
the presentation in this section is focused on basic model ( [ cghs ] ) and its extension for two hss , chiefly following works @xcite and spotsexact1 , for the settings with a single and double hs , respectively .
both analytical and numerical results are presented , which highlight the most fundamental properties of sds supported by the tightly localized gain embedded into lossy optical media .
stationary solutions to eq .
( [ cghs ] ) are looked for as @xmath101 where complex function @xmath102 satisfies an ordinary differential equation,@xmath103at@xmath104 supplemented by the boundary condition ( b.c . ) at @xmath45 , which is generated by the integration of eq .
( [ cghs ] ) in an infinitesimal vicinity of @xmath45,@xmath105assuming even stationary solutions , @xmath106 . as seen from the expression for @xmath107 in eq .
( [ psconstants ] ) [ recall that eqs .
( [ gl ] ) and ( [ cghs ] ) have opposite signs in front of @xmath108 , eq .
( [ ode ] ) with @xmath3 and @xmath2 can not be solved by a @xmath109 ansatz similar to that in eq .
( [ ps ] ) . as an alternative
, @xmath109 can be replaced by @xmath110 : @xmath111 ^{-\left ( 1+i\mu \right ) } , \label{sinh}\ ] ] where @xmath112 prevents the singularity
. this ansatz yields an exact _ codimension - one _ solution to eq .
( [ ode ] ) with b.c .
( [ bc ] ) , which is valid under a special constraint imposed on coefficients of the system:@xmath113parameters of this solution are @xmath114@xmath115the squared amplitude of the solution is@xmath116 the main characteristic of the localized beam is its total power , @xmath117for the solution given by eqs .
( [ sinh])-([xi]),@xmath118obviously , solution ( [ sinh ] ) exists if it yields @xmath119 and @xmath120 , i.e. , @xmath121the meaning of threshold condition ( [ thr ] ) is that , to support the stable pinned soliton , the local gain ( @xmath122 ) must be sufficiently large in comparison with the background loss , @xmath61 .
it is relevant to mention that , according to eq .
( [ codim ] ) , the exact solution given by eqs .
( [ sinh ] ) , ( [ sol3 ] ) , and ( [ xi ] ) emerges at threshold ( thr ) in the sdf medium , with @xmath13 , provided that @xmath123 . in the opposite case , @xmath124
, the threshold is realized in the sf medium , with @xmath12 .
the above analytical solution admits a nontrivial limit for @xmath126 , which implies that the local gain compensates only the nonlinear loss , accounted for by term @xmath127 in eq .
( [ cghs ] ) . in this limit ,
the pinned state is weakly localized , instead of the exponentially localized one ( [ sinh ] ) : @xmath128 with @xmath24 given by expression ( [ sol3 ] ) ( an overall phase shift is dropped here ) .
note that the existence of solution ( [ gamma=0 ] ) does not require any threshold condition , unlike eq .
( [ thr ] ) .
this weakly localized state is a physically meaningful one , as its total power ( [ p ] ) converges , @xmath129 .
note that this power does not depend on the local - potential strength , @xmath130 , unlike the generic expression ( [ ptot ] ) . in the limit case
when the loss and gain vanish , @xmath131 , solution ( [ sinh ] ) goes over into an exact one in the sdf medium ( with @xmath13 ) pinned by the attractive potential : @xmath132 which in interval @xmath133 of the propagation constant .
the total power of this solution is@xmath134 in the limit of @xmath135 , when amplitude ( [ amp0 ] ) and power ( [ p0 ] ) attain their maxima , the solution degenerates from the exponentially localized into a weakly localized one [ cf .
( [ gamma=0])],@xmath136whose total power converges , @xmath137 , as per eq .
( [ p0 ] ) .
the exact solutions given by eqs .
( [ u0])-([k=0 ] ) , which are generic ones in the conservative model [ no spacial constraint , such as ( [ codim ] ) , is required ] , may be used to construct an approximate solution to the full system of eqs .
( [ ode ] ) and ( [ bc ] ) , assuming that the gain and loss parameters , @xmath138 @xmath61 , and @xmath30 , are all small . to this end
, one can use the balance equation for the total power:@xmath139the substitution , in the zero - order approximation , of solution ( [ u0 ] ) , ( [ xi0 ] ) into eq .
( [ balance ] ) yields the gain strength which is required to compensate the linear and nonlinear losses in the solution with propagation constant @xmath35:@xmath140as follows from eq .
( [ gamma1 ] ) , with the decrease of @xmath35 from the largest possible value , @xmath141 , to @xmath81 , the necessary gain increases from the minimum , which exactly coincides with threshold ( [ thr ] ) , to the largest value at which the perturbative treatment admits the existence of the stationary pinned mode , @xmath142the respective total power grows from @xmath81 to the above - mentioned maximum , @xmath143 .
it is expected that , at @xmath122 exceeding the limit value ( [ gmax ] ) admitted by the stationary mode , the solution becomes nonstationary , with the pinned mode emitting radiation waves , which makes the effective loss larger and thus balances @xmath144 . however , this issue was not studied in detail .
the perturbative result clearly suggests that , in the lossy sdf medium with the local gain , the pinned modes exist not only under the special condition ( [ codim ] ) , at which they are available in the exact form , but as fully generic solutions too .
furthermore , the increase of the power with the gain strength implies that the modes are , most plausibly , stable ones . in the case of @xmath12 , which corresponds to the sf sign of the cubic nonlinearity , a commonly known exact solution for the pinned mode in the absence of the loss and gain , @xmath145 ,
is @xmath146 with the total power@xmath147which exists for propagation constants @xmath148 , cf .
( [ u0])-([p0 ] ) . in this case
, the power - balance condition ( [ balance ] ) yields a result which is essentially different from its counterpart ( [ gamma1]):@xmath149straightforward consideration of eq .
( [ gamma12 ] ) reveals the difference of the situation from that considered above for the sdf medium : if the strength of the nonlinear loss is relatively small,@xmath150the growth of power ( [ p02 ] ) from zero at @xmath151 with the increase of @xmath35 is initially ( at small values of @xmath152 ) requires not _ _
_ _ the increase of the gain strength from the threshold value ( thr ) to @xmath153 , but , on the contrary , _ decrease _ of @xmath122 to @xmath154 @xcite . only at @xmath155 ,
( [ epscr ] ) , the power grows with @xmath122 starting from @xmath156 .
it is possible to check the stability of the zero solution , which is an obvious prerequisite for the soliton s stability , as said above . to this end
, one should use the linearized version of eq .
( [ cghs]),@xmath157the critical role is played by localized eigenmodes of eq .
( [ linearized ] ) , @xmath158where @xmath159 is an arbitrary amplitude , localization parameter @xmath17 must be positive , @xmath160 is a wavenumber , and @xmath161 is a complex instability growth rate .
straightforward analysis yields @xcite @xmath162it follows from eq .
( [ lambda ] ) that the stability condition for the zero solution , @xmath163 , amounts to inequality @xmath164 which is exactly _ opposite _ to eq .
( [ thr ] ) . in fact , eq . ( [ ampl ] ) demonstrates that the exact pinned - soliton solution given by eqs .
( [ sinh])-([xi ] ) emerges , via the the standard _ forward _ ( alias _ supercritical _ ) _ pitchfork bifurcation _
@xcite , precisely at the point where the zero solution loses its stability to the local perturbation .
the above analysis of eq .
( [ gamma1 ] ) demonstrates that the same happens with the perturbative solution ( [ u0 ] ) , ( [ xi0 ] ) in the sdf model . on the contrary , the analysis of eq .
( [ gamma12 ] ) has revealed above that the same transition happens to the perturbative solution ( [ u02 ] ) , ( [ xi02 ] ) in the sf medium only at @xmath165 , see eq .
( epscr ) , while at @xmath165 the pitchfork bifurcation is of the _ backward _ ( alias _ subcritical _ bif ) type , featuring the power which originally grows with the _ decrease _ of the gain strength . accordingly , the pinned modes emerging from the subcritical bifurcation are unstable .
however , the contribution of the nonlinear loss ( @xmath2 ) eventually leads to the turn of the solution branch forward and its stabilization at the turning point .
for very small @xmath166 , the turning point determined by eq .
( [ gamma12 ] ) is located at @xmath167 , @xmath168 .
lastly , note that the zero solution is never destabilized , and the stable pinned soliton does not emerge , in the absence of the local attractive potential , i.e. , at @xmath169 . the numerical analysis of the model based on eq . ( [ cghs ] )
was performed with the delta - function replaced by its gaussian approximation,@xmath170with finite width @xmath171 .
the shape of a typical analytical solution ( sinh)-([ampl ] for the pinned soliton , and a set of approximations to it provided by the use of approximation ( [ delta ] ) , are displayed in fig .
all these pinned states are stable , as was checked by simulations of their perturbed evolution in the framework of eq .
( [ cghs ] ) with @xmath46 replaced by @xmath172 . ) -([ampl ] ) , and a set of approximations generated by the regularized delta - function defined as per eq .
( [ delta ] ) .
all the profiles represent stable solutions .
other parameters are @xmath173 @xmath174 @xmath175 , and @xmath176 , as per eq .
( [ codim]).,width=264 ] the minimum ( threshold ) value of the local - gain strength , @xmath122 , which is necessary for the existence of stable pinned solitons is an important characteristic of the setting , see eq .
( [ thr ] ) .
figure fig2 displays the dependence of the @xmath177 on the strength of the background loss , characterized by @xmath178 , for @xmath179 and three fixed values of the local - potential s strength , @xmath180 ( in fact , the solutions corresponding to @xmath181 are unstable , as they are repelled by the hs ) . in addition
, @xmath182 is also shown as a function of @xmath183 under constraint ( [ codim ] ) , which amounts @xmath184 , in the case of @xmath179 .
the corresponding analytical prediction , as given by eq . ( [ thr ] ) , is virtually identical to its numerical counterpart , despite the difference of approximation ( delta ) from the ideal delta - function . , which is necessary for the creation of stationary pinned solitons in the framework of eq .
( [ cghs ] ) , as a function of the background loss ( @xmath185 ) , with @xmath186 @xmath187 , and @xmath188 approximated as per eq .
( [ delta ] ) with @xmath189 .
the strength of the local potential , @xmath130 , is fixed as indicated in the box . for @xmath190 and @xmath191
the lines are guides for the eye , while the straight line for the case of @xmath192 , which corresponds to eq .
( codim ) with @xmath193 , is the analytical prediction given by eq .
( [ thr ] ) .
, width=264 ] figure [ fig2 ] corroborates that , as said above , the analytical solutions represent only a particular case of the family of generic dissipative solitons that can be found in the numerical form .
in particular , the solution produces narrow and tall stable pinned solitons at large values of @xmath194 .
all the pinned solitons , including weakly localized ones predicted by analytical solution ( [ gamma=0 ] ) for @xmath195 , are stable at @xmath153 and @xmath196 .
the situation is essentially different for large @xmath171 in eq .
( delta ) , i.e. , when the local gain is supplied in a broad region . in that case
, simulations do not demonstrate self - trapping into stationary solitons ; instead , a generic outcome is the formation of stable _ breathers _ featuring regular intrinsic oscillations , the breather s width being on the same order of magnitude as @xmath197 , see a typical example in fig .
it seems plausible that , with the increase of @xmath171 , the static pinned soliton is destabilized via the hopf bifurcation @xcite which gives rise to the stable breather . ) with @xmath188 replaced by approximation @xmath198 as per eq .
( [ delta ] ) with @xmath199 , the other parameters being @xmath200 , @xmath201 , and @xmath202.,width=302 ] the extension of eq .
( [ cghs ] ) for two mutually symmetric hss separated by distance @xmath203 was introduced in @xcite : @xmath204 u. \label{two1}\end{aligned}\ ] ] numerical analysis has demonstrated that stationary symmetric solutions of this equation ( they are not available in an analytical form ) are stable , see a typical example in fig .
[ fig4 ] , while all antisymmetric states are unstable , spontaneously transforming into their symmetric counterparts . ) with the delta - functions approximated as per eq .
( [ delta ] ) with @xmath189 .
other parameters are @xmath205 , @xmath206 , and @xmath207.,width=226 ] as shown above in fig .
[ fig3 ] , eq .
( [ cghs ] ) with the single hs described by expression ( [ delta ] ) , where @xmath171 is large enough , supports breathers , instead of stationary pinned modes .
simulations of eq .
( [ two1 ] ) demonstrate that a pair of such broad hss support unsynchronized breathers pinned by each hs , if the distance between them is large enough , hence the breathers virtually do not interact . a completely different effect is displayed in fig . [ fig5 ] , for two broad hss which are set closer to each other : the interaction between the breathers pinned to each
hs transforms them into a _
stationary _ stable symmetric mode .
it is relevant to stress that the transformation of the breather pinned by the isolated hs into a stationary pinned state does not occur for the same parameters .
this outcome is a generic result of the interaction between the breathers , provided that the distance between them is not too large . ) , into a stationary symmetric mode .
parameters are @xmath193 , @xmath208 , @xmath209 , @xmath201 , @xmath210 , and @xmath211.,width=302 ] in addition to the hss embedded into the medium with the uniform nonlinearity , more specific models , in which the nonlinearity is also concentrated at the hss , were introduced in @xcite and spotsexact2 . in particular , the respective system with the double hs is described by the following equation:@xmath212 \left [ \delta \left ( x+l\right ) + \delta \left ( x - l\right ) \right ] u , \label{ad - hoc}\end{aligned}\]]where @xmath70 and @xmath71 are coefficients of the localized kerr nonlinearity and cubic loss , respectively .
these settings may be realized if the nonlinear properties of the waveguides are dominated by narrow doped segments .
although model ( [ ad - hoc ] ) seems somewhat artificial , its advantage is a possibility to find both symmetric and antisymmetric pinned modes in an exact analytical form .
those include both fundamental modes , with exactly two local power peaks , tacked to the hss , and higher - order states , which feature additional peaks between the hss .
an essential finding pertains to the stability of such states , for which the sign of the cubic nonlinearity plays a crucial role : in the sf case [ @xmath72 in eq .
( [ ad - hoc ] ) ] , only the fundamental symmetric and antisymmetric modes , with two local peaks tacked to the hss , may be stable . in this case , all the higher - order multi - peak modes , being unstable , evolve into the fundamental ones . in the case of the sdf cubic nonlinearity [ @xmath73 in eq .
( [ ad - hoc ] ) ] , the hs pair gives rise to _ multistability _ , with up to eight coexisting stable multi - peak patterns , both symmetric and antisymmetric ones .
the system without the kerr term ( @xmath213 ) , the nonlinearity in eq .
( [ ad - hoc ] ) being represented solely by the local cubic loss ( @xmath214 ) is similar to one with the self - focusing or defocusing nonlinearity , if the linear potential of the hs is , respectively , attractive or repulsive , i.e. , @xmath215 or @xmath216 ( note that a set of two local repulsive potentials may stably trap solitons in an effective _ cavity _ between them @xcite ) .
an additional noteworthy feature of the former setting is the coexistence of the stable fundamental modes with robust breathers .
the nonlinear schrdinger equation ( nlse ) in the form of eq .
( [ eq ] ) is a unique example of a @xmath1-symmetric system in which a full family of solitons can be found in an exact analytical form @xcite . indeed , looking for stationary solutions with real propagation constant @xmath35 as @xmath217 , where the @xmath1 symmetry is provided by condition @xmath218 , one can readily find , for the sf and sdf signs of the nonlinearity , respectively , @xmath219with real constants @xmath220 and @xmath221 , cf .
( [ sinh ] ) .
the form of this solution implies that @xmath222 while jumps ( @xmath223 ) of the imaginary part and first derivative of the real part at @xmath224 are determined by the b.c .
produced the integration of the @xmath225- and @xmath226- functions in an infinitesimal vicinity of @xmath45 , cf .
( [ bc]):@xmath227the substitution of expressions ( [ sf ] ) and ( [ sdf ] ) into these b.c .
yields@xmath228which does not depend on @xmath35 and is the same for @xmath229 , and@xmath230which does not depend on the @xmath1 coefficient , @xmath61 .
the total power ( [ p ] ) of the localized mode is @xmath231as seen from eq .
( [ xi2 ] ) , the solutions exist at @xmath232as concerns stability of the solutions , it is relevant to mention that expression ( [ power ] ) with @xmath58 and @xmath59 satisfy , respectively , the vakhitov - kolokolov ( vk ) @xcite and anti - vk " @xcite criteria , i.e. , @xmath233which are necessary conditions for the stability of localized modes supported , severally , by the sf and sdf nonlinearities , hence both solutions have a chance to be stable .
as above [ see eq .
( [ delta ] ) ] , the numerical analysis of the model needs to replace the exact @xmath225-function by its finite - width regularization , @xmath234 . in the present context , the use of the gaussian approximation is not convenient , as , replacing the exact solutions in the form of eqs .
( [ sf ] ) and ( [ sdf ] ) by their regularized counterparts , it is necessary , _ inter alia _ , to replace @xmath235 by a continuous function realized as @xmath236 , which would be a non - elementary function .
therefore , the regularization was used in the form of the lorentzian,@xmath237with @xmath238 @xcite . the first result for the sf nonlinearity , with @xmath58 ( in this case , @xmath239
is also fixed by scaling ) is that , for given @xmath240 in eq .
( [ lorentz ] ) , there is a critical value , @xmath241 , of the @xmath56 coefficient , such that at @xmath242 the numerical solution features a shape very close to that of the analytical solution ( [ sf ] ) , while at @xmath243 the single - peak shape of the solution transforms into a _
double - peak _ one , as shown in fig .
[ fig6](a ) .
in particular , @xmath244 .
@xmath245{fig6a.pdf } \\ ( \mathrm{a } ) \\
\includegraphics[width=2.5in]{fig6b.pdf } \\
( \mathrm{b})% \end{array}% $ ] the difference between the single- and double - peak modes is that the former ones are completely stable , as was verified by simulations of eq .
( [ eq ] ) with regularization ( [ lorentz ] ) , while all the double - peak solutions are unstable .
this correlation between the shape and ( in)stability of the pinned modes is not surprising : the single- and double - peak structures imply that the pinned mode is feeling , respectively , effective attraction to or repulsion from the local defect .
accordingly , in the latter case the pinned soliton is unstable against spontaneous escape , transforming itself into an ordinary freely moving nlse soliton , as shown in fig .
[ fig7 ] .
( b).,width=288 ] figure [ fig8 ] summarizes the findings in the plane of @xmath246 for a fixed propagation constant , @xmath247 .
the region of the unstable double - peak solitons is a relatively narrow boundary layer between broad areas in which the stable single - peak solitons exist , as predicted by the analytical solution , or no solitons exist at all , at large values of @xmath248 . note also that the stability area strongly expands to larger values of @xmath61 as the regularized profile ( [ lorentz ] ) becomes smoother , with the increase of @xmath240 . on the other hand ,
the stability region does not vanish even for very small @xmath240 .
-symmetric solitons , separated by @xmath249 , in the plane of the regularization scale , @xmath240 , and the gain - loss parameter , @xmath250 for fixed @xmath247 , in the system described by eqs .
( [ eq ] ) and ( [ lorentz ] ) with @xmath251 ( the sf nonlinearity ) and @xmath252.,width=288 ] for the same model but with the sdf nonlinearity , @xmath253 , the results are simpler : all the numerically found pinned solitons are close to the analytical solution ( [ sdf ] ) , featuring a single - peak form , and are completely stable .
a generalization of model ( [ eq ] ) , including a nonlinear part of the trapping potential , was studied in @xcite too:@xmath254equation ( [ 02 ] ) also admits an analytical solution for pinned solitons , which is rather cumbersome .
it takes a simpler form in the case when the trapping potential at @xmath45 is purely nonlinear , with @xmath255 , @xmath256 , eq .
( [ xi2 ] ) being replaced by@xmath257 \label{xi3}\]][it is the same for both signs of the bulk nonlinearity , @xmath58 and @xmath258 , while the solution at @xmath259 keeps the form of eqs .
( [ sf ] ) or ( [ sdf ] ) , respectively ] , with total power@xmath260 .
\label{power2}\]]note that expressions ( [ xi3 ] ) and ( [ power2 ] ) depend on the @xmath0 strength @xmath61 , unlike their counterparts ( [ xi2 ] ) ( [ power ] ) .
furthermore , solution ( [ xi3 ] ) exists for all values of @xmath261 , unlike the one given by eq .
( [ xi2 ] ) , whose existence region is limited by condition ( [ > < ] ) .
the numerical analysis demonstrates that these solutions have a narrow stability area ( at small @xmath61 ) for the sf bulk medium , @xmath58 , and are completely unstable for @xmath253 .
as mentioned above , the first example of sdss supported by an hs in lossy media was predicted in the framework of the cmes ( [ o1 ] ) and ( [ o2 ] ) for the bg in a nonlinear waveguide @xcite .
it is relevant to outline this original result in the present review . unlike the basic models considered above , the cme system does not admit exact solutions for pinned solitons , because eqs .
( [ o1 ] ) and ( [ o2 ] ) do not have analytical solutions in the bulk in the presence of the loss terms , @xmath3 , therefore analytical consideration ( verified by numerical solutions ) is only possible in the framework of the perturbation theory , which treats @xmath61 and @xmath122 as small parameters , while the strength of the local potential , @xmath130 , does not need to be small .
this example of the application of the perturbation theory is important , as it provided a paradigm for the analysis of other models , where exact solutions are not available either @xcite . a stationary solution to eqs .
( [ o1 ] ) and ( [ o2 ] ) with @xmath262 and @xmath215 is sought for in the form which is common for quiescent bg solitons @xcite-@xcite : @xmath263where @xmath264 stands for the complex conjugate , @xmath220 is a parameter of the soliton family , and function @xmath102 satisfies equation@xmath265 u+3|u|^{2}u - u^{\ast } = 0 .
\label{reduced}\]]the integration of eq .
( [ reduced ] ) around @xmath45 yields the respective b.c . , @xmath266 , cf .
( [ bc])@xmath267 . as shown in @xcite , an exact soliton - like solution to eq .
( [ reduced ] ) , supplemented by the b.c .
, can be found , following the pattern of the exact solution for the ordinary gap solitons @xcite-@xcite:@xmath268 } \
, , \label{solution}\]]where offset @xmath221 , cf .
( [ sinh ] ) , ( [ sf ] ) , ( [ sdf ] ) , is determined by the relation @xmath269accordingly , the soliton s squared amplitude ( peak power ) is @xmath270from eq .
( [ tan ] ) it follows that the solution exists not in the whole interval @xmath271 , where the ordinary gap solitons are found , but in a region determined by constraint @xmath272 , i.e. , @xmath273 in turn , eq .
( [ interval ] ) implies that the solutions exist only for @xmath274 ( @xmath275 corresponds to the repulsive hs , hence the soliton pinned to it will be unstable ) . in the case of @xmath276 , eqs .
( [ o1 ] ) and ( [ o2 ] ) conserve the total power , @xmath277 ^{2}dx\ , .
\label{e}\]]in the presence of the loss and gain , the exact evolution equation for the total power is @xmath278 |_{x=0}\ , . \label{de
/ dt}\]]if @xmath61 and @xmath122 are treated as small perturbations , the balance condition for the power , @xmath279 , should select a particular solution , from the family of exact solutions ( [ solution ] ) of the conservative model , which remains , to the first approximation , a stationary pinned soliton , cf .
( [ balance ] ) .
the balance condition following from eq .
( [ de / dt ] ) demands @xmath280 \,.$ ] substituting , in the first approximation , the unperturbed solution ( [ solution ] ) and ( [ tan ] ) into this condition , it can be cast in the form of @xmath281 the pinned soliton selected by eq .
( [ complicated ] ) appears , with the increase of the relative gain strength @xmath282 , as a result of a bifurcation .
the inspection of fig .
[ fig9 ] , which displays @xmath283 vs. @xmath282 , as per eq .
( [ complicated ] ) , shows that the situation is qualitatively different for @xmath284 and @xmath285 . in the former case , a _ tangent _
( saddle - node ) bifurcation @xcite occurs at a minimum value @xmath286 of the relative gain , with two solutions existing at @xmath287 @xmath288 .
analysis of eq .
( complicated ) demonstrates that , with the variation of @xmath130 , the value @xmath286 attains an absolute minimum , @xmath289 , at @xmath290 . with the increase of @xmath282 , the lower unstable solution branch , that starts at the bifurcation point
[ see fig . [ fig9](a ) ] hits the limit point @xmath291 at @xmath292 , where it degenerates into the zero solution , according to eq .
( [ amplitude ] ) .
the upper branch generated , as a stable one , by the bifurcation in fig .
fig9(a ) continues until it attains the maximum possible value , @xmath293 , which happens at @xmath294 in the course of its evolution , this branch may acquire an instability unrelated to the bifurcation , see below . in the case
@xmath285 , the situation is different , as the saddle - node bifurcation is imaginable in this case , occurring in the unphysical region @xmath295 , see fig . [ fig9](b ) .
the only physical branch of the solutions appears as a stable one at point @xmath296 , where it crosses the zero solution , making it the unstable . however , as well as the above - mentioned branch , the present one may be subject to an instability of another type .
this branch ceases to be a physical one at point ( [ max ] ) . at the boundary between the two generic cases considered above ,
i.e. , at @xmath290 , the bifurcation occurs exactly at @xmath297 , see fig .
[ fig9](c )
. the situation is different too in the case @xmath298 [ see fig .
1(d ) ] , when the hs has no local - potential component , and eq .
( [ complicated ] ) takes the form of @xmath299 in this case , the solution branches do not cross axis @xmath300 in fig . [ fig9](d ) .
the lower branch , which asymptotically approaches the @xmath301 axis , is unstable , while the upper one might be stable within the framework of the present analysis .
however , numerical results demonstrate that , in the case of @xmath298 , the soliton is always unstable against displacement from @xmath45 .
as done above for the cgle model with the hs , see eqs .
( [ linearized ] ) - ( [ lambda ] ) , it is relevant to analyze the stability of the zero background , and the relation between the onset of the localized instability , driven by the hs , and the emergence of the stable pinned mode , in the framework of the cmes . to this end ,
a solution to the linearized version of the cmes is looked for as @xmath302 with @xmath303 .
a straightforward analysis makes it possible to eliminate constant @xmath304 and find the instability growth rate @xcite : @xmath305thus , the zero solution is subject to the hs - induced instability , provided that the local gain is strong enough : @xmath306note that the instability is impossible in the absence of the local potential @xmath307 . the instability - onset condition ( [ critical ] )
simplifies in the limit when both the loss and gain parameters are small ( while @xmath130 is not necessarily small ) : @xmath267@xmath308as seen in fig .
[ fig9 ] , the pinned mode emerges at critical point ( cr2 ) , for @xmath309 , while at @xmath284 it emerges at @xmath310 @xmath311 .
first , direct simulations of the conservative version of cmes ( [ o1 ] ) , ( [ o2 ] ) , with @xmath276 ( and @xmath215 ) and an appropriate approximation for the delta - function , have demonstrated that there is a very narrow stability interval in terms of parameter @xmath220 [ see eqs .
( [ reduced])-([interval ] ) , close to ( slightly larger than ) @xmath312 , such that the solitons with @xmath313 decay into radiation , while initially created solitons with @xmath314 spontaneously evolve , through emission of radiation waves , into a stable one with @xmath315 @xcite .
this value weakly depends on @xmath130 .
for instance , at @xmath316 , the stability interval is limited to @xmath317 , while at a much larger value of the local - potential strength , @xmath318 , the interval is located between boundaries @xmath319 . in this connection , it is relevant to mention that in the usual bg model , based on eqs .
( [ o1 ] ) , ( [ o2 ] ) with @xmath298 , the quiescent solitons are stable at @xmath320 barash , trillo
. direct simulations of the full cme system ( [ o1 ] ) , ( [ o2 ] ) , which includes weak loss and local gain , which may be considered as perturbations , produce stable dissipative gap solitons with small but persistent internal oscillations , i.e. , these are , strictly speaking , breathers , rather than stationary solitons @xcite .
an essential finding is that the average value of @xmath220 in such robust states may be essentially larger than the above - mentioned @xmath321 selected by the conservative counterpart of the system , see particular examples in table 1 . with the increase of the gain strength @xmath122 , the amplitude of the residual oscillations increases , and , eventually , the oscillatory state becomes chaotic , permanently emitting radiation waves , which increases the effective loss rate that must be compensated by the local gain .
taking values of @xmath61 and @xmath122 small enough , it was checked if the numerically found solutions are close to those predicted by the perturbation theory in the form of eqs .
( [ solution ] ) and ( [ tan ] ) , with @xmath220 related to @xmath61 and @xmath122 as per eq .
( complicated ) .
it has been found that the quasi - stationary solitons ( with the above - mentioned small - amplitude intrinsic oscillations ) are indeed close to the analytical prediction .
the corresponding values of @xmath220 were identified by means of the best fit to expression ( [ solution ] ) .
then , for the so found values of @xmath220 and given loss coefficient @xmath61 , the equilibrium gain strength @xmath122 was calculated as predicted by the analytical formula ( [ complicated ] ) , see a summary of the results in table 1 .
it is seen in the table that the numerically found equilibrium values of @xmath122 exceeds the analytically predicted counterparts by @xmath322 , which may be explained by the additional gain which is necessary to compensate the permanent power loss due to the emission of radiation . [ cols="<,<,<,<,<",options="header " , ] table 1 .
values of the loss parameter @xmath61 at which quasi - stationary stable pinned solitons were found in simulations of the cme system , ( [ o1 ] ) , ( [ o2 ] ) , by adjusting gain @xmath122 , for fixed @xmath323 .
values of the soliton parameter , @xmath220 , which provide for the best fit of the quasi - stationary solitons to the analytical solution ( [ solution ] ) are displayed too .
@xmath324 is the gain coefficient predicted , for given @xmath61 , @xmath220 , and @xmath325 , by the energy - balance equation ( [ complicated ] ) , which does not take the radiation loss into account .
the rightmost column shows the relative difference between the numerically found and analytically predicted gain strength , which is explained by the necessity to compensate additional radiation loss .
it has also been checked that the presence of nonzero attractive potential with @xmath215 is necessary for the stability of the gap solitons pinned to the hs in the bg model .
in addition to the analysis of stationary pinned modes , collisions between a gap soliton , freely moving in the weakly lossy medium , with the hs were also studied by means of direct simulations @xcite .
the collision splits the incident soliton into the transmitted and trapped components .
the 1d version of the discrete model based on eq .
( [ eq : gl ] ) , i.e. , @xmath326 \delta _ { n,0}u_{n } , \label{1d}\end{aligned}\]]makes it possible to gain insight into the structure of lattice solitons supported by the hot site " , at which both the gain and nonlinearity are applied , as in that case an analytical solution is available @xcite .
it is relevant to stress that the present model admits stable localized states even in the case of the unsaturated cubic gain , @xmath75 , see details below . the known _ staggering transformation _ @xcite ,
@xmath327 , where the asterisk stands for the complex conjugate , reverses the signs of @xmath92 and @xmath70 in eq .
( [ 1d ] ) . using this option ,
the signs are fixed by setting @xmath215 ( the linear discrete potential is attractive ) , while @xmath328 or @xmath329 corresponds to the sf and sdf nonlinearity , respectively .
separately considered is the case of @xmath213 , when the nonlinearity is represented solely by the cubic dissipation localized at the hs .
dissipative solitons with real propagation constant @xmath35 are looked for by the substitution of @xmath330 in eq .
( [ 1d ] ) .
outside of the hs site , @xmath331 , the linear lattice gives rise to the exact solution with real amplitude @xmath332 , @xmath333and complex @xmath334 , localized modes corresponding to @xmath335 .
the analysis demonstrates that the amplitude at the hs coincides with @xmath332 , i.e. , @xmath336 .
then , the remaining equations at @xmath78 and @xmath80 yield a system of four equations for four unknowns , @xmath332 , @xmath337 , @xmath338 , and @xmath35 : @xmath339this system was solved numerically
. the stability of the discrete sdss was analyzed by the computation of eigenvalues for modes of small perturbations @xcite and verified by means of simulations of the perturbed evolution .
examples of stable discrete solitons can be seen below in fig .
[ fig12 ] . in the linear version of the model , with @xmath340 in eq .
( [ 1d ] ) , amplitude @xmath332 is arbitrary , dropping out from eqs .
( [ eq : sys ] ) . in this case , @xmath122 may be considered as another unknown , determined by the balance between the background loss and localized gain , which implies structural instability of the stationary trapped modes in the linear model against small variations of @xmath122 . in the presence of the nonlinearity
, the power balance is adjusted through the value of the amplitude at given @xmath122 , therefore solutions can be found in a range of values of @xmath122 .
thus , families of pinned modes can be studied , using linear gain @xmath122 and cubic gain / loss @xmath71 as control parameters ( in the underlying photonic lattice , their values may be adjusted by varying the intensity of the external pump ) .
the most interesting results were obtained for the _ unsaturated nonlinear gain _ , i.e. , @xmath75 in eq .
( [ 1d ] ) .
figure [ fig10 ] shows amplitude @xmath332 of the stable ( solid ) and unstable ( dashed ) pinned modes as a function of linear gain @xmath44 and cubic gain .
the existence of stable subfamilies in this case is a noteworthy finding . in particular , at @xmath341 there exists a family of stable pinned modes in the region of @xmath342 , with the amplitude ranging from @xmath343 to @xmath344 . outside this region ,
solutions decay to zero if the linear gain is too weak ( @xmath345 ) , or blow up at @xmath346 .
figure [ fig10 ] shows that the bifurcation of the zero solution @xmath347 into the pinned mode takes place at a particular value @xmath348 , which is selected by the above - mentioned power - balance condition in the linear system .
the existence of the stable pinned modes in the absence of the gain saturation being a remarkable feature , the stability region is , naturally , much broader in the case of the cubic loss , @xmath74 .
figure [ fig11 ] shows the respective solution branches obtained with the sdf nonlinearity
. when the cubic loss is small , e.g. , @xmath349 , there are two distinct families of stable modes , representing broad small - amplitude ( @xmath350 ) and narrow large - amplitude ( @xmath351 ) ones .
these two stable families are linked by an unstable branch with the amplitudes in the interval of @xmath352 .
the two stable branches coexist in the interval of values of the linear gain @xmath353 , where the system is _
bistable_. figure [ fig12 ] shows an example of the coexisting stable modes in the bistability region . in simulations
, a localized input evolves into either of these two stable solutions , depending on the initial amplitude . while the pinned modes may be stable against small perturbations under the combined action of the self - defocusing nonlinearity ( @xmath329 ) and unsaturated nonlinear gain ( @xmath354 )
, one may expect fragility of these states against finite - amplitude perturbations .
it was found indeed that sufficiently strong perturbations destroy those modes @xcite . in the case of the sf nonlinearity , @xmath328 ,
the pinned - mode branches are shown in fig .
[ fig13 ] , all of them being unstable without the cubic loss , i.e. , at @xmath354 . for the present parameters , all the solutions originate , in the linear limit , from the power - balance point @xmath355 .
the localized modes remain stable even at very large values of @xmath194 .
lastly , in the case of @xmath213 , when the nonlinearity is represented solely by the cubic gain or loss , all the localized states are unstable under the cubic gain , @xmath75 , and stable under the cubic loss , @xmath74 .
this article presents a limited review of theoretical results obtained in recently studied 1d and 2d models , which predict a generic method for supporting stable spatial solitons in dissipative optical media , based on the use of the linear gain applied in narrow active segments ( hss , hot spots " ) implanted into the lossy waveguide . in some cases
, the unsaturated cubic localized gain may also support stable spatial solitons , which is a counter - intuitive finding . in view of the limited length of the article
, it combined a review of the broad class of such models with a more detailed consideration of selected 1d models where exact or approximate analytical solutions for the dissipative solitons are available .
naturally , the analytical solutions , in the combination with their numerical counterparts , provide a deeper understanding of the underlying physical models .
a relevant issue for the further development of the topic is a possibility of the existence of _ asymmetric modes _ supported by symmetric double hss , i.e. , the analysis of the spontaneous symmetry breaking in this setting @xcite .
thus far , such a result was not demonstrated in a clear form .
another challenging extension may be a possibility of _ dragging _ a pinned 1d or 2d soliton by a hs _ moving _ across a lossy medium .
i appreciate the invitation from prof .
g. swartzlander , the editor - in - chief of j. opt .
am . b , to draft and submit this mini - review .
my work on the present topic has greatly benefited from collaborations with many colleagues , including p. l. chu ( deceased ) , n. b. aleksi , j. atai , o. v. borovkova , k. w. chow , p. colet , m. c. cross , e. ding , r. driben , w. j. firth , d. gomila , y .- j .
he , y. v. kartashov , e. kenig , s. k. lai , c. k. lam , h. leblond , y. li , r. lifshitz , v. e. lobanov , w. c. k. mak , a. marini , t. mayteevarunyoo , d. mihalache , a. a. nepomnyashchy , p. v. paulau , h. sakaguchi , v. skarka , d. v. skryabin , a. y. s. tang , l. torner , c. h. tsang , v. a. vysloukh , p. k. a. wai , h. g. winful , f. ye , and x. zhang .
f. ye , d. mihalache , b. hu , and n. c. panoiu , subwavelength plasmonic lattice solitons in arrays of metallic nanowires " , phys .
lett . * 104 * , 106802 ( 2010 ) m. c. cross and p. c. hohenberg , pattern formation outside of equilibrium " , rev .
65 , 851 - 1112 ( 1993 ) .
p. v. paulau , d. gomila , p. colet , b. a. malomed , and w. j. firth , from one- to two - dimensional solitons in the ginzburg - landau model of lasers with frequency - selective feedback " , phys .
e * 84 * , 036213 ( 2011 ) .
d. k. serkland , g. d. bartolini , a. agarwal , p. kumar , and w. l. kath , `` pulsed degenerate optical parametric oscillator based on a nonlinear - fiber sagnac interferometer '' , opt
* 23 * , 795 - 797 ( 1998 ) .
f. k. abdullaev , v. v. konotop , m. salerno , and a. v. yulin , dissipative periodic waves , solitons , and breathers of the nonlinear schrdinger equation with complex potentials " , phys .
e * 82 * , 056606 ( 2010 ) .
v. skarka , n. b. aleksi , h. leblond , b. a. malomed , and d. mihalache , varieties of stable vortical solitons in ginzburg - landau media with radially inhomogeneous losses " , phys .
lett . * 105 * , 213901 ( 2010 ) .
z. lin , h. ramezani , t. eichelkraut , t. kottos , h. cao , and d. n. christodoulides , unidirectional invisibility induced by @xmath56-symmetric periodic structures " , phys .
* 106 * , 213901 ( 2011 ) .
n. i. karachalios , h. e. nistazakis , and a. n. yannacopoulos , asymptotic behavior of solutions of complex discrete evolution equations : the discrete ginzburg - landau equation " , discrete contin .
* 19 * , 711 - 736 ( 2007 ) .
a. christ , t. zentgraf , j. kuhl , s. g. tikhodeev , n. a. gippius , and h. giessen , optical properties of planar metallic photonic crystal structures : experiment and theory , phys .
b * 70 * , 125113 ( 2004 )
. y. s. bian , z. zheng , x. zhao , y. l. su , l. liu , j. s. liu , j. s. zhu , and t. zhou , guiding of long - range hybrid plasmon polariton in a coupled nanowire array at deep - subwavelength scale " , ieee phot . tech .
* 24 * , 1279 - 1281 ( 2012 ) .
x. zhang , j. chai , j. huang , z. chen , y. li , and b. a malomed , discrete solitons and scattering of lattice waves in guiding arrays with a nonlinear @xmath1-symmetric defect " , opt . exp . *
22 * , 13927 - 13939 ( 2014 ) .
v. skarka , n. b. aleksi , h. leblond , b. a. malomed , and d. mihalache , varieties of stable vortical solitons in ginzburg - landau media with radially inhomogeneous losses " , phys .
* 105 * , 213901 ( 2010 ) . | the creation of stable 1d and 2d localized modes in lossy nonlinear media is a fundamental problem in optics and plasmonics .
this article gives a short review of theoretical methods elaborated for this purpose , using localized gain applied at one or several hot spots " ( hss ) .
the introduction surveys a broad class of models for which this approach was developed .
other sections focus in some detail on basic 1d continuous and discrete systems , where the results can be obtained , partly or fully , in an analytical form ( verified by comparison with numerical results ) , which provides a deeper insight into the nonlinear dynamics of optical beams in dissipative nonlinear media . in particular , considered are the single and double hs in the usual waveguide with the self - focusing ( sf ) or self - defocusing ( sdf ) kerr nonlinearity , which gives rise to rather sophisticated results , in spite of apparent simplicity of the model ; solitons attached to a @xmath0-symmetric dipole embedded into the sf or sdf medium ; gap solitons pinned to an hs in a bragg grating ( bg ) ; and discrete solitons in a 1d lattice with a hot site " .
_ list of acronyms _ : 1d one - dimensional ; 2d two - dimensional ; b.c .
boundary conditions ; bg bragg grating ; cgle complex ginzburg - landau equation ; cme coupled - mode equations ; hs hot spot ; nlse nonlinear schrdinger equation ; @xmath1 parity - time ( symmetry ) ; sds spatial dissipative soliton ; sf self - focusing ; sdf self - defocusing ; vk vakhitov - kolokolov ( stability criterion ) ; ws warm spot ocis numbers : 190.6135 ; 240.6680 ; 230.4480 ; 190.4360 |
uniaxially anisotropic heisenberg antiferromagnets in a magnetic field have been studied quite extensively in the past , both experimentally and theoretically .
typically , they display , at low temperatures , the antiferromagnetic phase and , when increasing the field , the spin flop phase @xcite .
a prototypical model describing these phases is the heisenberg model with a uniaxial exchange anisotropy , the xxz model @xmath0 \ ; - \ ; h \sum\limits_{i } s_i^z\ ] ] where @xmath1 is the exchange coupling between classical spins , @xmath2 , of length one at neighboring sites , @xmath3 and @xmath4 , of a simple cubic lattice , @xmath5 is the exchange anisotropy , @xmath6 , and @xmath7 is the applied magnetic field along the easy axis , the @xmath8axis .
the phase diagram of the model has been studied several years ago , using mean field theory @xcite , monte carlo simulations @xcite , and high temperature series expansions @xcite , suggesting the transition between the antiferromagnetic ( af ) and spin flop ( sf ) phases to be of first order and the boundaries of the paramagnetic phase to the af and sf phases to be continuous transitions in the ising and xy universality classes .
based on renormalization group analysis in one loop order , a bicritical point in the heisenberg universality class has been proposed , at which the three different phases meet @xcite . however , this scenario has been scrutinized when doing renormalization group calculations in high loop order @xcite , where the bicritical point is found to be unstable against a tetracritical biconical point @xcite , which , in turn , may be unstable towards transitions of first order in the vicinity of the multicritical point of the three phases .
the seemingly conflicting descriptions may be reconciled by a new renormalization group analysis in two loop order @xcite . as has been noted very recently @xcite , not only af and sf phases , but also biconical ( bc ) structures ,
1 , may play an important role in the xxz model .
indeed , such bc structures are degenerate ground states at the critical field separating af and sf configurations at zero temperature . for the xxz model on a square lattice , these degenerate bc fluctuations seem to lead to a narrow disordered phase intervening between the af and sf phases at low temperatures @xcite .
the importance of bc structures for the three dimensional xxz antiferromagnet , where they are also present as degenerate ground states at the special field , had not been studied in any detail so far .
biconical structures may be ground states even in a finite range of magnetic fields , giving then rise to an ordered bc phase at low temperatures , when introducing in the xxz model further anisotropy terms or longer
range exchange interactions , as it is known for many years @xcite .
this feature has been confirmed in recent simulations when adding a quadratic single
ion anisotropy to the xxz model on the square lattice @xcite .
the related lowest order single
ion term of cubic symmetry may be written in the form @xcite @xmath9\ ] ] where @xmath10 denotes the strength of the cubic anisotropy. the sign of @xmath10 determines whether the spins tend to align along the cubic axes , for @xmath11 , or , for @xmath12 , in the diagonal directions of the lattice . because of these tendencies , the bc structures show no full rotational invariance in the @xmath13plane perpendicular to the easy axis , the @xmath8axis , in contrast to the xxz case .
when an ordered bc phase exists at low temperatures , for example due to the cubic anisotropy , intricate multicritical behavior may show up , including a tetracritical biconical point , at which the af , sf , bc , and paramagnetic phases meet , as has been discussed before @xcite , applying mean - field theory and renormalization group arguments .
experimentally , many antiferromagnets with uniaxial anisotropy have been investigated , quasi two
dimensional magnets @xcite as well as three dimensional magnets such as gdalo@xmath14 , nicl@xmath154h@xmath15o , mnf@xmath15 or mn@xmath15(si , ge)s@xmath16 @xcite . while we shall deal here with the theoretical analysis of the models
, results may turn out to be useful for interpreting specific experiments in future work .
the aim of our paper is to study , especially , the role of biconical , non collinear structures in three dimensional classical heisenberg antiferromagnets with uniaxial exchange anisotropy and cubic anisotropy . both ground state considerations and monte carlo techniques are applied .
plane for the full hamiltonian , @xmath17= @xmath18 + @xmath19 , with exchange anisotropy @xmath5= 0.8 .
the full circle denotes the highly degenerate point in the xxz model . ]
the paper is organized as follows : first , results of ground state calculations are presented , and then monte carlo findings on thermal properties and phase diagrams will be discussed , for the xxz model and its extension .
the article will be concluded by a short summary .
the ground states of the full hamiltonian , @xmath17= @xmath18 + @xmath20 , eqs . ( 1 ) and ( 2 ) , may be determined by minimization of the energy , e.g. , with respect to the azimuthal angle , @xmath21 , i.e. the angle between the projection of the spin vector in the @xmath13plane and the @xmath22axis , and the @xmath8component of the spin vector @xcite . as usual , we consider spin structures with two sublattices , @xmath23 and @xmath24 , where neighboring sites on the cubic lattice belong to different sublattices . in general , the minimization may be easily done numerically . in fig .
2 , resulting ground states in the @xmath25plane are depicted , setting the exchange anisotropy @xmath5 equal to 0.8 , as before @xcite .
, in the case of ( a ) @xmath26 , bc2 , and ( b ) @xmath27 , bc1 . , scaledwidth=90.0% ] in the case of a vanishing cubic anisotropy , @xmath28 , i.e. for the xxz antiferromagnet , one encounters , when increasing the field , af , sf , and ferromagnetic ground state configurations . in complete analogy to the model on a square lattice , the ground state is highly degenerate in bc structures at the critical field @xmath29 separating the af and sf structures , see figs . 1 and 2 . for cubic lattices
, one has @xmath30 .
the non collinear bc spin structures may be characterized by two different polar or tilt angles of the spin vectors on the two sublattices @xmath23 and @xmath24 , @xmath31 and @xmath32 , as shown in fig .
obviously , for the sf structures , the angles are identical , @xmath33 , depicted in fig .
1(c ) . at @xmath29 ,
the two tilt angles of the degenerate bc structures are interrelated by @xmath34 interpolating continuously between the af and sf configurations , when varying one of the tilt angles @xcite .
of course , the rotational invariance of the spin components in the @xmath13plane leads to an additional degeneracy in the bc and sf configurations .
let us now consider positive values of @xmath10 , @xmath35 .
for vanishing field , @xmath7= 0 , and @xmath36 , see fig .
2 , the simple af structure is replaced by a tilted antiferromagnetic ( taf ) configuration .
spins on the two sublattices point still in opposite directions , but not along the easy axis . indeed , the spins are oriented more and more towards diagonals of the lattice with increasing positive cubic anisotropy @xmath10 @xcite . applying , for @xmath26 , an external field , bc structures may occur ,
in which neighboring spins have different tilt angles , @xmath31 and @xmath32 .
in contrast to the xxz case , there is no full rotational symmetry in the azimuthal angles , @xmath21 , of the spins . instead , for @xmath35 , the spins order along the diagonals of the lattice , and @xmath21 is discretized , taking the values , @xmath37 , where @xmath38 is an integer .
the @xmath13spin components are ordered antiferromagnetically , @xmath39= @xmath40 .
the resulting discretized biconical structure , bc2 , is sketched in fig .
obviously , the sf configurations are discretized in the azimuthal angle in the same way as the bc structures . when fixing @xmath10 and varying the field @xmath7 , the two tilt angles , @xmath31 and @xmath32 , change continuosly when going from the af ( or taf ) to the bc2 structures . in the bc2 region ,
the tilt angles , @xmath31 and @xmath32 , vary continuously as well . on the other hand
, they seem to jump when going from the bc2 to the sf region .
note that the tilt angle of the sf structure , @xmath41 , at the border to the bc2 region , see fig .
2 , depends only weakly on the strength @xmath10 of the cubic anisotropy . for @xmath42
, the azimuthal angles of the spins , both for biconical and sf structures , take the values @xmath43 , @xmath38 being an integer , reflecting the fact that the cubic anisotropy now favors alignment of the spins along the cubic axes .
the @xmath13spin components are ordered antiferromagnetically , @xmath39= @xmath40 , as for @xmath35 .
biconical ground states occur for sufficiently strong negative values of @xmath10 , as shown in fig .
the resulting bc1 structure is shown in fig .
the tilt angles @xmath31 and @xmath32 change continuously in the bc1 region , when varying the cubic anisotropy @xmath10 and the field @xmath7 . at the boundaries of the bc1 region to the sf ,
af , and ferromagnetic regions the tilt angles seem to jump , except between the bc1 and sf regions near the triple point of these two and the ferromagnetic regions , see fig .
, as obtained from monte carlo simulations described in the text . ]
in our simulations of thermal properties of the xxz and the full model , @xmath17 , we applied the metropolis algorithm with single spin flips .
lattices of @xmath44 sites , employing periodic boundary conditions , with @xmath45 ranging from 4 to 32 , allowed to do useful finite size analyses . to obtain thermal averages , we performed , at fixed model parameters and temperature , several independent runs , with distinct random numbers , each run consisting of , at least , @xmath46 monte carlo steps per site for the larger systems .
we recorded standard thermodynamic quantities such as the specific heat , @xmath47 , the ( absolute ) total magnetization , @xmath48 , longitudinal , @xmath49 , and transverse , @xmath50 , staggered magnetizations , being the order parameters in the af and sf phases , as well as the corresponding susceptibilities , @xmath51 , @xmath52 , and @xmath53 .
in addition , we computed the binder cumulants @xcite of the two different order parameters , @xmath54 and @xmath55 , and related histograms . to gain microscopic insights , especially , on bc structures , we also recorded probability functions of the tilt angles , such as the probability @xmath56 for finding the two angles , @xmath31 and @xmath32 , at neighboring sites , and the probability @xmath57 for encountering the tilt angle @xmath58 @xcite . for finding the tilt angles @xmath31 and @xmath32 at neighboring sites in the xxz model with @xmath59 in the vicinity of the transition between the af and sf phases , at @xmath60 and @xmath61 .
the solid line corresponds to eq .
the system size is @xmath62 . ]
let us first consider the xxz model on a cubic lattice , eq .
its phase diagram has been determined , using monte carlo techniques , already some time ago @xcite .
certainly , in present simulations the accuracy has been improved significantly .
the resulting phase diagram is depicted in fig.4 , where the phase boundaries have been estimated by standard finite size extrapolations @xcite for thermodynamic quantities and the binder cumulant .
the phase boundary lines deviate somewhat from the ones estimated before @xcite , which we attribute to the improved statistics of the present study .
we also identified the type of transitions and , in the case of continuous transitions , their universality classes , by monitoring , e.g. , the finite size dependences of the peak heights of the susceptibilities and specific heat , to estimate critical exponents , see below . in agreement with previous suggestions @xcite
, the transition between the af and sf phases is found to be of first order , with , for instance , the maximal staggered susceptibilities increasing with system size proportionally to @xmath44 , characteristic for such transitions @xcite .
moreover , as a remainder of the degeneracy of the ground state at @xmath29 , biconical structures prevail near that transition at low temperatures , as one may easily observe in @xmath63 , see fig .
5 . as included in the figure , the dominant bc fluctuations are quite close to those expected from the degenerate ground states , eq .
note that @xmath63 displays simultaneously local maxima at positions belonging to the af and sf structures , as shown in fig .
5 . the phenomenon gets more pronounced when increasing the system size , and it reflects coexistence of the af and sf phases at a first order transition .
this behavior is in marked contrast to that of the xxz antiferromagnet on a square lattice , where the bc fluctuations seem to lead to a disordered phase between the af and sf phases @xcite .
the transition from the paramagnetic phase to the af phase is found to be in the ising universality class , while the transition from the paramagnetic phase to the sf phase is found to belong to the xy universality class .
for instance , we estimated canonical @xcite asymptotic critical exponents describing the size dependence of the peak height of the staggered susceptibilties , @xmath64 , @xmath65 , and analogously for the transverse susceptibility , see fig .
our estimates agree nicely with the known , accurate values obtained from renormalization group calculations in high loop order for both universality classes @xcite .
similarily , our estimates @xcite for the critical binder cumulants @xmath66 and @xmath67 at the two types of transitions are close to those obtained for the three
dimensional nearest
neighbor ising and xy models @xcite .
note , that the critical binder cumulant may change , within a given universality class , due to , e.g. , spatially anisotropic interactions @xcite .
this possible pitfall seems to play no role in the xxz model . of the staggered longitudinal susceptibility , to locate the multicritical point in the xxz model .
the arrow marks the asymptotic critical exponent in the ising universality class @xcite . ] analyzing critical exponents and critical binder cumulants , we locate the multicritical point at @xmath68 and @xmath69 , improving the previous estimate @xcite . to obtain the estimate
, we proceed as exemplified for @xmath52 in fig .
we determine the _ effective _ , size dependent critical exponent @xmath70 of the height of the peak in the staggered longitudinal susceptibility @xmath52 for the simulated system sizes at various fixed temperatures close to the boundary of the af phase .
one observes a pronounced increase of the effective exponent in a small interval of temperatures .
this behavior signals the change from a continuous transition of ising type , with @xmath71 @xcite , to a transition of first order , where the asymptotic exponent is expected to be 3 .
a pronounced increase occurs simultaneously , within the temperature resolution of the simulations , for the effective exponent of the staggered transverse susceptibility @xcite .
accordingly , it indicates a transition of first order between the af and sf phases .
the ratio of the two staggered susceptibilities exhibits an extremum which height is largely independent of system size in the vicinity of the multicritical point .
such a behavior seems to be consistent with a bicritical heisenberg point .
these observations and interpretations on the character of the multicritical point have to be viewed with care , having in mind controversial renormalization group arguments .
early renormalization group calculations in one loop order suggested a bicritical heisenberg point @xcite . later work , in five loop order , found the biconical point to be the stable one @xcite , however , not ruling out a triple point , at which at least one of the transition lines to the paramagnetic phase is , eventually close to the multicritical point , of first order .
a most recent analysis indicates that all three scenarios are possible , depending on the model parameters @xcite . in our case ,
@xmath59 , the monte carlo simulations seem to favor a bicritical point , with ising and xy lines of the paramagnetic phase meeting the line of first order between the af and sf phases .
of course , a crossover to a different multicritical behavior in the immediate vicinity of the multicritical point may happen .
simulational clarification of this aspect may need an enormous amount of computer time .
, with a positive cubic anisotropy @xmath10 at fixed field , @xmath72 = 1.8 .
] let us now turn to the monte carlo results for the full hamiltonian , @xmath17= @xmath18 + @xmath19 , eqs .
( 1 ) and ( 2 ) , including exchange and cubic anisotropies .
we simulated thermal properties for a few selected cases , where discretized biconical structures play an important role .
as before , we set the exchange anisotropy @xmath5 equal to 0.8 . for positive cubic anisotropy , @xmath26 , favoring orientations of the spins along the diagonals of the cubic lattice , a discretized ordered bc2 phase may arise from the corresponding ground state , see fig .
this behavior is illustrated in fig .
7 , when fixing the field , @xmath72= 1.8 , and varying @xmath10 . at small values of @xmath10
, there is an af ordering at low temperatures . above a critical value of @xmath10 , @xmath73 , the low
temperature phase is , indeed , of bc2 type . in that case
, increasing the temperature at fixed @xmath10 , one goes from the bc2 to the af phase before then entering the disordered paramagnetic phase .
the transitions between the two ordered phases and between the af and disordered phases have been located from monitoring the staggered magnetizations and susceptibilities , the specific heat , and binder cumulants .
an example is shown in fig.8 , where the staggered longitudinal and transverse magnetizations are displayed , as a function of temperature , at @xmath74= 1.0 . at the transition between the af and bc2 phases , the absolute staggered transverse magnetization , @xmath75 ,
drops rather sharply , with an accompanying anomaly in the absolute staggered longitudinal magnetization , @xmath76 .
the staggered longitudinal magnetization is the order parameter of the af phase , and it decreases rapidly on approach to the paramagnetic phase .
= 1.8 , and at fixed cubic anisotropy @xmath74= 1.0 .
systems with @xmath77 sites have been simulated .
note that error bars are small compared to symbol sizes . ]
as expected , the transition from the af to the disordered phase seems to be ising like , as follows , e.g. from the critical exponent describing the size dependence of the height of the peak in @xmath52 .
the transition from the bc2 to the af phase is found to be consistent with the xy universality class @xcite .
indeed , the transition between the af and biconical phases has been predicted to belong to the xy universality class in an early renormalization group analysis @xcite .
furthermore , the cubic anisotropy is expected @xcite to be an irrelevant perturbation in the three dimensional xy case . as depicted in fig.7
, there is no multicritical point of the three phases at the values of @xmath10 we studied .
a multicritical point may occur in the @xmath78plane , when fixing @xmath10 , @xmath35 . indeed ,
our simulation results , especially at @xmath79 and 1.5 @xcite , show a phase diagram comprising the af , bc2 , sf , and paramagnetic phases .
however , much more extensive simulations , being beyond the scope of the present study , would be needed to study the possible multicritical points in detail .
especially , the order of the transition between the bc2 and sf phases deserves a careful analysis .
a transition of first order , as suggested by the ground state analysis , would preclude a tetracritical point @xcite . finally ,
we briefly mention results for the case of negative cubic anisotropy , @xmath27 , favoring spin orientations along the cubic axes . as follows from the ground state considerations , see fig.2 , biconical structures only show up at relatively large negative values of @xmath10 .
moreover , the tilt angles usually seem to jump when going from the bc1 region to one of its neighboring , i.e. , ferromagnetic , af , or sf , regions .
our monte carlo simulations , especially at @xmath74= -2 and varying the field at fixed temperatures @xcite , suggest that there is an ordered bc1 phase which , however , gets destabilized at quite low temperatures .
thorough analyses of full phase diagrams and the types of the transition are well beyond the scope of the present work .
classical heisenberg antiferromagnets on cubic lattices with uniaxial exchange , @xmath5 , and cubic , @xmath10 , anisotropies in a magnetic field , @xmath7 , have been studied , using ground state considerations and monte carlo techniques .
in addition to antiferromagnetic ( af ) and spin flop ( sf ) structures non collinear , biconical ( bc ) structures are observed , depending on the strength of the anisotropies , the field , and temperature .
they may occur as ground states , which , at low temperatures , may contribute to thermal fluctuations of bc type or may give rise to ordered bc phases .
specifically , in the xxz antiferromagnet ( @xmath10= 0 ) biconical structures are present as degenerate ground states at the critical field separating the af and sf states . at low temperatures near the transition between the af and sf phases ,
bc fluctuations show up , but they do not lead to a distinct phase , in contrast , for instance , to the situation for the xxz magnet on a square lattice , where the af and sf phases are separated by a narrow disordered phase . moreover , in this study , the multicritical point of the three dimensional xxz model , at which the af , sf , and paramagnetic phases meet , has been located accurately for @xmath5= 0.8 .
that point seems to have bicritical character , being the intersection of two continuous transition lines between the af and sf phases to the paramagnetic phase with a transition line of first order between the af and sf phases .
note that we did not rule out a crossover to a different character in the immediate vicinity of the multicritical point .
the cubic anisotropy @xmath10 may lead , depending on its sign , to a tendency of the spin orientations along the diagonals of the lattice , @xmath35 , or along the cubic axes , @xmath27 . in both cases ,
bc as well as sf structures have now discretized preferred orientations of the @xmath13components of the spins , thereby breaking the rotational symmetry of the xxz model . for positive cubic anisotropy , @xmath26
, biconical configurations occur as ground states next to the sf and af phases , giving rise to an ordered bc2 phase .
the transition to the af phase is found to be consistent with being in the xy universality class , where the cubic anisotropy is expected to be an irrelevant perturbation .
phase diagrams with three ordered phases , af , sf , and bc2 , and a paramagnetic phase have been determined . at small negative cubic anisotropy , @xmath27
, there are no biconical ground states .
however , such structures , of bc1 type , may be stabilized at larger negative values of @xmath10 , between the ferromagnetic and the af or sf ground states .
the biconical ground states then give rise to an ordered bc1 phase at sufficiently low temperatures .
we should like to thank martin holtschneider , david landau , and david peters for useful discussions .
we also thank reinhard folk for very helpful conversations and information on pertinent calculations prior to publication .
l. nel , ann .
paris * 5 * , 232 ( 1936 ) . c. j. gorter and t. van peski - tinbergen , physica ( utr . ) * 22 * , 273 ( 1956 ) .
d. p. landau and k. binder , phys .
b * 17 * , 2328 ( 1978 ) .
o. g. mouritsen , e. k. hansen , and s. j. k. jensen , phys .
b * 22 * , 3256 ( 1980 ) .
m. e. fisher and d. r. nelson , phys .
* 32 * , 1350 ( 1974 ) . j. m. kosterlitz , d. r. nelson , and m. e. fisher , phys .
b * 13 * , 412 ( 1976 ) .
p. calabrese , a. pelissetto , and e. vicari , phys .
b * 67 * , 054505 ( 2003 ) .
r. folk , yu . holovatch , and g. moser , preprint ( 2008 ) .
m. holtschneider , s. wessel , and w. selke , phys .
b * 75 * , 224417 ( 2007 ) . m. holtschneider and w. selke , phys .
rev . b * 76 * ( r ) , 220405 ( 2007 ) .
m. holtschneider and w. selke , eur .
j. b * 62 * , 147 ( 2008 ) . c. zhou , d. p. landau , and t. c. schulthess , phys .
rev . b * 74 * , 064407 ( 2006 ) .
m. holtschneider , w. selke , and r. leidl , phys .
b * 72 * , 064443 ( 2005 ) . h. matsuda and t. tsuneto , prog .
supplement * 46 * , 411 ( 1970 ) . k .- s .
liu and m. e. fisher , j. low . temp
. phys . * 10 * , 655 ( 1973 ) .
f. keffer , in _ handbuch der physik _ , ed . by s.
flgge ( springer , berlin , 1966 ) , vol .
xviii , pt.2 , p.1 .
a. aharony , phys .
b * 8 * , 4270 ( 1973 ) . f. wegner , solid state commun .
* 12 * , 785 ( 1973 ) .
a. d. bruce and a. aharony , phys .
b * 11 * , 478 ( 1975 ) .
a. aharony , j. stat .
phys . * 110 * , 659 ( 2003 ) .
l. j. de jongh ( ed . ) , _ magnetic properties of layered transition metal compounds _ , ( klwer , dordrecht , 1990 ) . t. thio , c. y. chen , b. s. freer , d. r. gabbe , h. p. jenssen , m. a. kastner , p. j. picone , n. w. preyer , and r. j. birgeneau , phys .
b * 41 * , 231 ( 1990 ) .
t. kroll , r. klingeler , j. geck , b. bchner , w. selke , m. hcker , and a. gukasov , j. magn .
magn . mat .
* 290 * , 306 ( 2005 ) .
h. rohrer and c. gerber , phys .
lett . * 38 * , 909 ( 1977 ) . c. c. becerra , n. f. oliveira , a. paduan - filho , w. figueiredo , and m. v. souza , phys .
b * 38 * , 6887 ( 1988 ) .
g. p. felcher and r. kleb , europhys .
* 36 * , 455 ( 1996 ) .
k. ohgushi and y. ueda , phys .
lett . * 95 * , 217202 ( 2005 ) .
z. w. ouyang , v. k. pecharsky , k. a. gschneidner , d. l. schlagel , and t. a. lograsso , phys .
b * 76 * , 134415 ( 2007 ) .
g. bannasch , diploma thesis , rwth aachen ( 2008 ) .
k. binder , z. physik b * 43 * , 119 ( 1981 ) .
m. n. barber , in _ phase transitions and critical phenomena _ , ed . by c. domb and j. l. lebowitz ( academic press , new york , 1983 ) , vol .
k. vollmayr , j. d. reger , m. scheucher , and k. binder , z. physik b * 91 * , 113 ( 1993 ) .
a. pelissetto and e. vicari , phys
. rep . * 368 * , 549 ( 2002 ) .
m. hasenbusch , k. pinn , and s. vinti , phys .
b * 59 * , 11471 ( 1999 ) .
m. hasenbusch and t. trk , j. phys .
a * 32 * , 6361 ( 1999 ) .
x. s. chen and v. dohm , phys .
e * 70 * , 056136 ( 2004 ) ; v. dohm , phys .
e * 77 * , 061128 ( 2008 ) .
w. selke and l. n. shchur , j. phys . a : math .
gen.*38 * , l739 ( 2005 ) .
a. pelissetto and e. vicari , phys .
76 * , 024436 ( 2007 ) .
j. m. carmona , a. pelissetto , and e. vicari , phys .
b * 61 * , 15136 ( 2000 ) . | classical heisenberg antiferromagnets with uniaxial exchange anisotropy and a cubic anisotropy term in a field on simple cubic lattices are studied with the help of ground state considerations and extensive monte carlo simulations .
especially , we analyze the role of non collinear structures of biconical type occurring in addition to the well known antiferromagnetic and spin flop structures .
pertinent phase diagrams are determined , and compared to previous findings .
heisenberg antiferromagnet , cubic anisotropy , monte carlo simulation , biconical structures 05.10.ln , 75.40.cx , 75.10.hk , 75.50ee |
conventional logic and memory devices are built out of stable deterministic units such as standard mos ( metal oxide semiconductor ) transistors , or nanomagnets with energy barriers in excess of @xmath0 40@xmath160 kt . the objective of this paper is to introduce the concept of what we call `` p - bits '' representing unstable , stochastic units which can be interconnected to create robust correlations that implement precise logical functions .
remarkably this `` probabilistic spin logic '' or psl can also be used to implement the inverse function with multiple solutions which are all visited with equal probability . any random signal generator
whose randomness can be tuned with a third terminal should be a suitable building block for the kind of probabilistic spin logic ( psl ) proposed in this paper .
the icon in fig.1b represents our generic building block whose input @xmath3 controls the output @xmath4 according to the equation ( fig.1a ) @xmath5 where rand(@xmath11,+1 ) represents a random number uniformly distributed between @xmath11 and + 1 , and @xmath6 is assumed large enough that memory of past history has been lost .
if the input is zero , the output takes on a value of @xmath11 or + 1 with equal probability , as shown in the middle panel of fig .
[ fi : fig1 ] . a negative input
@xmath3 makes negative values more likely ( left panel ) while a positive input makes positive values more likely ( right panel ) .
[ fi : fig1]c shows @xmath7 as the input is ramped from negative to positive values . also shown is the time - averaged value of @xmath4 which equals @xmath8
. a possible physical implementation of p - bits could use stochastic nanomagnets with low energy barriers @xmath9 whose retention time @xcite : @xmath10 is very small , of the order of @xmath11 which is a material dependent quantity called the attempt time and is experimentally found to be @xmath12 @xcite among different magnetic materials .
such stochastic nanomagnets can be pinned to a given direction with spin - currents @xmath13 that are a factor of 40 less than those needed to switch 40 kt magnets .
we have performed detailed simulation of such magnets using the stochastic landau - lifshitz - gilbert ( s - llg ) equation for both perpendicular ( pma ) and in - plane ( i m a ) low barrier nanomagnets and the results are in good agreement with those presented in this paper using the simple generic model in eq .
( [ eq : sigmoid ] ) .
these magnet - specific results will be presented separately @xcite , @xcite .
all results in this paper are based on eq .
( [ eq : sigmoid ] ) in order to emphasize the generality of the concept of p - bits which need not necessarily be nanomagnet - based @xcite .
the sigmoidal tuning curve in fig.1c represents the essence of a p - bit , and it seems like a natural feature of nanomagnets driven by spin currents , but it is not too different from those discussed in the context of cmos @xcite . to ensure that individual p - bits can be interconnected to produce robust correlations , it is also important to have separate terminals marked w and r for writing ( more correctly for biasing ) and for reading as shown in fig .
[ fi : fig1]b . with i
m a nanomagnets this could be accomplished following existing experiments @xcite using the giant spin hall effect ( gshe ) .
recent experiments using a built - in exchange bias @xcite could make this approach applicable to pma as well .
note , however , that these experiments have all been performed with stable free layers , and have to be carried out with low barrier magnets in order to establish their suitability for the implementation of p - bits .
as the field progresses , one can expect the bias terminal to involve voltage control instead of current control , just as the output could invove quatities other than magnetization . _ ensemble - average versus time - average : _ a sigmoidal response was presented in @xcite for the ensemble - averaged magnetization of large barrier magnets biased along a neutral state . this
was proposed as a building block for both ising computers and directed belief networks and a recent preprint @xcite describes a similar approach applied to a graph coloring problem .
by contrast low barrier nanomagnets provide a sigmoidal response for the time - averaged magnetization and a suitably engineered network of such nanomagnets cycle through the @xmath14 collective states at ghz rates , with an emphasis on the `` low energy states '' which can encode the solution to the combinatorial optimization problems , like the traveling salesman problem ( tsp ) as shown in @xcite .
once the time - varying magnetization has been converted into a time - varying voltage through a read circuit , a simple rc circuit can be used to extract the answer through a moving time average .
for example , in fig .
[ fi : fig1]c the red trace was obtained from the rapidly varying blue trace using an rc circuit in a spice simulation .
the central feature underlying both implementations is the _ p - bit _ that acts like a tunable random number generator , providing an intrinsic sigmoidal response for the ensemble - averaged or the time - averaged magnetization as a function of the spin current .
it is this response that allows us to _ correlate _ the fluctuations of different p - bits in a useful manner by interconnecting them according to @xmath15 where @xmath16 provides a local bias to magnet @xmath17 and @xmath18 defines the effect of bit j to bit i , and @xmath19 sets a global scale for the strength of the interactions like an inverse `` pseudo - temperature '' .
equations ( [ eq : sigmoid]-[eq : weight ] ) are essentially the same as the defining equations for boltzmann machines introduced by hinton and his collaborators @xcite which have had enormous impact in the field of machine learning , but they are usually implemented in software that is run on standard cmos hardware .
equation ( [ eq : sigmoid ] ) arises naturally from the physics of low barrier nanomagnets as we have discussed above .
equation ( [ eq : weight ] ) represents the `` weight logic '' for which there are many candidates such as memristors @xcite , floating - gate based devices @xcite , domain - wall based devices @xcite , standard cmos @xcite .
the suitability of these options will depend on the range of j values and the sparsity of the j - matrix .
_ outline of paper : _ the objective of this paper is to show that p - bits can provide the building block for a probabilistic spin logic ( psl ) framework that is suitable not just for heuristic non - boolean logic but _ even for precise boolean logic _ by implementing relatively small gates as bidirectional boltzmann machines ( bm @xcite , fig .
[ fi : fig2]a ) which are then interconnected in a directed fashion to implement complex functionalities ( fig .
[ fi : fig2]b ) . in section ii we will briefly describe how a given truth table can be implemented reliably and reconfigurably ( fig .
[ fi : fig3 ] ) using bidirectional networks of p - bits for which we can define an energy functional @xmath20 such that the boltzmann law accurately describes the long - time average of the individual probabilities for the interconnected network to be in different states : @xmath21 fig .
[ fi : fig4 ] shows the close agreement between the probabilities obtained from a direct numerical solution of eq .
( [ eq : sigmoid]@xmath1[eq : weight ] ) over long times and those obtained from eq .
( [ eq : bl ] ) for an and gate , consistent with its designation as a bm .
unlike standard transistor - based implementations , these bms are reversible like the memcomputers proposed in @xcite .
they not only implement boolean functions providing the correct output for a given input , but also the _ inverse function _ providing the correct input(s ) for a given output , thus providing solutions to the boolean satisfiability problem ( fig .
[ fi : fig5 ] ) . in section iii
we show that full adders implemented as bms can be interconnected to implement relatively large logic operations like 32-bit adders .
this is particularly surprising since we do not expect a large collection of stochastic units to get correlated precisely enough to converge on the one correct answer out of @xmath2 possibilities !
note that the 32-bit adder is not fully bidirectional , even though it comprises full adders that are individually bidirectional .
the carry bits connect unidirectionally from the less significant to the more significant bit as dictated by the logic of addition .
interestingly , despite this deviation from complete bidirectionality , the overall adder seems to work in reverse : when we fix the sum bits and let the input bits `` float '' , the input bits fluctuate randomly , but in a correlated manner such that their sum is always precisely equal to the fixed sum as shown in fig .
[ fi : fig7 ] . finally ,
by combining individual full adders ( fa ) and and gates as a directed network of bms , we construct a 4-bit multiplier that performs integer factorization ( fig .
[ fi : fig8 ] ) when operated in reverse similar to memcomputing @xcite based on deterministic memristors . note , however , that the building blocks and operating principles of stochastic p - bits and memcomputing are very different .
) which ensures that the truth table corresponds to the low energy states of the boltzmann machines according to eq .
( [ eq : bl ] ) .
a handle bit of + 1 is introduced to each line of the truth table which can be biased to ensure that the complementary truth table does not appear along with the desired one .
this bit also allows a truth table to be electrically reconfigured into its complement . ]
we now present examples showing how any given truth table can be implemented reliably using boltzmann machines .
our approach , pictorially described in fig .
[ fi : fig3 ] , begins by transforming a given truth table from binary @xmath22 to bipolar @xmath23 variables .
the lines of the truth table are then required to be eigenvectors each with eigenvalue + 1 , all other eigenvectors are assumed to have 0 eigenvalues .
this leads to following prescription for j as shown in fig .
[ fi : fig3 ] : @xmath24 = \sum_{i , j}\mathrm{[s^{-1}]_{ij } u_i u_j^\dagger}\\ \label{eq : amit } \rm s_{ij } = u_i^\dagger u_j \end{aligned}\ ] ] where @xmath25 are the eigenvectors corresponding to lines in the truth table of a boolean operation and s is a projection matrix that accounts for the non - orthogonality of the vectors defined by different lines of the truth table .
note that the resultant j - matrix is always symmetric ( @xmath26 ) with diagonal terms that are subtracted in our models such that @xmath27 .
the number of p - bits in the system is made greater than the number of lines in a truth table through the addition of hidden layers ( fig .
[ fi : fig3 ] ) to ensure that the number of conditions we impose is less than the dimension of the space defined by the number of p - bits .
another important aspect is that an eigenvector @xmath28 implies that its complement @xmath29 is also a valid eigenvector .
however only one of these might belong to a truth table .
we introduce a `` handle '' bit to each @xmath28 that is biased @xmath30 to distinguish complementary eigenvectors .
these handle bits provide the added benefit of reconfigurability .
for example , and and or gates have complementary truth tables , and a given gate can be electrically reconfigured as an and or an or gate using the handle bit . note that this prescription for [ j ] is based on the principles developed originally for hopfield networks ( @xcite , and eq .
( 4.20 ) in @xcite ) .
however , other approaches are possible along the lines described in the context of ising hamiltonians for quantum computers @xcite .
we have tried some of these other designs for [ j ] and many of them lead to results similar to those presented here . for practical implementations , it will be important to evaluate different approaches in terms of their demands on the dynamic range and accuracy of the weight logic .
once a j - matrix and the h - vector are obtained for a given problem , the system is initialized by randomizing all @xmath4 at time , @xmath31 .
first , the current ( voltage ) that a given p - bit ( @xmath4 ) feels due to all the other @xmath32 is obtained from eq .
( [ eq : weight ] ) , and the @xmath4 value is updated according to eq .
( [ eq : sigmoid ] ) . next the procedure
is repeated for the remaining p - bits by finding the current they feel due to all other @xmath4 using the _ updated _ values of @xmath4 .
for this reason , the order of updating was chosen randomly in our models and we found that the order of updating has no effect in our results .
this is similar to the `` asynchronous '' updating of hopfield networks @xcite .
[ fi : fig4 ] shows the time evolution of an and gate designed as outlined in fig .
[ fi : fig3 ] with a total of 8 p - bits comprising 5-auxiliary bits in addition to the 3 input / output bits of an and gate .
initially for @xmath33 the interaction strength @xmath34 , making the pseudo - temperature of the system infinite , and the network produces uncorrelated noise , visiting each state with equal probability . in the second phase ( @xmath35 ) , the interaction strength is suddenly increased to @xmath36 , effectively `` quenching '' the network by reducing the temperature .
this correlates the system such that only the states corresponding to the truth table of the and gate are visited , each with equal probability when a long time average is taken .
the average probabilities in each phase are reproduced quantitatively by the boltzmann law defined by eq .
( [ eq : bl ] ) . in fig .
[ fi : fig5 ] , we show how a correlated network that produces a given truth table can be used to do directed computation analogous to standard cmos logic . by `` clamping '' the input bits of an or gate ( @xmath33 ) through their bias terminals , @xmath16 to ( a , b)=(+1,+1 ) ,
the system is forced to only one of the peaks of the truth table , effectively making c=1 .
the psl gates however exhibit a remarkable difference with standard logic gates , in that inputs and outputs are on an equal footing .
not only do clamped inputs give the corresponding output , _ a clamped output gives the corresponding input(s ) .
_ in the second phase ( @xmath35 ) the output of the or gate is clamped to + 1 , that produces three possible peaks for the input terminals , corresponding to various possible input combinations that are consistent with the clamped output ( a , b)=(0,1),(1,0 ) and ( 1,1 ) .
the probabilistic nature of psl allows it to obtain multiple solutions at the same time ( fig .
[ fi : fig5]c )
. it also seems to make the results more resilient to _ unwanted _ noise due to `` stray fields '' that are inevitable in physical implementations as shown in fig .
[ fi : fig6 ] . ) as indicated in the figure .
the noise is assumed to be uniformly distributed over all p - bits in a given network , and centered around zero with magnitude @xmath37 .
each gate is simulated 50,000 times for t=100 time steps to produce an error probability for a given noise value , and the maximum peak produced by the system is assumed to be an output that can be read with certainty .
the system shows robust behavior even in the presence of large levels of noise . ]
when constructing larger circuits composed of individual boltzmann machines , the reciprocal nature of the boltzmann machine often interferes with the directed nature of computation that is desired .
it seems advisable to use a hybrid approach .
for example in constructing a 32-bit adder we use full - adders ( fa ) that are individually bms with symmetric connections , @xmath38 .
but when connecting the carry bit from one fa to the next , the coupling element @xmath18 is non - zero in only one direction from the less significant to the more significant bit .
this directed coupling between the components distinguishes psl from purely reciprocal boltzmann machines . indeed
even the full adder could be implemented not as a boltzmann machine but as a directed network of more basic gates .
_ adder : _ fig .
[ fi : fig7 ] shows the operation of a 32-bit adder that sums two 32-bit numbers a and b to calculate the 33-bit sum s. in the initial phase ( @xmath39 ) we have @xmath34 corresponding to infinite temperature so that the system fluctuates among @xmath40 8 billion possibilities . at @xmath31
we increase @xmath19 to 70 , and the system unerringly settles down to the one unique solution within 10 to 15 time steps , as we have observed in many different runs , even while adding two binary numbers that propagates a carry all the way from the first bit to the 32@xmath41 bit , as in this example .
interestingly , although the overall system includes several unidirectional connections , it seems to be able to perform the inverse function as well . with
a and b clamped it calculates s = a+b as noted above .
conversely with s clamped , the input bits a and b fluctuate in a correlated manner so as to make their sum sharply peaked around s ! fig .
[ fi : fig7 ] shows the time evolution of the input bits that have broad distributions spanning a wide range .
their sum too shows a broad distribution for @xmath39 when @xmath42 , but once @xmath19 steps up at @xmath31 the distributions of a and b get strongly correlated making the distribution of a+b sharply peaked around the fixed value of s. _ factorizer : _ fig .
[ fi : fig8 ] shows how the reversibility of psl logic blocks can be used to perform integer factorization using a multiplier in reverse similar to memcomputing devices @xcite .
normally , the factorization problem requires specific algorithms and hardware @xcite to be performed in cmos - like hardware , but here we simply use a digital 4-bit multiplier working in _ reverse _ to achieve this operation , which would not be possible in standard cmos implementations . specifically with the output of the multiplier clamped to a given integer from 0 to 15
, the input bits float to the correct factors . as with other examples ,
the interconnection strength @xmath19 is stepped up from zero at @xmath31 .
if the output is set to 9 , both inputs float to 3 as shown in fig .
[ fi : fig8]b . with the output set to 6
, both inputs fluctuate between two values , 2 and 3 .
note that factors like @xmath43 do not show up , since encoding 9 in binary requires 4-bits ( 1001 ) and the input terminals only have 2-bits .
we have checked other cases where factorizing 3 shows both @xmath44 and @xmath45 , and factorizing zero shows all possible peaks since there are many solutions such that @xmath46 and so on .
we also kept the same directed connections between the full adders for the carry bits , making them a directed network of boltzmann machines , similar to the 32-bit adder .
moreover , we kept a directed connection _ from _ the full adders _ to _ the and gates as shown in fig .
[ fi : fig8]a since the information needs to flow from the output to the input in the case of factorization .
the input bits that go to multiple and gates are `` tied '' to each other with a positive exchange ( @xmath47 ) value much like 2-spins interacting ferromagnetically , however in psl we envision these interactions to be controlled purely electrically .
we have presented a framework for using probabilistic units or p - bits as a building block for deterministic computation by establishing robust correlations among them .
this requires the randomness of individual units to be tunable with an input signal , with a clear separation of input and output terminals as shown in fig .
[ fi : fig1 ] .
low barrier nanomagnets driven by spin currents generated through the spin hall effect could be used to provide a physical realization of a p - bit , but this is only one of many possibilities . to emphasize this generality we use a generic hardware - agnostic model defined by eqs .
( [ eq : sigmoid]-[eq : weight ] ) whose predictions are in good agreement with more specific simulations based on the s - llg equation for stochastic nanomagnets @xcite . *
it shows that unstable , stochastic units can be interconnected to create robust correlations that can even implement precise boolean functions .
a 32-bit adder is shown to converge unerringly to the one correct answer by sorting through @xmath48 possibilities at ghz speeds . *
it demonstrates that the proposed probabilistic spin logic ( psl ) can implement not just a function but also its inverse , which in general has multiple solutions .
a 4-bit multiplier working in the inverse mode is shown to function as a factorizer . *
it identifies an electrically tunable random bit generator as the building block which we call a p - bit that can be interconnected in large numbers to implement psl .
low barrier nanomagnets are suggested as a natural realization of p - bits , but not necessarily the only one . *
it defines a new direction for the field of spintronic and nanomagnetic logic by shifting the emphasis from stable high barrier magnets to stochastic low barrier magnets .
* it introduces the intriguing possibility of p - bits emerging as a robust version of qubits that could tackle some of the same challenging problems like factorization , but could be achieved at room temperature with present - day technology .
it is a real pleasure to acknowledge many helpful discussions with behtash behin - aein ( globalfoundries ) , ernesto e. marinero ( purdue university ) and with jaijeet roychowdhury ( uc berkeley ) .
this work was supported in part by c - spin , one of six centers of starnet , a semiconductor research corporation program , sponsored by marco and darpa , in part by the nanoelectronics research initiative through the institute for nanoelectronics discovery and exploration ( index ) center , and in part by the national science foundation through the ncn - needs program , contract 1227020-eec .
28ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo
secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop ( ) @noop ( ) , in @noop _ _ ( , ) pp .
@noop ( ) @noop * * , ( ) in @noop _ _ ( ) pp . @noop * * , ( ) @noop * * , ( ) @noop * * ( ) @noop ( ) @noop ( ) @noop ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop ( ) @noop ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | conventional logic and memory devices are built out of stable deterministic units such as standard mos ( metal oxide semiconductor ) transistors , or nanomagnets with energy barriers in excess of @xmath0 40@xmath160 kt . in this paper
we show that unstable , stochastic units can be interconnected to create robust correlations that implement precise functions .
we use a generic model for such
p - bits " to illustrate some of the functionalities that can be implemented with this `` probabilistic spin logic ( psl ) '' .
remarkably psl can also be used to implement the inverse function with multiple solutions which are all visited with equal probability .
the results from the simple generic model agree well with those obtained for low barrier magnets from the stochastic landau - lifshitz - gilbert ( sllg ) equation , but p - bits need not be magnet - based : any tunable random bit generator should be suitable .
we show that such p - bits can be interconnected symmetrically to construct hardware boltzmann machines ( bm ) that are designed to implement any given truth table .
such bms can be used to implement boolean functions by fixing the inputs to produce a given output .
they can also be run in reverse by fixing the outputs to produce _ all the consistent inputs_. we further show that bm units like individual full adders can be interconnected in a directed manner to implement large logic operations such as 32-bit binary addition , that precisely correlate hundreds of stochastic p - bits to pick out a single state out of @xmath2 possibilities !
interestingly some directed networks of bms can also perform the inverse function .
for example , a 4-bit multiplier acts as a factorizer in reverse : with the output clamped to an integer value , we show that the inputs float to the only factors consistent with that integer . |
distance and velocity measurements of millisecond pulsars can be used to constrain models of supernova dynamics , binary star evolution , pulsar emission physics , and the ionized interstellar medium .
they can characterize the millisecond pulsar population as a whole , and they can be used to probe the kinematic evolution of the millisecond pulsar population in the galaxy and its relation to other stellar populations .
pulsar timing allows for high precision measurement of millisecond pulsar positions , parallaxes , and proper motions .
pulse times of arrival ( toas ) measured over the course of a year vary in part due to the changing time - of - flight of the pulses across the solar system , i.e. , the roemer delay .
this variation is approximately sinusoidal with a period of one year .
the phase and amplitude of this pattern can be used to infer the ecliptic longitude and latitude of the pulsar , respectively . for many millisecond pulsars
, toas can be measured to a precision well under a microsecond , yielding position measurements with precision of order milliarcseconds or better .
measurements of positions over several years can be used to infer proper motions with precision of milliarcseconds per year .
such highly precise toa measurements can also be used to infer pulsar parallaxes , and hence distances , out to distances of order a kiloparsec .
proper motions and distances combine to yield two components of the pulsar velocity vectors .
canonical ( non - millisecond ) pulsars are well known to be high velocity objects .
for example , the proper motion study of @xcite found that canonical pulsars have mean speeds of @xmath8 km s@xmath4when measured in one dimension and @xmath9 km / s when measured in two dimensions ; from this they inferred that pulsars are formed in a thin disk with a mean birth speed of @xmath10 km / s .
these high velocities presumably result from asymmetries in supernova explosions , possibly combined with pre - supernova orbital and space motion of the pulsar progenitor .
in contrast , early studies of millisecond pulsars showed their space velocities to be much lower than those of canonical pulsars @xcite .
this has been confirmed by , among others , @xcite , whose analysis of recycled pulsars ( which they defined as having spin periods @xmath11 ms and spin - down rates @xmath12 ) found their one - dimensional and two - dimensional mean speeds to be @xmath13 km s@xmath4and @xmath14 km s@xmath4 , respectively .
@xcite , whose analysis of millisecond pulsars ( which they defined to have spin periods @xmath15 ms ) found their mean two - dimensional speed to be @xmath16km / s .
the relatively small space velocities of millisecond pulsars are likely a byproduct of their formation process . according to the conventional scenario ,
a millisecond pulsar is formed as a canonical pulsar and then spun - up to its millisecond rotational period via mass accretion from a binary companion , @xcite .
if a neutron star in a binary system is formed with a large kick , the binary is likely to be disrupted , and the binary accretion needed to form a millisecond pulsar can not take place . only binary systems with small or fortuitously oriented kicks survive and allow formation of millisecond pulsars .
alternatively , low - velocity kicks may result from electron - capture supernovae , and the millisecond pulsars may be formed via accretion - induced collapse of a white dwarf during mass transfer in a binary ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . one might expect correlations between millisecond pulsar kinematics and orbital properties , i.e. , different velocity distributions for isolated and binary millisecond pulsars ; and , among binary millisecond pulsars , a dependence of velocity on orbital period or mass function ( e.g. , * ? ? ? * ) .
indeed , @xcite found greater galactic heights for lower - mass binary millisecond pulsar systems .
@xcite compared isolated and binary millisecond pulsars and found that they had indistinguishable velocity distributions , but that observed binary pulsars had a larger galactic scale height , leading to speculation that the two populations have different luminosity distributions , with isolated pulsars weaker and therefore more difficult to detect at large distances ; however , @xcite argue that this could arise from selection effects in pulsar searches . @xcite , analyzing a larger sample of millisecond pulsars , also found the velocity distributions of binary and isolated millisecond pulsars to be indistinguishable .
stellar populations in the solar neighborhood have ellipsoidal velocity distributions .
@xcite show that analysis of these velocity distributions can give insight into the dynamical history of millisecond pulsars , including both the magnitudes of any kicks received at formation and the diffusion of these pulsars through the galaxy .
we adopt galactocentric coordinates , with @xmath17 , @xmath18 , and @xmath19 in the direction of the galactic center , the direction of galactic rotation , and the direction perpendicular to the galactic disk , respectively , , @xmath20 , and @xmath21 in the solar neighborhood .
we use alternate notation to acknowledge that some pulsars in our analysis are far from the solar neighborhood . ] and we denote their velocity dispersions by @xmath22 , @xmath23 , and @xmath24 . it is well established theoretically and observationally that , for any given stellar population , @xmath25 , with typical ratios @xmath26 and @xmath27 ( * ? ? ?
* ; for simplicity , we ignore vertex deviation ) .
analyzing millisecond pulsar velocities in galactocentric coordinates has three advantages .
( i ) it facilitates comparisons with other stellar populations .
( ii ) the consistency of the magnitudes of the different components can be used to cross - check the measured velocity distributions .
( iii ) if a component were biased by selection effects in particular , pulsar searches are preferentially made along the galactic plane , and one could imagine this biasing the distribution of @xmath19the other components might still give unbiased results
. the nanograv collaboration is undertaking long - term timing of several dozen millisecond pulsars for the purpose of detecting and characterizing gravitational waves via their perturbation of millisecond pulsar toas on time scales of tens of years ( frequencies of nanohertz ) .
the detection of gravitational wave signals requires the modeling and removal of all other phenomena that influence pulse arrival times , including solar system time - of - flight effects which yield the astrometric measurements described above . the nanograv nine - year data release @xcite reports timing observations of 37 pulsars collected over nine years , and includes pulse timing models for all of these pulsars . in the present paper , we analyze the astrometric parameters from that work . in [ sec : obs ] , we summarize the observations . in [ sec : measurements ] , we list the parameters resulting from the timing analysis . in [ sec : distances ] , we discuss the distances inferred from the timing measurements . in [ sec : mspvelocities ] , we combine our newly measured millisecond pulsar velocities with previous measurements to analyze the kinematics of the millisecond pulsar population as a whole . in [ sec : conclusion ] , we summarize our results . except where otherwise specified
, we define a millisecond pulsar to be a pulsar with rotational period of 20 ms or less , and a spin - down rate of less than @xmath28 , a definition that includes all pulsars observed in the nanograv program .
we exclude pulsars in globular clusters from our study , because no nanograv sources are in globular clusters , and because the formation and dynamics of millisecond pulsars in globular clusters are different than those of millisecond pulsars in the galactic disk .
we use data from the nanograv nine - year data set , which we briefly summarize here . for full details see @xcite . observations of 37 millisecond pulsars were made using the arecibo observatory ( ao ) and the green bank telescope ( gbt ) .
the project began with 15 pulsars in 2004 , and new pulsars were added to the project as they were discovered or when wide - band data acquisition systems enabled sources previously deemed unreliable to be precisely timed .
pulsars were chosen based on high timing precision , detectability over a wide frequency range , and expected timing stability .
the nine - year data set includes observations through 2013 . observed time spans of individual pulsars range from 0.6 yr to 9.2 yr .
each pulsar was observed at approximately monthly intervals . at every epoch ,
each pulsar was observed for approximately @xmath29 minutes each using two different radio telescope receivers at widely separated frequencies .
such dual - receiver observations allow for measurement and removal of interstellar and solar - system dispersive effects .
( in a small number of cases , dual - receiver observations were not available , but single - receiver observations were made over a wide observing band . )
observations in early years of the project used the astronomical signal processor ( asp ) at ao and the green bank astronomical signal processor ( gasp ) at the gbt ; each sampled up to 64 mhz bandwidth , depending on telescope receiver capability . in the later years
, observations used the puerto rican ultimate pulsar processor ( puppi ) at ao and the green bank ultimate pulsar processor ( guppi ) at the gbt ; each provided up to 800 mhz bandwidth , again depending on telescope receiver capability .
the data were polarization - calibrated , cleaned of radio frequency interference , and analyzed to produce toas using a standardized pipeline ] that made extensive use of the psrchive software package @xcite .
the sets of toas were fit to timing models using the tempo and tempo2 packages .
we found that these packages yielded essentially identical results , and we used tempo for the work in the present paper .
the timing models contain standard spin - down , astrometric , and ( where appropriate ) binary models .
the astrometric parameter measurements are the basis for the present paper and are described in more detail in
[ sec : measurements ] .
criteria for inclusion of specific parameters in the timing model for each pulsar are given in @xcite .
in the timing analysis , earth motion around the solar - system barycenter was modeled using the jpl de421 planetary ephemeris @xcite .
toas were measured using hydrogen - maser clocks at the observatories and were transformed to universal time and then to barycentric dynamical time using standard techniques .
separate values of dispersion measure were fit independently at every observing epoch , with a few exceptions ( see @xcite for details ) .
this results in timing parameter uncertainties that are larger than would be found using smoothed dispersion measure models , but is necessary to avoid contamination of our results by variations in interstellar and solar wind dispersion .
the latter , in particular , is not easily modeled , and is highly covariant with the astrometric parameters of interest to this paper .
a novel noise model was used to account for any aspects of the timing data that did not fit the standard timing model within the expected measurement uncertainties .
the noise model incorporated white noise , correlated noise in simultaneously collected toas , and red noise ( as needed for a few sources ) .
details are given in @xcite .
the full data set , including toas and timing parameters of all pulsars , is available online .
we made a small number of modifications to the nanograv nine - year data set for the present work .
the data sets of psrs j1741 + 1351 , j1853 + 1303 , j1910 + 1256 ,
j1944 + 0907 , and b1953 + 29 contain lengthy spans in which observations were made using only a single , narrow - band receiver , eliminating the possibility of monitoring and correcting for dispersion measure variation over those spans .
we excised the toa measurements from those data spans .
we re - evaluated the timing models for these pulsars using the remaining toas , following the same guidelines used by @xcite for noise models , parameter inclusion , etc .
the timing solution for psr j2317 + 1439 in the nanograv nine - year data release included secular variations of the laplace - lagrange orbital elements , which are formally significant in the fit , but which are physically implausible .
we removed those parameters from the timing solution .
we then re - evaluated the timing model using the guidelines of @xcite , following which we added shapiro delay parameters ( now significant ) and eliminated the secular variation in the projected orbital semi - major axis ( now not significant ) .
details of the analysis of this and other binaries will be given in a forthcoming paper ( fonseca et al .
, in preparation ) .
the modified timing solutions used for this paper are bundled with the nanograv nine - year data release files .
the timing analysis of the nanograv nine - year data set parameterizes pulsar positions in ecliptic coordinates .
these are natural coordinates for pulsar timing astrometry , as ecliptic longitude and latitude are nearly orthogonal parameters in the timing fit .
table [ tab : equatorial ] lists the ecliptic and equatorial positions of all of the pulsars in the data set .
the epoch of each position was chosen near the middle of its data set in order to minimize covariance between position and proper motion parameters .
the reference frame for these position measurements is the reference frame of jpl planetary ephemeris de421 , which is oriented to the international celestial reference frame @xcite .
the ephemeris coordinates were transformed into ecliptic coordinates by a rotation of @xmath30 , the iers2010 obliquity of the ecliptic . for observational convenience , table [ tab : equatorial ]
also lists the positions in equatorial coordinates .
the equatorial positions were determined by fitting timing solutions without rotating the planetary ephemeris into ecliptic coordinates .
as shown in figure [ fig : position_covariance ] , uncertainties in equatorial coordinates are generally larger than uncertainties in ecliptic coordinates .
this is particularly true when one ecliptic coordinate is measured much more precisely than the other ( e.g. , for pulsars along the ecliptic ) . proper motion , @xmath31 and @xmath32 ,
was included in the timing model for 35 of the 37 pulsars in our data set .
the two pulsars for which proper motion was excluded have less than one year of timing measurements in our data set , making it impossible to measure their proper motions .
lengths of all of the data sets are given in table [ tab : propmo ] .
the proper motions from timing analyses in both ecliptic and equatorial coordinates are listed in table [ tab : propmo ] . as with the position measurements ,
ecliptic coordinates provide the best separation of proper motion into orthogonal components , and hence the smallest uncertainties .
the table also includes proper motions in galactic coordinates .
the best previous proper motion measurements in ecliptic coordinates are listed in table [ tab : propmo ] .
all the new and previous measurements agree to within @xmath33 except for two pulsars ( psrs j1909@xmath03744 and j2145@xmath00750 ) which have discrepancies between @xmath33 and @xmath34 for at least one component of proper motion .
the proper motions of these two pulsars are known to high precision , and these small discrepancies have little practical impact . we speculate that the differences may be due to differences in dispersion measure variation models ( as we discuss for parallax measurements in [ sec : pxcomparison ] ) , or differences in solar system ephemerides used in the timing analysis .
our measured proper motions are illustrated in figure [ fig : galactic_velocity ] .
we discuss the derivation of two- and three - dimensional space velocities from these proper motions in
[ sec : mspvelocities ] .
we included parallax , @xmath35 , as a free parameter in the timing model for each pulsar , whether or not the measurement was statistically significant .
the timing analysis allowed both negative and positive parallaxes .
although negative parallaxes are non - physical , allowing them in the timing solutions provides a useful a check on the reliability of the measurements ( [ sec : pxlowsignificance ] ) .
the measured parallax values are listed in table [ tab : px ] and shown in figure [ fig : px ] .
previous measurements for these pulsars are also included in the table and figure .
there were no previously reported parallax measurements or limits for 20 of the 37 pulsars .
of these 20 sources , two provided significant new parallax measurements : psr j1918@xmath00642 , @xmath36mas ; and psr j2043 + 1711 , @xmath37mas . of the 17 pulsars with previous measurements or limits , we improved the precision of four measurements by a factor of two or better : psrs j0030 + 0451 , j0613@xmath00200 , j1600@xmath03053 , and j1614@xmath02230 . for psr j1713
+ 0747 , in addition to the previous timing parallax value of 0.94(5 ) given in table [ tab : px ] , there are two other previous measurements of interest .
one is from an analysis of 21 years of data from this pulsar by @xcite , which obtained a parallax of @xmath38 mas , identical to our own measurement .
this is not surprising , as the vast majority of the toas in the @xcite analysis were from the same nanograv nine - year data set that is being used for the present work .
the other previous parallax measurement of j 1713 + 0747 is from the interferometric vlba measurements of @xcite , which obtained @xmath39 mas , in reasonable agreement ( @xmath40 ) with our value . in most other cases
, our parallax values also agree with previous measurements . however , there are discrepancies of 2@xmath41 or more ( i.e. , disagreement at 95% confidence ) between our measurements and the previous measurements for three sources : psrs j1643@xmath01224 , b1855 + 09 , and j1909@xmath03744 .
a possible explanation for these discrepancies is the difference in treatment of dispersion measure ( dm ) variations .
small variations in the ionized component of the interstellar medium and the solar wind along the pulsar
earth line - of - sight significantly affect toas . to minimize contamination of the timing analysis from dm variations on the half - yearly time scale of the parallax signal ( as well as other time scales of interest to pulsar modeling ) , we fit for dm at every epoch of observation .
since the dm variations are covariant with the parallax signal in the timing fit , this increases formal uncertainties on our parallax measurements , but it also makes the measurements more robust . the previous parallax analysis of two of the discrepant sources psrs
j1643@xmath01224 , and b1855 + 09assumed a constant value of dm over the entire data set ( see references in table [ tab : px ] ) .
the third , psr j1909@xmath03744 @xcite , used the smoothed dm time series model of @xcite , which includes a smoothing time scale longer than the half - year period of the parallax signature in the timing model . to test the effect of using constant or smoothed dm variation models , we ran trials of timing analyses of these sources fitting ( i ) a constant dm and ( ii ) a model of dm as a up - to - fifth - order polynomial in time over the entire data sets .
these tests yielded changes in measured parallax values by amounts ranging from @xmath42 to @xmath43 , with three of the four values moving by at least @xmath44 .
we therefore believe that limited dm variation modeling is a plausible explanation for the discrepancies between previously measured parallax values and our measurements .
the timing analyses of a majority of the pulsars we analyzed did not yield significant parallax detections . to illustrate the nature of these non - detections , we sorted the pulsars by parallax measurement significance , @xmath45 , and plotted the cumulative distribution of these values in figure [ fig : px_cdf ] .
of the 37 parallax measurements , 12 are significant positive detections of parallax , @xmath46 .
no sources had significant negative detections of parallax , @xmath47 .
( although negative parallax values are unphysical , corresponding to pulse wavefronts with concave rather than convex curvature relative to the pulsar , they were allowed in the timing analysis .
a significant negative parallax measurement would indicate a problem with the data or the timing model . ) of the 25 pulsars with non - detections , only 7 measurements are negative , all but two within @xmath44 of zero , and the remaining 18 are positive .
this bias toward positive values indicates that most of these are indeed real , physical measurements , and significant parallax values should be attainable for many of them with only moderately larger data sets .
we now use the parallaxes in table [ tab : px ] to analyze distances to 30 of the 37 pulsars .
the timing data sets of the other 7 pulsars span less than 2 years , which is typically not sufficient time to disentangle position , proper motion , and parallax in the timing analysis .
indeed , none of these sources had nominally significant parallax measurements , and 3 of them had negative parallaxes . for the 12 pulsars with significant parallax measurements , @xmath48 , we calculated distances as listed in the upper portion of table [ tab : pxlimits ] . for each of these pulsars ,
the central distance value , upper limit , and lower limit given in the table were calculated using @xmath49 , where @xmath35 was the 84% , 50% , and 16% points in the measured parallax distribution , respectively , corresponding to the 16% , 50% and 84% points in the distance distribution .
for the 18 remaining pulsars , we used 95% confidence upper limits on parallax to compute 95% confidence lower limits on distance . to find the upper limits on parallax
, we first verified that changes in assumed parallax values in the timing solution yielded changes in goodness - of - fit @xmath50 values appropriate for parallax values drawn from a gaussian distribution .
we then found the 95% confidence upper limits on @xmath35 by calculating the value of @xmath35 at which 95% of the area under the gaussian distribution of the parallax was at lower values , but restricting the integral to positive values of parallax .
the resulting parallax limits are listed in table [ tab : pxlimits ] . the corresponding @xmath51 confidence lower limit on distance for each source
was then calculated to be the reciprocal of this value .
these are also listed in table [ tab : pxlimits ] .
the lutz - kelker bias @xcite may be significant for the distance measurements of pulsars with large uncertainties on parallax .
the bias can be incorporated as a correction to parallax and distance values through methods outlined by @xcite and @xcite .
the bayesian analysis used under these methods assumes a uniform prior distribution of a pulsar s position throughout three - dimensional space .
since there is more three - dimensional space at larger distances than at smaller distances , this has the effect of moving the posterior distribution of a pulsar s distance to values larger than that calculated from @xmath49 .
this is problematic for pulsars for which there are only upper limits on parallax , as this allows a finite probability of @xmath52 , which corresponds to infinitely large distances , at which there is infinite parameter space . to achieve a tractable result ,
some additional prior must be invoked to cut off the likelihood of the pulsar being at a large distance .
one possibility is to use a model of the galactic distribution of millisecond pulsars , but we want our results to stand independent of _ a priori _ beliefs about the galactic distribution of millisecond pulsars ; further , we found that attempts to use prior galactic distribution models tended to give results that were dominated by the details of the model .
another possibility is to model the sensitivity of pulsar search programs , which find nearby pulsars more readily than distant pulsars ; however , we found that use of such a prior gave results that were highly dependent on the choice of millisecond pulsar luminosity distribution , which is not well known . for these reasons , we elected not to apply any corrections to the distances in table [ tab : pxlimits ] . for sources for which we give lower limits on distances ( lower portion of table [ tab : pxlimits ] ) , this is a conservative approach from a lutz - kelker perspective , in the sense that the lutz - kelker bias would place these pulsars still further than our lower - limit distances .
for the twelve sources for which we give distance measurements ( upper portion of table [ tab : pxlimits ] ) , we tested the effect of lutz - kelker corrections by running the measurements through the lutz - kelker bias code from @xcite .
we found the differences in calculated distances to be small ; for example , six of the central distance measurement values were unchanged , and the other six all changed by less than @xmath44 ( typically much less ) .
in addition to timing parallax , there are other means by which distances can be inferred from pulsar timing .
one such approach uses the observed properties of binary pulsars .
the observed orbital period , @xmath53 , of a binary pulsar undergoes a time change @xmath54 due to mass loss from the system ( negligible here ) , general relativistic phenomena , and kinematic effects @xcite , @xmath55 the relativistic term , @xmath56 , is the orbital decay due to emission of gravitational radiation , which depends on the component masses and other orbital elements .
the kinematic term , @xmath57 , is due to the relative acceleration of the binary system and the sun , which causes changes in the doppler shift of the observed period .
the relevant expression from @xcite , here modified to apply to pulsars off the galactic plane @xcite , is @xmath58 where the terms on the right hand side are as follows .
the first term is the acceleration of the binary system in the galactic potential perpendicular to the disk , @xmath59 , which depends on the perpendicular distance @xmath60 , where @xmath61 is the galactic latitude , and @xmath62 is the speed of light .
the second term is the line - of - sight acceleration due to differential rotation in the galaxy ; here @xmath63 is the galactocentric distance of the solar system ; @xmath64 is the circular rotation speed of the galaxy at @xmath63 ; @xmath65 is the galactic longitude ; and @xmath66 .
the third term is the apparent acceleration due to proper motion @xcite .
each of these terms depends on the distance to the pulsar ; as @xcite pointed out , given a measurement @xmath67 and a calculated @xmath68 , one can solve equations [ eqn : pbdot1 ] and [ eqn : pbdot2 ] to infer the distance to the pulsar .
this method has been used to compute high - precision distances to psrs j0437 - 4715 @xcite and b1534 + 12 @xcite . in cases where the measurement of @xmath54 is marginal or not significant
, this method can be used to place an upper limit on the pulsar distance .
there are 25 binary pulsars in the nanograv nine - year data set .
a full analysis of these systems will be presented in fonseca et al .
( in preparation ) . for the present work
, we considered 22 of these pulsars which have been observed for at least two years .
for each of these ,
we ( i ) fit for @xmath69 in the timing solution of the pulsar , ( ii ) calculated the expected @xmath69 as a function of distance according to equations [ eqn : pbdot1 ] and [ eqn : pbdot2 ] , ( iii ) compared the calculated @xmath69 values with the measured value and its uncertainty to find a probability associated with that distance , and ( iv ) used the probabilities to calculate either the distance and its uncertainty or a 95% upper limit on the distance . for the @xmath69 calculations ,
the gr term is negligible for wide binary systems ; for tight binaries , we estimated @xmath70 using masses from analysis of the shapiro delay in each system ( fonseca et al .
, in prep ) , except for the case of psr j1012 + 5307 , where we used masses derived from the optical measurements of @xcite . for the acceleration toward the galactic disk , @xmath71
, we used the model of the galactic potential given in equation 17 of @xcite , based on the galactic potential of @xcite ; other galactic potential models , such as that of @xcite , give essentially identical results .
we used @xmath72 kpc and @xmath64=240 km s@xmath4 @xcite . for the proper motion term in the @xmath69 calculations
, we used the proper motions given in table [ tab : propmo ] .
we found that 15 of the 22 pulsars had distance upper limits greater than 15 kpc .
since such limits are of little interest , and since this is beyond any reasonable extrapolation of the galactic rotation and acceleration models , we did not analyze these further .
of the seven remaining pulsars , we measured significant distance constraints ( at least @xmath73 ) for two sources , and we placed distance upper limits on the other five sources ( @xmath74 in table [ tab : pbdotdistances ] . ) in all cases , these measurements are consistent with the distance measurements or limits we found via parallax measurement ( table [ tab : pxlimits ] ) .
the most precise @xmath69 distance measurement is that of psr j1909@xmath03744 , @xmath75 kpc .
this is a good match for our parallax distance of @xmath76 kpc . the best previous parallax measurement , @xmath77 mas ( @xmath73 , * ? ?
* ) gave a somewhat larger distance of @xmath78 kpc ( @xmath44 , * ? ? ?
as discussed above , we suggest the difference between previous and presently measured parallaxes may result from differences in the dispersion measure model used in the timing analysis .
the distance derived from @xmath69 is relatively impervious to changes in the dispersion measure model , so it provides a good verification of our parallax distance measurement .
finally , we note that @xmath69 values of two sources , psrs j1012 + 5307 and j1713 + 0747 , have been measured more precisely than in the present paper @xcite . those works focused on the use of @xmath69 measurements to test relativistic gravity rather than measuring distances .
using those @xmath69 measurements in our distance algorithms above gives distance upper limits of @xmath79 kpc and @xmath80 kpc , for psrs j1012 + 5307 and j1713 + 0747 , respectively .
these constraints , while weak , are consistent with our parallax distance measurements and limits .
another set of constraints on pulsar distances arises from their spin - down rates @xcite .
a variant on equation [ eqn : pbdot1 ] can be applied to a pulsar rotation period , @xmath81 , and spin - down rate , @xmath82 . the observed spin - down rate , @xmath83 depends on both the intrinsic spin - down rate , @xmath84 , and kinematic corrections : @xmath85 the kinematic term , @xmath86 , obeys the same expression as equation [ eqn : pbdot2 ] , substituting rotational for orbital parameters , @xmath87 if we presume that millisecond pulsars are powered by rotational energy loss , then @xmath88 . from this
, equation [ eqn : pdot1 ] implies @xmath89 , so the observed spin parameters place an upper limit on this kinematic term .
the right hand side of equation [ eqn : pdot2 ] is generally a monotonically increasing function of distance , so the upper limit on @xmath86 sets an upper limit on distance .
we used equations [ eqn : pdot1 ] and [ eqn : pdot2 ] to place upper limits on distances to the 30 pulsars in the nanograv nine - year data set that have been observed for at least 2 years , so that they have well - measured proper motions .
we used the galactic rotation and acceleration parameters described in [ sec : pbdotdistances ] .
we found that 12 of these pulsars had distance limits greater than 15 kpc ; we did not pursue these further .
the spin - down distance limits , @xmath90 , of the remaining 17 pulsars are listed in table [ tab : pdotdistances].3744 , in addition to the distance range allowed by the entry in the table [ tab : pdotdistances ] , 0 to 1.375 kpc , there are additional allowed solutions at @xmath91 kpc , where the flat galactic rotation curve model implies very large accelerations . since this pulsar
is well established to be much closer than 9 kpc , we do not include this solution in the table . ]
most of the spin - down distance constraints agree with the distances derived from parallax measurements .
indeed , all of the parallax distances measured with at least 3@xmath41 precision ( table [ tab : pxlimits ] ) fall within the spin - down distance limit constraints ( table [ tab : pdotdistances ] ) .
the pulsar for which the spin - down distance limit comes closest to the parallax distance is psr j1909@xmath92 , for which the spin - down limit is @xmath93 kpc and the parallax distance is @xmath94 kpc , providing a good check on the parallax distance . for the pulsars for which we have only an upper limit on parallax , and therefore a lower limit on distance ( table [ tab : pxlimits ] ) , all but one are in agreement with the spin - down distance limits .
in these cases , the pair of distance constraints bracket the actual distance to the pulsar . for psr j1024@xmath00719 ,
the spin - down distance upper limit , @xmath95 kpc , disagrees with the parallax distance lower limit , @xmath96 kpc .
this is perplexing .
the spin - down distance limit should be robust , presuming that the pulsar is losing energy and not subject to external acceleration .
we have run several tests on our parallax measurement , and we consistently find low parallax upper limits ( i.e. , large distance lower limits ) .
we discuss this source further in appendix [ app : j1024 ] .
in figures [ fig : sigpxvsdm ] and [ fig : pxlimitvsdm ] , we compare our measured distances and limits with the predictions of the ne2001 galactic electron density model @xcite given the measured dms of these pulsars .
note that previous parallax measurements or limits for psrs j1713 + 0747 , j1744 - 1134 ,
b1855 + 09 , and b1937 + 21 were used as input data to the ne2001 model , generally with much larger uncertainties than in the present work ( j. cordes , private communication ) .
we find that , in large part , distances predicted based upon dispersion measure and the ne2001 model were in agreement with parallax derived distances and limits . among the pulsars for which we have distance measurements ( not just limits ) ,
there were only two pulsars for which there is no possible distance that is both within the @xmath73 parallax distance measurement range and within @xmath97 of the ne2001 distance prediction , psrs j1614@xmath02230 and j1909@xmath03744 ( figure [ fig : sigpxvsdm ] ) .
in the case of parallax distance lower limits , the ne2001 electron density model fared equally as well .
apart from a few outlying pulsars , the limits were not too far from the distance derived via dispersion measure and the ne2001 model .
we calculated the dispersion measures for which the ne2001 model yields the distance measurements we derived from parallax . for psr j1614@xmath02230 , the dispersion measure required for agreement would shift from @xmath98 to @xmath99 , and for psr j1909@xmath03744 it would shift from @xmath100 to @xmath101 .
recent work on the galactic electron distribution can be found in , e.g. , @xcite , @xcite , and references therein .
new and refined pulsar distance measurements can contribute to this effort .
for example , @xcite reported that ne2001 underestimated distances to some pulsars at high galactic latitudes , implying the need for a larger disk scale height or disk electron density ( see also * ? ? ? * ) . in figure
[ fig : pxdm_gb ] , we plot @xmath102 vs. galactic latitude .
no trends are apparent in our measurements , though we have fewer high galactic latitude sources than @xcite and are limited by small number statistics . for those pulsars with dispersion distances that did not agree within two standard deviations of the parallax - derived measurements
, we find no correlation between their sky positions .
the sources that disagree are at varying galactic longitudes , although all are greater than @xmath103 from the galactic plane .
to analyze the kinematics of the millisecond pulsars , we used the measured proper motions and distances to calculate velocity vectors in galactocentric coordinates at the pulsars standards of rest . because there are not line - of - sight velocity measurements for most pulsars , complete three - dimensional velocity vectors are not obtainable , and our analysis for such pulsars uses only velocity components that are close to orthogonal to the line - of - sight
; we describe this further below . for this analysis , we used both measurements from the present work and previously reported measurements found in the literature .
we included all millisecond pulsars for which proper motions were available with 5@xmath41 significance in two orthogonal components .
when available , we used our proper motion measurements ( table [ tab : propmo ] ) ; for psrs j1012 + 5307 , j1738 + 0333 , and j1949 + 3106 , although they are included among our measurements , higher - precision measurements are available in the literature . for these and other pulsars
, we used the values listed in table [ tab : propmo_nonnanograv ] . for distances , we used measured parallax distances with 3@xmath41 or greater significance from our measurements ( table [ tab : pxlimits ] ) or from the literature ( table [ tab : px_nonnanograv ] ) . for pulsars without parallax distances , we estimated distances from dispersion measures and the ne2001 electron density model . for each pulsar , we calculated a two - dimensional velocity in the reference frame of the sun and transformed it to the galactocentric coordinates at the location of the pulsar . for this transformation , we used solar motion @xmath104 , @xmath105 , and @xmath106 , @xcite , and we used solar galactocentric distance @xmath107 and galactic rotation velocity @xmath108 @xcite .
the line - of - sight ( los ) velocity has been measured via optical observations for three pulsars of interest ( table [ tab : los_velocity ] ) , allowing three - dimensional galactocentric velocities to be calculated for these pulsars . for the remaining pulsars , we calculated and removed the line - of - sight velocity expected for the pulsar if at rest in its standard of rest .
this los correction is only a reasonable approximation for components of the galactocentric velocity that are nearly perpendicular to the los . in the calculations below
, we only include measurements of galactocentric velocity components that are at least @xmath109 from the direction of the los to the pulsar .
using the algorithm and criteria describe above , we calculated galactic radial velocities of 18 pulsars ( figure [ fig : rv_los ] ) , azimuthal velocities of 12 pulsars ( figure [ fig : azimuthv_los ] ) , and perpendicular velocities of 28 pulsars ( figure [ fig : zv_los ] ) .
the component values included in each velocity dispersion calculation are listed in table [ tab : velocity ] .
using the values in table [ tab : velocity ] , we find the dispersions of these velocity measurements to be : @xmath110 visual examination suggests that there are four outliers in the velocity component distributions : psrs b1957 + 20 ( @xmath17 , figure [ fig : rv_los ] ) ; j1909@xmath03744 ( @xmath111 , figure [ fig : azimuthv_los ] ) ; and j1944 + 0907 and j0610 - 2100 ( @xmath19 , figure [ fig : zv_los ] ) .
we used the median absolute deviation as a robust method , under the assumption of a gaussian velocity distribution , to test whether these points are outliers @xcite . for each pulsar
, we calculated @xmath112 , where @xmath113 is the velocity component of the pulsar of interest , @xmath114 is the median of all such measurements , and mad is the median absolute deviation .
we obtained values of @xmath115 , @xmath116 , @xmath117 , and @xmath118 for psrs b1957 + 20 , j1909@xmath03744 , j1944 + 0907 , and j0610 - 2100 , respectively .
based on the criterion @xmath119 , we argue that these are indeed outliers .
we discuss possible physical mechanisms for such outliers in section [ sec : conclusion ] .
if we exclude these three outlier pulsars from all velocity dispersion calculations , the dispersions of the velocity measurements are : @xmath120 using these dispersions , psr b1957 + 20 is @xmath121 sigma from the mean , psr j1909@xmath03744 is @xmath116 sigma from the mean , psr j1944 + 0907 is @xmath122 sigma from the mean , and psr j0610 - 2100 is @xmath123 sigma from the mean , reinforcing their status as outliers . a possible concern with the velocity dispersion measurements is selection effects due to non - uniform sky coverage of pulsar surveys .
search programs such as the palfa @xcite and pmb @xcite surveys have concentrated on low galactic latitudes . as an old population ,
millisecond pulsars found close to @xmath124 are more likely to have smaller velocities in the direction away from the plane .
thus , surveys focusing on the galactic plane might be susceptible to biased velocity distributions . however , there is no similar reason to expect azimuthal or radial velocity distributions to be susceptible to such selection effects , so the latter components are more robust measurements of the true underlying kinematics of the millisecond pulsar population .
velocity dispersions of other stellar populations in the galaxy are observed to have @xmath125 , with typical ratios
@xmath26 and @xmath27 @xcite .
the velocity dispersions of equation [ eqn : disp_exclude ] ( outliers excluded ) follow this pattern , suggesting that these calculated dispersions are relatively free of bias due to pulsar search selection effects and hence represent the intrinsic velocity dispersions of this population .
the velocity dispersions of equation [ eqn : disp_all ] ( all measurements ) do nt strictly follow these patterns : they have @xmath126 , and @xmath127 is a bit higher than for typical stellar populations .
however , this could easily be a case of small - number statistics : removing a single high - velocity source from the @xmath24 calculation would bring these numbers into alignment with other stellar populations .
previous works have studied the velocities distributions for isolated and binary millisecond pulsars and found no significant difference @xcite . to test whether there exists a difference in velocities between isolated and binary millisecond pulsars with our new proper motion values , we calculated the radial , azimuthal and perpendicular velocity dispersions for each subpopulation , recognizing that small number statistics allow only a rough comparison .
we obtained velocity dispersions of 55 and 43 km s@xmath4 for the radial velocity dispersions of isolated and binary millisecond pulsars , respectively .
we obtained dispersions of 33 and 42 km s@xmath4 for the isolated and binary azimuthal velocity dispersions and 18 and 25 km s@xmath4 for the isolated and binary perpendicular velocity dispersions . in all cases
the similarities between the isolated and binary dispersion values imply there is little to no difference between the two subpopulations .
these calculations excluded the outliers described above . including the outliers for each velocity component significantly changes the dispersions .
since the binary and isolated dispersions are calculated separately , including an outlier only affects either isolated or binary pulsar dispersion for a given component . for the radial component ,
the binary dispersion becomes 87 km s@xmath4 .
for the azimuthal component , the binary dispersion becomes 63 km s@xmath4 .
for the perpendicular component , the isolated dispersion becomes 72 km s@xmath4and the binary dispersion becomes 55 km s@xmath4 .
the consistency between the binary and isolated dispersions when the outliers are excluded , and the inconsistency ( with no underlying pattern ) when the outliers are included lends itself well to arguing that these pulsars are indeed outliers .
millisecond pulsars are an old population .
characteristic spin - down ages , @xmath128 , of the pulsars in the nanograv nine - year data set calculated using the observed spin parameters range from 0.2 to 29 gyr , with a median of 5 gyr . correcting for kinematic effects to estimate intrinsic spin - down ages ( [ sec : pdotdistances ] )
gives larger ages : 0.2 to 46 gyr with a median of 8 gyr for the 30 of our pulsars with more than two years of data ( and hence well - measured proper motions , as needed for the kinematic correction ) .
obviously , pulsars can not be older than a hubble time ; some , perhaps all of these ages must be overestimates , which is easily explained if millisecond pulsar spin periods at formation are not much shorter than their present - day spin periods ( e.g. , * ? ? ?
* ; * ? ? ?
nevertheless , it seems reasonable to assume that millisecond pulsars are typically at least several gyr old .
we will use ages of @xmath129 and 10 gyr in the comparisons below .
velocity dispersions of stellar populations are well known to correlate with age ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , with younger stars described as a thin disk and older stars described as a thick disk , although the mechanism behind the correlation is not well established ( e.g. , * ? ? ?
@xcite pointed out that the diffusion of millisecond pulsars into a thick disk contributes a significant portion of their observed velocities . here
we consider our millisecond pulsar velocity dispersion measurements in the context of dispersion - age relations developed from studies of optical stellar populations .
a caveat to this comparison is that optical stellar population modeling tends to use stars in the solar neighborhood , whereas our millisecond pulsar population is spread over a larger volume .
nevertheless , we have already seen that the ratios @xmath130 for millisecond pulsars follow those of other stellar populations , so it seems plausible that the magnitudes of the dispersions may be comparable as well .
@xcite use local stellar data to fit equations of the form @xmath131^{\beta}$ ] for each component of velocity dispersion . for radial , azimuthal , and perpendicular components
, they found @xmath132 , 0.430 , and 0.445 ; @xmath133 , 0.715 , and 0.001 ; and @xmath134 , 28.823 , and 23.831 km s@xmath4 , respectively .
for an age of @xmath135 , this gives @xmath136 km s@xmath4 , @xmath137 km s@xmath4 , and @xmath138 km s@xmath4 .
for an age of @xmath139 , it gives @xmath140 km s@xmath4 , @xmath141 km s@xmath4 , and @xmath142 km s@xmath4 .
@xcite , fit an equation to binned @xmath143-velocity vs. age data and find an empirical relation @xmath144 km s@xmath4 . for ages of @xmath129 and 10 gyr , this gives @xmath145 km s@xmath4and 30 km s@xmath4respectively . for typical millisecond pulsar ages ,
then , the velocity dispersions from the model of @xcite are only modestly different than our velocity dispersions in the measurements that exclude outliers ( equation [ eqn : disp_exclude ] ) , and the velocity dispersions of @xcite are a nearly perfect match . we conclude that , if this characterization of millisecond pulsar kinematics as a gaussian velocity distribution with a small number of outliers is correct , then the bulk of these objects need essentially no velocity boost to reach their observed velocities , since they are comparable to other stars of similar ages .
it is generally accepted that most neutron stars receive a kick at birth ( e.g. , * ? ? ?
when the neutron star has a stellar - mass companion , the resulting space motion of the center of mass will generally be smaller than in the case of an isolated neutron star due to the mass of the binary .
however , the observed outliers in our sample could be the result of fortuitously directed kicks that produce significant center - of - mass velocities .
furthermore , it has been proposed that o - ne - mg - core stars undergo electron capture supernovae with small velocity kicks , whereas iron - core stars undergo supernovae with large velocity kicks @xcite .
this dichotomy may also contribute to our observed velocity distribution .
on the other hand , if the outlier model is incorrect , and the velocity dispersions in equation [ eqn : disp_all ] are a more appropriate measure of the millisecond pulsar population , then their velocities are moderately larger than other stars of similar ages , but a significant portion of their velocities must still be attributable to the same mechanism that increases other stellar velocities over time ( whatever that mechanism is ) . in either model ,
unlike canonical pulsars , millisecond pulsar velocities are very low , and require small velocity boosts at most during their formation .
we have measured and refined distances to twelve millisecond pulsars and found distance limits on eighteen more ( table [ tab : pxlimits ] ) .
these distances will find uses in a variety of applications , from the physics of pulsars themselves ( e.g. , the distance is needed if using a measured flux density to calculate its luminosity ) , to the analysis of the ionized interstellar medium . in [ sec : ne2001 ] we focused on the latter application and found the distances predicted by the ne2001 electron density model in general agreement with those calculated from parallax . for 7 of the 37 nanograv millisecond pulsars , we also calculated distance or a @xmath51 upper limit on distance from the change in orbital period , @xmath69 .
this independent calculation of distance is in good agreement with the distances derived from parallax . as an additional independent method of calculating distances , we derived upper limits on distance for 17 of the nanograv millisecond pulsars from the rotational spin down rate , @xmath146 .
apart from psr j1024@xmath00719 ( [ app : j1024 ] ) , these upper limits are in agreement with the parallax distances and @xmath51 lower limits .
we have measured proper motions of 35 millisecond pulsars .
we used 30 of these ( from pulsars observed for more than two years ) , in combination with distance estimates and measurements found in the literature , to analyze pulsar motion in galactocentric coordinates .
we found the velocity dispersion components to follow similar proportions as other stellar populations ( equations [ eqn : disp_all ] and [ eqn : disp_exclude ] ) .
we propose two mathematical models of the velocity dispersion magnitudes . in one model
, the bulk of the millisecond pulsar population has essentially the same velocity distribution as other stellar populations with ages of several gyr , and a small number of pulsars are high - speed outliers .
we speculate that the outlier velocities could be due to fortuitously directed kicks during neutron - star formation , though we can not rule out either a different formation mechanism than the bulk of the millisecond pulsar population or an origin in a different dynamical population ( e.g. , halo stars ) . in the other model ,
the millisecond pulsars come from a single velocity distribution , which has dispersion much smaller than canonical pulsars , but moderately larger than other stellar populations with ages of several gyr . in this model ,
the pulsar velocities derive from a combination of their dynamical origin as thick disk objects and from modest velocity boosts during millisecond pulsar formation , presumably at the time they were formed as neutron stars in supernovae .
our goal has been to develop an empirical description of millisecond pulsar dynamics based on measured pulsar parameters alone .
a more comprehensive study would take into account the directions and sensitivities of pulsar search programs , the luminosity distribution of millisecond pulsars , millisecond pulsar birth locations and dynamical evolution , lutz - kelker bias , uncertainties in the dispersion distance model , and so on .
such studies are typically done via monte carlo simulations ; see @xcite for an overview and further references .
a conclusion from the present work is that the analysis of velocities in such studies should take into account the significant dispersion in millisecond pulsar velocities due simply to their large ages , and that the separation of velocities into galactocentric components ( not just total velocities ) is important . _
author contributions . _
the paper evolved from the undergraduate senior thesis of amm , supervised by djn .
amm undertook the bulk of the calculations ; generated most of the text , figures , and tables ; and coordinated the development of the paper .
djn contributed additional text , figures , tables , and analysis , including the appendix on psr j1024@xmath00719 .
ef contributed text and calculations to the binary distance analysis section .
this paper is based on positions , proper motions , and parallaxes derived from the nanograv nine - year data set @xcite .
za , kc , pbd , td , rdf , ef , meg , gj , mj , mtl , ll , mam , djn , ttp , smr , ihs , ks , jks , and wwz all ran observations and analyzed timing models for this data set .
additional specific contributions to the data set are summarized in @xcite ; here we note the particularly critical contributions of pbd , who generated the toas and developed novel techniques for removing systematics from the data , and jae , who developed and implemented the noise model used in the timing analysis .
we thank s. chatterjee , j. cordes , j. verbiest , j. lazio , and c. ng for useful discussions and comments on the manuscript .
the atnf pulsar catalogue @xcite , was an invaluable resource for this project , as was s. chatterjee s online pulsar parallax list .
the nanograv project receives support from national science foundation ( nsf ) pire program award number 0968296 and nsf physics frontier center award number 1430284 .
nanograv research at ubc is supported by an nserc discovery grant and discovery accelerator supplement and the canadian institute for advanced research .
part of this research was carried out at the jet propulsion laboratory , california institute of technology , under a contract with the national aeronautics and space administration .
tp was a student at the national radio astronomy observatory while this project was undertaken .
mtl was partially supported by nasa new york space grant award number nnx15ak07h .
the national radio astronomy observatory is a facility of the nsf operated under cooperative agreement by associated universities , inc .
the arecibo observatory is operated by sri international under a cooperative agreement with the nsf ( ast-1100968 ) , and in alliance with ana g. mndez - universidad metropolitana , and the universities space research association .
natexlab#1#1 , a. a. , ackermann , m. , atwood , w. b. , et al .
2009 , , 699 , 1171 , a. a. , ajello , m. , allafort , a. , et al .
2013 , , 208 , 17 , m. a. , cheng , a. f. , ruderman , m. a. , & shaham , j. 1982 , , 300 , 728 , j. , van kerkwijk , m. h. , koester , d. , et al .
2012 , , 423 , 3316 , z. , fruchter , a. s. , & taylor , j. h. 1994 , , 426 , l85 , z. , brazier , a. , burke - spolaor , s. , et al .
2015 , arxiv e - prints , arxiv:1505.07540 , m. , & binney , j. j. 2009 , , 397 , 1286 , e. d. , guillemot , l. , champion , d. j. , et al .
2013 , , 429 , 1633 bell , j. f. , & bailes , m. 1996 , apj , 456 , l33 , j. , & merrifield , m. 1998 , galactic astronomy ( princeton university press ) , p. j. , garnavich , p. m. , & koester , d. 1998 , , 298 , 207 , f. , nice , d. j. , shrauner , j. a. , & taylor , j. h. 1996 , , 469 , 819 , d. j. , lorimer , d. r. , mclaughlin , m. a. , et al .
2005 , , 363 , 929 , s. , brisken , w. f. , vlemmings , w. h. t. , et al .
2009 , , 698 , 250 , j. m. 2013 , in iau symposium , vol .
291 , iau symposium , ed .
j. van leeuwen , 211216 , j. m. , & chernoff , d. f. 1997 , , 482 , 971 , j. m. , & lazio , t. j. w. 2002 , arxiv astrophysics e - prints , astro - ph/0207156 damour , t. , & taylor , j. h. 1991 , apj , 366 , 501 , s. a. , & schrder , k .- p .
2010 , , 404 , 917 , w. , & binney , j. j. 1998 , , 298 , 387 , a. t. , verbiest , j. p. w. , tingay , s. j. , & bailes , m. 2008 , , 685 , l67 , a. t. , archibald , a. m. , brisken , w. f. , et al .
2012 , , 756 , l25 , j. s. , freire , p. c. c. , cordes , j. m. , et al .
2012 , , 757 , 89 , y. , yang , j. , campbell , r. m. , et al .
2014 , , 782 , l38 , c. m. , guillemot , l. , elik , . , et al .
2013 , , 430 , 571 , w. m. , williams , j. g. , & boggs , d. h. 2009 , interplanetary network progress report , 178 , c1 , e. , stairs , i. h. , & thorsett , s. e. 2014 , , 787 , 82 , p. c. c. , & tauris , t. m. 2014 , , 438 , l86 , p. c. c. , bassa , c. g. , wex , n. , et al .
2011 , , 412 , 2763 , p. c. c. , wex , n. , esposito - farse , g. , et al .
2012 , , 423 , 3328 , m. e. , stairs , i. h. , ferdman , r. d. , et al .
2011 , , 743 , 102 , l. , freire , p. c. c. , cognard , i. , et al .
2012 , , 422 , 1294 , g. , lorimer , d. r. , lyne , a. g. , & kramer , m. 2005 , , 360 , 974 , j. , & flynn , c. 2004 , , 352 , 440 , a. w. , bailes , m. , & ord , s. m. 2006 , , 369 , 1502 , x. , smith , d. a. , guillemot , l. , et al . 2014 , , 570 , a44 , g. h. , stappers , b. w. , bassa , c. g. , et al .
2010 , , 514 , a74 , m. , camilo , f. , johnson , t. j. , et al .
2012 , , 748 , l2 , j. , stocke , j. t. , froning , c. s. , gopakumar , a. , & joshi , b. c. 2012 , , 744 , 183 , b. , & thorsett , s. e. 2010 , , 715 , 335 , m. , & wolszczan , a. 2003 , , 591 , l147 kuijken , k. , & gilmore , g. 1989 , mnras , 239 , 571 , k. , wex , n. , jessner , a. , et al .
2009 , , 400 , 805 , p. , brazier , a. , hessels , j. w. t. , et al .
2015 , arxiv e - prints , arxiv:1504.02294 leys , c. , ley , c. , klein , o. , bernard , p. , & licata , l. 2013 , journal of experimental social psychology , 49 , 764 , o. , lewandowski , w. , wolszczan , a. , & wielebinski , r. 2005 , , 621 , 388 , a. n. , kipphorn , r. a. , nice , d. j. , et al .
2006 , , 642 , 1012 , d. r. 2013 , in iau symposium , vol . 291 , iau symposium , ed . j. van leeuwen , 237242 , d. r. , mclaughlin , m. a. , champion , d. j. , & stairs , i. h. 2007 , , 379 , 282 , d. r. , faulkner , a. j. , lyne , a. g. , et al .
2006 , , 372 , 777 , t. e. , & kelker , d. h. 1973 , , 85 , 573 , r. s. , boyles , j. , ransom , s. m. , et al .
2013 , , 763 , 81 , r. n. , hobbs , g. b. , teoh , a. , & hobbs , m. 2005 , , 129 , 1993 , c. , bailes , m. , bates , s. d. , et al .
2014 , , 439 , 1865 , d. j. , splaver , e. m. , & stairs , i. h. 2001 , , 549 , 516 , d. j. , & taylor , j. h. 1995 , , 441 , 429 , p. , langer , n. , poelarends , a. j. t. , et al . 2004 , , 612 , 1044 press , w. h. , teukolsky , s. a. , vetterling , w. t. , & flannery , b. p. 2007
, numerical recipes : the art of scientific computing , 3@xmath147 edition ( cambridge : cambridge university press ) , m. j. , menten , k. m. , brunthaler , a. , et al .
2014 , , 783 , 130 , d. h. f. m. 2012 , , 427 , 664 , r. , binney , j. , & dehnen , w. 2010 , , 403 , 1829 , s. , bland - hawthorn , j. , binney , j. , et al .
2014 , , 793 , 51 shklovskii , i. s. 1970 , sov .
astron . , 13 , 562 , e. m. 2004 , phd thesis , princeton university , k. , lynch , r. s. , ransom , s. m. , et al .
2014 , , 791 , 67 , f. k. , ray , a. , reisenegger , a. , et al .
2003 , , 406 , 245 , t. m. , & bailes , m. 1996 , , 315 , 432 , t. m. , langer , n. , & kramer , m. 2012 , , 425 , 1601 , t. m. , sanyal , d. , yoon , s .- c . , & langer , n. 2013 , , 558 , a39 , m. , sandhu , j. s. , bailes , m. , et al . 1999 , , 307 , 925 , e. p. j. 2011 , bulletin of the astronomical society of india , 39 , 1 , m. h. , breton , r. p. , & kulkarni , s. r. 2011 , , 728 , 95 , w. , demorest , p. , & oslowski , s. 2012 , astronomical research and technology , 9 , 237 , j. p. w. , lorimer , d. r. , & mclaughlin , m. a. 2010 , , 405 , 564 , j. p. w. , weisberg , j. m. , chael , a. a. , lee , k. j. , & lorimer , d. r. 2012 , , 755 , 39 , j. p. w. , bailes , m. , van straten , w. , et al . 2008 , , 679 , 675 , j. p. w. , bailes , m. , coles , w. a. , et al . 2009 , , 400 , 951 ,
a. , hoffman , i. m. , konacki , m. , anderson , s. b. , & xilouris , k. m. 2000 , , 540 , l41 , x. p. , hobbs , g. , coles , w. a. , et al .
2007 , , 378 , 493 , v. e. 2006 , , 638 , 951 , w. w. , stairs , i. h. , demorest , p. b. , et al .
2015 , arxiv e - prints , arxiv:1504.00662
as described in [ sec : pdotdistances ] , we have derived two distance limits for psr j1024@xmath00719 that contradict one other : an upper limit on distance from spin - down , @xmath95 kpc , and a lower limit on distance from parallax , @xmath96 kpc . to quantify this discrepancy , note that if the pulsar lies within the bound established by the spin - down distance , then the parallax measurement , @xmath148 mas is in error by at least @xmath149 .
the resolution of this conflict is not clear . in this appendix
, we summarize the reasons the spin - down distance measurement is robust , we describe tests of our parallax measurement , we summarize other observations of this pulsar , and we present an orbital model which is a candidate for resolving this discrepancy .
psr j1024@xmath00719 is isolated ( non - binary ) and has rotation properties typical of millisecond pulsars .
it has the largest proper motion among the sources in this paper , and one of the smallest dispersion measures . in [ sec : pdotdistances ] , the spin - down distance was calculated from equation [ eqn : pdot2 ] assuming the observed spin - down , @xmath150 , is entirely due to kinematic effects .
similar spin - down distance upper limits have been previously calculated for this pulsar @xcite . at the upper limit distance , 0.427 kpc ,
the contributions to @xmath146 of the three terms of the right hand side of equation [ eqn : pdot2 ] are @xmath151 for the @xmath59 term ; @xmath152 for the galactic rotation term ; and @xmath153 for the proper motion term , where we have used the pulsar period @xmath154 ms .
the proper motion term dominates the calculation , so the calculation is robust against uncertainties in the galactic potential and rotation and we can write , with error of no more than 2% , @xmath155 . as shown in table
[ tab:1024_previous ] , our measurements of @xmath146 and @xmath156 are in agreement with two previously published timing solutions for this pulsar . for these reasons ,
this upper limit determination is robust , as long as the pulsar is spinning down .
our parallax measurement for psr j1024@xmath00719 is @xmath148 mas .
there was one previously reported measurement , @xmath158 mas @xcite , which differs by @xmath40 from our value .
while this previous value is compatible with the parallax implied by the @xmath146 measurement , it has a large uncertainty and is only marginally significant . to test the validity of our parallax measurement
, we performed a series of tests on the data set as detailed below and as summarized in table [ tab:1024_px ] . for each test , the table shows the best - fit parallax ; the @xmath50 of the fit ; the number of degrees of freedom , @xmath159 ; and the reduced @xmath50 .
each of our tests used the noise model values for this source as determined by @xcite , which consists only of white - noise terms . in all tests
, we found the parallax to be consistent with our standard measured value and to be lower than that implied by the @xmath146 distance limit . _ data subset tests .
_ our data set consists of measurements taken with two instruments , gasp and guppi ( [ sec : obs ] ) .
gasp data were collected at three epochs , in years 2009.8 through 2010.1 , and guppi data were collected at 49 epochs , 2010.2 through 2013.8 .
( there are also a very small number of gasp toas within the guppi date range covering frequencies without good guppi toas . )
we ran independent timing solutions on ( i ) the guppi data only ; ( ii ) the first half of the data set , 2009.8 through 2011.8 ; and ( iii ) the second half of the data set , 2011.8 through 2013.8 . _ timing noise tests .
_ @xcite found no evidence for red noise in this data set . nevertheless , as a test , we ran independent timing solutions with extra spin frequency derivatives as a proxy for timing noise .
defining @xmath160 , we ran tests fitting for all the usual parameters and additionally ( i ) @xmath161 ; ( ii ) @xmath161 and @xmath162 ; ( iii ) @xmath161 through @xmath163 .
_ dispersion measure tests .
_ our standard solution fits for independent values of dispersion measure at every epoch .
these dispersion measure values are dominated by a linear trend ( * ? ? ?
we ran two tests in which we fit dispersion measure as a linear trend combined with a solar wind electron density with a @xmath164 falloff , with electron density at 1 au of @xmath165 of 0 and 10 @xmath166 in the two tests .
in such models , the solar wind electron density is highly covariant with the best - fit parallax value , so in principle one can adjust @xmath165 to attain any derived parallax value ; however , we found that the quality of the fit diminished significantly if @xmath165 was increased beyond 10 @xmath166 .
@xcite noted that the gamma - ray emission of this pulsar would be unusually high at this distance ( they used 0.410kpc ) , and that a closer distance ( 0.350 kpc ) would be needed for its gamma - ray luminosity to be similar to other millisecond pulsars .
this pulsar has also been detected in x - rays @xcite .
@xcite presented optical observations of the field of psr j1024@xmath00719 .
they detected two sources near the pulsar .
one was bright ( @xmath167 , @xmath168 , @xmath169 , @xmath170 ) , with a spectrum similar to a k - type dwarf star .
one was faint ( @xmath171 , @xmath172 , @xmath173 , @xmath174 ) .
the bight star may have a proper motion in a direction similar to the pulsar , although uncertainties are large .
since the pulsar is isolated , the bright source may be unassociated with the pulsar , but its presence is an interesting coincidence .
here we speculate what circumstances could reconcile the measurements if the pulsar is at the parallax lower limit distance , @xmath176 kpc . at this distance , according to equation [ eqn : pdot2 ] , the observed period derivative would be biased upward by @xmath177 , i.e. , this is a lower limit to the observed period derivative if equation [ eqn : pbdot2 ] fully describes the biases to the observed period derivative .
since the actual observed period derivative is @xmath178 , under this model , the terms of equation [ eqn : pbdot2 ] are not sufficient to explain the observed value .
an additional bias of @xmath179 is needed , which could arise from an additional acceleration of order @xmath180 .
such an acceleration could be caused by the potential of a globular cluster indeed , millisecond pulsars in globular cluster cores have a wide range of observed positive and negative @xmath146 values . however , there is no cluster in the direction of psr j1024-@xmath181 .
such an acceleration could also be caused by binary motion in a wide orbit .
the acceleration would change over the course of the orbit , causing a change in the observed @xmath146 , i.e. , a nonzero @xmath182 .
the difference in observed @xmath146 between @xcite and the nanograv nine - year value is @xmath183 over a time span of 9.6 yr ( using the centers of the observing data spans ) ; this corresponds to an approximate upper limit @xmath184 .
an orbit would have to involve @xmath146 values that varied on the scale of @xmath179 , so the time scale of such variations would be @xmath185 yr , or an orbital period of @xmath186 yr . for acceleration @xmath187 , where @xmath188 , this gives pulsar orbital radius @xmath189 .
for a 1.4 m@xmath190 pulsar , this would require a companion star of mass @xmath191 .
much larger masses would also satisfy the constraints ; intriguingly , this includes the mass of a k - type star as observed by @xcite ( [ sec:1024_other ] ) .
placing the pulsar at this large distance would present challenges , though .
its two - dimensional space velocity derived from proper motion would be 260 km s@xmath4 , higher than typical millisecond pulsars ; its dm would be much lower than that predicted by the ne2001 electron density model ; and its gamma - ray efficiency would be very high , much greater than the value calculated in @xcite .
lcccllc j@xmath192 & @xmath193 & @xmath194 & & 00:23:16.87910(3 ) & 09:23:23.871(1 ) & @xmath195 + j@xmath196 & @xmath197 & @xmath198 & & 00:30:27.42826(5 ) & 04:51:39.711(2 ) & @xmath199 + j@xmath200 & @xmath201 & @xmath202 & & 03:40:23.28818(2 ) & 41:30:45.2903(5 ) & @xmath203 + j@xmath204 & @xmath205 & @xmath206 & & 06:13:43.975631(3 ) & @xmath002:00:47.2223(1 ) & @xmath207 + j@xmath208 & @xmath209 & @xmath210 & & 06:45:59.081898(9 ) & 51:58:14.9208(1 ) & @xmath211 + j@xmath212 & @xmath213 & @xmath214 & & 09:31:19.1180(4 ) & @xmath019:02:55.015(6 ) & @xmath215 + j@xmath216 & @xmath217 & @xmath218 & & 10:12:33.43745(6 ) & 53:07:02.3071(7 ) & @xmath219 + j@xmath220 & @xmath221 & @xmath222 & & 10:24:38.670189(9 ) & @xmath007:19:19.5396(3 ) & @xmath223 + j@xmath224 & @xmath225 & @xmath226 & & 14:55:47.97069(2 ) & @xmath033:30:46.3833(6 ) & @xmath227 + j@xmath228 & @xmath229 & @xmath230 & & 16:00:51.903261(4 ) & @xmath030:53:49.3830(2 ) & @xmath231 + j@xmath232 & @xmath233 & @xmath234 & & 16:14:36.50708(2 ) & @xmath022:30:31.233(2 ) & @xmath235 + j@xmath236 & @xmath237 & @xmath238 & & 16:40:16.744825(3 ) & 22:24:08.84178(6 ) & @xmath239 + j@xmath240 & @xmath241 & @xmath242 & & 16:43:38.16140(2 ) & @xmath012:24:58.676(1 ) & @xmath219 + j@xmath243 & @xmath244 & @xmath245 & & 17:13:49.5331505(5 ) & 07:47:37.49284(2 ) & @xmath239 + j@xmath246 & @xmath247 & @xmath248 & & 17:38:53.96744(2 ) & 03:33:10.8824(5 ) & @xmath223 + j@xmath249 & @xmath250 & @xmath251 & & 17:41:31.144770(5 ) & 13:51:44.12241(15 ) & @xmath252 + j@xmath253 & @xmath254 & @xmath255 & & 17:44:29.407190(3 ) & @xmath011:34:54.6925(2 ) & @xmath256 + j@xmath257 & @xmath258 & @xmath259 & & 17:47:48.71665(1 ) & @xmath040:36:54.7795(7 ) & @xmath260 + j@xmath261 & @xmath262 & @xmath263 & & 18:32:27.5936(2 ) & @xmath008:36:54.98(4 ) & @xmath264 + j@xmath265 & @xmath266 & @xmath267 & & 18:53:57.318423(7 ) & 13:03:44.0596(3 ) & @xmath268 + b@xmath269 & @xmath270 & @xmath271 & & 18:57:36.390614(4 ) & 09:43:17.2075(1 ) & @xmath272 + j@xmath273 & @xmath274 & @xmath275 & & 19:03:05.79287(3 ) & 03:27:19.194(1 ) & @xmath276 + j@xmath277 & @xmath278 & @xmath279 & & 19:09:47.434674(1 ) & @xmath037:44:14.46667(7 ) & @xmath280 + j@xmath281 & @xmath282 & @xmath283 & & 19:10:09.70147(1 ) & 12:56:25.4727(2 ) & @xmath284 + j@xmath285 & @xmath286 & @xmath287 & & 19:18:48.033256(5 ) & @xmath006:42:34.8877(3 ) & @xmath288 + j@xmath289 & @xmath290 & @xmath291 & & 19:23:22.49331(2 ) & 25:15:40.6164(5 ) & @xmath292 + b@xmath293 & @xmath294 & @xmath295 & & 19:39:38.561227(2 ) & 21:34:59.12567(5 ) & @xmath296 + j@xmath297 & @xmath298 & @xmath299 & & 19:44:09.329903(7 ) & 09:07:23.0362(3 ) & @xmath252 + j@xmath300 & @xmath301 & @xmath302 & & 19:49:29.6379(4 ) & 31:06:03.795(5 ) & @xmath303 + b@xmath304 & @xmath305 & @xmath306 & & 19:55:27.87546(5 ) & 29:08:43.4464(5 ) & @xmath252 + j@xmath307 & @xmath308 & @xmath309 & & 20:10:45.920937(8 ) & @xmath013:23:56.0755(5 ) & @xmath310 + j@xmath311 & @xmath312 & @xmath313 & & 20:17:22.70503(1 ) & 06:03:05.5686(4 ) & @xmath314 + j@xmath315 & @xmath316 & @xmath317 & & 20:43:20.882230(3 ) & 17:11:28.92694(9 ) & @xmath318 + j@xmath319 & @xmath320 & @xmath321 & & 21:45:50.46089(3 ) & @xmath007:50:18.491(1 ) & @xmath322 + j@xmath323 & @xmath324 & @xmath325 & & 22:14:38.85097(2 ) & 30:00:38.1976(2 ) & @xmath326 + j@xmath327 & @xmath328 & @xmath329 & & 23:02:46.97878(3 ) & 44:42:22.0928(3 ) & @xmath203 + j@xmath330 & @xmath331 & @xmath332 & & 23:17:09.236644(9 ) & 14:39:31.2557(2 ) & @xmath333 j@xmath334 & 2.3 & @xmath335 & @xmath336 & @xmath337 & @xmath338 & @xmath339 & @xmath340 & & & + j@xmath341 & 8.8 & @xmath342 & @xmath343 & @xmath344 & @xmath345 & @xmath346 & @xmath347 & @xmath348 & @xmath349 & 1 + j@xmath350 & 1.7 & @xmath351 & @xmath352 & @xmath353 & @xmath354 & @xmath355 & @xmath356 & & & + j@xmath357 & 8.6 & @xmath358 & @xmath359 & @xmath360 & @xmath361 & @xmath362 & @xmath363 & @xmath364 & @xmath365 & 2 + j@xmath366 & 2.4 & @xmath367 & @xmath368 & @xmath369 & @xmath370 & @xmath371 & @xmath372 & @xmath373 & @xmath374 & 3 + j@xmath375 & 0.6 & & & & & & & & & + j@xmath376 & 9.2 & @xmath377 & @xmath378 & @xmath379 & @xmath380 & @xmath381 & @xmath382 & @xmath383 & @xmath384 & 4 + j@xmath385 & 4.0 & @xmath386 & @xmath387 & @xmath388 & @xmath389 & @xmath390 & @xmath391 & @xmath392 & @xmath393 & 2 + j@xmath394 & 9.2 & @xmath395 & @xmath396 & @xmath397 & @xmath398 & @xmath399 & @xmath400 & @xmath401 & @xmath402 & 5 + j@xmath403 & 6.0 & @xmath404 & @xmath405 & @xmath406 & @xmath407 & @xmath408 & @xmath409 & @xmath410 & @xmath411 & 2 + j@xmath412 & 5.1 & @xmath413 & @xmath414 & @xmath415 & @xmath416 & @xmath417 & @xmath418 & & & + j@xmath419 & 8.9 & @xmath420 & @xmath421 & @xmath422 & @xmath423 & @xmath424 & @xmath425 & @xmath426 & @xmath427 & 6 + j@xmath428 & 9.0 & @xmath429 & @xmath430 & @xmath431 & @xmath432 & @xmath433 & @xmath434 & @xmath435 & @xmath436 & 2 + j@xmath437 & 8.8 & @xmath438 & @xmath439 & @xmath440 & @xmath441 & @xmath442 & @xmath443 & @xmath444 & @xmath445 & 7@xmath446 + j@xmath447 & 4.0 & @xmath448 & @xmath449 & @xmath450 & @xmath451 & @xmath452 & @xmath453 & @xmath454 & @xmath455 & 8 + j@xmath456 & 2.3 & @xmath457 & @xmath458 & @xmath459 & @xmath460 & @xmath461 & @xmath462 & & & + j@xmath463 & 9.2 & @xmath464 & @xmath465 & @xmath466 & @xmath467 & @xmath468 & @xmath469 & @xmath470 & @xmath471 & 2 + j@xmath472 & 1.7 & @xmath473 & @xmath474 & @xmath475 & @xmath476 & @xmath477 & @xmath478 & & & + j@xmath479 & 0.6 & & & & & & & & & + j@xmath480 & 2.3 & @xmath481 & @xmath482 & @xmath483 & @xmath484 & @xmath485 & @xmath486 & @xmath487 & @xmath488 & 9@xmath446 + b@xmath489 & 8.9 & @xmath490 & @xmath491 & @xmath492 & @xmath493 & @xmath494 & @xmath495 & @xmath496 & @xmath497 & 2 + j@xmath498 & 4.0 & @xmath499 & @xmath500 & @xmath501 & @xmath502 & @xmath503 & @xmath504 & @xmath505 & @xmath506 & 10 + j@xmath507 & 9.1 & @xmath508 & @xmath509 & @xmath510 & @xmath511 & @xmath512 & @xmath513 & @xmath514 & @xmath515 & 2 + j@xmath516 & 4.7 & @xmath517 & @xmath518 & @xmath519 & @xmath460 & @xmath520 & @xmath521 & @xmath522 & @xmath523 & 9@xmath446 + j@xmath524 & 9.0 & @xmath525 & @xmath526 & @xmath527 & @xmath528 & @xmath529 & @xmath530 & @xmath531 & @xmath532 & 11 + j@xmath533 & 2.2 & @xmath534 & @xmath535 & @xmath536 & @xmath537 & @xmath538 & @xmath539 & @xmath540 & @xmath541 & 12 + b@xmath542 & 9.1 & @xmath543 & @xmath544 & @xmath545 & @xmath546 & @xmath547 & @xmath548 & @xmath549 & @xmath550 & 2 + j@xmath551 & 2.3 & @xmath552 & @xmath553 & @xmath554 & @xmath555 & @xmath556 & @xmath557 & @xmath558 & @xmath559 & 13 + j@xmath560 & 1.2 & @xmath561 & @xmath562 & @xmath563 & @xmath564 & @xmath565 & @xmath566 & @xmath488 & @xmath567 & 14 + b@xmath568 & 2.3 & @xmath569 & @xmath570 & @xmath571 & @xmath572 & @xmath573 & @xmath574 & @xmath575 & @xmath576 & 9@xmath446 + j@xmath577 & 4.1 & @xmath578 & @xmath579 & @xmath580 & @xmath581 & @xmath582 & @xmath583 & & & + j@xmath584 & 1.7 & @xmath585 & @xmath586 & @xmath587 & @xmath588 & @xmath589 & @xmath590 & & & + j@xmath591 & 2.3 & @xmath592 & @xmath593 & @xmath594 & @xmath595 & @xmath596 & @xmath597 & @xmath598 & @xmath599 & 15 + j@xmath600 & 9.1 & @xmath601 & @xmath602 & @xmath603 & @xmath604 & @xmath605 & @xmath606 & @xmath607 & @xmath608 & 2 + j@xmath609 & 2.1 & @xmath610 & @xmath611 & @xmath612 & @xmath613 & @xmath614 & @xmath615 & & & + j@xmath616 & 1.7 & @xmath617 & @xmath618 & @xmath619 & @xmath620 & @xmath621 & @xmath622 & & & + j@xmath623 & 8.9 & @xmath624 & @xmath625 & @xmath626 & @xmath627 & @xmath628 & @xmath629 & @xmath630 & @xmath631 & 16 lccc j@xmath334 & @xmath632 & & + j@xmath341 & @xmath633 & @xmath634 & 1 + j@xmath350 & @xmath635 & & + j@xmath357 & @xmath636 & @xmath637 & 2 + j@xmath366 & @xmath638 & @xmath639 & 3 + j@xmath375 & @xmath640 & & + j@xmath376 & @xmath641 & @xmath642 & 4 + j@xmath385 & @xmath643 & @xmath644 & 5 + j@xmath394 & @xmath645 & & + j@xmath403 & @xmath646 & @xmath647 & 2 + j@xmath412 & @xmath648 & @xmath649 & 6 + j@xmath419 & @xmath650 & @xmath651 & 7 + j@xmath428 & @xmath652 & @xmath653 & 2 + j@xmath437 & @xmath654 & @xmath655 & 2 + j@xmath447 & @xmath656 & @xmath657 & 8,9 + j@xmath456 & @xmath658 & & + j@xmath463 & @xmath659 & @xmath660 & 2 + j@xmath472 & @xmath661 & & + j@xmath479 & @xmath662 & & + j@xmath480 & @xmath663 & @xmath664 & 10 + b@xmath489 & @xmath665 & @xmath666 & 2 + j@xmath498 & @xmath667 & & + j@xmath507 & @xmath668 & @xmath669 & 2 + j@xmath516 & @xmath670 & @xmath671 & 10 + j@xmath524 & @xmath672 & & + j@xmath533 & @xmath673 & & + b@xmath542 & @xmath674 & & + j@xmath551 & @xmath675 & & + j@xmath560 & @xmath676 & & + b@xmath568 & @xmath677 & & + j@xmath577 & @xmath678 & & + j@xmath584 & @xmath632 & & + j@xmath591 & @xmath679 & & + j@xmath600 & @xmath680 & @xmath681 & 2 + j@xmath609 & @xmath682 & & + j@xmath616 & @xmath683 & & + j@xmath623 & @xmath684 & & j@xmath196 & @xmath685 & @xmath686 & @xmath687 + j@xmath204 & @xmath688 & @xmath689 & @xmath690 + j@xmath208 & @xmath691 & @xmath692 & @xmath693 + j@xmath228 & @xmath694 & @xmath695 & @xmath696 + j@xmath232 & @xmath697 & @xmath698 & @xmath699 + j@xmath243 & @xmath700 & @xmath701 & @xmath702 + j@xmath253 & @xmath703 & @xmath704 & @xmath705 + j@xmath277 & @xmath706 & @xmath707 & @xmath708 + j@xmath285 & @xmath709 & @xmath710 & @xmath711 + j@xmath315 & @xmath712 & @xmath713 & @xmath714 + j@xmath319 & @xmath715 & @xmath716 & @xmath717 + j@xmath330 & @xmath718 & @xmath719 & @xmath720 + j@xmath192 & @xmath721 & @xmath722 & @xmath693 + j@xmath216 & @xmath723 & @xmath724 & @xmath725 + j@xmath220 & @xmath726 & @xmath727 & @xmath728 + j@xmath224 & @xmath729 & @xmath730 & @xmath731 + j@xmath236 & @xmath732 & @xmath733 & @xmath734 + j@xmath240 & @xmath735 & @xmath736 & @xmath737 + j@xmath246 & @xmath738 & @xmath739 & @xmath740 + j@xmath249 & @xmath741 & @xmath742 & @xmath743 + j@xmath265 & @xmath744 & @xmath745 & @xmath746 + b@xmath269 & @xmath747 & @xmath748 & @xmath749 + j@xmath273 & @xmath750 & @xmath751 & @xmath752 + j@xmath281 & @xmath726 & @xmath727 & @xmath753 + j@xmath289 & @xmath754 & @xmath755 & @xmath696 + b@xmath293 & @xmath756 & @xmath757 & @xmath758 + j@xmath297 & @xmath759 & @xmath760 & @xmath761 + b@xmath304 & @xmath762 & @xmath763 & @xmath764 + j@xmath307 & @xmath765 & @xmath766 & @xmath767 + j@xmath323 & @xmath768 & @xmath769 & @xmath770 j@xmath412 & 8.687 & @xmath771 & @xmath772 + j@xmath507 & 1.533 & @xmath773 & @xmath774 + j@xmath357 & 1.199 & @xmath775 & @xmath776 + j@xmath376 & 0.604 & @xmath777 & @xmath778 + j@xmath419 & 175.461 & @xmath779 & @xmath780 + j@xmath524 & 10.913 & @xmath781 & @xmath782 + j@xmath600 & 6.839 & @xmath783 & @xmath784 lccc j0023 + 0923 & 3.050 & 11.421 & @xmath785 + j0030 + 0451 & 4.865 & 10.174 & @xmath786 + j0613 - 0200 & 3.062 & 9.590 & @xmath787 + j0645 + 5158 & 8.853 & 4.920 & @xmath788 + j1012 + 5307 & 5.256 & 17.127 & @xmath789 + j1024 - 0719 & 5.162 & 18.552 & @xmath790 + j1614 - 2230 & 3.151 & 9.624 & @xmath791 + j1640 + 2224 & 3.163 & 2.818 & @xmath792 + j1744 - 1134 & 4.075 & 8.934 & @xmath793 + j1909 - 3744 & 2.947 & 14.025 & @xmath794 + j1923 + 2515 & 3.788 & 9.553 & @xmath795 + j1944 + 0907 & 5.185 & 17.339 & @xmath796 + j2010 - 1323 & 5.223 & 4.824 & @xmath797 + j2043 + 1711 & 2.380 & 5.243 & @xmath798 + j2145 - 0750 & 16.052 & 29.790 & @xmath799 + j2214 + 3000 & 3.119 & 14.701 & @xmath800 + j2317 + 1439 & 3.445 & 2.430 & @xmath801 j@xmath802 & @xmath803 & @xmath804 & @xmath805 & @xmath806 & @xmath807 + j@xmath808 & @xmath809 & @xmath810 & @xmath811 & @xmath812 & @xmath813 + j@xmath814 & @xmath815 & @xmath816 & @xmath817 & @xmath818 & @xmath819 + j@xmath820 & @xmath821 & @xmath822 & @xmath823 & @xmath824 & @xmath825 + j@xmath826 & @xmath827 & @xmath828 & @xmath829 & @xmath830 & @xmath831 + j@xmath832 & @xmath833 & @xmath834 & @xmath835 & @xmath836 & @xmath837 + j@xmath376 & @xmath838 & @xmath839 & @xmath840 & @xmath841 & @xmath842 + j@xmath843 & @xmath844 & @xmath845 & @xmath846 & @xmath847 & @xmath848 + j@xmath849 & @xmath850 & @xmath851 & @xmath852 & @xmath853 & @xmath854 + j@xmath855 & @xmath856 & @xmath857 & @xmath858 & @xmath859 & @xmath831 + j@xmath860 & @xmath861 & @xmath862 & @xmath863 & @xmath864 & @xmath848 + j@xmath865 & @xmath866 & @xmath867 & @xmath868 & @xmath869 & @xmath825 + b@xmath870 & @xmath871 & @xmath872 & @xmath873 & @xmath874 & @xmath875 + j@xmath876 & @xmath877 & @xmath878 & @xmath879 & @xmath880 & @xmath848 + j@xmath881 & @xmath882 & @xmath883 & @xmath884 & @xmath885 & @xmath831 + j@xmath447 & @xmath886 & @xmath887 & @xmath888 & @xmath889 & @xmath890 + j@xmath891 & @xmath892 & @xmath893 & @xmath894 & @xmath895 & @xmath896 + j@xmath897 & @xmath898 & @xmath899 & @xmath900 & @xmath901 & @xmath902 + j@xmath903 & @xmath904 & @xmath905 & @xmath906 & @xmath880 & @xmath837 + j@xmath560 & @xmath907 & @xmath908 & @xmath909 & @xmath910 & @xmath911 + b@xmath912 & @xmath913 & @xmath914 & @xmath915 & @xmath916 & @xmath917 + j@xmath918 & @xmath919 & @xmath920 & @xmath921 & @xmath922 & @xmath923 + j@xmath924 & @xmath925 & @xmath926 & @xmath927 & @xmath928 & @xmath929 + j@xmath930 & @xmath931 & @xmath932 & @xmath933 & @xmath934 & @xmath825 + j@xmath935 & @xmath936 & @xmath937 & @xmath938 & @xmath939 & @xmath831 + j@xmath940 & @xmath941 & @xmath942 & @xmath943 & @xmath944 & @xmath945 lcc j@xmath814 & @xmath946 & @xmath807 + j@xmath947 & @xmath948 & @xmath813 + j@xmath376 & @xmath949 & @xmath819 + j@xmath843 & @xmath950 & @xmath825 + j@xmath849 & @xmath951 & @xmath831 + b@xmath952 & @xmath953 & @xmath837 + j@xmath447 & @xmath954 & @xmath842 + j@xmath930 & @xmath955 & @xmath848 j@xmath334 & @xmath959 & & + j@xmath357 & & & @xmath960 + j@xmath366 & & @xmath961 & + j@xmath385 & @xmath962 & & + j@xmath403 & & & @xmath963 + j@xmath412 & & @xmath964 & + j@xmath428 & & @xmath965 & + j@xmath463 & & @xmath966 & @xmath967 + j@xmath480 & & & @xmath968 + b@xmath969 & & & @xmath970 + j@xmath498 & @xmath971 & @xmath972 & @xmath973 + j@xmath507 & & @xmath974 & @xmath975 + j@xmath524 & & & @xmath976 + j@xmath533 & & & @xmath977 + b@xmath978 & @xmath979 & & @xmath980 + j@xmath551 & & & @xmath981 + j@xmath591 & @xmath982 & & @xmath983 + j@xmath623 & @xmath984 & & + j@xmath802 & @xmath985 & & + j@xmath808 & & & @xmath986 + j@xmath814 & @xmath987 & & + j@xmath820 & & & @xmath988 + j@xmath826 & @xmath989 & & + j@xmath832 & & @xmath990 & + j@xmath376 & @xmath991 & @xmath992 & @xmath993 + j@xmath843 & @xmath994 & & @xmath995 + j@xmath855 & @xmath996 & & @xmath997 + j@xmath860 & @xmath998 & & @xmath999 + b@xmath870 & @xmath1000 & @xmath1001 & + j@xmath876 & & & @xmath1002 + j@xmath881 & & & @xmath1003 + j@xmath447 & & & @xmath1004 + j@xmath891 & & & @xmath1005 + j@xmath897 & & & @xmath1006 + j@xmath903 & & & @xmath1007 + j@xmath560 & & & @xmath1002 + b@xmath912 & @xmath1008 & @xmath1009 & @xmath1010 + j@xmath918 & @xmath1011 & & @xmath1012 + j@xmath924 & @xmath1013 & & @xmath1014 + j@xmath930 & & @xmath1015 & + j@xmath935 & & @xmath1016 & + j@xmath940 & @xmath1017 & & lrrrrc @xcite & 1.853(6 ) & -34.9(4 ) & -47(1 ) & 2003.0 - 2005.3 + @xcite & 1.852(8 ) & -35.2(2 ) & -48.2(3 ) & 1996.1 - 2008.2 + nanograv nine - year release ( this work ) & 1.8551(1 ) & -34.2(1 ) & -48.0(3 ) & 2009.8 - 2013.8 ) .
parallax distances are plotted as @xmath73 uncertainties ; ne2001 dispersion model distances are plotted with 25% uncertainties .
two pulsars for which there is no distance that is both within @xmath73 of the parallax distance and within 25% of the ne2001 distance are indicated by name and by light blue uncertainty boxes .
previous parallax measurements of psrs j1713 + 0747 and j1744 - 1134 were used as input data to the ne2001 model ; our measurements of these pulsars are shown in blue .
all of our other measurements are shown in green .
[ fig : sigpxvsdm ] ] ) .
lower limit parallax distances are plotted as 95% confidence values ; ne2001 dispersion model distances are plotted with 25% uncertainties .
previous parallax measurements or limits for psrs b1855 + 09 and b1937 + 21 were used as input data to the ne2001 model ; our limits for these pulsars are shown in blue .
all of our other limits are shown in green .
[ fig : pxlimitvsdm ] ] | using the nine - year radio - pulsar timing data set from the north american nanohertz observatory for gravitational waves ( nanograv ) , collected at arecibo observatory and the green bank telescope , we have measured the positions , proper motions , and parallaxes for 37 millisecond pulsars .
we report twelve significant parallax measurements and distance measurements , and eighteen lower limits on distance .
we compare these measurements to distances predicted by the ne2001 interstellar electron density model and find them to be in general agreement .
we use measured orbital - decay rates and spin - down rates to confirm two of the parallax distances and to place distance upper limits on other sources ; these distance limits agree with the parallax distances with one exception , psr j1024@xmath00719 , which we discuss at length . using the proper motions of the 37 nanograv pulsars in combination with other published measurements , we calculate the velocity dispersion of the millisecond pulsar population in galactocentric coordinates .
we find the radial , azimuthal , and perpendicular dispersions to be @xmath1 , @xmath2 , and @xmath3 km s@xmath4 , respectively , in a model that allows for high - velocity outliers ; or @xmath5 , @xmath6 , and @xmath7 km s@xmath4 for the full population .
these velocity dispersions are far smaller than those of the canonical pulsar population , and are similar to older galactic disk populations .
this suggests that millisecond pulsar velocities are largely attributable to their being an old population rather than being artifacts of their birth and evolution as neutron star binary systems .
the components of these velocity dispersions follow similar proportions to other galactic populations , suggesting that our results are not biased by selection effects . |
the acoustic peaks imprinted on the temperature power spectrum of the cosmic microwave background ( cmb ) have now been measured with impressive precision by a number of experiments @xcite .
these observations place tight constraints on the values of many of the fundamental cosmological parameters . with the fifth year of integration from the wmap satellite
there is seemingly little room left for any deviation from the basic @xmath0cdm model @xcite .
however , degeneracies exist between some parameters which can not be broken by cmb data alone ( e.g. @xcite .
perhaps the two most important examples are the curvature of the universe and the equation of state of the dark energy , @xmath9 , where @xmath10 is the pressure of the dark energy and @xmath11 is its density .
meaningful constraints can not be obtained on these parameters using cmb data in isolation .
the full potential of the cmb measurements is realized when these data are combined with other observations , such as the hubble diagram of type ia supernovae and the large scale structure of the universe as traced by galaxies @xcite .
these complementary data sets come from the late universe compared with the cmb data , and the interpretation of the observations is more complicated and controversial .
type ia supernovae ( sne ) have been proposed as standard candles which can probe the luminosity distance - redshift relation . the first strong evidence in support of a cosmological constant came from combining the sne data with cmb measurements @xcite .
the type ia hubble diagram has come under intense scrutiny to uncover any hint of non - standardness arising from the nature of the host galaxy , possible evolution with redshift or variation in dust extinction @xcite .
a recent joint analysis of sne from different datasets suggests that the systematic error on the equation of state parameter from a joint cmb and sne analysis is comparable to the size of the random error @xcite .
the power spectrum of galaxy clustering has also been used in combination with cmb data ( e.g. percival et al .
according to the standard lore , the galaxy power spectrum on large scales is expected to have a simple relation to the underlying dark matter spectrum .
moreover , the shape of the spectrum is believed to closely follow that expected in linear perturbation theory for the matter , which can be readily computed given a set of values for the cosmological parameters . however , with the availability of improving measurements and more refined modelling of the galaxy power spectrum it has become clear that this simple picture is no longer sufficiently accurate to describe the data .
additional levels of modelling of the deviations from linear theory have to be incorporated into the analysis .
these include empirical models of the nonlinear distortion of the power spectrum @xcite and possible scale dependent biases between the clustering of galaxies and mass @xcite .
recent studies have cast doubt on the accuracy of these prescriptions @xcite . in principle , if the simple models described the form of the observed power spectrum , then the power spectra measured from different galaxy samples should yield equivalent constraints on cosmological parameters .
however , @xcite found a fundamental difference in the shapes of the galaxy power spectra measured from the two - degree galaxy redshift survey and the sloan digital sky survey , after attempting to correct the measured spectra .
these authors found that the red selection of the sdss galaxies results in a strong scale dependent bias ( see also * ? ? ? * ) .
similar scale dependent effects have been seen in the power spectrum of dark matter haloes and galaxies modelled in simulations @xcite .
recent analyses have not attempted to model the overall shape or amplitude of the power spectrum , as a consequence of the difficulties described above in interpreting the results from different samples . instead , attention has shifted to a pattern of oscillatory features called the baryonic acoustic oscillations ( bao ) , which are imprinted on the matter power spectrum .
these features have been advocated as a standard ruler which can be used to measure the distance - redshift relation , and hence constrain the dark energy equation of state @xcite .
the bao arise from oscillations in the baryon - photon fluid prior to matter - radiation decoupling .
this phenomenon gives rise to the peaks seen in the power spectrum of temperature fluctuations in the cmb . in the matter power spectrum ,
the oscillations have a much smaller amplitude as baryons only account for around 20 per cent of the total matter density of the universe .
furthermore , the oscillations in the matter spectrum are out of phase with those in the cmb @xcite .
the oscillation scale is related to the size of the sound horizon at recombination , which can be measured with high accuracy from the cmb @xcite .
the apparent size of the bao ruler depends upon the parameters @xmath12 and @xmath13 , as these determine the angular diameter distance out to a given redshift . in practice ,
the bao are not precisely a standard ruler at the level of precision demanded for their interpretation in future surveys .
nevertheless , by modelling the appearance of the bao accurately , they are still a valuable probe of cosmological parameters @xcite .
careful simulation work and modelling has shown that techniques can be developed which can overcome long - wavelength gradients in the power spectrum to yield robust constraints on the bao using the galaxy power spectrum @xcite .
the bao signal has been detected in both the 2dfgrs and sdss surveys @xcite .
the most powerful bao measurements currently come from samples of luminous red galaxies ( lrgs ) @xcite .
detections at a lower significance have also been reported using galaxy clusters @xcite .
the imprint of these features has even been found in the three point function of lrgs @xcite . despite this rapid progress ,
the conclusions drawn from measurements of the bao remain unclear .
for example , @xcite analysed the bao signal in a joint galaxy sample drawn from the sloan digital sky survey ( sdss ) data release five ( dr5 ) and the two - degree field galaxy redshift survey ( 2dfgrs ) . ignoring the information from the amplitude and long - wavelength shape of the power spectrum , and simply isolating the bao , the results of percival et al
. showed a 2.4@xmath14 discrepancy with the distance measurements inferred from the supernovae type ia ( sn ) data by @xcite , signaling a possible problem in the modelling of the bao data , or a challenge to the @xmath0cdm model .
much theoretical work has been devoted to uncovering scale dependent effects in the bao and in improving the modelling of the signal in galaxy surveys @xcite . in order to realize the full potential of the bao technique as a cosmological probe it is essential to quantify any systematics in the signal and to understand how the measurements relate to cosmological parameters .
first , let us debunk some possible preconceptions about bao .
as we have remarked above , the bao are not precisely a standard ruler . in the correlation function ,
the fourier transform of the power spectrum , the bao appear as a broad peak at large pair separations @xcite .
@xcite showed that even in linear perturbation theory , the maximum of the bao peak in correlation function does not coincide with the sound horizon scale at the percent level of accuracy required to fully exploit the measurements expected from forthcoming galaxy surveys .
furthermore , @xcite and @xcite have shown that both large volume numerical simulations and theoretical predictions based on renormalized perturbation theory ( rpt ) indicate that the bao peak in the correlation function is shifted and distorted in a non - trivial manner relative to the prediction of linear perturbation theory .
if unaccounted for , these shifts bias the constraints obtained by using the bao measurements as a standard ruler .
careful modelling of the correlation function is therefore required to extract the cosmological information encoded in large scale structure .
@xcite argued that the correlation function is less affected by scale dependent effects than the power spectrum and that a simple model for the correlation function proposed by @xcite , based on rpt @xcite , gives an essentially unbiased measurement of the dark energy equation of state .
this means that information from the large scale shape of the correlation function , in addition to the form of the bao peak , can be used to provide robust constraints on cosmological parameters .
the correlation function therefore provides a better constraint on the distance scale than the more conservative , `` bao only '' approach required when using the power spectrum ( i.e. which requires the long wavelength shape information to be discarded along with the amplitude ) . in this paper
we apply this new model to the shape of the redshift space correlation function , @xmath15 , of the sdss data release 6 ( dr6 ) lrgs measured by @xcite .
the lrg sample analysed by @xcite is twice the size of the sample used to make the first detection of the bao by @xcite , and has also yielded the first measurement of the radial bao signal , which constrains the hubble parameter @xcite .
we combine the lrg clustering information with the latest measurements of cmb and sne data .
we focus on the constraints on the dark energy equation of state and the curvature of the universe , which are the parameters where the extra information from the shape of @xmath15 can dramatically improve upon the cmb only constraints .
we also pay special attention to the consistency of the results obtained with different dataset combinations .
the outline of the paper is as follows . in section
[ sec : data ] , we describe the data used in our parameter estimation . in section [
sec : method ] we describe the details of our modelling of the shape of the redshift space correlation function and compare it with measurements in n - body simulations .
we also set out the different parameter spaces that we study and describe our methodology for parameter estimation . in section
[ sec : test ] we assess the impact of the details of the parameter estimation technique on the obtained constraints on cosmological parameters .
section [ sec : results ] presents our main results for the parameter constraints obtained by comparing theoretical models to the cmb data and the correlation function of the sdss - dr6 . in section [ sec : distances ] we focus on the constraints on certain distance combinations obtained from the shape of @xmath15 .
we summarize our conclusions in section [ sec : conclusions ] .
appendix a gives the theoretical motivation for the model we use to describe the correlation function and appendix b gives the covariance matrix for the distance constraints .
here we describe the different datasets that we use to constrain cosmological parameters .
the modelling of the correlation function of luminous red galaxies is described later on in section [ ssec : rpt ] .
the datasets described below are used in different combinations to check the consistency of the constraints returned .
luminous red galaxies ( lrgs ) are an efficient tracer of the large scale structure of the universe .
these galaxies are selected by color and magnitude cuts designed to identify intrinsically red , bright galaxies using sdss photometry ( see eisenstein et al 2001 for a complete description of the color cuts ) .
lrgs can be seen out to higher redshifts than galaxies in a simple magnitude limited catalogue , and so map a larger volume of the universe .
lrgs have a low space density compared to @xmath16 galaxies , which means that fewer redshifts have to be measured to map out the same volume . the low space density , which translates into a higher shot noise , is compensated for by the stronger than average clustering of lrgs , which maintains the signal - to - noise of the correlation function at a level which can be measured .
here we use the measurement of the 2-point correlation function of lrgs in redshift space made by @xcite shown by the black points in fig .
[ fig : corfunc ] .
these authors studied the clustering of lrgs in data release 6 ( dr6 ) of the sloan digital sky survey ( sdss ) , which has 75000 lrg galaxies spanning a volume of @xmath17 over the redshift interval @xmath18 . the comparison between dr6 and the result of @xcite
is plotted in fig.b4 of @xcite and fig .
[ fig : corfunc ] ; the two results are in good agreement with the dr6 result showing an improvement of a factor of about @xmath19 in the size of the errors .
@xcite carried out an extensive series of tests of their measurement of the lrg correlation function .
they found that systematic uncertainties in the estimation of the radial selection have a small impact on the results ( ie see their figure b2 ) .
the adoption of different weighting schemes also produces negligibly small changes ( their fig.b5 and b11 ) .
these careful tests indicate that this new estimation of the correlation function is robust and that any possible systematic effects are small compared to the error bars . here
we recap two aspects of the analysis of @xcite which are particularly relevant to this paper : the treatment of the survey mask and the estimation of the covariance matrix of the correlation function .
an accurate knowledge of the angular and radial selection of a galaxy catalogue , including the redshift completeness as a function of magnitude and position on the sky , is an essential prerequisite for a measurement of clustering .
this information allows the mean density of galaxies in the survey to be estimated . after the release of dr6
, @xcite provided this information in a readily usable form , translating the original mask files extracted from the nyu value - added galaxy catalog @xcite , from mangle into healpix format @xcite .
@xcite describe how they constructed a survey mask " for lrgs and tested the impact of the mask on clustering measurements using mock catalogues .
using the same techniques , we have also examined the correlation function of lrgs in dr7 , which has become available since the submission of @xcite . in fig .
[ fig : corfunc ] , we plot a summary of the possible systematics in the estimation of the correlation function .
the estimates corresponding to different versions of the survey mask and different ways of generating a catalogue of random points are in remarkably good agreement with one another .
furthermore , the results from dr6 and dr7 are in excellent agreement ; dr7 represents only a modest improvement over dr6 .
the results of @xcite are plotted in fig .
[ fig : corfunc ] using triangles .
their estimate is consistent with that of @xcite within the errors ( note that their binning is different , and also , perhaps , their normalisation , due to a small shift which could be attributed to systematics ( see section [ sec : kshift ] ) ) . whilst there is a modest improvement in the parameter constraints on using the dr7 measurement , in line with the incremental change in the number of lrgs and the solid angle covered , we have decided to retain the dr6 measurement in this paper , because the survey mask has been tested more extensively in that case .
the dr6 measurement represents a factor of @xmath20 more volume than that covered by the lrg sample analyzed by eisenstein et al .
( 2005 ) .
@xcite constructed a covariance matrix for the lrg correlation function using mock catalogues drawn from the marenostrum institut de ciencias del espacio ( mice ) n - body simulations @xcite .
the extremely large volume of this run ( box size @xmath21 ) allowed 216 essentially independent lrg dr6 mocks to be extracted .
@xcite investigated different methods to estimate the error on the correlation function ( see their appendix a ) .
they found that the jackknife error , an internal estimate made from the dataset itself ( jk ; see norberg et al .
2008 ) , gave a reasonable match to the diagonal elements of the covariance matrix obtained directly from the 216 mock catalogues .
however , the jk estimate of the off - diagonal elements of the covariance matrix is noisier than that obtained from the mocks . in this analysis
we construct the full covariance matrix of the measurement from the correlation matrix estimated from mock catalogues of dark matter haloes with similar clustering and abundance to the lrgs , rescaled by the jk estimate of the variance from the data , which has the advantage of being independent of the cosmological model adopted in the n - body simulation .
small systematic effects including , for example , the integral constraint , calibration errors or evolutionary effects , will , if unaccounted for , appear as an additive term in the measured correlation function : @xmath22 , where @xmath23 stands for the systematic error .
the simplest model for @xmath23 is a constant shift , @xmath24 , which we label here as the _ k - shift_. systematic shifts include both effects which are unaccounted for in the estimate of the radial selection function and angular calibration errors which can move galaxies in or out of the galaxy sample .
these can introduce spurious fluctuations in the observed density . a 1% variation in @xmath25 color or a 3% shift in @xmath26 magnitude can introduce a 10% modulation in the lrg target number density @xcite .
however , @xcite found that the large scale density variations in the final lrg sample are completely consistent with the predictions of biased @xmath0cdm models , showing that when averaged over large angular scales , the fluctuations in the photometric calibration of the sdss do not affect significantly the uniformity of the lrg sample .
even a spurious number density fluctuation as large as @xmath27 can only produce a shift @xmath28 . as illustrated in fig.[fig : corfunc ] the potential systematics that we are able to check seem to produce shifts of @xmath29 .
these are small changes , but could our analysis below be affected by such a systematic effect , if present ?
we have checked this explicitly by repeating our analysis allowing for a constant additive term @xmath30 as an extra free parameter , which we marginalize over .
we imposed a prior on this k - shift as large as @xmath31 .
this corresponds to 10% density fluctuations , well above the expected systematics in the lrg sample .
we find that even with this wide prior in @xmath30 , the marginalization over the additive k - shift does not change the obtained errors of the cosmological parameters and only changes the mean values by less than 0.5@xmath14 in the most extreme cases .
this small difference does not justify the inclusion of this extra parameter .
but note that if we want to address the question of what is the _ absolute _ goodness of fit , then it might be important to consider such a k - shift as an additional degree of freedom .
we will come back to this question in sec .
[ sec : conclusions ] .
the main piece of clustering information we shall use in this paper is the monopole or spherical average of the two - point redshift space correlation function described in the previous subsection .
however , there is also useful cosmological information in the form of the correlation function split into bins of pair separation parallel ( @xmath32 ) and perpendicular ( @xmath14 ) to the line of sight , @xmath33 .
@xcite found a significant detection of a peak along the line - of - sight direction , the position and shape of which is consistent with it being the baryonic acoustic peak .
@xcite measured the position of the radial bao peak ( rbao ) using the full lrg dr6 sample , and two sub - samples at low , @xmath34 , and high redshifts , @xmath35 ( see their table ii ) . in this paper
we make use of these two last measurements , which can be treated as independent due to the large separation between these redshift intervals .
when combined with a measurement of the sound horizon scale from the cmb , the position of the bao feature can be used as a standard ruler to obtain a constraint on the value of the hubble parameter , @xmath36 , as a function of redshift @xcite . here ,
instead of calibrating the radial bao distance @xmath37 with the cmb , we follow the same approach as @xcite and use the measurement of the dimensionless redshift scale @xmath38 in this equation @xmath39 is the hubble constant at the redshift of the measurement in the fiducial cosmology assumed by @xcite to convert the observed galaxy redshifts into comoving distances .
note that the factor @xmath40 in eq .
( [ eq : dimensionless ] ) ensures that @xmath41 does not depend on the fiducial cosmology assumed to obtain the rbao measurement .
section [ ssec : rbao_method ] describes the details of our modelling of these measurements . because of the narrow range of @xmath14 values ( @xmath42 ) used in the radial measurement ,
the rbao results are essentially independent from the monopole of the two - point function ; fewer than 1% of the bins considered in the spherically averaged correlation function are used in the estimate of the radial correlation . for this reason we treat these two datasets as independent .
we tested this explicitly using the mock lrg catalogues from @xcite ( the same set used to estimate the covarince matrix in the monopole ) , to evaluate the covariance of the radial and the monopole correlations .
[ fig : covrbao ] shows the normalized covariance : @xmath43}{\sigma_{\rm m}(i ) } \frac{[\xi_{\rm r}(j)^{k}-\widehat\xi_{\rm r}(j)]}{\sigma_{\rm r}(j)}\ ] ] where @xmath44 and @xmath45 are the monopole and radial correlation in the @xmath46-th and @xmath47-th bins respectivelly measured in the @xmath48-th mock catalogue @xmath49 and @xmath50 and @xmath14 are the corresponding mean and rms fluctuations over the @xmath51 realizations . as shown in fig.[fig : covrbao ] the covariance on the scales of interest in our analysis ( larger than @xmath52 ) is quite small ( less than the noise in the estimation , of about 10% ) which shows that the two datasets are indeed independent in practice .
this implies that the rbao measurements obtained from the radial correlation functions are independent from the monopole @xmath15 . ) . ]
the accuracy of recent observations of the cmb mean that this is the single most powerful dataset for constraining the values of cosmological parameters .
the compilation of cmb measurements we use includes the temperature power spectrum in the range @xmath53 and the temperature - polarization power spectrum for @xmath54 of the first five years of observations of the wmap satellite @xcite ; the bandpower temperature spectrum from the 2008 observations of the arcminute cosmology bolometer array receiver ( * ? ? ?
* ; * ? ? ?
* acbar , ) over the spherical harmonic range @xmath55 ; temperature and polarization data from the cosmic background imager ( * ? ? ?
* cbi , ) with @xmath56 ; observations from the 2003 flight of boomerang @xcite in the range @xmath57 and the recent results for the temperature and polarization power spectra measurements from quad @xcite over the range @xmath58 .
these measurements of the power spectrum of temperature fluctuations in the cmb cover the spherical harmonic range @xmath59 .
following @xcite , in order to avoid cross - correlations with the wmap data , we use only the bandpowers of the small scale cmb experiments that do not overlap with the signal - dominated wmap data .
note that the quad measurements were not available at the time of the `` wmap+cmb '' analysis carried out by komatsu et al .
( 2008 ) .
we also consider the constraints provided by the hubble diagram of type ia supernovae ( sn ) as provided by the union sample from @xcite .
this compilation is drawn from 13 independent datasets processed using the salt light curve fitter @xcite and analysed in a uniform way .
the sample contains a set of 57 low - redshift sn , the recent samples from the supernova legacy survey ( * ? ? ?
* snls , ) and the equation of state supernovae trace cosmic expansion ( essence , * ? ? ?
* ) , the high - redshift sample from the hubble space telescope @xcite , as well older datasets @xcite . the final sample is the largest available to date , comprising 307 sn which pass the selection criteria .
@xcite suggest a way to include the effect of systematic errors when fitting the sn data .
this estimation and those from other authors @xcite , predict different systematic errors in the estimated constraints on the dark energy equation of state , with values ranging from 5% to more than 10% .
these approaches differ in the choice of which potential sources of systematic errors are taken into account and the estimation of their likely magnitude .
it is our understanding that the community has not reached a consensus about the correct way to estimate the effect of the systematic errors in the analysis of the sn data .
for this reason we follow other authors ( e.g. * ? ? ? * ) and do not include the systematic errors in our constraints on cosmological parameters
. however , the inclusion of the systematic errors in the sn data has important implications for the derived values of the dark energy equation of state ( see section [ ssec : wde ] ) .
this should be borne in mind when comparing constraints obtained using sn data with those obtained using other datasets .
in this section we summarize the approach used to obtain constraints on cosmological parameters .
we start by describing he different parameter sets that we consider . the parametric model we use to describe the shape of the correlation function in redshift space
is presented [ ssec : rpt ] .
the theoretical motivation for this parametric form can be found in appendix [ ssec : shape ] . in section [ ssec :
rpt ] , we also compare the model with measurements made using numerical simulations . section [ ssec : rbao_method ] describes the model we implement to describe the radial bao measurements . the methodology we follow to explore and constrain the parameter spaces is discussed in section [ ssec : mechanics ] . in this paper
we make the basic assumption that the primordial density fluctuations were adiabatic , gaussian and had a power - law spectrum of fourier amplitudes , with a negligible contribution from tensor modes . from the analysis of the fifth year of wmap data
, @xcite did not detect any deviation from these hypotheses at the 99% confidence limit ( cl ) . within this framework
, a cosmological model can be defined by specifying the values of the following eight parameters : @xmath60 we now go through the parameters in the above list , defining each one and also explaining how the values of other parameters are obtained , which we will refer to as derived parameters .
the homogeneous background cosmology is described through the various contributions to the mass - energy density .
these are , in units of the critical density : @xmath61 , which describes the curvature of the universe ; @xmath62 , the density of the dark matter ( assumed cold , where @xmath63 is hubble s constant in units of @xmath64 ) and @xmath65 , the baryon density .
we assume that massive neutrinos make no contribution to the mass budget . for most of this paper ,
we assume that the dark energy component has a constant equation of state independent of redshift , with the ratio of pressure to density given by @xmath12 . in section [ ssec :
wa ] we relax this assumption and analyse models allowing for a time variation in this parameter . in this case
we use the standard linear parametrization given by @xcite @xmath66 where @xmath67 is the expansion factor and @xmath68 and @xmath69 are parameters .
the form of the initial fluctuations is described by two quantities ; the scalar spectral index , @xmath70 and the primordial amplitude of the scalar fluctuations @xmath71 .
these parameter values are quoted at the `` pivot '' scale wavenumber of @xmath72 .
we assume that the reionization of the neutral intergalactic medium occurred instantaneously , with an optical depth given by @xmath73 .
finally , @xmath74 gives the ratio of the sound horizon scale at the epoch of decoupling to the angular diameter distance to the corresponding redshift .
there are further basic quantities whose values can be derived from those listed in the set of eq .
( [ eq : param ] ) : @xmath75 the matter density parameter is given by @xmath76 .
the value of the hubble constant is derived from @xmath77 .
the energy - density of the dark energy is set by @xmath78 .
the results for @xmath71 can be translated into a constraint on @xmath79 , the _ rms _ linear perturbation theory variance in spheres of radius @xmath80mpc , using the matter fluctuation transfer function .
the redshift of reionization , @xmath81 , can be computed from the values of @xmath73 , the hubble constant and the matter and baryon densities @xcite .
the age of the universe is @xmath82 .
the @xmath0cdm cosmological model is the simplest model which can account for the wide variety of cosmological observations available today .
this model is characterized by six parameters : @xmath83 assuming @xmath84 and @xmath85 .
this parameter space is well constrained by the temperature and polarization power spectrum measurements from five years of integration of the wmap satellite @xcite . in section [ ssec : lcdm ]
we analyse the impact of the measurement of the shape of @xmath15 on the constraints in this parameter space . using the latest wmap data , @xcite placed strong constraints on the possible deviations from the @xmath0cdm model , namely non - gaussianity , the presence of isocurvature modes , deviations from a pure power law scalar primordial power spectrum , the presence of tensor modes , a non - negligible energy component in the form of massive neutrinos and parity - violation interactions .
however , there are two parameters that signal important deviations of the standard @xmath0cdm model that can not be tightly constrained from cmb data alone : the curvature of the universe @xmath61 and the dark energy equation of state @xmath12 . in order to assess the improvement on the constraints
once the information on the shape of @xmath15 is included in the analysis , we explore four parameter spaces which contain extensions of the simple @xmath0cdm set .
first we extend the parameter set of eq .
( [ eq : param6 ] ) by adding a constant dark energy equation of state @xmath86 fixing @xmath87 .
the results obtained in this case are shown in section [ ssec : wde ] . in section [ ssec : wa ] we include the parametrization of eq .
( [ eq : wa ] ) in our analysis and explore the extended parameter space @xmath88 where we also implement the hypothesis of a flat universe with @xmath87 .
we also analyse the effect of dropping this assumption .
first , in section [ ssec : omk ] , we include @xmath13 as a free parameter , with @xmath89 assuming that the dark energy is given by a cosmological constant ( or vacuum energy ) with @xmath85 .
finally , in section [ ssec : wok ] , we also allow this parameter to vary freely and we explore the full parameter space of eq .
( [ eq : param ] ) . in the linear perturbation theory regime
( valid when the fluctuation amplitude is small , for instance at high redshift or on very large scales ) , the shape of the matter correlation function is well understood and can be readily obtained using linear boltzmann solvers , such as camb @xcite .
the shape of the correlation function , in the context of a standard adiabatic cdm model , is sensitive to the values of the matter density , @xmath90 , the baryon density @xmath91 , the spectral tilt @xmath70 and the density parameter of massive neutrinos . the evolution with redshift of the correlation function is well understood in the linear regime .
in this case , each fourier mode of the density field evolves independently of the others and the shape of @xmath92 is unaltered , although the overall amplitude changes with time .
unfortunately this simple behaviour is modified by a number of nonlinear phenomena , which affect different scales at different epochs .
these include nonlinear effects generated by the latter stages of gravitational instability , redshift - space distortions caused by gravitationally induced motions and a possible non - trivial scale dependent bias relation between the distribution of galaxies and the underlying dark matter field ( see , for example , the step - by - step illustration of these effects given by * ? ? ?
nonlinear growth results in cross - talk between different fourier modes and introduces scale dependent patterns in the clustering , even on large scales .
this is particularly noticeable in the appearance of the bao bump , which is sensitive to the match between the amplitude and phases of the fluctuations around the peak scale .
due to the distortion of the fourier modes from their original values , nonlinear growth causes the bao bump to become smeared out and also to lose contrast .
such effects can not be ignored when modelling low - redshift data . ) .
the error bars correspond to the error on the mean of the ensemble and are obtained from the scatter among the @xmath93 realizations . the best fitting parametric model used in this work , eq . ( [ eq : xi_model ] ) , is shown by the solid blue line ( note the fit takes into account the covariance between the bins ) .
the dashed line corresponds to setting @xmath94 and highlights the importance of this term in matching the shape of the correlation function at @xmath95 .
the scaling uses @xmath96mpc . ] on large scales , ( @xmath97 ) , the correlation function falls sharply with increasing comoving pair separation , scaling roughly as a power law @xmath98 , with @xmath99 . at even larger separations ,
this behaviour is altered by the emergence of a bump known as the bao peak ( e.g. see fig .
[ fig : modelrzspace ] ) .
this feature , first measured for lrgs by @xcite , is centered at about @xmath100 and has a width of @xmath101 .
the amplitude of the peak in the lrg correlation function corresponds to a @xmath102 excess in the number of lrg pairs above the number expected in a random distribution ( e.g. * ? ? ?
* ) . to model the shape of the correlation function on large scales , we follow @xcite and @xcite and adopt the following parametrization : @xmath103 where @xmath104 , @xmath105 and @xmath106 are nuisance parameters , and the symbol @xmath107 denotes a convolution .
here @xmath108 is the derivative of the linear correlation function and @xmath109 is defined by the integral : @xmath110 with @xmath111 denoting the spherical bessel function of order one . the model in eq .
[ eq : xi_model ] is primarily motivated by rpt , where the matter power spectrum is written as @xmath112 , with @xmath113 a _ nonlinear _ growth factor and @xmath114 being the power generated by mode - coupling . to a very good approximation
@xmath113 is of gaussian form , while at large scales the leading order contribution of @xmath114 in real space is @xmath115 ( see appendix [ ssec : shape ] for a detailed discussion ) .
the gaussian degradation of the bao information was also shown to be a good description by @xcite .
we now demonstrate how accurately the model of eq .
( [ eq : xi_model ] ) can reproduce the spatial clustering of halo samples with comoving number densities similar to that of the lrgs in dr6 . to this end
we utilize an ensemble of @xmath93 realizations of collisionless dark matter n - body simulations .
each simulation contains @xmath116 particles in a comoving volume @xmath117 .
the cosmological parameters were set to @xmath118 , @xmath119 , @xmath120 and @xmath121 .
the initial power spectrum had spectral index @xmath122 and was normalized to give @xmath123 when linearly extrapolated to @xmath124 .
halos were identified using the friends - of - friends algorithm with linking - length parameter @xmath125 ( see * ? ? ?
* ; * ? ? ?
* for more details ) .
[ fig : modelrzspace ] shows the 2-point correlation function measured in redshift - space at @xmath124 for two non - overlapping samples of dark matter haloes of masses @xmath126<15\times 10^{13}$ ] and @xmath127 $ ] ( with number densities @xmath128=1.88\times 10^{-5}$ ] and @xmath129 respectively ) .
our redshift space measurements were done considering the contribution of peculiar velocities along one dimension in the simulation ( i.e. recreating a plane - parallel configuration ) .
@xcite discussed in depth the issue of wide angle redshift space distortions .
he found that opening angles @xmath130 would ensure the validity of the plane - parallel approximation . for the redshift range of the lrg sample considered in our analysis ( @xmath131 ) and the largest pair - distance separation allowed ( @xmath132 ) ,
the maximum opening angle is given by @xmath133 , which implies @xmath134 .
for the mean redshift of the sample the angle subtended by @xmath135 is @xmath136 , well within the validity of the plane - parallel approximation .
the error bars plotted show the error on the mean obtained from the ensemble of 50 simulations .
each simulation has a larger volume than the lrg sample we use from dr6 . therefore these errors are @xmath137 times smaller than they would be for the lrg sample .
this means that the deviations between models ( solid lines ) and simulation results ( points ) are not important for our purposes .
the parametrization given in eq .
( [ eq : xi_model ] ) corresponds to the solid blue line in fig .
[ fig : modelrzspace ] .
the corresponding best - fit @xmath138 values for the nuisance parameters @xmath139,b , a_{\rm mc})$ ] were @xmath140 and @xmath141 respectively .
clearly , the model can accurately describe the clustering of halos in redshift space .
@xcite also compared eq .
( [ eq : xi_model ] ) against measurements of the non - linear correlation function from a similar large ensemble of n - body simulations at various redshifts , and confirmed that this form gives an essentially unbiased measurement of the dark energy equation of state using both , real and redshift space information .
the dashed line in fig . [ fig : modelrzspace ] corresponds to eq .
( [ eq : xi_model ] ) with @xmath94 . from the plot
we can see that the inclusion of this term is the key to recovering the correct clustering shape at separations smaller than the bao bump .
in addition it contributes slightly to the shape of the bump and alleviates a systematic effect related to the position of the bao peak @xcite ( but this is sub - dominant for the survey volume being considered ) . in subsequent sections we check that this nuisance parameter is not degenerate with any of the cosmological parameters . on theoretical grounds ,
one expects the smoothing length @xmath142 to depend on cosmology ( i.e. aside from galaxy type or redshift ) with , for example , a @xmath143 increase in @xmath144 expected to increase @xmath105 by about @xmath145 @xcite . in view of this
, we decided to consider @xmath105 as a nuisance parameter at the expense of a possible increase in error bars .
note that this is at variance with the approach of @xcite , @xcite and @xcite who kept this length fixed . in summary
, for each cosmological model we compute the linear correlation function @xmath146 using camb to generate the corresponding transfer function , and simulate nonlinear effects through eq .
( [ eq : xi_model ] ) after computing @xmath147 and @xmath148 from eq .
( [ eq : xi1 ] ) . the value of @xmath41 can be computed as , @xmath149 where @xmath36 is the hubble constant at the mean redshift of the measurements ( @xmath150 and 0.43 ) , and @xmath151 is the comoving sound horizon at the drag epoch , which is given by @xmath152 where @xmath153 and @xmath154 for a cmb temperature @xmath155 .
the value of @xmath156 can be computed with high accuracy from the values of @xmath157 and @xmath158 using the fitting formulae of @xcite .
the radial bao scale measurements can be used to place constraints on cosmological parameters independently of other datasets .
being a purely geometrical test , as in the case of the sn data , the radial bao measurements are not sensitive to all the cosmological parameters of eq .
( [ eq : param ] ) since they contain no information about the primordial spectrum of density fluctuations , that is @xmath71 and @xmath70 , and of the optical depth to the last scattering surface @xmath73 . | we apply a new model for the spherically averaged correlation function at large pair separations to the measurement of the clustering of luminous red galaxies ( lrgs ) made from the sdss by @xcite .
our model takes into account the form of the bao peak and the large scale shape of the correlation function .
we perform a monte carlo markov chain analysis for different combinations of datasets and for different parameter sets . when used in combination with a compilation of the latest cmb measurements , the lrg clustering and the latest supernovae results give constraints on cosmological parameters which are comparable and in remarkably good agreement , resolving the tension reported in some studies . the best fitting model in the context of a flat
, @xmath0-cdm cosmology is specified by @xmath1 , @xmath2 , @xmath3 , @xmath4 and @xmath5 . if we allow the time - independent dark energy equation of state parameter to vary , we find results consistent with a cosmological constant at the @xmath6 level using all data sets : @xmath7 .
the large scale structure measurements by themselves can constrain the dark energy equation of state parameter to @xmath8 , independently of cmb or supernovae data .
we do not find convincing evidence for an evolving equation of state .
we provide a set of `` extended distance priors '' that contain the most relevant information from the cmb power spectrum and the shape of the lrg correlation function which can be used to constrain dark energy models and spatial curvature .
our model should provide an accurate description of the clustering even in much larger , forthcoming surveys , such as those planned with nasa s jdem or esa s euclid mission .
cosmological parameters , large scale structure of the universe |
the flow of two immiscible fluids through porous media arises in many important industrial and natural situations such as secondary oil recovery , ground water remediation , and geological @xmath0 storage .
such flows are known to be potentially unstable , especially when the displacing fluid is more viscous than the displaced one . there
exist some similarities between porous media and hele - shaw flows ( i.e. flow in a hele - shaw cell , see below ) ; for example the pressure drop in both such flows are governed by darcy s law for single fluid flow . due to this and the fact that it is significantly easier to study hele - shaw flows theoretically , numerically , and experimentally , there have been numerous theoretical and numerical studies even for hele - shaw flow of two immiscible fluids since the early 1950s , starting with the work of saffman and taylor @xcite .
there are many review articles on such studies , for example see @xcite .
these studies were originally motivated by displacement processes arising in secondary oil recovery , even though these studies have much wider appeal in the sciences and engineering . in the late 1970s , tertiary displacement processes involved in chemical
enhanced oil recovery generated interest in three - layer and multi - layer hele - shaw flows ( see @xcite ) . in this paper , we first briefly derive the non - standard eigenvalue problem . this eigenvalue problem has been derived earlier by the first author and his collaborators ; for example see @xcite .
but the difference is that the derivation presented here is more general and shows how to generate higher order correction terms if necessary in order to study the effect of nonlinear terms that may dominate the dynamics , particularly in view of the sensitivity of fingering problems to finite amplitude perturbations .
however , we do not study or discuss such nonlinear effects in this paper which will be taken up in the future as it falls outside the scope of this paper .
we then analytically study this non - standard eigenvalue problem using non - linear transformation for the case when the viscous profile of the middle layer is linear .
we will see below that this case is relatively hard to study in comparison to the case when the viscous profile is exponential which we have recently addressed in @xcite . the physical set - up consists of rectilinear motion of three immiscible fluids in a hele - shaw cell which is a device separating two parallel plates by a distance @xmath1 ( see fig . [
fig : hs_rect_linearprofile ] ) .
the fluid in the extreme left layer @xmath2 with viscosity @xmath3 extends up to @xmath4 , the fluid in the extreme right layer @xmath5 with viscosity @xmath6 extends up to @xmath7 , and the fluid in the middle - layer @xmath8 of finite length @xmath9 has a smooth viscous profile with viscosity increasing in the direction of displacement .
the interfacial tensions of the leading and the trailing interfaces are given by @xmath10 and @xmath11 respectively .
it is well established that this hele - shaw flow is similar to flow in homogeneous porous media with equivalent permeability @xmath12 . without any loss of generality ,
we take this to be one below .
the mathematical model considered here consists of conservation of mass , darcy s law and advection equation for viscosity .
thus we have @xmath13 due to the continuity equation , we can define the stream function @xmath14 such that @xmath15 and @xmath16 .
this then implies that @xmath17 since @xmath18 when @xmath19 , we consider a small perturbation of the basic scalar fields @xmath20 and @xmath21 of the form @xmath22 substituting into the original equations , we get the following @xmath23 and @xmath24 equations.@xmath23 equations : @xmath25 these equations provide the basic solution given by @xmath26 where @xmath27 is an arbitrary function of @xmath28 , meaning the viscous profile is fixed with respect to a moving frame moving at a constant velocity @xmath29.@xmath24 equations : @xmath30 now , introducing the moving frame change of variables , namely @xmath31 , we get the following system of equations .
@xmath32 taking cross derivatives of the first two equations with respect to @xmath33 and @xmath34 respectively and then subtracting the resulting equations from each other gives @xmath35 this combined with the equation @xmath36 leads to @xmath37 using the ansatz @xmath38 in the above equation together with the appropriate boundary conditions ( see @xcite ) give the following eigenvalue problem . @xmath39
+ \frac{sk^4}\sigma\right\}\\ & \mu^-_0(0)f_\xi(0 ) = f(0 ) \left\{-\mu_2k + \frac{u_0k^2}\sigma [ \mu_2-\mu^-_0(0 ) ] - \frac{tk^4}\sigma\right\}\end{aligned}\ ] ] where the viscous profile of the middle layer , namely , @xmath40 is an arbitrary function .
this is a non - standard eigenvalue problem in that the spectral parameter @xmath41 appe ars in the equation as well as in the boundary conditions .
recently , this problem has been numerically solved by daripa @xcite for a constant viscous profile and by daripa & ding @xcite for non - constant viscous profiles @xmath42 to determine the most optimal profile , i.e. , the least unstable profile .
this problem has been too difficult to solve analytically for non - constant profiles .
progress made in this direction for the linear viscous profile is presented below . in this paper
we consider a linear viscous profile for the intermediate fluid region given by @xmath43 where @xmath44 and @xmath45 are jump discontinuity values at the interfaces @xmath46 and @xmath47 , respectively . in the left region
the problem reduces to @xmath48 which has solution @xmath49 . in the right region
the problem reduces to @xmath50 which has solution @xmath51 in the intermediate region the problem reduces to @xmath52 where @xmath53 is the spectral parameter and @xmath54
we introduce the nonlinear transformation and change of variables given by @xmath55 after some manipulation of the eigenvalue problem using the above transformation , we obtain the following eigenvalue problem for the kummer s equation @xmath56 .
@xmath57 where a prime denotes derivative , @xmath58 the eigenvalue problem is a regular two point boundary value problem for each wave number @xmath59 .
one solution @xmath60 of the kummer s equation is given by @xmath61 where @xmath62 and @xmath63.this is an analytic solution .
it is easily seen that the derivative of this solution which we will need below for the dispersion relation is given by @xmath64 the linearly independent second solution is easily constructed by the method of frobenius .
avoiding all the details , the second solution @xmath65 is given by @xmath66 where @xmath67 .
its derivative which we will need below is then given by @xmath68 the general solution of the kummer s equation is then given by @xmath69 where @xmath70 and @xmath71 are arbitrary constants .
substituting the general solution into the two boundary conditions of the eigenvalue problem , we obtain the following linear system of equations for @xmath70 and @xmath71 .
@xmath72 therefore , for a non - trivial solution we have @xmath73 this formally gives the dispersion relation @xmath74 in terms of the problem data : @xmath75 and @xmath76 . in terms of the original variables
@xmath77 and @xmath34 , the fundamental solutions @xmath78 and @xmath79 are then given by ( see ) @xmath80 where @xmath81 also , recall that @xmath28 , @xmath82 and @xmath83 .
noticing that both series are centered at @xmath84 and applying the ratio test for series for @xmath78 above , we have @xmath85 for any @xmath34 fixed .
thus the series for @xmath78 converges absolutely @xmath86 , but the series _ evaluated _ at @xmath84 reduces to 1 .
this implies that the radius of this series is @xmath87 .
hence @xmath88 and since it is an alternating singular series , the error @xmath89 in approximating @xmath90 by terms up to @xmath91 is smaller than the last neglected term , namely @xmath92 for a fixed @xmath34 .
similarly , applying the ratio test to the series for @xmath79 , @xmath93 for any @xmath34 .
thus the series inside the brackets converges absolutely @xmath94 , but the series inside the brackets evaluated at @xmath95 reduces to 1 .
therefore , the radius of convergence of the series within the brackets is @xmath87 .
hence @xmath96 notice that @xmath79 has a branch point at @xmath97 . in any case , since the series within the brackets is an alternating sign series , if we truncate it , the error @xmath98 is smaller than the last neglected term , i.e. , @xmath99 the general solution of the ode @xmath100 is then given by @xmath101 .
the boundary values of @xmath102 follow from and whi ch are now given by @xmath103 substituting these in the boundary conditions @xmath104 , we obtain the following system of equations for the constants @xmath70 and @xmath71 .
@xmath105 where @xmath106 and @xmath107 . for the existence of nontrivial solutions
, we then have @xmath108 which gives us the dispersion relation in the form : @xmath109 . because of the nature of the series solutions given above , it is not possible to give this dispersion relation explicitly .
there are an infinite number of eigenvalues @xmath110 ( recall @xmath111 ) which can be ordered : @xmath112 . we know that these infinite number of eigenvalues should reduce to ( i ) only two in the limit @xmath113 corresponding to the constant viscosity of the intermediate layer fluid ( see daripa @xcite ) ; ( ii ) only one in the limit @xmath114 ( see saffman & taylor @xcite , daripa @xcite ) and ( iii ) only two in the limit of @xmath115 ( see daripa @xcite ) .
in fact , we also know the eigenvalues in these limiting cases from the pure saffman - taylor growth rate of individual interfaces . these results by no means are transparent from the solutions of the eigenvalue problem given in the previous section .
below , we show how to recover these limit solutions ( eigenvalues ) from the infinite number of eigenvalues for the linear viscous profile . in this case , the eigenvalue problem reduces to @xmath117 in this case , the change of variable introduced previously , namely @xmath118 , which converts the equation @xmath100 to kummer s equation , i s not well - defined .
therefore we work with the boundary value problem @xmath119 .
now , consider the general solution of the ode in ( [ slp3 ] ) @xmath120 such that @xmath121 therefore , we get @xmath122 we search for a solution of the boundary value problem ( [ slp3 ] ) of the form @xmath123 to find a solution of ( [ slp3 ] ) of this form , we start determining the coefficients using the shooting technique such that the boundary condition at @xmath46 is satisfied . obviously , the coefficients @xmath124 and @xmath125 depend on the parameter @xmath126 .
then , we find @xmath126 in such a way that the solution satisfies the boundary condition at @xmath47 .
hence , we look for @xmath124 and @xmath125 such that @xmath127 then it follows directly from ( [ cond1 ] ) that @xmath128 and @xmath129 therefore @xmath130 satisfies the ode in and the boundary condition at @xmath46 . from these
it follows that the spectrum of problem ( [ slp3 ] ) can be studied using the following algebraic equation ( see @xmath131 ) @xmath132 where @xmath133 is the function defined in .
evaluating @xmath134 and @xmath135 from and substituting directly in one obtains @xmath136 then , taking @xmath137 one obtains @xmath138 which is equivalent to the equation @xmath139 since @xmath140 ( @xmath141 or @xmath142 ) , it follows that @xmath143 now , using the definition of the coefficient @xmath144 given in ( [ coef ] ) we have @xmath145 \lambda =- k(\mu_{1}+\mu_{2})\ ] ] from which it follows that @xmath146 which is the formula for the growth rate of an interface with surface tension @xmath147 , which is what should be expected in this limit .
thus we recover the classical formula for the growth rate in this limit .
to take the limit when @xmath148 , we go back to equation ( [ explicitall ] ) and write it as follows @xmath149 now , taking the limit when @xmath150 , we obtain @xmath151 using @xmath152 and expressions for the coefficients from @xmath153 , we obtain @xmath154 finally , since @xmath155 ( @xmath141 or @xmath142 ) it follows that @xmath156 which gives the classical formula for saffman - taylor instability of the leading interface . similarly , we can recover the the classical formula for saffman - taylor instability of the trailin g interface by reversing the shooting technique ( see after ) , i.e. , first find the solution which is analogous to but satisfies the boundary condition at @xmath47 instead and then shoot to satisfy the boundary condition at @xmath46 ( i.e. , replace by a similar formula derived from the boundary condition at @xmath46 and follow the procedure ) . in this section , we study asymptotic limits ( @xmath158 and @xmath148 ) of the solutions to the eigenvalue problem . to this end
, we consider the following form of two linearly independent solutions of kummer s equation @xmath159 .
these are convenient for the asymptotic analysis presented below .
@xmath160 where @xmath161 is euler s digamma function ( see abramowitz @xcite , chapter 13 ) . to this end
, we follow the steps presented in the previous section [ subsection : constant - viscosity ] for the particular case @xmath140 ( @xmath141 or @xmath142 ) . from the transformation in
, it follows that @xmath162 is the general solution of the ode @xmath100 where @xmath163 and @xmath164 are chosen such that @xmath165 where @xmath166 . substituting in the boundary conditions , we obtain the following linear systems of equations @xmath167 @xmath168 solving the above two systems and using the relations ( see abramowitz @xcite , chapter 13 ) @xmath169 we obtain @xmath170 where @xmath171 is the determinant of the coefficient matrix of the system .
similar to the procedure of the previous section [ subsection : constant - viscosity ] , we find @xmath124 and @xmath125 so that @xmath172 and @xmath173
. therefore , it follows from and that @xmath174 and therefore @xmath175 and @xmath176 . substituting these constants in the function @xmath133 defined by , we obtain a solution of the ode that satisfies the boundary condition at @xmath46 of the eigenvalue problem .
since @xmath124 and @xmath125 depend on the spectral parameter @xmath126 , it follows that the eigenvalues of the problem can be obtained by studying the following algebraic equation which is a reformulation of the boundary condition at @xmath47 of the eigenvalue problem .
@xmath177 since the right - hand side of the above equation does not depend on @xmath9 , we need to study the asymptotic limits ( @xmath178 and @xmath148 ) of the lefthand side of ( [ al2 ] ) .
notice that the expression @xmath179 above is given by ( see ) @xmath180 therefore , we first find the asymptotic approximations for @xmath181 , and @xmath182 in both cases below before estimating the ratio @xmath183 using for its use in .
below , we write @xmath184 and @xmath185 where @xmath186 and @xmath187 are given by ( see ) , @xmath188 0.1truein ( when @xmath148 ) : it follows from abramowitz and stegun @xcite that @xmath189 using the identities from ( [ relspecialfunction ] ) we obtain @xmath190 using , and the relation @xmath191 in the expression @xmath192 for @xmath193 , we obtain @xmath194 which can be written as @xmath195 where @xmath196 ( see ) . using similar arguments it follows that @xmath197 for @xmath148 ,
see ( [ coef2 ] ) for the dependence of @xmath198 and @xmath9 of the coefficient @xmath199 and @xmath200 .
thus , using the above asymptotic results for the coefficient @xmath199 and @xmath200 and the asymptotic results for the confluent hypergeometric functions given in ( [ asymphypergeometric1 ] ) and ( [ asymphypergeometric2 ] ) , we get @xmath201 substituting this in , we obtain @xmath202 therefore , equation ( [ al2 ] ) becomes @xmath203 . using @xmath152 and expressions for the coefficients from @xmath153 , we obtain @xmath204 which is the classical formula for saffman - taylor instability of the leading interface . similarly , we can also recover the the classical formula for saffman - taylor instability of the trailing interface by reversing the shooting technique as discussed at the end of section [ subsection : constant - viscosity ] .
0.1truein ( when @xmath178 ) : similar to the previous case , we will first need to get asymptotic approximations for @xmath181 , and @xmath205 in this limit .
notice that in this case , singularities of the confluent hypergeometric function of the second kind will arise .
now , we give the following asymptotic results from abramowit z and stegun @xcite @xmath206 where we recall that @xmath207 and @xmath208 are defined by . similar to the calculations of the previous case @xmath148 , we present the dominant terms of the left hand side of ( [ al2 ] ) .
it is worth pointing out that due to , the derivative of the confluent hypergeometric function of the second kind is dominant . from the definition of the coefficients @xmath199 and @xmath200 given in and the asymptotic results presented in , we obtain @xmath209 we remark that @xmath210 and therefore the asymptotic result for @xmath211 follows from the definition of the coefficient @xmath212 and @xmath213 , see ( [ coef2 ] ) . from the forms of @xmath199 and @xmath200 , we get @xmath214 therefore @xmath215 similarly , we obtain @xmath216 using , , and , it follows that @xmath217 from the definition of @xmath207 and @xmath208 given in we obtain @xmath218 and therefore @xmath219,\ ] ] where @xmath175 and @xmath176 .
it then follows that @xmath220\lambda+\alpha_{2}(k).\ ] ] using this in equation ( [ al2 ] ) , we obtain @xmath221\lambda+\alpha_{2}(k)=\beta_{1}(k)\lambda-\beta_{2}(k)\ ] ] which is equivalent to @xmath222 after substituting the values of @xmath223 , @xmath224 , @xmath225 and @xmath226 and simplifying we obtain @xmath227 therefore , @xmath228 which is the formula for the growth rate of an interface with surface tension @xmath147 , which is what should be expected in this limit . thus we recover the classical formula for the growth rate in this limit .
we converted a non - standard eigenvalue problem arising in the linear stability analysis of a three - layer hele - shaw model of enhanced oil recovery to a boundary value problem for kummer s equation when the middle layer has a linear viscous profile .
we presented the general solution in terms of frobenius series and discussed the convergence properties of these series solutions .
we also formally gave the dispersion relation implicitly through the existence criterion for non - trivial solutions . in order to recover the well - known physical solutions for some limiting cases
, we rewrote the general solutions using a different set of fundamental solutions and analyzed these for those limiting cases : ( i ) when the viscous profile of the middle layer approaches a constant viscosity , both in the case of a fixed - length middle layer and also as the length of the middle layer appraoches infinity ; and ( ii ) when the length of the middle layer approaches zero .
we showed that we were thus able to recover the correct physical solutions .
this paper was made possible by an nprp grant # 08 - 777 - 1 - 141 to one of the authors ( prabir daripa ) from the qatar national research fund ( a member of the qatar foundation ) .
the second author ( oscar orellana ) acknowledges financial support through this grant for travel to tamuq , qatar for a two day workshop on international workshop on enhanced oil recovery and porous media flows " organized by the first author ( prabir daripa ) during july 31st and august 1 of 2013 .
the work of the second author ( oscar orellana ) was also supported in part by fondo nacional de desarrollo centifico y technologico ( fondecyt ) under grant 1141260 and universidad tecnica federico santa maria , valparaiso , chile .
the statements made herein are solely the responsibility of the authors .
kummer s equation has the general form @xmath229 where @xmath230 and @xmath231 the two linearly independent solutions are @xmath232 and @xmath233 where the general expression for @xmath232 is given by @xmath234 where @xmath235 the linearly independent second solution @xmath233 is similarly given by a series which can be easily constructed by the method of frobenius . | we present a non - standard eigenvalue problem that arises in the linear stability of a three - layer hele - shaw model of enhanced oil recovery .
a nonlinear transformation is introduced which allows reformulation of the non - standard eigenvalue problem as a boundary value problem for kummer s equation when the viscous profile of the middle layer is linear . using the existing body of works on kummer s equation
, we construct an exact solution of the eigenvalue problem and provide the dispersion relation implicitly through the existence criterion for the non - trivial solution .
we also discuss the convergence of the series solution .
it is shown that this solution reduces to the physically relevant solutions in two asymptotic limits : ( i ) when the linear viscous profile approaches a constant viscous profile ; or ( ii ) when the length of the middle layer approaches zero .
* mathematics subject classification ( 2010 ) : * 76e17 , 34l10 , 34l15 * keywords : * hele - shaw flows , non - standard eigenvalue problem , kummer s equation , linear stability |
deep surveys with and have provided a unique perspective on the cosmic evolution of accreting supermassive black holes ( smbhs ) and high - energy activity from normal galaxies ( e.g. , star - formation processes , evolving stellar binaries , and hot gas cooling ) over significant fractions of cosmic history ( see fig
. 1 and brandt & hasinger 2005 for a review ) . due to limitations originating from point - source confusion for at flux levels below @xmath0@xmath9 , the deepest views of the extragalactic universe have come exclusively from deep surveys with exposures .
the deepest surveys to date are the @xmath02 ms deep field - north (; alexander 2003 ) and @xmath02 ms deep field - south (; luo 2008 ) surveys ( hereafter , the deep fields ; cdfs ) .
these have been supplemented by additional deep surveys ( e.g. , the @xmath0250 ks extended [ ; lehmer 2005 ] and the 0.20.8 ms extended groth strip [ ; laird 2009 ] ) , which cover larger solid angles than the cdfs and provide important information on rarer source populations .
these deep surveys have been chosen to lie at high galactic latitudes in regions of the sky that have relatively low galactic columns ( @xmath1010@xmath11 @xmath12 ) and few local sources in the field , as to gain a relatively unobscured and unbiased perspective of the distant ( @xmath13 ) universe . despite these growing resources , there has not yet been an equivalent survey targeting very high - density regions of the @xmath14 universe where the most massive smbhs and galaxies are thought to be undergoing very rapid growth ( e.g. , kauffmann 1996 ; de lucia 2006 ) . in order to address this limitation and
study the growth of smbhs and galaxies as a function of environment , we have carried out a deep @xmath0400 ks survey ( pi : d. m. alexander ) of the highest density region in the @xmath1 ssa22 protocluster . plane .
each survey has a range of flux limits across its solid angle ; we have generally shown the most sensitive flux limit .
this plot has been adapted from fig . 1 of brandt & hasinger ( 2005 ) to show the part of parameter space most relevant for ssa22 ; see table 1 of brandt & hasinger ( 2005 ) for references and descriptions of many of the surveys plotted here .
the gray curves show contours regions of the flux limit versus solid angle plane where @xmath0100 , 1000 , and 10,000 sources are expected to be detected ( based on the bauer 2004 number counts relation ) .
, width=340 ] the @xmath1 ssa22 protocluster was originally identified by steidel ( 1998 ) , who found a significant overdensity ( @xmath046 times higher surface density than the field ) within an @xmath0@xmath15 arcmin@xmath7 region ( @xmath0@xmath16 comoving mpc@xmath7 at @xmath17 ) through spectroscopic follow - up observations of @xmath18 candidate lyman break galaxies ( lbgs ) .
theoretical modelling indicates that the protocluster will evolve into a rich cluster with a total mass @xmath19@xmath20 at the present day ( e.g. , coma ; see steidel 1998 for details ) . since its discovery
, the protocluster has also been found to contain a factor of @xmath06 overdensity of ly@xmath2 emitters ( laes ; steidel 2000 ; hayashino 2004 ; matsuda 2005 ) and many remarkable bright extended ly@xmath2-emitting blobs ( labs ; steidel 2000 ; matsuda 2004 ) , which are believed to be sites of massive galaxy formation powered by starburst / agn outflows ( e.g. , bower 2004 ; geach 2005 , 2009 ; wilman 2005 ) .
therefore , ssa22 is an ideal field for studying how smbh growth depends on environment in the @xmath21 1 universe .
in addition to the lbg and lae surveys , ssa22 has become a premier multiwavelength survey field .
the region has been imaged from space by in a sparse mosaic of 12 acs pointings using the f814w filter and by over the entire field in four irac bands ( 3.6 , 4.5 , 5.8 , and 8.0@xmath22 m ) and at 24@xmath22 m by mips ( see fig . 2 )
. additionally , the ground - based observations are extensive and include imaging at radio ( 1.4 ghz from the vla ; chapman 2004 ) , submm ( scuba 850@xmath22 m , aztec 1.1 mm , and laboca 870@xmath22 m ; chapman 2001 ; geach 2005 ; tamura 2009 ) , near - infrared ( @xmath23 , @xmath24 , and @xmath25 bands from ukirt and subaru ; lawrence 2007 ; uchimoto 2008 ) , and optical wavelengths ( e.g. , subaru @xmath26 , @xmath27 , @xmath28 , @xmath29 , @xmath30 , and nb497 bands ; hayashino 2004 ) . a variety of spectroscopic campaigns have been conducted ( e.g. , at kitt peak , subaru , keck , and the vlt ; steidel 2003 ; matsuda 2005 ; garilli 2008 ; chapman 2009 , in - preparation ) and others are currently underway . thus far
, we have utilised the and multiwavelength data presented here to conduct two scientific investigations including ( 1 ) lehmer ( 2009 ) , which finds that the growth of galaxies and smbhs in the @xmath1 protocluster environment is enhanced by a factor of over that found in @xmath8 field galaxies ( i.e. , in the cdfs ) , and ( 2 ) geach ( 2009 ) , which uses the data to study the host galaxies of labs and concludes that agn and star - formation activity provide more than enough uv emission to power the extended ly@xmath2 emission .
additional follow - up investigations and scientific studies are planned and will be presented in subsequent papers . in this paper , we present point - source catalogs and data products derived from the @xmath0400 ks data set .
the paper is orgainized as follows : in @xmath312 , we provide details of the observations and data reduction . in @xmath313 , we present point - source catalogs and details of our technical analyses . in @xmath314 ,
we present basic multiwavlength properties of the sources and conduct scientific analyses of sources within the ssa22 protocluster .
finally , in @xmath315 , we analyse the properties of the only obvious extended source in the field .
the observational procedures and data processing were similar in spirit to those presented in lehmer ( 2005 ) and luo ( 2008 ) ; however , we have made wider use of the data analysis software package acis extract version 3.131 ( broos 2002 ) while producing our point - source catalogs ( see @xmath313 for further details ) . throughout this paper , we assume the galactic column density along the line of sight to to be ( e.g. , stark et al .
the coordinates throughout this paper are j2000 .
@xmath32 = 70 , @xmath33 = 0.3 , and @xmath34 = 0.7 are adopted throughout this paper ( e.g. , spergel 2003 ) , which give the age of the universe as 13.5 gyr and imply a @xmath17 look - back time and spatial scale of 11.4 gyr and 7.6 kpc arcsec@xmath35 , respectively . note .
table a1 is presented in its entirety in the electronic version ; an abbreviated version of the table is shown here for guidance as to its form and content .
the full table contains 44 columns of information for the nine sources .
meanings and units for all columns have been summarized in detail in @xmath36 3.2.2 and appendix a. | we present point - source catalogs for a deep @xmath0400 ks exposure of the ssa22 field .
the observations are centred on a @xmath1 protocluster , which is populated by lyman break galaxies ( lbgs ) , ly@xmath2 emitters ( laes ) , and extended ly@xmath2-emitting blobs ( labs ) .
the survey reaches ultimate ( ) sensitivity limits of @xmath05.7 @xmath3 and @xmath03.0 @xmath4 for the and bands , respectively ( corresponding to @xmath5 and at @xmath1 , respectively , for an assumed photon index of @xmath6 ) .
these limits make ssa22 the fourth deepest extragalactic survey yet conducted , and the only one focused on a known high redshift structure . in total , we detect 297 point sources and identify one obvious bright extended source over a @xmath0330 arcmin@xmath7 region .
in addition to our catalogs , we provide all available optical spectroscopic redshifts and near - infrared and mid - infrared photometry available for our sources .
the basic and infrared properties of our sources indicate a variety of source types , although absorbed active galactic nuclei ( agns ) appear to dominate . in total
, we have identified 12 sources ( either via optical spectroscopic redshifts or lae selection ) at that are likely to be associated with the ssa22 protocluster .
these sources have and multiwavelength properties that suggest they are powered by agn with luminosities in the range of .
we have analysed the agn fraction of sources in the protocluster as a function of local lae source density and find suggestive evidence for a correlation between agn fraction and local lae source density ( at the @xmath096 per cent confidence level ) , implying that supermassive black hole growth at @xmath8 is strongest in the highest density regions .
[ firstpage ] cosmology : observations early universe galaxies : active galaxies : clusters : general surveys : general |
deep x - ray surveys , particularly with chandra , have had enormous recent successes in resolving the x - ray background and shedding light on the nature and evolution of accretion power in galaxies ( e.g. mushotzky et al .
2000 ; giacconi et al .
2001 ; barger et al .
2001 , 2002 ; cowie et al .
2002 , 2003 ) .
it should be borne in mind , however , that the number of fields for which both deep chandra data and comprehensive supporting mutliwavelength observations exists are relatively few . aside from the deepest surveys in the hdf - n ( brandt et al .
2001 ; alexander et al .
2003 hereafter a03 ) and cdf - s ( giacconi et al . 2001 , 2002 ; tozzi et al . 2001 ; rosati et al .
2002 ) , only two other fields have been presented with a chandra exposure in excess of 150 ks ( stern et al .
2002 ; wang et al .
2004 ) .
a particularly well studied field in wavebands outside the x - ray is the groth / westphal survey area ( hereafter gws ) .
the original groth strip " survey with hst ( groth et al .
1995 ) was complemented by a single deeper pointing , and further data have been obtained as part of the canada - france redshift survey ( cfrs ; lilly et al .
comprehensive imaging with hst / acs has now also been approved .
the cfrs data have themselves been superceded by the canada - france deep field ( cfdf ) survey ( mccracken et al .
2001 ) and together they provide multi - band imaging over much of the field .
it has also been covered by deep optical imaging , and followup spectropscopy , in a lyman break galaxy survey by steidel et al .
a further large optical spectroscopy effort is being undertaken as part of the deep and deep2 galaxy redshift surveys ( davis et al .
scuba has observed a large area of the groth strip to deep limits for the canada - uk deep submm survey ( eales et al .
2000 ; webb et al . 2003 ) , and there is also coverage by iso ( flores et al .
1999 ) and in the radio ( fomalont et al . 1991 ) . with a large number of ongoing or future programs for galaxy evolution in this field ( e.g. deep2 , spitzer , cfht , hst / acs ) , and current interest in the connection between agn and galaxy formation and evolution , deep x - ray data are clearly of great value .
some x - ray data have been obtained in this field with xmm ( waskett et al .
2003 ; miyaji et al .
2004 ) . here
we report chandra observations which probe @xmath10 times deeper in flux , obtained in a 200 ks exposure with the acis - i ccd camera as prime instrument .
this is currently the third deepest chandra exposure , after the hdf - n and cdf - s .
[ cols="^,^,^,^,^,^,^,^ " , ] | we present the results of a 200 ks chandra observation of part of the groth strip region , using the acis - i instrument .
we present a relatively simple method for the detection of point - sources and calculation of limiting sensitivities , which we argue is at least as sensitive and more self - consistent than previous methods presented in the literature .
158 distinct x - ray sources are included in our point - source catalogue in the acis - i area .
the number counts show a relative dearth of x - ray sources in this region .
for example at a flux limit of @xmath0 erg @xmath1 s@xmath2 around 20 per cent more soft band sources are detected in the hdf - n and almost @xmath3 per cent more in the elais - n1 field , which we have analysed by the same method for comparison .
we find , however , that these differences are consistent with poisson variations at @xmath4 significance , and therefore there is no evidence for cosmic variance based on these number counts alone .
we determine the average spectra of the objects and find a marked difference between the soft - band selected sources , which have @xmath5 typical of unobscured agn , and the hard - band selected sources , which have @xmath6 .
reassuringly , the sample as a whole has a mean spectrum of @xmath7 , the same as the x - ray background .
nonetheless , our results imply that the fraction of sources with significant obscuration is only @xmath8 per cent , much less than predicted by standard agn population synthesis models .
this is confirmed by direct spectral fitting , with only a handful of objects showing evidence for absorption . after accounting for absorption ,
all objects are consistent with mean intrinsic spectrum of @xmath9 , very similar to local seyferts .
the survey area is distinguished by having outstanding multi - waveband coverage .
comparison with these observations and detailed discussion of the x - ray source properties will be presented in future papers .
[ firstpage ] surveys galaxies : active x - rays : diffuse background x - rays : galaxies cosmology : observations |
supernovae and their remnants have long been identified as likely sites of galactic cosmic ray acceleration .
evidence for electron acceleration first came from observations of radio synchrotron radiation .
the observation of nonthermal x - ray emission from sn 1006 by _ asca _ has provided evidence for electron acceleration up to @xmath3 tev ( koyama 1995 ; reynolds 1998 ) .
the proton cosmic ray component is more difficult to observe , but it has long been recognized that pion decays from collisions with interstellar gas could give an observable flux of @xmath4 gev photons .
initial estimates of @xmath2-ray emission from supernova remnants ( snrs ) , concentrating on the component resulting from pion decays , were made by chevalier ( 1977 ) , blandford & cowie ( 1982 ) , and drury , aharonian , & vlk ( 1994 ) . when @xmath2-ray emission was apparently detected from snrs by _
cgro _ , the @xmath2-ray spectrum could not be fitted by a pure pion decay spectrum and some other component was needed ( esposito 1996 ; sturner 1997 ; gaisser , protheroe , & stanev 1998 ) .
in addition to the pion decays , the relevant processes are bremsstrahlung emission of relativistic electrons and inverse compton emission .
gaisser ( 1998 ) modeled these processes in detail in order to fit the observed @xmath2-ray spectra of the remnants ic 443 and @xmath2 cygni .
they assumed acceleration to a power law spectrum in the shock front and determined the spectral index , electron to proton ratio , and the upper energy cutoff .
the evolution of the relativistic component in a supernova remnant has been modeled by a number of groups .
sturner ( 1997 ) assumed particle acceleration in the shock front with an @xmath5 energy spectrum and followed the evolution of the particle spectrum .
their snr model was homogeneous : a factor 4 density jump at the shock front , constant density in the interior , and a constant magnetic field in the interior .
their shock acceleration model had a diffusion length parameter that was important for the maximum energy reached by the particles .
they modeled the nonthermal emission from ic 443 as synchrotron emission in the radio and bremsstrahlung in @xmath2-rays .
de jager & mastichiadis ( 1997 ) dealt with the same processes in w44 , as well as inverse compton emission .
they noted that the observed radio spectrum is flatter than would be expected from shock acceleration of newly injected particles and suggested that the particles originated from a pulsar in the supernova remnant .
recently , ostrowski ( 1999 ) showed that second - order electron acceleration by the turbulent medium just after the shock could flatten the spectra to account for the radio observations of ic 443 . baring ( 1999 ) presented calculations of the broad - band emission from nonlinear shock models of shell - type snrs .
they used sedov adiabatic shock dynamics in a homogeneous medium and a monte carlo simulation of the particle acceleration taking into account the nonlinear shock structure .
the set of models considered by baring ( 1999 ) covers the range of shock speeds 490 @xmath6 4000 km s@xmath1 and ambient medium number densities 10@xmath7 1 @xmath8 .
massive stars that are the likely progenitors of core collapse supernovae are expected to be spatially correlated with molecular clouds .
the remnants that are likely to be @xmath2-ray sources in _ cgro _ observations ( esposito 1996 ) also show evidence for interaction with molecular gas .
chevalier ( 1999 ) recently studied the evolution of supernova remnants in molecular clouds and concluded that many aspects of the multiwavelength observations could be understood in a model where the remnants evolve in the interclump medium of a molecular cloud , which has density of @xmath9 h atoms @xmath8 , and become radiative at radii @xmath10 pc .
the compression in the radiative shell is limited by the magnetic field .
molecular emission occurs when the radiative shell collides with molecular clumps .
ic 443 is the remnant with the best evidence for high energy emission and also shows clear evidence for interaction of the shock with a molecular cloud ( e.g. , burton 1990 ; van dishoeck 1993 ; cesarsky 1999 ) .
most of the current models of particle acceleration in snrs have dealt with adiabatic remnants in a homogeneous medium ( see , however , the model of boulares & cox 1988 for the cygnus loop and the model of jones & kang 1993 ) .
in contrast , our aims here are to examine in detail the nonthermal emission of radiative supernova remnants in molecular clouds using a kinetic model of electron injection , acceleration and propagation , including the inhomogeneous structure deduced from multiwavelength snr observations .
we use global mhd ( magnetohydrodynamic ) models of supernova remnants interacting with molecular clouds as described by chevalier ( 1999 ) , including the theory of radiative shock structure by shull & mckee ( 1979 ) and the kinetic model of electron injection and acceleration by shocks from bykov & uvarov ( 1999 ) .
the model described here is relevant to the mixed - morphology snrs ( rho & petre 1998 ) , if they are interacting with molecular clouds .
these remnants comprise a sizable fraction of the galactic snr population .
the plan of our paper is as follows . in 2 , we treat the energy spectrum of energetic electrons as they pass through the regions associated with a radiative shock wave .
we include the shock acceleration of particles from the thermal pool . in 3 , we present calculations of the nonthermal emission from the relativistic particles , including emission from shocks in clumps as well as emission from the radiative shock wave in the interclump region .
our model is specifically applied to the well - observed remnant ic 443 , although we expect it to more generally apply to other supernova remnants in molecular clouds .
we also examine the ionization by energetic particles and the energy involved in the relativistic particles .
a discussion of results and future propects is in 4 .
a detailed discussion of expected range of densities , shock velocities and magnetic fields determined from multiwavelength observations of ic 443 , w44 and 3c391 was given in the paper by chevalier ( 1999 , and references therein ) .
the structure of the flow expected in a radiative shock can be divided into the following regions ( see , e.g. , draine & mckee 1993 ) : + _ ( i ) _ preshock gas ionized and heated by _ uv _ ( ultraviolet ) emission and fast particles from the shock ; + _ ( ii ) _ shock transition region ; + _ ( iii ) _ postshock cooling layer ; + _ ( iv ) _ cold shell of swept - up gas ; + _ ( v ) _ hot low density gas interior to the shock .
+ radiative shocks are subject to thermal instabilities ( chevalier & imamura 1982 ; bertschinger 1986 ) as well as dynamical instabilities ( vishniac 1983 ; blondin et al .
although the effect of inhomogeneous magnetic fields on the stability of radiative shock flow is yet to be studied , mhd turbulence is expected in such a system .
the transport and acceleration of nonthermal particles in the violent snr environment are governed by the mhd flow and depend on the ionization structure and on the spectrum of mhd turbulence .
shocks with velocity @xmath11 1.1 generate sufficient ultraviolet _ uv _ radiation for full pre - ionization of hydrogen and helium ( he@xmath12 ) in the preshock region , where the gas temperature is typically @xmath13 k ( shull & mckee 1979 ) . here , @xmath14 is the shock velocity measured in units of @xmath15 , and @xmath16 is the ion number density measured in units of 10 @xmath8 .
shocks with @xmath17 1.5 propagating in a medium with @xmath18 1 photoionize gas ahead of the viscous jump to a depth @xmath19 @xmath20 .
the shock transition in such a case is collisionless and supercritical unless the preshock magnetic field is above @xmath21 4@xmath2210@xmath23 g. the geometry of the magnetic field is important for the efficiency of electron injection and acceleration .
it has been argued ( e.g. , levinson 1996 ; mcclements et al .
1997 ) that quasi - perpendicular shocks are an efficient source of freshly injected electrons if their speed @xmath24 ; the injection efficiency drops sharply below that threshold . on the other hand , bykov & uvarov ( 1999 ) have shown that quasi - parallel mhd shocks of moderate alfven mach number @xmath25 are suitable sites of electron injection and acceleration .
the snr shock waves in molecular clouds are expected to be in this regime of mach number .
thus we shall discuss below the quasi - parallel portion of the shock ( magnetic field inclinations @xmath26 ) propagating into a molecular cloud with @xmath27 .
nonresonant interactions of the electrons with fluctuations , generated by kinetic instabilities of the ions in the transition region of a quasi - parallel supercritical shock , play the main role in heating and preacceleration of the electrons .
the electron heating and preacceleration ( injection ) occur in the collisionless shock front transition region on a scale @xmath28 of several hundred inertial lengths of the ions , @xmath29 ( bykov & uvarov 1999 ) . here
@xmath30 cm and @xmath31 is the ion plasma frequency . for a snr shock in a molecular cloud ,
the scale @xmath32 10@xmath33 cm is much shorter then the scales of all the other regions . to accelerate electrons injected in the shock transition region to relativistic energy @xmath34
, mhd turbulence should fill the acceleration region , which has a scale @xmath35 .
here , @xmath36 , where @xmath37 is the diffusion coefficient of an electron of momentum @xmath38 and @xmath39_(i ) , ( iii ) _ refers to the regions defined above . to provide the conditions for efficient transformation of the mhd flow power to the accelerated electrons ,
the mhd turbulent fluctuation spectrum in the vicinity of the shock should extend to scales @xmath40 ( resonant scattering ) , where @xmath41 is the gyroradius of an electron with momentum @xmath38 . the mhd fluctuations responsible for the electron scattering are collisionless for particles with energies up to a gev because the ion - neutral collision length is above 10@xmath42 cm for typical parameters of the radiative shock described below ( a kev electron has @xmath43 10@xmath44^{-1}$ ] cm ) . to calculate the spectra of nonthermal electrons in the regions
_ i = i @xmath45 iv _ , we use a kinetic equation for the nearly - isotropic distribution function @xmath46 : @xmath47.\end{aligned}\ ] ] the fokker - planck type equation ( 1 ) takes into account diffusion and advection ( with bulk velocity @xmath48 $ ] ) of an electron in phase space due to particle interaction with mhd waves and the large scale mhd flow ( bykov & toptygin 1993 ) . here , @xmath49 is the momentum loss rate of an electron due to coulomb collisions in a partially ionized plasma ( e.g. , ginzburg 1979 ) .
the momentum diffusion coefficient @xmath50 is responsible for second order fermi acceleration , and @xmath37 is the fast particle spatial diffusion coefficient . for low energy electrons ( i.e. , @xmath51 ) , coulomb and ionization losses are important in the regions _
( i @xmath45 v ) _ , except for the narrow shock transition region _
( ii ) _ where acceleration is fast enough to overcome losses and nonthermal electron injection occurs . the characteristic energy @xmath52 depends on the plasma density and ionization state , magnetic field , and fast particle diffusion coefficients @xmath37 and @xmath53 .
the corresponding momentum @xmath54 can be estimated as the point where the last two terms in equation ( 1 ) are equal .
we used the energy @xmath52 calculated simultaneously with the electron distribution function as a convenient parameter to distinguish between different possible models of mhd turbulence in the postshock cooling layer of the radiative shock structure ( see below in this section ) .
the diffusion coefficients @xmath55 and @xmath53 depend on the spectrum of the collisionless mhd turbulence , which is poorly understood .
a plausible approximation for the diffusion coefficients @xmath37 was assumed here .
for all of the regions we used the following parameterization : @xmath56 here , @xmath57 is the momentum of the upstream thermal electrons , @xmath58 is the momentum corresponding to the upper cut - off energy @xmath59 ( see equation [ 3 ] ) , and @xmath60 and @xmath61 are defined in the next paragraph .
the standard relation for the momentum diffusion coefficient @xmath62 = @xmath63 ( e.g. , berezinsky 1990 ) was used . in the low energy regime @xmath64 ,
the particle transport is dominated by large scale turbulent advection ( bykov & toptygin 1993 ) .
the large scale turbulence is due to stochastic mhd plasma motions on scales @xmath65 longer than the particle mean free path due to resonant scatterings .
the sources of mhd turbulent motions are the shock wave instabilities mentioned at the beginning of this section .
the large scale vortex rms - velocity @xmath66 typically is a fraction of bulk speed @xmath67 .
the spatial diffusion is energy independent below @xmath60 and @xmath68 . for @xmath69
, electrons are scattered by resonant mhd waves presumably generated by the streaming instability of shock accelerated particles ( e.g. , blandford & eichler 1987 ; jones & ellison 1991 ) . in this case
@xmath66 is close to the alfven velocity @xmath70 and @xmath37 is the electron diffusion coefficient due to resonant wave - particle interactions .
if the power spectrum of magnetic field fluctuations of alfvenic turbulence for the resonant fluctuations is approximated as @xmath71 then the power - law index @xmath72 ( equation [ 2 ] ) is related to the index @xmath73 as @xmath74 .
an important particular case corresponding to @xmath75 can be written as @xmath76 .
the parameter @xmath77 determines the scattering `` strength '' and strong scattering ( bohm limit ) corresponds to @xmath78 1 . the diffusion model with @xmath75
was successfully used for modeling the observed anomalous cosmic ray proton fluxes in the interplanetary medium ( ellison 1999 ) and in the monte carlo simulations of nonlinear shocks in the shell type snrs ( e.g. , baring 1999 ) .
the momentum @xmath79 in our model can be estimated from the equation @xmath80 .
mhd waves in the partially ionized plasma of the radiative shock flow are subject to ion - neutral damping ( e.g. , vlk 1981 ) .
we define the momentum @xmath81 from the condition @xmath82 , where @xmath83 is determined from the relation @xmath84 and @xmath85 is the alfven wave damping rate due to ion - neutral collisions ( kulsrud & cesarsky 1971 ) . for @xmath86
, electrons are scattered by nonresonant small scale waves .
the limits imposed on the diffusive shock acceleration of particles due to ion - neutral mhd wave damping have been discussed by draine & mckee ( 1993 ) and drury ( 1996 ) .
they considered the spectra of alfven waves driven by the instabilities of shock accelerated ions in a partially ionized plasma .
drury ( 1996 ) obtained the upper cut - off energy of a proton due to ion - neutral wave damping @xmath87 where @xmath88 is the neutral particle density , @xmath89 is the ion density ( both are measured in @xmath90 ) , and @xmath91 is the total particle pressure normalized to 10% of the shock ram pressure .
the ionization structure of the preshock region is inhomogeneous so the ion number density , temperature and other numbers in equation ( 3 ) must be taken at an upstream distance @xmath92 from the collisionless shock transition .
application to our model yields @xmath93 10 gev if @xmath94 ( @xmath95 ) and @xmath96 @xmath97 s@xmath1 because of the high ionization in the preshock region _
( i ) _ and the postshock cooling layer _ ( iii ) _
provided by the _
uv _ radiation .
one can justify the same value of @xmath59 for the electrons because the synchrotron and inverse compton losses are relatively unimportant for @xmath93 10 gev . _
region i_. in the preshock region of the shock with @xmath98 1.1 , we assume a highly ionized plasma of temperature @xmath99 1 up to depths @xmath21 a few times 10@xmath100 @xmath20 upstream . for the ic 443 preshock region , we adopted a density @xmath101 @xmath8 and @xmath102 1.5 ( fesen & kirshner 1980 ) . the magnetic field @xmath103 g ( chevalier 1999 ) , while @xmath104 g ( the inclination is assumed to be @xmath105 ) .
the alfven velocity here is about 5 km s@xmath1 .
we assume a conservative diffusion model in the shock upstream region where @xmath55 has @xmath72 = 1 and the moderate scattering `` strength '' 10 @xmath106 100 ( ellison 1999 ) . _
region ii_. in the shock transition region , injection and heating of the electrons occur due to nonresonant interactions with strong mhd fluctuations generated by the ions . following the model developed by bykov & uvarov ( 1999 ) we take @xmath107 where @xmath108 and @xmath109 .
the alfvenic mach number of the shock , @xmath110 10 , is moderate for our situation .
nevertheless , the effect of the accelerated particle pressure on the shock structure could be important ( see , e.g. , jones & ellison 1991 ; baring 1999 ) .
an exact treatment of the effect is not feasible in our model because it would require a kinetic description of the ion injection .
we accounted for the effect by correction of the total shock compression ratio @xmath111 and by introducing an extended modified shock structure with a precursor and a subshock with compression ratio @xmath112 ( e.g. , jones & ellison 1991 ) . in our case with mach numbers
@xmath113 15 , the estimated extra compression was about 10% and thus @xmath114 4.4 . a subshock compression ratio @xmath115 3.3 is consistent with the assumption of substantial heating of the precursor gas due to wave dissipation ( berezhko & ellison 1999 ) .
these numbers are in agreement with those from a hybrid simulation of nonlinear shock structure ( giacalone 1997 ) .
we found that the resulting high energy emission spectra are sensitive to the particular choice of precursor flow structure in the mev regime .
the kev and gev emission is not sensitive to the structure if the compression ratios @xmath111 and @xmath116 are fixed . _
region iii_. the postshock cooling layer has about a column density @xmath11710@xmath100 @xmath20 of highly ionized plasma with an initial density about @xmath111 times that of the preshock one .
the alfven velocity in the highly ionized portion of the region _ ( iii ) _ is similar to that at the preshock region , @xmath21 5 km s@xmath1 .
large scale vorticity with an amplitude @xmath118 on scales @xmath119
cm may occur here because of shell instabilities of the radiative shock with @xmath17 1.5 ( see , e.g. , section 5 in the review by draine & mckee 1993 ) .
this would dominate the low energy electron propagation , providing a spatial diffusion coefficient of @xmath120 @xmath97 s@xmath1 for the radiative shock parameters described above .
coulomb losses are important here for electrons with energies below @xmath121 20 kev if the large scale turbulent velocity has a substantial longitudinal component with @xmath122 .
@xmath123 8 mev is expected in the absence of a longitudinal component of large scale turbulence in the postshock cooling layer because the coulomb losses are overcome by resonant interaction with alfven waves .
since the alfven velocity is relatively low ( @xmath124 ) , the value of @xmath52 is much higher in that case .
an accurate description of mhd turbulence in the postshock cooling layer is not available now .
thus we considered both limiting cases described above and show the uncertainty introduced by the lack of data concerning the large scale turbulence properties in fig . 1 . to avoid overestimating the nonthermal emission , we used a conservative minimum value of @xmath52 = 120 kev for the radiative shock structure described above , although one could expect @xmath125 20 kev in the most favorable case . _
region iv_. the density and temperature in the radiative shell depend on the preshock density , magnetic field and abundances ( shull & mckee 1979 ; chevalier 1999 ) . from the model of chevalier ( 1999 ) for the above preshock parameters , we adopted @xmath126 275 @xmath8 ( @xmath127 10@xmath128 @xmath20 ) and @xmath129 10@xmath130 k in the dense radiative shell .
the mass of the dense radiative shell is @xmath131 @xmath132 and its magnetic field is @xmath133 10@xmath134 g. the ionization in the dense radiative shell is supported at the level of about a percent by shock accelerated particles penetrating into the shell ( see section 3.5 for details ) .
the phase velocity of the high frequency collisionless alfven waves which are responsible for electron resonant scattering is about 50 km s@xmath1 ; it is determined by the number density of the ions .
the coulomb loss rate of the electrons is then dominated by the neutral component .
we numerically calculated the electron distribution function in regions _
( i - iv)_. the results were then used in calculations of the nonthermal emission .
the snr ic 443 ( g189.1 + 3.0 ) is a very good candidate for testing our model because the interaction of the snr with a molecular cloud was established from radio and infrared line observations ( e.g. , denoyer 1979 ; mufson 1986 ; burton 1990 ; van dishoeck 1993 ; richter 1995 ; claussen 1997 )
. the high energy fluxes are somewhat better determined than in other cases .
although we concentrate on ic 443 , we briefly mention the case of w44 , where observations of oh masers and co clumps again provide convincing evidence of molecular cloud interaction ( claussen et al 1997 ; frail & mitchell 1998 ) . to model the nonthermal emission from the shell of ic 443 we integrated the local emissivities over the radiative shock structure ( regions _ i @xmath45iv _ ) .
as a parcel of gas containing accelerated nonthermal electrons evolves through the radiative shock structure including the dense shell , the local emissivity from bremsstrahlung and inverse compton was calculated using standard theory ( see appendix a for details ) .
the electron - ion bremsstrahlung emission differential cross sections ( bethe - heitler formulae ) were taken from akhiezer & berestetsky ( 1957 ) taking into account the elwert factor , which is important for modeling the kev emission . the electron - electron bremsstrahlung contribution to the emissivity
was calculated using cross sections derived by haug ( 1975 ) .
for the inverse compton emissivity of ic 443 , we used the same description of the background photon field as gaisser ( 1998 ) .
the hard emission production cross sections used in our calculations are similar to those used recently by other authors ( asvarov 1990 ; sturner 1997 ; gaisser 1998 ; baring 1999 ) .
we also estimated the contribution of pion decays to the emission using a model for the proton component and the cross sections from dermer ( 1986 ) .
the resulting emission in the _ egret _ regime was roughly 7 times below that from electron bremsstrahlung , which is consistent with the results of sturner et al .
we do not discuss the pion component further .
3.55truein 0.2truein the synchrotron emission of the electrons calculated using the scheme described by ginzburg ( 1979 ) can be compared to the radio observations and to the upper limits at higher frequencies .
we also included free - free absorption of low frequency radio waves ( dotted line in fig .
the synchrotron losses of a relativistic electron were included following ginzburg ( 1979 ) ; bremsstrahlung energy losses are relatively unimportant in our case .
the emission was integrated over the regions of the radiative shock in order to model the spectrum of the entire shell .
iv _ is the dominant contributor to the synchrotron and bremsstrahlung emission .
the results in fig .
1 can be compared to the _ asca _ and _ egret _ observations of snrs , as well as to the significant upper limits that have been set at tev energies by whipple observations ( buckley 1998 ) and from the preliminary analysis of _ cgro osse _
observations presented by sturner ( 1997 ) .
we tried different parameter sets to fit the observations . in fig . 1
, @xmath135 fits to the observations of ic 443 are presented for the model described in the previous section .
the preshock density is taken to be @xmath136 and the assumed distance to ic 443 is 1.5 kpc ( e.g. , fesen & kirshner 1980 ) .
interstellar photoelectric absorption of the x - ray spectrum is accounted for with a line - of - sight column density @xmath137 @xmath20 using the cross sections from morrison & mccammon ( 1983 ) . for @xmath14 = 1.5 in the diffusion model described by equation ( 2 )
, we used @xmath138 @xmath97 s@xmath1 , @xmath72 = 1.0 , @xmath139 ) = 1 mev , and @xmath140 ) = 20 gev , which are compatible with theoretical estimates of resonant wave - particle interactions . the corresponding value @xmath141 30 . while in principle
@xmath142 can be determined from observations of heliospheric shocks , single spacecraft observations do not provide direct values and are subject to ambiguous interpretation .
values of @xmath143 are consistent with modeling of the highly oblique solar wind termination shock ( ellison 1999 ) , but smaller values ( @xmath144 ) have been inferred for interplanetary traveling shocks ( baring et al .
1997 ) . on the other hand ,
very weak scattering ( @xmath145 ) has been inferred from observations near corotating interaction regions ( e.g. , fisk , schwadron , & gloeckler 1997 ) and in the solar wind when the interplanetary magnetic field is nearly radial ( e.g. , mbius et al .
we use @xmath141 30 as a conservative value .
if strong mhd turbulence is present in the upstream shock region , the @xmath142 values could be closer to 1 .
this might allow an electron acceleration model similar to that described earlier , but for slower shock velocities ( @xmath146 ) .
the diffusion propagation model is not unique .
we obtained a similar fit to that presented in fig . 1 for a diffusion model with @xmath147 = 10@xmath128 @xmath97 s@xmath1 , @xmath72 = 0.7 , @xmath139 ) = 1 mev , @xmath140 ) = 20 gev .
as mentioned in 2 , there is an important uncertainty concerning the lack of a quantitative model for the generation of large scale mhd turbulence in the postshock cooling layer .
the uncertainty affects the hard x - ray spectral calculations . in view of this , we calculated the expected bremsstrahlung emission for two limiting cases covering the range of uncertainties .
curve 1 in fig .
1 corresponds to fully developed alfvenic large scale mhd turbulence in the postshock cooling layer ( @xmath52 = 120 kev ) .
curve 2 corresponds to the case of a lack of large scale mhd turbulence ( @xmath52 = 2 gev ) . while the gev bremsstrahlung emission ( as well as synchrotron and inverse compton emission )
is similar in the 2 cases , the hard x - ray emission is sensitive to the uncertainty
. we also obtained a similar fit to that presented in fig .
1 for a model with @xmath14 = 1.1 and a diffusion model described by @xmath148 @xmath97 s@xmath1 , @xmath72 = 1.0 , @xmath139 ) = 1 mev , and @xmath140 ) = 10 gev . the model with @xmath149 1.1 requires a somewhat higher power conversion efficiency than that with @xmath149 1.5 ( see 3.6 ) , but this value is preferred for the ic 443 radiative shock velocity ( chevalier 1999 ) .
the model of electron injection from the cloud thermal pool described above is based on the ionization structure of a radiative shock , which is very sensitive to the shock velocity if @xmath17 1 ( shull & mckee 1979 ; hollenbach & mckee 1989 ) . in this respect , it is instructive to consider a model in which high energy electrons are not related to freshly injected particles , but are reaccelerated cosmic ray electrons ( blandford & cowie 1982 ; chevalier 1999 ; cox 1999 ) . to model that case
we suppose that the far upstream electron flux is just that observed for galactic cosmic ray electrons near earth , but extrapolated back to the energy @xmath150 with the same slope as observed above a gev .
the calculations of @xmath135 for this case are shown in fig .
we consider the radiative shock structure with the parameters described in 2 ( the same as were used for fig .
the galactic cosmic ray electron flux in the far upstream region was taken to be 3@xmath151 e@xmath152 s@xmath1 @xmath20 gev@xmath1 sr@xmath1 ( e.g. , berezinsky 1990 ) for @xmath153 .
we assumed a flat cosmic ray electron spectrum below @xmath150 , corresponding to flattening due to coulomb losses in the interstellar medium .
the maximum energy of cosmic ray electrons reaccelerated by the radiative shock is @xmath93 10 gev ( see equation [ 3 ] ) , as in the case shown in fig .
the energy @xmath150 was considered as a free parameter here .
3.55truein we do not distinguish between electrons and positrons because there is no difference between them in the mhd shock acceleration process considered above and most of the emission from high energy leptons involved in our model remains the same for both kind of particles . because of interactions with the radiative shock structure , accelerated cosmic ray electrons could provide a good fit for both synchrotron radio and _ egret _ emission if @xmath154 10 mev for the radiative shell parameters described above , but fall short for @xmath155 100 mev .
the relatively low energies , @xmath156 10 mev , required for the ic443 shell are a potential problem for this model .
the radio synchrotron radiation from the galaxy indicates that the cosmic ray electron spectrum flattens to @xmath157 below energies @xmath158 gev ( webber 1983 ) .
if we use such a spectrum in our model , the nonthermal emission is significantly below that observed .
however , there are uncertainties about betatron - type acceleration in the shock compression , about the magnetic field strength , and about the volume of the emitting region .
duin & van der laan ( 1975 ) did successfully model the ic 443 radio flux with a cosmic ray compression model .
we found that models with a higher magnetic field and/or larger volume could reproduce the radio flux , but that the @xmath2-ray fluxes observed by _ egret _ could not be achieved .
cox ( 1999 ) have recently concluded that cosmic ray compression by radiative shocks could provide a good fit to radio and gamma - ray emission from w44 if the cosmic ray electron population in the vicinity of the snr is higher than that of measured in solar neighborhood .
they used a somewhat higher magnetic field ( @xmath159 200 @xmath160 g ) in the radiative shell of w44 than the 60 @xmath160 g field we assumed for the ic 443 shell .
the acceleration time is a constraining parameter in this model because of the relatively low shock speed @xmath27 in the dense medium .
the test particle shock acceleration time @xmath161 can be estimated from the equation @xmath162 where the velocities @xmath163 are measured in the shock rest frame ( e.g. , axford 1981 ) .
for the radiative shock parameters described above , the acceleration time for @xmath164 10 gev is about 3,000 years , for @xmath165 30 ( note that @xmath166 ) , consistent with the age of ic 443 estimated from the hydrodynamical model by chevalier ( 1999 ) . on the basis of _ ginga _ observations of ic 443 , wang ( 1992 ) suggested an age of about one thousand years and an association with the supernova of ad 837 .
the reasons for the low age are related to a thermal interpretation of the observed x - ray emission above 10 kev .
we shall argue below for the nonthermal origin of the hard emission in a radiative shock model .
a distinctive feature of the radiative shock snr shell model with direct injection of the electrons from the thermal pool is a hard spectrum of nonthermal emission extending from kev to gev energies ( fig .
1 ) . there is also a hard emission component due to inverse compton emission of relativistic electrons having a scale comparable to the scale of the whole radio image of the remnant .
observations of the radio emission from ic 443 were performed with different instruments over the last 35 years ( see , e.g. , erickson & mahoney 1985 ; green 1986 ; claussen 1997 and references therein ) .
the 151 and 1419 mhz radio maps ( green 1986 ) reveal that the most intense flux comes from the shell of the remnant .
we suggest that the eastern part of the shell is a radiative shock which has a high magnetic field and electron flux , dominating the synchrotron radiation from the remnant .
we calculated the synchrotron radiation from the remnant using the local emissivity given by equation ( a17 ) .
the integrated spectrum from the remnant is dominated by the emission from the radiative shell ( having magnetic field @xmath167 g ) if the magnetic field in the central parts of the remnant ( region _ v _ ) is moderate , @xmath168 g. for modeling the shell emission , we took into account the interstellar free - free absorption at low radio frequencies , assuming a column density to ic 443 of @xmath169 @xmath20 and an average interstellar medium temperature @xmath170 k. this column density is similar to that adopted for the recent analysis of _ asca _ data by keohane ( 1997 ) and is about four times less than the value used earlier by petre ( 1988 ) and wang ( 1992 ) .
we assumed that the ionization fraction is @xmath171% .
the free - free absorption inside the cool radiative shell is not very important above 50 mhz because the emission measure is about 1 @xmath172 pc .
the solid and dot - dashed lines on fig . 1 show the total synchrotron radiation from ic 443 ( dominated by shell emission in our model ) expected at the earth compared to the observational data ( erickson & mahoney 1985 ) . both of our models are illustrated : direct injection of thermal electrons ( fig . 1 ) and reacceleration of preexisting cosmic ray electrons by a radiative shock ( fig .
2 ) . a radiative shock with total compression ratio @xmath111 = 4.4 , including a shock precursor and subshock ( @xmath116 = 3.3 ) , was considered . an important feature of the radio observations of ic 443 and w44 is that their spectral indices are relatively flat , implying electron energy spectra flatter than those produced by standard strong shock acceleration in the test particle limit . in our model
, the calculated radio spectrum fits that observed from the ic 443 shell because of the large shock compression ratio @xmath173 and to a lesser extent because of the effect of second order fermi acceleration discussed recently by ostrowski ( 1999 ) .
molecular clouds show complex structure with an ensemble of dense massive clumps embedded in the interclump medium ( blitz 1993 ) .
we considered the possibility of radiative shock interaction with dense molecular clumps , resulting in high energy emission . in ic 443 and w44
, those regions do not appear to be significant sites of ghz radio continuum emission , but the clumps may contain most of the cloud mass and are possible sites of hard x - rays and mev @xmath2-ray emission . the interaction of a radiative shell with a molecular clump generates a dense slab bounded by two shocks ( chevalier 1999 ) .
the forward shock propagating into the clump is strong ( see also the simulations of jones & kang 1993 ) .
since the clump size is typically smaller or comparable to that of the radiative shock layer , it is embedded in a pool of nonthermal particles created by the radiative shock according to the scenario described above .
accelerated particles as well as ionizing radiation from the hot interior regions of the snr should provide ionization of the external portions of a dense clump at the level @xmath174% .
this is sufficient for a shock propagating into the dense clump with @xmath175 10@xmath176 @xmath90 and @xmath159 0.1 mg ( e.g. , claussen 1997 ) to be collisionless .
the percent ionization level provides alfven mach numbers @xmath110 1 and weak damping of collisionless alfven waves up to wavelengths @xmath21 10@xmath177 cm for a flow with velocity @xmath178 25 @xmath179 in the clump .
then the injection and acceleration processes considered above would be efficient and the ionization in the preshock region would be maintained self - consistently at least at a percent level even for a shock propagating inside the dense clump .
the gas ionization by uv radiation in the radiative shock precursor is sensitive to the shock speed if @xmath180 1.1 ( shull & mckee 1979 ; hollenbach & mckee 1989 ) .
this implies that a shock having @xmath180 1 propagating in a dense clump would be able to accelerate electrons only below @xmath181 1 gev ( see equation [ 3 ] ) .
we shall estimate below also the gas ionization rate due to accelerated nonthermal particles .
the lifetime of a pre - existing electron of mev energy against ionization losses in the interclump medium ( @xmath175 10 @xmath8 ) is about @xmath182 years . unless a nearby source of fresh accelerated electrons is present , galactic cosmic ray mev electrons are unlikely to be present in the cloud .
direct injection of electrons from the thermal pool of the weakly ionized precursor would be a dominant source of shock accelerated electrons up to the cut - off energy @xmath59 . because of a lack of accelerated gev electrons in that case ( @xmath180 1 ) the source would not appear as a bright radio and hard _ egret _ emission feature .
nevertheless , these shocks might manifest themselves as hard x - ray and mev emission sources with @xmath183 @xmath184 up to a few mev and with a soft spectrum at higher energies . a comparison of fig .
3 with fig .
1 shows that the hard x - ray flux at 7 kev and above could be dominated by the emission from the shocked clumps , which is consistent with _ asca _ observations of ic 443 presented by keohane ( 1997 )
. the hard x - ray nonthermal emission flux may be variable on a timescale of years .
a relatively high mev emissivity is a distinctive feature of shocked clumps , but they are difficult to observe
. they must be nearby ( within a few hundreds of parsecs ) to be detected with _
comptel / osse cgro_. for a forward shock of velocity @xmath180 1 propagating into a dense clump of number density @xmath185 10@xmath186 @xmath8 with upstream ionization level of a few percent ( at @xmath187/v_s$ ] distance ) , the upper cut - off energy @xmath188 0.5 gev .
this is about 40 times less than that for a radiative shock in the interclump medium .
the spatial diffusion @xmath189 below @xmath60 scales as @xmath190 .
since the scale of a postshock radiative cooling layer @xmath191 and @xmath65 is a fraction of that , we have @xmath192 10@xmath193 @xmath97 s@xmath1 in the clump case . the effective compression of the nonthermal electrons as well as the magnetic field by the system of two shocks bounding the slab could be high .
the magnetic field in the clump is expected to be high .
claussen ( 1997 ) derived line - of - sight magnetic field strengths of @xmath210.2 mg , remarkably uniform on the scale of several parsecs .
relativistic electrons accelerated by a shock propagating into a clump should produce radio emission with a flat spectrum up to some hundreds of mhz , which is lower than the maximum radio frequency from the shell because of the lower @xmath59 in the clumps .
the emission measure of a clump interacting with a radiative shock could be em@xmath194 1,000 @xmath172 pc , which is much higher than that of the shell .
thus , the internal free - free absorption could be important here providing a flatter radio spectrum in the 100s of mhz regime .
this is possibly a reason for the observed spatial variations of the radio spectral index in ic 443 ( e.g. , green 1986 ) .
we present in fig .
3 the calculated @xmath135 spectrum of nonthermal emission from a radiative shock - molecular clump interaction region . the forward shock velocity was high
@xmath15 , the initial number density in the clump 1,000 @xmath8 , and the shocked clump radius 0.5 pc .
the shock compression ratio was the same as that in the radiative shell model : @xmath195 3 and @xmath196 4.4 .
the ionization of gas in the shock upstream was about 5% ( at the distances @xmath187/v_s$ ] ) . to account for some uncertainty in the gas ionization , we present the radio spectra for two possible values of the emission measure em of the shocked clump : em = 110 @xmath172 pc ( solid line ) and em = 900 @xmath172 pc ( dashed line ) . the diffusion coefficient was @xmath192 10@xmath193 @xmath97 s@xmath1 , @xmath139 ) = 1 mev .
we illustrate the results for two possible maximal energies of the electrons ( depending on the time of shock - clump interaction ) .
curve 1 in fig .
3 corresponds to @xmath197 0.5 gev and @xmath197 0.05 gev for curve 2 .
the synchrotron radio spectrum for curve 2 is below the scale of the plot .
the radio spectrum corresponding to curve 1 is very hard because of internal absorption in the shocked clump medium .
3.55truein we presented above the results for nonthermal emission of a strong radiative shock of @xmath198 propagating into a molecular clump .
for the case of radiative shock interaction with the molecular clumps in ic 443 , lower shock velocities , @xmath17 0.3 , are more realistic ( e.g. , cesarsky 1999 ; chevalier 1999 ) .
the preshock ionization may be dominated mostly by nonthermal particles in such a shock because uv radiation is inefficient .
a self - consistent model of mhd collisionless shock propagation into a dense clump requires simultaneous simulations of the mhd turbulence spectral properties and the nonthermal particle kinetics .
instead , we suppose that a _ collisionless _ shock transition exists for @xmath17 0.3 and @xmath199 and assume the same diffusion model as described in 2 .
we calculated the upstream gas ionization , finding that the ionized fraction @xmath200 0.9 may hold up to depths about 10@xmath201 @xmath20 due to shock accelerated electrons .
the mach number of the shock with @xmath17 0.3 is typically below 5 because of heating of upstream gas up to 10@xmath176 k by accelerated particles and the compression ratio is lower than 4 .
the maximum energy of accelerated electrons in that case is typically below an mev .
we illustrate the calculated @xmath135 spectrum of nonthermal emission from a 30 @xmath179 velocity shock ( @xmath111 = 3 ) interacting with a molecular clump of density 10@xmath202 by curve 3 in fig .
3 ; this model corresponds to @xmath203 10@xmath204 @xmath97 s@xmath1 and @xmath139 ) = 10 kev .
such collisionless shocks , if they exist , would provide a softer spectrum of nonthermal radiation than that for high velocity shocks with @xmath98 1 . a more comprehensive study is required to model the nonthermal emission from the low velocity shocks in detail . a search for hard x - ray emission correlated with molecular emission predicted by
the simplified model described above is possible with high resolution instruments like _ chandra _ and _
xmm_. ic 443 was a target of x - ray observations with _ heao 1 _ ( petre 1988 ) , _ ginga _ ( wang 1992 ) , _ rosat _ ( asaoka & aschenbach 1994 ) and _ asca _ ( keohane 1997 ) .
_ asca gis _ observations discovered the localized character of the hard x - ray emission ( keohane 1997 ) .
most of the 2 - 10 kev _ gis _ photons came from an isolated emitting feature and from the southeast elongated ridge of hard emission .
the ridge is also coincident with the 95% confidence error circle of the _ egret cgro _ source ( esposito 1996 ) , while the isolated feature is outside the circle ( keohane 1997 ) .
the integrated 2 - 10 kev flux observed by _
asca gis _ was ( 5@xmath205 10@xmath206 erg @xmath20 s@xmath1 , representing more than 90% of the total flux from the snr ( keohane 1997 ) .
the hard x - ray ridge , as defined by keohane ( 1997 ) , is only a part of the more extended radio shell discussed in the previous section .
the flux density at 7 kev from the isolated emitting feature was about 4@xmath22 10@xmath134 photons @xmath20 s@xmath1 kev@xmath1 and that from the ridge was 2@xmath22 10@xmath134 photons @xmath20 s@xmath1 kev@xmath1 .
_ heao a-2 _ measured a flux at 210 kev of ( 7@xmath205 10@xmath206 erg @xmath20 s@xmath1 ( petre 1988 ) , while _ ginga _ measured a 220 kev flux of about 9@xmath22 10@xmath206 erg @xmath20 s@xmath1 ( wang 1992 )
. these numbers are consistent with the assumption that the integrated flux is dominated by the extended soft component coming from the central parts of the remnant .
the soft x - ray 0.23.1 kev surface brightness map of ic 443 from the _ einstein _ observatory ( petre 1988 ) shows bright features in the northeastern part of the remnant .
the presence of nearly uniform x - ray emission from the central part of the remnant is a characteristic feature of mixed - morphology snrs ( rho & petre 1998 )
. it might be due to the effect of thermal conduction ( cox 1999 ; shelton 1999 ) , although harrus ( 1997 ) argued for an alternative radiative shock model for w44 , another mixed - morphology snr .
that feature in our model corresponds to the emission from hot ( t@xmath207 k ) , low density gas interior to the shock ( region _ v _ ) . in the radiative shock model with direct injection from the thermal pool and second order fermi acceleration in postshock cooling layer ,
most of the hard x - ray emission ( @xmath208 kev ) comes from the shell - radiative shock structure with a rising @xmath135 x - ray spectrum ( curve 1 in fig . 1 for the shell and curves 1 and 2 in fig . 3 for a shocked molecular clump ) .
in addition to the shell , there is a more extended source of hard x - rays from inverse compton emission .
it is the dominant source of hard x - rays in the scenario with cosmic ray electron reacceleration ( fig .
2 ) and in the case of a lack of second order fermi acceleration in the radiative shell ( see curve 2 in fig .
1 ) . in the scenario with reacceleration of cosmic ray electrons , it is impossible to obtain the 7 kev flux observed by _
asca _ from the ic 443 ridge .
thus , a resolved hard x - ray image of the ic 443 remnant may be very informative in determining the electron acceleration scenario . the isolated emitting feature in the southern part of ic 443 detected by _ asca _ and brightest around 7 kev , has a flat spectrum low frequency radio continuum source ( green 1986 ) and is close to a region of excess h@xmath209 emission .
high spatial resolution observations of this isolated emitting feature are required to decide between a low - luminosity pulsar nebula and shock interaction models ( e.g. , radiative shock interaction with a dense molecular clump , fig .
observations of the hard emission with the sensitivity and resolution available with _
bepposax , chandra xro , xmm _ and _ glast _ would be required to constrain the extension and spectral properties of the hard x - ray emission from ic 443 .
nonthermal particles accelerated by a mhd collisionless shock wave provide an efficient ionizing agent .
we calculated above the nonthermal emission generated by accelerated energetic electrons ; one can also calculate the ionization produced by the same electrons in the molecular cloud .
this might connect the high energy observations with radio , ir and optical emission from shocked atomic and molecular gas . to estimate the gas ionization due to accelerated electrons in the radiative shock structure we used the electron spectra calculated with the kinetic model of electron acceleration and propagation described in 2 and electron ionization cross sections from appendix b. in fig .
4 , we present the calculated ionization rate @xmath210 and estimated ionized fraction @xmath211 in the shell of a radiative shock due to accelerated electrons for the case of @xmath212 for both models of large - scale turbulence in the postshock cooling layer considered in 3.1
. 4.55truein [ fig : fig4 ] there should also be contributions to @xmath211 from accelerated nucleons and from uv radiation which are not included in fig .
we accounted for only radiative recombination , assuming a small molecular fraction in the radiative shell at column densities below 2 @xmath213 @xmath20 . the solid curves in both panels of fig .
4 correspond to the case of fully developed alfvenic turbulence in the postshock cooling layer , which provides efficient second order fermi acceleration making coulomb losses relatively unimportant above @xmath52 = 120 kev .
this case corresponds to curve 1 in fig . 1 for emission from the shell with substantial hard x - ray emission . the dotted lines in fig .
4 correspond to the case of a lack of large scale mhd turbulence where the coulomb losses overcome the second order fermi acceleration up to @xmath52 = 2 gev . the nonthermal emission expected in this case
is illustrated by curve 2 in fig .
the radiative shell ionization by nonthermal electrons is sensitive to the turbulent structure of the postshock cooling layer and there is a correlation between the hard x - ray spectrum and the ionization structure of the radiative shell .
we shall consider the effect of nonthermal particles on the ionization structure of a shocked clump elsewhere .
the energy of the electron component for the parameter set used to compute the spectra in fig .
1 is substantial .
the power required for electron acceleration and maintenance for curve 1 in fig . 1 ,
@xmath214 erg s@xmath1 , is high because of the strong coulomb losses of kev electrons in the dense ionized medium , but it is consistent with the estimate of the total radiative losses from ic 443 , dominated by infrared emission ( mufson 1986 ; burton 1990 ) .
a similar power is required for both models 1 and 2 in fig .
3 , where radiative shock - molecular clump interaction is illustrated . for curve 2 ,
the power requirements are less , @xmath215 erg s@xmath1 .
the implied efficiency of power conversion from the mhd shock flow ( which can be estimated as 3 @xmath216 erg s@xmath1 ) to the nonthermal electrons is about 5% in these cases .
the model with @xmath217 would provide emission similar to that shown in fig . 1 , but requires about three times higher efficiency .
this is higher than the electron acceleration efficiency estimated from gev regime cosmic rays observed near the earth , but evolved snrs interacting with dense molecular gas probably can not be considered as the main source of the observed galactic cosmic rays because of the relatively low maximum energies of accelerated particles .
the pressure of the nonthermal electron component downstream of the shock ( region _ iv _ ) is @xmath218 7@xmath22 10@xmath206 erg @xmath8 .
the magnetic pressure in the dense shell is about 1.5 @xmath22 10@xmath219 erg @xmath8 , which is higher than the thermal gas pressure .
the uniform ( @xmath21 60 @xmath160 g ) and stochastic magnetic field components dominate the total pressure in the shell .
let us consider the energy requirements for the scenario with electron injection from preexisting cosmic rays ( fig .
the nonthermal electron pressure in this case is @xmath218 5.2@xmath22 10@xmath206 erg @xmath8 ( for @xmath154 8 mev ) , similar to that for the model with injection from the thermal pool , because it is dominated by gev particles .
an advantage of this scenario is that much less power is required .
it is about 4 @xmath22 10@xmath220 erg s@xmath1 , which is less than 10% of that for the scenario with injection from the thermal pool .
the difference is because coulomb losses are unimportant for the high energy electrons involved in that scenario .
however , a potential problem is the relatively low @xmath150 required .
the lifetime of a 20 mev electron against ionization losses in a neutral medium of number density @xmath21 1 @xmath8 is about 10@xmath221 years .
this implies that the source of mev cosmic ray electrons should exist within about 100 pc of the molecular cloud if the diffusion coefficient is @xmath117 10@xmath222 @xmath97 s@xmath1 at these energies .
for individual supernova remnants , the ambient density is an important parameter that can be estimated from multiwavelength observations ( chevalier 1999 ) .
another important parameter for modeling the nonthermal particles in a snr in a molecular cloud is the collisionless mhd turbulence spectrum , particularly in the radiative shock cooling layer .
a substantial level of mhd collisionless turbulence could overcome coulomb losses in the dense plasma downstream from the radiative shock .
an accurate model of mhd turbulence in the postshock cooling layer is not available now .
thus we considered both limiting cases described in the 2 and show in fig .
1 the uncertainty introduced by the lack of data concerning the mhd turbulence properties . coulomb losses are important in the postshock cooling region only for electrons with energies below @xmath121 20 kev if the large scale turbulent velocity has a substantial longitudinal component of @xmath223 .
a higher value ( @xmath123 8 mev ) is expected in the absence of a longitudinal component of large scale turbulence in the postshock cooling layer because the coulomb losses are overcome by resonant interaction with alfven waves .
since the alfven velocity is relatively low ( @xmath124 ) , the value of @xmath52 is much higher in that case .
we used a conservative minimum value of @xmath52 = 120 kev for the radiative shock structure described above , although one could expect @xmath125 20 kev in the most favorable case .
the nonthermal emission of snr in a molecular cloud ( especially in the hard x - ray regime ) and ionization structure of the radiative shock are sensitive to the mhd turbulence model .
spatially resolved observations of the nonthermal emission from snrs may be able to constrain models of mhd turbulence .
all of the snr candidates from the _ egret _ list : ic 443 , @xmath2 cyg , w44 , and monoceros ( esposito 1996 ) are old remnants interacting with molecular clouds . the emission from electrons accelerated by the radiative shock calculated for the ic 443 parameters presented in figs .
1 and 2 is in good agreement with that observed by _
egret _ ( esposito 1996 ) , as well as with the upper limits established by _ whipple _ ( buckley 1998 ) .
the high energy @xmath2-rays ( @xmath224 mev ) from the radiative shell should be spatially correlated with the radio emission .
the _ egret _
telescope detected an extended excess ( 95% confidence circle @xmath225 in radius ) from the monoceros snr correlated with the 1.42 ghz radio emission ( esposito 1996 ) .
this is in accordance with the model of a leptonic origin of gev emission from extended snrs interacting with clouds ( de jager & mastichiadis 1997 ) .
_ glast _ ( gamma - ray large area space telescope ) will be an excellent instrument for future @xmath2-ray observations , because it will have the capability of spatially resolving the @xmath2-ray emission from snrs .
spatially resolved observations of w44 , ic 443 , 3c391 and some other mixed - morphology snrs from the list given by rho & petre ( 1998 ) with _ bepposax , chandra xro , xmm _ would be very valuable tools to test our model .
the hard x - ray detector ( hxd ) aboard the forthcoming _ astro - e _ mission could be used for observations of the hard 10700 kev continuum with a field of view of 0.8@xmath226 fwhm at 60 kev and 2.8@xmath226 at 500 kev from the extended hard x - ray emission from ic 443 , w44 and the monoceros snr predicted by the radiative shock model ( see fig . 1 and fig .
3 ) . spatially resolved spectra are needed to distinguish the shell emission from the shocked clump emission .
molecular clumps interacting with moderately fast radiative shocks are also expected to be sources of hard x - rays and mev @xmath2-rays ( up to 100s of mev ) .
nonthermal continuum radio emission ( 100s of mhz ) with a sharply rising @xmath135 spectrum ( fig .
3 ) and a time dependent cut - off frequency is expected in the fast shock model . due to potentially substantial internal free - free absorption
, the spectrum of radio emission from a localized clump might constrain the ionized gas density in the clump . the mev spectrum of a localized clump
may be resolved with the forthcoming missions _ integral _
( e.g. , schnfelder 1999 ; winkler 1999 ) and _ glast_. with an expected angular resolution about 12@xmath227 fwhm , imager _ ibis _ aboard _
integral _ would allow detection of hard x- ray ( @xmath178 50 kev ) emission from molecular clouds within @xmath21 1 kpc .
a comparison of hard emission spectra with the radio spectrum can provide valuable information about the density and magnetic field in a clump .
a possible variability on a timescale of a few years for clump hard emission and radio emission ( on a longer time scale ) could further constrain the model . for low velocity ( below @xmath228 ) shocks interacting with a dense ( @xmath229 ) , magnetized ( @xmath230 0.5 mg ) molecular clump
, one may expect hard x - ray emission below the mev regime correlated with the regions of molecular emission of shocked gas .
nonthermal radio continuum is not expected in the case of low velocity shocks because of a lack of accelerated relativistic electrons at sufficiently high energies .
finally , we note that the high density of energetic particles in the vicinity of a shock wave in a molecular cloud can affect the ionization and thermal properties of the gas .
rich molecular spectra have been observed from ic 443 and 3c391 ( reach & rho 1999 ) .
it will be interesting to see whether there is a signature of the presence of nonthermal particles that can be discerned from the molecular spectra .
a.m.b thanks hans bloemen for useful discussions .
we are grateful to m.g . baring for comments and for making available the corrections to the e - e bremsstrahlung formulae and to the referee for very constructive comments .
the work of a.m.b and yu.a.u was supported by the intas grant 96 - 0390 and that of r.a.c . by nasa grant nag5 - 8232 .
given the electron distribution function , the high energy emission flux @xmath231 at a distance @xmath232 from the source can be calculated from the equation : @xmath233 where @xmath234 is the emissivity , which here includes bremsstrahlung , synchrotron radiation and the inverse compton effect .
we generally assumed a pure hydrogen composition in our calculations .
high energy particles penetrating through a partially ionized medium produce photons due to interactions with atoms , electrons and ions .
we assumed a maxwellian distribution for the ambient matter and in most cases took the target ambient particles to be at rest .
then , the emissivity can be calculated from @xmath235 where @xmath236 is the electron flux measured in units s@xmath1
@xmath20 kev@xmath1 sr@xmath1 , @xmath237 and @xmath238 are differential cross sections of one photon emission due to electron - proton and electron - electron interactions integrated over angles , and @xmath239 is the minimum energy of incoming electron that could generate a photon with energy @xmath240 . for e - p interactions , @xmath241 is a good approximation .
the correction due to finite proton mass is less than the uncertainty in the cross section . in the case of e - e interactions
, we used the following equation for @xmath239 : @xmath242.\ ] ] the inverse bremsstrahlung contribution is negligible unless there is a very high ( far in excess of 100 ) ratio of protons to electrons .
we used the following cross section approximations in our calculations : haug ( 1975 ) has obtained a general formula for the e - e bremsstrahlung cross section , but it is more convenient to use simplified approximations .
we used relativistic and non - relativistic asymptotic expressions , which were matched in the intermediate energy region ( see baring 1999 ) . in the relativistic limit , we used approximate formulae given by baier ( 1967 ) with a correction factor from baring ( 1999 ) : @xmath243 where @xmath244 , @xmath245 , @xmath246\left[\ln\left(2(\gamma_e - 1 ) \frac{\gamma_e- 1- \epsilon_{\gamma}}{\epsilon_{\gamma}}\right)-\frac12\right],\ ] ] @xmath247 , & \epsilon_{\gamma}\le\frac12,\\ \frac2{\epsilon_{\gamma}}\left[\left(4-\frac1{\epsilon_{\gamma } } + \frac1{4\epsilon_{\gamma}^2}\right)\ln(2\gamma_e)-2 + \frac2{\epsilon_{\gamma}}-\frac5{8\epsilon_{\gamma}^2}\right ] , & \epsilon_{\gamma}>\frac12 , \end{array } \right.\ ] ] @xmath248 is the classical electron radius , @xmath249 , and @xmath250 according to baring ( 1999 ) , these formulae agree with those from haug ( 1975 ) with an accuracy @xmath251 for electrons of energy @xmath252 mev . in the non - relativistic limit , we used equations from garibyan ( 1952 ) and fedyushin ( 1952 ) with correction coefficients from baring ( 1999 ) : @xmath253 where the photon energy is in the range @xmath254 and @xmath255 ( @xmath256 ) has the form : @xmath257 \sqrt{1-x}\\ & & + \left[12(2-x)-\frac{7x^2}{2-x}-\frac{3x^4}{(2-x)^3}\right ] \ln\left(\frac{1+\sqrt{1-x}}{\sqrt{x}}\right ) .
\end { array}\ ] ] the correction coefficients are : @xmath258 baring ( 1999 ) found less than a @xmath259 difference between equation ( a7 ) and the expressions of haug ( 1975 ) for electron energies @xmath260 kev . in the non - relativistic limit
, we took the e - p bremsstrahlung cross section from akhiezer & berestetsky ( 1957 ) : @xmath261 which is valid if @xmath262 ; here @xmath263 .
for the relativistic limit , we used born approximation formulae from akhiezer & berestetsky ( 1957 ) : @xmath264\right\ } , \end{array } \label{eq : lanah}\ ] ] where @xmath265 , @xmath266 , and @xmath267 . in the velocity range @xmath268 , we included the elwert factor from koch & motz ( 1959 ) : @xmath269 ) } { \beta_2 ( 1-exp[-2 \pi z \alpha / \beta_2)])}$ ] . in equation ( a11 ) ,
@xmath270 is assumed . in the intermediate energy range , we matched equation ( a10 ) and equation ( a11 ) .
we used standard formulae for synchrotron radiation in our calculations ( e.g. , ginzburg 1979 ) .
the emissivity can be written as : @xmath271 where @xmath272 is the angle between the electron velocity and the magnetic field , @xmath273 is the electron distribution function over energy and angle , @xmath274 , and @xmath275 is the mcdonald function . integrating equation ( 12 ) over angle under the assumption of a chaotic magnetic field orientation , we obtain the equation @xmath276 where @xmath277 is the electron distribution function as a function of energy and position and @xmath278 is the averaged magnetic field .
high energy electrons colliding with photons result in high energy photon production ( e.g. , jones 1968 ; gaisser 1998 ; sturner 1997 ; baring 1999 ) .
if the electron energy @xmath279ev ( @xmath240 is the energy of the emitted photon ) , then the formulae in the thompson limit are ( e.g. , ginzburg 1979 ) : @xmath280 where @xmath281 is the thompson cross section , @xmath282 is the background photon energy spectrum , and @xmath283 is the incoming electron flux .
if the electron energy @xmath284ev , we must use the klein - nishina formulae ( integrated over angles ) : @xmath285 @xmath286,\ ] ] @xmath287 in the above , @xmath245 , @xmath288 , @xmath289 is the electron lorentz factor , and @xmath290 is the classical electron radius .
the interstellar background photon spectrum , @xmath291 , was adopted from mezger ( 1982 ) and mathis ( 1983 ) taking into account infrared data from saken ( 1992 ) for the case of ic443 ( see also gaisser 1998 ) .
radio waves propagating in the partially ionized thermal plasma are subject to thermal free - free absorption . for @xmath292 ( the rayleigh
jeans regime ) , the free - free absorption coefficient @xmath293 ( cm@xmath1 ) can be calculated from @xmath295 using an approximate formula for the gaunt factor @xmath296 from rybicki & lightman ( 1979 ) appropriate to our parameter range , we obtained a simplified expression for the optical depth @xmath297 @xmath298 where the radio wave frequency @xmath299 is measured in units of 100 mhz , the plasma temperature in units of 100 k is @xmath300 , and the emission measure @xmath301 is in units of @xmath172 pc .
the primary ionization rate @xmath210 due to electron impact was calculated from the equation : @xmath302 where @xmath283 is the electron flux measured in units s@xmath1 @xmath20 kev@xmath1 sr@xmath1
, @xmath303 is the differential cross section for hydrogen atom primary ionization by electron - atom interactions integrated over atomic electron states , and @xmath239 is the minimum energy of the ionizing incoming electron . in the regime below 1 kev , we obtained the following fit to the ionization cross section @xmath303 using experimental data from fite & brackmann ( 1958 ) : @xmath304 where @xmath305 and @xmath306(ev ) is the incoming electron energy measured in ev .
the fitting coefficients are @xmath307=76.11 , @xmath308=14.34 , @xmath309=0.89 , @xmath310=3.82 , @xmath311=2.55 , and @xmath312=4.5 .
@xmath313 is the bohr radius . at high energies ( @xmath174 kev )
, the born approximation provides good accuracy ( e.g. , mott & massey 1965 ) : @xmath314 where @xmath315 } { \left [ \mu^4 + 2\mu^2 \left ( \rule{0em}{2ex } k^2+\kappa^2
\right ) + \left ( \rule{0em}{2ex } k^2-\kappa^2\right)^2 \right]^3 } \nonumber \\ & \times & \exp \ !
\left [ -\frac{2\mu}{\kappa}\ ; { \rm arctg}\ !
\left ( \frac{2\mu\kappa}{k^2-\kappa^2+\mu^2 } \right ) \right ] \frac1{1-\exp \ !
\left ( -2\pi \mu / \kappa \right ) } , \end{aligned}\ ] ] @xmath316 , @xmath317 is the incoming electron wave number , @xmath318 is the outgoing electron wave number , and @xmath319 is the emitted atomic electron wave number .
in addition , @xmath320 , @xmath321 , and @xmath322 .
we found a convenient approximation providing a reasonably accurate fit ( better than 10% above 1 kev and better than 5% above 3 kev ) to equations ( b3 ) and ( b4 ) : @xmath323 where @xmath324=97.72 and @xmath325=0.93 .
the gas ionization fraction @xmath326 ( we consider here a hydrogen plasma ) can be expressed through the ionization rate @xmath327 and recombination rate @xmath328 as : @xmath329 where @xmath330 .
we used the radiative recombination rates given by spitzer ( 1978 ) and dissociative recombination rates from schneider ( 1994 ) .
asvarov , a.i . ,
guseinov , o.h . ,
kasumov , f.k . , & dogiel , v.a .
1990 , a&a , 229 , 196 axford , w.i . 1981 , in proc .
17th icrc ( paris ) , 12 , 155 baier , v.n .
, fadin , v.s . , & khoze , v.a .
1967 , sov .
jetp , 24 , 760 baring , m.g . , ellison , d.c .
, reynolds , s.p . , grenier , i.a . , & goret , p. 1999 , apj , 513 , 311 baring , m.g . ,
ogilvie , k.w . ,
ellison , d.c .
, & forsyth , r.j .
1997 , apj , 476 , 889 berezhko , e.g. & ellison , d.c .
1999 , apj , 526 , 385 berezinsky , v.s . ,
bulanov , s.v . ,
dogiel , v.a .
, ginzburg , v.l . , & ptuskin , v.s .
1990 , astrophysics of cosmic rays ( amsterdam : north holland ) bertschinger , e. 1986 , apj , 304 , 154 blandford , r. , & eichler , d. 1987 , physics reports , 154 , 2 blandford , r.d . , & cowie , l.l .
1982 , apj , 260 , 625 blitz , l. 1993 , in protostars and planets iii , ed .
levy & j.i .
lunine , ( tucson : univ . of arizona ) , 125 blondin , j. m. , wright , e. b. , borkowski , k. j. , & reynolds , s. p. 1998 , apj , 500 , 342 boulares , a. , & cox , d.p .
1988 , apj , 333 , 198 buckley , j.h . ,
akerlof , c.w .
, carter - lewis , d.a .
1998 , a&a , 329 , 639 burton , m.g . , hollenbach , d.j . ,
haas , m.r . , & erickson , e.f .
1990 , apj , 355 , 197 bykov , a.m. , & toptygin , i.n .
1993 , physics uspekhi , 36 , 1020 bykov , a.m. , & uvarov , yu.a . 1999 , jetp , 88 , 465 cesarsky , d. , cox , p. , pineau des forets , g. , van dishoeck , e.f . , boulanger , f. , & wright , c.m . 1999 ,
a&a , 348 , 945 chevalier , r.a .
1977 , apj , 213 , 52 chevalier , r.a . 1999 , apj , 511 , 798 chevalier , r.a .
, & imamura , j.n .
1982 , apj , 261 , 543 claussen , m.j . ,
frail , d.a . ,
goss , w.m .
, & gaume , r.a .
1997 , apj , 489 , 143 cox , d.p . , shelton , r.l .
, maciejewski , w. , smith , r.k . ,
plewa , t. , pawl , a. , & ryczka , m. 1999 , apj , 524 , 179 de jager , o.c . , & mastichiadis , a. 1997 , apj , 482 , 874 denoyer , l.k .
1979 , apj , 232 , l165 dermer , c.d .
1986 , a&a , 157 , 223 draine , b.t . , &
mckee c.f .
1993 , ara&a , 31 , 373 drury , l.oc . , aharonian , f. , & vlk , h. 1994 , a&a , 267 , 959 drury , l.oc . ,
duffy , p. , & kirk , j.g 1996 , a&a , 309 , 1002 duin , r. m. , & van der laan , h. 1975 , a&a , 40 , 111 ellison , d.c .
, jones , f.c . , & baring , m.g . 1999 ,
apj , 512 , 403 erickson , w.c . , & mahoney , m.j .
1985 , apj , 290 , 596 fedyushin , b.k . 1952 , zhurn .
, 22 , 140 fesen , r.a .
, & kirshner , r.p .
1980 , apj , 242 , 1023 fisk , l.a . ,
schwadron , n.a . , & gloeckler , g. 1997 , geophys .
lett , 24 , 93 fite , w. l. , & brackmann , r. t. 1958 , phys .
, 112 , 1141 frail , d.a . , &
mitchell , g.f .
1998 , apj , 508 , 690 gaisser , t.k . ,
protheroe , r.j . , & stanev , t. 1998 , apj , 492 , 219 garibyan , g.m .
1952 , zhurn .
fiz . , 24 , 617 giacalone , j. , burgess , d. , schwartz , s.j . ,
ellison , d.c , & bennett , l. 1997 , j. geoph .
, 102 , 19789 harrus , i. , hughes , j.p . ,
singh , k.p .
, koyama , k. , & asaoka , i. 1997 , apj , 488 , 781 haug e. 1975 , z. naturforsch . , 30a , 1099 hollenbach , d. , & mckee , c.f .
1989 , apj , 342 , 306 jones , f.c . 1968 , phys .
rev . , 167 , 1159 jones , f.c . , & ellison , d.c .
1991 , space sci .
, 58 , 259 jones , t.w .
, & kang , h. 1993 , apj , 402 , 560 keohane , j.w .
, petre , r. , gotthelf , e.v . , ozaki , m. , & koyama , k. 1997 , apj , 484 , 350 koch , h.w . , & motz , j.w .
1959 , rev .
phys . , 31 , 920 koyama k. , petre r. , gotthelf , e.v . , hwang , u. , matsuura , m. , ozaki , m. , & holt , s.s .
1995 , nature , 378 , 255 kulsrud , r. , & cesarsky , c.j . 1971 , astroph .
lett . , 8 , 189 levinson , a. 1996 , mnras , 278 , 1018 mathis , j.s . ,
mezger , p.g . , & panagia , n. 1983 , a&a , 128 , 212 mcclements , k.g . , dendy , r.o . ,
bingham , r. , kirk , j.g . , & drury , l.oc .
1997 , mnras , 291 , 241 mezger , p.g . ,
mathis , j.s .
, & panagia , n. 1982 , a&a , 105 , 372 mbius , e. , rucinski , d. , lee , m.a . , & isenberg , p.a .
1998 , j.geophys .
res , 103 , 257 ostrowski , m. 1999 , a&a , 345 , 256 petre , r. , szymkowiak , a.e . ,
seward , f.d .
, & willingale , r. 1988 , apj , 335 , 215 reach , w.t . , & rho , j. 1999 , apj , 511 , 836 reynolds , s.p .
1998 , apj , 493 , 375 richter , m.j . ,
graham , j.r .
, & wright , g.s .
1995 , apj , 454 , 277 rho , j. , & petre , r. 1998 , apj , 503 , l167 rybicki , g.b .
, & lightman , a.p .
1979 , radiative process in astrophysics ( new york : wiley ) saken , j.m . ,
fesen , r.a . , & shull , j.m .
1992 , apjs , 82 , 715 schneider , i.f . , dulieu , o. , giusti - suzor , a. , & roueff , e. 1994 , apj , 424 , 983 schnfelder , v. 1999 , in libeb , cosmic rays and related x- and gamma - rays , ed .
r. ramaty , e. vangioni - flam , m. cass , & k. olive , ( san francisco : asp ) , 217 shelton , r.l . ,
cox , d.p .
, maciejewski , w. , smith , r.k . ,
plewa , t. , pawl , a. , & ryczka , m. 1999 , apj , 524 , 192 shull , j.m . , & mckee , c.f .
1979 , apj , 227 , 131 spitzer , l. 1978 , physical processes in the interstellar medium ( new york : wiley ) sturner , s.j . ,
skibo , j.g .
, dermer , c.d . , & mattox , j.r .
1997 , apj , 490 , 619 van dishoeck , e.f . ,
jansen , d.j .
, & phillips , t.g .
1993 , a&a 279 , 541 vishniac , e.t .
1983 , apj , 274 , 152 vlk , h.j . , morfill , g.e . ,
& forman , m.a .
1981 , apj , 249 , 161 wang , z.r . , asaoka , i. , hayakawa , s. , & koyama , k. 1992 , pasj , 44 , 303 webber , w. r. 1983 , in composition and origin of cosmic rays , ed . m. m. shapiro ( dordrecht : reidel ) , 83 winkler , c. 1999 , astroph .
lett . & communic . , 39 , 309 | in evolved supernova remnants ( snrs ) interacting with molecular clouds , a highly inhomogeneous structure consisting of a forward shock of moderate mach number , a cooling layer , a dense radiative shell and an interior region filled with hot tenuous plasma is expected .
we present a model of nonthermal electron injection , acceleration and propagation in that environment and find that these snrs are efficient electron accelerators and sources of hard x- and emission .
a forward shock of velocity @xmath0 100 km s@xmath1 with an ionized precursor propagating into the molecular cloud accompanied by magnetohydrodynamic turbulence provides a spatially inhomogeneous distribution of nonthermal electrons .
the energy spectrum of the nonthermal electrons is shaped by the joint action of first and second order fermi acceleration in a turbulent plasma with substantial coulomb losses .
bremsstrahlung , synchrotron , and inverse compton radiation of the nonthermal electrons produce multiwavelength photon spectra in quantitative agreement with the radio and the hard emission observed by _
asca _ and _ egret _ from ic 443 .
we distinguish interclump shock wave emission from molecular clump shock wave emission ; particles reach higher energies in the interclump shock and that is the likely source of @xmath2-ray emission and radio synchrotron emission .
spatially resolved x- and spectra from the supernova remnants ic 443 , w44 , and 3c391 as might be observed with _
bepposax , chandra xro , xmm , integral _ and _ glast _ would distinguish the contribution of the energetic lepton component to the @xmath2-rays observed by _
egret _ , constraining the cosmic ray nuclear component spectra in these snrs .
these data would provide a valuable tool for studying the complex structure of molecular clouds where snr radiative shocks interact with dense molecular clumps . |
-.05 in there are a number of wonderful sources available on hyperplane arrangements , most notably orlik - terao s landmark 1992 text @xcite . in the last decade alone
several excellent surveys have appeared : suciu s paper on aspects of the fundamental group @xcite , yuzvinsky s paper on orlik - solomon algebras and local systems cohomology @xcite , and several monographs devoted to connections to areas such as hypergeometric integrals @xcite , mathematical physics @xcite , as well as proceedings from conferences at sapporo @xcite , northeastern @xcite and istanbul @xcite .
the aim of this note is to provide an overview of some recent results and open problems , with a special emphasis on connections to computation .
the paper also gives a concrete and example driven introduction for non - specialists , but there is enough breadth here that even experts should find something new .
there are few proofs , but rather pointers to original source material .
we also explore connections of arrangements to other areas , such as de concini - procesi wonderful models , the feichtner - yuzvinsky algebra of an atomic lattice , the orlik - terao algebra and blowups , and plane curve singularities .
all computations in this survey can be performed using macaulay2 @xcite , available at : http://www.math.uiuc.edu/macaulay2/ , and the arrangements package by denham and smith @xcite .
let @xmath0 , and let @xmath1 be the symmetric algebra on @xmath2 : @xmath3 is a @xmath4-graded ring , which means that if @xmath5 and @xmath6 , then @xmath7 .
a graded @xmath1-module @xmath8 is defined in similar fashion .
of special interest is the case where @xmath9 is a field @xmath10 , so that each @xmath11 is a @xmath10vector space .
the free @xmath1 module with generator in degree @xmath12 is written @xmath13 , and in general @xmath14 .
the hilbert function @xmath15 the hilbert series @xmath16 [ exm : first ] @xmath17 $ ] , @xmath18 .
then @xmath19 & @xmath20 & @xmath19 + @xmath20 & @xmath21 & @xmath19 + @xmath22 & @xmath23 & @xmath20 + @xmath24 & @xmath25 & @xmath21 + @xmath26 & @xmath27 & @xmath23 + @xmath28 & @xmath29 & @xmath30 + the respective hilbert series are @xmath31 an induction shows that @xmath32 ; this makes it easy to compute the hilbert series of an arbitrary graded module from a free resolution . for @xmath33 , a minimal free resolution is @xmath34 } s(-2)^2 \xrightarrow{\left [ \!\begin{array}{cc } x^2 & xy \end{array}\ ! \right ] } s \longrightarrow s / i \longrightarrow 0.\ ] ] the map @xmath35 $ ] sends @xmath36 so in order to have a map of graded modules , the basis elements of the source must have degree two , explaining the shifts in the free resolution . taking the alternating sum of the hilbert series yields @xmath37 which agrees with the previous computation .
[ exm : twistedcubic ] the @xmath38 minors of @xmath39 $ ] define the twisted cubic @xmath40 $ ] .
@xmath41 } s(-2)^3 \xrightarrow{\left [ \!\begin{array}{ccc } y^2\!-\!xz & yz\!-\!xw & z^2\!-\!yw \end{array}\ ! \right ] } s \longrightarrow s / i \ ] ] the numerical information in a free resolution may be compactly displayed as a _
betti table _ : @xmath42 @xmath43 -.1 in in particular , the indexing begins at position @xmath44 and is read over and down .
so for the twisted cubic , @xmath45 we now give a quick review of arrangements .
let @xmath46 be an arrangement of complex hyperplanes in @xmath47 .
we assume @xmath48 is _ central _ and _ essential _ : the @xmath49 with @xmath50 are homogeneous , and the common zero locus @xmath51 .
the central condition means that @xmath48 also defines an arrangement in @xmath52 .
the main combinatorial object associated to @xmath48 is the intersection lattice @xmath53 , which consists of the intersections of elements of @xmath48 , ordered by reverse inclusion .
@xmath54 is the lattice element @xmath55 and the rank one elements of @xmath53 are the hyperplanes .
the mbius function @xmath56 : @xmath57 is defined by @xmath58 the poincar and characteristic polynomials of @xmath48 are defined as @xmath59 [ exm : nonfano ] the @xmath60 arrangement is @xmath61 .
projecting along @xmath62 gives a central arrangement in @xmath63 , hence a configuration of lines in @xmath64 .
this configuration corresponds to the figure below , but with the line at infinity ( which bounds the figure ) omitted . for the @xmath65 rank two elements of @xmath66 , the four corresponding to triple points have @xmath67 , and
the three normal crossings have @xmath68 .
thus , @xmath69 . adding the bounding line gives the non - fano arrangement nf , with @xmath70 .
in @xcite , orlik and solomon showed that the cohomology ring of the complement @xmath71 has presentation @xmath72 , with generators @xmath73 in degree @xmath20 and @xmath74 for additional background on arrangements , see @xcite .
let @xmath76 be a central arrangement . for each @xmath12 , fix @xmath77 , and define @xmath78 $ ] .
[ da ] the module of @xmath48-derivations ( or terao module ) is the submodule of @xmath79 consisting of vector fields tangent to @xmath48 : @xmath80 an arrangement is free when @xmath75 is a free @xmath1module . in this case , the degrees of the generators of @xmath75 are called the exponents of the arrangement . note that @xmath75 is always nonzero , since the euler derivation @xmath81 .
it is easy to show that @xmath82 where @xmath83 is the jacobian ideal of @xmath84 , and @xmath85 denotes the module of syzygies on @xmath83 : polynomial relations on the generators of @xmath83 .
@xmath86 is free iff there exist @xmath87 elements @xmath88 such that @xmath89 ) = c \cdot q_{{{\mathcal{a}}}}$ ] , for some @xmath90 .
[ exm : nonfanoii ] for example [ exm : nonfano ] , a computation shows that @xmath91 interestingly , the respective poincar polynomials factor , as @xmath92 this suggests the possibility of a connection between the exponents of a free arrangement and the poincar polynomial .
a landmark result in arrangements is : if @xmath93 , then @xmath94 [ exm : stanley][stanley ] for @xmath48 below , @xmath95 .
a computation shows that @xmath48 is not free , so factorization of @xmath96 is a necessary but not sufficient for freeness of @xmath48 .
a famous open conjecture in the field of arrangements is : if @xmath97 , then freeness of @xmath75 depends only on @xmath98 .
[ exm : zieglerab][ziegler s pair @xcite ] let @xmath48 be an arrangement of @xmath99 lines in @xmath64 , as below .
then @xmath75 depends on nonlinear geometry : if the six triple points lie on a smooth conic , we compute : @xmath100 & s(-7 ) \oplus s(-8 ) \ar[r ] & s(-5 ) \oplus s^3(-6 ) \ar[r ] & syz(j_{{{\mathcal{a } } } } ) \ar[r ] & 0},\ ] ] while if six triple points are not on a smooth conic , the resolution is:.06 in @xmath101 & s^4(-7 ) \ar[r ] & s^6(-6 ) \ar[r ] & syz(j_{{{\mathcal{a } } } } ) \ar[r ] & 0}.$ ] a version of terao s theorem applies to any arrangement : [ dp ] @xmath102 consists of @xmath103 such that @xmath104 [ solt ] @xmath105 relate the modules @xmath106 , for @xmath107 , to @xmath98 .
a closed subarrangement @xmath108 is a subarrangement such that @xmath109 for some flat @xmath110 .
the best result relating @xmath75 to @xmath98 is : if @xmath111 is a closed subarrangement , then @xmath112 .
find bounds on @xmath113 depending on @xmath98 a particularly interesting class of arrangements are graphic arrangements , which are subarrangements of @xmath114 .
given a simple ( no loops or multiple edges ) graph @xmath115 , with @xmath116 vertices and edge set @xmath117 , we define @xmath118 @xmath119 is supersolvable iff @xmath115 is chordal . if @xmath119 has an induced @xmath120-cycle , then @xmath121 .
the largest induced cycle of @xmath115 below is a 6-cycle .
a computation shows @xmath122 .
the largest induced cycle of @xmath115 below is a 4-cycle .
a computation shows @xmath123 .
find a formula for @xmath124 .
[ triple ] a triple @xmath125 of arrangements consists of a choice of @xmath126 , with @xmath127 a main tool for proving freeness is terao s addition - deletion theorem .
[ ad ] for a triple , any two of the following imply the third 1 .
2 . @xmath129 3 .
@xmath130 in example [ exm : nonfano ] , the @xmath60 arrangement is free with exponents @xmath131 .
let @xmath132 be the line at infinity , which meets @xmath60 in four points .
then @xmath133 is free , with exponents @xmath134 , so the non - fano arrangement is free with exponents @xmath135 , which agrees with our earlier computation .
example 4.59 of @xcite gives a free arrangement for which the addition - deletion theorem does not apply . as a corollary of theorem [ ad ] , terao showed that supersolvable arrangements are free .
[ ssolve ] an element @xmath110 of a lattice is modular if for all @xmath136 and all @xmath137 , @xmath138 .
a central arrangement @xmath48 is supersolvable if there exists a maximal chain @xmath139 of modular elements in @xmath140 . for line configurations in @xmath64 ,
the supersolvability condition simply means there is a singular point @xmath141 such that every other singularity of @xmath48 lies on a line of @xmath48 which passes through @xmath142 .
for example , the @xmath60 arrangement is supersolvable , since any triple point is such a singularity . for arrangements in @xmath64
, there is a beautiful characterization of freeness involving multiarrangements .
a multiarrangement @xmath143 consists of an arrangement @xmath48 , along with a multiplicity @xmath144 for each @xmath126 . @xmath145
@xmath146 is free if and only if 1 .
@xmath147 and 2 .
@xmath148 , where @xmath149 holds for all @xmath150 , with @xmath151 the necessity of these conditions was shown by ziegler in @xcite , and sufficiency was proved by yoshinaga in @xcite . in @xcite , yoshinaga gives a generalization to higher dimensions .
the exponents of free multiarrangements are not combinatorial : consider the two multiarrangements in @xmath152 , with underlying arrangements defined by @xmath153 with @xmath154 . to compute @xmath155
, we must find all @xmath156 such that @xmath157 thus , @xmath155 is the kernel of the matrix @xmath158.$ ] computations show that @xmath155 has exponents @xmath159 , and @xmath160 has exponents @xmath161 . there is an analog of theorem [ solt ] for multiarrangements . [ dp2
] @xmath162 consists of @xmath103 such that @xmath163 define @xmath164 if @xmath165 then @xmath166 in @xcite , abe - terao - wakefield prove an addition - deletion theorem for multiarrangements by introducing _
euler multiplicity _ for the restriction .
it follows from the hilbert - burch theorem that any @xmath167 is free , which leads to the question of whether there exist other arrangements which are free for any @xmath168 . in @xcite , abe - terao - yoshinaga
prove that any such arrangement is a product of one- and two dimensional arrangements .
nevertheless , several natural questions arise : characterize the projective dimension of @xmath169 .
define supersolvability for multiarrangements .
for a collection of hypersurfaces @xmath170 the module of derivations @xmath171 is obtained by substituting @xmath172 for @xmath173 in definition [ da ] .
it is not hard to prove that saito s criterion still applies .
are there other freeness theorems ? for the arrangement @xmath174 depicted below we compute that @xmath175 .
this example can be explained by an addition - deletion theorem @xcite , but there is subtle behavior related to singular points . for the remainder of this section , @xmath176 is reduced plane curve , and if @xmath177 is a singular point , translate so @xmath178 . [ qhomogsing ]
a plane curve singularity is quasihomogeneous if and only if there exists a holomorphic change of variables so that @xmath179 is weighted homogeneous : there exists @xmath180 such that @xmath181 is homogeneous .
the milnor number at @xmath44 is @xmath182 the tjurina number at @xmath44 is @xmath183 for a projective plane curve @xmath184 , it is easy to see that the degree of @xmath185
. let @xmath186 be as below : if @xmath142 is an ordinary singularity with @xmath120 distinct branches , then @xmath187 , so the sum of the milnor numbers is @xmath188 .
however , a computation shows that @xmath189 .
all singularities are ordinary , but the singularity at the origin is not quasihomogeneous .
if @xmath190 has an isolated singularity at the origin , then @xmath191 iff @xmath192 is quasihomogeneous . for arrangements of lines and conics
such that every singular point is quasihomogeneous , @xcite proves an addition / deletion theorem ; @xcite generalizes the result to curves of higher genus .
let @xmath186 be as below : @xmath171 has exponents @xmath131 , which can be shown using the aforementioned addition - deletion theorem .
change @xmath186 to @xmath193 via : @xmath194 a computation shows that @xmath195 is not free .
thus , for line - conic arrangements , freeness is not combinatorial .
define supersolvability for hypersurface arrangements
. give combinatorial bounds on @xmath196 .
analyze associated primes and @xmath197 modules of @xmath171 .
the orlik terao algebra is a symmetric analog of the orlik - solomon algebra .
while the orlik - solomon algebra records the existence of dependencies among sets of hyperplanes , the orlik - terao algebra records the actual dependencies .
if @xmath198 , then there exist @xmath199 with @xmath200 the orlik - terao ideal @xmath201 the orlik - terao algebra is @xmath202/i_{{{\mathcal{a}}}}.$ ] [ otexm1 ] @xmath203 , the only dependency is @xmath204 , so @xmath205 in @xcite , orlik and terao answer a question of aomoto by considering the artinian quotient @xmath206 of @xmath207 by @xmath208 .
they prove : @xmath209 .
[ teraoca ] @xmath210 it is not hard to show that @xmath211\stackrel{\phi}{\rightarrow } { { \mathbb{k}}}\bigg[\frac{1}{l_1},\ldots , \frac{1}{l_n}\bigg ] \rightarrow 0\ ] ] is exact , so @xmath212 is irreducible and rational . in any situation where weights of dependencies play a role ,
the orlik - terao algebra is the natural candidate to study .
one such situation involves 2-formality : @xmath48 is 2-formal if all dependencies are generated by dependencies among three hyperplanes .
if @xmath48 is @xmath213 , @xmath48 is 2-formal .
if @xmath48 is free , @xmath48 is 2-formal .
one reason that formality is interesting is that it is not a combinatorial invariant : in example [ exm : zieglerab ] , the arrangement for which the six triple points lie on a smooth conic is not 2-formal , and the arrangement for which the points do not lie on a smooth conic is 2-formal .
[ st2formal ] @xmath48 is 2-formal iff @xmath214 in @xcite , brandt and terao generalized the notion of 2-formality to @xmath215formality : @xmath48 is k - formal if all dependencies are generated by dependencies among @xmath216 or fewer hyperplanes .
brandt and terao prove that every free arrangement is @xmath215formal .
find an analog of theorem [ st2formal ] for @xmath215formality .
the configuration of example [ otexm1 ] consists of four generic lines : -.1 in -.1 in the orlik - terao ideal defines a cubic surface in @xmath217 , and a computation shows that @xmath218 has four singular points .
this can be interpreted in terms of a rational map .
let @xmath219 , and define @xmath220 $ ] .
@xmath221 restrict to the case @xmath222 , and let @xmath223 denote the blowup of @xmath64 at the singular points of @xmath48 , with @xmath224 denoting the pullback to @xmath225 of the class of a line on @xmath64 , and @xmath226 the exceptional divisors over singular points of @xmath48 .
let @xmath227 utilizing results of proudfoot - speyer @xcite showing that @xmath207 is cohen - macaulay and the riemann - roch theorem , @xcite shows that the map @xmath228 is determined by the global sections of @xmath229 , and that @xmath228 1 .
is an isomorphism on @xmath230 ) 2 .
contracts the lines of @xmath48 to points 3 .
blows up the singularities of @xmath48 .
[ cmregularity ] a graded @xmath1-module @xmath231 has castelnuovo - mumford regularity @xmath232 if @xmath233 for all @xmath234 and all @xmath235 . in terms of the betti table ,
the regularity of @xmath231 is the label of the last non - zero row , so in example [ exm : twistedcubic ] , @xmath236 has castelnuovo - mumford regularity one .
the regularity of @xmath207 is determined in @xcite : [ cmregofca ] for @xmath237 , @xmath207 is @xmath238regular . to see this , note that since @xmath207 is cohen - macaulay , quotienting @xmath207 by @xmath116 generic linear forms yields an artinian ring whose hilbert series is the numerator of the hilbert series of @xmath207 .
the regularity of an artinian module is equal to the length of the module , so the result follows from theorem [ teraoca ] .
a main motivation for studying @xmath207 is a surprising connection to nets and resonance varieties , which are the subject of [ resonancesection ] .
first , the definition of a net : [ net ] let @xmath239 .
a @xmath120-net in @xmath64 is a partition of the lines of an arrangement @xmath48 into @xmath120 subsets @xmath240 , together with a choice of points @xmath241 , such that : 1 . for every @xmath242 and every @xmath243 , @xmath244 .
2 . @xmath245 @xmath246 and every @xmath247 , there exists a unique @xmath248 with @xmath249 . in @xcite , libgober and yuzvinsky
show that nets are related to the first resonance variety @xmath250 . the definition of a net forces each subset @xmath240 to have the same cardinality , and if @xmath251 , the net is called a @xmath252-net .
using work of @xcite and @xcite , it is shown in @xcite that [ encpx ] existence of a @xmath252 net implies that there is a decomposition @xmath253 with @xmath254 and @xmath255 .
[ 1generic ] a matrix of linear forms is @xmath20-generic if it has no zero entry , and can not be transformed by row and column operations to have a zero entry . in @xcite , eisenbud shows that if a divisor @xmath256 on a smooth curve @xmath110 factors as @xmath257 , with @xmath258 having @xmath259sections and @xmath260 having @xmath28sections , then the ideal of the image of @xmath110 under the map defined by the global sections of @xmath256 will contain the @xmath261 minors of a @xmath20-generic matrix . using this result and theorem [ encpx ]
, it can be shown that @xmath262 contains the ideal @xmath263 of @xmath261 minors of a @xmath20-generic @xmath264 matrix @xmath8 .
so if @xmath265 , the eagon - northcott complex @xcite @xmath266 is a subcomplex of resolution of @xmath267 .
the geometric content of theorem [ encpx ] is that it implies @xmath218 lies on a _ scroll _ @xcite .
[ showmethm ] for the @xmath60 arrangement , the set of triple points @xmath268 gives a @xmath269 net , where the @xmath270 correspond to normal crossing points : @xmath271 , @xmath272 , @xmath273 .
@xmath274 so @xmath275 and @xmath276 contains the @xmath277 minors of a @xmath278 matrix , whose resolution appears in example [ exm : twistedcubic ] .
the graded betti diagram for @xmath279/i_{a}$ ] is @xmath280 from this , it follows that the free resolution of @xmath281 is a mapping cone resolution @xcite .
the geometric meaning is that @xmath282 is the intersection of a generic quadric hypersurface with the scroll .
since @xmath229 contracts proper transforms of lines to points , it is not very ample .
however , it follows from @xcite that @xmath283 is very ample , and gives a de - concini - procesi wonderful model ( see next section ) for the blowup .
determine the graded betti numbers of @xmath207 .
relate @xmath284 to the graded betti numbers of @xmath207 .
in @xcite , fulton - macpherson provide a compactification @xmath285 for the configuration space of @xmath28 marked points on an algebraic variety @xmath110 .
the construction is quite involved , but the combinatorial data is that of @xmath114 . in a related vein , in @xcite , de concini - procesi construct a wonderful model @xmath110 for a subspace complement @xmath286 : a smooth , compact @xmath110 such that @xmath287 is a normal crossing divisor .
here it is the combinatorics which are complex .
a key object in their construction is @xmath288 where @xmath115 is a building set . in @xcite , feichtner - kozlov generalize the construction of @xcite to a purely lattice - theoretic setting .
see @xcite for additional background on this section .
[ definition : nestset ] for a lattice @xmath289 , a building set @xmath115 is a subset of @xmath289 , such that for all @xmath290 , @xmath291 satisfies @xmath292 \simeq \prod_{j=1}^m [ \hat{0},x_j].$ ] a building set contains all irreducible @xmath290 . [ definition : nestset ] a subset @xmath231 of a building set @xmath115 is nested if for any set of incomparable @xmath293 with @xmath294 , @xmath295 exists in @xmath289 , but is not in @xmath115 .
nested sets form a simplicial complex @xmath296 , with vertices the elements of @xmath115 ( which are vacuously nested ) .
[ minbldgset ] the minimal building set for @xmath60 consists of the hyperplanes themselves , the triple intersections in @xmath297 , and the element @xmath298 .
since @xmath298 is a member of every face of @xmath296 , the nested set complex @xmath296 is the cone over there is an edge @xmath299 because there are no incomparable subsets with at least two elements , while @xmath300 is an edge because @xmath301 exists in @xmath289 ( it is a normal crossing ) , but is not in @xmath115 .
suppose @xmath289 is an atomic lattice , and @xmath115 a building set in @xmath289 . in @xcite , feichtner and yuzvinsky study a certain algebra associated to the pair @xmath302 : @xmath303/i , \mbox { with } x_g \mbox { of degree } 2.\ ] ] where @xmath276 is generated by @xmath304 [ thm : atomiclatticealg ] if @xmath48 is a hyperplane arrangement and @xmath115 a building set containing @xmath298 , then @xmath305 where @xmath306 is the wonderful model arising from the building set @xmath115 .
the importance of this is the relation to the knudson - mumford compactification @xmath307 of the moduli space of @xmath28 marked points on @xmath152 .
[ dpthm ] @xmath308 where @xmath115 is the minimal building set for @xmath309 .
a presentation for the cohomology ring of @xmath307 was first described by keel in @xcite ; the description which follows from @xcite is very economic . by theorem [ thm : atomiclatticealg ] and
@xcite , @xmath310 the nested set complex for @xmath60 and @xmath311 appears in example [ minbldgset ] , so that @xmath312 is the quotient of a polynomial ring @xmath1 with eleven generators by an ideal consisting of 6 linear forms ( one form for each hyperplane ) and 19 quadrics . to see that there are 19 quadrics , note that the space of quadrics in @xmath313-variables has dimension @xmath314 , and @xmath315 has @xmath316 edges ( recall that @xmath298 is not pictured ) .
a computation shows that @xmath317/i,\ ] ] where @xmath276 consists of all but one quadric of @xmath1 ( and includes all squares of variables ) .
this meshes with the intuitive picture : to obtain a wonderful model , simply blow up the four triple points , so that @xmath318 is the corresponding del pezzo surface @xmath319 , which has @xmath320 , agreeing with the computation .
analyze @xmath321 for other lattices .
let @xmath115 be a finitely - generated group , with normal subgroups , @xmath323 defined inductively by @xmath324 $ ] .
we obtain an associated lie algebra @xmath325 with lie bracket induced by the commutator map .
let @xmath326 denote the rank of the @xmath120-th quotient .
presentations for @xmath327 are given by randell @xcite , salvetti @xcite , arvola @xcite , and cohen - suciu @xcite .
for computations , the braid monodromy presentation of @xcite is easiest to implement . for a detailed discussion of @xmath327 ,
see suciu s survey @xcite .
the fundamental group is quite delicate , and in this section , we investigate properties of @xmath327 via the associated graded lie algebra @xmath328 the lefschetz - type theorem of hamm - le @xcite implies that taking a generic two dimensional slice gives an isomorphism on @xmath322 .
thus , to study @xmath327 , we may assume @xmath329 or @xmath64 .
as shown by rybnikov @xcite , @xmath327 is not determined by @xmath98 ; whereas the orlik - solomon algebra @xmath330 is determined by @xmath98 .
[ a3hs ] in example [ exm : nonfano ] , we saw that the hilbert series for @xmath60 is @xmath331 .
a computation shows that the lcs ranks begin @xmath332 for higher @xmath120 , @xmath333 , where @xmath334 is a witt number . in general
, we may encode the lcs ranks via @xmath335 for @xmath60 , this is @xmath336 expanding this and writing out the first few terms yields @xmath337 .1 in if we multiply this with @xmath338 the result is @xmath20 , and is part of a general pattern . for the arrangement @xmath339 ( graphic arrangement of @xmath340 ) @xmath341 this explains the computation of example [ a3hs ] .
we now compute the free resolution of the residue field @xmath342 as an @xmath258-module , where @xmath343 .
let @xmath344 [ a3tors ] for @xmath60 , we compute @xmath345 if @xmath242 , and @xmath346 the @xmath347 are the coefficients of the formal power series in example [ a3hs ] ! kohno s result was the first of a long line of results on lcs formulas for certain special families of arrangements 1 .
braid arrangements : kohno @xcite 2 .
fiber type arrangements : falk randell @xcite 3 .
supersolvable arrangements : terao @xcite 4 .
lower bound for @xmath348 : falk @xcite 5 .
koszul arrangements : shelton
yuzvinsky @xcite 6 .
hypersolvable arrangements : jambu papadima @xcite 7 .
rational @xmath213 arrangements : papadima yuzvinsky @xcite 8 .
mls arrangements : papadima suciu @xcite 9 .
graphic arrangements : lima - filho
schenck @xcite 10 .
no such formula in general : peeva @xcite let @xmath349 denote the free lie algebra on @xmath350 dualizing the cup product gives a map @xmath351 following chen @xcite , define the holonomy lie algebra @xmath352 where @xmath262 is the lie ideal generated by @xmath353 . as noted by kohno in @xcite , taking transpose of cup product shows that the image of @xmath354 is generated by @xmath355,\ ] ] where @xmath356 is a generator of @xmath357 corresponding to @xmath358 , and the set @xmath359 is a maximal dependent set of codimension two , so corresponds to an element of @xmath360 .
the upshot is that @xmath361 this was first made explicit by peeva in @xcite ; the proof runs as follows .
first , brieskorn @xcite showed that @xmath362 is formal , in the sense of @xcite . using sullivan s work and an analysis of the bigrading on hirsch extensions , kohno proved @xmath363 .
thus 1 .
@xmath364 , which follows from kohno s work and poincar - birkhoff - witt .
shelton - yuzvinsky show in @xcite that @xmath365 is the quadratic dual of the quadratic orlik - solomon algebra .
3 . results of priddy - lfwall show that the quadratic dual is related to diagonal yoneda ext - algebra via @xmath366 results of peeva @xcite and roos @xcite show that in general there does not exist a standard graded algebra @xmath367 such that @xmath368 . for any quotient of a free lie algebra ,
can we : find spaces for which there is a simple generating function for @xmath348 .
relate @xmath369 to @xmath370 , as in @xcite . as shelton - yuzvinsky proved in @xcite , the natural class of arrangements for which an lcs formula holds are arrangements for which @xmath258 is a koszul algebra , which we tackle next .
let @xmath371 denote the tensor algebra on @xmath372 .
a quadratic algebra is @xmath373 , with @xmath374 . a quadratic algebra @xmath258 has a quadratic dual @xmath375 : @xmath376 @xmath258 is koszul if @xmath377 , @xmath378 .
a quadratic algebra @xmath258 is koszul iff the minimal free resolution of the residue field over @xmath258 has matrices with only linear entries .
the hilbert series of @xmath379 is @xmath380 , and a computation shows that the minimal free resolution of @xmath10 over @xmath1 is the koszul complex , so @xmath381 . since @xmath382 we see that @xmath383 .
the hilbert series of @xmath384 is @xmath385 .
a computation shows that @xmath386 , which are the coefficients in an expansion of @xmath380 .
frberg @xcite proved that if @xmath276 is a quadratic monomial ideal then @xmath236 is koszul . by uppersemicontinuity @xcite , this means @xmath236 is koszul if @xmath276 has a quadratic grobner basis ( qgb ) .
see example [ knoqgb ] below for a koszul algebra having no qgb .
both @xmath1 and @xmath384 are koszul , and the relation between their hilbert series is explained by : if @xmath258 is koszul , so is @xmath387 , and @xmath388 the orlik - solomon algebra has a qgb iff @xmath48 is supersolvable .
a computation shows that the orlik - solomon algebra of @xmath60 has a quadratic grobner basis , so is koszul . for the non - fano arrangement , @xmath389 , so @xmath258 is not koszul .
[ knoqgb ] map @xmath390 \stackrel{\phi}{\longrightarrow } { { \mathbb{k}}}[x , y , z]$ ] using all cubic monomials of @xmath391 $ ] except @xmath392 , and let @xmath393 . then @xmath394 is koszul , but has no quadratic grobner basis .
for orlik - solomon algebras , does koszul imply supersolvable ? in the case of graphic arrangements , it does @xcite . find a combinatorial description of @xmath395 .
let @xmath258 be the orlik - solomon algebra of @xmath362 , with @xmath396 .
for each @xmath397 , we consider the aomoto complex @xmath398 , whose @xmath399 term is @xmath270 , and differential is @xmath400 : @xmath401 & a_0 \ar[r]^{a } & a_1 \ar[r]^{a } & a_2 \ar[r]^{a } & \cdots \ar[r]^{a } & a_{\ell}\ar[r ] & 0}.\ ] ] this complex arose in aomoto s work @xcite on hypergeometric functions , as well as in the study of cohomology with local system coefficients @xcite , @xcite . in @xcite , yuzvinsky showed that for a generic @xmath402 , the aomoto complex is exact ; the resonance varieties of @xmath48 are the loci of points @xmath403 for which @xmath398 fails to be exact , that is : -.2 in for each @xmath404 , @xmath405 in @xcite , falk gave necessary and sufficient conditions for @xmath406 .
a partition @xmath407 of @xmath48 is neighborly if for all @xmath408 and @xmath409 a block of @xmath407 , @xmath410 falk proved that components of @xmath250 arise from neighborly partitions ; he conjectured that @xmath250 is a union of linear components .
this was established ( essentially simultaneously ) by cohen
suciu @xcite and libgober - yuzvinsky @xcite .
libgober and yuzvinsky also showed that @xmath411 is a disjoint union of positive dimensional subspaces in @xmath412 , and cohen - orlik @xcite show that @xmath413 is also a subspace arrangement . on the other hand , as shown by falk in @xcite , in positive characteristic , the components of @xmath411 can meet , and need not be linear .
the approach of libgober yuzvinsky involves connecting @xmath250 to pencils / nets / webs and there is much recent work in the area , e.g. @xcite , @xcite @xcite .
of special interest is the following conjecture relating @xmath250 and the lcs ranks @xmath348 : if @xmath414 , then @xmath415 where @xmath416 is the fourth chen rank ( definition [ cranks ] ) .
[ exm : toyres ] let @xmath417 , and @xmath418 , with generators @xmath419 , so that @xmath420 since @xmath421 , the second relation is unnecessary .
to compute @xmath250 , we need only the first two differentials in the aomoto complex . using @xmath422 as a basis for @xmath423 , we find that @xmath424 . since @xmath425 in @xmath258 , @xmath426 , so that @xmath427 this means @xmath428 ; similar computations for the other @xmath429 show that the aomoto complex is @xmath430 } { { \mathbb{k}}}^4 \xrightarrow{\left [ \!\begin{array}{cccc } a_2+a_3 & -a_1 & -a_1 & 0\\ a_4 & 0 & 0 & -a_1\\ -a_2 & a_1+a_3 & -a_2 & 0 \\ 0 & a_4 & 0 & -a_2 \\ 0 & 0 & a_4 & -a_3 \end{array}\ ! \right ] } { { \mathbb{k}}}^5\ ] ] the rank of the first map is always one , @xmath431 is the locus where the second matrix has kernel of dimension at least two , so the @xmath432 minors must vanish .
a computation shows this locus is @xmath433 . letting @xmath434 ,
observe that @xmath406 iff there exists a @xmath435 so that @xmath436 , so that @xmath250 is the locus of decomposable 2-tensors in @xmath437 .
since @xmath437 is determined by the intersection lattice @xmath140 in rank @xmath438 , to study @xmath250 , it can be assumed that @xmath222 . while the first resonance variety is conjecturally connected ( under certain conditions ) to the lcs ranks , @xmath439 is _ always _ connected to the chen ranks introduced by chen in @xcite . [ cranks ] the chen ranks of @xmath115 are the lcs ranks of the maximal metabelian quotient of @xmath115 : @xmath440 where @xmath441 $ ] .
[ conj : chen ] let @xmath442 be an arrangement group , and let @xmath443 be the number of components of @xmath411 of dimension @xmath444 .
then , for @xmath445 : @xmath446 for the previous example , @xmath447 , so @xmath448 to discuss the chen ranks , we need some background .
the alexander invariant @xmath449 is a module over @xmath450 $ ] . for arrangements , @xmath450=$ ] laurent polynomials in @xmath28-variables . in @xcite
, massey showed that @xmath451 it turns out to be easier to work with the linearized alexander invariant @xmath260 introduced by cohen - suciu in @xcite @xmath452 where @xmath453 is built from the koszul differential and @xmath454 .
@xmath455 for @xmath456 , @xmath457 this shows that the chen ranks are combinatorially determined , and depend only on @xmath140 in rank @xmath438 . for the @xmath60 arrangement depicted in example [ showmethm ] ,
write @xmath458 . with this labelling @xmath459
from which a presentation for @xmath260 can be computed : @xmath460 a computation shows that @xmath461 is @xmath462 and that the hilbert series of @xmath260 is : @xmath463 on the other hand , the graded betti numbers @xmath464 are @xmath465 so the hilbert series for @xmath260 encodes the ranks of @xmath466 .
this suggests a connection between @xmath439 and @xmath466 , which we tackle in the next section . besides the connection to resonance varieties
, there is a second reason to study @xmath467 : the numbers @xmath468 studied in 7 grow very fast , while the numbers @xmath469 grow at a much slower rate . for the non - fano arrangement of example [ exm : nonfano]@xmath470 @xmath470 -.2 in @xmath471 @xmath472 .
the spaces @xmath467 and @xmath473 are related via the change of rings spectral sequence : @xmath474 for arrangements , details of this relationship are investigated in @xcite .
find a combinatorial description of @xmath475 .
if @xmath258 is koszul , does this provide data on @xmath475 ?
it is fairly easy to see that there is a connection between @xmath439 and linear syzygies , that is , to the module @xmath476 . since @xmath477 the relations @xmath478 yield linear syzygies on @xmath437 : @xmath479 * example [ exm : toyres ] * , continued . since @xmath480
, both @xmath481 and @xmath482 are in @xmath439 , as is the line connecting them : @xmath483 parametrically , this may be written @xmath484 so @xmath485 gives a family of linear syzygies on @xmath437 , parameterized by @xmath152 .
@xmath486 to make the connection between linear syzygies and the module @xmath260 precise , we need the following result : for an arrangement @xmath48 , the aomoto complex is exact , as a sequence of @xmath1-modules : @xmath487 & a_0 \otimes s \ar[r]^{\cdot a } & a_1\otimes s \ar[r]^(.62){\cdot a } & \cdots \ar[r]^(.37){\cdot a } & a_{\ell } \otimes s \ar[r ] & f(a ) \ar[r ] & 0 } .\ ] ] the linearized alexander invariant @xmath260 is determined by @xmath488 : @xmath489 furthermore , for @xmath456 , @xmath490 using this , it is possible to prove one direction of conjecture [ conj : chen ] for @xmath445 , @xmath491 prove the remaining direction of conjecture [ conj : chen ] .
what makes all this work is the bernstein - gelfand - gelfand correspondence , which is our final topic .
let @xmath492 and @xmath493 .
the bernstein - gelfand - gelfand correspondence is an isomorphism between derived categories of bounded complexes of coherent sheaves on @xmath494 and bounded complexes of finitely generated , graded @xmath384modules .
although this sounds exotic , from this it is possible to extract functors let @xmath498 be a finitely generated , graded @xmath384-module . then @xmath499 is the complex @xmath500 & s \otimes p_{i+1 } \ar[r]^(.56){\cdot a } & s \otimes p_{i } \ar[r]^(.44){\cdot a } & s \otimes p_{i-1 } \ar[r]^(.66){\cdot a } & \cdots } , \ ] ] where @xmath501 , so that @xmath502 @xmath512 } s^3 \xrightarrow{\left [ \!\begin{array}{ccc } -x_2 & x_1 & 0 \\ -x_3 & 0 & x_1 \\ 0 & -x_3 & x_2 \end{array}\
! \right ] } s^3 \xrightarrow{\left [ \!\begin{array}{ccc } x_3 & -x_2 & x_1 \end{array}\ ! \right ] } s^1\ ] ] if @xmath8 is a finitely generated , graded @xmath1-module , then @xmath513 is the complex @xmath500 & \hat{e } \otimes m_{i-1 } \ar[r]^(.56){\cdot a } & \hat{e } \otimes m_{i } \ar[r]^(.44){\cdot a } & \hat{e } \otimes m_{i+1 } \ar[r]^(.66){\cdot a } & \cdots } , \ ] ] where @xmath514 , so @xmath515 , and @xmath516 is @xmath10-dual to @xmath384 : @xmath517 just as @xmath518 , @xmath519 .
.05 in * example [ exm : first ] * , continued .
if @xmath520/\langle x_0x_1 , x_0 ^ 2 \rangle$ ] , then @xmath522 } e(3)^2 \xrightarrow{\left [ \!\begin{array}{cc } 0 & e_1 \end{array}\ ! \right ] } e(4)^1 \xrightarrow{\left [ \!\begin{array}{c } e_1 \end{array}\ ! \right ] } e(5)^1 \xrightarrow{\left [ \!\begin{array}{c } e_1 \end{array}\ ! \right ] } \cdots\ ] ] this complex is exact , except at the second step .
the kernel of @xmath523\ ] ] is generated by @xmath524 $ ] and @xmath525 $ ] , with relations @xmath526 , so that @xmath527 the betti table for @xmath8 is : @xmath528 -.2 in note that in this example , @xmath8 is @xmath20-regular.@xmath486 higher resonance may be interpreted via @xmath197 : @xmath537 furthermore , the differentials in free resolution of @xmath258 over @xmath384 can be analyzed using bgg and the grothendieck spectral sequence .
for any coherent sheaf @xmath538 on @xmath539 , there is a finitely generated , graded _
@xmath1-module @xmath8 whose sheafification is @xmath538 .
if @xmath538 has castelnuovo - mumford regularity @xmath444 , then the tate resolution of @xmath538 is obtained by splicing the complex @xmath540 : @xmath487 & \hat{e } \otimes m_{r } \ar[r]^{d^r } & \hat{e } \otimes m_{r+1 } \ar[r ] & \hat{e } \otimes m_{r+2 } \ar[r ] & \cdots } , \ ] ] with a free resolution @xmath541 for the kernel of @xmath542 : @xmath543 & & p_1 \ar[rr ] & & p_0 \ar[rr ] \ar[dr ] & & \hat{e } \otimes m_{r } \ar[rr ] & & \hat{e } \otimes m_{r+1 } \ar[rr ] & & \cdots \\ & & & & & \ker(d^r ) \ar[ur ] \ar[dr ] & & & & & \\ & & & & 0 \ar[ur ] & & 0 & & & & } \ ] ] * example [ exm : twistedcubic ] * , continued .
the betti table for the twisted cubic shows that @xmath236 has regularity one , which provides us the information needed to compute the tate resolution . plugging the resulting numbers into theorem [ thmefs ] shows that @xmath550 & @xmath19 & @xmath20 & @xmath22 + @xmath551 & @xmath552 & @xmath553 & @xmath22 & @xmath19 & @xmath19 & @xmath19 + @xmath554 & @xmath19 & @xmath19 & @xmath19 & @xmath20 & @xmath26 & @xmath65 + * conclusion * in this note we have surveyed a number of open problems in arrangements .
the beauty of the area is that these problems are all interconnected .
perhaps the most central objects are the resonance varieties , which are related to both the lcs ranks studied in 7 and 8 using koszul and lie algebras , and to the chen ranks .
the results of 5 tie resonance to the orlik - terao algebra , and @xcite notes that @xmath558 , so the orlik - terao algebra is also linked to @xmath75 and freeness . but freeness ties in to multiarrangements , and can be generalized to hypersurface arrangements , the topics of 3 and 4 .
to complete the circle , recent work of cohen - denham - falk - varchenko @xcite relates freeness to @xmath250 . in short , everything is connected ! .05 in * acknowledgements *
many thanks are due to the organizers of the mathematical society of japan seasonal institute : takuro abe , hiroaki terao , masahiko yoshinaga and sergey yuzvinsky organized a wonderful meeting .
i also am grateful to my research collaborators : g. denham , m. mustata , a. suciu , h. terao , s. tohaneanu , and m. yoshinaga .
as noted in the introduction , the calculations carried out in this survey may be performed using the hyperplane arrangements package of denham and smith @xcite in macaulay2 @xcite .
code for the individual examples is available at http://www.math.uiuc.edu/ schenck/. m. falk , h. terao eds . ,
_ arrangements
tokyo 1998 .
proceedings of the workshop on mathematics related with arrangements of hyperplanes , celebrating the 60th birthday of peter orlik _ mathematical society of japan , tokyo , 2000 .
e. feichtner , _ de concini - procesi wonderful arrangement models : a discrete geometer s point of view .
_ combinatorial and computational geometry , 333360 , math .
52 , cambridge univ . press , cambridge , 2005 . | this paper provides an overview of selected results and open problems in the theory of hyperplane arrangements , with an emphasis on computations and examples .
we give an introduction to many of the essential tools used in the area , such as koszul and lie algebra methods , homological techniques , and the bernstein - gelfand - gelfand correspondence , all illustrated with concrete calculations .
we also explore connections of arrangements to other areas , such as de concini - procesi wonderful models , the feichtner - yuzvinsky algebra of an atomic lattice , fatpoints and blowups of projective space , and plane curve singularities . |
stars are the result of the gravitational collapse of cold gas , but on large scales the conversion of cold gas into stars takes much longer than a collapse time .
early explanations for this low star formation rate ( sfr ) involved the magnetic support against the collapse of giant molecular clouds ( gmcs ) , while the more recent scenario of turbulent fragmentation relies on the support of gmcs by supersonic turbulence @xcite . however , numerical simulations of star formation in turbulent clouds have generally failed to yield sfrs as low as observed values ( see 4 ) .
this work shows that supersonic mhd turbulence results in a sfr as low as observed in gmcs , for characteristic values of the free - fall time divided by the dynamical time , @xmath0 , the alfvnic mach number , @xmath1 , and the sonic mach number , @xmath2 ( defined in 2 ) . using a very large set of adaptive - mesh - refinement ( amr ) simulations of driven mhd turbulence , including self - gravity and sink particles
, we quantify the dependence of the sfr per free - fall time , @xmath3 , on @xmath0 , confirming the results of our previous uniform grid simulations in the case of a very weak mean magnetic field .
the sfr per free - fall time is the efficiency factor of a theoretical schmidt - kennicut law , @xmath5 often used in models and simulations of star formation in galaxies .
we find that @xmath3 decreases exponentially with increasing @xmath0 , has no dependence on @xmath2 alone , and varies by less than a factor of two with variations in @xmath1 within reasonable values for star - forming regions .
we therefore propose a new empirical law of star formation that depends only on @xmath0 .
llccccccc + run & @xmath2 & @xmath1 & @xmath0 & @xmath6 & @xmath3 + [ 0.8ex ] + m10b0.3g12.5 & 10 & 33 & 3.09 & 20 & 0.00 + m10b0.3g25 & 10 & 33 & 2.18 & 14 & 0.10 + m10b0.3g50 & 10 & 33 & 1.54 & 10 & 0.20 + m10b0.3g100 & 10 & 33 & 1.09 & 7 & 0.41 + m10b0.3g200 & 10 & 33 & 0.77 & 5 & 0.46 + m10b0.3g400 & 10 & 33 & 0.54 & 3.5 & 0.59 + m10b2g12.5 & 10 & 5 & 3.09 & 20 & 0.00 + m10b2b25 & 10 & 5 & 2.18 & 14 & 0.03 + m10b2g50 & 10 & 5 & 1.54 & 10 & 0.09 + m10b2g100 & 10 & 5 & 1.09 & 7 & 0.22 + m10b2g200 & 10 & 5 & 0.77 & 5 & 0.28 + m10b2g400 & 10 & 5 & 0.54 & 3.5 & 0.38 + m10b2g800 & 10 & 5 & 0.39 & 2.5 & 0.47 + m20b1g50 & 20 & 20 & 3.09 & 10 & 0.008 + m20b1g100 & 20 & 20 & 2.18 & 7 & 0.10 + m20b1g200 & 20 & 20 & 1.54 & 5 & 0.16 + m20b1g400 & 20 & 20 & 1.09 & 3.5 & 0.28 + m20b1g800 & 20 & 20 & 0.77 & 2.5 & 0.44 + m20b4g50 & 20 & 5 & 3.09 & 10 & 0.006 + m20b4g100 & 20 & 5 & 2.18 & 7 & 0.04 + m20b4g200 & 20 & 5 & 1.54 & 5 & 0.10 + m20b4g400 & 20 & 5 & 1.09 & 3.5 & 0.20 + m20b4g800 & 20 & 5 & 0.77 & 2.5 & 0.28 + m20b16g50 & 20 & 1.25 & 3.09 & 10 & 0.008 + m20b16g100 & 20 & 1.25 & 2.18 & 7 & 0.06 + m20b16g200 & 20 & 1.25 & 1.54 & 5 & 0.15 + m20b16g400 & 20 & 1.25 & 1.09 & 3.5 & 0.25 + m20b16g800 & 20 & 1.25 & 0.77 & 2.5 & 0.30 + + 32m10b2g50 & 10 & 5 & 1.54 & 5 & 0.07 + 128m10b2g50 & 10 & 5 & 1.54 & 20 & 0.08 + 128m10b2g501e6 & 10 & 5 & 1.54 & 20 & 0.07 + 32m10b2g100 & 10 & 5 & 1.09 & 3.5 & 0.24 + 128m10b2g100 & 10 & 5 & 1.09 & 14 & 0.22 + 128m10b2g1001e6 & 10 & 5 & 1.09 & 14 & 0.22 + 32m10b2g800 & 10 & 5 & 0.39 & 1.25 & 0.46 + 128m10b2g800 & 10 & 5 & 0.39 & 5 & 0.49 + 128m10b2g8001e6 & 10 & 5 & 0.39 & 5 & 0.50 + + 32m10b2g50b & 10 & 5 & 1.54 & 5 & 0.09 + 32m10b2g50c & 10 & 5 & 1.54 & 5 & 0.12 + 32m10b2g50d & 10 & 5 & 1.54 & 5 & 0.11 + 32m10b2g50e & 10 & 5 & 1.54 & 5 & 0.05 + 32m10b2g100b & 10 & 5 & 1.09 & 3.5 & 0.22 + 32m10b2g100c & 10 & 5 & 1.09 & 3.5 & 0.31 + 32m10b2g100d & 10 & 5 & 1.09 & 3.5 & 0.26 + 32m10b2g100e & 10 & 5 & 1.09 & 3.5 & 0.17 + [ 0.4ex ] + [ t1 ] the simulations are carried out with the amr code ramses @xcite , modified to include random turbulence driving and sink particles , and optimized for supercomputers with large multi - core nodes through an openmp / mpi hybrid layout . as in @xcite
, we adopt periodic boundary conditions , isothermal equation of state , and solenoidal random forcing in fourier space at wavenumbers @xmath7 ( @xmath8 corresponds to the computational box size ) .
we choose a solenoidal force to make sure that dense , star - forming structures are naturally generated in the turbulent flow , rather than directly imposed by the external force .
the simulations are first run without self - gravity for approximately five dynamical times , starting with uniform density and magnetic fields , and random velocity field with power only at wavenumbers @xmath9 .
the driving force keeps the rms sonic mach number , @xmath10 , at the approximate values of either 10 or 20 , characteristic of gmcs on scales of order 10 pc ( @xmath11 is the three - dimensional rms velocity , and @xmath12 is the isothermal speed of sound ) . for the simulations with @xmath13 ,
we have chosen two different values of the mean magnetic field , corresponding to rms alfvnic mach numbers of @xmath14 and 5 , where @xmath15 , and @xmath16 is the alfvn velocity corresponding to the mean magnetic field and the mean density .
for the larger mach number , @xmath17 , we have adopted three different values of the mean magnetic field , corresponding to @xmath18 , 5 , and 1.25 . in the supersonic regime , the turbulent dynamo
is suppressed @xcite , and the magnetic energy growth saturates at approximately 2% of the kinetic energy of the turbulence , when the mean magnetic field is negligible and does not characterize the system @xcite . on the other hand ,
if the mean magnetic field is not negligible , as in our simulations and in real gmcs , it is readily amplified , well beyond the dynamo saturation level , by stretching and compression in the turbulent flow , and thus its saturated rms value depends directly on its mean value , as found in simulations of @xcite , and in equation ( 20 ) of @xcite .
therefore , the value of @xmath1 , defined with the mean magnetic field , characterizes the saturated regime and can be taken as a fundamental parameter of the system .
the final snapshots of these five turbulence simulations are used as initial conditions for the star - formation simulations with self - gravity and sink particles .
each initial condition is used for 5 to 7 simulations with different numerical values of the gravitational constant , corresponding to different values of @xmath0 , resulting in 28 star - formation simulations ( plus 6 larger runs used to test numerical convergence , and 11 smaller runs to test both convergence and variations of @xmath3 with initial conditions a total of 45 star - formation simulations ) . the dynamical time is defined as @xmath19 , where @xmath20 is the size of the computational domain , and @xmath21 , where @xmath22 is the mean density in the computational volume .
the simulations were run until the star formation efficiency value of sfe@xmath23 0.2 - 0.3 , where @xmath24 is the mass in the sink particles , and @xmath25 the total mass ( gas plus sinks ) , except the simulations with very low sfr , that were run until sfe@xmath26 0.05 - 0.1 .
the five initial turbulence simulations have a root grid of 128@xmath27 computational cells and 5 amr levels , thus a maximum resolution equivalent to 4,096@xmath27 cells . in all runs ,
the first amr level is created if the density is 2.5 times larger than the mean .
subsequent amr levels are created when the gas density increases by a factor of 4 relative to the previous level .
in contrast to @xcite and @xcite , we do not refine based on velocity or pressure gradients .
however , we verified that the tails of the density pdfs are sufficiently converged in the turbulence simulations used as initial conditions ( details will be given elsewhere ) .
when self - gravity is included , the root grid is reduced to 64@xmath27 cells , while the number of amr levels is increased from 5 to 8 , to achieve a maximum resolution equivalent to 16,384@xmath27 cells .
sink particles are created in cells where the gas density is larger than @xmath28 times the mean density ( much larger than the maximum density reached in the initialization runs without self - gravity ) , if the following additional conditions are met at the cell location : i ) the gravitational potential has a local minimum value , ii ) the three - dimensional velocity divergence is negative , and iii ) no other previously created sink particle is present within an exclusion radius , @xmath29 ( @xmath30 in these simulations , where @xmath31 is the highest - resolution cell size ) .
these conditions are similar to those in @xcite .
we have verified that they avoid the creation of spurious sink particles in regions where the gas is not collapsing . , and sink particles from the run 128m10b2g50 , at a time when sfe@xmath32 .
the color scale is linear in @xmath33 , and the tick values in the color bar give @xmath34 in numerical units , with mean gas density and computational box size equal to unity . ]
the numerical parameters characterizing the simulations are given in table 1 .
the simulation names contain the rms sonic mach number , and the numerical values of the mean magnetic field and of the gravitational constant . in table 1
, we give three non - dimensional parameters , @xmath2 , @xmath1 , and @xmath0 , that fully characterize the simulations .
the ratio @xmath0 expresses the relative strength of gravity and turbulence , like the virial parameter , @xmath35 , used in previous works .
although formally equivalent to @xmath35 , for example through the usual expressions valid for a uniform sphere , we prefer to use @xmath0 as the third parameter , in line with the choice of the mach numbers ( ratios of crossing times as well ) as the first two parameters . in table 1 , we also give the ratio between the minimum jeans length at each level , @xmath36 ( corresponding to the highest density before further refinement ) , and the computational cell size , @xmath37 , at that level . due to the refinement by a factor of two in size for every increase by a factor of four in density , the minimum number of cells per jeans length , @xmath6 , is the same for every level @xmath38 in a given simulation .
the last column of table 1 shows the values of @xmath3 obtained from the simulations . at the bottom of table 1
, we also list the simulations used to test numerical convergence and to study the variations of @xmath3 with initial conditions .
the names of these extra runs begin with their root grid size , either @xmath39 or @xmath40 computational cells . projected density and sink particles from one of these runs , 128m10b2g50 , are shown in figure 1 , at a time when sfe@xmath32 . and @xmath41 and 1.25 .
the values of @xmath0 are 0.77 , 1.09 , 1.54 , 2.18 , and 3.09 , from top ( high sfe ) to bottom . ]
figure 2 shows the time dependence of the sfe for the two sets of simulations with @xmath17 and @xmath41 and 1.25 . in some runs with high @xmath3
, the sfr tends to grow with time . to extract a representative value , we measure @xmath3 as the slope of the least - squares fit of the sfe versus @xmath42 , in the interval @xmath43 , as in @xcite .
the simulations with the lowest values of @xmath3 were not integrated up to sfe=0.2 .
however , they do not exhibit a systematic growth of the sfr with time , and they are run for many free - fall times ( up to 9 in some cases ) , so their average @xmath3 is nonetheless well defined .
the values of @xmath3 from all simulations are plotted in figure 3 , as a function of @xmath0 .
a striking result , stressed by the linear - log scale of figure 3 , is the approximately exponential dependence of @xmath3 on @xmath0 ( for fixed values of @xmath2 and @xmath1 ) , showing that supersonic mhd turbulence is very effective at slowing down the star formation process , which seems to be controlled primarily by the ratio of the characteristic times of gravity and turbulence .
another important result is the lack of a dependence of @xmath3 on @xmath2 , for fixed values of @xmath1 and @xmath0 .
furthermore , the dependence on @xmath1 is not very strong , considering the very large range of @xmath1 values of the simulations .
this is in part due to the complex dependence of @xmath3 on @xmath1 . down to @xmath41
, @xmath3 decreases with decreasing @xmath1 ( increasing magnetic field strength ) , while for further decreasing @xmath1 , @xmath3 starts to increase .
figure 3 shows that a simple exponential law , where @xmath3 is only controlled by the ratio @xmath0 , could describe all simulation results within a factor of two . for even weaker magnetic fields than considered here , @xmath3 would certainly be larger , as it should approach the values found in simulations without magnetic fields @xcite .
however , observations in star forming regions are consistent with @xmath44 @xcite , so it is likely that @xmath3 approaches its minimum value , which is approximately fit by a simple exponential law , @xmath45 , shown as a dashed line in figure 3 .
we therefore propose a new empirical law of star formation , based on this fit to the minimum @xmath3 : @xmath46 where @xmath47 accounts for mass loss through jets and winds , during the formation of a star .
this law only depends on the mean gas density and the rms velocity of a star - forming region , so it is easily implemented in analytical models and simulations of galaxy formation and evolution . , versus @xmath0 .
the symbols for each series of runs , where only the strength of gravity is changed and @xmath2 and @xmath1 are kept constant , are connected by a line , to better distinguish each series .
the two error bars give the mean of @xmath3 , plus and minus the rms values , for each group of five @xmath40-root - grid runs with identical parameters ( @xmath13 and @xmath41 ) , but different initial conditions .
the dashed line is an approximate exponential fit to the minimum value of @xmath3 versus @xmath0 . ]
equation ( [ sfrff ] ) is valid as long as the fraction of the system mass , @xmath48 , available to form stars is larger than the characteristic stellar mass , @xmath49 , which we can express as @xmath50 .
if @xmath3 predicted by equation ( [ sfrff ] ) is too small to satisfy this condition , then @xmath51 , simply because the total system mass is too small . adopting the characteristic stellar mass predicted by our turbulent fragmentation model @xcite , @xmath52 , which corresponds to approximately three times the mean bonnor - ebert mass , @xmath53 , evaluated for an external pressure equal to the ram pressure of the turbulence , and using the minimum @xmath3 derived here
, we find the condition can be expressed as @xmath54 .
as shown by table 1 , the weakest gravity we have simulated corresponds to @xmath55 , giving @xmath56 and 38.6 for the runs with @xmath13 and 20 respectively . in good agreement with these values , in each of the @xmath13 runs with @xmath55 , we get only a single star , so the sfr is practically zero , while in the three @xmath17 runs with @xmath55 we get approximately 10 - 20 stars per free - fall time . to test the numerical convergence of @xmath3
, we have rerun three of the simulations with @xmath13 and @xmath41 , using both twice smaller and twice larger root grid size ( @xmath40 and @xmath39 computational cells ) , and the same number of amr levels , corresponding to maximum resolutions equivalent to 8,192@xmath27 cells and 32,768@xmath27 cells .
these three larger runs were then rerun again using a 10 times larger value of the threshold density above which sink particles are created ( 10@xmath57 times the mean density , instead of @xmath28 ) .
as shown in table 1 , the values of @xmath3 in the runs with @xmath58 and 1.09 are practically independent of numerical resolution , showing that the correct sfr can be derived even when truelove s condition @xcite is violated , and that the value of @xmath28 times the mean density ( adopted in the main parameter study ) , above which sink particles are allowed to form , is sufficiently high to avoid dense regions that are not collapsing . in the case of @xmath59
, @xmath3 is approximately 20% lower in the @xmath39-root - grid simulations than in the corresponding @xmath60-root - grid run .
however , this low sfr is obtained from an average over many dynamical times . during the first dynamical time
, the sfr is virtually independent of resolution .
the corresponding simulation with @xmath40 root - grid size has the same value of @xmath3 as the largest simulations , showing that the difference from the @xmath60-root - grid run is not due to lack of convergence , but more likely to random differences between runs after a long time , as expected in chaotic systems .
the values of @xmath3 from the @xmath39-root - grid simulations can be seen in figure 3 . to investigate the effect of the initial conditions , we have carried out 10 simulations with twice smaller root grid size ( @xmath40 computational cells ) , and the same number of amr levels , five of them with @xmath61 , and the other five with @xmath59 , starting from different initial conditions of fully - developed turbulence , with @xmath13 and @xmath41 . as shown in table 1 , the maximum difference in @xmath3 from such small samples is approximately a factor of two .
the mean values of @xmath3 , plus and minus the rms values , are plotted as error bars in figure 3 .
the values of @xmath3 predicted by equation ( [ sfrff ] ) are consistent with those found in gmcs . as discussed in @xcite , accounting for the full uncertainty in the lifetime of the class ii phase ( @xmath62 myr ) , the results of @xcite implies @xmath63 - 0.12 . using the gmc sample by @xcite
, we derive @xmath64 ( 1-@xmath65 ) , resulting in @xmath63 - 0.12 based on our equation ( [ sfrff ] ) , with @xmath47 , in nice agreement with the values derived by @xcite ( or @xmath66 and @xmath67 - 0.19 if we double the cloud masses ) .
similar values are found also for the orion nebula cluster , @xmath68 - 0.09 , the object with the smallest error bars in @xcite .
based on the values of @xmath0 in the sample of @xcite we would predict an average value of @xmath69 , while , based on the free - free emission , murray finds an average value of @xmath70 .
the amr simulations with @xmath13 and @xmath14 presented here are consistent with our previous uniform grid simulations @xcite with @xmath71 and @xmath72 ( the values of @xmath35 given in the plot of figure 2 are derived from @xmath0 based on the case of a uniform sphere , while those reported in figure 7 of @xcite were based on the total mass in the computational box , and should be multiplied by a factor of 2 for the comparison ) .
the sfr model of @xcite adopted the simplifying assumption that , in the mhd case , the effective postshock ratio of gas to magnetic pressure , @xmath73 , and the density pdf do not depend on the mean magnetic field strength , for reasonable values of @xmath1 .
the complex , but weak dependence of @xmath3 on @xmath1 found in the simulations justifies that approximation .
however , by neglecting the dependence of @xmath73 on both @xmath1 and @xmath2 , the model yields a direct dependence on @xmath2 , not found in our amr simulations . equation ( 20 ) of @xcite , relating @xmath73 to @xmath2 and @xmath74 , the ratio of gas to magnetic pressure computed with the average magnetic field strength , @xmath75 , was derived under the assumption of flux freezing , neglecting the postshock thermal pressure and dynamical alignment of magnetic and velocity fields .
@xcite noticed that when @xmath73 is averaged at high densities , it becomes independent of @xmath74 , so they took @xmath76 , which also provided a good fit to the density pdfs .
however , they did not test the dependence of @xmath73 on @xmath2
. it may be more physical to retain the dependence of @xmath73 on @xmath2 even at high densities , because the postshock magnetic pressure tends to balance the ram pressure of the turbulence causing the shocks .
assuming @xmath77 instead of @xmath78 , we obtain @xmath3 practically independent of @xmath2 , as found in our amr simulations .
this assumption yields a density variance in mhd given by @xmath79 , for @xmath80 , while the approximation @xmath78 in @xcite gave @xmath81 ( this issue will be addressed in a future study ) .
although this correction eliminates the dependence on @xmath2 , the model does not predict the subtle dependence of @xmath3 on @xmath1 , as it does not account for the increasing anisotropy and dynamical alignment of magnetic and velocity fields , with decreasing @xmath1 .
therefore , the simple empirical law derived above is a better choice for practical applications .
the models by @xcite and @xcite do not include the effect of magnetic fields . as shown by our previous uniform - grid simulations
@xcite , even with a very weak mean magnetic field , corresponding to @xmath72 , @xmath3 is approximately three times smaller than in the non - magnetized case .
the amr simulations presented here show that , with more realistic , larger magnetic field values , @xmath3 decreases even further .
analytical models of star formation neglecting magnetic fields should not yield low values of @xmath3 consistent with observations of star - forming regions
. if they do , they are inconsistent with the numerical experiments .
the new amr simulations of star formation presented in this work show that i ) @xmath3 decreases exponentially with increasing @xmath0 , but is insensitive to changes in @xmath2 ( in the range @xmath82 ) , for constant values of @xmath0 and @xmath1 .
ii ) decreasing values of @xmath1 ( increasing magnetic field strength ) reduce @xmath3 , but only to a point , beyond which @xmath3 increases with a further decrease of @xmath1 .
iii ) for values of @xmath1 characteristic of star - forming regions , @xmath3 varies with @xmath1 by less than a factor of two .
we propose a simple law of star formation depending only on @xmath0 , based on the empirical fit to the minimum @xmath3 : @xmath4 .
this law shows that mhd turbulence is very effective at slowing down the star - formation process and can explain the low average sfr in molecular clouds . because it only depends on the mean gas density and rms velocity of a star - forming region ,
the star - formation law we propose is straightforward to implement in simulations and analytical models of galaxy formation and evolution .
future work should test its effect in simulations where the feedback of sns is accounted for , but the process of star formation is not spatially resolved and needs to be modeled .
we thank christoph federrath and the referee for useful comments on the manuscript .
pp is supported by the spanish micinn grant aya2010 - 16833 and by the fp7-people-2010-rg grant pirg07-ga-2010 - 261359 . th and n
are supported by the danish national research foundation , through its establishment of the centre for star and planet formation .
the simulations were carried out on the nasa / ames pleiades supercomputer . , ii , n. j. , dunham , m. m. , jrgensen , j. k. , enoch , m. l. , mern , b. , van dishoeck , e. f. , alcal , j. m. , myers , p. c. , stapelfeldt , k. r. , huard , t. l. , allen , l. e. , harvey , p. m. , van kempen , t. , blake , g. a. , koerner , d. w. , mundy , l. g. , padgett , d. l. , & sargent , a. i. 2009 , , 181 , 321 , a. g. , ustyugov , s. d. , norman , m. l. , & padoan , p. 2009
, in astronomical society of the pacific conference series , vol .
406 , numerical modeling of space plasma flows : astronum-2008 , ed . n. v. pogorelov , e. audit , p. colella , & g. p. zank , 15 | we show that supersonic mhd turbulence yields a star formation rate ( sfr ) as low as observed in molecular clouds ( mcs ) , for characteristic values of the free - fall time divided by the dynamical time , @xmath0 , the alfvnic mach number , @xmath1 , and the sonic mach number , @xmath2 . using a very large set of deep adaptive - mesh - refinement simulations ,
we quantify the dependence of the sfr per free - fall time , @xmath3 , on the above parameters .
our main results are : i ) @xmath3 decreases exponentially with increasing @xmath0 , but is insensitive to changes in @xmath2 , for constant values of @xmath0 and @xmath1 .
ii ) decreasing values of @xmath1 ( stronger magnetic fields ) reduce @xmath3 , but only to a point , beyond which @xmath3 increases with a further decrease of @xmath1 .
iii ) for values of @xmath1 characteristic of star - forming regions , @xmath3 varies with @xmath1 by less than a factor of two . we propose a simple star - formation law , based on the empirical fit to the minimum @xmath3 , and depending only on @xmath0 : @xmath4 . because it only depends on the mean gas density and rms velocity , this law is straightforward to implement in simulations and analytical models of galaxy formation and evolution . |
tlm98 showed that the outer ( adjustment ) radius of the transition layer ( cc ) @xmath21 measured in the dimensionless units with respect to schwarzschild radius @xmath22 , @xmath23 , anticorrelates with @xmath19number or photon index @xmath7 ( spectral state ) only .
thus @xmath15 ( or @xmath9 ) as a ratio @xmath24 should correlate with @xmath7 ( or @xmath19number ) where values @xmath15 related to the same @xmath7 for different sources should be inversely proportional to a mass of the central object ( black hole or neutron star ) .
note that a plasma velocity @xmath25 is also a function of @xmath7 only .
the comparison of the observed index
qpo frequency correlations for two different sources ( with two different masses ) should lead to determination of their relative masses with respect each other .
this is the main idea behind the method of weighing black holes ( tlm98 ) recently applied for bh mass determination in a number of galactic and extragalactic sources [ see tf04 ; @xcite ; @xcite ; @xcite and @xcite ] .
@xcite proposed a model of the truncated keplerian disk and hot corona located in the innermost part of the source using arguments in terms of radiation hydrodynamics .
@xcite , hereafter cgr99 , were the first who supported this model using the specific form of the power spectra of x - ray radiation .
particularly they noted that two - component x - ray spectra ( soft multicolour black - body and harder power law ) are frequently observed from accreting black holes .
these components are presumably associated with the different parts of the accretion flow ( optically thick and optically thin respectively ) in the vicinity of the compact source .
they also emphasized that for cygnus x-1 , the overall shape of the power density spectra in the soft and hard spectral states can be _ qualitatively _ explained if the geometrically thin disc is sandwiched by the geometrically thick corona extending in a radial direction up to a large distance from the compact object . in the hard state
the thin disc is truncated at some distance from the black hole followed by the geometrically thick flow .
the break in the pds is then associated with the characteristic frequencies in the accretion flow at the thin disc truncation radius .
however there is a difference between cgr99 s suggestions and the results of our paper .
we come to the model of the truncated disk plus corona using the fit of the theoretical model of the power spectrum ( see tsa07 ) to the data . whereas cgr99 proposed a _ qualitative _ explanation of the data using lyubarskii s idea on the formation of the power spectrum in x - ray binaries
[ see l97 ] .
our results confirm the cgr99 conclusion that a stable optically thick ( but geometrically thin ) disk extends down to small radii ( as indicated by the strong and stable soft component ) and whereas prominent variations of the harder component ( presumably related to compton cloud ) are present in the broad range of time - scales ( up to at least @xmath26 s ; see figs . 2 - 3 in tsa07 ) .
we also confirm the observed emergent spectrum consisting of two components ( soft stable component due to the disc emission and harder highly variable component due to comptonization in the corona ) .
the relative contribution of these two components to the luminosity of an averaged spectrum would then reflect the ratio of the energy releases in the disc and corona ( or mass accretion rates ) .
it is also worth noting that our presented analysis indicates that the high - frequency part of the pds ( qpo and turnover ) at about 10 hz and higher is possibly related to the local instabilities operating in the comptonization region of the energy release ( see cgr99 ) .
note that @xcite also pointed out the evolution of pdss in xte j15550 - 564 and xte j1650 - 500 .
in particular they showed a variable power in the comptonized component ( the pds high - frequency component ) when a given source evolves from hard to soft states .
however there is a difference in details between the cgr99 and results of our papers ( see also tsa07 ) .
for example the break in the pds frequency range between 0.01 and 1 hz is associated with the characteristic diffusion frequencies of the compton cloud but not as that in cgr99 `` with the characteristic frequencies in the accretion flow at the thin disc truncation radius . ''
@xcite , hereafter abl05 , suggested a model of the pds of cyg x-1 which successfully presented a natural transition from hard state through intermediate state to soft state and back , and also allowed them to study the behavior of the lorentzian components in detail . using this model consisting of a power - law and two lorentzian profiles ,
abl05 were able to fit the cyg x-1 pdss , and to follow the components from hard state through the transitions and back . by choosing this simple empirical approach for the pds data analysis abs05 could model the major features in all states with only three components .
the parameters used in abl05 were the width and peak frequency of the lorentzians , along with the power at the peak frequency , the parameter evolution and its correlation between each other can all be described by continuous functions .
abl05 also showed that the pds of cyg x-1 is dominated by the same two lorentzian components at all times .
they came to this conclusion based on the fact that the relation of the peak frequencies of these lorentzian components follows the pattern seen in both other bh and ns systems , and therefore they proposed that the physical processes responsible for them can not be explained by invoking magnetic fields , a solid surface , or an event horizon .
it is worth noting that we also find that the diffusion ( break ) and qpo frequency follows the bh - ns pattern ( see details in st06 and tsa07 ) .
in addition to this we demonstrate that there is a certain correlation between qpo low frequency and driving ( high ) frequency and photon index ( see fig .
2 ) and show a correlation between the diffusion time scale ( an inverse of the diffusion frequency ) and the driving frequency ( see fig .
furthermore we explain the nature of the observational appearances of the power and photon spectra evolution by the evolution of the truncated disk and corona configurations when a x - ray source evolves from hard state to soft states .
we come to the conclusion that the truncated disk and corona scenario should be a common model in bh and ns systems .
in fact , tlm98 showed that the bounded configuration ( compton cloud ) surrounding compact objects is the transition layer ( tl ) that is formed as a result of dynamical adjustments of a keplerian disk to the innermost sub- keplerian boundary conditions .
they argued that this type of adjustment is a generic feature of the keplerian flow in the presence of the sub - keplerian boundary conditions near the central object and that it does not necessarily require the presence or absence of a hard surface .
tlm98 concluded that an isothermal sub - keplerian tl between the ns surface and its last keplerian orbit forms as a result of this adjustment .
the tl model is general and is applicable to both ns and bh systems . \1 . that we explain the decay of the emergent time variability of x - ray emission in compact sources when these sources evolve from low / hard to high / soft states .
we find that the resulting power @xmath1 from cyg x-1 decays with the driving oscillation frequency @xmath2 as @xmath27 .
we show that the reciprocal of the diffusion time scale of the perturbation @xmath28 , the low frequency qpo @xmath15 , the driving oscillation frequency @xmath2 , inferred by fitting the cyg x-1 pdss with our diffusion model , increase when the source evolves from low / hard state to high / soft state .
this behavior of the pds characteristics implies that the compton cloud contracts towards softer spectral states .
the driving oscillations are probably caused by the local rayleigh - taylor instability which cumulative @xmath3 decreases when the effective area of the configuration producing the rt oscillations contracts .
the decay in driving power leads to the decay in the total observed variability power from the source . using the fact that @xmath28 , @xmath15 , @xmath2 increase with @xmath7 and @xmath3 , @xmath1 decrease with @xmath7 we conclude , as a result of our analysis , that the compton corona shrinks when cyg x-1 goes from low / hard state to high / soft state .
our extensive data analysis of the power spectra from cyg x-1 with an application of the method of @xmath19number determination developed in tsa07 indicates that @xmath8 related to compton cloud configuration increases from values about 10 in low / hard state to that about 70 in high / soft state .
we confirm the predictions by tlm98 that @xmath19number should increase with index and qpo frequencies .
thus one can conclude that the observable index - qpo correlation is probably driven by the increase of @xmath19number when the source evolves from low / hard state to high / soft state .
it is worth noting that inverse proportionality of the low - frequency qpo with respect bh mass in the index - qpo correlation leads to the method weighing bhs employing this index - qpo correlation . | a characteristic feature of the fourier power density spectrum ( pds ) observed from black hole x - ray binaries in low / hard and intermediate spectral states is a broad band - limited noise , characterized by a constant below some frequency ( a `` break '' frequency ) and a power law above this frequency .
it has been shown that the variability of this type can be produced by the inward diffusion of the local driving perturbations in a bounded configuration ( accretion disk or corona ) . in the framework of this model , the perturbation diffusion time @xmath0
is related to the phenomenological break frequency , while the pds power - law slope above the `` break '' is determined by the viscosity distribution over the configuration .
the perturbation diffusion scenario explains the decay of the power of x - ray variability observed in a number of compact sources ( containing black hole and neutron star ) during an evolution of theses sources from low / hard to high / soft states .
we compare the model predictions with the subset of data from cyg x-1 collected by the rossi x - ray time explorer ( rxte ) .
our extensive analysis of the cyg x-1 pdss demonstrates that the observed integrated power @xmath1 decreases approximately as a square root of the characteristic frequency of the driving oscillations @xmath2 .
the rxte observations of cyg x-1 allow us to infer @xmath3 and @xmath0 as a function of @xmath2 . using the inferred dependences of the integrated power of the driving oscillations @xmath3 and @xmath0 on @xmath2
we demonstrate that the power predicted by the model also decays as @xmath4 that is similar to the observed @xmath5 behavior .
we also apply the basic parameters of observed pdss , power - law index and low frequency quasiperiodic oscillations , to infer the reynolds ( re ) number from the observations using the method developed in our previous paper .
our analysis shows that re@xmath6number increases from values about 10 in low / hard state to that about 70 during the high / soft state . using the best - fit parameters of the pdss we can even infer the evolution of the reynolds number of the accretion flow in the compton cloud ( cc ) , with the change of photon index @xmath7 . in fact ,
tsa07 relate diffusion time @xmath0 with @xmath8 and a magneto - acoustic qpo frequency @xmath9 as @xmath10\left(\frac{r_0}{v_{r}}\right)= { 4\over{3}}{4\over{(4-\psi)^2}}\frac{\rm re}{a_{ma}\nu_{ma } } , \label{t0_mod}\ ] ] where @xmath11 and @xmath12 is a numerical coefficient .
formula ( [ t0_mod ] ) leads to the equation @xmath13 that allows us to infer a value of @xmath14 using the best - fit model parameters @xmath0 and the qpo low frequency @xmath15 which is presumably close to @xmath9 .
ultimately we can find the evolution of @xmath8 with the photon index @xmath7 given that @xmath15 , @xmath0 and the viscosity index @xmath16 evolve with @xmath7 ( see figs .
[ nu_dr_vs_gamma]-[t0_vs_vdr ] and [ psi_wrn ] ) . in figure [ re_vs_alpha ] we present the inferred reynolds number as a function of the photon index @xmath7 .
we use eq .
( [ re ] ) where we set @xmath17 ( see details of this determination of @xmath18 in tbw01 and tsa07 ) and the observable correlations of @xmath15 and @xmath0 with @xmath7 ( see figs .
[ nu_dr_vs_gamma ] and [ t0_vs_vdr ] ) .
one can see that @xmath19number steadily increases from 10 to 70 when the source evolves from low / hard state to high / soft state .
in contrast , tsa07 found that @xmath20 .
we note , however , that in tsa07 only wrn model was used without accounting for driving oscillation distribution , which significantly affects the resulting value for @xmath16 .
they also used a limited set of data . note the observed behavior of the re - number vs @xmath7 and mass accretion rate was predicted by tlm98 , where they formulated a transition layer model ( tlm ) and studied its consequences for observations .
it is important to emphasize that the @xmath19number along with the photon index @xmath7 , the low frequency qpo @xmath15 and the driving frequency @xmath2 can be considered as characteristics of the spectral state .
all of them correlate with each other . |
the spin of an electron moving in an electric field experiences a torque , which can be understood as arising from the magnetic field in its rest frame . in a two - dimensional electron gas
this velocity - dependent magnetic field produces the quantum spin hall effect,@xcite reminiscent of the quantum hall effect but without time - reversal symmetry breaking.@xcite the difference manifests itself in the integer quantized values @xmath1 of the dimensionless conductance . while there is no restriction on @xmath2 in the quantum hall effect ( qhe ) , only the two values @xmath3 appear in the quantum spin hall effect ( qshe ) . in both effects the current is carried by edge states , separated by an insulating bulk .
the insulator in the qshe is called a topological insulator@xcite if the topological quantum number @xmath4 and a trivial insulator if @xmath5 .
bulk states delocalize when the conductance switches between quantized values .
a distinguishing feature of the qshe is that the delocalized states can support metallic conduction ( conductance @xmath6 ) , while in the qhe the conductance remains @xmath7 .
the metallic conduction appears in extended regions of phase space , separated by a quantum phase transition ( anderson transition ) from the regions with a quantized conductance and an insulating bulk . for a trivial insulator
this is the familiar metal - insulator transition in a two - dimensional ( 2d ) electron gas with spin - orbit scattering.@xcite a fundamental question raised by the discovery of the qshe was whether the phase transition would depend on the topological quantum number . specifically , is the critical exponent @xmath8 of the diverging localization length different if the phase boundary separates a metal from a topological insulator , rather than a trivial insulator ?
a numerical simulation@xcite of the kane - mele model@xcite of the qshe gave an affirmative answer , finding a value @xmath9 substantially below the established result@xcite for the metal trivial - insulator transition .
in contrast , studies of the network model in the qshe universality class gave @xmath10 within numerical accuracy,@xcite consistent with analytical considerations from the nonlinear sigma model@xcite why the critical exponent should be @xmath1-independent .
in this work we study the metal - insulator transition in the qshe by means of an altogether different , _ stroboscopic _ model the periodically driven system known as the quantum kicked rotator.@xcite a key feature of this dynamical system is that spatial dimensions can be exchanged for incommensurate driving frequencies,@xcite allowing for the study of metal - insulator transitions in one spatial dimension.@xcite this proved very effective in the qhe,@xcite and also made it possible to experimentally study the 3d anderson transition in a 1d optical lattice.@xcite in the present paper we apply the same strategy to study the 2d qshe in a 1d system . in the next section
we show how the quantum kicked rotator can be extended to include the topological @xmath4 phase of the qshe .
we first construct this @xmath0 quantum kicked rotator in 2d and then carry out the mapping to 1d .
we calculate the phase diagram in sec .
[ sec : tidisorder ] , using a scattering formula for the topological quantum number valid for disordered systems.@xcite in sec .
[ ssec : scaling ] we determine the scaling law at the metal - insulator phase transition and compare the critical exponents @xmath8 for @xmath5 and @xmath4 .
we conclude in sec .
[ sec : disc ] .
we start from a translationally invariant 2d system and will add disorder later .
the minimal model hamiltonian @xmath12 of the qshe has four bands at each momentum @xmath13 , distinguished by indices @xmath14 ( up and down spins ) and @xmath15 ( _ s _ and _ p _ orbitals ) .
the pauli matrices @xmath16 and @xmath17 ( @xmath18 ) act on the spin and orbital degrees of freedom , respectively .
time - reversal symmetry is essential , @xmath19 inversion symmetry is not essential ( and will be broken anyway once we add disorder ) , but is assumed for convenience , @xmath20 the generic hamiltonian that satisfies the symmetries and has the form@xcite @xmath21 the real functions @xmath22 are even under inversion of @xmath23 , while the functions @xmath24 are odd . because the @xmath25-matrices anticommute , @xmath26 , the band structure is given by @xmath27 each band is twofold degenerate .
the band gap can close upon variation of a single control parameter at high - symmetry points @xmath28 in the first brillouin zone , which satisfy @xmath29 modulo a reciprocal lattice vector . at these time - reversally
invariant momenta the bloch wave function @xmath30 of the lower band has a definite parity @xmath31 under inversion , @xmath32 .
the @xmath0 topological quantum number @xmath1 follows from the fu - kane formula,@xcite @xmath33 a gap closing and reopening can switch the parity of the lower and upper bands , inducing a change in @xmath1 ( a topological phase transition ) .
the specific choice for the functions @xmath34 which we will study in the following is based on experience with the stroboscopic model of the qhe.@xcite we take @xmath35 where the vector @xmath36 ( of length @xmath37 ) has components @xmath38 for @xmath39 this is the bernevig - hughes - zhang model of the qshe,@xcite up to a function @xmath40 which flattens the bands without closing the band gap ( hence without affecting the topological quantum number ) . without the @xmath41-term
the spin degree of freedom @xmath14 is conserved and the qshe hamiltonian is identical to two copies of the qhe hamiltonian ( with opposite magnetic fields , to restore time - reversal symmetry ) .
when @xmath42 spin - orbit coupling mixes the spin - up and spin - down blocks of the hamiltonian .
( color online ) band structure for the clean hamiltonian @xmath43 , as calculated from eqs .
, , and , for parameter values @xmath44 , @xmath45 , @xmath46 , @xmath47 .
the dirac cones emerging at the time - reversally invariant points in the brillouin zone will touch when @xmath48 . ]
the band structure for one set of parameters is shown in fig .
[ fig : band ] .
the time - reversally invariant momenta @xmath49 in the first brillouin zone @xmath50 are at the four points @xmath51 , @xmath52 , @xmath53 , @xmath54 .
( we have set both @xmath55 and the lattice constant @xmath56 equal to unity . )
the topological quantum number depends only on the parameter @xmath57 , @xmath58 for @xmath59 or @xmath60 the band gap closes at @xmath51 or @xmath52 , with a switch in parity and a change of @xmath1 . for @xmath61
the gap closes at both points @xmath53 and @xmath54 at constant @xmath1 since the two parity switches cancel .
the time - dependent model is based on the quantum kicked rotator@xcite , which is a dynamical system designed to study the localization by disorder with great nuerical efficiency.@xcite the time dependent hamiltonian @xmath62 contains a disorder potential @xmath63 and a stroboscopic kinetic energy @xmath64 , @xmath65 ( the stroboscopic period @xmath15 has been set equal to unity . ) we take @xmath64 of the form and will specify @xmath66 later . the disorder strength is set by the relative importance of @xmath66 and @xmath64 .
the floquet operator @xmath67 describes the time evolution of the wavefunction over one period , @xmath68 here @xmath69 is the operator @xmath70 in the momentum representation .
the eigenvalues @xmath71 of the unitary operator @xmath72 define the quasi - energies @xmath73 .
we use the @xmath74-periodicity of @xmath64 to label the eigenstates @xmath75 of @xmath72 by a bloch vector @xmath76 in the brillouin zone @xmath77 . by construction , @xmath78 with @xmath79 a @xmath74-periodic eigenstate of @xmath80 this is the quantum kicked rotator with @xmath0 symmetry . by adding an incommensurate driving frequency to the quantum kicked rotator in two spatial dimensions , it is possible to simulate the system in one single dimension.@xcite for that purpose we take a potential of the form @xmath81 with @xmath82 an irrational number @xmath83 .
we consider states which at @xmath84 are plane waves in the @xmath85-direction , having a well - defined initial momentum @xmath86 . in the floquet operator the term linear in @xmath85
has the effect of shifting @xmath87 to @xmath88 , so that the 2d operator @xmath89 can be replaced at the @xmath90-th time step by the 1d operator @xmath91 there is considerable freedom in the choice of the potential @xmath92 in the remaining dimension .
the simple quadratic form @xmath93 provides sufficient randomness if @xmath94 form an incommensurate triplet.@xcite we take @xmath95 , @xmath96 .
the numerical simulation is performed by subsequent multiplication of the state @xmath97 with the series of floquet operators @xmath98 , @xmath99 , in the plane wave basis @xmath100 ( eigenstates of @xmath101 ) .
the integer @xmath102 is restricted to the values @xmath103 , with @xmath104 an even integer that sets the system size . as initial condition we choose @xmath105 spatially localized at @xmath106 .
the vector @xmath107 is a normalized vector of rank 4 with random components denoting the spin and orbital degrees of freedom .
the multiplication with the floquet operators , represented by @xmath108 unitary matrices , can be done very efficiently by means of the fast fourier transform algorithm .
( see the analogous calculation in the qhe for a more detailed exposition.@xcite ) the calculation is repeated for different values of @xmath109 and @xmath110 , to simulate a disorder average .
the disorder strength can be varied by varying @xmath111 , with _ small _ @xmath111 corresponding to _ strong _ disorder .
before embarking on a calculation of the scaling law and critical exponents , we first locate the metal - insulator transitions and identify the topological phases . to find the metal - insulator transitions , we calculate the time dependent diffusion coefficient @xmath112 from the mean squared displacement @xmath113 the brackets @xmath114 denote the expectation value in the state @xmath115 and the overline indicates the ensemble average over @xmath109 and @xmath110 .
we typically average over @xmath116 samples of size @xmath117 .
( color online ) _ bottom panel : _ time - dependent diffusion coefficient as a function of @xmath57 for @xmath118 , @xmath45 , @xmath46 , shown for times @xmath119 ( red ) , @xmath120 ( green ) and @xmath121 ( blue ) .
the points of intersection of these curves locate the metal - insulator transition . _
top panel : _ topological quantum number for the same parameter values , used to distinguish the topologically trivial ( @xmath5 ) and nontrivial ( @xmath4 ) insulators . in the metallic regions @xmath1 is not quantized . ]
( color online ) time - dependent diffusion coefficient as a function of @xmath122 for several values of @xmath57 near the metal topological - insulator transition .
( same parameters as in fig .
[ fig : dandtscan ] . )
the slope vanishes at the transition point @xmath123 . ] a representative series of scans of @xmath124 versus @xmath57 for different values of @xmath122 is shown in fig .
[ fig : dandtscan ] ( bottom panel ) . in fig .
[ fig : logscaling ] we show @xmath124 versus @xmath122 for different @xmath57 . in the insulating phase @xmath125
decays with increasing time , while in the metallic phase @xmath126 grows with increasing time.@xcite the metal - insulator transition at @xmath127 is signaled by a @xmath122-independent @xmath128 ( corresponding to a scale invariant diffusion coefficient ) . in this way we can locate the phase boundaries , but we can not yet distinguish topologically trivial and nontrivial insulators . for that purpose
we need the topological quantum number .
the formula for the topological quantum number of the translationally invariant system does not apply for nonzero disorder potential .
the scattering formulation@xcite continues to apply and is what we will use .
an alternative hamiltonian formulation for disordered systems has been given by prodan.@xcite the topological invariant is computed using the formalism described in ref . .
from the floquet operator we can construct a reflection matrix @xmath129 for a cylindrical system enclosing a flux @xmath130 .
the topological invariant then follows from the following combination of determinants and pfaffians , evaluated at @xmath131 and @xmath132 , @xmath133}{\text{pf}\,[\sigma_{y } r(0,0 ) ] } \frac{\sqrt{\det r(0,0)}}{\sqrt{\det r(0,\pi)}}. \label{eq : tidisordered}\ ] ] the results in fig .
[ fig : dandtscan ] ( top panel ) show the disorder - averaged @xmath57-dependence of @xmath1 for a system of size @xmath134 .
the value of @xmath1 is only quantized @xmath135 in the insulating regions . in the metallic regions @xmath1 averages to @xmath136 for a sufficiently large system,@xcite which is
not quite observed for our system sizes .
( color online ) phase diagram of the @xmath0 quantum kicked rotator in the @xmath57@xmath137 plane for fixed @xmath45 and @xmath46 . the topological quantum number @xmath138 distinguishes the topologically trivial and nontrivial insulating phases .
the metallic phase between the insulating phases disappears in the clean limit @xmath139 .
the two circles indicate the metal trivial - insulator and metal topological - insulator transitions that were studied to compare the critical exponents . ]
( color online ) phase diagram for fixed disorder strength ( @xmath118 ) , in the @xmath57@xmath41 plane for @xmath45 ( top panel ) and in the @xmath57@xmath140 plane for @xmath46 ( bottom panel ) .
the metallic phase is stabilized by increasing the spin - orbit coupling strength @xmath41 or by decreasing @xmath140 . ]
the phase diagram obtained in this way is shown in figs .
[ fig : pd ] and [ fig : gammapd ] . without disorder ,
the topological invariant gives a topological insulator for @xmath141 and a trivial insulator for @xmath142 . with disorder
a metallic phase appears between the insulating phases , provided that @xmath143 . for small @xmath41
we find an additional trivial insulating phase around @xmath61 , consistent with what was found in the quantum hall system in the presence of disorder @xcite . in some other models of the qshe a reentrant behavior @xmath144 with increasing disorder
is observed in some regions of parameter space.@xcite in our model a system which starts out topologically trivial in the clean limit stays trivial with disorder .
the premise of one - parameter scaling is expressed by the equation@xcite @xmath145 where @xmath72 is a universal scaling function .
the localization length @xmath146 has a power law divergence @xmath147 at the metal - insulator transition , with critical exponent @xmath148 .
we follow the established method of finite - size ( here : finite - time ) scaling to extract @xmath148 from the numerical data.@xcite we rewrite the scaling law as @xmath149 where @xmath150 is an analytic function of @xmath151 . by fitting the free parameters of the series expansion @xmath152 to the data for @xmath124 as a function of @xmath57 and @xmath122 , the critical exponent @xmath148
is obtained .
the extra term @xmath153 accounts for finite - time corrections to single - parameter scaling , of the form @xmath154 we have considered times up to @xmath119 for system sizes up to @xmath155 .
the number of terms @xmath156 , @xmath157 , and @xmath158 in the series expansions are systematically increased until the chi - square - value per degree of freedom ( @xmath159 ) is approximately unity , see table [ tab : critexps ] .
the calculation is carried out at the two points indicated by circles in fig .
[ fig : pd ] , one a metal trivial - insulator transition ( giving @xmath160 ) and the other a metal topological - insulator transition ( giving @xmath161 ) .
.[tab : critexps ] results of the finite - time scaling analysis described in the text . [ cols="<,^,^,^,^,^",options="header " , ]
in conclusion , we have presented a numerical method to study the metal - insulator transition in the quantum spin hall effect ( qshe ) , based on the incorporation of @xmath0 topological symmetry into the quantum kicked rotator .
we find that the critical exponent @xmath8 of the diverging localization length is the same whether the metal is approached from the topologically trivial insulator ( @xmath5 ) or from the topologically nontrivial insulator ( @xmath4 ) .
our results @xmath162 are in agreement with refs . , but not with ref .
( which found a much smaller @xmath9 for the topologically nontrivial insulator ) . since our stroboscopic model is fully independent of the network model used in refs .
, this is significant support for the insensitivity of @xmath148 to the topological quantum number @xmath1 .
a special feature of the quantum kicked rotator is that it allows the study of the qshe , as well as the qhe,@xcite in one spatial dimension , by exchanging a dimension for an incommensurate driving frequency.@xcite there is much interest in such one - dimensional models of topological phases,@xcite because they might be more easily realized in optical lattices of cold atoms than the original two - dimensional models .
critical exponents were not studied in these earlier investigations , which focused on the qshe in clean systems without disorder .
an interesting direction for future research is to study the edge states in the @xmath0 quantum kicked rotator , by replacing the periodic boundary condition used in this work by a zero - current boundary condition . while the localization length exponent @xmath148 does not depend on the topological quantum number , the edge state structure does depend on @xmath1 , with a characteristic multifractality at the metal - insulator transition.@xcite 99 c. l. kane and e. j. mele , phys .
lett . * 95 * , 146802 ( 2005 ) ; * 95 * , 226801 ( 2005 ) .
m. knig , s. wiedmann , c. brne , a. roth , h. buhmann , l. w. molenkamp , x .-
qi , and s .- c .
zhang , science * 318 * , 766 ( 2007 ) j. maciejko , t. l. hughes , and s .- c .
zhang , ann .
cond . mat .
* 2 * , 31 ( 2011 ) .
m. z. hasan and c. l. kane , rev .
phys . * 82 * , 3045 ( 2010 ) .
qi and s .- c .
zhang , rev .
phys . * 83 * , 1057 ( 2011 ) .
s. hikami , a. i. larkin , and y. nagaoka , prog .
. phys . * 63 * , 707 ( 1980 ) .
f. evers and a. d. mirlin , rev .
* 80 * , 1355 ( 2008 ) .
m. onoda , y. avishai , and n. nagaosa , phys .
lett . * 98 * , 076802 ( 2007 ) .
y. asada , k. slevin , and t. ohtsuki , phys .
b * 70 * , 035115 ( 2004 ) .
p. marko , and l. schweitzer , j. phys .
a * 39 * , 3221 ( 2006 ) .
h. obuse , a. furusaki , s. ryu , and c. mudry , phys .
b * 76 * , 075301 ( 2007 ) ; * 78 * , 115301 ( 2008 ) .
k. kobayashi , t. ohtsuki , and k. slevin , arxiv:1201.0244 .
s. ryu , c. mudry , h. obuse , and a. furusaki , new j. phys .
* 12 * , 065005 ( 2010 ) .
g. casati , b. v. chirikov , f. m. izrailev , and j. ford , lect .
notes phys .
* 93 * , 334 ( 1979 ) .
s. fishman , d. r. grempel , and r. e. prange , phys .
lett . * 49 * , 509 ( 1982 ) .
r. scharf , j. phys .
a * 22 * , 4223 ( 1989 ) d. l. shepelyansky , physica d * 8 * , 208 ( 1983 ) ; arxiv:1102.4450 .
g. casati , i. guarneri , and d. l. shepelyansky , phys .
lett . * 62 * , 345 ( 1989 ) .
g. lemari , h. lignier , d. delande , p. szriftgiser , and j. c. garreau , phys .
lett . * 105 * , 090601 ( 2010 ) . c. tian , a. altland , and m. garst , phys .
107 * , 074101 ( 2011 ) .
j. chab , g. lemari , b. grmaud , d. delande , p. szriftgiser , and j. c. garreau , phys .
* 101 * , 255702 ( 2008 ) .
j. p. dahlhaus , j. m. edge , j. tworzydo , and c. w. j. beenakker , phys .
b * 84 * , 115133 ( 2011 ) .
i. c. fulga , f. hassler , and a. r. akhmerov , arxiv:1106.6351 .
d. meidan , t. micklitz , and p. w. brouwer , arxiv:1107.2215 .
s. murakami , s. iso , y. avishai , m. onoda , and n. nagaosa , phys .
b * 76 * , 205304 ( 2007 ) .
l. fu and c. l. kane , phys .
b * 76 * , 045302 ( 2007 ) .
b. a. bernevig , t. l. hughes , and s .- c .
zhang , science * 314 * , 1757 ( 2006 ) .
f. borgonovi and d.l .
shepelyansky , physica d * 109 * , 24 ( 1997 ) .
y. asada , k. slevin , and t. ohtsuki , physica e * 34 * , 228 ( 2006 ) .
e. prodan , j. phys .
a * 44 * , 113001 ( 2011 ) .
e. prodan , phys .
b * 83 * , 195119 ( 2011 ) . j. li , r .-
chu , j. k. jain , and s .- q .
shen , phys .
102 * , 136806 ( 2009 ) . c. w. groth , m. wimmer , a. r. akhmerov , j. tworzydo , and c. w. j. beenakker , phys .
lett . * 103 * , 196805 ( 2009 ) .
a. yamakage , k. nomura , k. imura , and y. kuramoto , j. phys .
soc . jpn .
* 80 * , 053703 ( 2011 ) .
a. mackinnon and b. kramer , z. phys .
b * 53 * , 1 ( 1983 ) . k. slevin and t. ohtsuki , phys . rev .
b * 80 * , 041304 ( 2009 ) .
t. kitagawa , m. s. rudner , e. berg , and e. demler , phys .
a * 82 * , 033429 ( 2010 ) .
n. h. lindner , g. refael , and v. galitski , nature phys .
* 7 * , 490 ( 2011 ) .
y. e. kraus , y. lahini , z. ringel , m. verbin , and o. zilberberg , arxiv:1109.5983 .
f. mei , s .- l .
zhu , z .- m .
zhang , c. h. oh , and n. goldman , phys . rev .
a * 85 * , 013638 ( 2012 ) . | the quantum kicked rotator is a periodically driven dynamical system with a metal - insulator transition .
we extend the model so that it includes phase transitions between a metal and a _ topological _ insulator , in the universality class of the quantum spin hall effect .
we calculate the @xmath0 topological invariant using a scattering formulation that remains valid in the presence of disorder .
the scaling laws at the phase transition can be studied efficiently by replacing one of the two spatial dimensions with a second incommensurate driving frequency .
we find that the critical exponent does not depend on the topological invariant , in agreement with earlier independent results from the network model of the quantum spin hall effect . |
experimental advances involving atomic fermi gases enabled the control of interactions between atoms in different hyperfine states by using feshbach resonances @xcite .
these resonances can be tuned via an external magnetic field and allow the study of dilute many body systems with fixed density , but varying interaction strength characterized by the scattering parameter @xmath4 .
this technique allows for the study of new phases of strongly interacting fermions .
for instance , the recent experiments from the mit group @xcite marked the first observation of vortices in atomic fermi gases corresponding to a strong signature of superfluidity in the @xmath1-wave ( @xmath5 ) channel .
these studies combined @xcite correspond to the experimental realization of the theoretically proposed bardeen - cooper - schrieffer ( bcs ) to bose - einstein condensation ( bec ) crossover @xcite in three dimensional ( 3d ) @xmath1-wave superfluids .
recent extensions of these ideas include trapped fermions @xcite and fermion - boson models @xcite .
arguably one of the next frontiers of exploration in ultracold fermi systems is the search for superfluidity in higher angular momentum states ( @xmath6 ) .
substantial experimental progress has been made recently @xcite in connection to @xmath7-wave ( @xmath8 ) cold fermi gases , making them ideal candidates for the observation of novel triplet superfluid phases .
these phases may be present not only in atomic , but also in nuclear ( pairing in nuclei ) , astrophysics ( neutron stars ) , and condensed matter ( organic superconductors ) systems .
the tuning of @xmath7-wave interactions in ultracold fermi gases was initially explored via @xmath7-wave feshbach resonances in trap geometries for @xmath9 in ref .
@xcite and @xmath10 in ref .
@xcite . finding and sweeping through these resonances
is difficult since they are much narrower than the @xmath1-wave case , because atoms interacting via higher angular momentum channels have to tunnel through a centrifugal barrier to couple to the bound state @xcite .
furthermore , while losses due to two body dipolar @xcite or three - body @xcite processes challenged earlier @xmath7-wave experiments , these losses were still present but were less dramatic in the very recent optical lattice experiment involving @xmath9 and @xmath7-wave feshbach resonances @xcite .
furthermore , due to the magnetic dipole - dipole interaction between valence electrons of alkali atoms , the nonzero angular momentum feshbach resonances corresponding to projections of angular momentum @xmath11 [ @xmath12 are nondegenerate ( separated from each other ) with total number of @xmath13 resonances @xcite . therefore , in principle , these resonances can be tuned and studied independently if the separation between them is larger than the experimental resolution .
since the ground state is highly dependent on the separation and detuning of these resonances , it is possible that @xmath7-wave superfluid phases can be studied from the bcs to the bec regime . for sufficiently large splittings
, it has been proposed @xcite that pairing occurs only in @xmath14 and does not occur in the @xmath15 states .
however , for small splittings , pairing occurs via a linear combination of the @xmath14 and @xmath15 states .
thus , the @xmath14 or @xmath15 resonances may be tuned and studied independently if the splitting is large enough in comparison to the experimental resolution .
the bcs to bec evolution of @xmath16-wave ( @xmath17 ) superfluidity was discussed previously in the literature using continuum @xcite and lattice @xcite descriptions in connection to high-@xmath3 superconductivity .
more recently , @xmath7-wave superfluidity was analyzed at @xmath0 for two hyperfine state ( ths ) systems in 3d @xcite , and for single hyperfine state ( shs ) systems in two dimensions ( 2d ) @xcite , using fermion - only models .
furthermore , fermion - boson models were proposed to describe @xmath7-wave superfluidity at zero @xcite and finite temperature @xcite in 3d . in this manuscript , we present a generalization of the zero and finite temperature analysis of both ths pseudo - spin singlet and shs pseudo - spin triplet @xcite superfluidity in 3d within a fermion - only description .
our main results are as follows . through an analysis of the low energy scattering amplitude within a t - matrix approach
, we find that bound states occur only when the scattering parameter @xmath18 for any @xmath11 .
the energy of the bound states @xmath19 involves only the scattering length @xmath20 for @xmath5 .
however , another parameter @xmath21 related to the interaction range @xmath22 is necessary to characterize @xmath19 for @xmath6 .
therefore , all superfluid properties for @xmath6 depend strongly on @xmath23 and @xmath4 , while for @xmath5 they depend strongly only on @xmath20 but weakly on @xmath23 .
at zero temperature ( @xmath0 ) , we study the possibility of a topological quantum phase transition in @xmath6 atomic fermi gases during the evolution from bcs to bec regime @xcite .
we show that there is a fundamental difference between the @xmath5 and @xmath6 cases . in the @xmath1-wave ( @xmath5 )
case , there is no phase transition as the magnetic field is tuned through the feshbach resonance from the bcs to the bec limit
. that is , the zero temperature thermodynamic properties are analytic functions of the scattering length @xmath20 when the feshbach resonance is crossed . in this case
, the superfluid ground state does not change in any fundamental way as @xmath20 is varied .
this has been noted in the condensed matter literature long ago @xcite and it is referred to as the bcs - bec crossover problem .
however , for @xmath6 , we show that there is a phase transition as the magnetic field is swept through the @xmath6 feshbach resonance .
the phase transition does not occur when two body bound states are first formed , but occurs when the many body chemical potential crosses a critical value . to show that such a zero temperature ( quantum ) phase transition occurs in @xmath6 , we calculate the order parameter , chemical potential , quasiparticle excitation spectrum , momentum distribution , atomic compressibility , low energy collective excitations and average cooper pair size as a function of @xmath4 , and show that they are non - analytic at @xmath0 when the chemical potential @xmath24 crosses a critical value .
the symmetry of the order parameter remains unchanged through the transition , as the ground state wavefunction experiences a major rearrangement of its analytic structure .
in addition , the elementary excitations of the superfluid also change from gapless in the bcs side to fully gapped in the bec side leading to qualitatively different thermodynamic properties in both sides .
thus , we conclude that there is a potentially observable bcs - bec phase transition in @xmath6 atomic fermi gases in contrast to the bcs - bec crossover already found in @xmath1-wave ( @xmath5 ) gases . at finite temperatures ,
we develop a gaussian fluctuation theory near the critical temperature ( @xmath25 ) to analyze the number of unbound , scattering and bound fermions as well as the chemical potential .
we show that while the saddle point number equation is sufficient in weak coupling where all fermions are unbound , the fluctuation contributions have to be taken into account in order to recover the bec physics in strong couplings where all fermions are bound .
we also derive the time - dependent ginzburg - landau ( tdgl ) functional near @xmath26 and extract the ginzburg - landau ( gl ) coherence length and time .
we recover the usual tdgl equation for bcs superfluids in weak coupling , whereas in strong coupling we recover the gross - pitaevskii ( gp ) equation for a weakly interacting dilute bose gas .
the tdgl equation exhibits anisotropic coherence lengths for @xmath6 which become isotropic only in the bec limit , in sharp contrast to the @xmath5 case , where the coherence length is isotropic for all couplings .
furthermore , for any @xmath11 , the gl time is a complex number with a larger imaginary component for @xmath27 reflecting the decay of cooper pairs into the two - particle continuum with short lifetimes .
however , the imaginary component vanishes for @xmath28 and cooper pairs become stable with long lifetimes above @xmath26 .
the rest of the paper is organized as follows . in sec .
[ sec : hamiltonian ] , we analyze the interaction potential in both real and momentum space for nonzero orbital momentum channels .
we introduce the imaginary - time functional integration formalism in sec .
[ sec : functional ] , and obtain the self - consistency ( order parameter and number ) equations .
there we also discuss the low energy scattering amplitude of a finite range interaction for all possible angular momentum channels , and relate the self - consistency equations to scattering parameters . in sec .
[ sec : ground - state ] , we discuss the evolution from bcs to bec superfluidity at zero temperature .
there we analyze the order parameter , chemical potential , quasiparticle excitation spectrum , momentum distribution , atomic compressibility and ground state energy as a function of scattering parameters .
we also discuss gaussian fluctuations and low energy collective excitations at zero temperature in sec .
[ sec : gaussian.0 ] . in sec
[ sec : normal - state ] , we present the evolution of superfluidity from the bcs to the bec regimes near the critical temperature . there we discuss the importance of gaussian fluctuations , and analyze the number of unbound , scattering and bound fermions , critical temperature and chemical potential as a function of scattering parameters . in sec .
[ sec : tdgl ] , we derive tdgl equation and extract the gl coherence length and time .
there , we recover the gl equation in the bcs and the gp equation in the bec limit .
a short summary of our conclusions is given in sec .
[ sec : conclusions ] . finally , we present in appendices [ sec : appa ] and [ sec : appb ] the coefficients for the low frequency and long wavelength expansion of the action at zero and finite temperatures , respectively .
the hamiltonian for a dilute fermi gas is given by @xmath29 where @xmath30 labels the pseudo - spins corresponding to trapped hyperfine states and @xmath31 is the volume .
these states are represented by the creation operator @xmath32 , and @xmath33 . here ,
@xmath34 where @xmath35 is the energy and @xmath36 is the chemical potential of fermions .
the interaction term can be written in a separable form @xmath37 where @xmath38 is the spin and @xmath39 is the spatial part , respectively . in the case of ths case , where @xmath40 , both pseudo - spin singlet and pseudo - spin triplet pairings are allowed .
however , we concentrate on the pseudo - spin singlet ths state with @xmath41 in addition , we discuss the shs case ( @xmath42 ) , where only pseudo - spin triplet pairing is allowed , and the interaction is given by @xmath43 in this manuscript , we analyze ths singlet and shs triplet cases for all allowable angular momentum channels .
ths triplet pairing is more involved due to the more complex nature of the vector order parameters , and therefore , we postpone this discussion for a future manuscript . the two fermion interaction can be expanded as @xmath44 and should have the necessary symmetry under the parity operation , where the transformation @xmath45 or @xmath46 leads to @xmath47 for singlet , and @xmath48 for triplet pairing .
furthermore , @xmath47 is invariant under the transformation @xmath49 , and @xmath47 reflects the pauli exclusion principle . in order to obtain an approximate expression for the atomic interaction potential , we use the fourier expansion of a plane wave in 3d @xmath50 where @xmath51 is the spherical bessel function of order @xmath11 and @xmath52 is the spherical harmonic of order @xmath53 , in eq .
[ eqn : vreal ] to evaluate the matrix elements of the interaction potential in @xmath54-space @xmath55 here , @xmath56 , and @xmath57 denotes the angular dependence @xmath58 .
the @xmath59 dependent coefficients @xmath60 are related to the real space potential @xmath61 through the relation @xmath62 the index @xmath11 labels angular momentum states in 3d , with @xmath63 corresponding to @xmath64 channels , respectively . in the long wavelength limit ( @xmath65 ) , one can show that the @xmath66 dependence of this potential becomes exactly separable .
in fact , for @xmath67 and @xmath68 , the asymptotic expression of the spherical bessel function for small arguments can be used , giving @xmath69 , with the coefficient @xmath70 dependent on the particular choice of the real space potential . in the opposite limit , where @xmath71 and @xmath72 , the potential is not separable . in this case , @xmath60 mixes different @xmath66 and @xmath73 , and shows an oscillatory behavior ( which is dependent on the exact form of @xmath61 ) with a decaying envelope that is proportional to @xmath74 . under these circumstances
, we choose to study a model potential that contains most of the features described above .
one possibility is to retain only one of the @xmath11 terms in eq .
( [ eqn : vsh ] ) , by assuming that the dominant contribution to the scattering process between fermionic atoms occurs in the @xmath11th angular momentum channel . this assumption may be experimentally relevant since atom - atom dipole interactions split different angular momentum channels such that they may be tuned independently . using the properties discussed above , we write @xmath75 where @xmath76 is the interaction strength , and
the function @xmath77 describes the momentum dependence . here
, @xmath78 plays the role of the interaction range in real space and sets the scale at small and large momenta . in addition , the diluteness condition ( @xmath79 ) requires @xmath80 , where @xmath81 is the density of atoms and @xmath82 is the fermi momentum .
this function reduces to @xmath83 for small @xmath66 , and behaves as @xmath84 for large @xmath66 , which guarantees the correct qualitative behavior expected for @xmath60 according to the analysis above .
in this section , we describe in detail the ths singlet case for even angular momentum states . a similar approach for the shs triplet case for odd angular momentum states can be found in ref .
@xcite , and therefore , we do not repeat the same analysis here . however , we point out the main differences between the two cases whenever it is necessary . in the imaginary - time functional integration formalism ( @xmath85 , and @xmath86 ) , the partition function for the ths singlet case can be written as @xmath87 with action @xmath88\ ] ] where the hamiltonian for the @xmath11th angular momentum channel is @xmath89 here , @xmath90 and @xmath91 we first introduce the nambu spinor @xmath92 , where @xmath93 denotes both momentum and fermionic matsubara frequency @xmath94 , and use a hubbard - stratonovich transformation @xmath95 \nonumber \\ & & e^{\sum_{q}\left[\frac{|\phi(q)|^2}{\lambda } + b^\dagger(q)\phi(q ) + \phi^\dagger(q ) b(q ) \right]}\end{aligned}\ ] ] to decouple fermionic and bosonic degrees of freedom .
integration over the fermionic part [ @xmath96 leads to the action @xmath97 , \label{eqn : s_eff}\end{aligned}\ ] ] where @xmath98 , with bosonic matsubara frequency @xmath99 .
here , @xmath100\delta_{p , p'}\end{aligned}\ ] ] is the inverse nambu propagator , @xmath101 is the bosonic field , and @xmath102 and @xmath103 is the pauli spin matrix .
the bosonic field @xmath104 has @xmath105-independent @xmath106 and @xmath105-dependent @xmath107 parts . performing an expansion in @xmath108 to
quadratic order in @xmath107 leads to @xmath109 where the vector @xmath110 is such that @xmath111 $ ] , and @xmath112 are the matrix elements of the inverse fluctuation propagator matrix @xmath113 .
furthermore , @xmath114 is the saddle point action given by @xmath115 , \end{aligned}\ ] ] and the saddle point inverse nambu propagator is @xmath116 with saddle point order parameter @xmath117 notice that , @xmath118 may involve several different @xmath119 for a given angular momentum channel @xmath11 .
the matrix elements of the inverse fluctuation matrix @xmath120 are given by @xmath121 notice that while @xmath122 @xmath123 are even under the transformations @xmath124 and @xmath125 ; @xmath126 @xmath127 are even only under @xmath124 , having no defined parity in @xmath128 .
the gaussian action eq .
( [ eqn : gaction ] ) leads to the thermodynamic potential @xmath129 , where @xmath130 \big\rbrace , \label{eqn : sptp } \\ \omega_\ell^{\rm{fluct } } & = & \frac{1}{\beta } \sum_{q}\ln\det[\mathbf{f}^{-1}_{\ell}(q)/(2\beta ) ] \label{eqn : flucttp}\end{aligned}\ ] ] are the saddle point and fluctuation contributions , respectively . here , @xmath131^{\frac{1}{2}},\ ] ] is the quasiparticle energy spectrum .
having completed the presentation of the functional integral formalism , we discuss next the self - consistency equations for the order parameter and the chemical potential .
the saddle point condition @xmath132 leads to the order parameter equation @xmath133 which can be expressed in terms of experimentally relevant parameters via the @xmath134-matrix approach @xcite .
the low energy two body scattering amplitude between a pair of fermions in the @xmath11th angular momentum channel is given by @xcite @xmath135 where @xmath136 and @xmath4 are the effective range and scattering parameter , respectively . here
@xmath21 has dimensions of @xmath137 and @xmath4 has dimensions of @xmath138 , where @xmath139 is the size of the system .
the energy of the two body bound state is determined from the poles of @xmath140 , and is given by @xmath141 .
bound states occur when @xmath142 for @xmath5 , and @xmath143 for @xmath6 .
since @xmath144 , bound states occur only when @xmath18 for all @xmath11 , in which case the binding energies are given by @xmath145 notice that , only a single parameter ( @xmath20 ) is sufficient to describe the low energy two body problem for @xmath5 , while two parameters @xmath146 are necessary to describe the same problem for @xmath6 .
the point at which @xmath147 corresponds to the threshold for the formation of a two body bound state in vacuum . beyond this threshold ,
@xmath20 for @xmath5 and @xmath148 for @xmath6 are the size of the bound states . for any @xmath11
, the two body scattering amplitude is related to the @xmath134-matrix via @xmath149,\end{aligned}\ ] ] where the @xmath134-matrix is given by @xmath150 using the spherical harmonics expansion for both @xmath39 and @xmath151 leads to two coupled equations , @xmath152 relating @xmath153 and @xmath23 to @xmath4 and @xmath21 . except for notational differences , notice that these relations are identical to previous results @xcite .
after performing momentum integrations we obtain @xmath154 where @xmath155 and @xmath156 here @xmath157 is the gamma function .
notice that , @xmath158 diverges and changes sign when @xmath159 , which corresponds to the critical coupling for feshbach resonances ( the unitarity limit ) . [ htb ] [ htb ] in addition , the scattering parameter has a maximum value in the zero ( @xmath160 ) and a minimum value in the infinite ( @xmath161 ) coupling limits given respectively by @xmath162 the first condition eq .
( [ eqn : abcs ] ) ( when @xmath160 ) follows from eq .
( [ eqn : ar ] ) where @xmath163 has to be satisfied for all possible @xmath164 . however , there is no condition on @xmath165 for @xmath5 , and @xmath166 in the bcs limit .
the second condition eq .
( [ eqn : abec ] ) ( when @xmath161 ) follows from eq .
( [ eqn : ka ] ) , which is valid for all possible @xmath11 .
the minimum @xmath4 for a finite range interaction is associated with the pauli principle , which prevents two identical fermions to occupy the same state .
thus , while the scattering parameter can not be arbitrarily small for a finite range potential , it may go to zero as @xmath167 .
furthermore , the binding energy is given by @xmath168 when @xmath169 . in fig .
[ fig : swave.lambda ] , we plot the original interaction strength @xmath170 versus the scattering parameter @xmath171 for the @xmath1-wave ( @xmath5 ) channel .
notice that , @xmath172 in the bcs and @xmath173 in the bec limit .
a divergence in @xmath174 corresponds to a @xmath1-wave feshbach resonance occurring at @xmath175 . in fig .
[ fig : pwave.lambda ] , we plot the original interaction strength @xmath176 versus the scattering parameter @xmath177 for the @xmath7-wave ( @xmath8 ) channel . notice that , @xmath178 in the bcs and @xmath179 in the bec limit .
a divergence in @xmath177 corresponds to a @xmath7-wave feshbach resonance occurring at @xmath180 .
thus , the order parameter equation in terms of the scattering parameter is rewritten as @xmath181 } { 2e_\ell(\mathbf{k } ) } \right ] \nonumber \\ & & \hskip 2 mm \delta_{\ell , m'_\ell } \gamma_\ell^2(k ) y_{\ell , m_\ell}^*(\widehat{\mathbf{k}})y_{\ell , m'_\ell}(\widehat{\mathbf{k } } ) .
\label{eqn : opeqn}\end{aligned}\ ] ] this equation is valid for both ths pseudo - spin singlet and shs pseudo - spin triplet states .
however , there is one important difference between pseudo - spin singlet and pseudo - spin triplet states . for pseudo - spin singlet states ,
the order parameter is a scalar function of @xmath54 , while it is a vector function for pseudo - spin triplet states discussed next .
in general , the triplet order parameter can be written in the standard form @xcite @xmath182 where the vector @xmath183 $ ] is an odd function of @xmath54 .
thefore , all up - up , down - down and up - down components may exist for a ths pseudo - spin triplet interaction .
however , in the shs pseudo - spin triplet case only the up - up or down - down component may exist leading to @xmath184 .
thus , for the up - up case @xmath185 and @xmath186 , leading to @xmath187 , which breaks time reversal symmetry , as expected from a fully spin polarized state .
the corresponding down - down state has @xmath188 .
furthermore , the simplified form of the shs triplet order parameter allows a treatment similar to that of ths singlet states .
however , it is important to mention that the ths triplet case can be investigated using our approach , but the treatment is more complicated .
the order parameter equation has to be solved self - consistently with the number equation @xmath189 where @xmath190 is the full thermodynamic potential . in the approximations
used , @xmath191 has two contributions .
the saddle point contribution to the number equation is @xmath192 where @xmath193 is the momentum distribution given by @xmath194 .
\label{eqn : nk}\ ] ] for the shs triplet case , the summation over @xmath1 is not present in @xmath195 .
the fluctuation contribution to the number equation is @xmath196/\partial { \mu_\ell } } { \det \mathbf{f}^{-1}_{\ell}(q)},\ ] ] where @xmath113 is the inverse fluctuation matrix defined in eq .
( [ eqn : fluct.f11 ] ) and ( [ eqn : fluct.f12 ] ) . in the rest of the paper ,
we analyze analytically the superfluid properties at zero temperature ( ground state ) and near the critical temperatures for ths singlet ( only even @xmath11 ) and shs triplet ( only odd @xmath11 ) cases . in addition , we analyze numerically the @xmath1-wave ( @xmath5 ) channel of ths singlet and @xmath7-wave ( @xmath8 ) channel of shs triplet cases , which are currently of intense theoretical and experimental interest in ultracold fermi atoms .
at low temperatures , the saddle point self - consistent ( order parameter and number ) equations are sufficient to describe ground state properties in the weak coupling bcs and strong coupling bec limits @xcite
. however , fluctuation corrections to the number equation may be important in the intermediate regime @xcite .
ground state properties ( @xmath0 ) are investigated by solving saddle point self - consistency ( order parameter and number ) equations to obtain @xmath106 and @xmath24 , which are discussed next .
we discuss in this section @xmath106 and @xmath24 . in weak coupling
, we first introduce a shell about the fermi energy @xmath197 such that @xmath198 , inside of which one may ignore the 3d density of states factor ( @xmath199 ) and outside of which one may ignore @xmath200 . while in sufficiently strong coupling
, we use @xmath201 to derive the analytic results discussed below .
it is important to notice that , in strictly weak and strong coupling , the self - consistency equations eq .
( [ eqn : spnumbereqn ] ) and ( [ eqn : opeqn ] ) are decoupled , and play reversed roles . in weak ( strong )
coupling the order parameter equation determines @xmath106 ( @xmath24 ) and the number equation determines @xmath24 ( @xmath106 ) . in weak coupling , the number equation eq .
( [ eqn : spnumbereqn ] ) leads to @xmath202 for any @xmath11 where @xmath203 is the fermi energy . in strong coupling , the order parameter equation eq .
( [ eqn : opeqn ] ) leads to @xmath204 where @xmath205 and @xmath157 is the gamma function .
this calculation requires that @xmath206 for @xmath5 and that @xmath207 for @xmath6 for the order parameter equation to have a solution with @xmath208 in the strong coupling limit . in the bec limit
@xmath209 $ ] for @xmath5 .
notice that , @xmath210 when @xmath211 [ or @xmath212 , and thus , we recover the contact potential ( @xmath167 ) result . in the same spirit , to obtain the expressions in eq .
( [ eqn : mu_0 ] ) and ( [ eqn : mu_ell ] ) , we assumed @xmath213 .
notice that , @xmath214 in this limit for any @xmath11 .
on the other hand , the solution of the order parameter equation in the weak coupling limit is @xmath215 , \\ & & \exp\left[t_\ell \left(\frac{k_0}{k_{\rm f}}\right)^{2\ell -1 } - \frac{\pi}{2k_{\rm f}^{2\ell+1}|a_\ell|}\right],\end{aligned}\ ] ] where @xmath216 and @xmath217 .
these expressions are valid only when the exponential terms are small .
the solution of the number equation in the strong coupling limit is @xmath218 to order @xmath219 , where we assumed that @xmath220 for sufficiently strong couplings with @xmath213 .
[ htb ] [ htb ] next , we present numerical results for two particular states .
first , we analyze the ths @xmath1-wave ( @xmath221 ) case , where @xmath222 with @xmath223 second , we discuss the shs @xmath7-wave ( @xmath224 ) case , where @xmath225 with @xmath226 in all numerical calculations , we choose @xmath227 to compare @xmath1-wave and @xmath7-wave cases . in figs .
[ fig : swave.gap ] and [ fig : swave.mu0 ] , we show @xmath228 and @xmath229 at @xmath0 for the @xmath1-wave case .
notice that the bcs to bec evolution range in @xmath230 is of order @xmath231 .
furthermore , @xmath228 grows continuously without saturation with increasing coupling , while @xmath229 changes from @xmath232 to @xmath233 continuously and decreases as @xmath234 for strong couplings .
thus , the evolution of @xmath228 and @xmath229 as a function of @xmath235 is smooth . for completeness , it is also possible to obtain analytical values of @xmath20 and @xmath236 when the chemical potential vanishes .
when @xmath237 , we obtain for @xmath238 at @xmath239 \approx 0.554 , $ ] which also agrees with the numerical results . here
@xmath157 is the gamma function .
[ htb ] [ htb ] in figs .
[ fig : pwave.gap ] and [ fig : pwave.mu0 ] , we show @xmath240 and @xmath241 at @xmath0 for the @xmath7-wave case .
notice that the bcs to bec evolution range in @xmath242 is of order @xmath243 .
furthermore , @xmath240 grows with increasing coupling but saturates for large @xmath242 , while @xmath241 changes from @xmath232 to @xmath244 continuously and decreases as @xmath245 for strong couplings . for completeness , we present the limiting expressions @xmath246 , \\ in the weak and strong coupling limits , respectively . the evolution of @xmath240 and @xmath241 are qualitatively similar to recent @xmath0 results for ths fermion @xcite and shs fermion - boson @xcite models . due to the angular dependence of @xmath247 ,
the quasiparticle excitation spectrum @xmath248 is gapless for @xmath249 , and fully gapped for @xmath250 .
furthermore , both @xmath251 and @xmath241 are nonanalytic exactly when @xmath241 crosses the bottom of the fermion energy band @xmath252 at @xmath253
. the nonanalyticity does not occur in the first derivative of @xmath251 or @xmath241 as it is the case in 2d @xcite , but occurs in the second and higher derivatives .
thus , in the @xmath7-wave case , the bcs to bec evolution is not a crossover , but a quantum phase transition occurs , as can be seen in the quasiparticle excitation spectrum to be discussed next .
the quasiparticle excitation spectrum @xmath254^{1/2 } $ ] is gapless at @xmath54-space regions where the conditions @xmath255 and @xmath256 are both satisfied .
notice that the second condition is only satisfied in the bcs side @xmath257 , and therefore , the excitation spectrum is always gapped in the bec side ( @xmath208 ) . for @xmath5 ,
the order parameter is isotropic in @xmath54-space without zeros ( nodes ) since it does not have any angular dependence .
therefore , the quasiparticle excitation spectrum is fully gapped in both bcs ( @xmath258 ) and bec ( @xmath259 ) sides , since @xmath260 here , @xmath261
. this implies that the evolution of the quasiparticle excitation spectrum from weak coupling bcs to strong coupling bec regime is smooth when @xmath237 for @xmath5 pairing .
[ htb ] in fig .
[ fig : swave.qp ] , we show @xmath262 for an @xmath1-wave ( @xmath221 ) superfluid when a ) @xmath258 ( bcs side ) for @xmath263 , b ) @xmath237 ( intermediate regime ) for @xmath264 , and c ) @xmath259 ( bec side ) for @xmath265 .
notice that the quasiparticle excitation spectrum is gapped for all three cases .
however , the situation for @xmath6 is very different as discussed next . for @xmath6 ,
the order parameter is anisotropic in @xmath54-space with zeros ( nodes ) since it has an angular dependence .
therefore , while the quasiparticle excitation spectrum is gapless in the bcs ( @xmath266 ) side , it is fully gapped in the bec ( @xmath267 ) side , since @xmath268 this implies that the evolution of quasiparticle excitation spectrum from weak coupling bcs to strong coupling bec regime is not smooth for @xmath6 pairing having a nonanalytic behavior when @xmath269 .
this signals a quantum phase transition from a gapless to a fully gapped state exactly when @xmath270 drops below the bottom of the energy band @xmath269 .
[ htb ] in fig .
[ fig : pwave.qp ] , we show @xmath271 for a @xmath7-wave ( @xmath224 ) superfluid when a ) @xmath249 ( bcs side ) for @xmath272 , b ) @xmath252 ( intermediate regime ) for @xmath273 , and c ) @xmath250 ( bec side ) for @xmath274 .
the quasiparticle excitation spectrum is gapless when @xmath275 and @xmath276 are both satisfied in certain regions of @xmath54-space . for @xmath277 ,
these conditions are met only when @xmath278 and @xmath279 for a given @xmath241 .
notice that , these points come closer as the interaction ( @xmath241 ) increases ( decreases ) , and when @xmath252 they become degenerate at @xmath280 . for @xmath250 ,
the second condition can not be satisfied , and thus , a gap opens in the excitation spectrum of quasiparticles as shown in fig .
[ fig : pwave.qp]c .
the spectrum of quasiparticles plays an important role in the thermodynamic properties of the evolution from bcs to bec regime at low temperatures . for @xmath5 ,
thermodynamic quantities depend exponentially on @xmath134 throughout the evolution .
thus , a smooth crossover occurs at @xmath237 .
however , for @xmath6 , thermodynamic quantitites depend exponentially on @xmath134 only in the bec side , while they have a power law dependence on @xmath134 in the bcs side .
thus , a non - analytic evolution occurs at @xmath269 .
this can be seen best in the momentum distribution which is discussed next . in this section
, we analyze the momentum distribution @xmath281/2 $ ] in the bcs @xmath282 and bec sides @xmath283 , which reflect the gapless to gapped phase transition for nonzero angular momentum superfluids .
[ htb ] in fig .
[ fig : swave.md ] , we show @xmath284 for an @xmath1-wave ( @xmath221 ) superfluid when a ) @xmath258 ( bcs side ) for @xmath263 , b ) @xmath237 ( intermediate regime ) for @xmath264 , and c ) @xmath259 ( bec side ) for @xmath265 .
as the interaction increases the fermi sea with locus @xmath285 is suppressed , and pairs of atoms with opposite momenta become more tightly bound . as a result , @xmath286 broadens in the bec side since fermions with larger momentum participate in the formation of bound states .
notice that , the evolution is a crossover without any qualitative change .
furthermore , @xmath287 and @xmath288 can be trivially obtained from @xmath289 , since @xmath290 is symmetric in @xmath291 and @xmath292 .
[ htb ] in fig .
[ fig : pwave.md ] , we show @xmath293 for a @xmath7-wave ( @xmath224 ) superfluid when a ) @xmath249 ( bcs side ) for @xmath272 , b ) @xmath252 ( intermediate regime ) for @xmath273 , and c ) @xmath250 ( bec side ) for @xmath274 .
notice that @xmath293 is largest in the bcs side when @xmath294 , but it vanishes along @xmath294 for any @xmath295 in the bec side . as the interaction increases the fermi sea with locus @xmath296 is suppressed , and pairs of atoms with opposite momenta become more tightly bound . as a result , the large momentum distribution in the vicinity of @xmath280 splits into two peaks around finite @xmath297 reflecting the @xmath7-wave symmetry of these tightly bound states . furthermore ,
@xmath298/2 $ ] for any @xmath241 , and @xmath299 is trivially obtained from @xmath293 , since @xmath300 is symmetric in @xmath291 . here , @xmath301 is the sign function .
thus , @xmath300 for the @xmath7-wave case has a major rearrangement in @xmath54-space with increasing interaction , in sharp contrast to @xmath1-wave .
this qualitative difference between @xmath7-wave and @xmath1-wave symmetries around @xmath280 explicitly shows a direct measurable consequence of the gapless to gapped quantum phase transition when @xmath252 , since @xmath300 depends explicitly on @xmath302 .
these quantum phase transitions are present in all nonzero angular momentum states , and can be further characterized through the atomic compressibility as discussed in the next section . at finite temperatures ,
the isothermal atomic compressibility is defined by @xmath303 where @xmath31 is the volume and @xmath304 is the pressure of the gas .
this can be rewritten as @xmath305 where the partial derivative @xmath306 at @xmath307 is given by @xmath308 the expression above leads to @xmath309 in weak coupling bcs and @xmath310 in strong coupling bec limit for @xmath5 , where @xmath311 is the density of states per spin at the fermi energy .
notice that @xmath312 decreases as @xmath313 in strong coupling since @xmath314 .
however , we only present the strong coupling results for higher angular momentum states since they exhibit an interesting dependence on @xmath4 and @xmath23 . in the case of ths pseudo - spin singlet , we obtain @xmath315 for @xmath316 , while in the case of shs states we obtain @xmath317 for @xmath8 and @xmath318 for @xmath316 . here
@xmath205 and @xmath319 where @xmath157 is the gamma function .
notice that @xmath320 decreases as @xmath321 for @xmath8 since @xmath322 and @xmath323 is a constant for @xmath316 in strong coupling .
[ htb ] in fig .
[ fig : swave.ac ] , we show the evolution of @xmath324 for a @xmath1-wave @xmath325 superfluid from the bcs to the bec regime .
@xmath324 decreases continuously , and thus the evolution is a crossover ( smooth ) as can be seen in the inset where the numerical derivative of @xmath312 with respect to @xmath230 is shown @xmath326 \}$ ] .
this decrease is associated with the increase of the gap of the excitation spectrum as a function of @xmath230 . in this approximation ,
the gas is incompressible [ @xmath327 in the extreme bec limit . in fig .
[ fig : pwave.ac ] , we show the evolution of @xmath328 for a @xmath7-wave @xmath329 superfluid from the bcs to the bec regime .
notice that , there is a change in qualitative behavior when @xmath252 at @xmath253 as can be seen in the inset where the numerical derivative of @xmath320 with respect to @xmath242 is shown @xmath330 \}$ ] .
thus , the evolution from bcs to bec is not a crossover , but a quantum phase transition occurs when @xmath252 @xcite .
[ htb ] the non - analytic behavior occurring when @xmath269 can be understood from higher derivatives of @xmath331 with respect to @xmath24 @xmath332_{t,{\cal v } } = - 2n_\ell [ \kappa_\ell^t(t)]^2 + \frac{1}{n_\ell^2 } \left ( \frac{\partial^2 n_\ell } { \partial \mu_\ell^2 } \right)_{t,{\cal v}}.\ ] ] for instance , the second derivative @xmath333 $ ] tends to zero in the weak ( @xmath334 ) and strong ( @xmath335 ) coupling limits . on the other hand ,
when @xmath336 , @xmath337 is finite only for @xmath5 , and it diverges for @xmath6 .
this divergence is logarithmic for @xmath8 , and of higher order for @xmath316 .
thus , we conclude again that higher derivatives of @xmath338 are nonanalytic when @xmath269 , and that a quantum phase transition occurs for @xmath6 .
theoretically , the calculation of the isothermal atomic compressibility @xmath339 is easier than the isentropic atomic compressibility @xmath340 .
however , performing measurements of @xmath340 may be simpler in cold fermi gases , since the gas expansion upon release from the trap is expected to be nearly isentropic .
fortunately , @xmath340 is related to @xmath339 via the thermodynamic relation @xmath341 where @xmath342 since specific heat capacitites @xmath343 .
furthermore , at low temperatures ( @xmath344 ) the ratio @xmath345 , and therefore , @xmath346 .
thus , we expect qualitatively similar behavior in both the isentropic and isothermal compressibilities at low temperatures @xmath347
. the measurement of the atomic compressibility could also be performed via an analysis of particle density fluctuations @xcite .
as it is well know from thermodynamics @xcite , @xmath348 is connected to density fluctuations via the relation @xmath349 where @xmath350 is the average density of atoms . from the measurement of density fluctuations @xmath351
can be extracted at any temperature @xmath134 .
it is important to emphasize that in this quantum phase transition at @xmath269 , the symmetry of the order parameter does not change as is typical in the landau classification of phase transitions .
however , a clear thermodynamic signature occurs in derivatives of the compressibility suggesting that the phase transition is higher than second order according to ehrenfest s classification .
therefore , if the symmetry of the order parameter does not change when @xmath24 changes sign , what is changing ?
to address this question , the topology of momentum space is discussed next . in what follows
, we discuss the role of momentum space topology @xcite in the non - analytic behavior of the thermodynamic potential , when @xmath269 .
to investigate the role of topology , we make an immediate connection to the lifshitz transition @xcite in the context of ordinary metals at @xmath0 and high pressure . in the conventional lifshitz transition , the fermi surface @xmath352 changes its topology as the pressure @xmath304 is changed . for an isotropic pressure @xmath304 ,
the deviation @xmath353 from the critical pressure @xmath354 is proportional to @xmath355 where @xmath356 is the critical chemical potential at the transition point .
a typical example of the lifshitz transition is the disruption of a neck of the fermi surface which leads to a non - analytic behavior of the number of states @xmath357 inside the fermi surface . in this case
, @xmath357 behaves as @xmath358 for @xmath359 , and as @xmath360 for @xmath361 , in the vicinity of @xmath356 . here
, @xmath362 is the electronic compressibility , where @xmath363 is the critical density of electrons at the transition point .
notice that @xmath364 is nonanalytic , but it is not singular .
the quantity that signals a phase transition in this case is not @xmath364 , but the thermopower @xmath365 , which is proportional to @xmath366 , thus leading to @xmath367 . in the conventional lifshitz transition
, the system lowers its energy by @xmath368 and the transition is said to be of second and half order @xcite .
the topological transition proposed here is analogous to the lifshitz transition in the sense that the surface in momentum space corresponding to @xmath369 changes from a well defined set of @xmath54 points for @xmath266 to a null set for @xmath267 . here
, @xmath370 plays the role of @xmath371 and @xmath372 plays the role of the critical pressure @xmath354 . for the lifshitz transition in @xmath6 superfluids
, there is a non - analytic behavior in @xmath373 , and thus in @xmath374 .
this behavior in @xmath375 is due to the collapse of all order parameter nodes at @xmath376 , which produce a gap in the excitation spectrum @xmath377 and a massive discontinuous rearrangement of the momentum distribution @xmath378 in the ground state as @xmath379 .
a direct topological analogy with the standard lifshitz transition can be made by noticing the collapse of locus of zero quasiparticle excitation energy at @xmath380 in the excitation spectrum of the system .
generalized topological invariants can be invented along the lines of ref .
@xcite , however , we do not discuss them here .
instead , we analyze next the phase diagram at zero temperature . to have a full picture of the evolution from the bcs to the bec limit at @xmath0 , it is important to analyze thermodynamic quantities at low temperatures . in particular , it is important to determine the quantum critical region ( qcr ) where a qualitative change occurs in quantities such as the specific heat , compressibility and spin susceptibility . here
, we do not discuss in detail the qcr , but we analyze the contributions from quasiparticle excitations to thermodynamic properties .
however , the discussion can be extended to include collective excitations @xcite ( see sec . [
sec : gaussian.0 ] ) .
next , we point out a major difference between @xmath5 and @xmath6 states in connection with the spectrum of the quasiparticle excitations ( see sec .
[ sec : quasiparticle ] ) and their contribution to low temperature thermodynamics .
[ htb ] for @xmath5 , quasiparticle excitations are gapped for all couplings , and therefore , thermodynamic quantitites such as atomic compressibility , specific heat and spin susceptibility have an exponential dependence on the temperature and the minimum energy of quasiparticle excitations @xmath381 $ ] .
using eqs .
( [ eqn : swave.qp.bcs ] ) and ( [ eqn : swave.qp.bec ] ) leads to @xmath382 $ ] in the bcs side ( @xmath258 ) and @xmath383 $ ] in the bec side ( @xmath259 ) as shown in fig . [
fig : swave.phase ] , where @xmath261 .
notice that , there is no qualitative change across @xmath237 at small but finite temperatures .
this indicates the absence of a qcr and confirms there is only a crossover for @xmath1-wave ( @xmath5 ) superfluids at @xmath0 .
[ htb ] for @xmath6 , quasiparticle excitations are gapless in the bcs side and are only gapped in the bec side , and therefore , while thermodynamic quantitites such as atomic compressibility , specific heat and spin susceptibility have power law dependences on the temperature @xmath384 in the bcs side , they have exponential dependences on the temperature and the minimum energy of quasiparticle excitations @xmath385 $ ] in the bec side .
here , @xmath386 is a real number which depends on particular @xmath11 state . for @xmath8 , using eqs .
( [ eqn : pwave.qp.bcs ] ) and ( [ eqn : pwave.qp.bec ] ) leads to @xmath387 in the bcs side ( @xmath249 ) and @xmath388 in the bec side ( @xmath250 ) as shown in fig . [
fig : pwave.phase ] . notice the change in qualitative behavior across @xmath252 ( as well as other @xmath6 states ) at small but finite temperatures .
this change occurs within the qcr and signals the existence of a quantum phase transition ( @xmath0 ) for @xmath6 superfluids .
now , we discuss the thermodynamic potential @xmath389 at @xmath0 in the asmyptotic bcs and bec limits . in the weak coupling limit , we obtain @xmath390 which is identical to the full thermodynamic potential @xmath389 .
this indicates that @xmath391 is negligible in the bcs limit .
however , in the strong coupling limit , we obtain @xmath392 notice that , @xmath393 is the bosonic chemical potential and @xmath394 is the number of bosons . to evaluate @xmath395 ,
it is necessary to find the first nonvanishing correction for @xmath396 . in the specific case of @xmath5 ,
we obtain @xmath397 for the chemical potential and @xmath398 for the thermodynamic potential of the pairs . here
@xmath399 and @xmath400 is the scattering length and mass of the corresponding bosons .
a better estimate for @xmath401 can be found in the literature @xcite .
the main reason for this difference is that our theory does not include possible intermediate ( virtual ) scattering processes which renormalize @xmath402 .
this is also the case when we analyze the collective modes in sec .
[ sec : cm - sc ] and the tdgl equation in sec .
[ sec : tdgl.sc ] . using @xmath403 and the thermodynamic relation @xmath404 , where @xmath405 is the ground state energy , we obtain @xmath406 notice that this expression is identical to the thermodynamic potential of bosons @xmath407 where @xmath408 is the ground state energy .
therefore , the fermionic thermodynamic potential in the strong coupling limit should lead to the thermodynamic potential of real bosons ( @xmath409 ) . since @xmath410 we conclude from this thermodynamic argument that @xmath411 in the strong coupling limit
therefore , @xmath412 or @xmath413 which is consistent with quantum monte carlo calculations @xcite .
the fluctuation contribution to @xmath391 comes from the zero point energy of the collective excitations , which is discussed next .
the pole structure of @xmath414 determines the two - particle excitation spectrum of the superconducting state with @xmath415 , and has to be taken into account to derive @xmath416 .
the matrix elements of @xmath414 are @xmath417 for a given @xmath11 .
we focus here only on the zero temperature limit and analyse the collective phase modes . in this limit
, we separate the diagonal matrix elements of @xmath418 into even and odd contributions with respect to @xmath128 @xmath419 } \nonumber \\ & & \gamma_{\ell}^2(k ) y_{\ell , m_\ell}(\widehat{\mathbf{k } } ) y_{\ell , m_\ell'}^*(\widehat{\mathbf{k } } ) - \frac{\delta_{m_\ell , m_\ell'}{\cal v}}{4\pi \lambda_\ell } , \\ ( \mathbf{f}_{\ell , m_\ell , m_\ell'}^{-1})_{11}^o & = & - \sum_{\mathbf{k } } \frac{(\xi_+ \xi_- + e_+ e_-)(iv_j ) } { 2e_+ e_- [ ( iv_j)^2 - ( e_+ + e_-)^2 ] } \nonumber \\ & & \gamma_{\ell}^2(k ) y_{\ell , m_\ell}(\widehat{\mathbf{k } } ) y_{\ell , m_\ell'}^*(\widehat{\mathbf{k}}).\end{aligned}\ ] ] the off - diagonal term is even in @xmath128 , and it reduces to @xmath420 } \nonumber \\ & & \gamma_{\ell}^2(k ) y_{\ell , m_\ell}(\widehat{\mathbf{k } } ) y_{\ell , m_\ell'}^*(\widehat{\mathbf{k}}).\end{aligned}\ ] ] here the labels @xmath421 denote that the corresponding variables are functions of @xmath422 . in order to obtain the collective mode spectrum , we express @xmath423 / \sqrt{2}$ ] where @xmath424 , @xmath425 , @xmath426 and @xmath427 are all real .
notice that the new fields @xmath428 , $ ] and @xmath429 $ ] can be regarded essentially as the amplitude and phase fields respectively , when @xmath425 is small .
this change of basis can be described by the following unitary transformation @xmath430.\end{aligned}\ ] ] from now on , we take @xmath106 as real without loss of generality . the diagonal elements of the fluctuation matrix in the rotated basis are @xmath431 and @xmath432 and the off - diagonal elements are @xmath433 with the @xmath434 dependence being implicit .
the collective modes are determined by the poles of the propagator matrix @xmath435 for the pair fluctuation fields @xmath107 , which describe the gaussian deviations about the saddle point order parameter .
the poles of @xmath435 are determined by the condition @xmath436 , which leads to @xmath437 collective ( amplitude and phase ) modes , when the usual analytic continuation @xmath438 is performed . among them
, there are @xmath439 amplitude modes which we do not discuss here .
the easiest way to get the phase collective modes is to integrate out the amplitude fields to obtain a phase - only effective action .
notice that , for @xmath6 channels at any temperature , and for @xmath5 channel at finite temperature , a well defined low frequency expansion is not possible for @xmath27 due to landau damping which causes the collective modes to decay into the two quasiparticle continuum . a well defined expansion [ collective mode dispersion @xmath440 must satisfy the following condition @xmath441 thus
, a zero temperature expansion is always possible when landau damping is subdominant ( underdamped regime ) . to obtain the long wavelength dispersions for the collective modes at @xmath0 , we expand the matrix elements of @xmath442 to second order in @xmath443 and @xmath444 to get @xmath445 the expressions for the expansion coefficients
are given in app .
[ sec : appa ] . for @xmath5 , the coefficients @xmath446 and @xmath447 are diagonal and isotropic in @xmath448 , and @xmath449 vanishes . here
, @xmath450 is the kronecker delta .
thus , the collective mode is the isotropic goldstone mode with dispersion @xmath451 where @xmath452 is the speed of sound .
notice that the quasiparticle excitations are always fully gapped from weak to strong coupling , and thus , the goldstone mode is not damped at @xmath0 for all couplings . for @xmath6 ,
the dispersion for collective modes is not easy to extract in general , and therefore , we consider the case when only one of the spherical harmonics @xmath52 is dominant and characterizes the order parameter . in this case ,
@xmath453 = 0 due to the order parameter equation , and the collective mode is the anisotropic goldstone mode with dispersion @xmath454^\frac{1}{2 } , \\ { \cal c}_{\ell \ne 0 , m_\ell}^{i , j } & = & \left ( \frac{a_{\ell , m_\ell , m_\ell } q_{\ell , m_\ell , m_\ell}^{i , j } } { a_{\ell , m_\ell , m_\ell } r_{\ell , m_\ell , m_\ell } + b_{\ell , m_\ell , m_\ell}^2 } \right)^\frac{1}{2}. \label{eqn : soundv.l}\end{aligned}\ ] ] notice that the speed of sound has a tensor structure and is anisotropic .
furthermore , the quasiparticle excitations are gapless when @xmath266 , and thus , the goldstone mode is damped even at @xmath0 .
however , landau damping is subdominant and the real part of the pole dominates for small momenta .
in addition , quasiparticle excitations are fully gapped when @xmath267 , and thus , the goldstone mode is not damped . therefore , the pole contribution to @xmath455 comes from the goldstone mode for all couplings .
in addition , there is also a branch cut representing the continuum of two particle scattering states , but the contribution from the goldstone mode dominates at sufficiently low temperatures .
it is also illustrative to analyze the eigenvectors of @xmath456 in the amplitude - phase representation corresponding to small @xmath457 mode @xmath458 = \left [ \begin{array}{c } -i \frac{b_{\ell , m_\ell , m_\ell}}{a_{\ell , m_\ell , m_\ell } } w_{\ell , m_\ell}(\mathbf{q } ) \\ 1 \end{array } \right].\ ] ] notice that , when @xmath459 the amplitude and phase modes are not mixed .
next , we discuss the dispersion of collective modes in the weak and strong coupling limits , where the expansion coefficients are analytically tractable for a fixed ( @xmath460 ) state .
the @xmath1-wave @xmath325 weak coupling limit is characterized by the criteria @xmath258 and @xmath461 .
the expansion of the matrix elements to order @xmath462 and @xmath463 is performed under the condition @xmath464
|\delta_{0,0}|$ ] .
analytic calculations are particularly simple in this case since all integrals for the coefficients needed to calculate the collective mode dispersions are peaked near the fermi surface .
we first introduce a shell about the fermi energy @xmath465 such that @xmath466 , inside of which one may ignore the 3d density of states factor @xmath199 and outside of which one may ignore @xmath467 .
in addition , we make use of the nearly perfect particle - hole symmetry , which forces integrals to vanish when their integrands are odd under the transformation @xmath468 . for instance , the coefficient that couple phase and amplitude modes vanish ( @xmath469 ) in this limit .
thus , there is no mixing between phase and amplitude fields in weak coupling , as can be seen by inspection of the fluctuation matrix @xmath470 . for @xmath5 ,
the zeroth order coefficient is @xmath471 and the second order coefficients are @xmath472 here , @xmath473 is the fermi velocity and @xmath474 is the density of states per spin at the fermi energy . in weak coupling , since @xmath475 the sound velocity is simplified to @xmath476 $ ] for any @xmath11 . using the coefficients above in eq .
( [ eqn : soundv.0 ] ) , for @xmath5 , we obtain @xmath477 which is the well known anderson - bogoliubov relation . for @xmath6 ,
the expansion coefficients require more detailed and lengthy analysis , and therefore , we do not discuss here . on the other hand ,
the expansion coefficients can be calculated for any @xmath11 in the strong coupling bec regime , which is discussed next .
the strong coupling limit is characterized by the criteria @xmath208 , @xmath478 and @xmath479 .
the expansion of the matrix elements to order @xmath462 and @xmath463 is performed under the condition @xmath480
\ll |\mu_\ell|$ ] .
the situation encountered here is very different from the weak coupling limit , because one can no longer invoke particle - hole symmetry to simplify the calculation of many of the coefficients appearing in the fluctuation matrix @xmath481 .
in particular , the coefficient @xmath482 indicates that the amplitude and phase fields are mixed .
furthermore , @xmath483 , since this coefficient reduces to the order parameter equation in this limit . for @xmath5 , the zeroth order coefficient is @xmath484 the first order coefficient is @xmath485 and the second order coefficients are @xmath486 where @xmath487 using the expressions above in eq .
( [ eqn : soundv.0 ] ) , we obtain the sound velocity @xmath488 notice that the sound velocity is very small and its smallness is controlled by the scattering length @xmath20 .
furthermore , in the theory of weakly interacting dilute bose gas , the sound velocity is given by @xmath489 making the identification that the density of pairs is @xmath490 , the mass of the pairs is @xmath491 and that the bose scattering length is @xmath492 , it follows that eq .
( [ eqn : sound.0 ] ) is identical to the bogoliubov result @xmath493 .
therefore , our result for the fermionic system represents in fact a weakly interacting bose gas in the strong coupling limit . a better estimate for @xmath401 can be found in the literature @xcite .
this is also the case when we construct the tdgl equation in sec .
[ sec : tdgl.sc ] . for @xmath6 ,
the zeroth order coefficient is @xmath494 and the second order coefficients are @xmath495 where @xmath496 @xmath497 @xmath498 and @xmath499 here @xmath157 is the gamma function , and @xmath500 is an angular averaged quantity defined in app .
[ sec : appb ] . in strong coupling , since @xmath501 the sound velocity is simplified to @xmath502^{1/2 } $ ] for any @xmath11 . using the expressions above in eq .
( [ eqn : soundv.l ] ) , for @xmath6 , we obtain @xmath503 therefore , the sound velocity is also very small and its smallness is controlled by the interaction range @xmath23 through the diluteness condition i.e. @xmath504 , for @xmath6 .
notice that , the sound velocity is independent of the scattering parameter for @xmath6 .
now , we turn our attention to a numerical analysis of the phase collective modes during the evolution from weak coupling bcs to strong coupling bec limits .
we focus only on @xmath1-wave ( @xmath221 ) and @xmath7-wave ( @xmath224 ) cases , since they may be the most relevant to current experiments involving ultracold atoms . in fig .
[ fig : swave.collective ] , we show the evolution of @xmath452 as a function of @xmath235 for @xmath1-wave case .
the weak coupling anderson - bogoliubov velocity @xmath505 evolves continuously to the strong coupling bogoliubov velocity @xmath506 notice that the sound velocity is a monotonically decreasing function of @xmath235 , and the evolution across @xmath237 is a crossover .
[ htb ] in fig .
[ fig : pwave.collective ] , we show the evolution of @xmath507 as a function of @xmath242 for @xmath7-wave case .
notice that @xmath508 is strongly anisotropic in weak coupling , since @xmath509 and @xmath510 , thus reflecting the order parameter symmetry .
in addition , @xmath508 is isotropic in strong coupling , since @xmath511 for @xmath227 , thus revealing the secondary role of the order parameter symmetry in this limit .
the anisotropy is very small in the intermediate regime beyond @xmath250 .
notice also that , @xmath512 is a monotonically decreasing function of @xmath242 in bcs side until @xmath252 , where it saturates .
however , @xmath513 is a nonmonotonic function of @xmath514 , and it also saturates beyond @xmath252 .
therefore , the behavior of @xmath508 reflects the disapperance of nodes of the quasiparticle energy @xmath248 as @xmath241 changes sign .
[ htb ] these collective excitations may contribute significantly to the thermodynamic potential , which is discussed next .
in this section , we analyze corrections to the saddle point thermodynamic potential @xmath515 due to low energy collective excitations .
the evaluation of bosonic matsubara frequency sums in eq .
( [ eqn : flucttp ] ) leads to @xmath516 where @xmath517 \big\rbrace \label{eqn : omega.coll}\ ] ] is the collective mode contribution to the thermodynamic potential .
the prime on the summation indicates that a momentum cutoff is required since a long wavelength and low frequency approximation is used to derive the collective mode dispersion .
notice that the first term in eq .
( [ eqn : omega.coll ] ) contributes to the ground state energy of the interacting fermi system .
this contribution is necessary to recover the ground state energy of the effective bose system in the strong coupling limit as discussed in sec [ sec : gse ] .
the corrections to the saddle point number equation @xmath518 are due to the zero point motion ( @xmath519 and thermal excitation ( @xmath520 ) of the collective modes @xmath521 .
\label{eqn : n.coll}\end{aligned}\ ] ] here @xmath522 $ ] is the bose distribution . for @xmath5 ,
the last equation can be solved to obtain @xmath523 which vanishes at @xmath0 . here
@xmath524 is the zeta function .
similarly , @xmath525 has a power law dependence on @xmath134 , and therefore , vanishes at @xmath0 since the collective modes are not excited .
@xmath526 gives small contributions to the number equation in weak and strong couplings , but may lead to significant contributions in the intermediate regime for all @xmath11 .
the impact of @xmath526 on the order parameter and chemical potential in the intermediate regime may require a careful analysis of the full fluctuation contributions @xcite .
until now , we discussed the evolution of superfluidity from the bcs to the bec regime at zero temperature . in the rest of the manuscript
, we analyze the evolution of superfluidity from the bcs to the bec limit at finite temperatures .
in this section , we concentrate on physical properties near critical temperatures @xmath527 . to calculate @xmath528 , the self - consistency ( order parameter and number ) equations have to be solved simultaneously . at @xmath527 , then @xmath529 , and the saddle point order parameter equation eq .
( [ eqn : opeqn ] ) reduces to @xmath530 } { 2\xi_\ell(\mathbf{k } ) } \right ] .
\label{eqn : tceqn}\ ] ] this expression is independent of @xmath119 since the interaction amplitude @xmath153 depends only on @xmath11 .
similarly , the saddle point number equation reduces to @xmath531 , \label{eqn : tcnsp}\ ] ] where @xmath532 $ ] is the fermi distribution . notice that the summation over spins ( @xmath1 ) is not present in the shs case .
it is important to emphasize that the inclusion of @xmath533 around @xmath534 is essential to produce the qualitatively correct physics with increasing coupling , as discussed next .
to evaluate the gaussian contribution to the thermodynamic potential , we sum over the fermionic matsubara frequencies in eq .
( [ eqn : flucttp ] ) , and obtain the action @xmath535 where @xmath536 is the element of the fluctuation propagator given by @xmath537 this is the generalization of the @xmath5 case to @xmath6 , where @xmath538 . from @xmath539
, we obtain the thermodynamic potential @xmath129 , where @xmath515 is the saddle point contribution with @xmath540 , and @xmath541 \label{eqn : omega1}\ ] ] is the fluctuation contribution .
we evaluate the bosonic matsubara frequency ( @xmath128 ) sums by using contour integration , and determine the branch cut and pole terms .
we use the phase shift @xmath542\ ] ] to replace @xmath543 in eq .
( [ eqn : omega1 ] ) , leading to @xmath544 where @xmath545 and @xmath546 $ ] is the bose distribution .
notice that , this equation is the generalization of the @xmath1-wave ( @xmath5 ) case @xcite for @xmath6 .
furthermore , the phase shift can be written as @xmath547 where @xmath548 is the branch cut ( scattering ) and @xmath549 is the pole ( bound state ) contribution . here , @xmath550 is the heaviside function , @xmath551 with @xmath552 is the branch frequency and @xmath24 is the fermionic chemical potential .
the branch cut ( scattering ) contribution to the thermodynamic potential becomes @xmath553 for each @xmath554 , the integrand is nonvanishing only for @xmath555 since @xmath556 otherwise .
thus , the branch cut ( scattering ) contribution to the number equation @xmath557 is given by @xmath558 \nonumber \\ & & \widetilde{\varphi}_{\ell}(\mathbf{q } , \widetilde{w } ) , \label{eqn : sc}\end{aligned}\ ] ] where @xmath559 .
when @xmath560 , there are no bound states above @xmath561 and @xmath562 represents the correction due to scattering states .
however , when @xmath18 , there are bound states represented by poles at @xmath563 .
the pole ( bound state ) contribution to the number equation is @xmath564 \eta_{\ell}[\mathbf{q},{\cal w}_\ell(\mathbf{q})],\ ] ] where @xmath565 corresponds to the poles of @xmath566 and @xmath567 = { \rm res } \big\lbrace \frac{\partial \det l^{-1}_{\ell}[\mathbf{q},{\cal w}_\ell(\mathbf{q})]/\partial \mu_\ell } { \det l^{-1}_{\ell}[\mathbf{q},{\cal w}_\ell(\mathbf{q } ) ] } \big \rbrace \label{eqn : residue}\ ] ] is the residue .
heavy numerical calculations are necessary to find the poles as a function of @xmath554 for all couplings .
however , in sufficiently strong coupling , when @xmath568 in eq .
( [ eqn : l-1 ] ) , the pole ( bound state ) contribution can be evaluated analytically by eliminating @xmath153 in favor of the two body bound state energy @xmath569 in vacuum .
notice that , @xmath569 is related to the @xmath19 obtained from the t - matrix approach , where multiple scattering events are included
. however , they become identical in the dilute limit .
a relation between @xmath153 and @xmath569 can be obtained by solving the schroedinger equation for two fermions interacting via a pairing potential @xmath61 .
after fourier transforming from real to momentum space , the schroedinger equation for the pair wave function @xmath570 is @xmath571 using the fourier expansion of @xmath47 given in eq .
( [ eqn : vsh ] ) and choosing only the @xmath11th angular momentum channel , we obtain @xmath572 this expression relates @xmath573 to @xmath153 in order to express eq .
( [ eqn : bs ] ) in terms of binding energy @xmath573 .
notice that , this equation is similar to the order parameter equation in the strong coupling limit ( @xmath208 and @xmath574 ) , where @xmath575 therefore , @xmath576 as the coupling increases .
substitution of eq .
( [ eqn : binding ] ) in eq .
( [ eqn : residue ] ) yields the pole contribution which is given by @xmath577 , and the residue at this pole is @xmath578 = - 2 \sum_{m_\ell}$ ] .
therefore , the bound state contribution to the phase shift in the sufficiently strong coupling limit is given by @xmath579 which leads to the bound state number equation @xmath580 , \label{eqn : bs}\ ] ] where @xmath581 is the chemical potential of the bosonic molecules .
notice that , eq . ( [ eqn : bs ] ) is only valid for interaction strengths where @xmath582 .
thus , this expression can not be used over a region of coupling strengths where @xmath395 is positive . to obtain the evolution from bcs to bec , the number @xmath583 and order parameter [ eq . ( [ eqn : tceqn ] ) ]
equations have to be solved self - consistently for @xmath26 and @xmath24 .
first , we analyze the number of unbound , scattering and bound fermions as a function of the scattering parameter for the @xmath1-wave ( @xmath5 ) and @xmath7-wave ( @xmath8 ) cases . in fig .
[ fig : swave.tc.n ] , we plot different contributions to the number equation as a function of @xmath230 for the @xmath1-wave @xmath325 case .
notice that , @xmath584 ( @xmath585 ) dominates in weak ( strong ) coupling , while @xmath586 is the highest for intermediate couplings .
thus , all fermions are unbound in the strictly bcs limit ( not shown in the figure ) , while all fermions are bound in the strictly bec limit .
[ htb ] in fig .
[ fig : pwave.tc.n ] , we present plots of different contributions to the number equation as a function of @xmath242 for the @xmath7-wave @xmath329 case .
notice also that , @xmath587 ( @xmath588 ) dominates in weak ( strong ) coupling , while @xmath589 is the highest for intermediate couplings .
thus , again all fermions are unbound in the strictly bcs limit , while all fermions are bound in the strictly bec limit .
[ htb ] therefore , the total fluctuation contribution @xmath590 is negligible in weak coupling and @xmath338 is sufficient .
however , the inclusion of fluctuations is necessary for strong coupling to recover the physics of bec .
however , in the vicinity of the unitary limit [ @xmath591 , our results are not strictly applicable and should be regarded as qualitative .
next , we discuss the chemical potential and critical temperature . in weak coupling ,
we introduce a shell about the fermi energy @xmath197 , such that @xmath592 .
then , in eq .
( [ eqn : tceqn ] ) , we set @xmath593 = 1 $ ] outside the shell and treat the integration within the shell as usual in the bcs theory .
in strong coupling , we use that @xmath594 = |\mu_\ell| \gg t_{{\rm c},\ell}$ ] and set @xmath595 = 1 $ ] .
therefore , in strictly weak and strong coupling , the self - consistency equations are decoupled , and play reversed roles . in weak ( strong )
coupling the order parameter equation determines @xmath26 ( @xmath24 ) and the number equation determines @xmath24 ( @xmath26 ) . in weak coupling , the number equation @xmath596 leads to @xmath597 for any @xmath11 . in strong coupling
, the order parameter equation leads to @xmath598 where @xmath205 and @xmath157 is the gamma function .
this calculation requires @xmath599 for @xmath5 , and @xmath207 for @xmath6 for the order parameter equation to have a solution with @xmath208 . furthermore , we assume @xmath478 to obtain eqs .
( [ eqn : mu0.0 ] ) and ( [ eqn : mutc.0 ] ) .
notice that , @xmath214 in this limit .
on the other hand , the solution of the order parameter equation in weak coupling is @xmath600 , \\
t_{{\rm c},\ell } & \sim & \epsilon_{\rm f } \exp\left[t_\ell \left(\frac{k_0}{k_{\rm f}}\right)^{2\ell -1 } - \frac{\pi}{2k_{\rm f}^{2\ell+1}|a_\ell|}\right ] , \label{eqn : tc.weak}\end{aligned}\ ] ] where @xmath601 is the euler s constant , @xmath216 and @xmath217 .
these expressions are valid only when the exponential terms are small . in strong coupling
, the number equation @xmath602 leads to @xmath603^{\frac{2}{3 } } = \frac{0.218}{(\sum_{m_\ell})^{\frac{2}{3}}}\epsilon_{\rm f},\ ] ] where @xmath604 is the mass of the bosonic molecules . here
, @xmath605 is the density of fermions . for ths fermi gases ,
we conclude that the bec critical temperature of @xmath1-wave superfluids is the highest , and this temperature is reduced for higher angular momentum states . however , for shs fermi gases @xmath606^{\frac{2}{3 } } = \frac{0.137\epsilon_{\rm f}}{(\sum_{m_\ell})^\frac{2}{3}}.\ ] ] where @xmath607 and @xmath524 is the zeta function . here , the summation over @xmath119 is over degenerate spherical harmonics involved in the order parameter of the system , and can be at most @xmath608 . for shs states ,
we conclude that the bec critical temperature of @xmath7-wave superfluids is the highest , and this temperature is reduced for higher angular momentum states . for completeness , it is also possible to relate @xmath4 and @xmath26 when chemical potential vanishes . when @xmath336 , the solution of number equation eq .
( [ eqn : numbertc ] ) is highly nontrivial and it is difficult to find the value of the scattering parameter @xmath609 at @xmath336 .
however , the critical temperature in terms of @xmath609 can be found analytically from eq .
( [ eqn : tceqn ] ) as @xmath610 to order @xmath611 , where @xmath612 .
notice that , this relation depends on @xmath23 only through @xmath609 . on the other hand , if temporal fluctuations are neglected , the solution for @xmath613 from the saddle point self - consistency equations is @xmath614 and @xmath615 which leads to @xmath616 up to logarithmic accuracy .
therefore , @xmath613 grows without bound as the coupling increases . within this calculation , the normal state for @xmath617
is described by unbound and nondegenerate fermions since @xmath540 and @xmath618 .
notice that the saddle point approximation becomes progressively worse with increasing coupling , since the formation of bound states is neglected .
we emphasize that , there is no phase transition across @xmath613 in strong coupling .
however , this temperature is related to the pair breaking or dissociation energy scale . to see this connection
, we consider the chemical equilibrium between nondegenerate unbound fermions ( f ) and bound pairs ( b ) such that @xmath619 for ths singlet states and @xmath620 for shs triplet states .
notice that @xmath613 is sufficiently high that the chemical potential of the bosons and the fermions satisfy @xmath621 and @xmath622 at the temperature @xmath134 of interest .
thus , both the unbound fermions ( f ) and molecules ( b ) can be treated as classical ideal gases .
the equilibrium condition @xmath623 for these non - degenerate gases may be written as @xmath624^{3/2 } \big\rbrace - \widetilde{e}_{{\rm b},\ell } = 2 t \ln \big\lbrace n_{\rm f } [ 2\pi/(m_{\rm f }
t)]^{3/2 } \big\rbrace , $ ] where @xmath625 ( @xmath626 ) is the boson ( fermion ) density , @xmath627 ( @xmath628 ) is the boson ( fermion ) mass , and @xmath569 is the binding energy of a bosonic molecule .
the dissociation temperature above which some fraction of the bound pairs ( molecules ) are dissociated , is then found to be @xmath629 where we dropped a few constants of order unity .
therefore , the logarithmic term is an entropic contribution which favors broken pairs and leads to a dissociation temperature considerably lower than the absolute value of binding energy @xmath630 .
the analysis above gives insight into the logarithmic factor appearing in eq .
( [ eqn : t0 ] ) since @xmath631 .
thus , @xmath613 is essentially the pair dissociation temperature of bound pairs ( molecules ) , while @xmath26 is the phase coherence temperature corresponding to bec of bound pairs ( bosonic molecules ) .
[ htb ] in fig .
[ fig : swave.tc ] , we show @xmath632 for the @xmath1-wave ( @xmath221 ) case .
notice that @xmath632 grows from an exponential dependence in weak coupling to a constant in strong coupling with increasing interaction .
furthermore , the mean field @xmath633 and gaussian @xmath632 are similar only in weak coupling , while @xmath633 increases without bound as @xmath634 $ ] in strong coupling . when @xmath237 , we also obtain analytically @xmath635 from eq .
( [ eqn : tcc ] ) .
the hump in @xmath632 around @xmath636 is similar to the those in ref .
@xcite , and might be an artifact of the approximations used here . thus , a more detailed self - consistent numerical analysis is needed to determine if this hump is real .
[ htb ] in fig .
[ fig : swave.mutc ] , we show @xmath229 for the @xmath1-wave case , where it changes from @xmath232 in weak coupling to @xmath637 in strong coupling .
notice that , @xmath229 at @xmath632 is qualitatively similar to @xmath229 at @xmath0 , however , it is reduced at @xmath632 in weak coupling .
furthermore , @xmath229 changes sign at @xmath638 .
[ htb ] in fig .
[ fig : pwave.tc ] , we show @xmath639 for the @xmath7-wave ( @xmath224 ) case .
@xmath639 grows from an exponential dependence in weak coupling to a constant in strong coupling with increasing interaction . for completeness , we present the limiting expressions @xmath640 , \\
t_{{\rm c},1 } & = & \frac{2\pi}{m_{{\rm b } , 1 } } \left[\frac{n_1}{\zeta(3/2)}\right]^\frac{2}{3 } = 0.137\epsilon_{\rm f},\end{aligned}\ ] ] in the weak and strong coupling limits , respectively .
furthermore , the mean field @xmath641 and gaussian @xmath639 are similar only in weak coupling , while @xmath641 increases without bound as @xmath642 $ ] in strong coupling .
when @xmath252 , we also obtain analytically @xmath643 from eq .
( [ eqn : tcc ] ) .
the hump in the intermediate regime is similar to the one found in fermion - boson model @xcite . but to determine if this hump is real , it may be necessary to develop a fully self - consistent numerical calculation .
[ htb ] in fig .
[ fig : pwave.mutc ] , we show @xmath241 for the @xmath7-wave case , where it changes from @xmath232 in weak coupling to @xmath644 in strong coupling .
notice that , @xmath241 at @xmath639 is both qualitatively and quantitatively similar to @xmath241 at @xmath0 .
furthermore , @xmath241 changes sign at @xmath645 . for any given @xmath11 ,
mean field and gaussian theories lead to similar results for @xmath26 and @xmath613 in the bcs regime , while they are very different in the bec side . in the latter case
, @xmath613 increases without bound , however , gaussian theory results in a constant critical temperature which coincides with the bec temperature of bosons .
notice that the pseudogap region @xmath646 for @xmath5 state is much larger than @xmath6 states since @xmath647 grows faster than @xmath648 .
furthermore , similar humps in @xmath26 around @xmath147 are expected for any @xmath11 as shown for the @xmath1-wave and @xmath7-wave cases , however , whether these humps are physical or not may require a fully self - consistent numerical approach . as shown in this section ,
the frequency ( temporal ) dependence of fluctuations about the saddle point is crucial to describe adequately the bosonic degrees of freedom that emerge with increasing coupling . in the next section
, we derive the tdgl functional near @xmath26 to emphasize further the importance of these fluctuations .
our basic motivation here is to investigate the low frequency and long wavelength behavior of the order parameter near @xmath26 . to study the evolution of the time - dependent ginzburg - landau ( tdgl ) functional near @xmath26
, we need to expand the effective action @xmath650 in eq .
( [ eqn : s_eff ] ) around @xmath529 leading to @xmath651 here , @xmath107 is the pairing fluctuation field .
we first consider the static part of @xmath652 , and expand it in powers of @xmath653 to get @xmath654 next , we consider the time - dependence of the tdgl equation , where it is necessary to expand @xmath655 in powers of @xmath444 after analytic continuation @xmath415 . in the @xmath656 representation , the calculation above leads to the tdgl equation @xmath657\lambda_{\ell , m_{\ell_2}}(x ) = 0,\end{aligned}\ ] ] which is the generalization of the tdgl equation to higher momentum channels of ths singlet and shs triplet states .
notice that , for ths triplet states , there may be extra gradient mixing textures and fourth order terms in the expansion @xcite , which are not discussed here .
all static and dynamic expansion coefficients are presented in appendix [ sec : appb ] .
the condition @xmath658 with matrix elements @xmath659 is the thouless criterion , which leads to the order parameter equation .
the coefficient @xmath660 reflects a major difference between @xmath5 and @xmath6 cases . while @xmath661 is isotropic in space
, @xmath662 is anisotropic , thus characterizing the anisotropy of the order parameter .
the coefficient @xmath663 is positive and guarantees the stability of the theory .
the coefficient @xmath664 is a complex number .
its imaginary part reflects the decay of cooper pairs into the two - particle continuum for @xmath27 .
however , for @xmath208 , imaginary part of @xmath664 vanishes and the behavior of the order parameter is propagating reflecting the presence of stable bound states .
next , we present the asymptotic forms of @xmath665 ; @xmath663 ; @xmath660 and @xmath666 which are used to recover the usual ginzburg - landau ( gl ) equation for bcs superfluids in weak coupling and the gross - pitaevskii ( gp ) equation for a weakly interacting dilute bose gas in strong coupling
. the weak coupling bcs regime is characterized by @xmath27 and @xmath667 . for any given @xmath11 , we find the following values for the coefficients @xmath668 where @xmath669 with @xmath670 is the density of states per spin at the fermi energy . here
@xmath671 is the kronecker delta , and @xmath672 and @xmath673 are angular averaged quantities defined in app .
[ sec : appb ] .
notice that the critical transition temperature is determined by @xmath658 . in the particular case ,
where only one of the spherical harmonics @xmath52 is dominant and characterizes the order parameter , we can rescale the pairing field as @xmath674 to obtain the conventional tdgl equation @xmath675 \psi^{\rm w}_{\ell , m_\ell } = 0.\ ] ] here , @xmath676 with @xmath677 , @xmath678 = ( \xi^{\rm{gl}}_{\ell , m_\ell})_i^2 / \varepsilon_\ell $ ] is the characteristic gl length and @xmath679 is the characteristic gl time . in this limit , the gl coherence length is given by @xmath680^{1/2 } ( \epsilon_{\rm f } / t_{{\rm c},\ell } ) , $ ] which makes @xmath681 much larger than the interparticle spacing @xmath682 .
there is a major difference between @xmath5 and @xmath6 pairings regarding @xmath681 . while @xmath683 is isotropic
, @xmath684 is in general anisotropic in space ( see app . [
sec : appb ] ) .
thus , @xmath685 is isotropic and @xmath686 is not .
furthermore , @xmath687 showing that the dynamics of @xmath688 is overdamped reflecting the continuum of fermionic excitations into which a pair can decay .
in addition , there is a small propagating term since there is no perfect particle - hole symmetry . as the coupling grows , the coefficient of the propagating term increases while that of the damping term vanishes for @xmath28 .
thus , the mode is propagating in strong coupling reflecting the stability of the bound states against the two particle continuum .
the strong coupling bec regime is characterized by @xmath208 and @xmath689 . for @xmath5 , we find the following coefficients @xmath690 where @xmath691 similarly , for @xmath6 , we obtain @xmath692 where @xmath693 here @xmath205 and @xmath694 where @xmath157 is the gamma function . notice that , @xmath695 is isotropic in space for any @xmath11 .
thus , the anisotropy of the order parameter plays a secondary role in the tdgl theory in this limit . in the particular case , where only one of the spherical harmonics
@xmath52 is dominant and characterizes the order parameter , we can rescale the pairing field as @xmath696 to obtain the conventional gross - pitaevskii ( gp ) equation @xmath697 \psi^{\rm s}_{\ell , m_\ell } = 0\ ] ] for a dilute gas of bosons . here , @xmath698 is the chemical potential , @xmath699 is the mass , and @xmath700 is the repulsive interactions of the bosons .
we obtain , @xmath701 and @xmath702 for @xmath5 and @xmath6 , respectively . notice that the mass of the composite bosons is independent of the anisotropy and symmetry of the order parameter for
any given @xmath11 .
however , this is not the case for the repulsive interactions between bosons , which explicitly depends on @xmath11 . for @xmath5
, @xmath703 is directly proportional to the fermion ( boson ) scattering length @xmath20 ( @xmath704 ) , where @xmath705 is the boson - boson scattering lenth .
a better estimate for @xmath401 can be found in the literature @xcite . while for @xmath6 , @xmath706 is a constant ( independent of the scattering parameter @xmath4 ) depending only on the interaction range @xmath23 and the particular @xmath707 state . for a finite range potential ,
@xmath708 is small compared to @xmath232 , where @xmath709 is the density of bosons . in the @xmath5 case
@xmath710 is much smaller than unity . for @xmath6 and even , @xmath711 . in the case of shs states
where @xmath6 and odd , @xmath712 .
the results for higher angular momentum channels reflect the diluteness condition @xmath713 .
to calculate @xmath681 in the strong coupling limit , we need to know @xmath714 evaluated at @xmath26 ( see below ) .
the temperature dependence of @xmath24 in the vicinity of @xmath26 can be obtained by noticing that @xmath715 where @xmath716 .
$ ] this leads to @xmath717^{1/2 } $ ] in the bec regime . using the asymptotic values of @xmath706
, we obtain @xmath718^{1/2 } $ ] for @xmath5 and @xmath719^{1/2}(k_0/k_{\rm f})^{1/2 } $ ] for @xmath6 . therefore , @xmath681 is also much larger than the interparticle spacing @xmath682 in this limit , since @xmath720 for @xmath5 and @xmath721 for any @xmath11 . in the particular case ,
where only one of the spherical harmonics @xmath722 is dominant and characterizes the order parameter , we can define the gl coherence length as @xmath723 an expansion of the parameters @xmath724 and @xmath725 in the vicinity of @xmath26 leads to @xmath726 where the prefactor is the gl coherence length and given by @xmath727_{t = t_{{\rm c},\ell}}^{-1}.\ ] ] the slope of the coefficient @xmath724 with respect to @xmath134 is given by @xmath728 \nonumber \\ & & \hskip 1 cm \frac{\gamma_\ell^2(k)}{8\pi}.\end{aligned}\ ] ] here @xmath729 and @xmath730 are defined in app .
[ sec : appb ] .
notice that , while @xmath731 vanishes at @xmath649 in weak coupling , it plays an important role in strong coupling .
furthermore , while @xmath732 representing the phase coherence length is large compared to interparticle spacing in both bcs and bec limits , it should have a minimum near @xmath733 . the prefactor @xmath681 of the gl coherence length
must be compared with the average cooper pair size @xmath734 defined by @xmath735 where @xmath736 $ ] is the zero temperature pair wave function . in the bcs limit
, @xmath737 is much larger than the interparticle distance @xmath682 since the cooper pairs are weakly bound .
furthermore , for @xmath208 , we expect that @xmath737 is a decreasing function of interaction for any @xmath11 , since cooper pairs become more tightly bound as the interaction increases .
next , we compare @xmath738 and @xmath739 for @xmath1-wave ( @xmath5 ) and @xmath7-wave ( @xmath8 ) states .
[ htb ] in fig .
[ fig : swave.pairsize ] , a comparison between @xmath740 and @xmath741 is shown for @xmath1-wave ( @xmath221 ) .
@xmath742 changes from @xmath743 ( \epsilon_{\rm f}/t_{c,0 } ) $ ] in the bcs limit to @xmath744^{1/2 } = k_{\rm f}a_0/\sqrt{2 } $ ] in the bec limit as the interaction increases . here
@xmath601 is the euler s constant .
furthermore , when @xmath237 , we obtain @xmath745^{1/3}/4 \approx 1.29 , $ ] where @xmath157 is the gamma function .
notice that , @xmath742 is continuous at @xmath237 , and monotonically decreasing function of @xmath235 with a limiting value controlled by @xmath20 in strong coupling .
however , @xmath746 is a non - monotonic function of @xmath235 having a minimum around @xmath638 ( @xmath237 ) .
it changes from @xmath747^{1/2 } ( \epsilon_{\rm f}/t_{{\rm c},0 } ) $ ] in the bcs to @xmath748^{1/2 } $ ] in the bec limit as the coupling increases , where @xmath524 is the zeta function .
notice that , @xmath746 grows as @xmath749 in strong coupling limit .
[ htb ] in fig .
[ fig : pwave.pairsize ] , a comparison between @xmath750 and @xmath751 is shown for @xmath7-wave ( @xmath224 ) .
notice that , @xmath752 is nonanalytic at @xmath252 , and is a monotonically decreasing function of @xmath242 with a limiting value controlled by @xmath753 in strong coupling .
this nonanalytic behavior is associated with the change in @xmath248 from gapless ( with line nodes ) in the bcs to fully gapped in the bec side .
however , @xmath754 is a non - monotonic function of @xmath242 having a minimum around @xmath645 ( @xmath252 ) .
it changes from @xmath755^{1/2 } ( \epsilon_{\rm f}/t_{{\rm c},1 } ) $ ] in the bcs to @xmath756^{1/2 } $ ] in the bec limit as the coupling increases .
notice that , @xmath757 saturates in strong coupling limit reflecting the finite range of interactions .
it is important to emphasize that @xmath758 shown in figs .
( [ fig : swave.pairsize ] ) and ( [ fig : pwave.pairsize ] ) is only qualitative in the intermediate regime around unitarity @xmath147 since our theory is not strictly applicable in that region .
in this manuscript , we extended the @xmath1-wave ( @xmath5 ) functional integral formalism to finite angular momentum @xmath11 including two hyperfine states ( ths ) pseudo - spin singlet and single hyperfine states ( shs ) pseudo - spin triplet channels .
we analyzed analytically superfluid properties of a dilute fermi gas in the ground state ( @xmath0 ) and near critical temperatures ( @xmath25 ) from weak coupling ( bcs ) to strong coupling ( bec ) as a function of scattering parameter ( @xmath4 ) for arbitrary @xmath11 .
however , we presented numerical results only for ths @xmath1-wave and shs @xmath7-wave symmetries which may be relevant for current experiments involving atomic fermi gases .
the main results of our paper are as follows .
first , we analyzed the low energy scattering amplitude within a t - matrix approach .
we found that bound states occur only when @xmath18 for any @xmath11 .
the energy of the bound states @xmath19 involves only the scattering parameter @xmath20 for @xmath5 .
however , another parameter related to the interaction range @xmath22 is necessary to characterize @xmath19 for @xmath6 .
therefore , all superfluid properties for @xmath6 depend strongly on @xmath23 and @xmath4 , while for @xmath5 they depend strongly only on @xmath20 but weakly on @xmath23 .
second , we discussed the order parameter , chemical potential , quasiparticle excitations , momentum distribution , atomic compressibility , ground state energy , collective modes and average cooper pair size at @xmath0 .
there we showed that the evolution from bcs to bec is just a crossover for @xmath5 , while the same evolution for @xmath6 exhibits a quantum phase transition characterized by a gapless superfluid on the bcs side to a fully gapped superfluid on the bec side .
this transition is a many body effect and takes place exactly when chemical potential @xmath270 crosses the bottom of the fermion band ( @xmath269 ) , and is best reflected as non - analytic behavior in the ground state atomic compressibility , momentum distribution and average cooper pair size .
third , we discussed the critical temperature , chemical potential , and the number of unbound , scattering and bound fermions at @xmath527 .
we found that the critical bec temperature is the highest for @xmath5 .
we also derived the time - dependent ginzburg - landau functional ( tdgl ) near @xmath26 and extracted the ginzburg - landau ( gl ) coherence length and time .
we recovered the usual tdgl equation for bcs superfluids in the weak coupling limit , whereas in the strong coupling limit we recovered the gross - pitaevskii ( gp ) equation for a weakly interacting dilute bose gas .
the tdgl equation exhibits anisotropic coherence lengths for @xmath6 which become isotropic only in the bec limit , in sharp contrast to the @xmath5 case , where the coherence length is isotropic for all couplings .
furthermore , the gl time is a complex number with a larger imaginary component for @xmath27 reflecting the decay of cooper pairs into the two particle continuum .
however , for @xmath208 the imaginary component vanishes and cooper pairs become stable above @xmath26 . in summary ,
the bcs to bec evolution in higher angular momentum ( @xmath6 ) states exhibit quantum phase transitions and is much richer than in conventional @xmath5 @xmath1-wave systems , where there is only a crossover .
these @xmath6 states might be found not only in atomic fermi gases , but also in nuclear ( pairing in nuclei ) , astrophysics ( neutron stars ) and condensed matter ( high-@xmath759 and organic superconductors ) systems .
we would like to thank nsf ( dmr-0304380 ) for financial support .
from the rotated fluctuation matrix @xmath760 expressed in the amplitude - phase basis , we can obtain the expansion coefficients necessary to calculate the collective modes at @xmath0 . in the long wavelength @xmath761 , and low frequency limit @xmath762 the condition @xmath763
while there is no landau damping and a well defined expansion is possible for @xmath5 case for all couplings , extra care is necessary for @xmath6 when @xmath27 since landau damping is present .
the coefficients necessary to obtain the matrix element @xmath769 are @xmath770 corresponding to the @xmath771 term , @xmath772 \nonumber \\ & + & 3\ddot{\delta}_\ell^{i , j } e_\ell^2(\mathbf{k } ) \xi_\ell(\mathbf{k } ) \delta_\ell(\mathbf{k } ) + 5\dot{\xi}_\ell^i \dot{\xi}_\ell^j \delta_\ell^2(\mathbf{k } ) \xi_\ell^2(\mathbf{k } ) \nonumber \\ & + & \dot{\delta}_\ell^i \dot{\delta}_\ell^j \xi_\ell(\mathbf{k } ) \left[\xi_\ell^2(\mathbf{k } ) - 4\delta_\ell^2(\mathbf{k})\right ] \nonumber \\ & + & ( \dot{\xi}_\ell^i \dot{\delta}_\ell^j + \dot{\xi}_\ell^j \dot{\delta}_\ell^i ) \delta_\ell(\mathbf{k } ) \left[2\delta_\ell^2(\mathbf{k } ) - 3\xi_\ell^2(\mathbf{k})\right ] \big\rbrace \nonumber \\ & & \gamma_\ell^2(k ) y_{\ell , m_\ell}(\widehat{\mathbf{k } } ) y_{\ell , m_\ell'}^*(\widehat{\mathbf{k}}),\end{aligned}\ ] ] corresponding to the @xmath773 term , and @xmath774 corresponding to the @xmath463 term . here
@xmath775 is the kronecker delta .
the coefficients necessary to obtain the matrix element @xmath776 are @xmath777 corresponding to the @xmath771 term , @xmath778 corresponding to the @xmath773 term , and @xmath779 corresponding to the @xmath463 term . the zeroth order coefficient @xmath786 is diagonal in @xmath119 and @xmath787 , and is given by @xmath788,\ ] ] where @xmath789}$ ] .
the second order coefficient @xmath790 evaluated at @xmath768 is given by @xmath791\delta_{m_\ell , m'_\ell } \delta_{i , j } \nonumber \\ & + & \alpha_{\ell , m_\ell , m'_\ell}^{i , j } \frac{\beta^2 k^2 { \cal x}_\ell(\mathbf{k } ) { \cal y}_\ell(\mathbf{k})}{16 m \xi_\ell(\mathbf{k } ) } \big\rbrace\gamma_\ell^2(k),\end{aligned}\ ] ] where @xmath792 $ ] and the angular average @xmath793 here , @xmath794 , @xmath795 , @xmath796 and @xmath797 .
in general , @xmath798 is a fourth order tensor for fixed @xmath11 .
however , in the particular case where only one of the spherical harmonics @xmath52 is dominant and characterizes the order parameter , @xmath799 is diagonal in @xmath119 and @xmath800 . in this case
, we use gaunt coefficients @xcite to show that @xmath801 is also diagonal in @xmath802 and @xmath803 leading to @xmath804 . the coefficient of fourth order term is approximated at @xmath805 , and given by @xmath806 \gamma_\ell^4(k),\end{aligned}\ ] ] where the angular average @xmath807 to extract the time - dependence , we expand @xmath808 in powers of @xmath444 after the analytic continuation @xmath415 .
we use the relation @xmath809 , where @xmath304 is the principal value and @xmath810 is the delta function , to obtain @xmath811 \gamma_\ell^2(k ) \right]\end{aligned}\ ] ] keeping only the first order terms in @xmath444 leads to @xmath812 , where @xmath813\end{aligned}\ ] ] is also diagonal in @xmath119 and @xmath787 . here
@xmath670 is the density of states per spin at the fermi energy , @xmath814 is the interaction symmetry in terms of energy and @xmath550 is the heaviside function .
99 c. a. regal , m. greiner , and d. s. jin , phys .
lett . * 92 * , 040403 ( 2004 ) .
m. greiner , c. a. regal , and d. s. jin , nature ( london ) * 426 * , 537 ( 2003 ) .
k. e. strecker , g. b. partridge , and r. g. hulet , phys .
91 * , 080406 ( 2003 ) .
m. w. zwierlein , c. a. stan , c. h. schunck , s. m. f. raupach , s. gupta , z. hadzibabic , and w. ketterle , phys .
lett . * 91 * , 250401 ( 2003 ) .
t. bourdel , l. khaykovich , j. cubizolles , j. zhang , f. chevy , m. teichmann , l. tarruell , s. j. j. m. f. kokkelmans , c. salomon , phys .
lett . * 93 * , 050401 ( 2004 ) .
m. bartenstein , a. altmeyer , s. riedl , s. jochim , c. chin , j. hecker denschlag , and r. grimm , phys .
lett . * 92 * , 120401 ( 2004 ) .
j. kinast , s. l. hemmer , m. e. gehm , a. turlapov and j. e. thomas , phys .
lett . * 92 * , 150402 ( 2004 ) .
m. w. zwierlein , j. r. abo - shaeer , a. schirotzek , c. h. schunck , and w. ketterle , nature * 435 * , 1047 ( 2005 ) .
d. m. eagles , phys . rev . * 186 * , 456 ( 1969 ) .
a. j. leggett , in _ modern trends in the theory of condensed matter _ , edited by a. peralski and r. przystawa ( springer - verlag , berlin 1980 ) .
p. nozieres and s. schmitt - rink , j. low .
* 59 * , 195 ( 1985 ) .
c. a. r. s de melo , m. randeria , and j. r. engelbrecht , phys .
* 71 * , 3202 ( 1993 ) .
j. r. engelbrecht , m. randeria , and c. a. r. s de melo , phys .
b * 55 * , 15153 ( 1997 ) .
a. perali , p. pieri , l. pisani , and g. c. strinati , phys .
lett . * 92 * , 220404 ( 2004 ) .
y. ohashi and a. griffin , phys .
a * 67 * , 033603 ( 2003 ) .
m. holland , s. j. j. m. f. kokkelmans , m. l. chiofalo , and r. walser , phys .
lett . * 87 * , 120406 ( 2001 ) .
e. timmermans , k. furuya , p. w. milonni , and a. k. kerman , phys .
a * 285 * , 228 ( 2001 ) y. ohashi and a. griffin , phys .
. lett . * 89 * , 130402 ( 2002 ) . c. a. regal , c. ticknor , j. l. bohn , and d. s. jin , phys .
lett . * 90 * , 053201 ( 2003 ) .
c. ticknor , c. a. regal , d. s. jin , and j. l. bohn , phys .
a * 69 * , 042712 ( 2004 ) j. zhang , e. g. m. van kempen , t. bourdel , l. khaykovich , j. cubizolles , f. chevy , m. teichmann , l. tarruell , s. j. j. m. f. kokkelmans , and c. salomon , phys .
a * 70 * , 030702 ( 2004 ) . c. h. schunck , m. w. zwierlein , c. a. stan , s. m. f. raupach , w. ketterle , a. simoni , e. tiesinga , c. j. williams , and p. s. julienne , phys
a * 71 * , 045601 ( 2005 ) .
k. gnter , t. stferle , h. moritz , m. khl , and t. esslinger , phys .
95 * , 230401 ( 2005 ) .
j. l. bohn , phys .
a * 61 * , 053409 ( 2000 ) .
v. gurarie , l. radzihovsky , and a. v. andreev , phys .
lett . * 94 * , 230403 ( 2005 ) .
chi - ho cheng and s .- k .
yip , phys .
lett . * 95 * , 070404 ( 2005 ) .
l. s. borkowski and c. a. r. s de melo , cond - mat/9810370 .
r. d. duncan and c. a. r. s de melo , phys .
rev b * 62 * , 9675 ( 2000 ) .
jorge quintanilla , balazs l. gyorffy , james f. annett , and jonathan p. wallington , phys .
b * 66 * , 214526 ( 2002 ) .
b. c. den hertog , phys
b * 60 * , 559 ( 1999 ) .
n. andrenacci , a. perali , p. pieri , and g. c. strinati , phys .
b * 60 * , 12410 ( 1999 ) .
tin - lun ho and roberto b. diener , phys .
* 94 * , 090402 ( 2005 ) .
s. s. botelho and c. a. r. s de melo , cond - mat/0409368 .
m. iskin and c. a. r. s de melo , phys .
b * 72 * , 224513 ( 2005 ) . s. s. botelho and c. a. r. s de melo , j. low temp . phys . *
140 * , 409 ( 2005 ) .
y. ohashi , phys .
* 94 * , 050403 ( 2005 ) .
m. iskin and c.a.r .
s de melo , phys .
lett . * 96 * , 040402 ( 2006 ) .
g. e. volovik , cond - mat/0601372 . l. d. landau and e. m. lifshitz , _ quantum mechanics _ , permagon , oxford ( 1994 ) .
a. j. leggett , rev .
47 , 331 ( 1975 ) .
p. pieri , l. pisani and g. c. strinati , phys .
b. * 70 * , 094508 ( 2004 ) .
m. greiner , c. a. regal , j. t. stewart , and d. s. jin , phys .
94 * , 110401 ( 2005 ) .
e. altman , e. demler , and m. lukin , phys .
a * 70 * , 013603 ( 2004 ) .
f. reif , _ fundamentals of statistical and thermal physics _ , chap . 8 , mcgraw - hill , tokyo ( 1981 )
. g. e. volovik , _ exotic properties of superfluid @xmath815 _ , world scientific , singapore ( 1992 ) .
i. m. lifshitz , sov .
jetp * 11 * , 1130 ( 1960 ) . [ zh . eksp .
. fiz . * 38 * , 1569 ( 1960 ) ] .
a. a. abrikosov , _ fundamental theory of metals _ , pages 111 - 114 , north - holland ( 1988 ) .
p. pieri and g. c. strinati , phys .
b * 61 * , 15370 ( 2000 ) .
d. s. petrov , c. salomon , and g. v. shlyapnikov , phys .
* 93 * , 090404 ( 2004 ) .
i. v. brodsky , a. v. klaptsov , m. yu .
kagan , r. combescot , x. leyronas , cond - mat/0507240 .
j. levinsen , v. gurarie , cond - mat/0510672 . g. e. astrakharchik , j. boronat , j. casulleras , and s. giorgini , phys .
lett . * 93 * , 200404 ( 2004 ) .
g. b. arfken and h. j. weber , _ mathematical methods for physicists _ , harcourt academic press , new york ( 2001 ) . | we analyze the evolution of superfluidity for nonzero orbital angular momentum channels from the bardeen - cooper - schrieffer ( bcs ) to the bose - einstein condensation ( bec ) limit in three dimensions .
first , we analyze the low energy scattering properties of finite range interactions for all possible angular momentum channels .
second , we discuss ground state ( @xmath0 ) superfluid properties including the order parameter , chemical potential , quasiparticle excitation spectrum , momentum distribution , atomic compressibility , ground state energy and low energy collective excitations . we show that a quantum phase transition occurs for nonzero angular momentum pairing , unlike the @xmath1-wave case where the bcs to bec evolution is just a crossover .
third , we present a gaussian fluctuation theory near the critical temperature ( @xmath2 ) , and we analyze the number of bound , scattering and unbound fermions as well as the chemical potential .
finally , we derive the time - dependent ginzburg - landau functional near @xmath3 , and compare the ginzburg - landau coherence length with the zero temperature average cooper pair size . |
dynamical properties of nonrelativistic quantum systems may be described by the schrdinger equation @xcite , in which the time - dependent wavefunction @xmath1 for the one - dimensional system with the potential @xmath2 is described by @xmath3\psi(x , t ) .
\label{eq : a1}\end{aligned}\ ] ] it is generally difficult to obtain exact solutions of the schrdinger equation which are available only for limited cases like a harmonic oscillator ( ho ) system . for general quantum systems ,
various approaches such as perturbation and spectral methods have been developed to obtain approximate solutions @xcite . from eq .
( [ eq : a1 ] ) , we may derive equations of motion for @xmath4 and @xmath5 expressed by @xmath6 where the bracket @xmath7 denotes the expectation value .
although equations of motion given by eq .
( [ eq : a2 ] ) are closed within @xmath4 and @xmath8 for a ho system , they generally yield equations of motion including higher - order fluctuations such as @xmath9 , @xmath10 and @xmath11 where @xmath12 and @xmath13 .
it is necessary to develop an approximate method to close or truncate a hierarchical chain of equations of motion .
the gaussian wavepacket method ( gwm ) is one of such methods whose main aim is a semi - classical description of quantum systems ( for a recent review on gwm , see ref .
if the wavefuction is gaussian at @xmath14 in a ho system , it remains at all @xmath0 .
heller @xcite proposed that even for more realistic potentials , we may adopt a ( thawed ) gaussian wavepacket given by @xmath15 \right ] , \label{eq : a3}\end{aligned}\ ] ] where @xmath16 and @xmath17 are time - dependent complex parameters .
heller @xcite derived equations of motion for @xmath4 , @xmath5 , @xmath16 and @xmath17 , employing an assumption that the potential expanded in the taylor series at @xmath18 may be truncated by @xmath19 where @xmath20 signifies the @xmath21th derivative of @xmath2 .
the concept of the gaussian wavepacket has been adopted in many fields @xcite .
dynamics is well described by gwm for a ho system where motions of fluctuations are separated from those of @xmath4 and @xmath5 , leading to the uncertainty relation : @xmath22 .
various types of variants of gwm such as the frozen @xcite and generalized gaussian wavepacket methods @xcite have been proposed @xcite . among them , we pay our attention into the time - dependent variational approximation ( tdva ) which employs the normalized squeezed coherent - state gaussian wavepacket given by @xcite @xmath23 , \label{eq : a5}\end{aligned}\ ] ] @xmath24 and @xmath25 being time - dependent parameters . for
the introduced squeezed coherent state , equations of motion given by eq .
( [ eq : a2 ] ) are closed within @xmath4 , @xmath5 , @xmath26 and @xmath27 [ see eqs .
( [ eq : g17])-([eq : g19 ] ) ] .
a comparison between heller s gwm and tdva is made in refs .
@xcite .
there have been many studies on gwm which is applied to ho , anharmonic oscillator ( ao ) and morse potentials @xcite
. however , gwm has some difficulty when applied to a potential @xmath2 including terms of @xmath28 with @xmath29 . although it has been claimed that gwm yields a fairly good result for ao systems @xcite , we wonder whether it actually works for double - well ( dw ) systems .
dw potential models have been employed in a wide range of fields including physics , chemistry and biology ( for a recent review on dw systems , see ref .
lin and ballentine @xcite , and utermann , dittrich and hnggi @xcite studied semi - classical properties of dw systems subjected to periodic external forces , calculating the husimi function @xcite .
their calculations showed a chaotic behavior in accordance with classical driven dw systems .
igarashi and yamada @xcite studied a coherent oscillation and decoherence induced by applied polychromatic forces in quantum dw system . by using tdva , pattanayak and
schieve @xcite pointed out that a chaos is induced by quantum noise in dw systems without external forces although classical counterparts are regular .
this is in contrast to the usual expectation that quantum effects suppress classical chaos .
chaotic - like behavior was reported in a square dw system obtained by the exact calculation @xcite .
quantum chaos pointed out in ref .
@xcite is still controversial @xcite .
quite recently , hasegawa has studied effects of the asymmetry on the specific heat @xcite and tunneling @xcite in the asymmetric dw systems , by using the spectral method ( sm ) in which expansion coefficients are evaluated for energy matrix elements with a finite size of @xmath30 [ eqs .
( [ eq : b7 ] ) and ( [ eq : b8 ] ) ] . model calculations in refs . @xcite have pointed out intrigue phenomena which are in contrast with earlier relevant studies .
it is worthwhile to examine the validity of tdva applied to dw systems with the use of quasi - exact sm @xcite , which is the purpose of the present paper .
such a study has not been reported as far as we are aware of .
it is important to clarify the significance of tdva for dw systems .
the paper is organized as follows . in section 2 ,
we mention the calculation method employed in our study .
we consider quantum systems described by the symmetric dw ( sdw ) model . in solving dynamics of a gaussian wavepacket in the sdw
, we have adopted the two methods : sm and tdva . in section 3 ,
we report calculated results of the magnitude of wavefunction ( @xmath31 ) , an expectation value of @xmath32 ( @xmath4 ) , the auto - correlation function ( @xmath33 ) and the uncertainty product ( @xmath9@xmath10 ) . in section 4 we apply our method also to an ao model .
section 5 is devoted to our conclusion .
we consider a dw system whose hamiltonian is given by @xcite @xmath34 where @xmath35 here @xmath36 , @xmath32 and @xmath37 express mass , position and momentum , respectively , of a particle , @xmath2 stands for the dw potential , and @xmath38 is the ho hamiltonian with the oscillator frequency @xmath39 .
the sdw potential @xmath2 has stable minima at @xmath40 and an unstable maximum at @xmath41 with the potential barrier of @xmath42 .
a prefactor of @xmath43 in eq .
( [ eq : h1 ] ) is chosen such that the dw potential @xmath2 has the same curvature at the minima as the ho potential @xmath44 : @xmath45 .
figure [ fig1 ] expresses the adopted quartic dw potential @xmath2 with @xmath46 and @xmath47 in eq .
( [ eq : h1 ] ) .
eigenfunction and eigenvalue for @xmath38 are given by @xmath48 where @xmath49 stands for the @xmath50th hermite polynomial . and
@xmath47 in eq .
( [ eq : h1 ] ) , dashed curves expressing eigenvalues of @xmath51 ( @xmath52 ) .
, width=264 ] various approximate analytical and numerical methods have been proposed to solve the schrdinger equation given by eq .
( [ eq : a1 ] ) @xcite .
assuming @xmath53 , we first solve the steady - state schrdinger equation , @xmath54 , with the eigenvalue @xmath55 .
the stationary wavefunction @xmath56 is expanded in terms of @xmath57 @xmath58 leading to the secular equation @xmath59 with @xmath60 where @xmath61 is the maximum quantum number .
for the time - dependent state , we adopt sm in which the eigenfunction @xmath1 is expanded in terms of @xmath57 with finite @xmath61 @xmath62 time - dependent expansion coefficients @xmath63 obey equations of motion given by @xmath64 equation ( [ eq : b8 ] ) expresses the ( @xmath65 ) first - order differential equations , which may be solved for given initial conditions of @xmath66 .
initial values of expansion coefficients @xmath66 are determined by @xmath67 for a given gaussian wavepacket @xmath68 [ eq .
( [ eq : a5 ] ) ] @xmath69 , \label{eq : b11}\end{aligned}\ ] ] where @xmath70 and @xmath71 are initial position and momentum , respectively , and @xmath72 and @xmath73 are assumed initial parameters at @xmath74 .
once solutions of @xmath63 in eq .
( [ eq : b8 ] ) are obtained , the wavefunction @xmath1 may be constructed by eq .
( [ eq : b7 ] ) .
matrix elements @xmath75 in eq .
( [ eq : b9 ] ) may be analytically evaluated , and various time - dependent averages such as @xmath4 and @xmath5 are expressed in terms of @xmath63 ( see the appendix ) .
we expect that sm with @xmath30 adopted in our numerical calculations is fairly accurate @xcite .
some results of sm have been cross - checked , by solving the schrdinger equation with the use the mathematica resolver for the partial differential equation .
equations of motion in eq .
( [ eq : a2 ] ) are expressed by @xmath76 equations ( [ eq : g1])-([eq : g4 ] ) include higher - order fluctuations which are not closed in general .
it is possible to construct various approximations depending on how many terms are taken into account in eqs .
( [ eq : g1])-([eq : g4 ] ) .
if we neglect the second term of eq .
( [ eq : g2 ] ) , eqs .
( [ eq : g1 ] ) and ( [ eq : g2 ] ) form classical equations of motion . when we neglect the second term in eq .
( [ eq : g2 ] ) and truncate eqs .
( [ eq : g3 ] ) and ( [ eq : g4 ] ) at @xmath77 , eqs .
( [ eq : g1])-([eq : g4 ] ) reduce to equations of motion in heller s gwm .
equations of motion including up to fourth - order corrections were obtained in ref .
@xcite . to close a hierarchal chain of equations of motion
, tdva assumes that a wavepacket is expressed by the normalized squeezed coherent state given by eq .
( [ eq : a5 ] ) , implying relations @xcite @xmath78 where @xmath24 and @xmath25 are time - dependent parameters .
note that eqs .
( [ eq : g12])-([eq : g13b ] ) yield the uncertainty product expressed by @xmath79 these lead to equations of motion given by @xmath80 alternatively , eqs .
( [ eq : g5a])-([eq : g5c ] ) may be rewritten as @xmath81 which show a closure of equations of motion within @xmath4 , @xmath5 , @xmath9 and @xmath82 .
-@xmath37 phase space of the sdw model for various initial energies of @xmath83 , marks @xmath84 denoting positions of initial states adopted in model calculations . ,
width=264 ] we apply our calculation method to the sdw potential given by eq .
( [ eq : h1 ] ) .
we have calculated energy matrix elements of @xmath85 , by using eq .
( [ eq : h6 ] ) with @xmath86 and @xmath30 . obtained eigenvalues
are @xmath87 0.450203 , 0.474126 , 1.09262 , 1.39334 and 1.91286 for @xmath88 to 4 , respectively , which are plotted by dashed curves in fig .
the ground state ( @xmath89 ) and first excited state ( @xmath90 ) , which are quasi - degenerate , are below the potential barrier of @xmath47 .
the energy gap between ground and first excited states is @xmath91 .
low - lying eigenvalues calculated with @xmath30 are in good agreement with those obtained with @xmath92 @xcite .
figure [ fig2 ] shows the classical @xmath32-@xmath37 phase space for initial energies of @xmath93 , 1.0 and 1.125 .
marks @xmath94 in fig .
[ fig2 ] show two initial states in the @xmath32-@xmath37 phase space adopted in our calculations .
calculated results for the two initial states of @xmath95 and @xmath96 will be separately reported in the following .
we have adopted the gaussian wavepacket @xmath68 locating at the stable point of the left well with @xmath97 , and @xmath98 and @xmath99 at @xmath74 , which yields the minimum uncertainty product of @xmath100 .
initial coefficients @xmath66 calculated by eq . (
[ eq : b10 ] ) are real with appreciable magnitudes for @xmath101 .
a norm of the initial gaussian wavepacket is @xmath102 [ eq . ( [ eq : c2 ] ) ] . after solving ( @xmath65 ) first - order differential equations for @xmath63 given by eq .
( [ eq : b8 ] ) for initial values of @xmath66 , we obtain the time - dependent eigenfunction @xmath1 expressed in terms of @xmath63 in eq .
( [ eq : b7 ] ) .
figure [ fig3 ] shows the 3d plot of @xmath31 calculated by sm .
we note that the gaussian wavepacket in sm quickly spreads as the time develops . in order to scrutinize the behavior of @xmath31 at small @xmath103 , its time dependence at @xmath104
is plotted by bold solid curves in fig .
[ fig4 ] , where solid curves denote results of tdva .
the gaussian wavepacket becomes widespread even at @xmath105 in sm , and its trend becomes more significant with increasing @xmath103 .
figure [ fig4 ] clearly shows that @xmath31 in sm is quite different from that in tdva and that @xmath1 can not be expressed by a single gaussian except at @xmath74 . as functions of @xmath32 and @xmath103 calculated by sm for the sdw model ( @xmath106 ) .
, width=453 ] dependence of @xmath31 at various @xmath103 of the sdw model calculated by sm ( bold solid curves ) and tdva ( solid curve ) ( @xmath106 ) . , width=377 ] of the sdw model calculated by ( a ) sm and ( b ) tdva ( @xmath106 ) . , width=264 ] vs. @xmath5 plot of the sdw model calculated by ( a ) sm and ( b ) tdva , time step being @xmath107 for @xmath108 ( @xmath106 ) . , width=264 ] and ( b ) @xmath109 of the sdw model calculated by sm ( @xmath106 )
. note that @xmath110 equals to @xmath109 in tdva [ eq .
( [ eq : g13 ] ) ] , which is not realized in ( a ) and ( b ) . , width=264 ] of the sdw model calculated by ( a ) sm and ( b ) tdva ( @xmath111 ) .
, width=264 ] the difference between sm and tdva is more clearly seen in the time - dependent expectation value of @xmath4 .
figure [ fig5](a ) shows @xmath4 of sm expressing a tunneling of a particle with the period of about 260 , which is consistent with the period estimated from the energy gap by @xmath112 . on the contrary , @xmath4 of tdva in fig .
[ fig5](b ) shows more rapid oscillation with a period of about @xmath113 .
figures [ fig6](a ) and [ fig6](b ) show @xmath4 vs. @xmath5 plots calculated by sm and tdva , respectively . the @xmath4 vs. @xmath5 plot of sm in fig .
[ fig6](a ) is quite different from that of tdva in fig .
[ fig6](b ) .
figure [ fig7](a ) shows the uncertain product of @xmath114 calculated by sm , which expresses a measure of quantum fluctuation .
it starts from the minimum uncertainty of @xmath115 at @xmath14 , and with increasing @xmath103 it grows and oscillates between about 5 and 17 with the period of about 130 . for a comparison
, we plot @xmath109 in fig .
[ fig7](b ) .
tdva assumes the equality of @xmath116 as given by eq .
( [ eq : g13 ] ) .
figures [ fig7](a ) and [ fig7](b ) , however , imply that this equality is not satisfied in sm .
figures [ fig8](a ) and [ fig8](b ) show the auto - correlation functions @xmath117 calculated by sm and tdva , respectively , with eq .
( [ eq : c2 ] ) in the appendix .
@xmath117 of sm , which is unity at @xmath74 , oscillates between about 0.1 and 0.7 with a period of about 260 .
the result of sm in fig .
[ fig8](a ) is again quite different from that of tdva in fig . [
fig8](b ) .
next we adopt a gaussian wavepacket with a different initial state of @xmath96 but with the same @xmath98 and @xmath99 at @xmath74 .
initial coefficients @xmath66 calculated by eq .
( [ eq : b10 ] ) are complex with appreciable magnitudes for @xmath119 .
the initial state of @xmath118 locates near a top of the potential barrier ( see fig .
[ fig2 ] ) .
note that in the classical calculation , the @xmath32 vs. @xmath37 plot forms a cocoon shape extending from @xmath120 to @xmath121 and from @xmath122 to 1.5 , as shown in fig .
then at @xmath0 , a particle starting from @xmath96 rolls down the potential up to @xmath121 and then approaches @xmath120 after passing through @xmath123 in the classical calculation .
however , this classical behavior is quite different from quantum results calculated by sm and tdva .
the 3d plot of @xmath31 of sm shown in fig .
[ fig9 ] has appreciable magnitudes at @xmath124 for @xmath125 .
bold solid curves and solid curves in fig .
[ fig10 ] show @xmath31 calculated by sm and tdva , respectively .
@xmath31 of sm , which is distorted and spreads at @xmath0 , is different from the relevant result of tdva .
an expectation value of @xmath4 of sm in fig .
[ fig11](a ) does not so much depart from the initial point of @xmath126 in contrast to that of tdva shown in fig .
[ fig11](b ) .
figures [ fig12](a ) and [ fig12](b ) show @xmath4 vs. @xmath5 plots calculated by sm and tdva , respectively .
the result of sm in fig .
[ fig12](a ) exhibits a random - like motion , which is different from a quasi - periodic motion of tdva in fig . [
fig12](b ) .
figures [ fig13](a ) and [ fig13](b ) show @xmath110 and @xmath109 , respectively , calculated by sm .
we note that @xmath127 in sm , which is in contrast with eq .
( [ eq : g13 ] ) in tdva . as functions of @xmath32 and @xmath103 calculated by sm for the sdw model ( @xmath128 ) . ,
width=415 ] dependence of @xmath31 at various @xmath103 of the sdw model calculated by sm ( bold solid curves ) and tdva ( solid curves ) ( @xmath128 ) .
, width=377 ] of the sdw model calculated by ( a ) sm and ( b ) tdva ( @xmath128 ) .
, width=264 ] vs. @xmath5 plot of the sdw model calculated by ( a ) sm and ( b ) tdva , time step being @xmath107 for @xmath108 ( @xmath128 ) . ,
width=264 ] and ( b ) @xmath109 of the sdw model calculated by sm ( @xmath128 ) .
note that @xmath129 equals to @xmath109 in tdva [ eq .
( [ eq : g13 ] ) ] , which is not held in sm . ,
refs . @xcite showed that by using a change of variables given by @xmath130 eqs .
( [ eq : g5a])-([eq : g5c ] ) are transformed to @xmath131 it was shown that fluctuation variables @xmath132 and @xmath133 are conjugate and that the effective hamiltonian may be expressed in the extended phase space spanned by @xmath4 , @xmath5 , @xmath132 and @xmath133 as given by @xcite @xmath134 we should note that the effective hamiltonian given by eq .
( [ eq : g25 ] ) relies on the identities given by eqs .
( [ eq : g12])-([eq : g13b ] ) which are based on the assumed squeezed gaussian wavepacket given by eq .
( [ eq : a5 ] ) . if these identities are not held as our sm calculation suggests , the effective hamiltonian given by eq .
( [ eq : g25 ] ) is not valid in dw systems . as functions of @xmath32 and @xmath103 of the ao model with ( a ) @xmath135 and ( b ) @xmath136 calculated by sm ( @xmath137 ) .
, width=377 ] dependence of @xmath31 at @xmath138 ( @xmath139 ) of the ao model with @xmath136 calculated by sm ( bold solid curves ) and tdva ( solid curve ) ( @xmath137 ) , where @xmath140 @xmath141 is a period for a ho system ( @xmath142 ) .
, width=377 ] in the ao model with @xmath136 calculated by ( a ) sm and ( b ) tdva ( @xmath137 ) .
, width=264 ] we have studied gaussian wavepacket dynamics of quantum dw systems in the preceding section . it is worthwhile to examine also an ao model given by @xmath143 where @xmath144 expresses a degree of anharmonicity .
we have repeated calculations , by using sm and tdva with necessary modifications .
figure [ fig14](a ) shows the 3d plot of @xmath31 for @xmath135 with an assumed gaussian state for @xmath145 , and @xmath98 and @xmath99 at @xmath74 . in the case of a ho potential ( @xmath146 ) , @xmath31 is periodic with a period of @xmath147 .
it is shown in [ fig14](a ) that @xmath31 for a small @xmath135 is nearly periodic as that for @xmath146 at @xmath148 .
for a larger @xmath136 , however , this periodicity is destroyed and the wavepacket spreads in the non - gaussian form as fig .
[ fig14](b ) shows .
this is more clearly realized in fig .
[ fig15 ] where @xmath31 calculated by sm ( bold solid curves ) are quite different from their counterparts obtained by tdva ( solid curve ) .
our result of sm in fig .
[ fig15 ] is consistent with that in ref .
@xcite which studied effects of anharmonicity and interactions in dw systems .
figures [ fig16](a ) and [ fig16](b ) show time dependences of @xmath4 calculated by sm and tdva , respectively . @xmath4 oscillates with a period of about @xmath149 in both results .
however , a period of its envelope variation in sm ( @xmath150 ) is larger than that in tdva ( @xmath151 ) : the former corresponds to the revival time after which a wavepacket periodically returns to the initial shape .
by using sm and tdva , we have calculated time dependences of wavefunctions , averages of position and momentum , auto - correlation function , and uncertainty product in quantum sdw systems .
the validity of tdva has been examined by comparisons between results of sm and tdva .
we have obtained following results : \(1 ) an initial gaussian wavepacket in dw systems of sm spreads and deforms at @xmath0 where a wavepacket can not be expressed by a single gaussian in contrast to tdva , \(2 ) time dependences of expectation values of @xmath4 and @xmath5 , and the auto - correlation function in sm are quite different from their counterparts in tdva , and \(3 ) the identity relation for uncertainty product assumed in tdva [ eq .
( [ eq : g13 ] ) ] is not satisfied in sm .
the item ( 1 ) holds also in asymmetric dw systems @xcite .
the item ( 2 ) implies that the tunneling phenomenon characteristic in dw systems can not be well accounted for in tdva ( fig .
[ fig5 ] ) @xcite .
the item ( 3 ) suggests that the effective hamiltonian in the extended phase space given by eq .
( [ eq : g25 ] ) does not hold in dw systems because it is derived with the squeezed gaussian wavepacket with assumptions given by eqs .
( [ eq : g12])-([eq : g13b ] ) in tdva .
gwm is best applied to dynamics in ho and ao with a small anharmonicity , for which it provides us with an efficient and physically - transparent calculation method .
our calculations , however , point out that gwm is not a good approximation for dw systems . for a better description of quantum dw systems , it might be necessary to adopt extended gwms with superimposed multiple gaussian wavepackets ( see ref .
@xcite , related references therein ) , which are much sophisticated and complicated than the original heller s gwm @xcite .
this work is partly supported by a grant - in - aid for scientific research from ministry of education , culture , sports , science and technology of japan .
matrix elements @xmath152 in eq .
( [ eq : b9 ] ) are given as follows : we rewrite the potential @xmath2 as @xmath153 with @xmath154 by using relations given by @xmath155 we obtain the symmetric matrix elements @xmath152 for @xmath156 given by @xmath157 \:\delta_{n , k } \nonumber \\ & + & \left[a_3 \left(\frac{g}{2}\right)^{3/2 } n \sqrt{n } + a_1 \left ( \frac{g}{2 } \right)^{1/2 } \sqrt{n } \right ] \delta_{n-1,k } \nonumber \\ & + & \left [ \frac{a_4 g^2}{8 } ( n-1)\sqrt{n(n-1 ) } + \frac{a_2 ' \:g}{4 } \sqrt{n(n-1 ) } \right ] \:\delta_{n-2,k } \nonumber \\ & + & \frac{a_3}{3 } \left(\frac{g}{2}\right)^{3/2 } \sqrt{n(n-1)(n-2)}\ : \delta_{n-3,k } \nonumber \\ & + & \frac{a_4 g^2}{16 } \sqrt{n(n-1)(n-2)(n-3 ) } \:\delta_{n-4,k } , \label{eq : h6}\end{aligned}\ ] ] where @xmath158 . various time - dependent quantities
may be expressed in terms of @xmath63 as follows : after some manipulations with the use of the relations given by eqs.([eq : c1a ] ) and ( [ eq : c1b ] ) , the auto - correlation function is given by @xmath159 and expectation values such as @xmath160 and @xmath161 are expressed by @xmath162 , \label{eq : c3a } \\ \langle p(t )
\rangle & = & i \sqrt{\frac{\hbar^2}{2 g } } \sum_n \left[\sqrt{n+1}\ : c_{n+1}^*(t ) c_n(t ) - \sqrt{n}\ : c_{n-1}^*(t ) c_n(t ) \right ] , \\ \langle x(t)^2 \rangle & = & \left ( \frac{g}{2 } \right ) \sum_n [ \sqrt{(n+1)(n+2 ) } \:c_{n+2}^*(t ) c_n(t)+(2 n+1 ) \:c_{n}^*(t ) c_n(t ) \nonumber \\ & + & \sqrt{n(n-1 ) } \:c_{n-2}^*(t ) c_n(t ) ] , \\
\langle p(t)^2 \rangle & = & - \left ( \frac{\hbar^2}{2 g } \right ) \sum_n [ \sqrt{(n+1)(n+2 ) } \:c_{n+2}^*(t ) c_n(t)-(2 n+1 ) \:c_{n}^*(t ) c_n(t ) \nonumber \\ & + & \sqrt{n(n-1 ) } \:c_{n-2}^*(t ) c_n(t ) ] , \\
\langle x(t)p(x)+p(t)x(t ) \rangle & = & i \:\hbar \sum_n [ \sqrt{(n+1)(n+2 ) } \:c_{n+2}^*(t ) c_n(t ) \nonumber \\ & - & \sqrt{n(n-1 ) } \:c_{n-2}^*(t ) c_n(t ) ] .
\label{eq : c3b}\end{aligned}\ ] ] | we have examined the validity of the time - dependent variational approximation ( tdva ) to the gaussian wavepacket method ( gwm ) for quantum double - well ( dw ) systems , by using the quasi - exact spectral method ( sm ) .
comparisons between results of wavefunctions , averages of position and momentum , the auto - correlation function , and an uncertainty product calculated by sm and tdva have been made .
it has been shown that a given initial gaussian wavepacket in sm is quickly deformed at @xmath0 where a wavepacket can not be expressed by a _ single _
gaussian , and that assumptions on averages of higher - order fluctuations in tdva are not justified .
these results cast some doubt on an application of tdva to dw systems .
gaussian wavepacket dynamics in anharmonic potential systems is studied also .
keywords : gaussian wavepacket , time - dependent variational approximation , spectral method , double - well potential |
the fokas - lenells equation ( fl)@xcite @xmath0 is one of the important models from both mathematical and physical considerations . in eq.([fl ] ) , @xmath1 represents complex field envelope and asterisk denotes complex conjugation , and subscript @xmath2 ( or @xmath3 ) denotes partial derivative with respect to @xmath2 ( or @xmath3 ) .
the fl equation @xcite is related to the nonlinear schrdinger ( nls ) equation in the same way that the camassa - holm equation is associated with the kdv equation . in optics , considering suitable higher order linear and nonlinear optical effects
, the fl equation has been derived as a model to describe femtosecond pulse propagation through single mode optical silica fiber and several interesting solutions have been constrcuted @xcite . here , considering the physical significance of the fl equation and inspired by the importance of the recent interesting developments in the analysis of rogue waves of the nls and the dnls equations
, we shall construct rogue wave solutions of the fl equation , by using the darboux transformation ( dt)@xcite . our construction reveals that there exists a difference between the fl system and other integrable models , like the ablowitz - kaup - newell - segur ( akns ) @xcite system and the kaup - newell ( kn ) @xcite system . before discussing the dt of the fl system ,
let us briefly discuss the importance of rogue waves in mathematics and physics .
rogue waves have recently been the subject of intensive investigations in oceanography@xcite , where they occur due to modulation instability @xcite , either through random initial condition @xcite or via other processes @xcite .
the first order rogue wave usually takes the form of a single peak hump with two caves in a plane with a nonzero boundary .
one of the possible generating mechanisms for rogue waves is through the creation of breathers@xcite which can be formed due to modulation instability .
then , larger rogue waves can emerge when two or more breathers collide with each other @xcite .
rogue waves have also been observed in space plasmas@xcite , as well as in optics when propagating high power optical radiation through photonic crystal fibers @xcite . considering higher order effects in the propagation of femtosecond pulses ,
rogue waves have been reported in the hirota equation and the nls - mb system@xcite .
this paper is organized as follows .
section 2 , presents a simple approach to dt for the fl system , the determinant representation of the 1-fold dt , and formulae of @xmath4}$ ] and @xmath5}$ ] expressed in terms of eigenfunctions of the associated spectral problem .
the reduction of dt to the fl equation is also discussed by choosing appropriate pairs of eigenvalues and eigenfunctions .
section 3 presents the construction of rogue wave solutions by using a taylor series expansion about the breather solution generated by dt from a periodic seed solution with constant amplitude .
finally , in section 4 we conclude our results .
let us start from the non - trivial flow of the fl ( fokas - lenells ) system@xcite in the following modified form , @xmath6 @xmath7 which are exactly reduced to the fl eq.([fl ] ) for @xmath8 while the choice @xmath9 would lead to eq.([fl ] ) with change in sign of the nonlinear term .
the lax pairs corresponding to coupled fl eq.([sy1 ] ) and ( [ sy2 ] ) can be given by the fl spectral problem of the form @xcite
@xmath10 @xmath11 with @xmath12 @xmath13 here @xmath14 , an arbitrary complex number , is called the eigenvalue(or isospectral parameter ) , and @xmath15 is called the eigenfunction associated with @xmath14 of the fl system .
equations ( [ sy1 ] ) and ( [ sy2 ] ) are equivalent to the compatibility condition @xmath16=0 $ ] of ( [ sys11 ] ) and ( [ sys22 ] ) .
now , let us consider a matrix @xmath17 of gauge transformation for the spectral problem ( [ sys11 ] ) and ( [ sys22 ] ) with the following form @xmath18}=t~\psi,\ ] ] and then @xmath19}}_{x}=u^{[1]}~\psi^{[1]},\ \ u^{[1]}=(t_{x}+t~u)t^{-1}.\ ] ] @xmath20}}_{t}=v^{[1]}~\psi^{[1 ] } , \ \ v^{[1]}=(t_{t}+t~v)t^{-1}.\ ] ] + by cross differentiating ( [ bh1 ] ) and ( [ bh2 ] ) , we obtain @xmath21}}_{t}-{v^{[1]}}_{x}+[{u^{[1]}},{v^{[1]}}]=t(u_{t}-v_{x}+[u , v])t^{-1}.\ ] ] this implies that , in order to prove eqs.([sy1 ] ) and ( [ sy2 ] ) are invariant under the transformation ( [ bh3 ] ) , it is crucial to construct a matrix @xmath22 so that @xmath23}$ ] and @xmath24}$ ] have the same forms as that of @xmath25 and @xmath26 . at the same time
the old potentials ( or seed solutions)(@xmath1 , @xmath27 ) in spectral matrixes @xmath25 and @xmath26 are mapped into new potentials ( or new solutions)(@xmath4}$ ] , @xmath5}$ ] ) in terms of transformed spectral matrixes @xmath23}$ ] and @xmath24}$ ] .
this newly obtained matrix @xmath22 is a darboux transformation of the fl system of eqs .
( [ sy1 ] ) and ( [ sy2 ] ) . based on the dt for the nls@xcite and the dnls@xcite
, we suppose that a trial darboux matrix @xmath22 in eq.([bh3 ] ) is assumed to be in the form @xmath28 here @xmath29 are undetermined coefficients and they are function of ( @xmath2 , @xmath3 ) , which will be expressed in terms of the eigenfunctions associated with @xmath30 and @xmath31 in the fl spectral problem and @xmath32 and @xmath33 are constants . for our further analysis , setting two eigenfunctions @xmath34 as @xmath35 after tedious but straightforward calculations , we get the one - fold darboux transformation of the fl system as @xmath36 and then the corresponding new solutions @xmath4}$ ] and @xmath5}$ ] are given by @xmath37}=q \dfrac{a_{-1}}{d_{-1}}+\dfrac{b_{0}}{d_{-1 } } , r^{[1]}=r \dfrac{d_{-1}}{a_{-1}}+\dfrac{c_{0}}{a_{-1}}.\end{aligned}\ ] ] with the condition @xmath38 and @xmath39 having obtained the values of all the unknown coefficients , we are now in a position to consider the reduction of the dt so that @xmath4}=(r^{[1]})^*$ ] , then the dt of the fl equation is given . under the reduction condition
@xmath40 with @xmath41 is an eigenfunction associated with the eigenvalue @xmath42 , then @xmath43 and @xmath44 .
it is easy to check @xmath4}=(r^{[1]})^*$ ] with the help of following choice of eigenvalues and eigenfunctions : @xmath45 thus , under the above choice of eigenfunctions and eigenvalues , the resulting @xmath46 in eq.([onefolddt ] ) can be called as one - fold darboux transformation of the fl equation .
further , taking one pair of eigenvalues @xmath47 and @xmath48 and their corresponding eigenfunctions , the final form of one - fold dt eq.([onefolddt ] ) reduces to a new solution of the form @xmath49}=q+\dfrac{({\lambda_{2}}^{2}-{\lambda_{1}}^{2 } ) \phi_{1}{\varphi_{1}}^{\ast}}{\lambda_2\lambda_1(-\lambda_{2}\varphi_{1 } { \varphi_{1}}^{\ast}+\lambda_{1}\phi_{1}{\phi_{1}}^{\ast})},\ ] ] notice that the denominator of @xmath4}$ ] has a nonzero imaginary part if @xmath50 is not zero , which clearly indicates that the new solution @xmath4}$ ] is a non - singular solution .
using the results of dt constructed above , breather solutions of fl equation can be generated by assuming a periodic seed solution of eq.([q2j2 ] ) , then we can construct the rogue waves of the fl equation from a taylor series expansion of the breather solutions . for this purpose , assuming @xmath51 and @xmath52 as two real constants , then @xmath53 is a periodic solution of the fl equation , which will be used as a seed solution of the dt . substituting the above form of @xmath1 into the spectral problem eq.([sys11 ] ) and eq.([sys22 ] ) , and using the method of separation of variables and the superposition principle , the eigenfunction @xmath54 associated with @xmath30
are given explicitly .
for simplicity , setting @xmath55 , taking @xmath56 back into eq.([q2j2 ] ) , then the final form of the breather solution is obtained in the form @xmath57}=\lefteqn{(\dfrac{-2\alpha_{1}\beta_{1}(w_1\cosh(h_1)+w_2sinh(h_1)+w_3\cos(h_2)+w_4sin(h_2))}{w_5\cosh(h_1)+w_6sinh(h_1)+w_7\cos(h_2)+w_8sin(h_2)}{}}\nonumber\\ & & { } \mbox{\hspace{0.5cm}}-\dfrac{\sqrt{a+2{\alpha_1}^{2}-2{\beta_1}^{2}}}{a})\exp(i(ax+(a+2-\dfrac{2({\beta_{1}}^{2}-{\alpha_{1}}^{2})t}{a^2}))),\end{aligned}\ ] ] where @xmath58 from eq.([q2j3 ] ) , we can easily get the ma breathers@xcite(time periodic breather solution ) and the akhmediev breathers @xcite ( space periodic breather solution ) solution . in general , the solution in eq.([q2j3 ] ) evolves periodically along the straight line with a certain angle between @xmath2 and @xmath3 axis
. the dynamical evolution of @xmath59}|^2 $ ] in eq.([q2j3 ] ) is plotted in figure 1 . finally , a taylor series expansion about the breather solution of the fl equation is used to construct the rogue wave of the fl equation in the following way . by letting @xmath60 in @xmath61
, we obtain the rogue wave solution in the form @xmath62}_{rw}=\alpha_{1 } \dfrac{s_2}{s_3}\exp{s_1}\ ] ] with @xmath63 by letting @xmath64 , so @xmath65}_{rw}|^{2}\rightarrow $ ] @xmath66 .
the maximum amplitude of @xmath65}_{rw}|^{2}$ ] occurs at @xmath67 and @xmath68 and is equal to @xmath69,and the minimum amplitude of @xmath65}_{rw}|^{2}$ ] occurs at @xmath70 and @xmath71 and is equal to @xmath72 . through figure 2 and figure 3 of
@xmath65}_{rw}|^2 $ ] , the main features ( such as large amplitude and local property on ( x - t ) plane ) of the rogue wave are clearly shown .
thus , in this paper , considering fl system of equation which describes nonlinear pulse propagation through single mode optical fiber , the determinant representation of the one - fold dt for the fl system is given eqs.([onefolddt ] ) and ( [ stt ] ) . by choosing pair of eigenvalues in the form @xmath73 and assuming suitable seed solutions , the breather solution and rogue waves of the fl equation are derived in eq .
( [ q2j3 ] ) and eq .
( [ focas ] ) .
our results provide an alternative possibility to observe rogue waves in optical system . from the one - fold dt , it is interesting to observe that the dt of the fl system exhibits the following novelty in comparison with other integrable models like the akns and the kn systems : the dt matrix of the fl system has three different terms depending on @xmath14 .
thus , the dt as well as the rogue wave of the fl system present novel features , in comparison with the dt and rogue wave solutions of the standard integrable systems like the akns and the kn systems . the construction of higher order rogue wave of the fl equation by using the determinant representation of the dt will be published elsewhere . *
acknowledgments * this work is supported by the nsf of china under grant no.10971109 and k.c.wong magna fund in ningbo university .
jingsong he is also supported by program for ncet under grant no.ncet-08-0515 and natural science foundation of ningbo under grant no.2011a610179 .
we thank prof .
yishen li(ustc , hefei , china ) for his useful suggestions on the rogue wave . j.
he thank prof . a.s.fokas and dr . dionyssis mantzavino(cambridge university ) for many helps on this paper .
kp wishes to thank the dst , dae - brns , ugc , csir , government of india , for the financial support through major projects .
99 a.s.fokas,physica.d .
87 ( 1995 ) 145 .
j.lenells and a.s.fokas,inverse.probl . 25 ( 2009 ) 115006 .
j.lenells and a.s.fokas,nonlinearity . 22 ( 2009 )
20 ( 2010 ) 709 .
j.lenells , stud .
math.123 ( 2009 ) 215 .
g.neugebauer and r.meinel , phys.lett.a.100 ( 1984 ) 467 . v. b. matveev , m.a .
salle , darboux transfromations and solitons(springer - verlag , berlin)(1991 ) .
j.s.he , l.zhang , y.cheng and y.s.li , science in china series a : mathematics.12 ( 2006 ) 1867 .
kenji imai , j.phys.soc.japan.68 ( 1999 ) 355 .
h steudel , j.phys.a : math.gen.36 ( 2003 ) 1931 . s.w .
xu , j.s.he and l.h.wang , j. phys .
a : math . theor .
44 ( 2011 ) 305203 .
m.j.ablowitz , d.j.kaup , a.c.newell and h.segur , phys.rev.lett . 31
( 1973 ) 125 .
m.j.ablowitz and p.a.clarkson , solitons , nonlinear evolution equations and inverse scattering ( cambridge : cambridge university press ) ( 1991 ) .
d.j.kaup and a.c.newell , j.math.phys . 19 ( 1978 ) 798 .
c.kharif and e.pelinovsky , eur.j.mech.b ( fluids ) . 22 ( 2003 ) 603 .
c.kharif , e.pelinovsky and a.slunyaev , rogue waves in the ocean ( berlin : springer)(2009 ) .
n. akhmediev , a. ankiewicz and m.taki , phys .
a. 373 ( 2009 ) 675 .
n. akhmediev , j. m. soto - crespo and a. ankiewicz , phys .
a. 373 ( 2009 ) 2137 .
a.chabchoub , n.p.hoffmann and n.akhmediev , phys .
106 ( 2011 ) 204502 .
i.didenkulova1 and e.pelinovsky , nonlinearity . 24
( 2011 ) r1 .
d. h. peregrine , water waves , j. aust . math .
b , appl . math .
25 ( 1983 ) 16 .
kristian , b.dysthe and k.trulsen , phys scri .
82 ( 1999 ) 48 .
v.e.zakharov and a.i.dyachenko , eur.j.mech.b ( fluids ) . 29
( 2008 ) 127 .
v.e.zakharov and l.a.ostrovsky , phys.d .
238 ( 2009 ) 540 .
n.n.akhmediev and v.i.korneev , theor .
69 ( 1986 ) 1080 .
y.c.ma , stud.appl.math . 60
( 1979 ) 43 .
n. akhmediev , a.ankiewicz and j. m. soto - crespo , phys .
e. 80 ( 2009 ) 026601 . zhenya.yan , phys .
lett.a . 374
( 2009 ) 672 .
boling.guo,liming.ling and q.p.liu , phys.rev.e 85 ( 2012 ) 026607 .
j.heller and l.kaplan , nonlinearity . 24
( 2011 ) r67 .
y.otha , j.k.yang , proc .
r. soc . a 468 ( 2012 ) 1716 .
xuan ni , wen - xu wang and ying - cheng lai , europhysics.lett .
96 ( 2011 ) 44002 .
n. akhmediev , j .
m. soto - crespo and a. ankiewicz , phys .
a. 80 ( 2009 ) 043818 .
v. fedun , m.s.ruderman and r. erdlyi , phys .
lett.a . 372
( 2008 ) 6107 .
ruderman , euro .
jour , 185 ( 2010 ) 57 .
w.m.moslem , p.k.shukla and b.eliasson , euro.phys.lett .
96 ( 2011 ) 25002 .
d.r.solli , c.ropers , p.koonath and b.jalali , nature , 450 ( 2007 ) 1054 .
b.kibler , j. fatome , c.finot , g.millot,f .
dias , g. genty , n. akhmediev and j.m .
dudley , nature.physics .
6 ( 2010 ) 790 .
f.t.arecchi , u. bortolozzo , a .
montina and s. residori , phys .
106 ( 2011 ) 153901 .
adrian ankiewicz , j. m. soto - crespo and nail akhmediev , phys .
e. 81 ( 2010 ) 046602 .
yongsheng tao and jingsong he , phys .
e 85 ( 2012 ) 026601 .
jingsong he , shuwei xu and kuppuswamy porseizan , j.phys .
81 ( 2012 ) 033002 . | the fokas - lenells ( fl ) equation arises as a model eqution which describes for nonlinear pulse propagation in optical fibers by retaining terms up to the next leading asymptotic order ( in the leading asymptotic order the nonlinear schrdinger ( nls ) equation results ) . here
we present an explicit analytical representation for the rogue waves of the fl equation .
this representation is constructed by deriving an appropriate darboux transformation ( dt ) and utilizing a taylor series expansion of the associated breather solution . when certain higher - order nonlinear effects are considered , the propagation of rogue waves in optical fibers is given .
corresonding author : email : [email protected] , tel : 86 - 574 - 87600739 , fax : 86 - 574 - 87600744 ] + * key words * : nonlinear schrdinger equation , fokas - lenells equation , + darboux transformation , breather solution , rogue wave . *
pacs(2010 ) numbers * : 02.30.ik , 42.81.dp , 52.35.bj , 52.35.sb , 94.05.fg
* msc(2010 ) numbers * : 35c08 , 37k10 , 37k40 |
we have performed band - structure calculations using a @xmath61 @xmath49-mesh ( 32 atoms in the unit cell ) for the hf and bitc methods to confirm the convergence with respect to the number of @xmath49-points . for the bitc calculations , the band gap and the relative energy of the zn-@xmath0-band bottom to the valence - band top are 2.92 and @xmath62 ev , respectively , for a @xmath61 @xmath49-mesh , whereas they are 3.09 and @xmath63 ev , respectively , for a @xmath51 @xmath49-mesh ( 108 atoms in the unit cell ) . for the hf calculations , the corresponding values are 11.53 and 10.07 ev , respectively , for a @xmath61 @xmath49-mesh , whereas they are 11.42 and @xmath64 ev , respectively , for a @xmath51 @xmath49-mesh .
the difference between the values calculated using the two @xmath49-meshes is less than 0.2 ev for all of these values . in our previous study on the tc method @xcite
, we saw that the calculated band gap @xmath65 with @xmath46 @xmath49-points obeys an approximate relation : @xmath66 where @xmath67 is a constant . using this approximate relation for the band energies , the residual finite - size error in our calculations using a @xmath51 @xmath49-mesh
is estimated to be less than 0.1 ev .
therefore , we can say that the convergence with respect to the number of @xmath49-points is sufficiently achieved .
although it is possible that the approximate relation that is used above does not hold for this case @xcite , we can expect that the convergence will be sufficient for our discussion because the finite - size error is expected to be significantly reduced in a @xmath51 @xmath49-mesh from its value for a @xmath61 one .
we do not show the band structure calculated using the tc method without the bi - orthogonal extension in the main text .
this is because we found that the imaginary part of the eigenvalues of the zn-@xmath3 bands obtained by solving the one - body scf equation became as large as 0.1 ev in the tc method , which is more than 1,000 times larger than that in the bitc method . because the non - hermiticity in the ( bi)tc formulation originates from the similarity - transformation ,
the presumed equivalency between the tc solution and the true many - body eigenstate is suggested to break when the imaginary parts of the eigenvalues become too large .
the difference between the band structures obtained with the tc and bitc methods was small for weakly correlated systems except lif that is rather localized - electron system @xcite .
the differences in the electron densities calculated with the lda , hf , and bitc methods are shown in fig . s1 . in fig .
s1 , we can verify the tendencies explained in the main text , i.e. , that the bitc electron density tends to be localized near the atomic sites .
in addition , we can see an increase of the bitc electron density along the zn - o bonds near the oxygen atoms . | obtaining accurate band structures of correlated solids has been one of the most important and challenging problems in first - principles electronic structure calculation .
there have been promising recent active developments of wave function theory for condensed matter , but its application to band - structure calculation remains computationally expensive . in this letter , we report the first application of the bi - orthogonal transcorrelated ( bitc ) method : self - consistent , free from adjustable parameters , and systematically improvable many - body wave function theory , to solid - state calculations with @xmath0 electrons : wurtzite zno . we find that the bitc band structure better reproduces the experimental values of the gaps between the bands with different characters than several other conventional methods .
this study paves the way for reliable first - principles calculations of the properties of strongly correlated materials . to reveal fertile and nontrivial physics in condensed matter
, first - principles electronic - structure calculation has established itself as an indispensable tool in recent studies . for this purpose , density functional theory ( dft )
@xcite has played a leading role and has been applied to various materials ; however , the limitations of this theoretical framework have come to light .
one of the major problems is an inaccurate description of strong electron correlations , e.g. , in transition metal oxides .
the @xmath1 method @xcite is a promising way to ameliorate the inaccuracy of the band structures and has been applied to several solids , including @xmath0-electron systems . however , because the @xmath1 method is often applied without satisfying self - consistency , a nontrivial dependence on the initial dft calculations is introduced .
it has also been reported that the @xmath1 method sometimes exhibits severe difficulty in obtaining converged results @xcite , owing to its perturbative nature .
another possible choice is to construct effective models from dft that include correlation terms such as hubbard @xmath2 and to solve the models using elaborated methodologies @xcite such as dynamical mean - field theory @xcite .
however , the exact correspondence between the effective models and the first - principles hamiltonian is a nontrivial problem .
recently , wave function theory ( wft ) , which had been mainly applied to molecular systems and established itself as the gold standard in theoretical chemistry @xcite , has become a promising alternative to dft for accurate descriptions of electron correlation in solids @xcite . among the most powerful frameworks in wft
are first - principles quantum monte carlo ( qmc ) methods @xcite , such as the variational monte carlo ( vmc ) , diffusion monte carlo ( dmc ) , auxiliary - field quantum monte carlo ( afqmc ) , and full - configuration - interaction ( fci ) qmc methods .
other kinds of wft , called post - hartree - fock ( post - hf ) methods , have also been applied to condensed matter in recent years @xcite . however , their targets are in most cases limited to solids with small unit cells , owing to their expensive computational cost . in addition , the correlated band structure , which is quite useful in various kinds of theoretical analyses , is not easily obtained in many wfts .
for example , calculation of the band structure in the framework of vmc or dmc requires a large number of single - point calculations of the excited states , which is a clear difference from a mean - field - like approach such as dft , whereby the whole band structure is obtained at once . from this viewpoint , the transcorrelated ( tc )
method @xcite is a fascinating wft that can be applied to solids with reliable accuracy and moderate computational cost @xcite .
the tc method adopts the so - called jastrow ansatz , which is based on a promising strategy often adopted in several wfts such as qmc methods to describe strong electron correlations ; i.e. , the electron - electron distance is included into many - body wave functions .
explicitly correlated electronic - structure theory @xcite in quantum chemistry also adopts this strategy .
in fact , the gutzwiller- and jastrow - correlation factors have often been used to describe strong electron correlation , including the mott physics in systems such as the hubbard model @xcite .
it is also important to note that , unlike several wfts , the whole band structure is obtained at once by solving a one - body self - consistent - field ( scf ) equation in the tc method , as described later in this letter .
moreover , the tc method is deterministic , i.e. , free from the statistical error , unlike the qmc methods .
accurate calculations for the hubbard model @xcite and molecular systems @xcite were also reported using the tc method or other theories that have a close relationship with the tc method . however , insofar as solid - state calculations are concerned , the tc method has so far been applied only to weakly correlated systems . in this letter , we present the first application of the tc method to the band - structure calculation of a @xmath0-electron system : wurtzite zno .
@xmath3 transition metal oxides have posed theoretical challenges for first - principles band - structure calculations , as it is well known that popular approximations such as the local - density approximation ( lda ) fail to provide their band structures accurately , as we shall see later .
we find that the tc method with the bi - orthogonal formulation ( the bitc method ) @xcite successfully reproduces the experimental band structure of zno .
we also clarify how the jastrow factor improves the first - principles description of the correlated electronic states through comparison of the band structures and electron densities among the bitc and other methods .
the central concept of the tc method is to make use of the similarity transformation of the many - body hamiltonian with the jastrow factor @xmath4 , @xmath5 where the correlated wave function is represented as @xmath6 and @xmath7 is called the tc hamiltonian . here
, @xmath8 denotes a pair of space and spin coordinates : @xmath9 . by adopting the so - called slater - jastrow ansatz ,
@xmath10 becomes a slater determinant consisting of one - electron orbitals , @xmath11 : @xmath12 $ ] , and eq . ( [ eq1 ] ) yields an scf equation for one - electron orbitals that experience the effective interaction described with the tc hamiltonian : @xmath13 \notag \\ & - \frac{1}{2}\sum_{j=1}^n \sum_{k=1}^n \int \mathrm{d}x_2 \mathrm{d}x_3\ \phi_j^*(x_2)\phi_k^*(x_3)v_{\mathrm{3body}}(x_1,x_2,x_3 ) \notag \\ & \times \mathrm{det } \left [ \begin{array}{rrr } \phi_i(x_1 ) & \phi_i(x_2 ) & \phi_i(x_3 ) \\
\phi_j(x_1 ) & \phi_j(x_2 ) & \phi_j(x_3 ) \\
\phi_k(x_1 ) & \phi_k(x_2 ) & \phi_k(x_3 ) \end{array } \right ] = \sum_{j=1}^n \epsilon_{ij } \phi_j(x_1 ) , \label{eq : scf}\end{aligned}\ ] ] where @xmath14 , @xmath15 , and @xmath16 are the external potential including the nucleus - electron interaction @xcite and the two- and three - body effective interactions in the tc hamiltonian , defined as @xmath17,\label{eq:2body}\end{aligned}\ ] ] and @xmath18 respectively .
as is evident , the hf method can be regarded as the tc method with @xmath19 .
owing to this effective one - body picture , it is possible to treat the many - body correlation with moderate computational cost .
in addition , one can obtain the band structure of the quasiparticles by using the real part of the eigenvalues of the @xmath20 matrix on the right - hand side of eq .
( [ eq : scf ] ) .
one of the authors proved in prior work @xcite that such a use of the @xmath20 matrix as quasiparticle energies is consistent with koopmans theorem .
we note that one can systematically improve the accuracy of the tc method by utilizing quantum chemical methodologies such as the coupled - cluster and configuration interaction methods to go beyond a single slater determinant ( e.g. refs .
@xcite ) . here , the jastrow function , @xmath21 , is set to the following simple form without adjustable parameters : @xcite @xmath22 where @xmath23 ( @xmath24 is the number of valence electrons in the simulation cell , @xmath25 is the volume of the simulation cell ) and @xmath26 ( spin parallel:@xmath27 ) , @xmath28 ( spin antiparallel:@xmath29 ) .
the long - range asymptotic form of this function describes the screening effect of the electron - electron coulomb interaction @xcite .
the short - range behavior of the exact jastrow function should obey the cusp condition @xcite .
the jastrow ansatz adopted here works well for state - of - the - art qmc methods @xcite .
although our choice of the jastrow function is rather simple , we shall see that , nevertheless , it works well not only for weakly correlated systems @xcite but also for the 3@xmath0-electron system .
of course , it is possible to improve a quality of the many - body wave function by using a complicated jastrow factor , but we adopted this simple trial wave function to realize moderate computational cost . in this study @xcite , we adopted the bitc method @xcite , in which the left one - electron orbitals , @xmath30 in the left slater determinant x @xmath31 $ ] , replace the bra orbitals in the scf equation ( [ eq : scf ] ) , while the ket orbitals remain @xmath11 .
because the tc hamiltonian is non - hermitian , @xmath11 and @xmath30 become different .
we do not show the band structure calculated using the tc method without the bi - orthogonal extension here , because of the large imaginary part of the eigenvalues @xcite .
@xcite ( green circles ) , bitc ( solid line ) , dmc @xcite ( orange circles ) , and hf ( solid line ) methods .
experimental data taken from ref .
@xcite are shown with black dots in the middle figure .
blue broken lines show the other experimental data for the positions of the conduction - band minimum , the o-@xmath32 bottom @xcite , and the zn-@xmath3 peak position @xcite , which might correspond to the averaged position of the zn-@xmath3 bands .
the valence - band maximum energy is set to zero.,width=325 ] figure [ fig:1 ] presents the band structures of zno calculated with the lda , all - electron @xmath33 starting from lda using the lapw @xcite method , bitc , and hf methods .
the characteristic energy values in these band structures , as evaluated at the @xmath34 point , are listed in table [ table : band ] .
table [ table : band ] also lists calculated values with various other methods . by focusing on the gaps between the bands with different characters as listed in table [ table : band ] ,
the bitc band structure exhibits better accuracy than many other conventional methods , including the @xmath33 method starting from lda , and accuracy comparable to that for the most accurate varieties of the @xmath1 scheme , i.e. , the @xmath33 method starting from hse03 and the self - consistent @xmath1 methods such as the qs@xmath1 method @xcite .
it is noteworthy that both the @xmath1 and bitc methods yield such successful results despite being based on conceptually different formulations .
it is important that the bitc method is based on the self - consistent formulation and thus is independent of dft calculations , whereas the @xmath33 method strongly depends upon the unperturbed dft calculations , as seen in table [ table : band ] .
we should also note that , while some methods employ parameters @xmath2 and @xmath35 , which are difficult to determine in the _ ab initio _ way , the bitc method does not use such parameters .
consistency of the calculated band gaps among wfts that use similar trial wave functions ( afqmc , dmc , and bitc ) is also remarkable .
we again stress that the whole band structure is not easily obtained in qmc simulations and requires many single - point excited - state calculations .
we can see that the band structures calculated with the bitc and hf methods are very similar , except for the band gap .
therefore , the main role of the jastrow factor used in this study on the band structure seems to be improvement of the size of the band gap through the screening effect of the electron electron interaction , which is described with the long - ranged asymptotic behavior of the jastrow factor . to obtain a more accurate band structure , e.g. , with respect to the depth of the zn-3@xmath0 bands , more elaborated jastrow factors , as used in qmc studies @xcite , will be necessary .
because the zn-@xmath3 bands are almost flat , a key point might be accurate description of the atomic states , which is an important issue for future investigation . for the @xmath1 method ,
it was pointed out that the shallow zn-@xmath0 bands can strengthen the @xmath36-@xmath0 hybridization , and thus can result in underestimation of the band gap @xcite .
a similar situation might also be realized in the bitc band structure , whereas the zn-@xmath0 bands are rather deep @xcite and so the band gap can be overestimated .
however , the bitc band gap is also affected by the @xmath37 parameter in the jastrow factor , as mentioned in the above comparison between the bitc and hf band structures . because the @xmath37 parameter used in this study
was determined by rpa analysis of the uniform electron gas , it can cause overscreening in the insulator @xcite , thereby decreasing the band gap .
one possibility is that these two factors are canceled here , but more detailed investigation on other materials is also an important future issue @xcite .
figure [ fig:2 ] presents the electron densities calculated with the lda , hf , and bitc methods on the line shown in the crystal structure . the electron density obtained by the bitc method
is defined as @xmath38 re@xmath39 $ ] , where the condition @xmath40 is satisfied due to the bi - orthonormalization condition @xmath41 .
we can see that the electron densities of these methods are almost the same .
however , a slight increase of the electron density at the atomic sites is observed for the bitc method compared with the others .
such a tendency is consistent with the fact that the strong divergence of the electron - electron coulomb repulsion is alleviated for the effective two - body interaction in the similarity - transformed tc hamiltonian , because the jastrow factor satisfies the cusp condition .
more concretely , the @xmath42 and @xmath43 terms in eq .
( [ eq:2body ] ) yield @xmath44 divergence with a different sign than the electron - electron coulomb repulsion .
as can be seen in the proof of the cusp condition @xcite , the true many - body wave function should exhibit deformation described with the two - body degrees of freedom near the electron - electron coalescence point , which can not be represented solely with one - body degrees of freedom .
this is a characteristic advantage of the slater - jastrow - type wave function for the description of localized electronic states , such as in strongly correlated systems .
it is noteworthy that the atomic calculations of the tc method also exhibit a similar tendency for localization @xcite .
finally , we mention the computational effort required for the bitc calculation .
computation takes place on time scales given by @xmath45 , where @xmath46 , @xmath47 , and @xmath48 are the numbers of @xmath49-points , occupied bands , and plane waves , respectively @xcite .
this is the same order as that for the hf or hybrid dft calculations with a prefactor about 20 to 40 @xcite .
the bitc calculation involves neither the frequency index nor the convergence with respect to the number of conduction bands , unlike some perturbative methodologies such as the @xmath1 method .
as can be seen from the fact that the hybrid dft calculations have now been applied to various periodic systems , the computational cost of the bitc method is reasonable for solid - state calculations .
one remaining obstacle for wide application of the bitc method is that we use the norm - conserving pseudopotential with a very high cutoff energy to handle the semicore states at present .
however , this problem is not inherent to the bitc method and can be overcome in principle by the development of a pseudopotential formalism such as the paw method @xcite adapted to the tc method , which is an important future issue @xcite .
we also note that the deformation of the electron density near atomic sites shown in fig .
[ fig:2 ] also implies the importance of careful treatment for the core states .
this can also be a common problem for qmc calculations using the jastrow correlation factor . to conclude , we apply the bi - orthogonal version of the tc method to wurtzite zno and find that it well reproduces the experimental band structure .
our study encourages further investigation of other strongly correlated materials using the bitc method .
a part of the calculations was performed using supercomputers at the supercomputer center , institute of solid state physics , the university of tokyo .
this study was supported by grant - in - aid for young scientists ( b ) ( no .
15k17724 ) from the japan society for the promotion of science , mext element strategy initiative to form core research center , crest , jst , and computational materials science initiative , japan .
999 p. hohenberg and w. kohn , phys . rev .
* 136 * , b864 ( 1964 ) .
w. kohn and l. j. sham , phys . rev . *
140 * , a1133 ( 1965 ) .
l. hedin , phys . rev . * 139 * , a796 ( 1965 ) .
m. s. hybertsen and s. g. louie , phys .
rev .
lett . * 55 * , 1418 ( 1985 ) .
m. s. hybertsen and s. g. louie , phys .
rev .
b * 34 * , 5390 ( 1986 ) .
b .-
c .
shih , y. xue , p. zhang , m. l. cohen , and s. g. louie , phys .
rev .
lett . * 105 * , 146401 ( 2010 ) .
m. imada and t. miyake , j. phys .
. jpn . * 79 * , 112001 ( 2010 ) .
a. georges , g. kotliar , w. krauth , and m. j. rozenberg , rev .
mod .
phys . * 68 * , 13 ( 1996 ) .
g. kotliar , s. y. savrasov , k. haule , v. s. oudovenko , o. parcollet , and c. a. marianetti , rev .
mod .
phys . * 78 * , 865 ( 2006 ) .
k. held , adv .
phys . * 56 * , 829 ( 2007 ) .
a. szabo and n. s. ostlund , _ modern quantum chemistry : introduction to advanced electronic structure theory _ , ( macmillan , new york , 1982 ) . g. h. booth , a. grneis , g. kresse , and a. alavi , nature ( london ) * 493 * , 365 ( 2013 ) . w. m. c. foulkes , l. mitas , r. j. needs , and g. rajagopal , rev
. mod .
phys . * 73 * , 33 ( 2001 ) .
j. koloren and l. mitas , rep .
. phys . * 74 * , 026502 ( 2011 ) .
m. marsman , a. grneis , j. paier , g. kresse , j. chem .
phys . * 130 * , 184103 ( 2009 ) .
a. grneis , m. marsman , g. kresse , j. chem . phys . * 133 * , 074107 ( 2010 ) . j. j. shepherd , a. grneis , phys .
rev .
lett . * 110 * , 226401 ( 2013 ) .
j. j. shepherd , t. m. henderson , g. e. scuseria , phys .
. lett . * 112 * , 133002 ( 2014 ) .
j. j. shepherd , t. m. henderson , g. e. scuseria , j. chem .
phys . * 140 * , 124102 ( 2014 ) .
s. f. boys and n. c. handy , proc .
r. soc .
london ser .
a * 309 * , 209 ( 1969 ) ; _ ibid . _ * 310 * , 43 ( 1969 ) ; _ ibid . _ * 310 * , 63 ( 1969 ) ; _ ibid . _
* 311 * , 309 ( 1969 ) .
n. c. handy , mol .
phys .
* 21 * , 817 ( 1971 ) . s.
ten - no , chem .
. lett . * 330 * , 169 ( 2000 ) ; _ ibid . _ 175 ( 2000 ) .
n. umezawa and s. tsuneyuki , j. chem . phys . *
119 * , 10015 ( 2003 ) .
r. sakuma and s. tsuneyuki , j. phys .
. jpn . * 75 * , 103705 ( 2006 ) .
m. ochi , k. sodeyama , r. sakuma , and s. tsuneyuki , j. chem . phys . * 136 * , 094108 ( 2012 ) .
m. ochi , k. sodeyama , and s. tsuneyuki , j. chem . phys . * 140 * , 074112 ( 2014 ) .
m. ochi , y. yamamoto , r. arita , and s. tsuneyuki , j. chem . phys . * 144 * , 104109 ( 2016 ) .
m. ochi and s. tsuneyuki , j. chem .
theory comput .
* 10 * , 4098 ( 2014 ) .
m. ochi and s. tsuneyuki , chem .
. lett . * 621 * , 177 ( 2015 ) .
w. kutzelnigg , theor .
chim .
acta * 68 * , 445 ( 1985 ) .
h. yokoyama and h. shiba , j. phys .
. jpn . * 56 * , 1490 ( 1987 ) .
m. capello , f. becca , m. fabrizio , s. sorella , and e. tosatti , phys .
rev .
lett . * 94 * , 026406 ( 2005 ) .
h. yokoyama , t. miyagawa , and m. ogata , j. phys .
. jpn . * 80 * , 084607 ( 2011 ) .
t. miyagawa and h. yokoyama , j. phys .
. jpn . * 80 * , 084705 ( 2011 ) .
s. tsuneyuki , prog .
theor .
. suppl . * 176 * , 134 ( 2008 ) .
j. m. wahlen - strothman , c. a. jim ' enez - hoyos , t. m. henderson , and g. e. scuseria , phys . rev .
b * 91 * , 041114(r ) ( 2015 ) . h. luo , w. hackbusch , and h .- j .
flad , mol . phys . * 108 * , 425 ( 2010 ) . h. luo , j. chem . phys . * 133 * , 154109 ( 2010 ) . t. yanai and t. shiozaki , j. chem . phys .
* 136 * , 084107 ( 2012 ) .
o. hino , y. tanimura , and s. ten - no , j. chem . phys . *
115 * , 7865 ( 2001 ) .
note that the external potential @xmath50 does not include the dft potential unlike some other post - kohn - sham treatments .
d. m. ceperley , phys .
rev .
b * 18 * , 3126 ( 1978 ) .
d. m. ceperley and b. j. alder , phys .
rev .
lett .
* 45 * , 566 ( 1980 ) .
d. bohm and d. pines , phys . rev . * 92 * , 609 ( 1953 ) .
t. kato , commun .
pure appl . math . *
10 * , 151 ( 1957 ) .
r. t. pack and w. b. brown , j. chem .
phys .
* 45 * , 556 ( 1966 ) . to fully satisfy the cusp condition for the singlet and triplet spin pairs
, one should adopt the operator representation for the jastrow function @xcite , which introduces the troublesome non - terminating series of the effective interaction in the tc hamiltonian .
thus , we used this approximated cusp condition in the same manner as in many qmc studies @xcite . s. ten - no , j. chem . phys . * 121 * , 117 ( 2004 ) . the following hf and ( bi)tc calculations were performed using the tc++ code @xcite , which makes use of the lda @xcite orbitals obtained using the tapp code @xcite as an initial guess of the hf and ( bi)tc orbitals .
singularities of the electron - electron coulomb repulsion term and the jastrow function in the @xmath49-space were handled using a method proposed by gygi and baldereschi @xcite using an auxiliary function of the same form as ref .
@xcite . for calculations of wurtzite zno
, we used a @xmath51 @xmath49-mesh , a cutoff energy of 361 ry , and 150 bands to expand the subspace for block - davidson diagonalization in the hf and ( bi)tc methods ( @xmath52 with the notation used in ref . [ ] ) .
lattice parameters were extracted from the experiment : @xmath53 , @xmath54 , and @xmath55 @xcite . the same nonlocal norm - conserving pseudopotentials @xcite , where the semicore @xmath56 and @xmath57 orbitals of zinc atoms were treated as valence electrons , were used for our lda , hf , and ( bi)tc calculations .
@xcite j. p. perdew and a. zunger , phys .
rev .
b * 23 * , 5048 ( 1981 ) .
j. yamauchi , m. tsukada , s. watanabe , and o. sugino , phys .
rev .
b * 54 * , 5586 ( 1996 ) .
o. sugino and a. oshiyama , phys .
rev .
lett . * 68 * , 1858 ( 1992 ) .
f. gygi and a. baldereschi , phys .
rev .
b * 34 * , 4405 ( 1986 ) .
s. massidda , m. posternak , and a. baldereschi , phys .
rev .
b * 48 * , 5058 ( 1993 ) .
h. karzel , w. potzel , m. kfferlein , w. schiessl , m. steiner , u. hiller , g. m. kalvius , d. w. mitchell , t. p. das , p. blaha , k. schwarz , and m. p. pasternak , phys .
rev .
b * 53 * , 11425 ( 1996 ) .
n. troullier and j. l. martins , phys .
rev .
b * 43 * , 1993 ( 1991 ) .
see supplemental material for more information .
o. k. andersen , phys .
rev .
b * 12 * , 3060 ( 1975 ) .
s. v. faleev , m. van schilfgaarde , and t. kotani , phys .
rev .
lett . * 93 * , 126406 ( 2004 ) .
m. usuda , n. hamada , t. kotani , and m. van schilfgaarde , phys .
rev .
b * 66 * , 125101 ( 2002 ) .
g. zwicker and k. jacobi , solid state commun . * 54 * , 701 ( 1985 ) . , edited by k. h. hellwege and o. madelung , landolt - brnstein , new series , group iii ( springer , berlin , 1982 ) , vols .
17a and 22a .
l. y. lim , s. lany , y. j. chang , e. rotenberg , a. zunger , and m. f. toney , phys . rev .
b * 86 * , 235113 ( 2012 ) .
a. r. h. preston , b. j. ruck , l. f. j. piper , a. demasi , k. e. smith , a. schleife , f. fuchs , f. bechstedt , j. chai , and s. m. durbin , phys .
rev .
b * 78 * , 155114 ( 2008 ) .
t. kotani , m. van schilfgaarde , and s. v. faleev , phys .
rev .
b * 76 * , 165106 ( 2007 ) .
m. shishkin , m. marsman , and g. kresse , phys .
rev .
lett . * 99 * , 246403 ( 2007 ) .
f. ma , s. zhang , and h. krakauer , new j. phys . *
15 * , 093017 ( 2013 ) .
j. yu , l. k. wagner , and e. ertekin , j. chem .
phys . * 143 * , 224707 ( 2015 ) .
l. ley , r. a. pollak , f. r. mcfeely , s. p. kowalczyk , and d. a. shirley , phys .
rev .
b * 9 * , 600 ( 1974 ) . r. a. powell , w. e. spicer , and j. c. mcmenamin , phys
. rev .
lett .
* 27 * , 97 ( 1971 ) . c. j. vesely , r. l. hengehold , and d. w. langer , phys .
rev .
b * 5 * , 2296 ( 1972 ) . also for weakly correlated systems , we have observed that localized states ( e.g. , the 3@xmath58 states of bulk silicon ) tend to become deeper in their band structures @xcite . in our observation ,
the similar trend is also realized for other semiconductors and insulators with @xmath59 electrons , which include both the ionic and covalent crystals .
refs .
@xcite and @xcite used the lattice parameters that are a bit different from ours .
however , such difference yields a difference in the band gap and the @xmath0-band position for less than 0.1 ev in lda calculations .
therefore , we can conclude that the difference of the lattice parameters among several references has a minor effect on the comparison presented in table 1 .
k. momma and f. izumi , j. appl . crystallogr . * 44 * , 1272 ( 2011 ) .
this scaling is valid only when a single slater determinant is used ( i.e. multi - determinant methodologies are not used ) and the jastrow function is represented as @xmath60 ( i.e. not a general two - body function ) .
the ratio of computational time required for the tc and hf calculations is about 15 to 20 @xcite . as for the bitc method ,
this ratio becomes about 20 to 40 owing to the nonequivalency of the bra and ket orbitals .
p. e. bl " ochl , phys .
rev .
b * 50 * , 17953 ( 1994 ) .
g. kresse and d. joubert , phys .
rev .
b * 59 * , 1758 ( 1999 ) .
our present implementation requires about 20,000 core hours for solving the bitc - scf equation , which is determined by a very large cutoff energy of plane waves .
if we can make the cutoff energy an order of magnitude lower , which is a usual demand for other methods , by development of efficient core treatment , this cost will be reduced also by an order of magnitude . |
in geometric group theory ( and many other fields of mathematics ) , one of the main methods for understanding a group is to look at the spaces it can act on .
( for example , speakers at this conference have discussed actions of groups on @xmath2-hyperbolic spaces , @xmath3 cube complexes , euclidean buildings , and other spaces of geometric interest . ) in these lectures , we consider only very simple spaces , namely , the real line @xmath4 and the circle @xmath5 .
also , we consider only a single , very interesting class of groups , namely , the arithmetic groups .
more precisely , the topic of these lectures is : [ mainques ] let @xmath6 be @xmath7 , or some other arithmetic group . 1
. does there exist a faithful action of @xmath6 on @xmath4 ?
2 . does there exist a faithful action of @xmath6 on @xmath5 ?
all actions are assumed to be continuous , so the questions ask whether there exists a faithful homomorphism @xmath8 , where @xmath9 or @xmath5 .
( recall that a homomorphism is _ faithful _ if its kernel is trivial . ) a fundamental theorem in the subject tells us that the two seemingly different questions in [ mainques ] are actually the same for most arithmetic groups ( if , as is usual in geometric group theory , we ignore the very minor difference between a group and its finite - index subgroups ) : [ ghysfp ] let @xmath10 , or some other irreducible arithmetic group , such that no finite - index subgroup of @xmath6 is isomorphic to a subgroup of @xmath11 . then : @xmath12 ( @xmath13 ) suppose @xmath14 is a finite - index subgroup of @xmath6 that acts on @xmath4 .
then @xmath14 also acts on the one - point compactification of @xmath4 , which is homeomorphic to @xmath5 .
( note that this argument is elementary and very general .
it is the opposite direction of the theorem that requires assumptions on @xmath6 , and sometimes requires passage to a finite - index subgroup . )
( @xmath15 ) suppose @xmath14 is a finite - index subgroup of @xmath6 that acts on @xmath5 . a major theorem proved independently by ghys @xcite and burger - monod @xcite tells us that that the action must have a finite orbit .
( we will say a bit about the proof of this theorem in [ amenlect , bddcoholect ] . )
this means that a finite - index subgroup @xmath16 of @xmath14 has a fixed point in @xmath5 .
let @xmath17 be a point in @xmath5 that is fixed by @xmath16 .
then @xmath18 is a @xmath16-invariant subset , so its complement is also invariant .
this implies that @xmath16 acts on @xmath19 , which is homeomorphic to @xmath4 .
thus , in most cases , it does not matter which of the two versions of [ mainques ] we consider . for now ,
let us look at actions on @xmath4 . to avoid minor complications , we will assume , henceforth , that @xmath20 this means that an action of @xmath6 on @xmath21 is a faithful homomorphism @xmath22 , where @xmath23 is the group of _ orientation - preserving _ homeomorphisms of @xmath21 . since @xmath23 is a subgroup of index @xmath24 in the group of all homeomorphisms , this is just another example of ignoring the difference between a group and its finite - index subgroups .
the expository paper @xcite covers the main topics of these lectures in somewhat more depth .
see @xcite and @xcite for introductions to the general theory of group actions on the circle ( not just actions of arithmetic groups ) , and see @xcite for an introduction to arithmetic groups .
the following result provides an obstruction to the existence of an action on @xmath4 .
[ torsionnoact ] if a group has a nontrivial element of finite order , then the group does not have a faithful action on @xmath4 .
it suffices to show that every nontrivial element @xmath25 of @xmath26 has infinite order . since @xmath25 is nontrivial , there is some @xmath27 , such that @xmath28 .
assume , without loss of generality , that @xmath29 .
the fact that @xmath25 is an orientation - preserving homeomorphism of @xmath4 implies that it is an increasing function : @xmath30 therefore ( letting @xmath31 and @xmath32 ) , we have @xmath33 .
in fact , by induction , we have @xmath34 so @xmath35 for every @xmath36 .
this implies @xmath37 , so @xmath25 is not the identity map .
since @xmath38 is arbitrary , this means that @xmath25 has infinite order .
[ slnznoactbcstorsion ] if @xmath39 , then @xmath40 does not have a faithful action on @xmath4 .
it is easy to find a nontrivial element of finite order in @xmath40 .
for example , the matrix @xmath41 is in @xmath42 and has order @xmath24 .
it is not difficult to show that every arithmetic group has a finite - index subgroup that has no elements of finite order ( * ? ? ?
* lem . 4.19 , p. 232 ) .
this means that [ torsionnoact ] does not provide any obstruction at all to the existence of actions of sufficiently small finite - index subgroups of @xmath6 .
for example : it is well known that some finite - index subgroups of @xmath42 are free groups .
( in fact , every torsion - free subgroup is free ( * ? ? ? * eg . 1.5.3 , p. 11 , and prop . 18 , p. 36 ) . )
any such subgroup has _
many _ faithful actions on @xmath4 : [ freegrpactsonr ] show that every finitely generated free group has a faithful action on @xmath4 . here is a much less trivial class of arithmetic groups that act on @xmath4 : [ arithsl2cacts ] if @xmath6 is any arithmetic subgroup of @xmath43 , then some finite - index subgroup of @xmath6 has a faithful action on @xmath4 .
a very recent and very important theorem of agol @xcite tells us there is a finite - index subgroup @xmath14 of @xmath6 , such that there is a surjective homomorphism @xmath44 . since @xmath45 has an obvious nontrivial action on @xmath4 ( by translations ) , this implies that @xmath14 also acts nontrivially on @xmath4 ( by translations ) .
however , additional effort is required to obtain an action that is faithful . a classic theorem of burns - hale @xcite provides a cohomological condition that implies the existence of a faithful action : @xmath46 ( see [ loexers]([loexers - burnshale ] ) ) .
agol s theorem tells us @xmath47 is nonzero , which establishes the hypothesis for the special case where @xmath48 . by using @xmath49-manifold topology and a fairly simple argument about euler characteristics , a theorem of boyer - rolfsen - wiest ( * ? ? ?
. 3.1 and 3.2 ) promotes this nonvanishing to obtain the condition for all @xmath50 , and thereby yields a faithful action on @xmath4 .
[ egnotact ] we have seen that some finite - index subgroups of @xmath42 have actions on @xmath4 . to obtain arithmetic groups that do _ not _
act on @xmath4 ( even after passing to a finite - index subgroup ) , we need a bigger group . 1
. [ egnotact - sln ] one approach would be to take larger matrices ( not just @xmath51 ) . later in this lecture
, we will see that this works : if @xmath52 , then no finite - index subgroup of @xmath40 has a faithful action on @xmath4 .
[ egnotact - sl2o ] another possible approach would be to keep the same size of matrix , but enlarge the ring of coefficients : instead of only the ordinary ring of integers @xmath45 , consider a ring a algebraic integers @xmath53
. outlines a proof that this approach also works : if @xmath54 is a real algebraic integer that is irrational ( for example , we could take @xmath55 ) , then no finite - index subgroup of @xmath1 \bigr)$ ] acts faithfully on @xmath4 .
in the spirit of [ egnotact ] , it is conjectured that every `` irreducible '' arithmetic group that acts on @xmath4 is contained in a very small lie group , like @xmath43 : [ noactconj ] if @xmath6 is an `` irreducible '' arithmetic group , then @xmath56 unless @xmath6 is an arithmetic subgroup of a `` very small '' lie group . for the interested reader
, the remainder of this makes the conjecture more precise .
however , we will only look at examples of arithmetic groups , not delving deeply into their theory , so , for our purposes , a vague understanding of the conjecture is entirely sufficient . saying that @xmath6 is _ irreducible _ means that no finite - index subgroup of @xmath6 is a direct product @xmath57 ( where @xmath58 and @xmath59 are infinite ) .
the following simple observation shows that the problem reduces to this case .
[ dirprodfaithful ] show that the direct product @xmath57 has a faithful action on @xmath4 if and only if @xmath58 and @xmath59 both have faithful actions on @xmath4 . technically speaking , instead of saying that the lie group is `` very small , '' we should say that it is a semisimple lie group whose `` real rank '' is only @xmath60 . in other words , up to finite index
, it belongs to one of the four following families of groups ( up to local isomorphism ) : * @xmath61 ( the isometry group of hyperbolic @xmath38-space @xmath62 ) , or * @xmath63 ( the isometry group of complex hyperbolic @xmath38-space ) , or * @xmath64 ( the isometry group of quaternionic hyperbolic @xmath38-space ) , or * @xmath65 ( the isometry group of the hyperbolic plane over the octonions , also known as the cayley plane " ) . since @xmath43 is locally isomorphic to @xmath66 , this list does include the examples in [ arithsl2cacts ]
. applies only to actions on @xmath4 , not actions on @xmath5 , because some arithmetic groups of large real rank do act on the circle .
namely , if @xmath67 is a semisimple lie group that has @xmath11 as one of its simple factors , then every arithmetic subgroup of @xmath67 acts on the circle ( by linear - fractional transformations ) .
however , it is conjectured that these are the only such arithmetic groups of large real rank @xcite .
the following translates [ noactconj ] into a purely algebraic question about the existence of a certain structure on the group @xmath6 .
let @xmath6 be a group .
nobreaktrue * a _ total order _ on a set @xmath68 is an transitive , antisymmetric binary relation @xmath69 on @xmath68 , such that , for all @xmath70 , we have @xmath71 * when @xmath69 is a total order on a group @xmath6 , we can ask that the order structure be compatible with the group multiplication : @xmath69 is _ left - invariant _ if , for all @xmath72 , we have @xmath73 see @xcite for more about the theory of left - invariant total orders .
[ actifflo ] let @xmath6 be a countable group .
then @xmath74 ( @xmath13 ) if no nontrivial element of @xmath6 fixes @xmath75 , then we may define @xmath76 and this is a left - invariant total order .
( recall that each element of @xmath6 acts on @xmath4 via in increasing function , so if @xmath77 , then @xmath78 . if @xmath79 , the tie can be broken by choosing some other @xmath27 and comparing @xmath80 with @xmath81 .
( @xmath15 ) note that @xmath6 acts faithfully ( by left translation ) by automorphisms of the ordered set @xmath82 , which is isomorphic ( as an ordered set ) to a subset of @xmath83 .
if it is isomorphic to all of @xmath83 , then @xmath6 acts on the dedekind completion , which is homeomorphic to @xmath4 .
there is actually no loss of generality in assuming that @xmath84 , because @xmath85 can be replaced with a left - invariant ordering of @xmath86 that is order - isomorphic to @xmath87
. therefore , [ noactconj ] can be restated as follows : if @xmath6 is an irreducible arithmetic group , then @xmath88 unless @xmath6 is an arithmetic subgroup of a `` very small '' lie group .
[ prodpos ] suppose @xmath69 is a left - invariant total order on @xmath6 ( and @xmath89 is the identity element of @xmath6 ) . show that if @xmath90 with @xmath91 , then @xmath92 and @xmath93 .
we can now prove [ egnotact]([egnotact - sln ] ) : [ slnznotlo ] if @xmath6 is a finite - index subgroup of @xmath40 , with @xmath52 , then there does not exist a left - invariant total order on @xmath6 .
the proof is based on understanding a certain famous subgroup @xmath94 of @xmath0 : [ heisnotn ] let @xmath94 be the _ discrete heisenberg group _ , which means @xmath95 for convenience , let us also fix names for some particular elements of @xmath94 : @xmath96 ( note that @xmath97 is a generating set for @xmath94 . )
[ heisexers ] nobreaktrue 1 .
[ heisexers - z=[x , y ] ] show @xmath98 \in z(h)$ ] , where * @xmath99 = x^{-1 } y^{-1 } x y$ ] is the _ commutator _ of @xmath100 and @xmath101 , and * @xmath102 is the _ center _ of @xmath94 .
2 . [ heisexers-[xyyk ] ] show @xmath103 for @xmath104 .
[ heisexers - orderable ] ( optional ) show @xmath94 has a left - invariant total order .
+ suppose @xmath69 is a left - invariant total order on a group @xmath6 .
for @xmath90 , we write @xmath105 if @xmath106 is _ infinitely smaller _ than @xmath107 .
i.e. , @xmath108 here is the key fact that will be used in the proof : [ loheis ] if @xmath69 is any left - invariant total order on @xmath94 , then either @xmath109 or @xmath110 . assume , for simplicity , that @xmath111 .
( this actually causes no loss of generality , since there is no harm in replacing some or all of @xmath100 , @xmath101 , and @xmath112 by their inverses .
this is because we can retain the relation @xmath99 = z$ ] by interchanging @xmath100 and @xmath101 if necessary , since @xmath113 = z^{-1}$ ] . ) from [ heisexers]([heisexers-[xyyk ] ] ) , we have @xmath114 note that the exponent of @xmath112 is quadratic in @xmath38 ( and negative ) .
now suppose @xmath115 and @xmath116 .
then there exist @xmath117 , such that @xmath118 and @xmath119 .
therefore @xmath120 so , for all @xmath121 , we have @xmath122 this is a contradiction .
suppose there is a left - invariant total order on @xmath123 .
( for simplicity , we are writing the proof as if @xmath6 is the entire group @xmath0 , and leave it as an exercise for the reader to modify the proof to work for finite - index subgroups . ) in [ heisnotn ] , we gave names to three particular elements of @xmath6 that have a single off - diagonal @xmath60 .
for this proof , we actually want to name all six such elements : we call them @xmath124 , @xmath125 , @xmath126 , @xmath127 , @xmath128 , @xmath129 , where @xmath130 { \lower0.75pt\hbox{\ovalbox{4 } } } & * & { \lower0.75pt\hbox{\ovalbox{3 } } } \\[2pt ] { \lower0.75pt\hbox{\ovalbox{5 } } } & { \lower0.75pt\hbox{\ovalbox{6 } } } & * \end{bmatrix } .\ ] ] thus , for example , @xmath131 , @xmath132 , and @xmath133 , so @xmath134 is the heisenberg group
. actually , there are six copies of the heisenberg group in @xmath6 ( see [ sl3zpfexers]([sl3zpfexers - heis ] ) ) : @xmath135 \bigl\langle { \lower0.75pt\hbox{\ovalbox{4 } } } , { \lower0.75pt\hbox{\ovalbox{5 } } } , { \lower0.75pt\hbox{\ovalbox{6 } } } \bigl\rangle , & \bigl\langle { \lower0.75pt\hbox{\ovalbox{5 } } } , { \lower0.75pt\hbox{\ovalbox{6 } } } , { \lower0.75pt\hbox{\ovalbox{1 } } } \bigl\rangle , & \bigl\langle { \lower0.75pt\hbox{\ovalbox{6}}},{\lower0.75pt\hbox{\ovalbox{1 } } } , { \lower0.75pt\hbox{\ovalbox{2 } } } \bigl\rangle .
\end{matrix } \end{aligned}\ ] ] since @xmath136 is a heisenberg group , [ loheis ] tells us that either @xmath137 or @xmath138 .
assume , without loss of generality , that @xmath139 also , since @xmath140 is also a heisenberg group , [ loheis ] tells us that either @xmath141 or @xmath142 .
however , we know @xmath138 , which implies @xmath143 .
so we must have @xmath144 continuing in this way , using the other heisenberg groups in succession , we have @xmath145 by transitivity , this implies @xmath146 , which is a contradiction .
[ sl3zpfexers ] nobreaktrue 1 .
[ sl3zpfexers - heis ] verify that each of the subgroups listed in ( [ sixheisinsl3 ] ) is isomorphic to the heisenberg group .
more precisely , for @xmath147 , verify that there is an isomorphism @xmath148 , such that @xmath149 2 .
[ sl3zpfexers - finind ] the given proof of [ slnznotlo ] is incomplete , because it assumes that @xmath6 is all of @xmath0 .
( and we already knew from [ slnznoactbcstorsion ] that @xmath0 has no faithful action on @xmath4 . )
modify the proof so it is valid when @xmath6 is a finite - index subgroup of @xmath0 .
nobreaktrue 1 . the proof of [ slnznotlo ] can be adapted to show that finite - index subgroups of the group @xmath150 do not have left - invariant total orders @xcite .
2 . from [ loexers]([loexers - subgrp ] ) , we see that if @xmath6 contains a finite - index subgroup of either @xmath0 or @xmath150 , then @xmath6 does not have a left - invariant total order . in the terminology of arithmetic groups , this exactly means @xcite : @xmath151 3 .
the argument of [ slnznotlo ] relies on the existence of heisenberg groups in @xmath6 , so it does not apply to @xmath1 \bigr)$ ] .
( heisenberg groups are nonabelian nilpotent groups , but every nilpotent subgroup of @xmath43 is abelian . )
we will use a quite different argument to discuss these groups in [ bddgenlect ] .
another important case in which the argument of [ slnznotlo ] can not be applied is when @xmath152 is compact .
this is because every nilpotent subgroup of @xmath6 is virtually abelian .
[ cocpctopen ] find an arithmetic group @xmath6 , such that @xmath152 is compact , and no finite - index subgroup of @xmath6 has a faithful action on @xmath4 .
most large arithmetic groups have kazhdan s property @xmath153 .
( we refer the reader to e.breuillards lectures in this volume for further discussion of property @xmath153 . ) therefore , a negative answer to the following well - known question would be a major advance toward settling [ cocpctopen ] ( and many other interesting cases of [ noactconj ] ) : does there exist an infinite group with kazhdan s property @xmath153 that has a faithful action on @xmath4 or @xmath5 ?
the answer is negative for actions on the circle if we require our actions to act by homeomorphisms that have continuous second derivatives : infinite groups with kazhdan s property @xmath153 do not have faithful , @xmath154 actions on @xmath5 . [ loexers ] a group that has a left - invariant total order is said to be _ left - orderable_. nobreaktrue 1 .
[ loexers - subgrp ] show that every subgroup of a left - orderable group is left - orderable .
[ loexers - abelian ] show torsion - free , _
groups are left - orderable .
[ loexers - extension ] show that if @xmath155 is a normal subgroup of @xmath6 , such that @xmath155 and @xmath156 are left - orderable , then @xmath6 is left - orderable .
[ loexers - nilpotent ] show torsion - free , _ nilpotent _ groups are left - orderable .
[ loexers - solvable ] ( harder ) show that some torsion - free , _ solvable _ group is _ not _ left - orderable .
[ loexers - exps ] show that @xmath6 is left - orderable if and only if , for every finite sequence @xmath157 of nontrivial elements of @xmath6 , there exists @xmath158 , such that the semigroup generated by @xmath159 does not contain @xmath89 .
[ loexers - locally ] show _ locally _ left - orderable @xmath160 left - orderable .
[ loexers - residually ] show _ residually _ left - orderable @xmath160 left - orderable .
[ loexers - burnshale ] ( burns - hale ( * ? ? ?
2 ) ) show that if @xmath161 for every nontrivial , finitely generated subgroup @xmath50 of @xmath6 , then @xmath6 is left - orderable .
[ bddgenlect ] in [ lolect ] , we showed that finite - index subgroups of @xmath162 do not have faithful actions on @xmath4 . in this lecture ,
we prove the same conclusion for appropriate groups of @xmath51 matrices . throughout this lecture
, @xmath54 is a algebraic integer that is real and irrational .
( actually , we do not need to require @xmath54 to be real unless it satisfies a quadratic equation with rational coefficients . )
[ sl2anoact ] if @xmath6 is a finite - index subgroup of @xmath1 \bigr)$ ] , then @xmath6 does not have a faithful action on @xmath4 .
the proof has two ingredients : _ bounded generation _ and _ bounded orbits_. both are with respect to _
unipotent subgroups_. let @xmath163 @xmath164 and @xmath165 @xmath166 .
these are `` unipotent '' subgroups of @xmath1 \bigr)$ ] .
any subgroup of @xmath167 that is conjugate to a subgroup of @xmath168 is said to be _ unipotent _ , but we do not need any unipotent subgroups other than @xmath168 and @xmath169 .
now that we know what unipotent subgroups are , let us see what `` bounded generation '' means .
a basic theorem of undergraduate linear algebra says that every invertible matrix can be reduced to the identity matrix by row operations ( or by column operations , if you prefer those ) .
also , since performing a row operation ( or column operation ) is the same as multiplying by an `` elementary matrix , '' this implies the important fact that every invertible matrix is a product of elementary matrices . in other words , @xmath170 however ,
in your undergraduate course , the scalars were assumed to be in a _ field _ ( probably either @xmath4 or @xmath171 ) , but our matrices have their entries in a ring of integers ( namely , either @xmath45 or @xmath172 $ ] ) , which is not a field .
fortunately , this is not a problem : [ rowreduceeg ] the matrix @xmath173 is a fairly typical element of @xmath42 .
let us see that it can be reduced to the identity matrix , by using only _ integer _ row operations .
more precisely , the only allowable operation is adding an integer ( @xmath45 ) multiple of one row to another row .
( in linear algebra , a few additional operations are usually allowed , such as multiplying a row by a scalar , but we will not permit those operations . ) using @xmath174 to denote applying a row operation , we see that the matrix can indeed be reduced to the identity : @xmath175 here is the general case : [ rowreducesl2z ] every matrix in @xmath176 can be reduced to the identity matrix by integer row operations .
essentially , we apply the euclidean algorithm : * choose the smallest nonzero entry in the first column , and use a row operation to subtract an appropriate integer multiple of it from the other entry in the first column .
now it is the other entry that is the smallest in the first column .
* by repeating this process , we will eventually reach a situation with only one nonzero entry in the first column .
since the determinant is @xmath60 , this entry must be a unit in the ring @xmath45 , which means that the entry is either @xmath60 or @xmath177 . by performing just a few more row operations , we can assume it is @xmath60 , and that it is in the top - left corner .
* now the matrix is upper triangular , with a @xmath60 in the top - left corner . since the determinant is @xmath60
, there must also be a @xmath60 in the bottom right corner .
so one additional row operation yields the identity matrix .
here is another way of saying the same thing : [ uvgensl2z ] @xmath168 and @xmath169 generate @xmath42 .
adding an integer multiple of one row to another row is the same as multiplying on the left by a matrix in either @xmath168 or @xmath169 . in [ rowreduceeg ] , we used only a few ( namely , @xmath178 ) row operations to reduce the matrix to the identity by using the procedure outlined in the proof of [ rowreducesl2z ] .
however , it is easy to construct examples of matrices for which this procedure will use an arbitrarily large number of steps . it would be much better to have a more clever algorithm that can reduce every matrix to the identity in no more than , say , @xmath179 row operations .
unfortunately , this is _ impossible : _ [ reducesl2zunbdd ] for every @xmath180 , there is a matrix in @xmath42 that can not be reduced to the identity with less than @xmath180 integer row operations . here is another way of saying this : [ uvgensl2z ] tells us that every element of @xmath42 can be written as a word in the elements of @xmath168 and @xmath169 .
what [ reducesl2zunbdd ] tells us is that there is no uniform bound on the length of the word : some elements of @xmath40 require a word of length more than a hundred , others require length more than a million , others require length more than a trillion , and so on . in other words ,
if @xmath181 , then [ uvgensl2z ] tells us , for some @xmath38 , there are sequences @xmath182 and @xmath183 , such that @xmath184 however , [ reducesl2zunbdd ] tells us that there is no uniform bound on @xmath38 that is independent of @xmath185 .
that is , although @xmath168 and @xmath169 _ generate _
@xmath42 , they do not _ boundedly _ generate @xmath42 .
here is the official definition : suppose @xmath186 are subgroups of @xmath6 .
we say @xmath186 _ boundedly generate _
@xmath6 if there is some @xmath38 , such that @xmath187 in other words , for every @xmath188 , there exist sequences @xmath189 , such that @xmath190 the key point ( which is what distinguishes this from just saying the subgroups _ generate _
@xmath6 ) is that there is an upper bound on @xmath38 that is independent of @xmath185 .
( then , since @xmath191 is allowed to be the identity element , we can take the same value of @xmath38 for all @xmath185 . )
[ bddgenexers ] when @xmath6 is boundedly generated by cyclic subgroups ( i.e. , @xmath192 , with each @xmath193 cyclic ) , we usually just say @xmath6 is _ boundedly generated_. nobreaktrue 1 .
[ bddgenexers - quotient ] assume @xmath6 is boundedly generated ( by cyclic subgroups ) , and @xmath155 is a normal subgroup of @xmath6 .
show that @xmath156 is boundedly generated ( by cyclic subgroups ) .
[ bddgenexers - modnth ] assume @xmath6 is boundedly generated ( by cyclic subgroups ) , and @xmath121 .
show @xmath194 has finite index in @xmath6 .
[ bddgenexers - finind ] let @xmath14 be a finite - index subgroup of @xmath6 . show that @xmath6 is boundedly generated ( by cyclic subgroups ) if and only if @xmath14 is boundedly generated ( by cyclic subgroups ) 4 .
[ bddgenexers - sl2z ] prove [ reducesl2zunbdd ] .
[ bddgenexers - variablepowers ] ( harder ) assume @xmath6 is boundedly generated ( by cyclic subgroups ) , and @xmath38 is any function from @xmath6 to @xmath195 .
show @xmath196 has finite index in @xmath6 .
although there is no bound on the number of @xmath45 operations needed to reduce a @xmath51 matrix to the identity , we can find a bound if we allow slightly more scalars in our operations : [ sl2obddgen ] the subgroups @xmath168 and @xmath169 boundedly generate @xmath197 } \bigr)$ ] .
in other words , there is some @xmath38 , such that @xmath198 } \bigr ) = ( { \overline{u}}{\underline{v}})^n = { \overline{u}}\ , { \underline{v}}\ , { \overline{u}}\ , { \underline{v}}\ , \cdots\ , { \overline{u}}\ , { \underline{v}}.\ ] ] the proof of the general case of [ sl2obddgen ] is nonconstructive , so it does not provide an explicit bound on @xmath38 . in cases where a bound is known , it depends on @xmath54 and can be arbitrarily large if the algebraic integer @xmath54 is very complicated .
however , it is believed that there should be a uniform bound that is independent of @xmath54 .
in fact , we will see below that if certain number - theoretic conjectures are true , then less than 10 row operations should always suffice .
the known proofs of [ sl2obddgen ] are long and complicated , so we will not try to explain them .
instead , we will give a very short and simple proof that relies on an unproved conjecture in number theory .
let @xmath199 and @xmath200 be nonzero integers .
we say @xmath199 is a _ primitive root _
modulo @xmath200 if @xmath201 @xmath49 is a primitive root modulo @xmath202 , because @xmath203 is a list of all the nonzero residues modulo @xmath202 .
although it is still an open problem , we will assume : let @xmath204 , such that @xmath205 and @xmath199 is not a perfect square .
then there exist infinitely many primes @xmath200 , such that @xmath199 is a primitive root modulo @xmath200 .
although this conjecture is still an open problem , it is implied by a certain generalization of the riemann hypothesis @xcite .
dirichlet proved there are infinitely many primes in any ( appropriate ) arithmetic progression @xcite , and we will assume a stronger form of artin s conjecture that says @xmath200 can be chosen to be in any such arithmetic progression : [ artinarithprogconj ] let @xmath206 , such that * @xmath207 , and * @xmath205 and @xmath199 is not a perfect @xmath208th power for any @xmath209 .
then there exist infinitely many primes @xmath200 in the arithmetic progression @xmath210 , such that @xmath199 is a primitive root modulo @xmath200 .
we assume @xmath199 is not a perfect @xmath208th power in order to avoid the following obstruction : if @xmath209 , @xmath211 , @xmath212 , and @xmath199 is a perfect @xmath208th power , then @xmath199 is not a primitive root modulo any prime in the arithmetic progression @xmath213 . to avoid the need for any algebraic number theory
, we will prove a slight variation of [ sl2obddgen ] that replaces the algebraic number @xmath54 with a rational number , namely , @xmath214 : if @xmath17 is any prime , then every matrix in @xmath215 \bigr)$ ] can be reduced to the identity matrix with a bounded number of @xmath216 $ ] row operations .
let @xmath217 \bigr)$ ] .
assume , for simplicity , that @xmath218 . we explain how to reduce this matrix to the identity with only five row operations : @xmath219 \1 ) letting @xmath220 in [ artinarithprogconj ] , we know there is some @xmath221 , such that @xmath222 is prime , and @xmath17 is a primitive root modulo @xmath200 .
our first row operation adds @xmath223 times the second row to the first row .
\2 ) now , since @xmath17 is a primitive root modulo @xmath200 , we know there exists @xmath224 , such that @xmath225 .
this means there exists @xmath226 , such that @xmath227 .
our second row operation adds @xmath228 times the first row to the second row .
\3 ) every element of @xmath216 $ ] is a multiple of @xmath229 ( since @xmath17 is a unit in this ring ) .
therefore , our third row operation can add anything at all to the top - left entry , so we can change this matrix entry to anything we want .
we change it to a @xmath60 ( by subtracting @xmath230 times the second row from the first row ) .
\4 ) now , the fourth row operation can use the @xmath60 in the top - left corner to kill the bottom - left entry by subtracting @xmath229 times the first row from the second row .
\5 ) since the original matrix @xmath231 is in @xmath215 \bigr)$ ] , and the row operations we applied do not affect the determinant , we know that the determinant of this matrix is @xmath60 . therefore , the bottom - right entry must be @xmath60 .
so the fifth row operation can kill the top - right entry .
[ sl3zbddgen ] if @xmath52 , then every matrix in @xmath40 can be reduced to the identity by using no more than @xmath232 integer row operations .
thus , although the euclidean algorithm will use an unbounded number of row operations , carter and keller showed that the extra freedom provided by larger matrices can be exploited to find a different algorithm that uses only a bounded number .
we have seen that the subgroups @xmath168 and @xmath169 boundedly generate @xmath233 \bigr)$ ] .
the other ingredient in our proof of [ sl2anoact ] is that these subgroups have bounded orbits : [ bddorbs ] let @xmath6 be a finite - index subgroup of @xmath233 \bigr)$ ] . if @xmath6 acts on @xmath4 , then every @xmath168-orbit is a bounded set , and every @xmath169-orbit is a bounded set . before discussing the proof of this theorem , let us explain how it is used : suppose @xmath6 has a nontrivial action on @xmath4 .
( this will lead to a contradiction . )
pretend , for simplicity , that @xmath6 is all of @xmath233 \bigr)$ ] , instead of being a finite - index subgroup .
[ assumenofp ] we may assume the action has no fixed points .
let @xmath234 be the set of fixed points of the @xmath6-action on @xmath4 .
then @xmath234 is obviously a closed set , so its complement is open .
( also , the complement is nonempty , because the @xmath6-action is nontrivial . )
thus , if we let @xmath235 be a connected component of the complement , then @xmath235 is an open interval in @xmath4 .
it is easy to see that @xmath235 is @xmath6-invariant ( because , by definition , the endpoints of @xmath235 are fixed points for @xmath6 ) .
so @xmath6 acts on @xmath235 . by definition , @xmath235
is contained in the _ complement _ of the set of fixed points , so the @xmath6-action on @xmath235 has no fixed points . since @xmath235 is homeomorphic to @xmath4 , this provides an action of @xmath6 on @xmath4 with no fixed points .
[ orbitsbdd ] we show that every @xmath6-orbit on @xmath4 is bounded .
fix some @xmath27 . * from [ bddorbs ] , we know that the @xmath169-orbit of @xmath17 is bounded , so it has a finite infimum @xmath236 and a finite supremum @xmath237 .
thus , @xmath238 is contained in the compact interval @xmath239 $ ] . * from [ bddorbs ]
, we know that @xmath168-orbit of @xmath236 has a finite infimum @xmath240 , and the @xmath168-orbit of @xmath237 has a finite supremum @xmath241 . since every element of @xmath6 acts via an order - preserving homeomorphism of @xmath4 , we know that @xmath242 $ ] . * continuing in this way , we see that @xmath243 is contained in a compact interval @xmath244 $ ] for every @xmath121 .
however , since @xmath168 and @xmath169 boundedly generate @xmath6 ( see [ sl2obddgen ] ) , we know there is some @xmath38 , such that @xmath245 .
therefore , the @xmath6-orbit of @xmath17 is a bounded set .
we obtain a contradiction .
fix some @xmath27 . from [ orbitsbdd
] , we know @xmath246 is a bounded set , so it has a finite supremum @xmath107 . since @xmath246 is a @xmath6-invariant set , and @xmath107 is a point that is defined from this set , we know that the point @xmath107 must be fixed by @xmath6 .
this contradicts [ assumenofp ] , which tells us that there are no fixed points . instead of actually proving [ bddorbs ]
, we will prove a simpler version that replaces @xmath54 with the rational number @xmath214 ( much as in [ bddgensl2asect ] ) : let @xmath247 \bigr)$ ] ( or a finite - index subgroup ) , where @xmath17 is prime .
if @xmath6 acts on @xmath4 , then every @xmath168-orbit is a bounded set , and every @xmath169-orbit is a bounded set .
pretend , for simplicity , that @xmath6 is all of @xmath248 \bigr)$ ] , instead of being a finite - index subgroup . for @xmath249 $ ] ,
let @xmath250 also let @xmath251 a simple calculation shows @xmath252 suppose some @xmath168-orbit is not a bounded set .
then either it is not bounded above , or it is not bounded below .
assume , without loss of generality , that it is not bounded above .
since @xmath169 is conjugate to @xmath168 , this implies that some @xmath169-orbit is also not bounded above .
in fact , it can be shown that there is a single point @xmath253 , such that * both the @xmath168-orbit and the @xmath169-orbit of @xmath100 are not bounded above , and * @xmath254 fixes @xmath100 .
fix @xmath255 with @xmath256 .
since the @xmath169-orbit of @xmath100 is not bounded above , we can choose @xmath257 with @xmath258 .
then , since @xmath254 is order - preserving , we have @xmath259 however , as @xmath260 , we have @xmath261 therefore @xmath262 is less than the finite number @xmath263 .
this is a contradiction .
[ bddgenisom ] assume @xmath6 is boundedly generated ( by cyclic subgroups ) .
show that if @xmath6 acts by _ isometries _ on a metric space @xmath21 , and every cyclic subgroup has a bounded orbit on @xmath21 , then every @xmath6-orbit on @xmath21 is bounded .
each provide examples of arithmetic groups that can not act on the line . in both cases ,
the proof was based on unipotent elements , which means that they only apply to arithmetic groups that are _ not _ cocompact ( cf .
[ cocpctopen ] ) . to complete the treatment of such groups
, it will suffice to consider only a few more examples : if @xmath6 is any noncocompact , irreducible arithmetic group , and @xmath264 , then @xmath6 contains a finite - index subgroup of either @xmath265 \bigr)$ ] ( for some @xmath54 ) or a noncocompact arithmetic subgroup of either @xmath266 or @xmath267 .
combining this with [ sl2anoact ] establishes the following observation : proving the following very special case would establish [ noactconj ] under the additional assumption that @xmath6 is not cocompact .
[ lattsl3noact ] noncocompact arithmetic subgroups of @xmath266 and @xmath267 have no faithful action on @xmath4 .
one possible approach is to use bounded generation : let @xmath6 be a noncocompact arithmetic subgroup of @xmath266 or @xmath267 .
if some finite - index subgroup of @xmath6 is boundedly generated by unipotent subgroups , then @xmath6 does not have a faithful action on @xmath4 .
this implies that [ lattsl3noact ] is a consequence of the following fundamental conjecture in the theory of arithmetic groups : [ rapinchukconj ] if @xmath6 is a noncocompact , irreducible arithmetic group , and @xmath264 , then @xmath6 contains a finite - index subgroup that is boundedly generated by unipotent subgroups . in fact , to establish [ lattsl3noact ] ( and therefore also the entire noncocompact case of [ noactconj ] ) , it would suffice to prove the special case of [ rapinchukconj ] in which @xmath6 is an arithmetic subgroup of either @xmath266 or @xmath267 .
[ amenlect ] amenability is a very fundamental notion in group theory there are literally dozens of different definitions that single out exactly the same class of groups .
we will discuss just a few of these many viewpoints , and , for simplicity , we will restrict our attention to _ discrete _ groups that are countable , ignoring the important applications of this notion in the theory of topological groups .
much more information can be found in the monographs @xcite and @xcite .
let us begin with an amusing example that illustrates one of the many definitions .
[ ponzifreegrp ] consider the free group @xmath268 , and let us assume that every element of the group starts with $ 1 .
thus , if @xmath269 denotes the amount of money possessed by element @xmath185 at time @xmath270 , then @xmath271 now , everyone will pass their dollar to the person next to them who is closer to the identity .
( that is , if @xmath272 is a reduced word , with @xmath273 for each @xmath274 , then @xmath185 passes its dollar to @xmath275 .
the identity element has nowhere to pass its dollar , so it keeps the money it started with . ) then , letting @xmath276 denote the amount of money possessed now ( at time @xmath277 ) , we have @xmath278 thus , everyone has more than doubled their money .
furthermore , this result was achieved by moving the money only a bounded distance .
such an arrangement is called a _ ponzi scheme _ on the group @xmath279 : a _ ponzi scheme _ on a group @xmath6 is a function @xmath280 , such that : 1 .
@xmath281 for all @xmath188 @xmath282 and 2 .
there is a finite subset @xmath283 of @xmath6 , such that @xmath284 for all @xmath188 @xmath285 from [ ponzifreegrp ] , we know there is a ponzi scheme on the free group @xmath279 .
however , not all groups have a ponzi scheme : [ abelnoponzi ] there does not exist a ponzi scheme on the abelian group @xmath286 .
+ more generally , we will see later that no solvable group has a ponzi scheme ( even though solvable groups can have exponential growth ) .
this is because solvable groups are `` amenable , '' and the nonexistence of a ponzi scheme can be taken as the definition of amenability : [ ponziiffnotamenthm ] there exists a ponzi scheme on @xmath6 if and only if @xmath6 is not amenable .
this theorem provides a nice description of what it means for a group to _ not _ be amenable , but it does not directly provide any positive information about a group that _ is _ amenable .
most of the other definitions we discuss are better for that .
instead of using ponzi schemes , we will adopt the following definition : @xmath287 to see what this means , let us consider an example : let @xmath288 , where @xmath289 and @xmath290 .
if we let @xmath234 be a large ball in @xmath6 , then @xmath234 is very close to being invariant under the left - translation by @xmath106 and @xmath107 : @xmath291 where @xmath292 can be as small as we like , if we take @xmath234 to be sufficiently large .
we say that @xmath234 is `` almost invariant : '' let @xmath6 be a group , and fix a finite subset @xmath283 of @xmath6 and some @xmath293 .
a finite , nonempty subset @xmath234 of @xmath6 is _ almost invariant _ if @xmath294 [ amendefn ] @xmath6 is _ amenable _ if and only if @xmath6 has almost - invariant finite subsets ( for all finite @xmath283 and all @xmath293 ) .
[ folnerex ] use [ amendefn ] to show : 1 .
[ folnerex - free ] the free group @xmath279 is _ not _ amenable .
[ folnerex - union ] if @xmath6 is amenable , @xmath283 is a finite subset of @xmath6 , and @xmath293 , then there exists a finite subset @xmath234 of @xmath6 , such that @xmath295 , where @xmath296 3 .
[ folnerex - qi ] amenability is invariant under quasi - isometry .
( this means that you should assume @xmath58 is quasi - isometric to @xmath59 , and prove that @xmath58 is amenable if and only if @xmath59 is amenable . ) 4 .
[ folnerex - noponzi ] amenable groups do not have ponzi schemes .
[ folnerex - ponzi ] nonamenable groups have ponzi schemes .
nobreaktrue 1 .
since the notion of almost invariant " depends on the choice of @xmath283 and @xmath292 , many authors say that @xmath234 is `` @xmath297-invariant . ''
an almost - invariant set can also be called a `` flner set . ''
more precisely , a sequence @xmath298 of nonempty , finite subsets of @xmath6 is said to be a _ flner sequence _ if , for every finite subset @xmath283 of @xmath6 , every @xmath293 , and every sufficiently large @xmath38 , the set @xmath299 is @xmath297-invariant .
( the set @xmath299 is often called a `` flner set . '' ) thus , [ amendefn ] can be restated as saying that @xmath6 is amenable if and only if it has a flner sequence .
in many situations , it is difficult to directly employ the almost - invariant sets provided by [ amendefn ] .
this provides some consequences that are often easier to apply .
for example , every bounded function on @xmath6 has an average value : a _ mean _ on @xmath300 is a linear functional @xmath301 , such that @xmath302 satisfies two axioms that would be expected of the average value of @xmath25 : * _ the average value of a constant function is that constant _ @xmath303 and * _ the average value of a positive - valued function can not be negative _
@xmath304 the mean is _ left - invariant _ if @xmath305 for all @xmath306 and all @xmath188 , where @xmath307 .
[ amenavgval ] @xmath6 is amenable if and only if there exists a left - invariant mean on @xmath300 .
choose a sequence @xmath308 of almost - invariant sets with @xmath309 as @xmath260 , and let @xmath310 that is , @xmath311 is the average value of @xmath25 on the finite set
@xmath299 , so @xmath312 is obviously a mean on @xmath300 .
since the set @xmath299 is almost invariant , the mean @xmath312 is close to being left - invariant . to obtain perfect left - invariance
, we take a limit : @xmath313 where @xmath314 is a subsequence chosen so that the limit exists
. however , if we choose different subsequences for different functions @xmath25 , then the limit may not be linear or left - invariant we need to be consistent in our choice of @xmath302 for all @xmath25 .
this can be accomplished in various ways : * ( logician s approach ) an _ ultrafilter _ on @xmath315 tells us which subsequences are `` good '' and which are `` bad . ''
so the choice of an ultrafilter easily leads to a consistent value for @xmath302 . *
( analyst s approach ) define a linear functional @xmath316 that * * takes the value @xmath60 on the constant function @xmath60 , and * * is @xmath75 on every function of the form @xmath317 .
+ then the _ hahn - banach theorem _ tells us that @xmath316 extends to a linear functional defined on all of @xmath300 . *
( other viewpoints ) use _ zorn s lemma _ , _ tychonoff s theorem _ , or some other version of the _ axiom of choice_. the following consequence is very important in the theory of group actions : [ acthasinvtmeas ] suppose nobreaktrue * @xmath6 is amenable , and * @xmath6 acts on a compact metric space @xmath21 ( by homeomorphisms ) . then there exists a @xmath6-invariant borel probability measure @xmath318 on @xmath21 .
fix a basepoint @xmath319 . for any @xmath320
, we can define a function @xmath321 by restricting @xmath322 to the @xmath6-orbit of @xmath323 .
more precisely , @xmath324 since @xmath322 is continuous and @xmath21 is compact , we know that @xmath322 is bounded . so @xmath325 is also bounded .
therefore , [ amenavgval ] tells us that it has an average value @xmath326 , which we call @xmath327 . since @xmath328 is a mean , it is easy to see that @xmath318 is a positive linear functional of finite norm ( in fact , @xmath329 ) .
so the riesz representation theorem tells us that @xmath318 is a borel measure on @xmath21 .
since @xmath328 is translation - invariant and @xmath330 , we see that @xmath318 is translation - invariant and @xmath331 , so @xmath318 is a translation - invariant probability measure .
the converse is true : if every @xmath6-action on every compact metric space has a @xmath6-invariant probability measure , then @xmath6 is amenable .
so this is another possible choice for the definition of amenability .
will discuss the `` bounded cohomology group '' @xmath332 , which is defined exactly like the usual group cohomology , except that all cochains are required to be bounded functions .
this notion provides another definition of amenability : @xmath6 is amenable if and only if @xmath333 for every @xmath6-module @xmath334 that is the dual of a banach space .
recall that if @xmath6 is a finite group , and @xmath334 is a @xmath6-module ( such that multiplication by the scalar @xmath335 is invertible ) , then one can prove @xmath336 by averaging : for an @xmath38-cocycle @xmath337 , define @xmath338 then @xmath339 is an @xmath340-cochain , and @xmath341 .
so @xmath54 is a coboundary , and is therefore trivial in cohomology .
since @xmath6 is amenable , we can do exactly this kind of averaging for any _ bounded _ cocycle .
see [ hbb(amen;r)=0 ] for more details .
when @xmath6 is amenable , [ amenavgval ] allows us to take the average value of the characteristic function of any subset of @xmath6 .
this leads to vonneumann s original definition of amenability @xcite : [ vonneumann ] @xmath6 is amenable if and only if there exists a finitely additive , translation - invariant probability measure that is defined on all of the subsets of @xmath6 .
more precisely , if we let @xmath342 be the collection of all subsets of @xmath6 , then the conclusion means there is a function @xmath343 $ ] , such that : * @xmath344 if @xmath345 and @xmath346 are disjoint , * @xmath347 , and * @xmath348 for all @xmath188 and @xmath349 .
this definition was motivated by vonneumann s interest in the famous _ banach - tarski paradox_. the subjects are connected via the following notion : a _ paradoxical decomposition _ of @xmath6 is a representation @xmath350 such that , for some @xmath351 , we have @xmath352 [ vonneumannex ] nobreaktrue 1 .
[ vonneumannex - noparadox ] show that if @xmath6 is amenable , then @xmath6 does not have a paradoxical decomposition .
2 . [ vonneumannex - free ] find an explicit paradoxical decomposition of a free group .
[ vonneumannex - paradox ] show that if @xmath6 is not amenable , then @xmath6 has a paradoxical decomposition .
+ nobreaktrue 1 .
vonneumann used the german word `` messbar '' ( which can be translated as `` measurable '' ) , not the currently accepted term `` amenable , '' and his condition was not proved to be equivalent to [ amendefn ] until much later ( by flner @xcite ) .
2 . see @xcite for much more about the banach - tarski paradox , paradoxical decompositions , and the relevance of amenability .
we have now seen several proofs that the existence of almost - invariant sets implies some other notion that is equivalent to amenability . here a proof that goes the other way .
if there is an invariant mean @xmath328 on @xmath300 , then @xmath6 has almost - invariant finite sets .
the dual of @xmath353 is @xmath300 , so @xmath353 is dense in the dual of @xmath300 , in an appropriate weak topology .
hence , there is a sequence @xmath354 , such that @xmath355 . since @xmath328 is invariant , we may choose some large @xmath38 so that @xmath356 is close to being invariant .
then , for an appropriate @xmath357 , the finite set @xmath358 is almost invariant .
much of the following exercise can be proved fairly directly by using almost - invariant sets , but it will be much easier to use other characterizations of amenability for some of the parts .
[ egamenex ] show that all groups of the following types are amenable : 1 .
[ egamenex - finite ] finite groups 2 . [ egamenex - cyclic ] cyclic groups 3 .
[ egamenex - product ] @xmath359 4 .
[ egamenex - abelian ] abelian groups 5 .
[ egamenex - extension ] amenable by amenable 6 .
[ egamenex - solv ] solvable groups 7 .
[ egamenex - subgrp ] subgroups of amenable groups 8 .
[ egamenex - quotient ] quotients of amenable groups 9 .
[ egamenex - locally ] locally amenable groups 10 .
[ egamenex - limit ] direct limits of amenable groups 11 .
[ egamenex - subexp ] groups of subexponential growth [ amenclasses ] nobreaktrue 1 .
[ amenclasses - elem ] from [ egamenex ] , we see that any group obtained from finite groups and abelian groups by repeatedly taking extensions , subgroups , quotients , and direct limits must be amenable .
these `` obvious '' examples of amenable groups are said to be _
elementary amenable_. 2 .
[ amenclasses - grigorchuk ] the so - called `` grigorchuk group '' is an example of a group with subexponential growth that is not elementary amenable @xcite .
[ amenclasses - basilica ] a group is said to be _ subexponentially amenable _ if it can be constructed from groups of subexponential growth by repeated application of extensions , subgroups , quotients , and direct limits .
the `` basilica group '' is an example of an amenable group that is not subexponentially amenable @xcite .
thus , the following obvious inclusions are proper : @xmath360 \{\text{abelian}\ } \subsetneq\{\text{solvable}\ } \subsetneq \end{matrix } \left\ { \begin{matrix } \text{elementary } \\
\text{amenable } \end{matrix } \right\ } \subsetneq \left\ { \begin{matrix } \text{subexponentially } \\
\text{amenable } \end{matrix } \right\ } \subsetneq \left\ { \begin{matrix } \text{amenable } \end{matrix } \right\ } .\ ] ] by combining [ folnerex]([folnerex - free ] ) with [ egamenex]([egamenex - subgrp ] ) , we see that if @xmath6 has a nonabelian free subgroup , then @xmath6 is not amenable .
the converse is often called the `` vonneumann conjecture , '' but it was shown to be false in 1980 when olshanskii proved that the `` tarski monster '' is not amenable .
this is a group in which every element has finite order , so it certainly does not contain free subgroups @xcite .
n.monod @xcite has recently constructed counterexamples that are much less complicated . in the theory of topological groups ,
it is _ not _ true that every subgroup of an amenable group is amenable only the _ closed _ subgroups need to be amenable .
in particular , many amenable topological groups contain nonabelian free subgroups ( but such subgroups can not be closed ) .
we are discussing amenability in these lectures because it plays a key role in the proofs of ghys @xcite and burger - monod @xcite that large arithmetic groups must always have a finite orbit when they act on the circle ( cf .
[ ghysfp ] ) . here is a much simpler example of the connection between amenability and finite orbits : [ amenoncircle ] suppose * @xmath6 is amenable , and * @xmath6 acts on @xmath5 ( by orientation - preserving homeomorphisms )
. then either 1 .
the abelianization of @xmath6 is infinite , or 2 .
[ amenoncircle - finorb ] the action has a finite orbit . from [ acthasinvtmeas ]
, we know there is a @xmath6-invariant probability measure @xmath318 on @xmath5 .
assume @xmath318 has an atom .
this assumption means there exists some point @xmath361 that has positive measure : @xmath362 . since @xmath318 is @xmath6-invariant , every point in the orbit of @xmath17 must have the same measure .
however , since @xmath318 is a probability measure , we know that the sum of the measures of these points is finite
. therefore the orbit of @xmath17 must be finite ( since the sum of infinitely many copies of the same positive number is infinite ) .
assume @xmath318 has no atoms .
assume , for simplicity , that the support of @xmath318 is all of @xmath5 .
( that is , no nonempty open interval has measure @xmath75 . )
then the assumption of this case implies that , after a continuous change of coordinates , the measure @xmath318 is simply the lebesgue measure on @xmath5 .
since @xmath6 preserves this measure ( and is orientation - preserving ) , this implies that @xmath6 acts on the circle by rotations .
since the group of rotations is abelian , we conclude that the abelianization of @xmath6 is infinite .
( or else the image of @xmath6 in the rotation group is finite , which means that every orbit is finite . ) if we assume that @xmath6 is infinite and finitely generated , then the conclusion of the can be strengthened : it can be shown that the abelianization of @xmath6 is infinite @xcite , so there is no need for alternative ( [ amenoncircle - finorb ] ) .
large arithmetic groups always have finite abelianization , so it might seem that the theorem of ghys and burger - monod could be obtained directly from [ amenoncircle ] . unfortunately , that is not possible , because arithmetic groups are not amenable ( since they contain free subgroups ) . instead , ghys s proof is based on the following more sophisticated observations : [ amenonconvex ] suppose * @xmath6 is amenable , * @xmath6 acts by ( continuous ) linear maps on a locally convex vector space @xmath334 , and * @xmath363 is a nonempty , compact , convex , @xmath6-invariant subset of @xmath334 .
then @xmath6 has a fixed point in @xmath363 .
@xmath6 acts on the compact set @xmath363 by homeomorphisms , so [ acthasinvtmeas ] provides a @xmath6-invariant probability measure @xmath318 on @xmath363 .
let @xmath17 be the center of mass of @xmath318 .
then @xmath17 is fixed by @xmath6 , since @xmath318 is @xmath6-invariant . also , since @xmath363 is convex , we know @xmath364 .
has a converse : if @xmath6 has a fixed point in every nonempty , compact , convex , @xmath6-invariant set , then @xmath6 is amenable .
so this fixed - point property provides yet another possible definition of amenability .
[ furstenberglemma ] suppose * @xmath6 is an arithmetic subgroup of @xmath365 , * @xmath6 acts on @xmath5 ( by homeomorphisms ) , * @xmath366 $ \begin{bmatrix } * & * & * \\ & * & * \\ & & * \end{bmatrix}$ } \subset g$ ] , and * @xmath367 , with the natural weak topology .
then there exists a @xmath6-equivariant measurable function @xmath368 .
let @xmath369 ( where functions that differ only on a set of measure @xmath75 are identified ) .
it is easy to see that @xmath370 is convex .
then , since the banach - alaoglu theorem tells us that weak@xmath371-closed , convex , bounded sets are compact , we see that @xmath370 is compact in an appropriate weak topology .
also , @xmath372 acts continuously on @xmath370 , via @xmath373 .
we know that solvable groups are amenable ( see [ egamenex]([egamenex - solv ] ) ) .
although we have only been considering discrete groups , the same is true for topological groups in general .
so @xmath372 is amenable ( because it is solvable ) .
therefore ( a generalization of ) [ amenonconvex ] tells us that @xmath372 has a fixed point .
this means there is a @xmath6-equivariant map @xmath374 , such that @xmath375 ( a.e . ) . ignoring a minor issue about sets of measure @xmath75
, this implies that @xmath376 factors through to a well - defined @xmath6-equivariant function @xmath377 .
we omit the proof of the main step in ghys s argument : [ ghysconstant ] the function @xmath378 provided by [ furstenberglemma ] is constant ( a.e . ) . from this
, it is easy to complete the proof : if @xmath6 is any arithmetic subgroup of @xmath266 , then every action of @xmath6 on the circle has a finite orbit . from [ ghysconstant ]
, we know there is a constant function @xmath379 that is @xmath6-equivariant ( a.e . ) . *
since @xmath378 is constant , its range is a single point @xmath318 ( a.e . ) .
* since @xmath378 is @xmath6-equivariant , its range is a @xmath6-invariant set .
so @xmath318 is @xmath6-invariant . since @xmath380 , then the proof of [ amenoncircle ] shows that either 1 .
the abelianization of @xmath6 is infinite , or 2 .
the action has a finite orbit .
since the abelianization of every arithmetic subgroup of @xmath266 is finite , we conclude that there is a finite orbit , as desired .
see @xcite for a nice exposition of ghys s proof for the special case of lattices in @xmath381 .
( a slightly modified proof of the general case that reduces the amount of case - by - case analysis is in @xcite . ) a quite different ( and very interesting ) version of the proof is in @xcite . [ bddcoholect ] m.burger and n.monod @xcite developed a sophisticated machinery to calculate bounded cohomology groups , and used it to prove that actions of arithmetic groups on the circle have finite orbits .
( see @xcite for an exposition . )
we will discuss only some elementary aspects of bounded cohomology , and describe how it is related to actions on the circle , without explaining the fundamental contributions of burger - monod .
see @xcite for a more comprehensive introduction to bounded cohomology and its applications .
( almost all of the information in this lecture can be found there . )
the widespread interest in this subject was inspired by a paper of gromov @xcite .
for a discrete group @xmath6 , the cohomology group @xmath382 is defined as follows . *
any function @xmath383 is an _
@xmath38-cochain _ , and the set of these cochains is denoted @xmath384 . * a certain _ coboundary operator _
@xmath385 is defined .
here are the definitions for the smallest values of @xmath38 : @xmath386 * then @xmath387 ( note that , for simplicity , we take the coefficients to be @xmath4 , not a general @xmath6-module . ) the _ bounded cohomology _ group @xmath388 is defined in exactly the same way as @xmath389 , except that all cochains are required to be _ bounded _ functions .
@xmath390 and @xmath391 are very easy to compute : * it is easy to check that @xmath392 * the same calculation shows that @xmath393 is the set of _ bounded _ constants .
then , since it is obvious that every constant is a bounded function , we have @xmath394 . *
it is easy to check that @xmath395 . *
the same calculation shows that @xmath391 is the set of _ bounded _ homomorphisms into @xmath4 . since a homomorphism into @xmath4 can never be bounded ( unless it is trivial ) , this means @xmath396 .
thus , @xmath390 and @xmath391 give no information at all about @xmath6 .
( one of them is always @xmath4 , and the other is always @xmath397 . )
so @xmath388 is only interesting when @xmath39 .
these groups are not easy to calculate : [ bddcoho(f2 ) ] for the free group @xmath279 , we have : @xmath398 find some countable group @xmath6 , such that you can calculate @xmath388 for all @xmath38 ( and @xmath399 for some @xmath38 ) .
bounded cohomology is easy to calculate for amenable groups : [ hbb(amen;r)=0 ] if @xmath6 is amenable , then @xmath400 for all @xmath38 . from [ amenavgval ]
, we know there is a left - invariant mean @xmath401 any element of @xmath388 is represented by a bounded function @xmath383 , such that @xmath402 . to simplify the notation ,
let us assume @xmath403 .
for each @xmath188 , we can define a bounded function @xmath404 by @xmath405 .
then , by defining @xmath406 , we have @xmath407 .
now , for @xmath408 , we have @xmath409 applying @xmath328 to both sides ( considered as functions of @xmath100 ) , and recalling that @xmath328 is left - invariant , we obtain @xmath410 so @xmath411 .
therefore @xmath412 = 0 $ ] in @xmath413 .
nobreaktrue 1 .
the bounded cohomology @xmath414 of a topological space @xmath21 is defined by stipulating that a cochain in @xmath415 is _ bounded _ if it is a bounded function on the space of singular @xmath38-simplices .
( brooks @xcite , gromov @xcite ) @xmath416 .
3 . forgetting that the cochains are bounded yields a comparison homomorphism @xmath417 .
it is very interesting to find situations in which this map is an isomorphism .
( thurston ) if @xmath418 is a closed manifold of negative curvature , then the comparison map @xmath419 is _ surjective _ for @xmath420 .
however , it can fail to be injective .
if @xmath421 is a homomorphism , then , for each @xmath188 , covering - space theory tells us that @xmath422 can be lifted to a homeomorphism @xmath423 of the universal cover , which is @xmath4 .
however , the lift depends on the choice of a basepoint , so it is not unique lifts can differ by an element of @xmath424 . specifically ,
if @xmath425 is another lift of @xmath422 , then @xmath426 therefore , for any @xmath427 , there exists @xmath428 , such that @xmath429 because @xmath430 and @xmath431 are two lifts of @xmath432 .
it is easy to verify that @xmath180 is a @xmath24-cocycle : @xmath433 and that choosing a different lift @xmath423 only changes @xmath180 by a coboundary .
therefore , @xmath180 determines a well - defined cohomology class @xmath434 , which is called the _ euler class _ of @xmath435 .
[ eulertrivial ] show that the euler class of @xmath435 is trivial if and only if @xmath435 lifts to a homomorphism @xmath436 .
the euler class can also be defined more naturally , by noting that if we let @xmath437 be the set consisting of all possible lifts of all elements of @xmath438 , then we have a short exact sequence @xmath439 with @xmath45 in the center of @xmath437 .
any such central extension is determined by a well - defined cohomology class @xmath440 , and the euler class is obtained using the homomorphism @xmath435 to pull this class back to @xmath6 .
[ eulerisbdd ] choose a basepoint in @xmath4 ( say , @xmath75 ) , and assume the lift @xmath423 is chosen with @xmath441 for all @xmath188 .
show @xmath180 is bounded .
although we only defined bounded cohomology with real coefficients , the same definition can be applied with @xmath45 in place of @xmath4 . therefore , if we choose @xmath180 as in [ eulerisbdd ] , then it represents a bounded cohomology class @xmath412 \in { h_b}^2(\gamma ; { \mathbb{z}})$ ] , which is called the _ bounded euler class _ of the action .
it can be shown that the bounded euler class is a well - defined invariant of the action ( independent of the choice of basepoint , etc . ) .
[ bddeulerclassfp ] the bounded euler class is trivial if and only if @xmath6 has a fixed point in @xmath5 .
( @xmath15 ) we may assume the fixed point is the basepoint @xmath442 .
then we may choose @xmath423 with @xmath443 .
so @xmath444 for all @xmath445 . ( @xmath13 )
we have @xmath446 for some bounded @xmath447 .
letting @xmath448 , we have * @xmath449 , so @xmath450 is a lift of @xmath6 to @xmath451 , and * @xmath452 . hence , the @xmath453-orbit of @xmath75 is a bounded set , so it has a supremum in @xmath4 .
this supremum is a fixed point of @xmath453 , so its image in @xmath5 is a fixed point of @xmath6 .
if @xmath454 , then every orientation - preserving action of @xmath6 on @xmath5 has a fixed point .
the following result is easier to apply , because it uses real coefficients for the cohomology , instead of integers : [ h2rvanishfp ] if * @xmath455 , * @xmath456 , and * @xmath6 is finitely generated , then every orientation - preserving action of @xmath6 on @xmath5 has a finite orbit . the short exact sequence @xmath457 yields a long exact sequence of bounded cohomology : @xmath458 by assumption , the group at the right end is @xmath75 , so the map on the left is surjective .
therefore , the bounded euler class is the coboundary of some ( bounded ) @xmath60-cocycle @xmath459 .
i.e. , @xmath54 is a homomorphism to @xmath460 . since @xmath456 ( and @xmath6
is finitely generated ) , we know @xmath54 is trivial on some finite - index subgroup @xmath461 of @xmath6 .
then the bounded euler class @xmath462 is trivial on @xmath461 , so [ bddeulerclassfp ] tells us that @xmath461 has a fixed point @xmath17 .
since @xmath461 has finite index , we see that the @xmath6-orbit of @xmath17 is finite . if @xmath6 is any arithmetic subgroup of @xmath381 , with @xmath52 , then every action of @xmath6 on @xmath5 has a finite orbit .
burger and monod showed ( in a much more general setting ) that the comparison map @xmath463 is injective .
since it is known that @xmath464 ( if @xmath38 is sufficiently large ) , we conclude that @xmath455 .
the other hypotheses of [ h2rvanishfp ] are well known to be true .
to calculate @xmath413 , we would like to understand the kernel of the comparison map @xmath463 . for this
, we introduce some notation : nobreaktrue * a function @xmath465 is a : * * _ quasimorphism _ if @xmath466 is bounded ( as a function of @xmath467 ) ; * * _ near homomorphism _ if it is within a bounded distance of a homomorphism .
* we use @xmath468 and @xmath469 to denote the space of quasimorphisms and the space of near homomorphisms , respectively . note that @xmath470 .
[ kernelcomparison ] the kernel of the comparison map @xmath463 is @xmath471 let @xmath180 be a bounded @xmath24-cocycle , such that @xmath180 is trivial in @xmath472
. then @xmath473 , for some @xmath465 .
thus , for all @xmath427 , we have @xmath474 so @xmath54 is a quasimorphism .
this establishes that @xmath468 maps onto the kernel of the comparison map , via @xmath475 .
now suppose @xmath476 , such that @xmath462 is trivial in @xmath413 .
the triviality of @xmath462 means there is a bounded function @xmath477 , such that @xmath478 .
then @xmath479 , so @xmath480 is a homomorphism .
since @xmath180 is bounded , this means that @xmath54 is within a bounded distance of a homomorphism ; i.e. , @xmath481 .
we can construct many quasimorphisms on the free group @xmath279 : * as a warm - up , recall that there is an obvious homomorphism @xmath482 , defined by letting @xmath483 be the ( signed ) number of occurrences of @xmath106 in the reduced representation of @xmath100 .
for example , @xmath484 there is an analogous homomorphism @xmath485 , and every homomorphism @xmath486 is a linear combination of these two . * similarly , for any nontrivial reduced word @xmath487 , we can let @xmath488 be the ( signed ) number of disjoint occurrences of @xmath487 in the reduced representation of @xmath100 .
for example , @xmath489 this is a quasimorphism .
[ wordquasi ] verify that @xmath490 is a quasimorphism , for any reduced word @xmath487 . with these quasimorphisms in hand ,
it is now easy to prove a fact that was mentioned in [ bddcoho(f2 ) ] : [ h2b(free ) ] show that @xmath491 is infinite - dimensional .
[ quasimexers ] nobreaktrue 1 .
[ quasimexers - kernelfd ] show that if @xmath6 is boundedly generated ( by cyclic subgroups ) , then the kernel of the comparison map @xmath463 is finite - dimensional .
[ quasimexers - f2notbddgen ] show the free group @xmath279 is not boundedly generated ( by cyclic subgroups ) .
[ quasimexers - commutator ] show that every quasimorphism is bounded on the set of commutators @xmath492 .
[ quasimexers - amenable ] show that if @xmath6 is amenable group , and the abelianization of @xmath6 is finite , then @xmath6 does not have unbounded quasimorphisms .
[ quasimexers - sl3znoquasi ] show that @xmath0 has no unbounded quasimorphisms .
[ [ freegrpactsonr ] ] [ freegrpactsonr ] + + + + + + + + + + + + + + + + choose any faithful action of @xmath279 on the circle .
( for example , use the _ ping - pong lemma _
2.3.9 ) , which provides a sufficient condition for two homeomorphisms to generate a free group , or note that @xmath279 is a subgroup of @xmath494 , which acts faithfully on the circle by linear - fractional transformations . )
since @xmath279 is free , we can lift this to an action on the line ( simply by choosing any lift of the two generators @xmath106 and @xmath107 ) .
since it projects down to a faithful action on the circle , this action on the line must also be faithful . since any open interval is homeomorphic to @xmath4 , we may let @xmath58 act on the open interval @xmath495 ( fixing all points in the complement ) , and let @xmath59 act on the open interval @xmath496 ( fixing all points in the complement ) .
these actions commute , so they define an action of @xmath57 .
since @xmath98 $ ] , we have @xmath498 . by induction on @xmath223
( and using the fact that @xmath112 commutes with @xmath100 ) , then @xmath499 . by induction on @xmath500 ( and using the fact that @xmath112 commutes with @xmath101 ) , then @xmath501 for @xmath502 . either calculate that @xmath504 = \ovalbox{$k$}$ ] , and that @xmath505 commutes with both @xmath506 and @xmath507 , or observe that some permutation matrix conjugates the ordered triple @xmath508 to @xmath509 .
let @xmath511 and @xmath512 be left - invariant total orders on @xmath155 and @xmath156 , respectively .
then we define @xmath513 \text{$gn = hn$ \ and \
$ h^{-1 } g \prec _ * e$ } .
\end{matrix}\ ] ] the left - invariance of @xmath511 and @xmath512 implies the left - invariance of @xmath69 .
recall that the _ ascending central series _
@xmath514 is defined inductively by @xmath515 .
fix @xmath516 and let @xmath517 . for any nontrivial @xmath518
, there exists @xmath188 , such that @xmath519 $ ] is a nontrivial element of @xmath520 , which is torsion - free ( by induction ) . since @xmath521 = [ \overline{g } , z]^n$ ]
, this implies that @xmath522 is torsion - free .
it is also abelian , so we can apply [ loexers]([loexers - abelian ] ) and [ loexers]([loexers - extension ] ) .
here is the general philosophy .
assume @xmath67 is a nontrivial , finitely generated , left - orderable group .
if @xmath67 is solvable ( or , more generally , if @xmath67 is `` amenable '' ) , then it can be shown that the abelianization of @xmath67 is infinite @xcite .
so any torsion - free solvable group with finite abelianization provides an example .
( @xmath15 ) the condition implies it is possible to choose a semigroup @xmath372 in @xmath6 , such that @xmath526 and , for every nonidentity element @xmath185 of @xmath6 , either @xmath527 or @xmath528 .
define @xmath529 .
details can be found in ( * ? ? ?
* thm . 3.1.1 , p. 45 ) .
let @xmath157 be nontrivial elements of @xmath6 .
for each @xmath274 , there is a left - orderable group @xmath193 , and a homomorphism @xmath530 , such that @xmath531 . for the resulting homomorphism @xmath25 into @xmath532
, we have @xmath533 for all @xmath274
. now apply [ loexers]([loexers - exps ] ) .
given nontrivial elements @xmath157 of @xmath6 , the assumption provides a nontrivial homomorphism @xmath534 .
assume @xmath435 is trivial on @xmath535 , and nontrivial on the rest .
by induction on @xmath38 , we can choose @xmath536 , such that the semigroup generated by @xmath537 does not contain @xmath89 .
for @xmath538 , choose @xmath523 so that @xmath539 . then the semigroup generated by @xmath157 does not contain @xmath89 , so [ loexers]([loexers - exps ] ) applies .
( details can be found on page 50 of @xcite . )
call the subgroup @xmath155 .
then @xmath155 is a normal subgroup , so [ bddgenexers]([bddgenexers - quotient ] ) tells us that @xmath156 is boundedly generated by cyclic groups . however , every element of @xmath156 has finite order , so all of these cyclic groups are finite .
( @xmath15 ) for a normal subgroup @xmath155 of @xmath6 , it is easy to see that if @xmath155 and @xmath156 are boundedly generated , then @xmath6 is boundedly generated . also note that finite groups are ( obviously ) boundedly generated .
( @xmath13 ) to present the main idea with a minimum of notation , let us assume @xmath543 is the product of just two cyclic groups .
let @xmath544 , and let @xmath545 be a set of coset representatives for @xmath546 in @xmath547 .
there exists a finite - index subgroup @xmath548 of @xmath94 , such that the conjugate @xmath549 is contained in @xmath14 for every @xmath550 .
let @xmath551 be a set of coset representatives for @xmath548 in @xmath94 .
given @xmath552 , we may choose @xmath553 and @xmath554 , such that @xmath555 therefore , if we let @xmath556 be a list of the elements in @xmath557 , then @xmath558 let @xmath6 be a free subgroup of finite index in @xmath42 .
then [ quasimexers]([quasimexers - f2notbddgen ] ) tells us that @xmath6 is not boundedly generated , so [ bddgenexers]([bddgenexers - finind ] ) implies that @xmath42 also is not boundedly generated by cyclic groups .
since @xmath168 and @xmath169 are cyclic , this completes the proof .
the argument is somewhat similar to [ bddgenexers]([bddgenexers - finind])(@xmath13 ) .
let @xmath14 be the subgroup that is under consideration , and let us assume , for simplicity , that @xmath559 is the product of just two cyclic groups .
let @xmath544 , and let @xmath545 be a set of coset representatives for @xmath546 in @xmath547 .
the key point is to observe that , by definition , @xmath14 contains a finite - index subgroup of each @xmath560 , so we may choose a finite - index subgroup @xmath548 of @xmath94 , such that @xmath549 is contained in @xmath14 for every @xmath550 .
let @xmath551 be a set of coset representatives for @xmath548 in @xmath94 .
for @xmath188 , we have @xmath561 therefore , @xmath562 contains a set of coset representatives , so the index of @xmath14 is at most @xmath563 .
the triangle inequality implies that _ every _ orbit of every cyclic group is bounded .
now , for any @xmath564 , any @xmath565 , and any cyclic subgroup @xmath193 of @xmath6 , this implies there exists @xmath566 , such that @xmath567 is contained in the ball @xmath568 . by the triangle inequality
, we have @xmath569 . by induction , if @xmath540 , then @xmath570 .
suppose @xmath418 is a ponzi scheme on @xmath6 , and @xmath284 for all @xmath185 .
let @xmath223 be the maximum word length of an element of @xmath283 .
then , since @xmath418 is ( at least ) 2-to-1 , we know that the ball of radius @xmath571 has at least twice as many elements as the ball of radius @xmath199 .
so @xmath6 has exponential growth .
assume , without loss of generality , that ( at least ) @xmath572 of the elements of @xmath234 do _ not _ start with @xmath573 .
then @xmath572 of @xmath574 starts with @xmath106 and @xmath572 of @xmath575 starts with @xmath107 .
if @xmath234 is almost invariant , this implies that almost half of @xmath234 starts with both @xmath106 and @xmath107 .
suppose @xmath6 is not amenable .
then there exist @xmath283 and @xmath292 , such that @xmath581 , for every finite subset @xmath234 of @xmath6 . choosing @xmath38 large enough that @xmath582 , then we have @xmath583 .
suppose @xmath418 is a ponzi scheme on @xmath6 , and @xmath284 for all @xmath185 . from [ folnerex]([folnerex - union ] ) ( and replacing @xmath234 with @xmath584 to convert left - translations into right - translations ) , we know there is a finite set @xmath234 , such that @xmath585 .
this is impossible , since @xmath418 is ( at least ) 2-to-1 and @xmath586 . in the special case
where @xmath587 for all @xmath274 , the hint is known as _ hall s marriage theorem _ , and can be found in many combinatorics textbooks .
the general case is proved similarly .
since there are only finitely many possibilities for each set @xmath588 , a standard diagonalization argument shows that the result is also valid for an infinite sequence @xmath589 of finite sets and an infinite sequence @xmath590 . from the hint to [ folnerex]([folnerex - qi ] ) , there exists @xmath591 , such that @xmath592 for every finite @xmath593 . for @xmath594 ,
let @xmath595 and @xmath596 . then there exists @xmath597 , such that @xmath598 and the sets @xmath599 are pairwise disjoint .
define @xmath600 for all @xmath601 .
let @xmath602 where @xmath318 is a finitely additive , translation - invariant probability measure on @xmath6 .
then , since @xmath603 are pairwise disjoint , we have @xmath604 on the other hand , since @xmath605 , we have @xmath606 and , similarly , @xmath607 .
this is a contradiction .
let @xmath608 , @xmath609 , @xmath610 , and @xmath611 be the reduced words that start with @xmath106 , @xmath573 , @xmath107 , or @xmath612 , respectively .
( also add @xmath89 to one of these sets . )
then @xmath613 let @xmath418 be a ponzi scheme , and choose @xmath283 such that @xmath614 .
let @xmath328 contain a single element of @xmath615 , for every @xmath616 , and let @xmath617 be the complement of @xmath328 .
then , for @xmath618 , let @xmath619 and @xmath620 . by construction ,
these sets are pairwise disjoint , and we have @xmath621 .
it is perhaps easiest to apply [ amenonconvex ] . since @xmath155 is amenable , it has fixed points in @xmath363 .
the set @xmath625 of such fixed points is a closed , convex subset . also , it is @xmath6-invariant ( because @xmath155 is normal ) .
so @xmath6 acts on @xmath625 . since @xmath155 is trivial on this set , the action factors through to @xmath156 , which must have a fixed point
this is a fixed point for @xmath6 .
this is another case that is difficult to do with almost - invariant sets . instead ,
note that if there is a ponzi scheme on some subgroup of @xmath6 , then it could be reproduced on all of the cosets , to obtain a ponzi scheme on all of @xmath6 .
this establishes the contrapositive .
this is immediate from [ acthasinvtmeas ] ( or [ amenonconvex ] ) , because any action of @xmath156 is also an action of @xmath6 .
it also follows easily from [ vonneumann ] , since any subset of @xmath156 pulls back to a subset of @xmath6 .
given @xmath283 and @xmath292 , let @xmath94 be the subgroup generated by @xmath283 , so @xmath94 is finitely generated .
if @xmath94 is amenable , then it contains an almost - invariant set , which is also an @xmath297-invariant set in @xmath67 .
in fact @xmath637 never differs by more than @xmath60 from @xmath638 .
there is a difference only if some occurrence of @xmath487 ( or @xmath639 ) overlaps the boundary between @xmath100 and @xmath101 , and there can not be two such occurrences that are disjoint .
write @xmath641 .
any quasimorphism on @xmath6 is determined , up to bounded error , by its restriction to the cyclic subgroups @xmath642 . also , it is not difficult to show that every quasimorphism @xmath643 is a near homomorphism .
( or this can be deduced from [ hbb(amen;r)=0 ] and [ kernelcomparison ] . )
so the restriction of @xmath25 to each @xmath193 is a near homomorphism .
since the homomorphisms from @xmath45 to @xmath4 form a one - dimensional space , we conclude that the dimension of @xmath644 is at most @xmath199 .
let @xmath647 be a quasimorphism . from [ hbb(amen;r)=0 ] and [ kernelcomparison ]
, we see that @xmath25 is within bounded distance of a homomorphism . however , since the abelianization of @xmath6 is finite , there are no nontrivial homomorphisms @xmath648
. therefore @xmath25 is bounded .
every elementary matrix is a commutator ( recall that @xmath649 = z^k$ ] ) , so [ quasimexers]([quasimexers - commutator ] ) implies that @xmath25 is bounded on the set of elementary matrices .
since every element of @xmath0 is the product of a bounded number of these elementary matrices ( see [ sl3zbddgen ] ) , a simple estimate shows that @xmath25 is bounded .
.ghys : groupes dhomomorphismes du cercle et cohomologie borne , in : _ the lefschetz centennial conference , part iii ( mexico city , 1984 ) . _ _ contemp .
_ , vol .
58 , part iii .
soc . , providence , 1987 , pp .
isbn : 0 - 8218 - 5064 - 4 r.grigorchuk and i.pak : groups of intermediate growth : an introduction , _ enseign .
( 2 ) 54 ( 2008 ) , no .
34 , 251272 .
m.gromov : volume and bounded cohomology . _ publ .
ihes _ 56 ( 1982 ) 599 .
v.m.kopytov and n.ya.medvedev : _ right - ordered groups .
_ plenum , new york , 1996 .
isbn : 0 - 306 - 11060 - 1 l.lifschitz and d.w.morris : bounded generation and lattices that can not act on the line , _ pure appl .
q. _ 4 ( 2008 ) , no .
1 , part 2 , 99126 .
n.monod : _ continuous bounded cohomology of locally compact groups .
_ springer , berlin , 2001 .
isbn : 3 - 540 - 42054 - 1 n.monod : an invitation to bounded cohomology , in : _ proc .
congress math . , madrid , spain , 2006 _ , vol . 2 , pp .
http://www.mathunion.org/icm/icm2006.2/main/icm2006.2.1183.1212.ocr.pdf d.w.morris : amenable groups that act on the line , _ algebr .
_ 6 ( 2006 ) , 25092518 .
d.w.morris : bounded generation of @xmath651 ( after d. carter , g. keller and e. paige ) , _ new york j. math . _ 13 ( 2007 ) 383421 .
http://nyjm.albany.edu/j/2007/13-17.html a.yu.olshanskii and m.sapir : non - amenable finitely presented torsion - by - cyclic groups , _ electron .
_ 7 ( 2001 ) , 6371 .
http://dx.doi.org/10.1090/s1079-6762-01-00095-6 [ http://dx.doi.org/10.1090/s1079-6762-01-00095-6 ]
a.l.t.paterson : _
amenability_. american mathematical society , providence , ri , 1988 .
isbn : 0 - 8218 - 1529 - 6 j.p.pier : _ amenable locally compact groups_. wiley , new york , 1984 .
isbn : 0 - 471 - 89390 - 0 v.platonov and a.rapinchuk : _ algebraic groups and number theory , _ academic press , new york , 1994 .
isbn : 0 - 12 - 558180 - 7 .
s. wagon : _ the banach - tarski paradox_. cambridge u. press , cambridge , 1993 .
isbn : 0 - 521 - 45704 - 1 d.witte : arithmetic groups of higher @xmath652-rank can not act on @xmath60-manifolds , _ proc .
_ 122 ( 1994 ) 333340 . | the group @xmath0 can not act ( faithfully ) on the circle ( by homeomorphisms ) .
we will see that many other arithmetic groups also can not act on the circle
. the discussion will involve several important topics in group theory , such as ordered groups , amenability , bounded generation , and bounded cohomology .
provides an introduction to the subject , and uses the theory of left - orderable groups to prove that @xmath0 does not act on the circle . discusses bounded generation , and proves that groups of the form @xmath1 \bigr)$ ] do not act on the real line .
are brief introductions to amenable groups and bounded cohomology , respectively .
they also explain how these ideas can be used to prove that actions on the circle have finite orbits .
an appendix provides hints or references for all of the exercises .
these notes are slightly expanded from talks given at the park city mathematics institute s graduate summer school in july 2012 .
the author is grateful to the pcmi staff for their hospitality , the organizers for the invitation to take part in such an excellent conference , and the students for their energetic participation and helpful comments that made the course so rewarding ( and improved these notes ) .
[ lolect ] |
let @xmath0 denote the set of rooted plane trees of size @xmath1 .
simply generated trees are families of trees obtained by assigning weights @xmath2 to the elements @xmath3 using a degree function @xmath4 ( see @xcite ) .
basically , the weight @xmath2 of some @xmath5 is obtained by multiplying the factors @xmath6 over the nodes @xmath7 of @xmath8 , where @xmath9 denotes the outdegree of @xmath7 .
our main topic is the study of generating functions @xmath10 associated with multiplicative functions @xmath11 defined recursively by using a sequence of real numbers @xmath12 .
we call such multiplicative functions _ bare green functions _ : @xmath13 represents the sum of the feynman amptitudes associated to the relevant diagrams of size @xmath1 in some field theory , and the generating function is then a part of the perturbative expansion of the solution of some equation describing the system ( see @xcite ) . in section [ runge ] ,
we give an equation satisfied by @xmath14 when the weights @xmath15 come from some master function @xmath16 , with @xmath17 , @xmath18 .
we use series indexed by trees , the so - called b - series , as defined in @xcite , to show in theorem [ differentialsys ] that @xmath14 solves @xmath19 where @xmath20 is the differential operator @xmath21 .
@xcite considers a similar problem for _ additive tree functionals _ @xmath22 defined on varieties of increasing trees , like @xmath23 .
assuming some constraints on the degree function @xmath24 , it is proven that the exponential generating function @xmath25 is given by the formula @xmath26 where @xmath27 and @xmath28 solves @xmath29 .
we also consider a central functional called the _ tree factorial _
, denoted by @xmath30 in the sequel , which is relevant in various fields , like algorithmics @xcite , stochastics @xcite , numerical analysis ( see for example @xcite ) , and physics @xcite .
we focus on its negative powers @xmath31 , @xmath32 , which do not admit a master function when @xmath33 .
@xcite solved the case @xmath34 by using the so - called butcher s group ( see for example @xcite ) .
we provide in theorem [ inversion ] a differential equation for the associated generating function , @xmath35 . in section [ wignerprocesses ]
, we define special multiplicative functionals for which the weights @xmath15 are related to the covariance function @xmath36 of some gaussian process , as @xmath37 , for some positive constant @xmath38 .
we show that the generating function @xmath14 is related to the mean normalized trace of products of large symmetric random matrices having independent and indentically distributed versions of the process as entries .
theorem [ t.wig ] gives then a differential equation for the evolution of the trace of a stationary wigner processes .
it follows that most of the examples given in @xcite can be expressed in terms of traces of large random matrices . in section [ triangular ] ,
we show how b - series can be useful for studying traces of triangular operators appearing in free probability .
a rooted tree @xmath39 is a triple @xmath40 such that i ) @xmath41 is a non - empty directed tree with node set @xmath42 and edge set @xmath43 , ii ) all edges are directed away from the root @xmath44 .
the set of rooted trees of order @xmath1 is denoted by @xmath45 , and the set of rooted trees is @xmath46 .
a rooted plane tree @xmath5 is a quadruple @xmath47 satisfying i ) and ii ) and iii ) @xmath48 is a collection of @xmath49 linear orders . given @xmath50 , let @xmath51 be the set of children of @xmath7 .
@xmath52 is the outdegree of @xmath7 .
a rooted planar tree can be seen in the plane with the root in the lowest position , such that the orders @xmath53 coincide with the left - right order .
next consider the partial ordering @xmath54 defined by @xmath55 if and only if @xmath56 lies on the path linking @xmath36 and @xmath7 .
given @xmath50 and @xmath57 let @xmath58 be the subtree of @xmath8 rooted at @xmath7 spanned by the subset @xmath59 .
a rooted labelled tree is a quadruple @xmath60 satisfying i ) and ii ) , with a labelling @xmath61:=\{1,\cdots,\vert v\vert\}$ ] such that @xmath62 when @xmath63 .
the set of rooted labelled trees of order @xmath1 is denoted by @xmath64 .
let @xmath65 .
this family is a special variety of increasing trees , as defined in @xcite .
we next assign weights to the elements of @xmath0 , the set of rooted planar trees of order @xmath1 : the resulting family of trees is said to be simply generated ( see @xcite ) . given a sequence @xmath66 of real numbers with @xmath67 ,
define recursively the weight @xmath2 of @xmath68 as @xmath69 where @xmath70 are the @xmath71 subtrees of @xmath8 rooted at @xmath72 .
let @xmath73 be the generating function of the weight sequence @xmath74 .
our favourite example is @xmath75 , with @xmath76 , @xmath77 ( see @xcite for various interesting choices ) .
we will be concerned with functionals @xmath11 , where @xmath78 , called _ bare green functions_. this terminology is taken from quantum field theory where bare green functions occur during the action of the renormalization group ( see for example @xcite , 4.2 or @xcite ,
let @xmath79 denote the set of bare green functions .
any element @xmath80 is given through a sequence of functions @xmath81 , @xmath82 , which are usually laurent series in some variable @xmath83 ( see for example @xcite ) . in
what follows , we simply write the sequence as @xmath12 . [ bare ] the bare green function @xmath80 , @xmath11 , associated with the sequence of functions @xmath84 is defined recursively as @xmath85 where @xmath70 are the @xmath71 subtrees of @xmath8 rooted at @xmath72 , and where @xmath86 denotes the number of nodes of @xmath8 .
notice that the value of @xmath87 at @xmath5 does not depend on the linear orders and is independent of the labellings .
when dealing with rooted trees , we will adopt the notation @xmath88 for the operation of grafting the rooted trees @xmath70 , that is by considering the tree @xmath8 obtained by the creation of a new node @xmath36 ( the root ) and then joining the roots of @xmath89 to @xmath36 .
bare green functions appeared also in the probabilistic literature in specific situations .
the basic example , in algorithmics @xcite , in numerical analysis ( see @xcite ) , in stochastics @xcite and in physics ( see for example @xcite ) is the _ tree factorial _ , defined by [ factorial ] let @xmath39 with @xmath90 .
then the tree factorial @xmath8 ! is the functional @xmath80 defined by @xmath91 , associated with the sequence @xmath92 given by @xmath93 .
[ add1 ] it should be pointed out that the functional acting on trees , given as @xmath23 , for @xmath80 with @xmath94 , @xmath95 , is an inductive map or an additive tree functional , as defined in @xcite .
interestingly , @xmath96 is used in @xcite to define a probability measure on random search binary trees , and @xcite provide precise asymptotics for @xmath97 .
we first give some basic results on tree factorials , symmetry factors , and generating functions associated with bare green functions .
[ factors ] let @xmath98 .
then @xmath99 is the number of rooted labelled trees @xmath100 of shape @xmath39 , where the shape of a labelled tree @xmath101 is @xmath102 , @xmath103 is the number of rooted plane trees of shape @xmath8 , and @xmath104 is the symmetry factor of the tree , to be defined later
. moreover , let @xmath105 be the weight function associated with elements of @xmath106 , with weights given by @xmath107 .
notice that @xmath108 is the connes - moscovici weight in quantum field theory ( see @xcite ) .
the symmetry factor satisfies the recursive definition : @xmath109 @xmath110 where the indices @xmath111 means that @xmath8 is obtained by grafting @xmath112 times the tree @xmath113 , and so on , where we assume that the @xmath114 are all different as rooted trees .
[ identities ] let @xmath39 .
then @xmath115 and @xmath116 _ proof : _ ( [ syma ] ) is well known ( see for example @xcite ) .
suppose that @xmath39 is such that @xmath117 , the grafting of @xmath112 times the tree @xmath113 , and so on , where we set that the trees @xmath70 are different as rooted trees
. then @xmath118 using the recursive definition of @xmath2 and the definition of @xmath119 , we have @xmath120 therefore @xmath121 and the results follows from the recursive definition of the symmetry factor
. then @xmath122 where we have used ( [ symb ] ) of lemma [ identities ] .
consider the generating function @xmath123 given @xmath39 , the ratio @xmath124 is associated with the weight sequence @xmath125 ; using the expansion @xmath126 , we see that @xmath127 .
consider the _ elementary differentials _ @xmath128 ( see section [ runge ] ) defined by [ elementary ] @xmath129 when @xmath130 , where * denotes the tree of a single node .
for a map @xmath131 , a formal power series of the form @xmath132 is called a b - series @xcite . [ add2 ] when @xmath133 , the series @xmath14 is given by @xmath134 set @xmath135 , @xmath136 , and consider the degree function @xmath137 .
following @xcite , @xmath14 solves @xmath138 ( see also @xcite ) .
we shall see in the next section that it is a natural consequence of b - series expansions of solutions of ordinary differential equations .
consider a dynamical system on @xmath139 @xmath140 for some smooth @xmath141 .
the solution of this dynamical system has a b - series expansion of the form @xmath142 where the elementary differentials @xmath128 is defined recursively by @xmath143 when @xmath90 .
these kinds of expansions have been treated in great detail in @xcite and @xcite and developped independently in combinatorics ( see for example @xcite ) .
suppose that @xmath144 for simplicity .
butcher considered what happens with numerical approximations of the exact solution , the runge - kutta methods , which are themselves b - series @xcite ; here we focus on specific b - series , which are associated to bare green functions .
let @xmath80 be such that there exists a power series @xmath145 with @xmath146 bare green functions satisfying ( [ choice ] ) are used in practical situations in quantum field theory ( see @xcite , 4 and @xcite ,
consider euler s operator @xmath147 , with @xmath148 , for any polynomial @xmath149 , and consider the formal operator @xmath150 acting on monomials as @xmath151 given a power series @xmath152 converging for @xmath153 , we can define @xmath154 , which converges for @xmath153 when the sequence @xmath155 grows subexponentially .
we will not focus on convergence questions here , and work at the formal level .
let @xmath87 be a bare green function with weights @xmath156 , such that ( [ choice ] ) holds for some power series @xmath157 .
it should be pointed out that @xcite deal with the master function @xmath157 , but do not give explicitely an equation for @xmath14 .
the next theorem provides an equation ; its proof uses explicitely b - series .
[ differentialsys ] the formal power series @xmath158 solves @xmath159 .
_ proof : _ @xmath160 for given @xmath161 , set @xmath90 .
then @xmath162 , @xmath163 , and @xmath164 .
the associated term becomes @xmath165 next , every rooted tree @xmath39 can be decomposed uniquely as @xmath166 @xmath167 , meaning that @xmath8 is obtained by grafting @xmath112 times @xmath113 and so on , where the @xmath114 are different as rooted trees , with @xmath168 . collecting the terms associated with @xmath8 , we get the contribution @xmath169 where * means that the sum is taken over all the collections @xmath170 such that @xmath171 .
the above sum reduced then to @xmath172 where we use the first identity of lemma [ identities ] . using the recursive definition of the symmetry factor @xmath173
, we obtain @xmath174 we thus get that the contribution associated with @xmath175 is given by @xmath176 therefore @xmath177 [ add3 ] as we have observed in remark [ add1 ] , @xmath23 is an inductive map when the weights @xmath15 are positive .
it turns out that the exponential generating function associated with @xmath178 can be given as an integral transform for varieties of increasing trees ( see for example section [ s.intro ] ) .
this is the topic of @xcite .
[ gencat ] when @xmath179 , with @xmath180 , and @xmath75 , one has @xmath181 , the catalan number of order @xmath1 , with @xmath182 .
then @xmath183 is solution of the differential equation @xmath184 , that is @xmath185 .
the unique solution with @xmath186 satisfies @xmath187 , or @xmath188 , corresponding to a well known result .
@xcite , 5.3 , considers the case where @xmath189 , which is not of the form given in ( [ choice ] ) : in this situation , @xmath190 , with @xmath191 .
the solution is obtained by using the stucture of the so - called butcher s group of b - series ( that is series of the form ( [ bseries ] ) , where the group structure in given in @xcite ) by tensoring known b - series : [ m2 ] consider the bare functional given by @xmath192 , with @xmath193 . following brouder , the associated b - series , as given in ( [ bseries ] ) , is solution of the second order differential equation @xmath194 when @xmath195 , the solution is given by @xmath196 giving @xmath197 we study the general moment problem @xmath198 , @xmath199 , by working directly on a suitable differential equation as follows : the operator @xmath150 takes the form @xmath200 .
assume that the differential operator @xmath150 is invertible .
then the formal systems becomes @xmath201 consider again the second moment problem for the tree factorial , with @xmath192 and @xmath202 .
choose @xmath157 such that @xmath191 ; the inverse operator might be equal to @xmath203 , and , if this is the case , @xmath204 with @xmath205 , see example [ m2 ] .
more generally , if one considers the moment of order @xmath206 of the inverse tree factorial , the choice @xmath207 should give @xmath208 .
our result , theorem [ inversion ] below shows that the formalism of inversion is correct in term of power series .
this result sheds some light and extends the computations done in @xcite for the second moment , and its proof avoids computations in the butcher s group . [ inversion ] the b - series @xmath209 associated with the moment of order @xmath210 of the inverse tree factorial satisfies the differential equation @xmath211 _ proof : _ let @xmath212 then @xmath213 for given @xmath161 , set @xmath90 .
then @xmath162 , @xmath163 , and @xmath214 .
the associated term becomes @xmath215 proceeding as in the proof of theorem [ differentialsys ] , we get that the contribution associated with @xmath175 is given by @xmath216 on the other hand , @xmath217 in the next section , we show that traces of certain products of wigner matrices ( see for example @xcite ) provide natural examples of bare green functions .
[ d.goe ] the @xmath218-dimensional random matrices @xmath219 are called wigner matrices of variance @xmath220 if the following holds . *
each @xmath221 is symmetric , that is , @xmath222 . * for @xmath223 , the random variables @xmath224 are independent and centered . * for @xmath225 , @xmath226 .
* for any @xmath227 , @xmath228 , where @xmath229 is independent of @xmath223 .
[ d.goep ] the sequence @xmath230 of @xmath218-dimensional random matrices , indexed by @xmath231 , is called a wigner process of variance @xmath220 and correlation function @xmath36 , @xmath232 , @xmath233 and @xmath234 if the following holds . *
each @xmath235 is a wigner matrix of variance @xmath220 in the sense of definition [ d.goe ] . * for @xmath223 , each process
@xmath236 is independent of the others . * for @xmath225 , the process @xmath236 is @xmath36-correlated , that is , for any @xmath237
, @xmath238 a wigner process is stationary when @xmath36 is such that @xmath239 .
let @xmath240 be a sequence of random diagonal matrices , with independent and identically distributed entries of law @xmath241 , having finite moment @xmath242 , @xmath231 , with @xmath243 .
let @xmath244 and set @xmath245 * involutions , dyck paths and rooted plane trees * for @xmath231 , @xmath246:=\{1,2,\ldots , k\}$ ] , @xmath247 is the set of the involutions of @xmath246 $ ] with no fixed point , @xmath248 is the subset of @xmath247 of the involutions @xmath173 with no crossing .
this means that the configurations @xmath249 do not appear in @xmath250 .
let @xmath251 denote the fact that @xmath252 .
let @xmath253 be the set of the dyck paths of length @xmath254 , that is , of the sequences @xmath255 of nonnegative integers such that @xmath256 , @xmath257 , @xmath258 $ ] .
thus , exactly @xmath259 indices @xmath260 $ ] correspond to ascending steps @xmath261 , that is , to steps when @xmath262 .
we denote this by @xmath263 .
the @xmath259 others indices correspond to descending steps , that is , to steps when @xmath264 , and we denote this by @xmath265 .
we make use of bijections between @xmath253 and @xmath266 @xcite .
if @xmath267 , @xmath268 is an involution which maps each element of @xmath269 to a smaller element of @xmath270 .
thus , @xmath271 . more specifically , if @xmath265 , @xmath272 is the greatest @xmath273 such that @xmath274 . finally , the set @xmath275 is in bijection with @xmath276 , the set of rooted plane trees on @xmath277 nodes , where the bijection is given by the walk on the tree from the right to the left ( see for example @xcite ) .
let @xmath278 denote the involution of @xmath279 corresponding to @xmath5 .
given @xmath280 , consider the walk on @xmath8 from the right to the left : every edge @xmath281 with @xmath282 , is crossed at some instant @xmath283 $ ] as @xmath284 and at a later time @xmath285 $ ] as @xmath286 .
clearly , @xmath287 , where @xmath288 is the subtree of @xmath8 rooted at node @xmath289 , that is the subgraph of @xmath8 induced by the nodes @xmath56 with @xmath290 .
@xmath278 is such that @xmath291 and vice versa .
bijections between @xmath276 , @xmath275 and @xmath292 ( @xmath293 and @xmath294 ) .
[ specialbare ] assume that the covariance @xmath36 is such that @xmath295 . then , the functional @xmath296 given by the weights @xmath297 is such that @xmath298 _ proof : _ let @xmath5 .
let @xmath299 be the instants where the oriented edges @xmath284 and @xmath286 , @xmath282 , are crossed during the walk on the tree .
@xmath300 , and thus @xmath301 . finally , @xmath302 , as required . as we have just seen , every wigner process with covariance @xmath36 such that @xmath295 produces a bare green function @xmath296 .
the converse is not true , that is , there exists @xmath80 such that @xmath87 is not of the form @xmath303 for some covariance function @xmath36 .
set @xmath304 .
let @xmath305 be the generating function of the weight sequence @xmath306 , and let @xmath307 , @xmath5 be the associated weight function .
[ t.wig ] let @xmath308 be a stationary wigner process of covariance function @xmath36 and variance @xmath220 , and let @xmath240 be a sequence of random diagonal matrices , independent of the wigner process , with i.i.d .
entries @xmath309 of law @xmath241 , with @xmath310 and finite moments @xmath311 , @xmath136 . then @xmath312 and @xmath313 , @xmath314 .
assume that the covariance is such that there exists a power series @xmath315 with @xmath316 , @xmath136 .
then the formal power series @xmath317 solves @xmath318 moreover @xmath319 [ treefact ] let @xmath96 .
if a tree @xmath8 has @xmath1 nodes and @xmath320 edges , then the requirement @xmath321 is satisfied iff @xmath322 , that is @xmath36 must be such that @xmath323 , @xmath324 . by construction , @xmath325 and therefore @xmath326 .
@xmath327 is positive definite , which implies that @xmath96 is element of @xmath328 .
next , from theorem [ inversion ] , the generating function @xmath329 is solution of the system @xmath330 .
assume that @xmath241 is the point mass @xmath331 , that is each matrix @xmath240 is the identity matrix of size @xmath218 , with @xmath332 .
the solution of the system is @xmath333 , where @xmath334 is the series given in example [ gencat ] .
on the other hand , proposition [ specialbare ] and theorem [ t.wig ] show that @xmath335 .
therefore the limiting mean normalized trace @xmath336 of the product of correlated random matrices @xmath337 is such that @xmath338 , where @xmath339 denotes a normal n(0,1 ) random variable .
[ gencat2 ] consider as in example [ gencat ] the special case where @xmath179 .
the associated inductive parameter ( see remark [ add3 ] ) is the tree size .
the covariance @xmath36 is constant with @xmath340 , and @xmath341 .
then the generating function @xmath14 is solution of the fixed point equation @xmath342 ( either by theorem [ differentialsys ] or by ( [ fondlemma ] ) ) .
notice that in this situation , @xmath343 , and thus @xmath344 describes the mean normalized moment of the spectral measure of the random matrix @xmath345 .
this example can be extended by considering @xmath346 , for some @xmath347 .
when @xmath240 is the identity matrix , @xmath209 is related to the rogers - ramanujan continued fraction @xcite , and corresponds to the generating function associated with path length , see @xcite .
_ proof of theorem [ t.wig ] : _ the first part is a generalization of theorem 1 of @xcite .
set @xmath348 , and @xmath349 .
the mean normalized trace adds the contributions @xmath350 , for paths @xmath351 , with @xmath352 $ ] and @xmath353 .
the @xmath354 are centered , so that any edge @xmath355 appearing once appears at least twice .
given @xmath356 , define @xmath357 and @xmath358 when @xmath359 , and @xmath360 otherwise , and consider the walk @xmath361 defined by @xmath362 , with @xmath363 .
the support of @xmath356 is @xmath364 , of size @xmath365 , with @xmath366 .
the contribution @xmath367 is independent of the labels @xmath368 ; they are @xmath369 labellings giving the same walk @xmath370 , with the same contribution .
thus , the normalization @xmath371 shows that the only walks surviving in the large @xmath218 limit are those with @xmath372 .
this shows that @xmath373 when @xmath259 is odd .
concerning @xmath374 , @xmath375 means that every edge occuring in the path occurs exactly twice , in opposite directions .
@xmath370 is a dyck path of @xmath275 ; let @xmath5 be the associated rooted plane tree , with involution @xmath278 .
using the right to left walk on @xmath8 and the independence of the random variables , the contribution @xmath367 of any path leading to @xmath370 or @xmath8 is @xmath376 where @xmath377 . from proposition [ specialbare ] ,
one obtains @xmath378 , with @xmath379 , as required .
( [ fondlemma ] ) is a consequence of the multiplicative form of bare green functions and of lemma 1.9 , chap .
iii.1 of @xcite .
these results show that the elements of @xmath328 appear naturally in the computation of normalized traces of products of large random matrices ( see for example @xcite ) . in the next section
we illustrate b - series by considering triangular operators from free probability .
let @xmath380 be an algebra and @xmath381 be a @xmath380 bi - module .
let @xmath382 be a bilinear map .
we follow @xcite by defining the product @xmath383 , @xmath384 , and setting @xmath385 @xmath386 @xmath387 let @xmath388 be an involution of @xmath389 $ ] without fixed point and without crossing . given a word @xmath390 in @xmath391 , @xmath173 induces parentheses on @xmath392 , and the preceedings rules permit the evaluation of this parenthized word .
this extends to a map @xmath393 on @xmath394 .
sniady defines such maps to prove a conjecture of dykema and haagerup on generalized circular elements .
let @xmath395 be an operator valued probability space , that is @xmath391 is a unital * -algebra , @xmath396 an unital * -subalgebra and @xmath397 be a conditional expectation ( linear , @xmath398 , and @xmath399 , @xmath400 ) . the triangular operator @xmath407 of dykema and haagerup is obtained from @xmath408 $ ]
, the * -algebra of complex polynomials of one variable by setting @xmath409(x)=\int_x^1 b(s){\rm d}s,\ ] ] @xmath410(x)=\int_0^x b(s){\rm d}s,\ ] ] @xmath411(x)=[\kappa(t^*,bt^*)](x)=0.\ ] ] @xmath407 is the limit for the convegence of * -moments of large upper triangular random matrices @xmath412 ( @xcite ) .
define a trace @xmath413 as ( see @xcite ) @xmath414 in what follows , we use p - series ( where p stands for partitioned differential systems , see @xcite ) .
we follow @xcite , and adapt his notations to p - series . given
some function @xmath74 , and two kernels @xmath415}$ ] and @xmath416}$ ] , consider the iterated integrals @xmath417 and @xmath418 which are functionals over @xmath419 defined by @xmath420 , and , for @xmath88 , @xmath421 @xmath422 _ proof : _ the word @xmath426 is of the generic form with @xmath427 ( definition [ generalized ] ) .
let @xmath428 with associated involution @xmath278 ( see section [ wignerprocesses ] ) .
let @xmath429 and @xmath430 be the instants where the walk on @xmath8 crosses the oriented edges @xmath284 and @xmath286 , with @xmath282 .
we colour these edges by giving colour 1 to @xmath284 when the symbol in @xmath431 located at position @xmath429 is @xmath407 , and give the colour * otherwise . clearly , both edges have different colours , and the elements of the set of edges @xmath432 ( the children of @xmath7 in @xmath8 ) have the same colour .
the result is then a consequence of the definition of the product with the rules i ) , ii ) and iii ) .
iterated integrals are naturel objects to consider in the setting of butcher s theory .
for example , in the framework of theorem [ differentialsys ] , the iterated integrals @xmath433 defined by @xmath434 , when @xmath88 , are such that @xmath435 , @xmath436 .
this result shows that the generating function of the * -moments of the operator @xmath443 can be obtained by solving the system given in proposition [ integralsystem ] .
we recover in this way a result of @xcite , lemmas 8.5 and 8.8 .
[ dk8 ] in the setting of corollary [ solutiona ] , the generating function @xmath444 solves @xmath445 where @xmath446 , that is , @xmath447 and @xmath448 are inverse with respect to composition .
moreover @xmath449 . _
proof : _ we solve the integral system by looking for solutions of the form @xmath450 and @xmath451 , with @xmath452 and @xmath453 .
we deduce that @xmath454 is solution of the equation @xmath455 .
the formula for the moments of @xmath443 is a consequence of lagrange s inversion formula .
20 f. bergeron , f. flajolet and b. salvy varieties of increasing trees in
notes in comp .
_ , * vol . 581 * , j
c raoult ed . , 24 - 48 , 1992 .
biane permutations suivant le type dexcdances et le nombre dinversions et interprtation combinatoire dune fraction continue de heine , * 14*(4):277 - 284 , 1993 .
d.j . broadhurst and d. kreimer renormalization automated by hopf algebra , no.6 : 581 - 600 , 1999 .
butcher an algebraic theory of integration methods , * 26*:79 - 107 , 1972 .
butcher wiley , chichester , 2003 . c. brouder
runge - kutta methods and renormalization , * 12 * : 521 , 2000 . a. connes and d. kreimer hopf algebras , renormalization and noncommutative geometry , * 199
* : 203 - 242 , 1998 .
r. delbourgo and d. kreimer using the hopf algebra structure of qft in calculations , * 60*:105025 , 1999 .
r. dobrow and j. fill on the markov chain for the move - to - root rule for binary search trees .
_ , * 5 * : 1 - 19 , 1995 . k. dykema and u. haagerup dt - operators and decomposability of voiculescu s circular operator .
arxiv : math.oa/0205077 , 2002 .
j. fill on the distribution of binary search trees under the random permutation model .
, * 8(1)*:1 - 25 , 1996 . f. flajolet and r. sedgewick , book in preparation , available at http://pauillac.inria.fr/algo/flajolet/publications/books.html .
e. hairer , s. nrsett ans g. wanner springer , berlin , second edition . 1993 .
e. hairer , c. lubich and g. wanner springer series in computational mathematics 31 , 2002 .
d. kreimer chen s iterated integral represents the operator product expansion , * 3 * : 627 - 670 , 1999 .
p. leroux and g. viennot combinatorial resolution of systems of differential equations i : ordinary differential equations . in
_ combinatoire numrative_(1986 ) , g. labelle and p. leroux , eds , no . 1234 in lecture notes in mathematics , springer verlag , p. 210 - 245 .
p. leroux and g. viennot combinatorial resolution of systems of differential equations .
separation of variables .
, * 72*:237 - 250 , 1988 . h. mahmoud .
wiley , new york , 1992 . c. mazza and d. piau products of large correlated symmetric matrices and q - catalan numbers ,
* 124*:574 - 594 , 2002 .
a. meir and j. moon on the altitude of nodes in random trees , * 30 * : 997 - 1015 , 1978 .
a. meir and j. moon on the log - product of the subtree - sizes of random trees , * 12 * : 197 - 212 , 1997 .
p. sniady multinomial identities arising from the free probability theory , * 101*(1):1 - 19 , 2003 .
r. speicher a new example of `` independence '' and `` white noise '' , * 84*(2):141 - 159 , 1990 .
a. soshnikov universality at the edge of the spectrum in wigner random matrices .
, * 207*(3):697 - 733 , 1999 .
l. takacs a bernoulli excursion and its various applications , * 23*:557 - 585 , 1991 . | we consider simply generated trees , like rooted plane trees , and consider the problem of computing generating functions of so - called bare functionals , like the tree factorial , using b - series from butcher s theory .
we exhibit a special class of functionals from probability theory : the associated generating functions can be seen as limiting traces of product of semi - circular elements .
key words and phrases .
b - series , random matrices , rooted plane trees , runge - kutta methods , simply generated trees .
2000 ams subject classification .
primary 05a15 secondary 05c05 , 30b10 , 46l54 , 81s25 |
the imf is one of the most important factors influencing the evolution of galaxies . in particular
, possible transitions of the stellar imf from that of the first ( metal - free ) stars to that of the present - day ( metal - rich ) stars have been given much attention .
the typical first generation stars are thought to be predominantly of high - mass ( e.g. @xmath1 according to * ? ? ?
* ) from studies of star formation in metal - free environments
. however , the actual form of the imf for such stars is poorly understood and is still controversial ( see e.g. * ? ? ? * and references therein ) .
in addition , there are few studies dealing with the transition of the imf during the early epoch of the galaxy .
if the imf is not universal then a question arises - how and when did the imf change to the present - day imf which is biased towards low - mass stars ( @xmath2 ) @xcite ?
extremely metal - poor stars ( emp stars ) observed in the galaxy can help to answer this question because we can estimate the number of former agb stars that contributed to the change of the surface abundances evident in these stars .
spectroscopy of emp stars @xcite reveals that carbon - enhanced stars are very common amongst emp stars ( 20 % or more @xcite ) when compared to the higher - metallicity subgiant ch stars ( @xmath3 % ) @xcite .
the large frequency of carbon - enhanced metal - poor ( cemp ) stars is argued to be a consequence of binary mass transfer from agb stars that have undergone carbon dredge - up into their envelopes @xcite .
the mechanism for carbon enhancement , the third dredge - up ( tdu ) , is the currently accepted theory that explains the origin of classical ba ii stars and ch stars ( see , e.g. * ? ? ?
in the same manner , the origins of some cemp stars are thought to be explained by the binary scenario , i.e. , the observed stars have binary companions that are now unseen white dwarfs .
one of the advantages of this scenario is that we do not expect all the observed emp stars to show carbon enhancement on their surface .
this is qualitatively consistent with the recent discovery of emp stars in the lowest metallicity range without large enhancements of carbon @xcite .
binarity has been statistically confirmed for the population of cemp stars that show enhancements of _
s_-process elements @xcite , as is expected from theory .
these objects are called cemp-_s _ stars , in contrast to the other subclass of cemp stars , the cemp - no stars , which do not show any enhancement of _ s_-process elements .
the origin of cemp - no stars is still controversial ( see , e.g. , * ? ? ?
observationally , there is another class of emp stars that show enhancements of nitrogen greater than the enhancement of carbon .
these stars are called nitrogen - enhanced metal - poor ( nemp ) stars and are thought to be produced via binary mass transfer from intermediate - mass agb stars which have converted c to n via hot bottom burning ( hbb ) during the agb phase .
previous works exploring the possibility that the imf was different in the early universe @xcite dealt only with the statistics of cemp stars .
( @xcite also pointed out that the different imf is suggested by the number of low - mass star survivors estimated from the survey areas of hk and hamburg - eso surveys . ) in particular , the imf which was proposed to have a peak at around @xmath4 @xcite has been criticized due to the possible overproduction of nemp stars @xcite .
however recent work on pulsation driven mass loss in agb stars at metallicities below @xmath0}}\sim -2.5 $ ] @xcite suggests that nemp star formation should have been suppressed in the early universe .
that study showed that the growth of pulsation during the agb phase is too small to trigger dust - driven mass loss in low - metallicity , intermediate - mass stars . with
mass - loss suppressed , the degenerate cores of these stars would have grown to be massive enough for the stars to explode as type i.5 supernovae , i.e. , the mass of their cores would have grown to approach the chandrasekhar mass limit @xcite .
the purpose of this study is to explore the possibility that the imf is different at low - metallicity by including as many ingredients related to stellar evolution and binary evolution as possible . in particular , this is the first study to include characteristic agb yields at @xmath0 } } <
-2.5 $ ] ( see , e.g. , * ? ? ?
we use the monte carlo method to simulate the evolution of binary systems by generating binaries with a range of initial primary mass , secondary mass , and binary period .
the mass of the primary is described by a log - normal function or power - law function .
the former is defined by @xmath5 .
the median mass and the dispersion are set at ( @xmath6 ( in @xmath7 ) , @xmath8 ) = ( 5 , 0.6 ) , ( 10 , 0.4 ) , ( 20 , 0.45 ) , ( 30 , 0.5 ) , and ( 50 , 0.6 ) , respectively , in order to reproduce cemp - no / cemp-_s _ ratio following @xcite .
power - law functions are defined by @xmath9 , where we choose @xmath10 , 0.85 , 0.35 , and 0.0 .
the lower and upper cut - off mass are set at @xmath11 and @xmath12 , respectively .
the mass of the secondary is subject to the mass ratio function which was taken from the @xcite or @xcite . for the @xcite distribution we adopted a flat distribution by ignoring the preference for like - mass pairs . for the period distribution of binaries
, we adopted a fitting formula for nearby main - sequence stars @xcite .
the simulations generate binary systems for metallicities in the range @xmath13}}\leq -1 $ ] in steps of 0.5 dex in each simulation .
the total number of binary systems for a given metallicity is typically one million , which produces more than 1,000 observable giants .
we note here that the simulations are compared to observations of giant stars ( figure [ fig : cfreq ] ) .
giants were chosen so that the uncertainty concerning the depth of the convective envelopes for main - sequence stars is reduced , and because the uncertainty in the elemental abundances derived for dwarfs is considerable .
we assume that the stars with @xmath14 evolve to the final stage of their evolution . in our simulations , we consider , for the first time , another channel of cemp star formation , i.e. , hydrogen ingestion into the helium flash convective zone for agb stars at @xmath0}}\leq -2.5 $ ] ( he - flash driven deep mixing ; he - fddm ) @xcite .
previous studies have not taken this channel into account @xcite .
surface chemical abundances of primary stars are taken from stellar models after the he - fddm at the beginning of the thermally pulsing agb ( tp - agb ) phase @xcite , or after the tdu during the tp - agb phase .
the surface abundance evolution for all the stars was taken from the literature @xcite . for models undergoing the he - fddm event
, we took the chemical abundances just after the dredge - up since this event dominates the change of the surface composition .
for the tdu event , we took the average abundances of chemical yields provided by @xcite . for models with @xmath0 } } < -2.5 $ ] with the mass range responsible for the tdu , we adopted the same yields as in @xcite by adopting the values at @xmath0}}= -2.3 $ ] of the same mass .
we also considered the effect of carbon enhancement by super - agb stars that host oxygen - neon - magnesium cores @xcite , although it turned out to be a minor contribution to cemp stars in our simulations .
a linear interpolation is made for the surface abundances between the model grids of initial mass and metallicity .
binary mass transfer events are treated as instantaneous episodes .
they are described by the masses of the binary components and the binary separation assuming kepler rotation without any orbital ellipticity .
if the binary separation is larger than the sum of the roche radius of the primary star and the radius of the secondary star during the main sequence phase , wind accretion is assumed to be at play .
otherwise , roche lobe overflow ( rlof ) applies to the system . in this study
, the occurrence of mass loss in agb stars is controlled by a parameter introduced to take into account of the possible suppression of mass loss at low metallicity @xcite . here
we assume that stars with @xmath15 and @xmath0}}\leq -2.5 $ ] do not eject any envelope and explode as type i.5 supernovae , which is chosen to be consistent with the measured abundances using the saga database @xcite .
the saga database includes 451 giant stars with carbon abundances derived from spectra with a resolution of @xmath16 .
the database covers the data published up to 2011 ( august , 2012 update ) . for wind accretion episodes we assume that the mass accretion onto the secondary star is subject to bondi - hoyle accretion with the constant stellar wind velocity of 20 km / s from the primary star .
the amount of mass lost from the primary is estimated from initial - mass final - mass relations for white dwarfs @xcite .
these treatments are the same as in our previous work @xcite and the details are given in the reference @xcite . for rlof ,
the roche radius is calculated from a commonly used approximation , @xmath17 , where @xmath18 , @xmath19 , and @xmath20 are the radius of a primary , binary separation , and mass ratio , respectively , @xcite . if the condition for rlof accretion is satisfied , we assume a constant mass accretion of @xmath21 onto the secondary occurs , following @xcite .
there is however no supporting evidence for this exact amount .
an important point here is that the common envelope phase can be avoided only for binaries that have a mass ratio very close to unity .
thus the effect of rlof accretion is actually always negligible in this study , and most binaries undergoing rlof are excluded from the observable population due to merging .
the definition of cemp giants is given by @xmath22}}\geq 0.7 $ ] ( * ? ? ?
* see also aoki et al .
2007 ) in the mass range 0.82 to @xmath23 , considering the stellar lifetimes in the galactic halo population ( comparable to the age of the universe ) .
self - pollution by agb stars is excluded due to their negligible contribution to the population compared with red giants .
we designate cemp star candidates as nemp stars if the initial mass is in the range in which hbb operates .
this mass range is a parameter of our model and is set at @xmath24 in our fiducial models . for the definition of cemp-_s
_ and cemp - no stars , we use parameters for mass and metallicity .
we tentatively define a critical mass , set at @xmath25 , below ( above ) which the model stars become cemp-_s _
( no ) stars during the agb phase . for cemp - no stars
, it is required that the model stars should have @xmath0}}\leq -2.5 $ ] to be consistent with observations .
note that the origin of cemp - no stars is still controversial ( see , e.g. * ? ? ?
* ) and will be discussed in a separate paper .
see also discussions in @xcite for more detail .
the frequency of cemp stars as a function of metallicity is given in figure [ fig : cfreq ] for observations and for our models with different adopted imfs .
the cemp fraction cemp / emp and the fractions cemp - no / cemp and nemp / cemp from our models are shown in table [ tab : model ] for a selected set of parameters .
these ratios are averaged values from the saga database for metallicities below @xmath0}}= -2 $ ] .
they are obtained by taking a weighted average of the ratios in each of the metallicity bins below @xmath0}}= -2 $ ] , with the weight being the number of stars in each bin .
model b is our fiducial model as it satisfactorily reproduces the observations .
it was chosen to reproduce the distribution of carbon abundance as shown in figure [ fig : distc ] .
it is to be noted that the cemp fractions of stars taken from the saga database can be affected by selection bias , although they are consistent with the values derived for other homogeneous datasets @xcite .
the results presented in the top panel of fig .
[ fig : cfreq ] demonstrate that the cemp fraction for emp stars is well reproduced by the high - mass star dominated imf ( model b ) , while the salpeter imf ( model g ) produces a very low cemp fraction compared with observations .
this is consistent with previous work @xcite , although we note that those studies did not consider the effects of stellar evolution at @xmath0 } } < -2.5 $ ] and assumed that all the stars were born as binaries .
the cemp fraction produced by the model with the flat imf ( model j ) is also below the observed value for @xmath0}}\leq -2 $ ] .
these results show that in order to reproduce the cemp fraction at @xmath0}}\leq -2 $ ] , the emp population should be dominated by massive stars that end their lives as type ii supernovae i.e. a high - mass star dominated imf is required . for more metal - rich populations ( @xmath0 } }
> -2 $ ] ) such as the inner halo population , a flat imf or low - mass star dominated imf gives a better agreement with observations .
this implies that an imf transition occurred at @xmath0}}\sim -2 $ ] , perhaps in association with the structure formation process of our galaxy ( see also * ? ? ?
* ) . in the bottom panel of fig .
[ fig : cfreq ] , we present a model with a transition of the imf at @xmath0}}= -2 $ ] . here
we briefly mention the parameter dependencies of the simulations , although these are thoroughly discussed in a separate paper ( suda et al .
, in prep . ) . for the choice of the period distribution function ,
the distribution derived by @xcite gives preference to short period binaries , which results in efficient channels for the common envelope phase and cemp star formation .
we adopted this distribution because it is the only study that considers a population of metal - poor binary stars .
using the other two binary period distributions @xcite will decrease the fraction of cemp stars by up to @xmath2610% for the high - mass imf and @xmath263% for the low - mass imf .
our fiducial model adopts other input physics and parameters so that the number of cemp stars can be maximized with reasonable choices .
the choice of the mass ratio function does not significantly alter the fraction of cemp stars as discussed in @xcite and @xcite .
the boundary masses for the occurrence of hbb and suppression of mass loss significantly affect the fraction of nemp stars .
these values are still open questions and can not be determined at this stage ( see * ? ? ?
* ; * ? ? ?
the mass boundary for the progenitors of cemp - no stars is also an open question and can change the ratio cemp - no / cemp .
there are four observable constraints on the simulations .
three of them are shown in table [ tab : model ] , and the other one is the abundances of emp stars .
firstly , it can be seen that the cemp fraction in our models can never reach as high as the observed 25 per cent when using a low - mass imf ( see also fig .
[ fig : cfreq ] ) . in order to check the robustness of the conclusion
, we have tested extreme assumptions for the model parameters , which include ( 1 ) a binary rate of 100 % , ( 2 ) excluding the contribution of supernova binaries because the speed of companion stars could have been higher than the velocity dispersion of the host halo after the disruption of binaries @xcite , and ( 3 ) the carbon abundance of primary star enhanced by a factor of 10 .
the superposition of all these assumptions still fails to explain the observed cemp frequencies at low - metallicity when a low - mass imf is used .
a significant increase in the cemp fraction can only be obtained by assuming that all the emp stars are born as binary systems , although this gives the cemp fraction well below 20 % at @xmath0}}= -2.5 $ ] .
a 100% binary fraction is however highly unlikely @xcite .
secondly , we find that the ratio cemp - no / cemp is extremely low for the models using the low - mass imf . to obtain a ratio cemp - no / cemp @xmath27
, we have to set the lower mass boundary for the formation of cemp - no stars as small as @xmath28
. however , this then leads to a mismatch with the observed ratio cemp-_s _ /emp .
thus it is evident that other sources of cemp - no stars , for example stars that formed out of an ism pre - enriched with carbon , are needed if we adopt a low - mass imf @xcite .
we note however that any source of pre - enrichment would require more massive stars , i.e. , a massive - star dominated imf , although its shape depends on the efficiency of low - mass star formation out of ejecta from the massive stars .
thirdly , the ratios of nemp to cemp stars in models with low - mass imfs are too small to be compatible with the observations .
the observed ratio is around @xmath29 in the current available data @xcite , while it was previously thought to be much smaller @xcite .
the mass range of nemp star progenitors would need to be @xmath30 , without the assumption of mass loss suppression , to obtain a nemp / cemp ratio @xmath260.1 .
this mass range would need a higher efficiency of hbb at low metallicity as has been reported by some studies @xcite .
however , this range will also reduce the cemp fraction dramatically , and hence can not coexist with the requirement for the large frequency of cemp stars .
finally , the number of the progenitors of type i.5 supernovae dominates over , or is comparable to , that of type ii supernovae when a low - mass imf is used , which is not supported observationally since galactic halo stars do not show the low values of that would be expected from type i.5 supernovae @xcite .
thus our simulations arrive at the robust conclusion that population simulations using low - mass imfs can not match the observational constraints given by galactic emp stars .
our results add to previous evidence for a non - universality of the imf .
other evidence comes from studies such as the population of white dwarfs @xcite , those involving simulations of star formation ( see , e.g. * ? ? ?
* ) , and observations of early - type galaxies ( see , e.g. * ? ? ?
our study shows that the observed population in the galactic halo can be explained by a transition of the imf from one biased towards high masses to one biased toward lower masses at a metallicity of @xmath0}}\sim -2 $ ] .
we propose that the statistics of cemp and nemp stars in the entire range of metallicity in our galaxy and dwarf galaxies in the local group should be investigated to reveal the star formation history of galaxies .
we have compared the results of our binary population synthesis model with the observations of metal - poor halo stars in the galaxy .
we find that a low - mass biased imf , similar to the present one , is incompatible with the four observable quantities related to the cemp subclasses . on the other hand , considering a high - mass biased imf , together with the occurrence of sn i.5 in a certain mass and metallicity range ( supported by recent simulations of pulsation - driven mass - loss in low metallicity stars ) , a more reasonable agreement between models and observations
our results imply that the transition of the imf occurred around a metallicity of @xmath0}}\sim -2 $ ] , which should be tested by star formation theory and observations .
these results also give important clues in the most metal poor regimes as to the mass and metallicity limits of stars that experience hbb , tdu and stellar winds , or that allow the formation of cemp ( -_s _ and -no ) stars , nemp stars , sn i.5 and snii .
our understanding of these limits has been traditionally hampered by the scarcity of observations .
we hope these results can help to improve the understanding of the chemical evolution of our galaxy .
the authors thank daniela carollo and david yong for kindly providing the data of cemp fraction .
we also thank takayuki saitoh for useful comments and suggestions . this work has been supported by grant - in - aid for scientific research ( 18104003 ) , from japan society of the promotion of science . t.s .
acknowledges the support of the institutional program for young researchers overseas visits by jsps , which enhanced the results and discussions in this work . | the stellar initial mass function ( imf ) plays a crucial role in determining the number of surviving stars in galaxies , the chemical composition of the interstellar medium , and the distribution of light in galaxies .
a key unsolved question is whether the imf is universal in time and space .
here we use state - of - the - art results of stellar evolution to show that the imf of our galaxy made a transition from an imf dominated by massive stars to the present - day imf at an early phase of the galaxy formation .
updated results from stellar evolution in a wide range of metallicities have been implemented in a binary population synthesis code , and compared with the observations of carbon - enhanced metal - poor ( cemp ) stars in our galaxy .
we find that applying the present - day imf to galactic halo stars causes serious contradictions with four observable quantities connected with the evolution of agb stars .
furthermore , a comparison between our calculations and the observations of cemp stars may help us to constrain the transition metallicity for the imf which we tentatively set at @xmath0}}\approx -2 $ ] .
a novelty of the current study is the inclusion of mass loss suppression in intermediate - mass agb stars at low - metallicity .
this significantly reduces the overproduction of nitrogen - enhanced stars that was a major problem in using the high - mass star dominated imf in previous studies .
our results also demonstrate that the use of the present day imf for all time in chemical evolution models results in the overproduction of type i.5 supernovae .
more data on stellar abundances will help to understand how the imf has changed and what caused such a transition .
[ firstpage ] stars : evolution stars : agb and post - agb galaxy : halo binaries : general stars : formation stars : population ii |
increasingly often , researchers are confronted with monitoring the states of nodes in large computer , social , or power networks where these states dynamically change due to viruses , rumors , or failures that propagate according to the graph topology @xcite . this class of network dynamics has been extensively modeled as a percolation phenomenon , where nodes on a graph can randomly `` infect '' their neighbors .
percolation across networks has a rich history in the field of statistical physics , computer science , and mathematical epidemiology . here
, researchers are typically confronted with a network , or a distribution over the network topology , and extract fixed point attractors of node configurations , thresholds for phase transitions in node states , or distributions of node state configurations @xcite . in the field of fault detection , the nodes or edges can `` fail '' , and the goal is to activate a subset of sensors in the network which yield high quality measurements that identify these failures @xcite . while the former field of research concerns itself with extracting _
offline _ statistics about properties of the percolation phenomenon on networks , devoid of any measurements , the latter field addresses _ online _ measurement selection tasks . here
, we propose a methodology that actively tracks a causal markov process across a complex network ( such as the one in figure [ sf ] ) , where measurements are adaptively selected .
we extract conditions such that the updated posterior probability of all nodes `` infected '' is driven to one in the limit of large observation time .
in other words , we derive conditions for the existence of an epidemic threshold on the updated posterior distribution over the states . the proposed percolation threshold should more accurately reflect the true conditions that cause a phase transition in a network , e.g. , node status changing from healthy / normal to infected / failed , than traditional thresholds derived from conditions on predictive distributions which are devoid of observations or controls . since most practical networks of interest are large , such as electrical grids , it is usually infeasible to sample all nodes continuously , as such measurements are either expensive or bandwidth is limited . given these , or other resource constraints
, we present an information theoretic sampling strategy that selectively targets specific nodes that will yield the largest information gain , and thus , better detection performance . the proposed sampling strategy balances the trade - off between trusting the _ predictions _ from the known model dynamics ( from percolation theory ) and expending precious resources to select a set of nodes for measurement . we present the adaptive measurement selection problem and give two tractable approximations to this subset selection problem based upon the joint and marginal posterior distribution , respectively .
a set of decomposable bayesian filtering equations are presented for this adaptive sampling framework and the necessary tractable inference algorithms for complex networks are discussed .
we present analytical worst case performance bounds for our adaptive sampling performance , which can serve as sampling heuristics for the activation of sensors or trusting predictions generated from previous measurements . to the author s knowledge ,
this is the first attempt to extract a percolation threshold of an actively monitored network using the updated posterior distribution instead of the observation independent predictive distributions .
the objective of actively monitoring the @xmath0 node network is to recursively update the posterior distribution of each hidden node state given various measurements .
specifically , the next set of @xmath1 measurement actions ( nodes to sample ) , @xmath2 , at next discrete time are chosen such that they yield the highest quality of _ information _ about the @xmath0 hidden states .
the condition on @xmath3 simulates the reality of fixed resource constraints , where typically only a small subset of nodes in a large network can be observed at any one time . here , the hidden states are discrete random variables that correspond to the states encoded by the percolation process on the graph . here , the graph @xmath4 , with @xmath5 representing the set of nodes and @xmath6 corresponding to the set of edges .
formally , we will assume a state - space representation of a discrete time , finite state , partially observed markov decision process ( pomdp ) .
here , @xmath7 represents the joint hidden states , e.g. , healthy or infected @xmath8 represents the @xmath1 observed measurements obtained at time @xmath9 , e.g. , biological assays or pinging an ip address , and @xmath10 represents the @xmath1 actions taken at time @xmath9 , i.e. , which nodes to sample . here ,
@xmath11 , continuous / categorical valued vector of measurements , which is induced by action @xmath12 , @xmath13 , with @xmath14 confined to be the set of all @xmath0 individuals in the graph , and @xmath15 .
since the topology of @xmath16 encodes the direction of `` flow '' for the process , the state equations may be modeled as a decomposable partially observed markov process : @xmath17 here , @xmath18 is the neighborhood of @xmath19 , @xmath20 is a non - random vector - valued function , @xmath21 is measurement noise , and @xmath22 is a stochastic equation encoding the transition dynamics of the markov process ( see figure [ 2tbn ] for a two node graphical model representation ) . and @xmath23 for @xmath24,width=168 ] in our proposed framework for actively monitoring the hidden node states in the network , the posterior distribution is the sufficient statistic for inferring these states .
the general recursion for updating the joint posterior probability given all past and present observations is given by the standard bayes update formula : @xmath25 with @xmath26 the chapman - kolmogorov equations provide the connection between the posterior update ( [ intract ] ) and the distribution resulting from the standard percolation equations .
in the former , the updates are conditional probabilities that are conditional on past observations , while in the latter , the updates are not dependent on observations .
the local interactions in the graph @xmath16 imply the following conditional independence assumptions : @xmath27 @xmath28 where the likelihood term is defined in ( [ y ] ) and the transition dynamics are defined in ( [ z ] ) .
this decomposable structure allows the belief state ( posterior excluding time @xmath9 observations ) update , for the @xmath29 node in @xmath16 , to be written as : @xmath30 with the parent set , @xmath31 .
unfortunately , for highly connected nodes in @xmath16 , this marginal update becomes intractable .
it thus must be approximated @xcite . in most real world situations ,
acquiring measurements from all @xmath0 nodes at any time @xmath9 is unrealistic , and thus , a sampling policy must be exploited for measuring a subset of nodes @xcite . since we are concerned with monitoring the states of the nodes in the network , an appropriate reward is the expected information gain between the _ updated _ posterior , @xmath32 , and the belief state , @xmath33 : @xmath34\ ] ] @xmath35 with @xmath36-divergence @xmath37 \right)\ ] ] for distributions @xmath38 and @xmath39 with identical support . the reward in ( [ ig ] ) has been widely applied to multi - target , multi - sensor tracking for many problems including , sensor management and surveillance @xcite .
note that @xmath40 , where @xmath41 is the kullback - leibler divergence between @xmath38 and @xmath39 .
the expectation in ( [ ig ] ) is taken with respect to the conditional distribution @xmath42 given the previous measurements @xmath43 and actions @xmath44 . in practice , the expected information divergence in ( [ ig ] ) must be evaluated via monte - carlo methods .
also , the maximization in ( [ ig ] ) requires enumeration over all @xmath45 actions ( for subsets of size @xmath1 ) , and therefore , we must resort to greedy approximations .
we propose incrementally constructing the set of actions at time @xmath9 , @xmath44 , for @xmath46 , according to : @xmath47.\ ] ] both ( [ ig ] ) and ( [ approxig ] ) are selecting the nodes to sample which yield maximal divergence between the percolation prediction distribution ( belief state ) and the updated posterior distribution , averaged over all possible observations .
thus ( [ ig ] ) provides a metric to assess whether to trust the predictor and defer actions until a future time or choose to take action , sample a node , and update the posterior . since the expected @xmath36-divergence in ( [ ig ] )
is not closed form , we could resort to numerical methods for estimating this quantity .
alternatively , one could specify an analytical lower - bound that could be used in - lieu of numerically computing the expected information gain in ( [ ig ] ) or ( [ approxig ] ) .
we begin by noting that the expected divergence between the updated posterior and the predictive distribution ( conditioned on previous observations ) differ only through the measurement update factor , @xmath48 ( ( [ ig ] ) re - written ) : @xmath49 \nonumber\\ = \mathbb{e}_{g_{k|k-1 } } \left [ \frac{1}{\alpha-1 } \mbox{log } \
\mathbb{e}_{p_{k|k-1 } } \left [ \left ( \frac { f_k } { g_{k|k-1 } } \right)^{\alpha } \right ] \right]\label{alphadiv}.\end{gathered}\ ] ] so , if there is significant overlap between the likelihood distributions of the observations , the expected divergence will tend to zero , implying that there is not much value - added in taking measurements , and thus , it is sufficient to use the percolation predictions for inferring the states
. it would be convenient to interchange the order of the conditional expectations in ( [ alphadiv ] ) .
it is easily seen that jensen s inequality yields the following lower bound for the expected information gain @xmath50 \nonumber\\ \ge \frac{1}{\alpha-1 } \mbox{log } \
\mathbb{e}_{p_{k|k-1 } } \left [ \mathbb{e}_{g_{k|k-1 } } \left [ \left ( \frac { f_k } { g_{k|k-1 } } \right)^{\alpha } \right ] \right].\end{gathered}\ ] ] here , the inner conditional expectation can be obtained from @xmath51 , which has a closed form for common distributions ( e.g. , multivariate gaussians ) @xcite .
for tracking the percolation process across @xmath16 , we have discussed recursive updating of the belief state . however , computing these updates exactly is in general intractable . for the remainder of the paper
, we will use ( [ y ] ) and ( [ z ] ) to directly update the marginal posterior distribution using the following matrix representation : @xmath52 with updated marginal posterior @xmath53^t$ ] with @xmath54 , @xmath55 , and marginal belief state @xmath56^t$ ] with @xmath57 .
note that for @xmath58 , @xmath59 , and @xmath60 .
given that we can find an efficient way of updating @xmath61 , according to the transition dynamics ( [ z ] ) , we can solve a modified version of ( [ approxig ] ) , for @xmath62 : @xmath63\ ] ] @xmath64 one interesting property of the bayesian filtering equations is that the updated posterior can be written as a perturbation of the predictive percolation distribution through the following relationship ( @xmath65 omitted for clarity ) : @xmath66 hence , when the sensors do a poor job in discriminating the observations , @xmath67 , we have @xmath68 .
it is of interest to determine when there is significant difference between the posterior update and the prior update specified by the standard percolation equations .
recall that the updated posterior is , in the mean , equal to the predictive distribution , @xmath69 = \textbf{p}_{k|k-1}$ ] .
the total deviation of the updated posterior from the percolation distribution can be summarized by computing the trace of the following conditional covariance : @xmath70 \right ) = \label{cond_cov } \\
\mbox{tr } \left ( \mathbb{e } \left [ \left ( \textbf{p}_k - \mathbb{e } \left [ \textbf{p}_k | \textbf{y}^{k-1}_0 \right ] \right ) \left ( \textbf{p}_k - \mathbb{e } \left [ \textbf{p}_k | \textbf{y}^{k-1}_0 \right ] \right)^t | \textbf{y}^{k-1}_0 \right ] \right ) .
\nonumber\end{gathered}\ ] ] using ( [ mat_update ] ) and properties of the trace operator , we obtain the following measure of total deviation of the updated posterior from the predictive distribution in terms of @xmath71 and @xmath72 : @xmath73 \right ) = \mbox{tr } \left ( \mathbb{e } \left [ \left ( \textbf{d}_k - \textbf{i}\right)^2 | \textbf{y}^{k-1}_0 \right ] \textbf{p}_{k|k-1 } \right)\end{gathered}\ ] ] with @xmath74 . the conditional expectation in ( [ deviation ] ) is the pearson @xmath75 divergence between distributions @xmath76 and @xmath77 , for all @xmath19 .
this joint measure of deviation is analytical for particular families of distributions and thus can be used as an alternative measure of divergence for activation of sensors @xcite .
there has recently been significant interest in deriving the conditions of a percolation / epidemic threshold in terms of transition parameters and the graph adjacency matrix spectra for two state causal markov processes @xcite .
such thresholds yield conditions necessary for an epidemic to arise from a small number of `` infections '' .
knowledge of these conditions are particularly useful for designing `` robust '' networks , where the probability of epidemics is minimized .
percolation thresholds are typically obtained by extracting the sufficient conditions of the network and model parameters for the node states to be driven to their stationary point , with high probability .
the probability of these events are usually computed using the observation independent percolation distribution @xcite .
we use the results in @xcite to derive a percolation threshold based upon the updated posterior distribution assuming a restricted class of two - state markov processes .
these conditions should more accurately model the _ current _ network threshold since the posterior distribution tracks a particular `` disease '' trajectory better than the observation independent percolation distribution .
formally , @xmath78 , @xmath79 is the conditional likelihood for node @xmath19 , @xmath80 , and @xmath81 .
here , we will assume that @xmath82 is the unique absorbing state of the system .
the bayes update for @xmath83 can be written as ( @xmath19 subscript omitted for clarity ) : @xmath84 there are three different sampling / observation dependent possibilities for each individual at time @xmath9 : case ( 1 ) , @xmath19 is not sampled and therefore , @xmath85 , case ( 2 ) , @xmath86 , and case ( 3 ) , @xmath87 . we first derive a tight - upper bound for cases ( 2 ) and ( 3 ) of the form @xmath88 .
for the remainder of the analysis we will assume that @xmath89 for cases ( 2 ) and ( 3 ) ( see appendix ) .
using the upper - bounds derived in the appendix , and after gathering all @xmath0 nodes , we have the following element - wise upper - bound on the updated belief state : @xmath90 with @xmath91 and @xmath92 where @xmath93 is the indicator function for the event @xmath94 .
thus far , we have established , under the assumptions of @xmath89 , an upper - bound for the updated posterior in terms of observation likelihoods and the belief state ( [ upper_pos ] ) . next , consider the restricted class of two - state markov processes on @xmath16 , for which we can produce a bound of the form @xmath95 where * s * contains information about the transition parameters and the topology of the network .
it turns out that the @xmath96 model of mathematical epidemiology falls within this restricted class of percolation problems @xcite .
the @xmath96 model on a graph @xmath16 , assumes that each of the @xmath0 individuals are in states @xmath97 or @xmath98 , where @xmath97 corresponds to _ susceptible _ and @xmath98 corresponds to _ infected_. at any time @xmath9 , an individual can receive the infection from their neighbors , @xmath99 , based upon their states at @xmath100 . under this @xmath96 model in @xcite @xmath101
where the markov transition parameters @xmath102 is the probability of @xmath19 transitioning from @xmath98 to @xmath97 , @xmath103 is the probability of transmission between neighbors @xmath19 and @xmath23 , and * a * is the graph adjacency matrix ( see figure [ sis_chain ] ) . returning to the derivation , using the bound ( [ dynamic_bound ] ) , we have , by induction , the following recursion : @xmath104 where we have lumped the higher order modes and higher order cross - terms into @xmath105 .
the _ dominant mode of decay _ of the updated posterior may be found by investigating the following eigen - decomposition : @xmath106 with @xmath107^t$ ] ( @xmath98 at @xmath108 element ) . without loss of generality
, we can assume the eigenvalues of @xmath109 are listed in decreasing order , @xmath110 . now rewriting ( [ bs ] )
, we have @xmath111 where @xmath112 and the @xmath113 variables corresponds to the higher order terms . inserting ( [ spectral ] ) into ( [ recursion ] ) , and matching the largest eigenvalues of @xmath114 with @xmath115 we obtain @xmath116 thus , at large @xmath9 , the dominant mode of the posterior goes as @xmath117 ( the modes in @xmath118 decay faster than the dominant mode presented above ) .
we can see that if the spectral radius of @xmath109 is less than one , @xmath119 , then for large @xmath9 , @xmath120 , which is the unique absorbing state of the system .
this epidemic threshold condition on @xmath115 has been previously established for unforced @xmath96-percolation processes @xcite .
however , in the tracking framework , the rate at which the posterior decays to the _ susceptible _ state is perturbed by an additional measurement dependent factor , @xmath121 .
this measurement - dependent dominant mode of the posterior should more accurately model the true dynamic response of the node states better than that in @xcite since the posterior better tracks the truth than the unforced predictive distribution .
additionally , this dominant mode of the updated posterior distribution allows one to simulate the response of the percolation threshold to intervention and control actions which are designed to increase the threshold , such that the probability of epidemics is minimized .
markov chain for node @xmath19 interacting with the infected states of its neighbors , width=172 ] here , we present results of simulations of our adaptive sampling for the active tracking of a causal markov ground truth process across a random 200 node , scale - free network ( figure [ sf ] ) . since the goal in tracking is to accurately classify the states of each node , we are interested in exploring the detection performance as the likelihood of an epidemic increases through the percolation threshold for this graph
. one would expect different phase transitions ( thresholds ) in detection performance for various sampling strategies , ranging from the lowest threshold for unforced percolation distributions to highest for a continuous monitoring of all @xmath0 nodes .
we will present a few of these detection surfaces that depict these phase transitions for the unforced percolation distribution , random @xmath122 node sampling , and our proposed information theoretic adaptive sampling of @xmath122 . here
, we will restrict our simulations to the two - state @xmath96 model of mathematical epidemiology described above .
, width=240 ] the sensor models ( [ y ] ) , are of the form of two - dimensional multivariate guassians with common covariance and shifted mean vector .
the transition dynamics of the @xmath29 individual ( [ z ] ) , for the @xmath96 model is given by : @xmath123.\ ] ] where @xmath124 is the indicator function of @xmath19 being infected at time @xmath100 . the transmission term between @xmath19 and @xmath99
is known the reed - frost model @xcite . since the tail of the degree distribution of our synthetic scale - free graph contains nodes with degree greater than 10 , updating ( [ marginal ] )
exactly is unrealistic and we must resort to approximate algorithms . here
, we will assume the mean field approximation used by @xcite for this @xmath96 model , resulting in the following marginal belief state update for the @xmath29 node of infected ( @xmath125 ) : @xmath126.\ ] ] equation ( [ mfapprox ] ) allows us to efficiently update the marginal belief state directly for all @xmath0 nodes which are then used for estimating the best @xmath1 measurements using ( [ ig_marg ] ) .
[ aur_perc ] and time , title="fig:",width=321 ] [ aur_rnd ] and time , title="fig:",width=321 ] [ aur_adaptive ] and time , title="fig:",width=321 ] as we are interested in detection performance as a function of time and epidemic intensity , the area under the roc curve ( aur ) is a natural statistic to quantify the detection power ( detection of the infected state ) .
the aur is evaluated at each time @xmath9 , each @xmath96 percolation intensity parameter @xmath127 and over 500 random initial states of the network . for the @xmath96 model
, @xmath128 is the single parameter ( aside from the topology of the graph ) that characterizes the intensity of the percolation / epidemic .
it is useful to understand how the detection performance varies as a function of epidemic intensity , as it indicates how well the updated posteriors are playing `` catch - up '' in tracking the true dynamics on the network . for this @xmath96 model ,
the percolation threshold is defined as @xmath129 where @xmath130 is the spectral radius of the graph adjacency matrix , @xmath131 @xcite .
values of @xmath128 greater than @xmath132 imply that any infection tend to become an epidemic , whereas those values less than @xmath132 imply that small epidemics tend to die out . for the network under investigation ( figure [ sf ] ) , @xmath133 .
we see from figure [ aur_perc ] that a phase transition in detection power ( aur ) for the unforced percolation distribution does indeed coincide with the epidemic threshold @xmath132 .
while the epidemic threshold for the random and adaptive sampling policies is still @xmath133 , the measurements acquired allow the posterior to better track the truth , but only up to their respective phase transitions in detection power ( see figures [ aur_rnd ] and [ aur_adaptive ] ) .
figure [ aur_adaptive ] confirms that the adaptive sampling better tracks the truth than randomly sampling nodes , while pushing the phase transition in detection performance to higher percolation intensities , @xmath128 .
we see that the major benefit of the adaptive sampling is apparent when conditions of the network are changing moderately , at medium epidemic conditions . beyond a certain level of percolation intensity ,
more resources will need to be allocated to sampling to maintain a high level of detection performance .
a heuristic sampling strategy based on the topology of @xmath16 was also explored ( results not shown ) by sampling the `` hubs '' ( highly - connected nodes ) . however , detection performance was only slightly better than random sampling and poorer than our adaptive sampling method .
-axis ) of a given degree ( @xmath134-axis ) over time ( @xmath135-axis ) for @xmath122 adaptive sampling strategy : a. ) @xmath136 , b. ) @xmath137 , and c. ) @xmath138,width=345 ] it is often useful for developing sampling heuristics and offline control / intervention policies to inspect what _ type _ of nodes , topologically speaking , is the adaptive sampling strategy targeting , under various network conditions ( different values of @xmath128 ) . in figure [ sampling ] , the relative frequency of nodes sampled with a particular degree is plotted against time ( under the @xmath122 adaptive sampling strategy ) for three different values of @xmath128 ( over 500 random initial conditions of the network ) . for the larger of the three values explored ( @xmath139 )
we see that the sampling is approximately uniform across the nodes of each degree on the graph ( figure [ sampling](c ) ) .
therefore , under extremely intense epidemic conditions , the adaptive sampling strategy is targeting all nodes of each degree equally , and therefore , it is sufficient to perform random sampling .
for the two lower values of @xmath128 , figure [ sampling](a ) and figure [ sampling](b ) ( near @xmath132 ) , we see that adaptive policy targets highly connected nodes more frequently than those of lesser degree and thus , it is more advantageous to exploit such a strategy , as compared to random sampling ( see aur surface in figure [ aur_adaptive ] ) .
in this paper , we have derived the conditions for a network specific percolation threshold using expressions for the updated posterior distribution resulting from actively tracking the process .
these conditions recover the unforced percolation threshold derived in @xcite but with an additional factor involving sensor likelihood terms due to measurements obtained throughout the monitoring . a term of the form @xmath117 ( derived in ( [ bound ] ) )
was shown to be the dominant mode of the updated posterior dynamic response to active intervention of immunizing the nodes ( holding node states constant ) .
the conditions of the percolation using the updated posterior should more accurately model the phase transition corresponding to a particular disease trajectory and therefore , enable a better assessment of immunization strategies and any subsequent observations resulting from such actions .
the framework presented above , along with the new posterior percolation threshold , should provide additional insight into active monitoring of large complex networks under resource constraints .
in case ( 2 ) , when @xmath86 , we can re - write ( [ bayes_geo ] ) in terms of an _ alternating geometric series _ : @xmath140 p_{k|{k-1 } } \nonumber \\ & \le & \frac { f^{(1)}_k } { f^{(0)}_k } \left [ 1 + \frac { |\delta f_k| } { f^{(0)}_k } p_{k|{k-1 } } \right ] p_{k|{k-1 } } \end{aligned}\ ] ] where we have used the fact that @xmath141 . recalling that @xmath142 for @xmath143 , we have @xmath144 p_{k|{k-1}}.\ ] ] @xmath145 p_{k|{k-1 } } \nonumber \\ = \frac { f^{(1)}_k } { f^{(0)}_k } \left [ 1 + \frac { |\delta f_k| } { f^{(0)}_k } p_{k|{k-1 } } + \sum_{l=2}^{\infty } \left(\frac { |\delta f_k| } { f^{(0)}_k } p_{k|{k-1 } } \right)^l \right ] p_{k|{k-1}}. \end{gathered}\ ] ] | percolation on complex networks has been used to study computer viruses , epidemics , and other casual processes . here , we present conditions for the existence of a network specific , observation dependent , phase transition in the updated posterior of node states resulting from actively monitoring the network . since
traditional percolation thresholds are derived using observation independent markov chains , the threshold of the posterior should more accurately model the true phase transition of a network , as the updated posterior more accurately tracks the process .
these conditions should provide insight into modeling the dynamic response of the updated posterior to active intervention and control policies while monitoring large complex networks . |
a massive star with a main - sequence mass @xmath11 ends up as a core - collapse supernova ( ccsn ) .
core collapse is inaugurated by electron capture for a star with an o+ne+mg core ( @xmath12 ) or fe photodisintegration for a star with an fe core ( @xmath13 ) .
an explosion mechanism of ccsne is still under investigation , in particular , for the ccsn of an star with an fe core ( fe ccsn ) .
recently , sophisticated multidimensional simulations discovered that neutrino - driven convection and/or standing - accretion shock instability enhances neutrino heating and initiates an outward flow ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for recent reviews ) .
however , explosion energies @xmath14 are about one orders of magnitude smaller than a canonical value of a normal ccsn ( @xmath15 erg , e.g. , sn 1987a * ? ? ?
the fate of the less - massive star with the o+ne+mg core is different from that of the star with an fe core @xcite .
the o+ne+mg core is supported by electron degenerate pressure .
the mass and density of the o+ne+mg core increase through phases of shell burning of he and h. as the o+ne+mg core grows , an envelope undergoes mass loss to reduce the h mass and he dredge - up to enhance he abundance .
eventually , the star becomes a super - asymptotic giant branch ( sagb ) star ( e.g. , * ? ? ?
when the central density exceeds a critical value ( @xmath16 ) , electrons begin to be captured by magnesium , the degenerate pressure decreases , and thus the o+ne+mg core collapses gravitationally @xcite .
ensuing core bounce and neutrino heating can eject the envelope and part of the o+ne+mg core because of weak inward momentum carried by the low - density sagb envelope @xcite .
the explosion is called an electron - capture supernova ( ecsn ; * ? ? ?
* ; * ? ? ?
in contrast to the fe ccsn , the explosion of the ecsn is realized by first - principle simulations @xcite and a two - dimensional simulation demonstrates that the explosion takes place almost spherically @xcite .
this is the only example of the ccsn being self - consistently produced .
however , a low explosion energy derived from the simulations ( @xmath17 erg ) discriminates the ecsn from the normal ccsne .
the first - principle hydrodynamics simulation enables self - consistent studies , e.g. , on nucleosynthesis and observational features .
explosive nucleosynthesis in the ecsn is calculated by @xcite and @xcite based on @xcite .
they find a small amount of @xmath2ni in the ejecta [ @xmath18 and large production of elements with @xmath19 .
the small @xmath2ni mass is also decidedly different from the normal ccsne with @xmath20 .
however , a theoretical light curve of the ecsn has not been presented and the observational features of the ecsn are not yet theoretically clarified .
there are several sne which are suggested to be the ecsne .
one of the old examples is the crab nebula being a remnant of sn 1054 .
this is suggested from high he abundance @xcite , small ejecta mass @xcite , and low kinetic energy @xcite .
one of the recent examples is sn 2008s @xcite , which is suggested from a dust - surrounding bright progenitor and its faintness and slow evolution .
2008s - like objects have also been discovered ( e.g. , * ? ? ? * ; * ? ? ?
* ) . in this letter
, we adopt the explosion properties derived by the state - of - the - art first - principle simulation and present the first self - consistent multicolor light curves of ecsne .
we also investigate the contribution from a continuous energy release from a central remnant as the crab pulsar .
finally , we compare the multicolor light curves with observations of ecsn candidates .
we take an o+ne+mg core model with 1.377@xmath21 at a presupernova stage from @xcite and @xcite .
the model is a core of a star with @xmath22 which is calculated from an he star with 2.2@xmath21 .
a mass range of stars with the o+ne+mg core is @xmath0 but a progenitor of the ecsn should possess an sagb envelope ( see * ? ? ? * for a review ) , of which mass and abundance are influenced by @xmath23 , mass loss , and third dredge - up associated with thermal pulses .
however , the mass loss is highly uncertain and no calculation of full thermal pulses has been available . for almost fully convective envelope models of the progenitor , therefore , we adopt various envelope mass @xmath24 ( @xmath25 ) and hydrogen abundance @xmath26 ( @xmath27 ) by constructing hydrostatic and thermal equilibrium envelopes with binding energies of @xmath28 erg ( e.g. , @xcite ) .
density structures are shown in figure [ fig : presn ] .
the luminosity of the progenitor models is @xmath29 erg s@xmath5@xcite and their radii are @xmath30 km .
figure [ fig : hr ] demonstrates that the luminosity is roughly consistent with the progenitor of sn 2008s @xcite and is located at a bright tip of the sagb stars @xcite .
the explosion is initiated by a thermal bomb ) . ] with the explosion energy derived by the first - principle simulation ( @xmath31 erg , e.g. , @xcite ) .
the subsequent evolution is followed by a multigroup radiation hydrodynamical code stella @xcite , in which one - group @xmath32-ray transfer is calculated and in situ absorption of positron is assumed for energy deposition from @xmath2ni-@xmath2co radioactive decay .
an abundance distribution in the o+ne+mg core after the explosion and mass of heavy elements are taken from the model st in @xcite , which yields @xmath3 of @xmath2ni .
we note that no @xmath2ni is synthesized in the envelope due to the low temperature of @xmath33 k. we also investigate the contribution from the pulsar spin - down luminosity that could be bright at birth .
for example , the initial spin - down luminosity of the crab pulsar was @xmath34 erg s@xmath5 .
erg s@xmath5 , @xcite ) with an equation @xmath35 , where @xmath36 , @xmath37 and @xmath38 are initial spin - down luminosity , a spin - down timescale , and a braking index of the crab pulsar , respectively . ] since the pulsar spin - down luminosity can vary on individual ecsne and the deposition efficiency of an energy released from the pulsar is unknown , we expediently adopt the initial spin - down luminosity of the crab pulsar and assume that the released energy is fully deposited at the bottom of the ejecta ( `` full deposition '' ) or deposited pursuant to the same one - group transport as @xmath32-rays from the radioactive decay ( `` one - group transport '' ) .
in the explosion of the model with @xmath39 and @xmath40 ( figs . [ fig : str]a-[fig : str]b ) , a shock wave is accelerated up to @xmath41 km s@xmath5 during the first @xmath42 s due to the steep density gradient at the bottom of the envelope and decelerated down to @xmath43 km s@xmath5 by shock emergence from the stellar surface .
such drastic deceleration of the shock develops severe rayleigh - taylor instability . at the shock emergence , the ecsn flashes ( fig .
[ fig : lcdep]a ) .
the phenomenon is called a shock breakout , bolometrically brightest .
hereafter , we set @xmath44 at the shock breakout .
a peak wavelength at the shock breakout is @xmath45 and bolometric luminosity is as bright as @xmath46 erg s@xmath5 . adopting a dust evaporation radius @xcite and evaporation temperature for a short flash @xcite
, the shock breakout can destroy the circumstellar dust up to @xmath47 km for a case of carbon dust . as the sn ejecta cools down
, the ecsn enters a plateau phase lasting @xmath48 days .
the plateau luminosity is @xmath49 erg s@xmath5 , which is as bright as typical type ii plateau supernovae ( sne ii - p , e.g. , sn 2004et , @xcite ) , and a photospheric velocity at the plateau is @xmath50 km s@xmath5 , which is slightly slower than typical sne ii - p .
the plateau luminosity is fainter and the duration is longer for an explosion of a star with larger @xmath24 and higher @xmath26 .
the plateau is followed by a tail .
the luminosity suddenly fades by @xmath9 mag at the transition because of the small @xmath51 .
the tail luminosity declines gradually at a rate of @xmath52 mag day @xmath5 which is slightly faster than the energy release rate from the @xmath2codecay .
the multicolor light curves of the model with @xmath39 and @xmath40 are shown in figure [ fig : lcdep]b .
the photospheric temperature decreases after the shock breakout .
thus , the peak wavelength shifts to redder bands with time and the optical luminosity increases gradually .
the light curves in the bluer bands peak at earlier epochs , e.g. , @xmath53 days in the @xmath54 band and @xmath55 days in the @xmath56 band .
in contrast to typical sne ii - p , the @xmath54 and @xmath56 light curves do not show a plateau and begin to decline immediately after the peak .
the light curves in the bands redder than the @xmath57 band peak at the end of the plateau .
the multicolor light curves drop at the transition to the tail as the bolometric light curve does .
the decline rates of the tail are higher for bluer bands .
the pulsar spin - down only influences the tail because the spin - down luminosity of the crab pulsar is much fainter than the plateau luminosity .
the ecsn shines with the energy release from the @xmath2codecay at the beginning of the tail and then the energy release from the newborn pulsar becomes dominant at a late phase ( @xmath58 days , fig .
[ fig : lcdep]c ) .
the tail luminosity is floored in the full deposition model , while the decline rate is lowered in the one - group transport model .
the tail luminosity can be @xmath59 mag brighter than the model without the pulsar contribution at @xmath60 days .
the crab nebula is the most famous and conspicuous supernova remnant .
optical and uv observations illustrate that the crab nebula is notably he - rich [ @xmath61 , e.g. , @xcite ] .
in addition to the he - rich abundance , a small ejecta mass ( @xmath62 , @xcite ) and low kinetic energy ( @xmath63 erg , is assumed . the kinetic energy is doubled by adopting @xmath62 . ]
e.g. , @xcite ) suggest that the crab nebula is a remnant of the ecsn .
an expansion of the crab nebula was discovered in the early 1920s @xcite and , in the same year , the proximity of the crab nebula to sn 1054 was indicated @xcite .
an explosion date estimated by turning back the expansion is consistent with sn 1054 @xcite .
hence , it is widely believed that the expanding crab nebula is a remnant of sn 1054 .
a sudden appearance of sn 1054 was recorded in medieval times and its optical light curve is enscrolled in historiographies @xcite .
they described the dates of the first and last sightings and its brightness .
however , the medieval observations were rough and unconfident .
thus , reliability of the archives is deeply scrutinized by @xcite . referring the conclusion of that paper , we adopt the following three points with errorbars of @xmath64 mag and @xmath65 days ;
( 1 ) sn 1054 was as bright as venus ( optical magnitude @xmath66 to @xmath67 ) on july 4 , 1054 and might be visible earlier than july 4 , 1054 ( e.g. , may 10 , 1054 ) , ( 2 ) sn 1054 was visible in daytime for 23 days from july 4 , 1054 and @xmath68 on july 27 , 1054 , and ( 3 ) sn 1054 disappeared in night on april 6 , 1056 with @xmath69 .
we correct the distance ( @xmath70 kpc , @xcite ) and reddening [ @xmath71 , @xcite ] to the crab nebula .
we compare an `` optical '' and a root - mean - square deviation of @xmath72 @xcite . ]
light curve of the model with @xmath39 and @xmath40 against the observations of sn 1054 ( fig .
[ fig : lccomp]a ) . here
, an optical peak of the model is set to be july 4 , 1054 .
the ecsn model well reproduces the bright and short plateau of sn 1054 .
the epoch of the transition from the plateau to the tail corresponds to the date at which sn 1054 disappeared from the daytime sky .
the tail luminosities of the full deposition and one - group transport models are brighter and fainter than sn 1054 , respectively . if the deposition efficiency is the middle of these models , a gradually - declining optical light curve can cross with the last observation of sn 1054 .
we note that the last point of sn 1054 is not compatible with the ecsn model without the pulsar contribution as suggested in @xcite .
sn 2008s is suggested to be the ecsn because the progenitor is bright and surrounded by dust and because sn 2008s is faint and evolves slowly @xcite .
however , the plateau luminosity of the ecsn is as bright as that of normal sne ii - p due to the large presupernova radius ( fig .
[ fig : lcdep]a ) . thus , we conclude that the ecsn model based on the first - principle simulation @xcite is incompatible with the faintness of sn 2008s ( fig . [
fig : lccomp]b ) .
however , there is a caveat that the plateau luminosity and duration depend on @xmath14 and @xmath24 ( @xmath73 and @xmath74 , @xcite ) . if an explosion energy may be different for individual ecsne , e.g. , due to rotation , the faint and slowly - evolving sn 2008s could be explained . applying the scaling laws and taking @xmath75 erg s@xmath5 and @xmath76 days of sn 2008s
, we can derive its explosion properties ; @xmath77 erg and @xmath78 .
a series of first - principle simulation and light curve calculations are required to confirm that such the explosion is feasible and explains the multicolor light curves of sn 2008s .
we present multicolor light curves of ecsne with various @xmath24 and @xmath26 based on the results of the first - principle simulation @xcite .
we demonstrate that the shock breakout has a peak luminosity of @xmath4 erg s@xmath5 and can evaporate circumstellar dust up to @xmath79 km for the case of carbon dust and that the plateau luminosity and the duration of the ecsne are @xmath49 erg s@xmath5 and @xmath48 days , respectively .
the brighter and shorter plateau is realized by the model with smaller @xmath24 and lower @xmath26 .
the plateau is followed by the tail with the luminosity drop by @xmath9 mag .
the tail luminosity declines by @xmath52 mag day @xmath5 for the model without the pulsar contribution , while , if the pulsar contributes to the light curve , the tail light curve is floored or the decline rate is lowered .
the contribution from the pulsar spin - down luminosity as bright as the crab pulsar is prominent only in the tail .
the bright and short plateau and low explosion energy of the ecsn model are consistent with sn 1054 . the plateau is not reproduced with a low - energy explosion of a red supergiant star with heavy @xmath24 .
the tail luminosity of sn 1054 could be explained by the spin - down luminosity of the newborn crab pulsar .
the deceleration of the shock wave in the ecsn ( figs .
[ fig : str]a-[fig : str]b ) is also favorable to produce filamentary structures observed in the crab nebula ( e.g. , @xcite ) .
thus the observed features of sn 1054 are naturally reproduced by the ecsn model .
the luminosity of interaction of the ecsn with a circumstellar medium reaches only @xmath80 of the plateau of sn 1054 for a typical sagb star with the mass loss rate of @xmath81 @xcite .
thus an extremely dense and confined ( @xmath82 a few @xmath83 cm ) circumstellar matter is required to explain the bright and short plateau of sn 1054 @xcite .
the optical light curve of sn 1054 constrains the initial spin - down luminosity of the crab pulsar .
radio observations suggest @xmath84 yr and @xmath85 erg s@xmath5 to produce relic relativistic electrons @xcite .
however , such high @xmath36 leads to much brighter optical luminosity than sn 1054 at april 6 , 1056 .
our result favors @xmath86 yr and @xmath87 erg s@xmath5 .
these are also supported by a broad spectral evolution model of the pulsar wind nebula @xcite .
the typical ejecta velocity of the ecsn model with @xmath39 and @xmath40 is @xmath88 km s@xmath5 , which is consistent with the low expansion velocity of the crab nebula @xcite . on the other hand ,
the velocity of a wind blown from the progenitor is @xmath89 km s@xmath5 . adopting the duration of the sagb phase ( @xmath90 yr , @xcite )
, the sagb wind extends only upto @xmath91 km ( @xmath92 pc ) , which is similar to an apparent size of the crab nebula ( e.g. , @xcite ) .
the radius is smaller than the typical ejecta location of @xmath93 km ( @xmath94 pc ) .
therefore , the forward shock should locate in a low - density circumstellar wind blown at the core he - burning phase or in an interstellar medium .
this could be a reason why the forward shock of the crab nebula has not been found .
on the other hand , the plateau of the ecsn model is much brighter than that of sn 2008s .
if the explosion energy of ecsne is exactly @xmath1 erg as derived by the first - principle simulation @xcite , the multicolor light curves of the ecsn are inconsistent with those of sn 2008s
. however , there could be a caveat that the explosion energy may vary on individual ecsne .
the ecsn explosion with @xmath77 erg and @xmath78 might be compatible with sn2008s .
if sn 2008s is the ecsn with @xmath77 erg and @xmath78 , we can speculate on shock breakout luminosity and @xmath51 ; according to analytic dependence @xcite , the luminosity of the shock breakout is @xmath95 erg s@xmath5 .
carbon dust at @xmath96 km is evaporated by the shock breakout .
the size of a dust - free cavity is roughly consistent with that estimated from a midinfrared observation of sn 2008s @xcite .
the explosion energy is comparable to a gravitational binding energy of the progenitor at @xmath97 .
thus , sn 2008s is likely to eject materials above the outer edge of the core and yield @xmath98 , which is slightly smaller than an estimate from the observed tail of sn 2008s ( @xcite ; but see also @xcite ) .
therefore , we propose that the light curve tail of sn 2008s is powered by a spin - down luminosity of a newborn neutron star as sn 1054 .
we note a caveat that sn 2008s is a type iin sn and thus could be contributed to by interacting with a circumstellar medium . the number of sne sharing the characteristics of the ecsn , i.e. , a bright and short plateau and faint tail , is small , if the sn 2008s - like transients are not the ecsne .
we also note that faint and low - energy sne ii - p , e.g. , sn 1997d and sn 1999br @xcite , have a slower photospheric velocity at the plateau ( @xmath99 km s@xmath5 ) than the ecsn .
the scarcity of the ecsne might stem from a small mass range of a star ending up as ecsne at solar metallicity ( see * ? ? ? * for a review ) . since a significant contribution of ecsne to the production of @xmath100ca , @xmath101zn , and @xmath102zr
is suggested @xcite , chemical evolution models taking into account the scarcity is required to study the origin of these isotopes .
we would like to thank shuta j. tanaka and keiichi maeda for fruitful discussions on the crab pulsar and sne ii - p , respectively , and shinya wanajo for providing the abundance distribution .
data analysis were in part carried out on the general - purpose pc farm at center for computational astrophysics , national astronomical observatory of japan .
this research has been supported in part by the rfbr - jsps bilateral program , world premier international research center initiative , mext , japan , and by the grant - in - aid for scientific research of the jsps ( 23224004 , 23540262 , 23740157 ) and mext ( 23105705 ) .
is supported by the grants of the government of the russian federation 11.g34.31.0047 , rf sci .
school 5440.2012.2 and 3205.2012.2 and by a grant iz73z0 - 128180/1 of the swiss national science foundation ( scopes ) . , a. , dessart , l. , & livne , e. 2007 , in american institute of physics conference series , vol .
937 , supernova 1987a : 20 years after : supernovae and gamma - ray bursters , ed . s. immler , k. weiler , & r. mccray , 370380 , f. r. , & green , d. a. 2002 , historical supernovae and their remnants , by f. richard stephenson and david a. green .
international series in astronomy and astrophysics , vol .
oxford : clarendon press , 2002 , isbn 0198507666 , 5 | an electron - capture supernova ( ecsn ) is a core - collapse supernova ( ccsn ) explosion of a super - asymptotic giant branch ( sagb ) star with a main - sequence mass @xmath0 .
the explosion takes place in accordance with core bounce and subsequent neutrino heating and is a unique example successfully produced by first - principle simulations .
this allows us to derive a first self - consistent multicolor light curve of a ccsn . adopting the explosion properties derived by the first - principle simulation , i.e.
, the low explosion energy of @xmath1 erg and the small @xmath2ni mass of @xmath3 , we perform a multigroup radiation hydrodynamics calculation of ecsne and present multicolor light curves of ecsne of sagb stars with various envelope mass and hydrogen abundance .
we demonstrate that a shock breakout has a peak luminosity of @xmath4 erg s@xmath5 and can evaporate circumstellar dust up to @xmath6 cm for a case of carbon dust , that plateau luminosity and plateau duration of ecsne are @xmath7 erg s@xmath5 and @xmath8 days , respectively , and that a plateau is followed by a tail with a luminosity drop by @xmath9 mag .
the ecsn shows a bright and short plateau that is as bright as typical type ii plateau supernovae , and a faint tail that might be influenced by spin - down luminosity of a newborn pulsar .
furthermore , the theoretical models are compared with ecsn candidates : sn 1054 and sn 2008s .
we find that sn 1054 shares the characteristics of the ecsne . for sn 2008s
, we find that its faint plateau requires an ecsn model with a significantly low explosion energy of @xmath10 erg . |
understanding the origins of single - spin asymmetries in high - energy processes is one of the main goals in qcd and hadronic physics @xcite .
the longitudinal target spin asymmetry ( ltsa ) @xcite , as the first unambiguous single - spin phenomenon measured in semi - inclusive deep - inelastic scattering ( sidis ) , has attracted a lot of theoretical attentions .
experimentally , the ltsa is usually measured by using the target longitudinally polarized with respect to the incoming beam direction , which makes the interpretation of the asymmetry more complicated .
this is because in this case the measured asymmetry contains contributions from both the transverse and longitudinal polarization components with respect to the virtual photon direction .
the small but non - vanishing transverse polarization component , although suppressed by a factor @xmath6 compared to the longitudinal component , can contribute to the measured asymmetry through the collins @xcite or sivers @xcite effects , which are leading - twist observables .
this feature , especially the former one , has been considered in several estimates @xcite in terms of the transversity distribution and collins fragmentation functions @xcite . apart from this ,
there are also competing contributions that originate from the various convolutions of the twist-3 quark distribution functions or fragmentation functions with their twist-2 counterparts , due to the presence of the large longitudinal component .
therefore , the asymmetry from those contributions , usually denoted by @xmath2 where @xmath7 is the azimuthal angle of the final hadron around the photon direction , is the purely twist-3 results , compared to the leading - twist ones . in this paper , we will study the ltsa based upon the contributions from the subleading - twist effects .
although those effects encode important twist-3 dynamics of the nucleon , they are barely studied in phenomenology and have not been fully considered in the previous theoretical analyses . at the twist-3 level , there are four individual contributions @xcite that can give rise to the ltsa .
two of them are from the the twist-3 fragmentation functions , while another two are from the twist-3 distribution functions . in this work
, we will focus on the effects of the twist-3 quark distributions , that is , the distributions @xmath8 and @xmath9 .
the former one is a t - odd transverse momentum dependent ( tmd ) distribution and can be viewed as the twist-3 analogy of the sivers function ; it generates the ltsa through the coupling with the unpolarized fragmentation function @xmath10 .
the later one is a chiral - odd distribution , therefore its contribution to the ltsa involves another chiral - odd fragmentation function , the collins function @xmath11 @xcite . in our study
we will consider both the @xmath12 term @xcite and the @xmath13 term @xcite .
currently the knowledge on the twist-3 distributions is very limited . in order to quantitatively
predict the ltsa in sidis contributed by the subleading - twist effect , we will employ the spectator - diquark models to calculate the tmd distributions @xmath0 and @xmath1 for the @xmath14 and @xmath15 valence quarks inside the proton .
as the asymmetry we calculate , the @xmath2 , corresponds to the case in which the nucleon is longitudinally polarized with respect to the virtual photon momentum , it is not easy to compare our result directly with the experiments in which the longitudinal spin of the target is originally along the beam direction
. fortunately , there are available data @xmath2 for @xmath3 and @xmath4 production from the hermes collaboration , which are extracted from the original data by subtracting the contributions from the sivers effect and the collins effect , using the data on the transversely polarized target @xcite .
the extraction is valid up to corrections of order @xmath16 , with @xmath17 the polar angle between the incoming beam direction and the virtual photon direction .
new data on @xmath2 are also coming from the clas at the jefferson lab .
these make the comparison between theoretical estimates and experimental measurements possible .
the remained content of the paper is organized as follows . in section
ii , we calculate the twist-3 tmd distributions @xmath0 and @xmath1 using the spectator - diquark models , and we will use two different models to obtain two sets of tmd distributions for comparison . in section iii , we estimate the ltsa contributed by the @xmath12 and @xmath13 terms at the hermes kinematics , and compare the results with the hermes data
. we will also give the prediction on the ltsa at clas and compass .
we summarize the paper in section iv .
+ in this section , we present our calculation on the twist-3 tmd distributions @xmath8 and @xmath9 for the @xmath14 and @xmath15 valence quarks .
the tmd @xmath8 has been calculated in a scalar diquark model @xcite , while @xmath9 has been calculated by the spectator model @xcite and the bag model @xcite . although @xmath9 has the integrated version @xmath18 @xcite , here we trace its @xmath19-dependence in order to calculate its contribution to @xmath2 .
the gauge - invariant quark - quark correlator for the longitudinally polarized nucleon can be expressed as @xmath20}(x,\bm k_t)&=\int { d\xi^- d^2\xi_t\over ( 2\pi)^3}e^{ik\cdot\xi } \langle p s_l|\bar{\psi}_j(0)\mathcal{l}[0 ^ -,\infty^-]\nonumber\\ & \times \mathcal{l}[\bm 0_t,\bm \xi_t]\mathcal{l}[\infty^-,\xi^-]\psi_i(\xi)|p s_l\rangle\ , , \label{phi}\end{aligned}\ ] ] where @xmath21 $ ] , corresponding to the sidis process , denotes that the gauge - link appearing in @xmath22 is future - pointing ; @xmath23 and @xmath24 are the momenta of the struck quark and the target nucleon , respectively .
the tmd distributions @xmath1 and @xmath0 may be obtained from the correlator via the following traces : @xmath25}i\sigma^{+-}\gamma_5],\label{eq : hltr}\\ s_l\frac{{\epsilon}_{t}^{\alpha\rho } k_{t \rho}^{}}{p^+ } \ ,
f_l^{\perp}(x,\bm{k}_{t}^{2 } ) & = -\frac{1}{2}{\operatorname*{tr}\nolimits}[\phi^{[+]}\gamma^{\alpha } ] .
\label{phitr2}\end{aligned}\ ] ] the correlator ( [ phi ] ) can be calculated by the spectator models @xcite .
here we consider the contributions of both the scalar diquark and the axial - vector diquark , and use the dipolar form factor for the nucleon - quark - diquark couplings .
the corresponding diagrams used the in the calculation are shown in fig .
[ fig : diagrams ] . in the lowest - order expansion of the gauge - link , which is equivalent to set @xmath26
, we apply the diagram in the upper panel of fig .
[ fig : diagrams ] to obtain the correlator contributed by the scalar diquark component : @xmath27}{({\boldsymbol{k}_t}^2+l_s^2)^4 } , \label{lophis}\end{aligned}\ ] ] and by the axial - vector diquark component : @xmath28}{({\boldsymbol{k}_t}^2+l_v^2)^4 } , \label{lophiv}\end{aligned}\ ] ] where @xmath29 and @xmath30 are the normalization constant , @xmath31 is the polarization sum ( the propagator ) of the axial - vector diquark , and @xmath32 ( @xmath33 or @xmath34 ) has the form @xmath35 with @xmath36 the cutoff parameters for the quark momentum , and @xmath37 the mass for the diquarks . to calculate the t - odd tmd distribution @xmath0 , one has to consider the nontrivial effect of the gange - link @xcite , that is , the final - state interaction between the struck quark and the spectator . following refs .
@xcite , we perform the calculations initially with abelian gluons and generalize the result to qcd at the end . in our calculation
we expand the gauge - link to one - loop order , as shown by the diagram in the lower panel of fig .
[ fig : diagrams ] .
we note that in refs .
@xcite , an eikonal model including a whole ladder of non - perturbative gluon exchanges was used to describe the final state interactions relevant for t - odd tmds . after some algebra we obtain the expressions for the correlator contributed by the scalar diquark and the arxial - diquark component at this order : @xmath38 \over \bm q_{t}^2 ( ( \bm{k}_t-\bm{q}_t)^2+l_s^2)^2 } , \label{phis1}\\ \phi^{(1)}_{v } ( x,\bm k_t ) & \equiv -i e_q n_v^2 { ( 1-x)^2\over 128\pi^3 ( p^+)^2}{1\over ( \bm{k}_t^2+l_v^2)^2}\nonumber\\ & \times\int { d^2 \bm q_t\over ( 2\pi)^2 } \ , d_{\rho\alpha}(p - k)\ , ( -i\gamma^{+,\alpha\beta } ) \nonumber\\ & \times d_{\sigma\beta}(p - k+q ) \nonumber\\ & \times { \left[({\kern 0.2 em k\kern -0.45em /}-q{\!\!\!/}+m ) \gamma^\sigma(m-{\kern 0.2 em p\kern -0.50em /})\gamma_5{\kern 0.2 em s\kern -0.50em /}\gamma^\rho ( { \kern 0.2 em k\kern -0.45em /}+m)\right ] \over \bm q_t^2 ( ( \bm{k}_t-\bm{q}_t)^2+l_v^2)^2},\end{aligned}\ ] ] where @xmath39 is understood , @xmath40 is the charge for the quarks , @xmath41 is polarization sum of the vector diuqark which will be determined later , and @xmath42 or @xmath43 is the vertex between the gluon and the scalar diquark or the axial - vector diquark : @xmath44\label{gamma},\end{aligned}\ ] ] where @xmath45 denotes the charge of the scalar / axial - vector diquark .
we point out that , although in the calculation we initially apply abelian gluons which is very similar to the case of photons , here @xmath40 and @xmath46 are not necessarily to be the electric charges of the quark and diquark .
in fact they can be viewed as the color charges " which satisfy the relation @xmath47 @xcite .
this result comes from the fact that the proton is a color neutral particle , therefore in an abelian theory the charges of the two components ( quark and diquark ) should have the same size but opposite sign . in the following
we will calculate @xmath1 and @xmath0 using two different spectator models .
the first model is the one developed in ref .
@xcite , the second model is adopted in ref .
there are two differences between these two models , one is the option for the propagator of the axial - vector diquark , the other is the relation between the quark flavors under study and the diquark types .
the two models have also been applied to calculate @xcite the tmd distributions @xmath48 and @xmath49 , which are crucial in the understanding of the beam spin asymmetry measured @xcite in sidis . substituting ( [ lophis ] ) into ( [ eq : hltr ] ) and ( [ phis1 ] ) into ( [ phitr2 ] ) , we obtain the contributions from the scalar diquark to @xmath1 and @xmath0 : @xmath50 , \label{hls } \\
f_l^{\perp\,s } ( x , k_t^{\,2})&=-\frac{n_s^2(1-x)^2}{32\pi^3}{e_s e_q\over4\pi}{1\over l ( l+\vec k_t^{\,2})^3}\nonumber\\ & \times \left[(1-x)((1 + 2x)m^2 + 2 mm -\lambda_s^2)\right .
\nonumber\\ & \left.-(1+x)m_s^2\right],\label{fls}\end{aligned}\ ] ] and the expressions are the same in the two models . in the first spectator model @xcite
we apply here , the propagator @xmath31 is chosen as @xmath51 ^ 2 } \,n_{-\mu } n_{-\nu } , \label{d1}\end{aligned}\ ] ] which is the summation over the light - cone transverse polarizations of the axial - vector diquark @xcite , and has been applied to calculate the leading - twist tmd distributions in ref . @xcite . with the propagator ( [ d1 ] ) ,
we arrive at the following expressions for @xmath1 and @xmath0 from the axial - vector diquark component in the first model : @xmath52,\label{hv1}\\ f_l^{\perp\,v}(x,\bm k_{t}^2 ) & = \frac{n_v^2(1-x)^2}{32\pi^3}{e_v e_q\over4\pi } \biggl\{\frac{1}{l_v^2(\bm k_t^2+l_v^2)^3}\,\nonumber\\ & \times\left[\left(1+x-2x^2\right)m^2 + 2(1-x)mm \right.\nonumber\\ & \left .- m^2-(1+x)m_v^2+x\lambda_v^2\right]\nonumber\\ & -\frac{x}{(1-x)(\bm k_t^2+l_v^2)^2\bm k_t^2}\ln\left({\bm k_t^2+l_v^2\over l_v^2}\right)\biggr\}. \label{fv1}\end{aligned}\ ] ] also , in this model , a general relation between quark flavors and diquark types is adopted @xcite : @xmath53 where @xmath54 and @xmath55 denote the vector isoscalar diquark @xmath56 and the vector isovector diquark @xmath57 , respectively , and @xmath58 , @xmath59 and @xmath60 are the parameters of the model . in ref .
@xcite , these parameters as well as the mass parameters ( such as the diquark masses @xmath37 , cut - off parameters @xmath36 ) are fitted from the zeus @xcite and grsv01 @xcite distribution sets .
particularly , the mass parameters for different vector diquark types are treated differently , that is , the two isospin states of the vector diquark are distinguished . finally , to connect the
abelian " version of the gluon interaction to the qcd color interaction , we apply the replacement @xcite @xmath61 and @xmath62 , following the choice in refs .
. then we obtain the numerical results of @xmath1 and @xmath0 and label them as the set i tmd distributions . in fig .
[ fig : hlflset1 ] we plot the @xmath63-dependence ( at @xmath64 ) and @xmath19-dependence ( at @xmath65 ) of the functions @xmath66 and @xmath67 for @xmath68 and @xmath15 in set i. the second form for the propagator of the vector diquark is @xmath69 which was applied in ref .
. then we obtain the alternative expressions for @xmath70 and @xmath71 : @xmath72,\label{hv2}\\ f_l^{\perp \,v}(x,\bm k_{t}^2 ) & = \frac{n_v^2(1-x)}{64\pi^3}{e_v e_q\over4\pi } \biggl\{\frac{(1-x)}{l_v^2(\bm k_t^2+l_v^2)^3}\,\nonumber\\ & \times\left[(2 + 2x-3x^2)m^2 + 2mm(1-x)-m^2 \right.\nonumber\\ & \left.-(1-x)\lambda_v^2-{x\over(1-x)}\bm k_t^2-{2 + 2x - x^2\over 1-x}m_v^2\right]\nonumber\\ & -\frac{x}{(\bm k_t^2+l_v^2)^2\bm k_t^2}\ln\left({\bm k_t^2+l_v^2\over l_v^2}\right)\biggr\}. \label{fv2}\end{aligned}\ ] ] which are different from the forms obtained in the first model .
different from eq .
( [ ud ] ) , another way that has been applied to construct distributions of the @xmath14 and @xmath15 valence quarks can be expressed as follows @xcite : @xmath73 here the coefficients @xmath74 , @xmath75 and @xmath76 in front of @xmath77s are obtained from the su(4 ) spin - flavor symmetry of the proton wave function . in this case , the mass parameters for different axial diquark are the same .
we use the relation in ( [ set2 ] ) together with the expressions ( [ hls ] ) , ( [ fls ] ) , ( [ hv2 ] ) and ( [ fv2 ] ) to obtain another set of tmd distributions ( labeled as set ii ) . in this calculation
we apply the values for the parameters from ref . @xcite . in fig .
[ fig : hlflset2 ] we plot the @xmath63 dependence ( at @xmath78 ) and @xmath19 dependence ( at @xmath65 ) of @xmath66 and @xmath67 for @xmath68 and @xmath15 in set ii . comparing fig .
[ fig : hlflset1 ] with fig .
[ fig : hlflset2 ] , we can see that the tmd distributions in set i are very different from those in set ii .
in set i the tmd distributions peak at the small @xmath63 region , while in set i they peak at the moderate @xmath63 region .
the @xmath19 dependence of the tmd distributions shows that the @xmath19 widths for @xmath66 and @xmath67 in set i is larger than those in set ii , as the sizes of the tmd distributions in set ii decrease very quickly with increasing @xmath19 .
plane ) , while the detected hadron momentum together with the @xmath79 axis identify the hadron production plane , the longitudinal spin of the nucleon is along the @xmath80 axis . ] in this section , we will perform the phenomenological analysis on the ltsa for pions in sidis scattering on a longitudinally polarized nucleon target : @xmath81 at the kinematics of hermes , clas and compass .
the kinematics of sidis can be expressed by the following invariant variables : @xmath82 here @xmath83 is the momentum of the virtual photon , and @xmath84 is the invariant mass of the hadronic final state .
the reference frame we adopt in this work is shown in fig .
[ sidisframe ] , where the momentum of the virtual photon is along the @xmath79 axis .
the longitudinal polarization of the target is along the @xmath80 axis . in this frame the transverse momentum of pion
is denoted by @xmath85 , while the azimuthal angle of the pion around the momentum of the virtual photon is defined as @xmath86 .
the differential cross section of sidis scattering on a longitudinally polarized target can be expressed as @xcite : @xmath87 where @xmath88 and @xmath89 are the spin - averaged and spin - dependent structure functions , respectively .
the ellipsis stands for the leading - twist contribution to the @xmath90 moment , which will not be considered in this paper .
the ratio of the longitudinal and transverse photon flux is given as : @xmath91 in the parton model , based on the tree - level factorization adopted in ref .
@xcite , the two structure functions in eq .
( [ eq : cs ] ) can be expressed as the convolutions of twist-2 and twist-3 tmd distribution and fragmentation functions . with the help of the notation @xmath92 & = x\sum_q e_q^2\int d^2\bm k_t\int d^2 \bm p_t\delta^2(z\bm k_t-\bm p_t+\bm p_t ) \nonumber\\ & \times w(\bm k_t ,
\bm p_t)f^q(x,\bm k_t^2 ) d^q(z,\bm p_t^2),\end{aligned}\ ] ] and the reference frame we choose , @xmath88 and @xmath89 have the following forms @xcite @xmath93 , \label{eq : fuu}\\ f^{\sin { \phi_h}}_{ul } & = \frac{2m}{q } \ , \mathcal{c}\ , \left[\frac{\boldsymbol{\hat{p}_{t } } \cdot \boldsymbol{p_t}}{z m_h } \left ( { x}\ , h_l h_1^{\perp } + \frac{m_h}{m}\,g_{1l}\ , \frac{\tilde{g^{\perp}}}{z } \right)\right.\nonumber\\ & \left.+\frac{\boldsymbol{\hat{p}_{t}}\cdot \boldsymbol{k_t}}{m}\left({x}\ , f_l^{\perp } d_1-\frac{m_h}{m}\ , h_{1l}^{\perp } \frac{\tilde{h}}{z } \right)\right ] , \label{eq : ful}\end{aligned}\ ] ] where @xmath94 with @xmath95 , and @xmath96 is the mass of the final - state hadron .
the asymmetry @xmath97 as a function of @xmath98 therefore can be expressed as @xmath99 the @xmath63-dependent and the @xmath79-dependent asymmetries can be defined in a similar way . to obtain eq .
[ eq : ful ] , we have assumed that the tmd factorization can be generalized to the twist-3 level .
however , when dealing with higher - twist observables in the tmd framework , one should keep in mind that it is not at all clear from a theoretical point of view if tmd factorization is valid .
in fact , even the tree - level parton model results at twist-3 were questioned , as shown in @xcite .
nevertheless , we would like to adopt a more phenomenological way , i.e. , to use the tree level result in ref .
@xcite to perform the estimate , since there is no alternative theoretical approach for the ltsa so far .
( [ eq : ful ] ) shows that @xmath100 receives four contributions from the convolutions of the twist-3 tmd distribution and fragmentation functions with the twist-2 ones . in the following calculation
, we will apply the wandzura - wilczek approximation @xcite to neglect the contributions from the twist-3 t - odd fragmentation functions @xmath101 and @xmath102 , that is , we assume that the sizes of the functions with a tilde are small . therefore in our analysis
we restrict the scope on the @xmath12 term and the @xmath13 term . for the twist-3 tmd distrbutions @xmath0 and @xmath1
, we apply the results obtained in the previous section . as for the collins functions for different pions ,
we adopt the following relations : @xmath103 where @xmath104 and @xmath105 are the favored and unfavored collins functions , for which we use the standard parameterisation set from ref .
@xcite .
for the tmd fragmentation function @xmath106 that couples with the distribution @xmath0 , we assume its @xmath107 dependence has a gaussian form @xmath108 where @xmath109 is the gaussian width for @xmath110 . following the fitted result in ref .
@xcite , we choose @xmath111 @xmath112 in the calculation . for the integrated fragmentation function @xmath113 , we will adopt the leading order set of the dss parametrization @xcite . finally , throughout the paper , we consider the following kinematical constraints @xcite on the intrinsic transverse momentum of the initial quarks in our calculation : @xmath114 the first constraint in eq .
( [ constraints ] ) is obtained by requiring the energy of the parton to be less than the energy of the parent hadron , while the second constraint arises from the requirement that the parton should move in the forward direction with respect to the parent hadron @xcite . for the region @xmath115
, there are two upper limits for @xmath116 applied in the region @xmath115 at the same time ; it is understood that the smaller one should be chosen .
we adopt the following kinematics to calculate numerically the ssas @xmath2 at hermes @xcite : @xmath117 where @xmath118 is the momentum of the final - state pion . in the left , central , and right panels of fig .
[ fig : hermes ] , we show the results of the ltsas for @xmath3 , @xmath4 and @xmath5 as functions of @xmath63 and @xmath79 .
the thick and thin curves correspond to the asymmetries calculated from the tmd distributions in set i and set ii , respectively .
the dashed curves show the asymmetries contributed by the @xmath12 term , while the solid curves denote the total contribution of the @xmath12 and @xmath13 terms .
the difference between the solid curve and the dashed curve gives the asymmetry from the @xmath13 term . in the case of @xmath3 , @xmath4 production ,
we compare our results with the hermes data @xcite , which have been obtained by subtracting the contribution from the transverse spin component .
the results calculated from both sets of tmd distributions show that the asymmetry for @xmath3 is positive , while that for @xmath4 is negative , consistent with the hermes data . in the case of @xmath3 production , the asymmetry calculated from the set i agree with the data better than from the set ii tmd distributions . in the case of @xmath4 production ,
the asymmetry contributed by the @xmath12 term is positive , while the that contributed by the @xmath13 term is largely negative , leading to a negative asymmetry in total . in the case of @xmath5 production ,
the asymmetries is positive , similar to the case of @xmath3 , but the size is smaller .
an important feature in the @xmath5 asymmetry is that the contribution from the @xmath12 term dominates , and that the one from the @xmath13 term is almost negligible , especially in the @xmath63-dependent asymmetry .
this is very different from the case of charged pions , for which the contributions from the @xmath12 term and the @xmath13 term are comparable . therefore the measurement on the asymmetry @xmath2 for @xmath5 at hermes will provide great opportunity to access the distribution @xmath0 .
the asymmetry @xmath2 for all three pions is also being measured at clas @xcite by using an electron beam ( @xmath119 ) scattered off the longitudinally polarized proton target .
the kinematical cuts at clas applied in our calculation are : @xmath120 in fig .
[ fig : clas ] , we show the asymmetry @xmath2 for @xmath3 , @xmath4 and @xmath5 at clas .
for all three pions the asymmetries are sizable and could be measured at clas . for @xmath3 and @xmath5 production , the asymmetries calculated from set i tmd distributions are larger than those from set ii distributions ;
while for @xmath4 production , the size of the asymmetry from set ii tmd distributions is slightly larger than that from set i. in the case of @xmath5 production , again the @xmath121 term dominates the asymmetries as functions of @xmath63 and @xmath98 ; although there is sizable contribution from the @xmath13 term at the high @xmath79 region .
compass has measured @xcite the @xmath122 azimuthal asymmetries of charged hadrons produced off the longitudinally polarized deuteron target with a 160 gev muon beam , where @xmath123 is the azimuthal angle of the final - state hadron around the momentum of the lepton beam .
the azimuthal asymmetry @xmath2 at compass might be extracted by the method in ref .
thus , we calculate the asymmetry @xmath2 of charged and neutron pions produced off a deuteron target at compass . we adopt the following kinematics in the estimation : @xmath124 in fig .
[ fig : compass ] , we plot the asymmetry @xmath2 for @xmath3 , @xmath4 and @xmath5 vs @xmath63 , @xmath79 and @xmath98 . we find that in the case of deuteron target , again , for @xmath5 production the dominant contribution comes from the @xmath12 term , while the contribution from the @xmath13 term is consistent with zero .
therefore , it would be feasible to extract the flavor dependence of @xmath0 by combining the asymmetries for @xmath5 using a deuteron target together with a proton target .
in this work , we studied the subleading - twist effect on the azimuthal asymmetry of pions produced in sidis with longitudinal polarized nucleon targets .
we considered the contributions from the twist-3 tmd distributions @xmath0 and @xmath1 simultaneously . in our calculation ,
two different sets of twist-3 distribution functions are employed to analyze the asymmetry @xmath2 for @xmath3 , @xmath4 and @xmath5 at hermes , clas and compass . by comparing our results with the hermes measurement in which a proton target is used , we find that the sign of the asymmetries for @xmath3 and @xmath4 is consistent with the hermes data , and the @xmath12 and @xmath13 terms both give substantial contributions to @xmath2 . in the case of @xmath4 production ,
the @xmath13 term is largely negative , leading to the negative sign of the total asymmetry .
we also find that the @xmath12 term dominates in the asymmetry for @xmath5 , especially at hermes and compass where the beam energies are higher .
thus the measurement of @xmath2 for @xmath5 on the proton target , as well as on the deuteron target , will provide great opportunity to access @xmath0 , including its flavor separation .
future data on @xmath2 for charged and neutral pions from clas at 5.5 gev may be used to clarify the role of twist-3 tmd distribution functions on the longitudinal target spin asymmetry .
the author is grateful to h. avakian and w. gohn for useful discussions .
this work is partially supported by the national natural science foundation of china ( grant nos .
11005018 and 11120101004 ) , by the fundamental research funds for central universities ( grant no .
2242012r3007 ) , and by the qing lan project .
v. barone , a. drago , and p. g. ratcliffe , phys .
rep . * 359 * , 1 ( 2002 ) .
u. dalesio and f. murgia , prog .
. phys . * 61 * , 394 ( 2008 ) .
v. barone , f. bradamante , and a. martin , prog .
65 * , 267 ( 2010 ) .
d. boer , m. diehl , r. milner , r. venugopalan , w. vogelsang , a. accardi , e. aschenauer , and m. burkardt _ et al .
_ , arxiv:1108.1713 .
a. airapetian _
( hermes collaboration ) , phys .
84 * , 4047 ( 2000 ) .
a. airapetian _
( hermes collaboration ) , phys .
d * 64 * , 097101 ( 2001 ) .
a. airapetian _
( hermes collaboration ) , phys .
b * 562 * , ( 2003 ) .
a. bacchetta , m. diehl , k. goeke , a. metz , p. j. mulders , and m. schlegel , j. high energy phys .
02 * 2007 * 093 .
a. metz and m. schlegel , eur .
j. a * 22 * , 489 ( 2004 ) .
a. airapetian _
_ ( hermes collaboration ) , phys . rev . lett . *
94 * , 012002 ( 2005 ) .
z. lu and i. schmidt , phys .
b * 712 * , 451 ( 2012 ) .
r. jakob , p. j. mulders , and j. rodrigues , nucl
a626 * , 937 ( 1997 ) .
h. avakian , a.v .
efremov , p. schweitzer , and f. yuan , phys .
d * 81 * , 074035 ( 2010 ) .
a. bacchetta , a. schfer , and j.j yang , phys . lett .
b * 578 * , 109 ( 2004 ) .
l. p. gamberg , g. r. goldstein and m. schlegel , phys .
d * 77 * , 094016 ( 2008 ) .
s. j. brodsky , d. s. hwang , and i. schmidt , phys .
b * 530 * , 99 ( 2002 ) ; nucl .
b642 * , 344 ( 2002 ) .
x. ji and f. yuan , phys .
b * 543 * , 66 ( 2002 ) .
j. c. collins , phys .
b * 536 * , 43 ( 2002 ) .
h. avakian _
( clas collaboration ) , phys .
d * 69 * 112004 ( 2004 ) .
a. airapetian _
( hermes collaboration ) , phys .
b * 648 * , 164 ( 2007 ) .
m. aghasyan _ et al .
_ , phys .
b * 704 * , 397 ( 2011 ) .
w. gohn _ et al . _ [ clas collaboration ] , arxiv:1402.4097 [ hep - ex ] .
s. j. brodsky , d. s. hwang , b. -q .
ma , and i. schmidt , nucl .
phys . * b593 * , 311 ( 2001 ) . s. chekanov _
( zeus collaboration ) , phys .
d * 67 * , 012007 ( 2003 ) .
m. glck , e. reya , m. stratmann , and w. vogelsang , phys .
d * 63 * , 094005 ( 2001 ) .
m. anselmino , m. boglione , u. dalesio , s. melis , f. murgia and a. prokudin , phys .
d * 87 * , 094019 ( 2013 ) .
m. anselmino , m. boglione , u. dalesio , a. kotzinian , f. murgia , and a. prokudin , phys .
d * 71 * , 074006 ( 2005 ) .
d. de florian , r. sassot , and m. stratmann , phys .
d * 75 * , 114010 ( 2007 ) . | we study the single - spin asymmetries of pions produced in semi - inclusive deep - inelastic scattering on the longitudinally polarized nucleon targets .
we particularly consider the effects of the twist-3 transverse - momentum dependent distribution functions @xmath0 and @xmath1 to the asymmetry .
we calculate the asymmetric moment @xmath2 for @xmath3 , @xmath4 and @xmath5 produced off the proton target at hermes and compare the results with the hermes data .
we also present the prediction of the same asymmetries for different pions at the kinematics of clas 5.5 gev on a proton target , as well as at compass on a deuteron target . |
the baxter - belavin elliptic classical @xmath0 matrix is a quasi - classical limit of the famous baxter s elliptic @xmath1 matrix.@xcite it has a unique position in the classification of the classical @xmath0 matrices by belavin and drinfeld.@xcite etingof introduced the elliptic knizhnik - zamolodchikov ( kz ) equation as a system of differential equations satisfied by the twisted trace of product of vertex operators ( intertwining operators of representations ) of affine lie algebras @xmath2.@xcite this system is formally obtained by replacing the rational @xmath0 matrix in the ordinary kz equation by the elliptic one .
kuroki and the author constructed the _ twisted wess - zumino - witten ( wzw ) model _ on elliptic curves and gave to etingof s equations a geometric interpretation as flat connections.@xcite as is well - known , the elliptic @xmath0 matrix degenerates to the trigonometric @xmath0 matrix when the elliptic modulus @xmath3 tends to @xmath4 .
hence it is natural to expect that the twisted wzw model becomes a degenerate model which is related to the trigonometric @xmath0 matrix when the elliptic curve degenerates to a singular curve . the purpose of this and the subsequent papers is to show that this is indeed the case and to analyse the degenerate wzw model . in the present paper
we examine the degenerate wzw model and prove the factorisation theorem of tsuchiya - ueno - yamada type.@xcite the spaces of conformal blocks and conformal coinvariants of the twisted wzw model over the singular curve factorise into those of the orbifold wzw model , which has been intensively studied in other context .
see , e.g. , ref . @xcite . actually ,
though based on a different standpoint , etingof s work@xcite could be considered as a precursor to them . in the forthcoming papers we study the behaviour of the sheaves of conformal blocks and conformal coinvariants of the twisted wzw model around the boundary of the moduli of elliptic curves , especially the factorisation property of these sheaves .
the integrable systems arising from this wzw model , i.e. , the trigonometric gaudin model and the trigonometric kz equations , shall be also analysed . in this work ,
we use the formulation of wzw models in terms of the affine lie algebras .
the reader can translate it into the formulation by the vertex operator algebras@xcite without difficulties .
this paper is organised as follows : after reviewing the twisted wzw model on elliptic curves in [ sec : twwzw - ell ] , we define the trigonometric twisted wzw model and corresponding wzw model on the orbifold in
[ sec : trig , orb - wzw ] .
the space of conformal coinvariants of the trigonometric twisted wzw model is decomposed into the `` direct integral '' of the spaces of conformal coinvariants of the orbifold wzw model .
this general factorisation theorem is proved in [ sec : factorisation1 ] .
if certain conditions are imposed on the modules inserted to the degenerate curve , a refined factorisation theorem can be proved .
this is done in [ sec : factorisation2 ] .
the simplest non - trivial example of the factorisation is given in
[ sec : weyl - mod ] . the last section [ summary+remarks ]
summarises the results and lists several problems .
we use the following notations besides other ordinary conventions in mathematics .
* @xmath5 , @xmath6 : fixed integers .
@xmath7 will be the rank of the lie algebra and @xmath8 will be the number of the marked points on a curve .
* let @xmath9 be an algebraic variety , @xmath10 be a sheaf on @xmath9 and @xmath11 be a point on @xmath9 .
+ @xmath12 , @xmath13 , @xmath14 , @xmath15 : the structure sheaf of @xmath9 , the sheaf of holomorphic 1-forms on @xmath9 , the stalk of @xmath10 at @xmath11 , the maximal ideal of the local ring @xmath16 , respectively .
+ when @xmath10 is an @xmath12-module ; + @xmath17 , @xmath18 : fibre of @xmath10 at @xmath11 , @xmath15-adic completion of @xmath14 , respectively .
* we shall use the same symbol for a vector bundle and for a locally free coherent @xmath12-module consisting of its local holomorphic sections .
in this section we briefly review the twisted wzw model on elliptic curves . in the next section the degeneration of this model , the trigonometric twisted wzw model ,
shall be given .
we follow ref .
@xcite but make slight changes of normalisation and notations because we need to use the results of etingof@xcite later .
we fix an invariant inner product of @xmath19 by @xmath20 define matrices @xmath21 and @xmath22 by @xmath23 where @xmath24
. then we have @xmath25 and @xmath26 .
let @xmath27 be the elliptic curve with modulus @xmath3 : @xmath28 .
we define the group bundle @xmath29 with fibre @xmath30 and its associated lie algebra bundle @xmath31 with fibre @xmath32 over @xmath33 by @xmath34 where the equivalence relations @xmath35 and @xmath36 are defined by @xmath37 ( because @xmath38 , the group bundle @xmath29 is _ not _ a principal @xmath39-bundle .
is a @xmath40 principal bundle .
hurtubise and markman ( for example , see ref .
@xcite ) take this viewpoint , but we prefer to regard it as a group scheme over the elliptic curve by the reason we stated in 0 of ref . @xcite . ] ) let @xmath41 , which satisfies @xmath42 global meromorphic sections of @xmath31 are linear combinations of @xmath43 ( @xmath44 , @xmath45 ) , where @xmath46 is a meromorphic function with quasi - periodicity , @xmath47 the twisted lie algebra bundle @xmath31 has a natural connection , @xmath48 .
we define the invariant @xmath49-inner product on @xmath31 ( regarded as an @xmath49-lie algebra sheaf ) by @xmath50 for sections @xmath51 , @xmath52 of @xmath31 , where the symbol @xmath53 denotes the adjoint representation of the @xmath49-lie algebra @xmath31 .
this inner product is invariant under the translations with respect to the connection @xmath54 : @xmath55 under the trivialisation of @xmath31 defined by the construction , the connection @xmath56 and the inner product @xmath57 on @xmath31 coincide with the exterior derivative by @xmath58 and the inner product defined by respectively .
for each point @xmath11 on @xmath33 , we define lie algebras , @xmath59 where @xmath60 is the sheaf of meromorphic functions on @xmath33 and @xmath61 means the completion of the stalk at @xmath11 with respect to the maximal ideal @xmath15 of @xmath62 . in other words , it is isomorphic to the loop lie algebra @xmath63 , where @xmath64 is the coordinate of @xmath11 , but the isomorphism depends on choices of the coordinate @xmath58 and the trivialisation of @xmath31 .
the subspace @xmath65 \label{def : loop+}\ ] ] of @xmath66 is a lie subalgebra .
let us fix mutually distinct points @xmath67 on @xmath33 whose coordinates are @xmath68 and put @xmath69 .
we shall also regard @xmath70 as a divisor on @xmath33 ( i.e. , @xmath71 ) .
the lie algebra @xmath72 has a 2-cocycle defined by @xmath73 where @xmath74 and @xmath75 is the residue at @xmath76 .
( the symbol `` @xmath77 '' stands for `` cocycle defining the affine lie algebra '' . )
we denote the central extension of @xmath78 with respect to this cocycle by @xmath79 : @xmath80 where @xmath81 is a central element .
explicitly the bracket of @xmath79 is represented as @xmath82 = ( [ a_i , b_i]^\circ)_{i=1}^l \oplus { { { \mathrm c}_{\mathrm a}}}(a , b ) { { \hat k}}\quad \text{for $ a , b\in{{\mathfrak g}}^d$ , } \label{def : aff - alg - str}\ ] ] where @xmath83^\circ$ ] are the natural bracket in @xmath84 .
the lie algebra @xmath85 for a point @xmath11 is nothing but the affine lie algebra @xmath86 of type @xmath87 ( a central extension of the loop algebra @xmath88 ) .
the affine lie algebra @xmath89 can be regarded as a subalgebra of @xmath79 .
the subalgebra @xmath90 of @xmath84 ( cf . ) can be also regarded as a subalgebra of @xmath89 and @xmath79 .
let @xmath91 be the space of global meromorphic sections of @xmath31 which are holomorphic on @xmath33 except at @xmath70 : @xmath92 there is a natural linear map from @xmath91 into @xmath78 which maps a meromorphic section of @xmath31 to its germ at @xmath76 s .
the residue theorem implies that this linear map is extended to a lie algebra injection from @xmath91 into @xmath79 , which allows us to regard @xmath91 as a subalgebra of @xmath79 .
[ def : cc , cb : ell ] the space of _ conformal coinvariants _ @xmath93 and that of _ conformal blocks _ @xmath94 over @xmath33 associated to @xmath89-modules @xmath95 with the same level @xmath96 are defined to be the space of coinvariants of @xmath97 with respect to @xmath91 and its dual : @xmath98 ( in ref .
@xcite , @xmath93 and @xmath94 are called the space of _ covacua _ and that of _ vacua _ respectively . )
in this section we define the degeneration of the twisted wzw model in the previous section , [ sec : twwzw - ell ] , i.e. , the trigonometric twisted wzw model on a singular `` curve '' .
( exactly speaking , the model is defined on a `` stack '' , whatever it is . ) as is shown in the sections coming later , [ sec : factorisation1 ] and [ sec : factorisation2 ] , if we desingularise the `` curve '' , the spaces of conformal coinvariants and conformal blocks are factorised into those of the orbifold wzw model , the definition of which is also given in this section .
when the modulus @xmath3 of the elliptic curve @xmath99 tends to @xmath4 , one of the cycle of the torus is pinched and we obtain a singular curve of the `` croissant bread '' shape . such an object is constructed simply by identifying the north pole ( @xmath100 ) and the south pole ( @xmath101 ) of the riemann sphere @xmath102 . unfortunately , the twisted bundle @xmath31 over @xmath99 does not simply degenerates to a bundle over such a singular curve .
we need to take several steps : 1 .
lift the bundle @xmath31 to a covering space of @xmath99 where the pull - back of @xmath31 become trivial .
the covering is @xmath103-fold and the covering space is an elliptic curve isomorphic to @xmath99 .
2 . take such degeneration of the covering elliptic curve that it becomes the @xmath103-fold covering of the singular curve mentioned above .
this covering space consists of @xmath5 copies of @xmath104 connected to form a ring .
the `` degenerate bundle '' @xmath31 over the singular curve is understood as the trivial bundle over the singular covering space constructed above with the cyclic group action together . -covering . ]
a `` manifold with local finite group action '' is called an _ orbifold _ , but in our case the underlying space has a singularity , hence it should be regarded as a _ stack _ in algebraic geometry .
it is not our task ( and certainly beyond the reach of the author s ability ) to construct a general theory of the wzw model over stacks .
we shall define the twisted wzw model over the degeneration of the elliptic curve below by down - to - earth terminology only .
let us denote the standard coordinate of @xmath102 by @xmath105 .
the cyclic group @xmath106 acts as @xmath107 ( @xmath108 ) and the quotient @xmath109 is an ordinary orbifold .
the degenerate elliptic curve @xmath110 is defined by identifying @xmath101 and @xmath100 ( i.e. , the fixed points of the @xmath111 action ) of @xmath112 . not defining what the twisted bundle @xmath113 over @xmath110 and the twisted bundle @xmath114 over @xmath112 are , we directly define what their meromorphic sections are in the following way .
( in fact , specifying sections is equivalent to defining bundles in algebraic geometry . )
[ def : gtrig , gorb ] a @xmath115-valued meromorphic function @xmath116 on @xmath104 is a meromorphic section of @xmath114 if and only if it satisfies @xmath117 the same @xmath116 is a section of @xmath113 if and only if it satisfies @xmath118 in addition to .
the relations and are degenerate form of .
hereafter we fix mutually distinct points @xmath67 on the orbifold @xmath112 with coordinates @xmath119 ( modulo @xmath111-action ) .
the algebras @xmath120 , @xmath121 , @xmath122 and @xmath123 are defined from @xmath113 and @xmath114 in the same way as @xmath84 and @xmath90 for the elliptic curve , and .
when @xmath124 ( @xmath125 ) , they are isomorphic to @xmath84 and @xmath90 since the definition is local . for the orbifold wzw model ,
we need similar lie algebras @xmath126 , @xmath127 , @xmath128 , @xmath129 sitting at the singular points , @xmath101 and @xmath100 . formally their definition is the same as and with @xmath11 replaced by @xmath101 and @xmath100 and @xmath33 by @xmath112 .
the mode expansion of @xmath126 has a different form from the usual loop algebra , namely , @xmath130 the algebra @xmath128 has the same form .
the central extension of @xmath84 , @xmath131 , is defined in the same way as in [ sec : twwzw - ell ] by using the cocycle .
the central extension of @xmath126 and @xmath128 are defined essentially in the same manner .
difference is that we have to change the normalisation of the central element @xmath132 : @xmath82 = [ a , b]^\circ \oplus { { { \mathrm c}_{\mathrm a}}}_{,0}(a , b ) { { \hat k } } , \quad { { { \mathrm c}_{\mathrm a}}}_{,0}(a , b):= \frac{1}{n}{\mathop{\operatorname{res}}\nolimits}_{t=0 } ( \nabla a|b ) \label{def : aff - alg - str:0}\ ] ] for @xmath133 , where @xmath134^\circ$ ] is the natural loop - algebra bracket in @xmath135 . the origin of the factor @xmath136 shall be explained soon later .
the definition of @xmath137 is exactly the same .
we often denote @xmath126 and @xmath128 by @xmath138 and @xmath139 , following 1 of ref .
@xcite , where @xmath140 is shown to be isomorphic to the ordinary affine lie algebra @xmath86 .
the suffix @xmath22 is put to emphasise the twisting of the algebra by @xmath22 , .
let @xmath70 be the divisor @xmath141 .
the definition of the loop algebra @xmath78 and its central extension does not change so far as @xmath70 does not contain @xmath101 and @xmath100 .
when @xmath70 contains @xmath101 or @xmath100 or both , then the change of the normalisation of the cocycle at @xmath101 and @xmath100 should be taken into account .
( when we talk about @xmath110 or @xmath113 , it is always assumed that @xmath70 contains neither @xmath101 nor @xmath100 . ) instead of the subalgebra @xmath91 , , we define the following subalgebras of @xmath79 : @xmath142 their lie algebra structures are defined by the pointwise lie bracket of @xmath115-valued functions .
it is not trivial that they are subalgebras of @xmath79 , since the residue theorem can not be directly applied . for @xmath143 , it is easy to see that @xmath144 where @xmath145 denotes the action of @xmath146 on the point @xmath76 .
since @xmath147 contains all poles of @xmath148 on @xmath104 , we can apply the residue theorem here .
this proves that @xmath149 defined by is a lie subalgebra of @xmath79 . in the case of @xmath150 ,
when @xmath70 does not contain @xmath101 nor @xmath100 , the proof is the same .
when @xmath70 contains , e.g. , @xmath101 , the central extension should be @xmath132 times @xmath151 the last equality is the consequence of the residue theorem applied to @xmath152 over @xmath104 . hence , also in this case @xmath150 is a subalgebra of @xmath79 .
other cases ( when @xmath70 contains @xmath100 ) are proved in the same way .
this is the reason why we have to put @xmath136 in , which appeared first in ref .
@xcite , lemma 1.1 , where etingof wrote down explicit formulae of the isomorphism of @xmath140 and the ordinary affine lie algebra @xmath86 .
we are now ready to define the spaces of conformal coinvariants and conformal blocks for @xmath110 and @xmath112 .
[ def : cc , cb : trig , orb ] ( i ) the space of _ conformal coinvariants _ @xmath153 and that of _ conformal blocks _ @xmath154 over @xmath110 associated to @xmath89-modules @xmath95 with the same level @xmath96 are defined to be the space of coinvariants of @xmath155 with respect to @xmath149 and its dual : @xmath156 \(ii ) the space of _ conformal coinvariants _ @xmath157 and that of _ conformal blocks _ @xmath158 over @xmath112 associated to @xmath89-modules @xmath95 with the same level @xmath96 are defined to be the space of coinvariants of @xmath155 with respect to @xmath150 and its dual : @xmath159 in this paper we consider only _ smooth modules _ of affine lie algebras @xmath89 , @xmath160 , etc .
: let @xmath161 \oplus { { \mathbb c}}\hat k$ ] be an affine lie algebra .
( we identify @xmath89 , @xmath160 and so on with this algebra by means of suitable isomorphisms . ) a @xmath86-module @xmath162 is called _ smooth _
if for any @xmath163 , there exists @xmath164 such that , for @xmath165 , @xmath166 , and @xmath167 , @xmath168 where @xmath169 is the loop parameter of @xmath170 . not only laurent polynomials but also laurent series in @xmath171[[\xi]]$ ] can act on smooth modules .
therefore , if @xmath95 ( @xmath172 ) are smooth @xmath89 modules , the actions of @xmath149 and @xmath150 on @xmath173 are well - defined .
concrete examples of spaces of conformal coinvariants shall be given in [ sec : weyl - mod ] .
the relation of @xmath174 and @xmath175 is our main subject in this paper and studied in the next sections , [ sec : factorisation1][sec : factorisation2 ] .
how @xmath176 for elliptic curves degenerates to these spaces at the boundary of the moduli space of the elliptic curves shall be clear in the forthcoming papers .
in this section the general factorisation theorem of the space of conformal coinvariants @xmath174 is shown .
the result is not so sharp as that in the next section , but the idea of the factorisation becomes clear . to state the theorem ,
we need the notion of the verma module of @xmath140 , i.e. , @xmath160 and @xmath137 .
recall that @xmath160 has a natural triangular decomposition : @xmath177 where @xmath178 consists of @xmath115-valued power series in @xmath105 with strictly positive powers while @xmath179 consists of series in @xmath105 with negative powers .
@xmath180 is the same as the cartan subalgebra of @xmath115 , which is spanned by @xmath181 ( @xmath182 ) .
since @xmath180 is canonically identified with the cartan subalgebra @xmath183 of @xmath115 , we often drop the superscript @xmath184 .
the algebra @xmath137 has the same decomposition but with opposite powers of @xmath105 : @xmath185 is the subspace of @xmath128 with negative powers of @xmath105 and @xmath186 is that with negative powers of @xmath105 , because @xmath185 is the space of @xmath115 valued functions which vanish at @xmath100 . to avoid repeating the same statements for @xmath160 and @xmath137 , we identify them in the canonical way and denote the above decomposition as @xmath187 and define @xmath188 .
[ def : verma ] ( i ) let @xmath189 is an element of @xmath190 and @xmath191 is an arbitrary complex number .
the _ verma _ module of @xmath140 of highest weight @xmath189 and level @xmath191 is the @xmath140-module @xmath192 where @xmath193 is the left ideal generated by @xmath194 , @xmath195 for @xmath196 and @xmath197 .
the cyclic vector @xmath198 is denoted by @xmath199 .
equivalently @xmath200 is defined as the @xmath140-module induced from the one - dimensional @xmath201-module : @xmath202 where @xmath199 satisfies @xmath203 , @xmath204 and @xmath205 .
( see ref .
@xcite , ( 1.3 ) . )
\(ii ) the @xmath140-module @xmath206 where @xmath207 , is called the _
universal verma _ module of level @xmath191.@xcite the cyclic vector @xmath208 is denoted by @xmath209 .
equivalently @xmath210 is defined as the @xmath140-module induced from the one - dimensional @xmath211-module : @xmath212 where @xmath209 satisfies @xmath213 and @xmath214 . since @xmath215 \subset j_{\gamma,\lambda , k}$ ] and @xmath216 \subset j(0)$ ] ,
we can define right @xmath217-actions on @xmath200 and @xmath210 by @xmath218 \mapsto [ x h ] , \label{right - h - action : verma}\ ] ] for @xmath196 and @xmath219 , where @xmath220 $ ] denotes the equivalence class of @xmath221 modulo @xmath193 or @xmath222 .
the general factorisation theorem is as follows .
hereafter we always assume that @xmath223 do not contain neither @xmath101 nor @xmath100 .
hence @xmath224 is a divisor of @xmath110 as well as that of @xmath112 .
[ thm : general - factorisation ] let @xmath95 be a smooth @xmath89-module of level @xmath191 for @xmath172 and @xmath162 be the tensor product @xmath173 endowed with the @xmath79-module structure . then , @xmath225 here @xmath226 and @xmath227 are the universal verma modules defined above for @xmath228 and for @xmath229 respectively and @xmath230 is the right action of @xmath231 on @xmath232 defined by @xmath233 for @xmath234 , @xmath235 , @xmath236 and @xmath237 .
note that the left action of @xmath150 and the right action of @xmath183 on @xmath238 commute with each other and thus @xmath230 descends to the action on @xmath175 .
this statement is rigorous but not very illuminating , so we rephrase it in a little bit incorrect but more impressive way . by virtue of the poincar - birkhoff - witt theorem , @xmath200 and @xmath210
are isomorphic to @xmath239 and @xmath240 respectively as linear spaces .
being commutative , the universal enveloping algebra @xmath241 is a polynomial ring over the dual space @xmath242 .
hence @xmath210 is spanned by @xmath243 , where @xmath244 and @xmath245 is a polynomial in @xmath246 . evaluating @xmath247 at @xmath248
, we obtain an element of @xmath200 . in this sense
, @xmath210 may be approximately regarded as a direct integral : @xmath249 ( this is not correct since the direct integral should be spanned by @xmath250-functions instead of polynomials . )
using this ( approximate ) identification , we can rewrite @xmath251 as @xmath252 the statement of theorem [ thm : general - factorisation ] becomes @xmath253 where @xmath254 .
the proof of ( from theorem [ thm : general - factorisation ] ) is the same as that of .
the isomorphism is an analogue of the factorisation theorem of ordinary wzw models.@xcite in fact , under certain finiteness conditions for @xmath95 s , almost all @xmath255 vanish and the direct integral is replaced by finite direct sum as we will prove in [ sec : factorisation2 ] . the idea of the proof of theorem [ thm : general - factorisation ] is essentially the same as that of ref .
see also ref .
we make use of techniques in ref .
@xcite . to begin with ,
observe that there is an exact sequence , @xmath256 where @xmath257 using this exact sequence , we prove the theorem in two steps . in the first step , we show that a linear map @xmath258 \in { \operatorname{cc}}_{\mathrm{orb}}(m(0)_0 { \otimes}m { \otimes}m(0)_\infty)$ ] induces an isomorphism @xmath259 in the second step we show that @xmath260 ( cf .
for the definition of @xmath261 ) or , equivalently , @xmath262 hereafter @xmath263 denotes @xmath264 . combining these two statements with the exact sequence ,
we obtain the proof of the theorem . to prove that @xmath265 in is an isomorphism , we use a lemma by kazhdan and lusztig.@xcite [ lem : kl ] let @xmath266 be a lie algebra over @xmath267 and @xmath268 , @xmath269 be its lie subalgebra such that @xmath270 .
assume that an @xmath269-module @xmath271 is given and put @xmath272 .
then the canonical homomorphism @xmath273 induces an isomorphism @xmath274 see
a.7 of ref .
@xcite for the proof .
we apply this lemma to @xmath275 where @xmath276 and @xmath277 are the canonical cyclic vectors of @xmath226 and @xmath227 .
it is easy to see from or that the lie algebra @xmath270 is equal to @xmath278 . since @xmath279 acts trivially on the first and the last factor of @xmath238 , we have @xmath280 on the other hand , @xmath281 hence @xmath282 , which proves by lemma [ lem : kl ]
. next we prove .
namely , we prove the following two claims : @xmath283 to prove , assume @xmath284 and @xmath236 .
we show that @xmath285 belongs to @xmath286 .
since @xmath287 , we have @xmath288 the conditions and imply that @xmath289 .
hence the right hand side of belongs to @xmath290 which proves .
conversely , assume @xmath235 , @xmath236 , @xmath237 and @xmath234 .
we show that @xmath291 belongs to @xmath292 to prove .
thanks to the fact we proved in the first step of this proof of theorem [ thm : general - factorisation ] , there exists @xmath293 such that @xmath294 hence we have @xmath295 modulo @xmath296 . here @xmath297 is an element of @xmath149 satisfying @xmath298 and @xmath299 . when @xmath300 is @xmath181 ( cf . ) , such @xmath297 can be written down explicitly : @xmath301 where @xmath302 is the coordinate of @xmath76 for a certain @xmath303 ( @xmath304 ) .
for general @xmath300 , @xmath297 can be constructed as a linear combination of @xmath305 .
the last expression in is an element of @xmath306 , which proves .
thus as well as theorem [ thm : general - factorisation ] is proved .
the statement of the previous section can be refined for special modules .
in fact , the direct integral in is replaced by finite direct sum when we impose following finiteness conditions on @xmath95 s : for each @xmath172 there exists a finite dimensional @xmath115-module @xmath307 which generates @xmath95 over @xmath89 and @xmath308 . throughout this section
we assume this condition , which automatically implies that @xmath95 s are smooth .
let us denote @xmath309 and the set of @xmath115-weights of @xmath271 by @xmath310 . for @xmath311 , we define @xmath312 they are well - defined since @xmath313 is invertible on @xmath314 .
[ thm : factorisation : finite ] under the above assumption on @xmath95 , @xmath153 is finite dimensional and there is a linear isomorphism @xmath315 where @xmath316 and @xmath317 are inserted at @xmath101 and @xmath100 of @xmath112 respectively . for simplicity of notation , we denote @xmath181 by @xmath318 ( @xmath319 ) .
as mentioned before , they form a basis of @xmath183 .
we also introduce ideals @xmath320 and @xmath321 of the polynomial ring @xmath322 as follows : @xmath323 the main part of the proof of the theorem is to show the following inclusion : @xmath324 where @xmath325 and @xmath326 is the kernel of the canonical surjection @xmath327 , namely , @xmath328 when the inclusion is proved , we shall see that we may replace @xmath251 in by @xmath329 , where @xmath330 and @xmath331 which are finite direct sums of verma modules .
the space of conformal coinvariants @xmath332 decomposes into finite direct sum of @xmath333 and , factoring by @xmath334 , we shall obtain .
the proof of the first statement , @xmath335 , is given after lemma [ lem : decomp - gd ] .
let us show that @xmath336 namely , for @xmath337 and @xmath338 , @xmath339 belongs to @xmath326 .
thanks to the existence of @xmath340 satisfying , we may assume @xmath341 and @xmath342 , because @xmath343 .
hence the problem reduces to show @xmath344 by the following lemma , we have only to show for @xmath345 .
[ lem : decomp - gd ] ( i ) @xmath346 .
( ii ) @xmath347 .
\(i ) the space @xmath78 is spanned by @xmath348 and the elements of the form @xmath349 , where @xmath350 , @xmath351 , @xmath352 is the local coordinate at @xmath353 and @xmath354 sits at the @xmath303-th component .
the latter elements can be replaced with the following which belong to @xmath149 : if @xmath355 , @xmath356 and if @xmath357 , @xmath358 .
thus we have @xmath359 .
it is an exercise of complex analysis to show that @xmath360 .
\(ii ) by the assumption @xmath361 , @xmath362 and the poincar - birkhoff - witt theorem , we have @xmath363 . by this lemma
, every element @xmath163 is decomposed as @xmath364 , @xmath365 , @xmath366 .
the finite - dimensionality of @xmath153 immediately follows from this expression .
note that @xmath149 acts trivially on @xmath276 and @xmath277 .
hence @xmath367 belongs to @xmath368 .
@xmath326 becomes a right @xmath322-module by the action @xmath369 of @xmath234 .
therefore @xmath370 belongs to @xmath326 , which allows us to assume that @xmath371 in .
we further reduces the problem and show for a weight vector in @xmath271 , @xmath372 .
@xmath326 being a right @xmath322-module , as we mentioned above , it is enough to show that @xmath373 for @xmath319 .
since the constant function with value @xmath318 belongs to @xmath150 , we have @xmath374 modulo @xmath326 , which proves and consequently
. similarly we can prove that @xmath375 and the inclusion is proved .
now we have the following commutative diagram with exact rows .
@xmath376 here @xmath377 and @xmath378 .
( the exactness of the middle row is due to theorem [ thm : general - factorisation ] . )
applying the nine lemma in the homological algebra to this diagram ( or chasing the diagram ) , we have @xmath379 it is easy to see that @xmath380 because the left @xmath79-action and the right @xmath183-action on @xmath381 commute and the right @xmath183-action is commutative . by and we have @xmath382 [ lem : barm(0 ) ] @xmath383 ( cf . . ) for two distinct weights @xmath189 and @xmath384 in @xmath310 , we have @xmath385 , since @xmath386 is a basis of @xmath183 and the map @xmath387 is injective .
( see . ) hence the `` chinese remainder theorem '' implies @xmath388 regarding both hand sides as @xmath389-modules , on which @xmath127 acts trivially and @xmath132 acts as @xmath191 , we induce them up to @xmath160-modules and we obtain the first isomorphism in . the second isomorphism in is proved in the same manner .
due to lemma [ lem : barm(0 ) ] we may rewrite as @xmath390 we show that the component @xmath391 of the above decomposition vanishes unless @xmath392 . let us take @xmath393 . for any @xmath394 , we have @xmath395 and @xmath396 , which lead to @xmath397 due to .
hence if @xmath398 , there exists @xmath234 such that @xmath399 , which means @xmath400 .
therefore @xmath401 this completes the proof of theorem [ thm : factorisation : finite ] .
let us directly compute the space of conformal coinvariants for weyl modules as a simple example and see how the factorisation theorem theorem [ thm : factorisation : finite ] decompose @xmath174 into @xmath175 s .
let @xmath307 be a finite - dimensional @xmath115-module and regard them as @xmath402-module of level @xmath191 as before .
the _ weyl module _ is the @xmath89-module induced up from @xmath307 : @xmath403 the tensor product @xmath404 is naturally induced up from @xmath405 and , by virtue of lemma [ lem : decomp - gd ] ( i ) , we have @xmath406 hence the space of conformal coinvariants is isomorphic to @xmath271 itself : @xmath407 on the other hand , let us compute the component @xmath408 of the factorisation , .
put @xmath409 , @xmath410 , where @xmath411 and @xmath412 are subalgebras of @xmath160 and @xmath137 defined at the beginning of [ sec : factorisation1 ] and @xmath413 .
then @xmath414 we apply lemma [ lem : kl ] to these algebras and the @xmath269-module @xmath415 .
the module @xmath416 in the lemma is @xmath417 the space of coinvariants `` @xmath418 '' in the lemma is @xmath419 take a weight vector @xmath420 of weight @xmath384 . the action of @xmath394 on @xmath421 is equal to the multiplication of @xmath422 .
namely , if @xmath423 , @xmath424 , while if @xmath425 , @xmath426 contains @xmath427 .
consequently , @xmath428 the coinvariants `` @xmath429 '' in lemma [ lem : kl ] is @xmath430 because of .
thus we have proved @xmath431 summarising , the factorisation for the weyl modules @xmath432 is the same as the weight space decomposition of @xmath271 , @xmath433 .
we formulated the twisted wzw model on a degenerate elliptic curve @xmath110 and the orbifold @xmath112 , definition [ def : cc , cb : trig , orb ] .
theorem [ thm : general - factorisation ] asserts that the space of conformal coinvariants ( and thus the space of conformal blocks ) on @xmath110 factorises into that on @xmath112 .
this factorisation theorem holds for any smooth modules , but the result can be refined for modules generated by @xmath115-modules , theorem [ thm : factorisation : finite ] .
the factorisation for the case of weyl modules is nothing more than the weight space decomposition of the tensor product of @xmath115-module , ,
. from these results arise several problems : * in ref .
@xcite the factorisation of the space of conformal coinvariants / blocks on singular curves together with the existence of a flat connection ( the kz connection ) leads to the locally freeness of the sheaf of conformal coinvariants / blocks .
one of our next tasks is to follow this line to analyse the structure of the kz connection , which might help finding the integral representation of the solutions of the elliptic kz equations . * the weights @xmath434 and @xmath435 in theorem [ thm : factorisation : finite ]
look quite strange .
for example , @xmath436 when @xmath437 ( @xmath438 ) and @xmath316 can hardly be integrable when @xmath200 is .
but such weights are inevitable due to the charge conservation .
the same kind of phenomenon is implicit but exists also in ref .
@xcite , where the solution of the elliptic kz equations is constructed as a twisted trace of vertex operators .
representation theoretical meaning of @xmath316 is in question .
* we computed the space of conformal coinvariants for the weyl modules in [ sec : weyl - mod ] .
this case is important in the study of the elliptic gaudin model.@xcite the case for integrable modules instead of weyl modules would also be important .
the author expresses his gratitude to akihiro tsuchiya who explained details of his works , michio jimbo , tetsuji miwa , hiroyuki ochiai , kiyoshi ohba , nobuyoshi takahashi and tomohide terasoma for discussion and comments .
he also thanks gen kuroki for discussions during the previous collaborations .
christian schweigert and martin halpern who informed the references for the orbifold wzw model are gratefully acknowledged . the atmosphere and environment of mathematical scienences research institute ( berkeley , u.s.a . ) , institute for theoretical and experimental physics ( moscow , russia ) and the vith international workshop `` conformal field theory and integrable models '' ( chernogolovka , russia ) were very important .
the author thanks their hospitality .
s. hamidi , c. vafa , _ nucl .
_ * b279 * 465513 ( 1987 ) ; l. borisov , m. b. halpern , c. schweigert , _ internat . j. modern phys .
_ * a13 * 125168 , ( 1998 ) ; m. b. halpern , j. e. wang , 97162 ( 2001 ) ; o.ganor , m.b.halpern , c.helfgott , n.a .
obers , _ j. high energy phys . _ * 0212 * 019 , ( 2002 ) ; m.b.halpern , c. helfgott , twisted einstein tensors and orbifold geometry , preprint hep - th/0212275 , to appear in int
. j. modern phys .
a ; j.de boer , m.b.halpern , n.a.obers , _ j. high energy phys .
_ * 0110 * 011 , ( 2001 ) .
k. ueno , in _
vector bundles in algebraic geometry ( durham , 1993 ) _ , ed . by n.
j. hitchin , p. e. newstead and w. m. oxbury , _ london math .
lecture note ser .
_ * 208 * , ( cambridge univ . press , cambridge , 1995 ) pp .
283345 , y. shimizu , k. ueno , _ moduli theory iii _ , ( iwanami , tokyo , 1999 ) gendai suugaku no tenkai series ( in japanese ) ; _ advances in moduli theory _ , ( english translation by american mathematical society , providence , u.s.a . 2002 ) translations of mathematical monographs , * 206 * , iwanami series in modern mathematics . | trigonometric degeneration of the baxter - belavin elliptic @xmath0 matrix is described by the degeneration of the twisted wzw model on elliptic curves .
the spaces of conformal blocks and conformal coinvariants of the degenerate model are factorised into those of the orbifold wzw model . |
stimulated by
t hooft s @xcite and polyakov s @xcite research van nieuwenhuizen et al .
@xcite studied the influence of gravity on magnetic monopoles , but their results do neither guarantee the existence nor the stability of the solutions as shown in @xcite , @xcite , @xcite .
later breitenlohner et al . @xcite and others ( lee et al .
@xcite , ortiz @xcite ) investigated
t hooft s and polyakov s monopoles ( @xcite , @xcite ) in curved space using numerical methods . introducing the dimensionless parameters @xmath0 ( with the planck mass @xmath1 ) and @xmath2 , where @xmath3 and @xmath4 denote the mass of the corresponding yang - mills field and higgs field respectively , breitenlohner et al .
@xcite presented the following results : for @xmath5 the monopole mass varies only over a finite domain .
choosing the parameter @xmath6 solutions seem to exist for @xmath7 . for @xmath8
there are two solutions with different masses .
considering the mass as a function of @xmath9 , one finds that it attains a maximum @xmath10 for @xmath11 , and that it decreases towards @xmath12 asymptotically as @xmath9 tends to @xmath13 .
this situation is shown in * figure 1*. each point represents a solution calculated numerically .
the right plot of * figure 1 * is a magnification of the left plot near the point @xmath14 .
the series of points marked by @xmath15 will be denoted by `` upper branch '' , the part @xmath16 by `` lower branch '' .
= 6.3 cm = 6.3 cm in this paper stability of the nonabelian gravitating magnetic monopoles is investigated using the technique of spherically symmetric perturbations .
stability considerations of similar nature have been conducted by zhou and straumann @xcite .
the stability problem requires the numerical behaviour of the monopole solutions to be known .
the construction of the monopole solutions in question ( @xcite ) is briefly reviewed . in order to investigate their stability the monopoles
are subjected to small perturbations in time , which have a fourier expansion due to its linearity . eliminating the gravitational degrees of freedom , the remaining terms in the action
are written as a quadratic form .
then the application of jacobi s criterion ( @xcite , @xcite ) suggests instability on the branch @xmath15 and stability under spherically symmetric perturbations on the lower branch .
note however that this does not imply stability in general . in the final section
the results are summarized and discussed .
the t hooft polyakov monopole in curved space is a static finite energy solution of an so(3 ) gauge theory with a triplet of scalar fields . using a
coordinate free notation the action reads @xmath17 where @xmath18 , @xmath19 and @xmath20 denote the curvature scalar , the yang - mills field and the scalar field respectively . @xmath21 and @xmath22 are the gravitational constant and the gauge coupling .
@xmath23 is the hodge star operator and @xmath24 is a potential term which will be specified later .
the gravitational field is described by a static spherically symmetric metric tensor given by the line element @xmath25 in schwarzschild coordinates : @xmath26 the minimal ansatz for the yang - mills potential @xmath27 which is spherically symmetric is denoted in polar coordinates and abelian gauge as follows ( @xcite,@xcite , @xcite ) : @xmath28 the @xmath29 , @xmath30 , are the generators of the fundamental representation of so(3 ) . using the potential @xmath27 the field @xmath31 is expressed by @xmath32.\ ] ] the higgs field is chosen to be @xmath33 in order to reduce the four dimensional action to a one dimensional integral the curvature scalar , the yang - mills , and the higgs terms are substituted into the action ( [ wifu ] ) : @xmath34 with @xmath35 the term @xmath36 represents the potential @xmath37 .
the equations of motion ( eymh equations ) deduced from the action ( [ rerewifu ] ) contain the parameters and as well .
their values determine the stability properties of the monopoles .
the `` background solution '' @xmath38 corresponding to global regular monopoles has been obtained solving the equations of motion with a multiple shooting method @xcite .
the static background solution @xmath39 is subjected to a spherically symmetric perturbation @xmath40 , i.e. the perturbed and the background solutions are related as follows : @xmath41 where @xmath39 is the vector of the functions @xmath42 and @xmath43 and @xmath44 denotes the vector @xmath45 of the radial parts of the perturbations .
it is convenient to introduce @xmath43 and @xmath46 : @xmath47 instability in this context means the growth of the exponential term in time .
the analysis is thus reduced to the question whether admissible perturbations , i.e. perturbations fulfilling the boundary conditions , with an imaginary frequency exist . substituting @xmath48 into the action ( [ wifu ] ) one obtains a two dimensional reduced action differing from ( [ rerewifu ] ) only by the time - dependent additional term @xmath49 the part of the reduced action quadratic in the fields and their derivatives turns out to be @xmath50 taking into account the boundary conditions at @xmath51 , the variation with respect to @xmath52 yields @xmath53 where @xmath54 is the integrand of and @xmath55 , and @xmath56 denote the derivatives with respect to @xmath57 and @xmath58 .
( [ dgls ] ) is solved by @xmath59 this result is used to eliminate @xmath60 and @xmath52 in @xmath61 .
partial integration and introduction of @xmath62 and the variable @xmath63 by @xmath64 leads to the normal form of a 2-channel schrdinger problem @xmath61 : @xmath65 here @xmath66 the background solution is unstable , iff the schrdinger problem admits a bound state , that is a solution with @xmath67 .
the standard technique for a 2-channel schrdinger problem is to calculate the equations of motion from ( [ 2-k ] ) , which then define an eigenvalue problem with the eigenvalue @xmath68 . instead of analyzing the spectrum of this operator the positive definiteness of ( [ 2-k ] ) without the last term
is investigated applying jacobi s criterion @xcite , @xcite .
although in @xcite it is formulated for compact intervals only the proof applies in this case to an infinite interval of integration , too , in view of the asymptotic behaviour of @xmath69 .
the discussion of jacobi s criterion in @xcite also provides a normalizable perturbation with negative energy .
this criterion states that the monopole solutions are unstable iff @xmath70
@xmath71 , @xmath72 , @xmath73 , and @xmath74 denote the solutions of two linearly independent initial value problems for the equations of motion in the fields @xmath75 and @xmath76 ( jacobi equation ) .
the initial values are chosen such that they are compatible with the symmetries of the background solution .
a high order runge - kutta - fehlberg method @xcite has been employed for the numerical integration of this system of differential equations .
first we discuss the results in the prasad - sommerfield limit .
the numerical analysis shows that the branch @xmath15 in * figure 1 * is unstable in the sense of linear perturbation theory , since the determinant has a zero in the interior of the integration interval . in *
figure 2 * the determinant ( [ det ] ) is shown for some values of the parameter on the upper branch branch . for @xmath77 ,
that is the value of @xmath78 where the function @xmath79 has a minimum ( see @xcite ) , @xmath80 attains a local maximum .
a second maximum at greater values of @xmath81 is observed for some values of @xmath82 . for near @xmath83 the second
maximum dominates , whereas , if is approximately @xmath13 the left peak is well developed . for decreasing values of ( @xmath84 )
the zero of the determinant moves to the left .
it is convenient to rescale the determinant by @xmath85 in order to present the functions for all -values of the upper branch in one plot ( see * figure 3 * ) .
the pronounced minimum in the left part of * figure 3 * is due to the rescaling of @xmath86 .
it suggests that @xmath87 , i.e. in the neighbourhood of this point the determinant increases more slowly than @xmath88 .
for @xmath83 the determinant has a zero at infinity in agreement with a theorem in @xcite which states that a change of the stability properties can only happen at a point of bifurcation in the masses . from this theorem
it follows that the whole lower branch in * figure 1 * consists of stable solutions , because of the stability of the prasad - sommerfield solution in flat space ( @xmath89 ) .
= 6.3 cm = 6.3 cm = 6.3 cm the bifurcation point is determined using jacobi s criterion
. for increasing values of the method becomes more and more sensitive which means that the error becomes smaller and smaller .
the determinant corresponding to solutions on the lower branch is plotted in * figure 4*. as expected these functions have no zeros .
the curve with the steepest ascent belongs to @xmath90 .
* figure 5 * and * 6 * show the results for some selected values of @xmath91 .
one finds the same qualitative behaviour as for @xmath92 . in * figure 5 *
the determinant is plotted for given and the corresponding maximal @xmath9-values .
the lowest curve belongs to @xmath93 .
* figure 6 * uses the same @xmath94-values , but @xmath9-values on the upper branch .
increasing the zero of the determinant moves to the left .
some numerical results are collected in the following table .
@xmath95 is obtained by linear extrapolation : [ cols="<,<,<,<,<",options="header " , ] = 6.3 cm = 6.3 cm
the author is indebted to p. breitenlohner and d. maison for numerous discussions on the subject and to h. ewen , p. forgcs and b. schmidke for reading the manuscript .
she would also like to thank george lavrelashvili .
99 t hooft g. , _ nucl .
* b 79 * , 276 ( 1974 ) .
polyakov , a. m. , _ jetp lett . _ * 20 * , 194 ( 1974 ) .
van nieuwenhuizen p. , wilkinson d. , perry m. j. , _ phys .
_ * d 13 * , 778 ( 1976 ) .
breitenlohner p. , forgcs p. , maison d. , _ nucl .
* b 383 * , 357 ( 1992 ) .
lee k. , nair v. p. ,
weinberg e. j. , _ phys . rev . _ * d 45 * , 2751 ( 1992 ) .
lee k. , nair v. p. , weinberg e. j. , _ phys .
lett . _ * 68 * , 1100 ( 1992 ) .
ortiz m. e. , _ phys .
_ * d 45 * , 2586 ( 1992 ) .
zhou z .- h .. , straumann n. , _ phys .
* b 237 * , 353 ( 1990 ) .
gelfand i.m .
, fomin s.v . : _ calculus of variations_. englewood cliffs n.j . , prentice - hall ( 1963 ) .
wu t. t. , yang c. n. in : _ properties of matter under unusual conditions_. interscience new york ( 1969 ) .
forgcs p. , manton n. s. , _ commun .
phys _ * 72 * , 15 ( 1980 ) .
breitenlohner p. , forgcs p. , maison d. , preprint mpi - ph/93 - 41 submitted to _ commun .
phys _ ( 1993 ) .
stoer j. , bulirsch r. : _ numerische mathematik 2_. springer - verlag berlin ( 1990 ) .
baacke j. , _
c - particles and fields _ * 53 * , 399 ( 1992 )
. straumann n. : _ general relativity and relativistic astrophysics_. springer - verlag ( 1984 ) . | the behaviour of magnetic monopole solutions of the einstein - yang - mills - higgs equations subject to linear spherically symmetric perturbations is studied . using jacobi s criterion some of the monopoles
are shown to be unstable .
furthermore the numerical results and analytical considerations indicate the existence of a set of stable solutions .
-0.6 cm 0ex 0ex 1.5 mm epsf to * * 1.0truecm * helia hollmann * max - planck - institut fr physik , werner - heisenberg - institut fhringer ring 6 , 80805 munich , germany 1 |
thermodynamic observables in qcd such as the energy density @xmath15 and the pressure @xmath16 as functions of temperature @xmath17 and baryon chemical potential @xmath18 provide fundamental information for studying the physics of relativistic heavy - ion collisions and compact stars .
because of their importance , high precision lattice simulations of @xmath15 and @xmath16 in su(3 ) gauge theory @xcite and in full qcd @xcite on the lattice at finite @xmath17 have been carried out extensively for the past few decades . in most of these studies
the integral method @xcite is adopted , where @xmath15 and @xmath16 are obtained by integrating so - called the interaction measure @xmath19 calculated on the lattice .
recently , a new method to calculate thermodynamic quantities has been proposed @xcite on the basis of the gradient flow @xcite . in this method
, one makes use of the renormalized energy - momentum tensor ( emt ) operator @xmath20 constructed from the `` flowed field '' at nonzero flow - time @xmath21 @xcite .
once emt is defined , @xmath15 and @xmath16 can be calculated by simply taking thermal averages at any given temperature , @xmath22 this method has been tested for the thermodynamics of su(3 ) gauge theory in ref .
@xcite for the first time with @xmath23@xmath24 corresponding to the lattice spacing @xmath25@xmath26 and the spatial ( temporal ) size @xmath27 ( @xmath28 , @xmath29 , @xmath30 ) .
it was found that the @xmath15 and @xmath16 obtained by the gradient flow with small statistics can be comparable to those obtained by the integral method with high statistics .
an extension of this method to full qcd has been also formulated @xcite and numerical results were reported recently @xcite . in the present paper ,
we report high precision calculation of the thermodynamics of su(3 ) gauge theory with the gradient flow .
numerical analyses are carried out with @xmath0@xmath1 corresponding to the lattice spacing @xmath2@xmath3 .
the spatial ( temporal ) sizes are chosen to be @xmath4 , @xmath5 , @xmath6 ( @xmath7 , @xmath8 , @xmath9 , @xmath10 , @xmath11 ) with the aspect ratio @xmath12 .
the double extrapolation , @xmath13 ( the continuum limit ) followed by @xmath14 ( the zero flow - time limit ) , is taken using our high precision data .
we note that such a double limit could not be taken in ref .
@xcite due to the coarse lattice .
the lattice spacing @xmath31 required for these analyses has been determined on the basis of the gradient flow ( see ref .
@xcite and appendix a.1 ) . after taking the double limit
, the final results of @xmath15 and @xmath16 above the critical temperature @xmath32 reach a few percent precision with both statistical error and systematic errors .
the latter errors are associated with the @xmath33 and @xmath34 extrapolations as well as the scale setting and lambda parameter .
our high precision results based on the gradient flow are found to be in good agreement with the previous high precision results with the integral method .
this paper is organized as follows . in the next section
we introduce the gradient flow and the emt operator used in our study . after describing the setup of numerical simulations in sec .
[ sec : setup ] , the numerical results are presented in sec .
[ sec : num ] .
the last section is devoted to a short summary . in appendix
[ sec : scale ] , the analyses of the lattice spacing and the lambda parameter are described in detail .
let us first recapitulate the essential features of the yang
mills gradient flow @xcite and its application to define the renormalized emt @xcite .
the gradient flow of the yang mills field is generated by the differential equation with a flow time @xmath21 , which has a dimension of inverse mass squared , @xmath35 here the yang
mills action @xmath36 and the field strength @xmath37 are composed of the flowed field @xmath38 , which is a function of @xmath21 and the 4-dimensional euclidean coordinate @xmath39 .
color indices are suppressed for simplicity .
the initial condition at @xmath40 is taken to be @xmath41 with @xmath42 being the ordinary gauge field in 4-dimensional euclidean spacetime . with eq .
( [ eq : gf ] ) , the gauge field flows along the steepest descent direction of @xmath36 as @xmath21 increases . at the tree level , eq .
( [ eq : gf ] ) is rewritten as @xmath43 which is the diffusion - type equation .
therefore , the gradient flow for @xmath44 acts as a cooling of the gauge field with the smearing radius @xmath45 in the 4-dimensional euclidean spacetime . in ref .
@xcite , it is proved in pure gauge theory that all composite operators composed of @xmath38 take finite values for @xmath44 .
also , the idea of the gradient flow can be generalized to gauge theory with fermions @xcite . in the present study ,
we use the emt defined by the short flow - time expansion @xcite .
let us consider a composite local operator @xmath46 defined from the field @xmath38 at positive flow time @xmath44 .
the short flow - time expansion @xcite asserts that in the small @xmath21 limit this operator can be written as a superposition of local operators of the original gauge theory at @xmath40 as @xmath47 { } \sum_i c_i(t ) o_i^{\rm r}(x ) , \label{eq : sfte}\end{aligned}\ ] ] where @xmath48 in the right - hand side are renormalized operators of the original gauge theory at @xmath40 with the subscript @xmath49 denoting a set of operators , while @xmath50 are associated c - number coefficients calculable in perturbation theory for small @xmath21 . in order to define the emt using eq .
( [ eq : sfte ] ) , we consider the short flow - time expansion of dimension - four gauge - invariant operators @xcite . in pure gauge theory ,
there are two such operators ; @xmath51 since they are gauge invariant , only gauge invariant operators appear in the right - hand side of eq .
( [ eq : sfte ] ) : such an operator with dimension - zero is an identity operator , while operators with dimension - four are emts @xmath52 . up to this order , the short flow - time expansion of eqs .
( [ eq : u ] ) and ( [ eq : e ] ) thus gives @xmath53 + o(t ) , \label{eq : ut}\\ e(t , x ) & = \left\langle e(t , x)\right\rangle_0 + \alpha_e(t)t_{\rho\rho}(x ) + o(t ) .
\label{eq : et}\end{aligned}\ ] ] we normalize emt so that the vacuum expectation values vanish , @xmath54 .
this determines the coefficient of the unit operator in the right - hand side of eq .
( [ eq : et ] ) to be @xmath55
. the unit operator does not appear in eq .
( [ eq : ut ] ) since @xmath56 is traceless .
contributions from the operators of dimension six or higher are proportional to @xmath21 or higher from the dimensional reason , and thus they are suppressed for small @xmath21 . combining relations eqs .
( [ eq : ut ] ) and ( [ eq : et ] ) , we have @xmath57 with @xmath58 .
\label{eq : tildet^r}\end{aligned}\ ] ] the coefficients @xmath59 and @xmath60 are calculated perturbatively in the @xmath61 scheme in ref .
@xcite , @xmath62 , \label{eq : a_u } \\
\alpha_e(t ) & = \frac{1}{2b_0}\left[1 + 2b_0\bar{s}_2 \bar{g}(1/\sqrt{8t})^2+o(\bar{g}^4)\right ] , \label{eq : a_e}\end{aligned}\ ] ] where @xmath63 denotes the running gauge coupling in the @xmath61 scheme with @xmath64 and @xmath65 with @xmath66 , @xmath67 with @xmath68 .
we note here that ( i ) the right - hand side of eq .
( [ eq : t^r ] ) is independent of the regularization because of its uv finiteness , so that one can take , e.g. , the lattice regularization scheme , and ( ii ) the small @xmath21 expansion of @xmath69 and @xmath70 implies @xmath71 the thermodynamic quantities are obtained from the expectation values of diagonal elements of the emt as in eq .
( [ eq : e , p ] ) .
a combination of @xmath15 and @xmath16 called the interaction measure @xmath72 is related to the trace of the emt ( the trace anomaly ) : @xmath73 also , the entropy density @xmath74 at zero chemical potential is given by @xmath15 and @xmath16 as @xmath75 in the practical numerical analysis , we calculate eq . ( [ eq : tildet^r ] ) on a flowed gauge field with @xmath44 . with finite @xmath31 , the lattice gauge field has to be smeared by the gradient flow sufficiently to suppress the lattice discretization effect .
since the smearing length of the gradient flow is given by @xmath45 , this condition requires @xmath76 . on the other hand
, the value of @xmath45 has to be small enough compared with half the temporal extent of the lattice , @xmath77 , so that the smearing by the gradient flow does not feel the periodic boundary condition .
from these requirements , the measurement has to be performed in the range @xmath78 when eq .
( [ eq : tildet^r ] ) shows approximate linear dependence as shown in eq .
( [ eq : tildett ] ) in a range of @xmath21 given by eq .
( [ eq : t - range ] ) , one can take the small @xmath21 limit and obtain eq .
( [ eq : t^r ] ) .
the linear dependence eq .
( [ eq : tildett ] ) can also be violated for large @xmath21 when the perturbative results of the coefficients in eqs .
( [ eq : a_u ] ) and ( [ eq : a_e ] ) are no longer applicable .
this happens when @xmath45 approaches the lambda parameter @xmath79 . for the measurement of @xmath72
, we have to calculate @xmath80 to carry out vacuum subtraction .
this means that the numerical analysis for vacuum configuration is needed in addition to @xmath81 simulation . on the other hand
, the analysis of @xmath82 , which depends only on the traceless part does not require the vacuum subtraction and hence can be performed solely with a nonzero @xmath17 simulation .
this is an advantage of our method compared with the integral method . without vacuum subtraction
is recently proposed in ref .
simulation parameters @xmath83 , @xmath84 and the number of configurations for nonzero temperature simulations at @xmath85 .
the * symbol in the far right column shows the set of configurations that the corresponding vacuum simulation ( @xmath86 ) is available . [ cols="^,^,>,>,>,^",options="header " , ] by the method ii with pad improvement as a function of lattice spacing @xmath87 .
the continuum limit is shown at @xmath88 .
the finest lattice data at @xmath89 deviates from fitting line .
we note that the continuum extrapolation is consistent within the statistical error without using the result at @xmath90 .
[ fig : lambdamsbariipade ] , scaledwidth=45.0% ] g. boyd , j. engels , f. karsch , e. laermann , c. legeland , m. lutgemeier and b. petersson , nucl .
b * 469 * , 419 ( 1996 ) [ hep - lat/9602007 ] .
m. okamoto _ et al . _
[ cp - pacs collaboration ] , phys .
d * 60 * , 094510 ( 1999 ) [ hep - lat/9905005 ] ; y. namekawa _ et al . _
[ cp - pacs collaboration ] , phys .
d * 64 * , 074507 ( 2001 ) .
t. umeda , s. ejiri , s. aoki , t. hatsuda , k. kanaya , y. maezawa and h. ohno , phys .
d * 79 * , 051501 ( 2009 ) [ arxiv:0809.2842 [ hep - lat ] ] .
s. borsanyi , g. endrodi , z. fodor , s. d. katz and k. k. szabo , jhep * 1207 * , 056 ( 2012 ) [ arxiv:1204.6184 [ hep - lat ] ] .
l. giusti and m. pepe , pos lattice * 2015 * , 211 ( 2016 ) [ arxiv:1511.03710 [ hep - lat ] ] .
s. borsanyi , z. fodor , c. hoelbling , s. d. katz , s. krieg and k. k. szabo , phys . lett .
b * 730 * , 99 ( 2014 ) [ arxiv:1309.5258 [ hep - lat ] ] .
h. suzuki , ptep * 2013 * , no .
8 , 083b03 ( 2013 ) [ erratum : ptep * 2015 * , no .
7 , 079201 ( 2015 ) ] [ arxiv:1304.0533 [ hep - lat ] ] .
m. asakawa _ et al . _
[ flowqcd collaboration ] , phys .
d * 90 * , no . 1 , 011501 ( 2014 ) [ erratum : phys .
d * 92 * , no . 5 , 059902 ( 2015 ) ] [ arxiv:1312.7492 [ hep - lat ] ] .
m. lscher , jhep * 1008 * , 071 ( 2010 ) [ arxiv:1006.4518 [ hep - lat ] ] .
r. narayanan and h. neuberger , jhep * 0603 * , 064 ( 2006 ) [ hep - th/0601210 ] . reviewed in , m. lscher , pos lattice * 2013 * , 016 ( 2014 ) [ arxiv:1308.5598 [ hep - lat ] ] .
s. borsanyi , s. drr , z. fodor , s. d. katz , s. krieg , t. kurth , s. mages and a. schfer _ et al .
_ , arxiv:1205.0781 [ hep - lat ] .
z. fodor , k. holland , j. kuti , d. nogradi and c. h. wong , jhep * 1211 * , 007 ( 2012 ) doi:10.1007/jhep11(2012)007 [ arxiv:1208.1051 [ hep - lat ] ] . h. makino and h. suzuki , ptep * 2014 * , no .
6 , 063b02 ( 2014 ) [ erratum : ptep * 2015 * , no .
7 , 079202 ( 2015 ) ] [ arxiv:1403.4772 [ hep - lat ] ] .
e. itou , h. suzuki , y. taniguchi and t. umeda , pos lattice * 2015 * , 303 ( 2016 ) [ arxiv:1511.03009 [ hep - lat ] ] .
m. asakawa , t. hatsuda , t. iritani , e. itou , m. kitazawa and h. suzuki , arxiv:1503.06516 [ hep - lat ] .
m. lscher and p. weisz , jhep * 1102 * , 051 ( 2011 ) [ arxiv:1101.0963 [ hep - th ] ] .
m. lscher , jhep * 1304 * , 123 ( 2013 ) [ arxiv:1302.5246 [ hep - lat ] ] . l. del debbio , a. patella and a. rago , jhep * 1311 * , 212 ( 2013 ) [ arxiv:1306.1173 [ hep - th ] ] .
z. fodor , k. holland , j. kuti , s. mondal , d. nogradi and c. h. wong , jhep * 1409 * , 018 ( 2014 ) [ arxiv:1406.0827 [ hep - lat ] ] .
n. kamata and s. sasaki , arxiv:1609.07115 [ hep - lat ] .
k. a. olive _ et al . _
[ particle data group collaboration ] , chin .
c * 38 * , 090001 ( 2014 ) and 2015 update .
l. giusti and h. b. meyer , jhep * 1301 * , 140 ( 2013 ) [ arxiv:1211.6669 [ hep - lat ] ] .
m. kitazawa _
[ flowqcd collaboration ] , in preparation .
s. borsanyi , s. drr , z. fodor , c. hoelbling , s. d. katz , s. krieg , t. kurth and l. lellouch _ et al .
_ , jhep * 1209 * , 010 ( 2012 ) [ arxiv:1203.4469 [ hep - lat ] ] .
r. sommer , pos lattice * 2013 * , 015 ( 2014 ) [ arxiv:1401.3270 [ hep - lat ] ] .
m. lscher and s. schaefer , jhep * 1107 * ( 2011 ) 036 [ arxiv:1105.4749 [ hep - lat ] ] .
g. mcglynn and r. d. mawhinney , phys .
d * 90 * ( 2014 ) no.7 , 074502 [ arxiv:1406.4551 [ hep - lat ] ] . m. g. endres , r. c. brower , w. detmold , k. orginos and a. v. pochinsky , phys .
d * 92 * ( 2015 ) no.11 , 114516 [ arxiv:1510.04675 [ hep - lat ] ] . s. aoki , h. fukaya , s. hashimoto and t. onogi , phys . rev .
d * 76 * , 054508 ( 2007 ) doi:10.1103/physrevd.76.054508 m. c , c. consonni , g. p. engel and l. giusti , phys .
d * 92 * ( 2015 ) no.7 , 074502 [ arxiv:1506.06052 [ hep - lat ] ] . m. gckeler , r. horsley , a. c. irving , d. pleiter , p. e. l. rakow , g. schierholz and h. stben , phys .
d * 73 * , 014513 ( 2006 ) [ hep - ph/0502212 ] . | the energy density and the pressure of su(3 ) gauge theory at finite temperature are studied by direct lattice measurements of the renormalized energy - momentum tensor obtained by the gradient flow .
numerical analyses are carried out with @xmath0@xmath1 corresponding to the lattice spacing @xmath2@xmath3 .
the spatial ( temporal ) sizes are chosen to be @xmath4 , @xmath5 , @xmath6 ( @xmath7 , @xmath8 , @xmath9 , @xmath10 , @xmath11 ) with the aspect ratio , @xmath12 .
double extrapolation , @xmath13 ( the continuum limit ) followed by @xmath14 ( the zero flow - time limit ) , is taken using the numerical data . above the critical temperature ,
the thermodynamic quantities are obtained with a few percent precision including statistical and systematic errors .
the results are in good agreement with previous high - precision data obtained by using the integral method . |
mercury is the target of two current space missions ( see e.g. @xcite ) .
the first one , messenger ( nasa ) , already performed three flybys on january 14 , october 6 , 2008 , and september 29 , 2009 , before orbit insertion in march 2011 .
the second one , bepicolombo ( esa / jaxa ) , is planned to be launched in 2014 and to reach mercury in 2020 .
the preparation of these two missions motivated an in - depth study of the rotation of mercury .
the rotation of mercury is a unique case in the solar system because of its @xmath0 spin - orbit resonance , mercury performing exactly 3 rotations during 2 revolutions about the sun @xcite .
it corresponds to an equilibrium state @xcite known as cassini state 1 .
recently , radar earth - based measurements by @xcite detected a 88-day longitudinal libration of mercury with an amplitude @xmath1 of @xmath2 arcsec .
this amplitude being nearly twice too high to be consistent with a rigid mercury , it is the signature of an at least partially molten core .
if we consider mercury as a 2-layered body with a rigid mantle and a spherical liquid core that does not follow the short - period ( @xmath3 days ) excitations and does not interact with the mantle , we can derive from this amplitude the inertia of the mantle plus crust .
in particular , naming @xmath4 the inertial polar momentum of the mantle and @xmath5 the inertial momenta of mercury , we have @xcite : @xmath6 where @xmath7 is a second - degree coefficient of the gravitational potential of mercury , @xmath8 its mass , @xmath9 its radius , and @xmath10 its orbital eccentricity .
it leads to @xmath11 with @xmath12 .
if we take @xmath13 @xcite and @xmath14 @xcite we get @xmath15 .
recent studies in one @xcite and two @xcite degrees of freedom have theoretically estimated the longitudinal librations of mercury .
they highlighted in particular the possibility of a resonance with the jovian perturbation , whose period is @xmath16 years , that could potentially raise the amplitude of a long - term ( @xmath17 years ) libration .
other periodic terms of a few arcsec have been estimated .
this model also predicts that the latitudinal motion of mercury should adiabatically follow the cassini state 1 ( @xcite , @xcite ) , with short - period librations of about 10 milli - arcsec @xcite . in all these studies ,
the core - mantle interactions are neglected .
recently , @xcite explored the dynamics of the rotation of mercury , including core - mantle interactions in the sonyr model @xcite .
we here propose an alternative study , starting from the hamiltonian formulation of @xcite and highlighting the dynamical implications of core - mantle interactions , by considering mercury as composed of a rigid mantle and a triaxial ellipsoidal cavity filled with inviscid fluid of constant uniform density and vorticity .
the differential equations ruling the motion of a 2-layered body with a rigid mantle and a liquid non - spherical core have been derived by @xcite and @xcite .
more recently , @xcite gave a hamiltonian formulation of this problem , that @xcite applied to the rotational dynamics of io , assuming that the core and the mantle were aligned and proportional .
here , we generalize the model of henrard , allowing the core to be non - proportional and non - spherical . [ cols="^,^ " , ] we see on this table the overwhelming predominance of the @xmath18-d contribution , i.e. the rotational period .
as expected , it is excited by the proximity of the @xmath19 secondary resonance with the proper frequency @xmath20 .
we can see that the amplitude associated is roughly the mean value of @xmath21 as can be read from figure [ fig : polarcore ] ( @xmath22 arcsec = @xmath23 arcmin ) .
we also note some similarities with the precession of the rotation axis of the mantle ( table [ tab : polarmantle ] ) , the frequencies involved being the same ones .
in this paper we have investigated the 4-degree of freedom behaviour of a rotating mercury composed of a rigid mantle and a fluid ellipsoidal core , using both analytical and numerical tools with good agreement .
we have emphasized the influence of the proximity of a resonance with the spin of mercury , that can raise the velocity field of the fluid constituting the core .
we can not exclude a possibility of indirect detection of this effect by measuring mercury s magnetic field .
we have also shown the variations of the behaviour of the obliquity of the mantle with respect to the polar flattening of the core , this flattening being linked with the distance of the system from the resonance .
these variations should be negligible unless the core is trapped into the resonance with the spin of mercury .
we have also shown that neither the observations of the longitudinal motion of mercury , nor of its polar one , could be inverted to get the shape of the core .
however , they will give information on its size ( i.e. the parameter @xmath24 ) .
future works should take the viscosity of the fluid into account .
it is assumed to alter the response of the core of mercury to slow ( i.e. @xmath25-year period ) excitations , the planet being therefore assumed as rigid . as a consequence ,
a study of the spectral response of the rotation of mercury on periodic solicitations with respect to the viscosity is worth studying .
this study benefited from the financial support of the contract prodex c90253 `` romeo '' from belspo .
benot noyelles was also supported by the agenzia spaziale italiana ( asi grant `` studi di esplorazione del sistema solare '' ) , and thanks alessandra celletti and luciano iess for their reception in rome .
we also thank nicolas rambaux , marie yseboodt and tim van hoolst for fruitful discussions .
seidelmann p.k . ,
archinal b.a . ,
ahearn m.f . ,
conrad a. , consolmagno g.j . , hestroffer d. , hilton j.l . , krasinsky g.a . , neumann g. , oberst j. , stooke p. , tedesco e.f . , tholen d.j . ,
thomas p.c . , williams i.p .
, 2007 , celes .
astron . , 98 , 155 | mercury is the target of two space missions : messenger ( nasa ) which orbit insertion is planned for march 2011 , and esa / jaxa bepicolombo , that should be launched in 2014 .
their instruments will observe the surface of the planet with a high accuracy ( about 1 arcsec for bepicolombo ) , what motivates studying its rotation .
mercury is assumed to be composed of a rigid mantle and an at least partially molten core .
we here study the influence of the core - mantle interactions on the rotation perturbed by the solar gravitational interaction , by modeling the core as an ellipsoidal cavity filled with inviscid fluid of constant uniform density and vorticity .
we use both analytical ( lie transforms ) and numerical tools to study this rotation , with different shapes of the core .
we express in particular the proper frequencies of the system , because they characterize the response of mercury to the different solicitations , due to the orbital motion of mercury around the sun .
we show that , contrary to its size , the shape of the core can not be determined from observations of either longitudinal or polar motions .
however , we highlight the strong influence of a resonance between the proper frequency of the core and the spin of mercury that raises the velocity field inside the core .
we show that the key parameter is the polar flattening of the core .
this effect can not be directly derived from observations of the surface of mercury , but we can not exclude the possibility of an indirect detection by measuring the magnetic field .
[ firstpage ] planets and satellites : individual : mercury planets and satellites : interior |
diverse physical phenomena such as traffic flow @xcite , force propagation in granular media @xcite , clustering of buses @xcite , aggregation and fragmentation of clusters @xcite , phase separation dynamics @xcite , shaken granular gases @xcite and sandpile dynamics @xcite share one common feature their microscopic dynamics involves stochastic transport of ` mass ' , or some conserved quantity , from one point in space to another . to simplify analysis ,
continuous space is typically replaced by ( or `` binned '' into ) discrete sites .
several such lattice models of stochastic mass transport have been introduced and studied , most notably the zero - range process ( zrp ) @xcite , and the asymmetric random average process ( arap ) @xcite .
these models are defined by specifying the microscopic dynamics , i.e. the basic stochastic rules for mass transport .
given the dynamics , there are two principal theoretical issues : ( i ) to identify the steady state if there is any , i.e. to find the invariant measure and ( ii ) once the steady state is known , to understand various physical properties in the steady state , e.g. the phenomenon of ` condensation ' that happens when a finite fraction of the total mass condenses onto a single site @xcite .
it turns out that the step ( i ) itself is often very difficult and the exact steady state is known in only very few cases @xcite . in many of these known cases , the steady state is _
factorisable_. this means that the steady state probability @xmath2 of finding the system with mass @xmath3 at site 1 , mass @xmath4 at site 2 etc is given by a product of ( scalar ) factors @xmath5 one factor for each of the @xmath0 sites of the system e.g. for a homogeneous system where all sites @xmath6 have equivalent connectivities @xmath7 where @xmath8 is a normalisation which ensures that the integral of the probability distribution over all configurations containing total mass @xmath9 is unity , hence @xmath10 \,\delta \!\left ( \sum_{i=1}^l m_i - m\right ) \;. \label{z}\ ] ] here , the @xmath11-function has been introduced to guarantee that we only include those configurations containing mass @xmath9 in the integral .
the single - site weights , @xmath12 are determined by the details of the mass transfer rules and for a heterogeneous system may depend on the site @xmath6 .
the advantage of having a factorisable steady state is that the step ( ii ) mentioned above is often easier to carry out explicitly .
this has been demonstrated recently by an exact analysis of the condensation phenomenon that occurs in a general class of mass transport models @xcite .
this raises a natural question : when does the steady state in these mass transport models factorise ?
this issue was recently addressed in the context of a sufficiently general ` mass transport model ' in one dimension , that includes , as special cases , the previously studied zrp , arap and the chipping model @xcite . in this model
a mass @xmath13 resides at each site @xmath6 of a one dimensional lattice with periodic boundary conditions . at each time , a stochastic portion , @xmath14 of the mass @xmath13 at site @xmath6 , chosen from a distribution @xmath15 , is chipped off to site @xmath16 . the distribution @xmath17 was called the chipping kernel and by choosing its form appropriately one can recover the zrp , the arap and the chipping model of @xcite .
even though the model above is defined in discrete time where all sites are updated in parallel , by appropriately choosing the chipping kernel it is easy to study the continuous time limit , which corresponds to a random sequential update sequence , as a special case @xcite .
similarly , one can also recover , as a special case , the model with only discrete masses as in zrp .
thus the discrete time dynamics generalises continuous time dynamics but a continuous mass variable generalises discrete mass .
a natural question , first addressed in @xcite , is what should be form of the chipping kernel @xmath17 for the the final steady state to be factorisable . in that study
, it was proved that the _ necessary and sufficient _
condition for a factorised steady state in the one dimensional directed case defined above is that the chipping kernel is of the form @xmath18 where @xmath19 and @xmath20 are arbitrary non - negative functions .
then the single - site weight in eq .
( [ prob ] ) is given by @xmath21 furthermore , given a chipping kernel @xmath17 , sometimes it is hard to verify explicitly that it is of the form ( [ phi1 ] ) and thereby to identify the functions @xmath22 and @xmath23 in order to construct the weight @xmath12 in eq .
( [ ffact ] ) .
this problem was circumvented by devising a test @xcite to check if a given explicit @xmath17 satisfies the condition ( [ phi1 ] ) or not .
further , if it `` passes this test , '' the weight @xmath12 can be found explicitly by a simple quadrature @xcite . finally , for any desired function @xmath12 , one can construct dynamical rules ( i.e. , @xmath24 ) that will yield @xmath12 in a factorised steady state .
it was further demonstrated in ref .
@xcite that the corresponding necessary and sufficient condition in the case of random sequential dynamics in continuous time can be easily obtained , by taking a suitable limit , from the condition for the parallel dynamics manifest in eq .
( [ phi1 ] ) .
this is done by choosing the chipping kernel as @xmath25\ , \delta(\tilde m ) + \alpha(\tilde m|m)\ , dt \label{ct1}\ ] ] for small time increment @xmath26 and @xmath27 is the dirac delta function .
the function @xmath28 denotes the ` rate ' at which a mass @xmath29 is transferred from a site with mass @xmath30 to its right neighbour . note that the form in eq .
( [ ct1 ] ) ensures the normalization , @xmath31 .
then , the necessary and the sufficient condition for factorisable steady state , derived from the more general condition in eq .
( [ phi1 ] ) , is that the rate @xmath28 must be of the following form @xcite @xmath32 where @xmath33 and @xmath20 are arbitrary non - negative functions .
the corresponding steady state weight is then simply , @xmath34 .
the condition ( [ phi1 ] ) for factorisability in the mass transport model was derived only in one dimension and also only for unidirectional mass transport ( from site @xmath6 to site @xmath16 ) . a natural question ,
therefore , is whether one can generalise this condition to higher dimensional lattices , or to arbitrary graphs where mass transport can take place , in general , between any pair of sites @xmath6 and @xmath35 .
recently , greenblatt and lebowitz were able to derive a sufficiency condition @xcite for factorisability in the mass transport model with nearest neighbour mass transport on a regular @xmath1-dimensional lattice with periodic boundary conditions , but considered only the case of random sequential dynamics .
they showed that if @xmath36 is the probability of mass @xmath37 being chipped off a site with mass @xmath30 to a nearest neighbour in the direction @xmath38 ( there being @xmath39 nearest neighbours on a hypercubic lattice in @xmath1 dimensions ) , then the sufficient condition for factorisability is a direct generalization of the condition in eq .
( [ ct2 ] ) , namely that the rate function must be of the form @xmath40 for each @xmath38 , where @xmath41 for each @xmath38 and @xmath20 are arbitrary functions .
the steady state weight is simple , @xmath34 .
however , it was not possible to prove that the condition ( [ gl1 ] ) is also necessary @xcite , except in the case of generalized zero - range processes .
the purpose of this paper is to derive a more general sufficiency condition , valid for arbitrary graphs where the mass transport takes place not necessarily between nearest neighbours and for the more general case of parallel dynamics .
our results boil down to equations ( [ phi ] ) and ( [ cond ] ) .
the former yields a sufficiency condition and the latter an additional consistency condition which must be satisfied . in the special case of random sequential dynamics on a regular hypercubic lattice with nearest neighbour mass transport , our sufficiency condition reduces to the one derived by greenblatt and lebowitz .
our results , however , are considerably more general . the paper is organized as follows . in section 2
, we define the mass transport model on an arbitrary graph . in section 3 , we derive a sufficient condition on the chipping kernels , valid for parallel dynamics on an arbitrary graph @xmath42 , that would gaurantee that the steady state is factorisable .
we show that there are some additional consistency conditions that need to be satisfied in general and we demonstrate explicitly how these conditions are satisfied in several examples . in section 4 , we extend our approach to random sequential dynamics on an arbitrary graph .
we conclude with a summary and discussion in section 5 .
we consider a fixed arbitrary graph @xmath42 consisting of @xmath0 nodes labelled @xmath43 and a set of directed links or channels between certain pairs of nodes . at a given time @xmath44 ,
a node @xmath6 has mass @xmath45 where @xmath13 are continuous variables .
we consider discrete time dynamics where at each step the masses at all the nodes are updated in parallel according to the following rules .
we first define a @xmath46 mass - transfer matrix @xmath47 $ ] as follows .
an element @xmath48 of the matrix @xmath47 $ ] is identically zero at all times if there is no directed link from site @xmath6 to site @xmath35 on @xmath49 .
if there is a directed link from @xmath6 to @xmath35 , then @xmath50 is a non - negative stochastic variable that represents the mass transferred from site @xmath6 to site @xmath35 during one update . in fig .
1 we give an example which we refer to for illustrative purposes throughout this section . the diagonal element @xmath51 represents the mass that stays at site @xmath6 at the end of the single update .
we assume that the dynamics of mass transport conserves the total mass .
thus if @xmath52 represents the masses before the update , by virtue of mass conservation , the row sums of the matrix @xmath47 $ ] are given by , @xmath53 ( see fig .
1 ) . similarly ,
the column sum corresponding to a node @xmath6 , @xmath54 represents the mass at site @xmath6 just after the update . note that , by definition , some elements of the matrix @xmath47 $ ] are permanently zero ( when there is no directed channel available for mass transfer between a pair of sites ) . on the other hand ,
when there is an available channel from @xmath6 to @xmath35 , the matrix element @xmath48 is a stochastic variable which is chosen in the following way .
for each node @xmath6 , we define a generalized ` chipping kernel ' @xmath55 that represents the joint distribution of the masses transported from site @xmath6 in a single update . here the set @xmath56 runs over only those sites @xmath35 which are connected to @xmath6 via a directed link , and in addition it includes the diagonal element @xmath51 . in other words , @xmath56 is simply the set of non - zero elements in the @xmath6-th row of the mass - transfer matrix @xmath47 $ ] .
for example , in fig . 1 where we have a graph of four nodes along with the directed links , we need to define four chipping kernels as follows : @xmath57 , @xmath58 , @xmath59 and @xmath60 respectively .
the chipping kernels must be normalized to unity at all sites @xmath6 , i.e. @xmath61 where it is implied that the index @xmath35 in the sum as well as in the product runs over all the sites that @xmath6 feeds into , including @xmath6 itself .
thus in this model the amount of mass transported from a given site @xmath6 in one update depends only on that site @xmath6 , and not , e.g. on the destination sites to which the mass is transported .
also , we assume that the chipping kernels @xmath55 do not contain the time @xmath44 explicitly .
note also that this chipping kernel @xmath55 generalises that of @xcite sufficiently to include hypercubic lattices@xcite or more complicated graphs .
an example graph with four nodes labelled @xmath62 , @xmath63 , @xmath64 and @xmath65 and directed links between certain pairs of nodes
. the associated @xmath66 mass - transfer matrix @xmath47 $ ] is shown , whose element @xmath48 denotes the stochastic mass transferred from site @xmath6 to site @xmath35 in one single update , provided there is a directed link between the two sites .
if there is no directed link , the corresponding matrix element is always identically zero .
the diagonal element @xmath51 is the amount of mass that stays at site @xmath6 during the update .
the row sum and the column sum associated with any node @xmath6 , @xmath53 and @xmath67 , represent respectively the mass at @xmath6 before and after the update.,width=384 ] the chipping kernels thus specify the dynamics , i.e. the mass update rules .
given these kernels , we next ask what is the steady state joint probability distribution of masses @xmath68 , i.e. @xmath69 .
in particular , our goal is to determine the properties of the chipping kernels @xmath55 required in order to guarantee that the steady state joint probability distribution is factorisable , i.e. of the form @xmath70\ ,
\delta\left(\sum_{i=1}^l m_i -m\right ) \label{gpm1}\ ] ] where the normalization constant is given by @xmath71\ , \delta \left(\sum_{i=1}^l m_i -m\right ) . \label{gpt1}\ ] ] besides , if the steady state factorises as in eq .
( [ gpm1 ] ) , we would also like to know the single - site weights @xmath72 in terms of the prescribed chipping kernels @xmath55 .
note that on an arbitrary graph @xmath42 , the single - site weights @xmath72 are , in general , different from site to site .
hence there is an additional subscript @xmath6 in @xmath72 . on a homogeneous graph , where all sites have equivalent connectivities ,
the weight function @xmath12 does not depend on the site @xmath6 explicitly as in eq .
( [ prob ] ) .
since the dynamics conserves the total mass , at any time @xmath44 we can write @xmath73 where @xmath9 is the total mass and @xmath74 is the unnormalized weight at time @xmath44 .
below , we will first write down the general evolution equation of the weight @xmath74 under the mass transport rules prescribed by the chipping kernels . while the notations that we will use for a general graph @xmath42 may seem a bit complicated , it is instructive to keep the simple example in fig . 1 in mind and use it as a guide to the general notations .
let us consider a single update from time @xmath44 to time @xmath75 .
let @xmath76 denote the masses at time @xmath44 before the update and @xmath77 denote the masses at time @xmath75 after the update . in terms of the elements of the mass - transfer matrix
@xmath47 $ ] , we thus have , @xmath78 and @xmath79 for all @xmath6 . the master equation for the evolution of the weight then reads @xmath80 f(\underline{m},t ) \nonumber\end{aligned}\ ] ] where the product over @xmath35 runs over only the sites to which the site @xmath6 feeds into , i.e. , when there is a directed link between sites @xmath6 and @xmath35 ( note that this set includes the site @xmath6 itself ) . for other sites that are not connected to @xmath6 by a directed link , the corresponding matrix element @xmath81 identically ( see fig . 1 )
and hence they are not integration variables in eq .
( [ meq ] ) .
our strategy now is to assume that the steady state factorises and to determine a sufficent condition for this assumption to hold .
first , we take the @xmath82 limit on both sides of eq .
( [ meq ] ) and assume that the steady state weight factorises @xmath83 and write @xmath84 thus @xmath85 \;. \label{meq2}\ ] ] now we laplace transform this equation to obtain @xmath86 . \label{meq2a}\ ] ] where we have defined @xmath87 .
next we trivially perform the integration over the @xmath13 variables on the rhs of eq .
( [ meq2a ] ) and rearrange the @xmath88 to get @xmath89 \equiv \prod_i x_i(\{s_k\}_i ) \;. \label{lt}\ ] ] where @xmath90 indicates that @xmath91 runs over the sites to which @xmath6 is connected by a directed link , including the site @xmath6 itself . for example , for the graph in fig . 1
, we have @xmath92 one solution ( not the most general as we shall discuss in section [ sec : conc ] ) to ( [ lt ] ) is @xmath93 so that @xmath94 as before , the product in eq .
( [ ansatz ] ) runs over only those sites @xmath35 that the site @xmath6 feeds into ( including itself ) , whereas in eq . ( [ g ] )
the product runs over only those sites @xmath35 that feed onto @xmath6 ( including the site @xmath6 ) . for example , for the graph in fig . 1
, we will have @xmath95 and correspondingly @xmath96 since the rhs of eq .
( [ g ] ) is a product of laplace transforms , its inverse laplace transform @xmath72 therefore has the convolution form @xmath97 ( m)\ ; \label{f}\ ] ] where the notation indicates a product over sites @xmath35 feeding into @xmath6 which is a multiple convolution and the function @xmath98 is the inverse laplace transform of @xmath99 , i.e. @xmath100 again , going back to fig . 1 , we will have , for example , by inverting eq .
( [ ans2fig1 ] ) the following convolution for the site labelled @xmath62 @xmath101 and similarly for the sites labelled @xmath63 , @xmath64 and @xmath65 .
having obtained the form of @xmath102 , let us ask what does this imply for the chipping kernels . to see that , we go back to the definitions of @xmath103 in eq .
( [ lt ] ) . substituting the ansatz for @xmath103 in eq .
( [ ansatz ] ) on the rhs of eq .
( [ lt ] ) we see that for each @xmath6 @xmath104 for example , for the site @xmath62 in fig . 1 , we have @xmath105 and similarly for the sites @xmath63 , @xmath64 and @xmath65 .
( [ lt2 ] ) immediately implies for the general case @xmath106 where @xmath35 runs over the sites which site @xmath6 feeds into and @xmath107 is the laplace inversion of @xmath108 as in eq .
( [ invlt1 ] ) .
for example , for the site @xmath62 in fig .
1 we have @xmath109 substituting eq .
( [ prodp1 ] ) in eq .
( [ phimun ] ) we get the following expression for the chipping kernel @xmath110(m_i)}\;. \label{phi}\ ] ] in the numerator the product is over sites @xmath35 which site @xmath6 feeds into via a directed link connecting @xmath6 to @xmath35 ( this includes site @xmath6 itself ) , and one has the constraint @xmath111 . in the denominator , on the other hand , the product is over sites @xmath91 feeding into site @xmath6 ( including once again the site @xmath6 itself ) . going back again to the example in fig . 1 , we have for the site @xmath62 the following chipping kernel @xmath112 where @xmath113 .
similarly one can easily express the chipping kernels for the other sites @xmath63 , @xmath64 and @xmath65 in terms of the functions @xmath107 .
however this is not quite the end of the story since the chipping kernel @xmath114 must obey a key consistency condition , namely its normalization : eq .
( [ normphi1 ] ) .
this implies from eq .
( [ phi ] ) that for each site @xmath6 @xmath115(m ) = \left [ \prod_{*j}v_{ij } \right](m ) \label{cond}\ ] ] again , the convolution on the lhs is over sites @xmath35 feeding into site @xmath6 but the convolution on the rhs is over sites @xmath35 which @xmath6 feeds into . for example , for the site @xmath62 in fig .
1 , this consistency condition reads @xmath116 one can write down similar consistency conditions for sites @xmath63 , @xmath64 and @xmath65 also . in summary
, a sufficient condition for factorisation is that the chipping kernels are product form as in eq .
( [ phi ] ) where @xmath107 are non - negative functions that satisfy the consistency conditions in eq .
( [ cond ] ) , but otherwise are arbitrary .
if this condition holds , then the steady state will factorise with single - site weight @xmath102 given by the convolution formula in eq .
( [ f ] ) .
these conditions form the central result of this paper .
the conditions arise from making the ansatz in eq .
( [ ansatz ] ) which immediately implies eq .
( [ f ] ) for @xmath72 and eq.([prodp1 ] ) for @xmath117 .
these two equations , when substituted in eq .
( [ phimun ] ) , gives eq .
( [ phi ] ) for the chipping kernel .
finally , eq . ( [ cond ] ) arises from the requirement that the chipping kernel in eq .
( [ phi ] ) is normalized to unity .
condition ( [ phi],[cond ] ) is certainly a sufficient condition . whether it is also a necessary condition
remains a non - trivial issue .
in the special case of one dimensional graph with unidirectional transport , it was explicitly proved in ref .
@xcite that this is also a necessary condition .
similarly , this is true ( see below ) in the case of a complete graph where every site is connected to every other site .
however , it is unlikely , in general , that ( [ phi ] ) is also necessary . in section [ sec : conc ] , we provide an explicit example showing solutions to ( [ lt ] ) which are not of the form ( [ phi ] ) .
although we are unable , for a technical reason to be discussed later , to prove that this class of solutions are also `` legitimate , '' we feel that valid solutions different from ( [ phi ] ) should exist . beyond this simple example , finding the most general valid solution to eq .
( [ lt ] ) for the arbitrary graph seems to pose significant challenges .
let us now discuss several special cases where one can directly verify that the consistency conditions in eq .
( [ cond ] ) are satisfied automatically by the functions @xmath107 due to the nature of the underlying graph @xmath42 , thereby guaranteeing factorisability provided the chipping kernels are of the product form as in eq .
( [ phi ] ) .
* example 1 : mass transport in one dimension over extended range : * consider a one dimensional latice with periodic boundary conditions where mass from a given site can get transported up to a range @xmath118 to the left or to the right . in this case
@xmath119 and we have a translationally invariant chipping kernel i.e. @xmath120 is the same function of @xmath121 arguments for each site .
then the consistency condition ( [ cond ] ) reads @xmath122(m_i)=[v_{i - l , i}*v_{i - l+1,i}*\cdots*v_{i+l-1,i } * v_{i+l , i}](m_i ) \label{1dxr}\ ] ] which is automatically satisfied due to the translational invariance @xmath123 .
* example 2 : symmetric chipping rule : * it is easy to see that if the chipping rules are symmetric , i.e. @xmath124 , then the consistency condition ( [ cond ] ) holds automatically .
* example 3 : a more general chipping rule : * a more general way to meet the condition ( [ cond ] ) is to assume that the chipping rules are such that at each site @xmath6 , for every directed link @xmath125 out of site @xmath6 , there is an incoming directed link @xmath126 to @xmath6 from some other site @xmath91 with @xmath127 . for example , in the @xmath62-d example above , this is achieved through @xmath128 . similarly , one can achieve this on a hypercubic lattice in arbitrary dimensions and on any graph where the number of links out of a site equals the number of links into it .
* example 4 : complete graph with permutationally invariant chipping kernel : * so far we have proved that eq .
( [ phi ] ) is a sufficient condition for factorisability on any arbitrary graph provided we can satisfy the consistency condition ( [ cond ] ) for each site . as discussed before , to prove that this condition is also necessary seems hard for a general graph .
however , one can prove this for a complete graph where every site is connected to every other site via a directed link _ and _ the chipping kernel is permutationally invariant .
this can be proved as follows . on a complete homogeneous graph
the functions @xmath129 in eq .
( [ lt ] ) do not depend explicitly on the site index @xmath6 . moreover , since every site is connected to every other sites and using the fact that the chipping kernel is permutationally invariant eq .
( [ lt ] ) simply becomes @xmath130}^l .
\label{cg1}\ ] ] clearly @xmath131 with @xmath132 arbitrary , is _
one _ solution to eq .
( [ cg1 ] ) and thereby @xmath133^l$ ] . as in the general case , this will then lead to the sufficiency condition ( [ phi ] ) . to prove that @xmath134 is also the most general form of the solution that one can write down for eq .
( [ cg1 ] ) , we take logarithm on both sides of eq .
( [ cg1 ] ) and then take derivatives with respect to @xmath88 and @xmath135 with @xmath136 .
this gives @xmath137 for any @xmath136 .
it is then easy to see that the most general solution of the partial differential equation ( [ cg2 ] ) is of the form , @xmath138 where @xmath139 are arbitrary functions .
since the graph is homogeneous , we also have @xmath140 independent on the site index @xmath6 .
thus @xmath141 with @xmath132 being an arbitrary function , is the only solution of eq .
( [ cg1 ] ) that respects homogeneity .
since this solution finally leads to the sufficiency condition , the uniqueness of its form guarantees then that eq .
( [ phi ] ) is both necessary and sufficient .
note that the consistency condition ( [ cond ] ) is automatically satisfied in this case .
the sufficiency condition in eq .
( [ phi ] ) and the associated consistency condition in eq .
( [ cond ] ) , derived above for parallel update dynamics in discrete time , can be easily extended to the case of random sequential dynamics .
this can be achieved by letting the probability of the chipping event in a time step @xmath142 so that , to leading order in @xmath26 for small @xmath26 , at most one chipping event can occur in the whole graph @xmath42 per update , i.e. the chipping events occur sequentially one per update . in addition , taking @xmath143 one can obtain the continuous time limit where chipping events occur with ` rates ' per unit time .
thus , the corresponding sufficiency condition for the random sequential dynamics will be specified in terms of the chipping ` rate ' kernels , rather than the chipping probablity kernels @xmath144 in parallel dynamics . to translate the sufficiency condition in eq .
( [ phi ] ) valid for the probability kernels into one that is valid for ` rate ' kernels , we first redefine the functions @xmath145 , for all @xmath136 , in the following way @xmath146 where @xmath147 are arbitrary functions .
the diagonal functions @xmath148 are left unchanged . with this re - definition of @xmath107 , the steady state weight in eq .
( [ f ] ) becomes @xmath149 + o(dt^2 ) \label{wrs1}\ ] ] where the sum over @xmath35 in the second term runs over all sites @xmath150 that feed into site @xmath6 on @xmath42 . using the re - defined @xmath107 in eq .
( [ vrs1 ] ) one can similarly rewrite the chipping kernels in eq .
( [ phi ] ) .
since the diagonal elements play a special role , it is convenient to redefine the chipping kernel only in terms of non - diagonal elements , i.e. , @xmath151 where @xmath152 denotes the set of matrix elements in the row @xmath6 without the diagonal element @xmath51 .
we are allowed to get rid of the diagonal element using the row sum , @xmath153 .
for example , for the graph in fig . 1
, we will rewrite , @xmath154 . substituting eq .
( [ vrs1 ] ) in eq .
( [ phi ] ) and taking the limit @xmath155 , one gets @xmath156(m_i)\right]\ , \prod_{j\ne i } \delta(\mu_{ij } ) + \nonumber \\ & + & dt\ , \left[\sum_{j\ne i } \frac{x_{ij}(\mu_{ij})\ , v_{ii}(m_i-\mu_{ij})}{v_{ii}(m_i ) } \prod_{k\ne j}\delta(\mu_{ik})\right ] + o({dt}^2 ) \label{phirs1}\end{aligned}\ ] ] where @xmath157(m)= \int_0^m x(\sigma ) y(m-\sigma ) d\sigma$ ] denotes the convolution integral .
the notation in eq .
( [ phirs1 ] ) may look a bit complicated , but actually it s rather simple .
for example , for the graph in fig .
1 , the kernel in eq .
( [ phirs1 ] ) for the site labelled @xmath62 reads , @xmath158(m_1 ) + [ v_{11}*x_{21}](m_1)\right)\right]\ , \delta(\mu_{12})\,\delta(\mu_{14 } ) + \nonumber\\ & + & dt\ , \left[\frac{x_{12}(\mu_{12 } ) v_{11}(m_1-\mu_{12})}{v_{11}(m_1)}\,\delta(\mu_{14 } ) + \frac{x_{14}(\mu_{14 } ) v_{11}(m_1-\mu_{14})}{v_{11}(m_1)}\,\delta(\mu_{12})\right ] .
\nonumber\end{aligned}\ ] ] taking @xmath155 limit in eq .
( [ phirs1 ] ) we then obtain the continuous time limit where a portion @xmath48 gets chipped off the site @xmath6 with mass @xmath13 , to another site @xmath35 which is connected to @xmath6 by a directed link from @xmath6 to @xmath35 on @xmath42 , with a rate @xmath159 .
also , we notice that in the limit @xmath155 , the steady state weight @xmath72 is given simply from eq .
( [ wrs1 ] ) @xmath160 however , as in the case of parallel update , we need the functions @xmath161 to satisfy certain additional consistency conditions .
this is because the chipping kernel @xmath162 , when integrated over its arguments @xmath152 must give unity . integrating eq .
( [ phirs1 ] ) then gives the required consistency condition that must be satisfied for each node @xmath6 , @xmath163(m_i ) = \sum_{j\ne i } [ x_{ji}*v_{ii}](m_i ) , \label{consrs1}\ ] ] where the sum on the lhs runs over all sites @xmath35 that the site @xmath6 feeds into ( excluding the site @xmath6 itself ) , but on the rhs the sum runs over all sites @xmath35 that feed into site @xmath6 ( excluding the site @xmath6 itself ) . as an example of this condition for the site labelled @xmath62 in fig . 1
, we have @xmath164(m_1 ) + [ x_{14}*v_{11}](m_1)= [ x_{21}*v_{11}](m_1)+ [ x_{31}*v_{11}](m_1 ) .
\label{consrsfig1}\ ] ] similar consistency conditions can be wriiten down for sites @xmath63 , @xmath64 and @xmath65 in fig
thus , in summary , for the mass transport model on an arbitrary graph @xmath42 with random sequential dynamics specified by the rates @xmath165 of mass @xmath166 to be transferred from site @xmath6 to site @xmath35 ( provided there is a directed link between @xmath6 and @xmath35 on @xmath42 ) , a sufficiency condition for the steady state to be factorisable with weights @xmath72 is that @xmath165 is of the form @xmath167 for all @xmath6 , where @xmath161 and @xmath168 are functions that must satisfy the consistency conditions in eq .
( [ consrs1 ] ) , but otherwise arbitrary .
also , the corresponding steady state weight is then given simply by @xmath169 .
note that though formally eq .
( [ condf1 ] ) looks like a direct generalization of the one dimensional condition in eq .
( [ ct2 ] ) , the additional consistency conditions in eq .
( [ consrs1 ] ) are nontrivial to satisfy . for the one dimensional example in eq .
( [ ct2 ] ) , the corresponding consistency condition , @xmath170(m_i)=[x_{i-1,i}*v_i](m_i)$ ] is automatically satisfied due to the translational invariance , @xmath171 . it is easy to verify that the consistency condition in eq .
( [ consrs1 ] ) is automatically satisfied in the four example cases of section [ sec : spec ] ( as it should be since random sequential dynamics is just a limit of the discrete time case ) .
let us do this explicitly for the case where the graph @xmath42 is a homogeneous hypercubic lattice with periodic boundary conditions and mass transfer takes place only between nearest neighbours .
for a hypercubic lattice , from each site @xmath6 there are @xmath39 outgoing links to the @xmath39 nearest neighbours of @xmath6 . similarly , there are exactly @xmath39 incoming links to site @xmath6 from its nearest neighbours . then the sufficiency condition in eq .
( [ condf1 ] ) can be written in a simplified notation @xmath172 where the index @xmath173 runs over the @xmath39 directions @xmath174 , @xmath175 denotes the rate of transfer of mass @xmath166 from site @xmath6 with mass @xmath13 in the direction @xmath38 and @xmath176 denotes the function associated with the link @xmath177 .
similarly , the the consistency condition in eq .
( [ consrs1 ] ) can be rewritten as @xmath178(m_i ) = \sum_{j\in { \rm neighbours\,\,\ , of}\,\,\ , i } [ x_{j ,- q}*v_{ii}](m_i ) .
\label{conscubic}\ ] ] we next use the fact that the lattice is homogeneous , i.e. it is translationally invariant in all directions .
clearly then @xmath179 due to translational invariance in the @xmath38-th direction . in that case
the condition in eq .
( [ conscubic ] ) is clearly satisfied at all @xmath6 .
also , due to the translational invariance , the rate function @xmath180 does not depend on the site index @xmath6 .
by the same requirement , @xmath181 .
thus , the sufficiency condition in eq .
( [ condf1 ] ) simply reads , @xmath182 where @xmath183 and @xmath20 are arbitrary non - negative functions .
the consistency conditions are automatically satisfied as proved above .
the steady state single site weight @xmath5 is simply , @xmath184 , and naturally it does not depend on the site index @xmath6 explicitly .
the condition in eq .
( [ conscubic2 ] ) is precisely that derived by greenblatt and lebowitz .
thus , our general sufficiency condition , valid for an aribtrary graph @xmath42 , recovers this special case when @xmath42 is a homogeneous hypercubic lattice with nearest neighbour mass transport .
we can also check that the steady state for the continuous time zero - range process on an arbitrary graph is recovered .
the zero - range process involves discrete masses and is specified by the rates for a a unit of mass to hop from site @xmath6 to @xmath35 . in the case where these rates are given by @xmath185 where @xmath186 is the total rate for the unit mass leaving site @xmath6 and @xmath187 is the probability that the random destination for a hop from site @xmath6 is @xmath35 , the steady state factorises with single - site weight @xmath188 where @xmath189 is the steady state probability of a single random walker moving from site @xmath6 to @xmath35 with probability @xmath187 @xcite . to make the connection between the forms ( [ zrpag ] ) and ( [ condf1 ] ) we identify @xmath190 inverting the latter equality to express @xmath191 in terms of @xmath192 yields , by virtue of ( [ wrs2 ] ) , the single - site weight ( [ zrpagf ] ) .
finally the consistency condition ( [ consrs1 ] ) becomes @xmath193 and is satisfied with @xmath194 since we have @xmath195
in this work we have derived the sufficent condition ( [ phi ] ) along with a consistency condition ( [ cond ] ) for factorisation of the general ( discrete time , continuous mass ) mass transport model on an arbitrary graph . in this case
the single - site weight is given by ( [ f ] ) .
we gave in section [ sec : spec ] specific examples of geometries where the additional consistency condition associated with the sufficient condition is automatically fulfilled .
moreover on a complete graph with permutationally invariant chipping functions we showed that the sufficent condition is , in fact , also necessary .
of course a significant improvement would be to generalize condition ( [ phi],[cond ] ) to a necessary and sufficient condition . to illustrate the challenges involved in accomplishing this task , we offer another simple example where we can derive a condition that is both necessary and sufficient .
this example , a seemingly trivial generalization of the one in @xcite , involves a one - dimensional lattice with chipping to both nearest neighbour sites and we also assume that the chipping kernel is translationally invariant ( @xmath196 in example 1 of section [ sec : spec ] ) .
then , due to the translational invariance , ( [ lt ] ) becomes @xmath197 using a similar procedure to that employed in example 4 of section [ sec : spec ] taking the logarithm of ( [ nesex ] ) then successive derivatives with respect to @xmath88 and @xmath135 we find that the general solution to ( [ nesex ] ) is @xmath198 where @xmath199 , @xmath200 , @xmath201 and @xmath202 are arbitrary functions , independent of @xmath6 .
inverting the laplace transforms to give @xmath203 would yield a considerably more complicated form for the chipping kernel @xmath204 than ( [ phi ] ) . in principle , therefore , there could be a whole family of chipping kernels , generated by the choice of @xmath202 , which give rise to the same steady state ( i.e. the same single - site weight @xmath12 ) as that for @xmath205 .
thus , @xmath202 could be thought of as a `` gauge function '' .
in addition , for each choice of @xmath202 , one has to ensure the consistency condition , namely that @xmath206 is correctly normalized to unity . on the face of it
, it may appear that this imposes a formidable constraint on the choice of @xmath202 .
however , we now show that this consistency condition does not impose any additional constraint on @xmath202 .
in other words , if the consistency condition is ensured for @xmath205 , then it is automatically satisfied for all other choices of @xmath202 . to see this ,
consider the expression of the chipping kernel ( [ phiex ] ) .
now , the denominator @xmath5 is , of course , independent of the choice of @xmath202 .
so , to prove that the consistency condition of normalization of @xmath24 does not impose any additional constraint on @xmath202 , one has to prove that the integral @xmath207 is independent of the choice of @xmath202 .
now , taking the laplace transform of eq .
( [ h1 ] ) one obtains @xmath208 where we have used definition ( [ lt ] ) in going from the first line to the second and eq .
( [ exgen ] ) in going from the second line to the third .
but this expression on the rhs does not contain the `` gauge function '' @xmath202 thereby proving that the integral of @xmath209 is independent of the choice of @xmath202 .
thus the consistency condition is automatically fulfilled for arbitrary choices of @xmath202 as long as it is ensured for @xmath205 , which is indeed the case as shown in sec .
[ sec : spec ] example 1 .
a more serious constraint on @xmath202 is that the inverse laplace transform of ( [ exgen ] ) must be _ non - negative_. the case @xmath205 in ( [ exgen ] ) obviously imposes the trivial constraint that the inverse laplace transform of the @xmath210 s be non - negative .
it remains to be shown whether there is also a class of @xmath211 which lead to valid chipping kernels .
if it could be shown that this class is non - empty , the one - dimensional @xmath196 example we have discussed would show explicitly that it is _ not _ necessary for @xmath24 to be of the form ( [ phi ] ) . mapping
the extent of this class of @xmath202 ( if indeed it is non - empty ) remains a daunting task .
finally we note that having a factorised steady state opens the door for the study of condensation as in @xcite
. thus one should be able to analyse condensation in various geometries or even on scale - free networks @xcite . | we study a general mass transport model on an arbitrary graph consisting of @xmath0 nodes each carrying a continuous mass .
the graph also has a set of directed links between pairs of nodes through which a stochastic portion of mass , chosen from a site - dependent distribution , is transported between the nodes at each time step .
the dynamics conserves the total mass and the system eventually reaches a steady state .
this general model includes as special cases various previously studied models such as the zero - range process and the asymmetric random average process .
we derive a general condition on the stochastic mass transport rules , valid for arbitrary graph and for both parallel and random sequential dynamics , that is sufficient to guarantee that the steady state is factorisable .
we demonstrate how this condition can be achieved in several examples .
we show that our generalized result contains as a special case the recent results derived by greenblatt and lebowitz for @xmath1-dimensional hypercubic lattices with random sequential dynamics . |
many fascinating magnetic properties of solids are related to quantum effects@xcite . in particular , due to the pauli principle , the interplay between interactions and kinetic energy can induce a strong antiferromagnetic spin exchange .
such exchange leads to a remarkable dynamics for the spin degrees of freedom . on simple structures
, the antiferromagnetic exchange can stabilize an antiferromagnetic order . by variations in dimensionality and connectivity of the lattice
a variety of complex phenomena can arise .
recently , among those effects two fascinating situations in which the interaction strongly favors the formation of dimers have been explored in detail .
the first situation concerns a high dimensional system in which the antiferromagnetic coupling can lead to a spin liquid state made of singlets along the dimers .
in such a spin liquid the application of a magnetic field leads to the creation of triplons which are spin-@xmath2 excitations . the triplons which behave essentially like
itinerant bosons can condense leading to a quantum phase transition that is in the universality class of bose - einstein condensation@xcite ( bec ) .
such transitions have been explored experimentally and theoretically in a large variety of materials , belonging to different structures and dimensionalities@xcite . on the other hand ,
low dimensional systems behave quite differently .
quantum fluctuations are extreme , and no ordered state is usually possible . in many quasi one - dimensional systems the ground state properties are described by luttinger liquid ( ll ) physics@xcite that predicts a _ quasi _ long range order .
the elementary excitations are spin-@xmath0 excitations ( spinons ) .
they behave essentially as interacting spinless fermions .
this typical behaviour can be observed in spin ladder systems in the presence of a magnetic field .
although such systems have been studied theoretically intensively for many years in both zero@xcite and finite magnetic field@xcite , a _ quantitive _ description of the ll low energy physics remained to be performed specially for a direct comparison with experiments . quite recently the remarkable ladder compound@xcite @xmath1 , usually called bpcb ( also known as ( hpip)@xmath3cubr@xmath4 ) , has been investigated .
the compound bpcb has been identified to be a very good realization of weakly coupled spin ladders .
the fact that the interladder coupling is much smaller than the intraladder coupling leads to a clear separation of energy scales . due to this separation
the compound offers the exciting possibility to study both the phase with luttinger liquid properties typical for low dimensional systems _ and _ the bec condensed phase typical for high dimensions .
additionally , the magnetic field required for the realization of different phases lies for this compound in the experimentally reachable range .
the ll predictions have been quantitatively tested for magnetization and specific heat@xcite , nuclear magnetic reasonance@xcite ( nmr ) and neutron diffraction@xcite measurements .
additionally the bec transition and its corresponding order parameter have experimentally been observed by nmr@xcite and neutron diffraction measurements@xcite .
in addition , the excitations of this compound have recently been observed by inelastic neutron scattering@xcite experiments ( ins ) .
these are directly related to the dynamical correlations of spin ladders in a magnetic field .
these dynamical correlations have hardly been investigated so far .
the direct investigation of such excitations is of high interest , since they not only characterize well the spin system , but the properties of the triplon / spinon excitations are also closely related to the properties of some itinerant bosonic / fermionic systems . indeed using such mappings @xcite of spin systems to itinerant fermionic or bosonic systems
, the quantum spin systems can be used as quantum simulators to address some of the issues of itinerant quantum systems .
one of their advantage compared to regular itinerant systems is the fact that the hamiltonian of a spin system is in general well characterized , since the spin exchange constants can be directly measured .
the exchange between the spins would correspond to short range interactions , leading to very good realization of some of the models of itinerant particles , for which the short range of the interaction is usually only an approximation . in that respect quantum spin systems play a role similar to the one of cold atomic gases @xcite , in connection with the question of itinerant interacting systems . in this paper
, we present a detailed calculation of the properties of weakly coupled spin-@xmath0 ladders .
we focus in particular on their dynamics and their low energy physics providing a detailled analysis and a _ quantitative _ description necessary for an unbiased comparison with experiments .
more precisely , we explore the phase diagram of such a system computing static quantities ( magnetization , specific heat , bec critical temperature , order parameter ) and the nmr relaxation rate , using a combination of analytic ( mostly luttinger liquid theory and bethe ansatz ( ba ) ) and numerical ( mostly density matrix renormalization group ( dmrg ) and quantum monte carlo ( qmc ) ) techniques .
we compare our results with the various measurements on the compound bpcb .
a short account of some of these results in connection with measurements on bpcb was previously published in refs . .
we here extend these results and give the details on how the theoretical results were obtained .
motivated by recent experimental measurements , we further investigate the excitation spectra and dynamical correlation functions at high and intermediate energies , for which a theoretical description is very challenging .
we show how for the low energy part of the spectrum it is possible to use the mapping to low energy effective theories such as the ll or to a spin chain which can be solved by bethe ansatz techniques@xcite .
such a technique does not work , however , for energies of the order of the magnetic exchange of the system . in this manuscript
we thus complement such analytical approaches by a dmrg analysis .
we use the recent real - time variant to obtain the dynamics@xcite in real time and the dynamical correlation functions .
the same technique can also be used to obtain finite temperature results@xcite .
this allows to obtain an accurate computation of the excitation spectra and correlation functions in the high energy regime .
we use different analytical approaches to interpret our numerical results .
the plan of the paper is as follows .
[ sec : coupledladder ] defines the model of weakly coupled spin ladders .
its basic excitations and phase diagram are introduced as well as the spin chain mapping which proves to be very helpful for the physical interpretations .
[ sec : methods ] briefly recalls the different analytical ( ll , ba ) and numerical ( dmrg , qmc ) techniques which we used to obtain the results described in the present paper .
[ sec : staticproperties ] gives a detailed characterization of the phase diagram focusing on the static properties and the nmr relaxation rate .
[ sec : dynamicalcorrelation ] presents the computed dynamical correlations of a single spin ladder at different magnetic fields and couplings .
the numerical calculations are compared to previous results ( link cluster expansion , spin chain mapping , weak coupling approach ) and analytical descriptions ( ll , t - j model ) .
[ sec : experimental ] directly compares the computed quantities to experimental measurements .
in particular the theoretical spectra are compared to the low energy ins measurements on the compound bpcb .
it also provides predictions for the high energy part of the ins cross section . finally , sec .
[ sec : conclusions ] summarizes our conclusions and discusses further perspectives .
in this section we introduce the theoretical model of weakly coupled spin-@xmath0 ladders in a magnetic field .
we recall its low - temperature phase diagram , paying special attention to the regime of strong coupling along the rungs of the ladder .
this regime is particularly interesting since it is realized in the spin - ladder compound @xmath1 , customarily called bpcb .
we discuss briefly the energy scales for bpcb in the present section , leaving more detailled discussions for sec .
[ sec : experimental ] .
the hamiltonian we consider is @xmath5 here @xmath6 is the hamiltonian of the single ladder @xmath7 and @xmath8 is the strength of the interladder coupling .
the operator @xmath9 acts at the site @xmath10 ( @xmath11 ) of the leg @xmath12 ( @xmath13 ) of the ladder @xmath7 .
often we will omit ladder indices from the subscripts of the operators ( in particular , replace @xmath14 with @xmath15 ) to lighten notation .
@xmath16 ( @xmath17 ) are conventional spin-@xmath0 operators with @xmath18=is^z_{l , k}$ ] , and @xmath19 . the hamiltonian @xmath6 of the spin-@xmath0 two - leg ladder illustrated in fig .
[ fig : singleladder ] is @xmath20 where @xmath21 ( @xmath22 ) is the coupling constant along the rungs ( legs ) and @xmath23 the magnetic field , @xmath24 is applied in the @xmath25 direction , and @xmath26 is the @xmath25-component of the total spin operator @xmath27 .
since @xmath6 has the symmetry @xmath28 , @xmath29 , we only consider @xmath30 .
the relation between @xmath31 and the physical magnetic field in experimental units is given in eq . .
( @xmath32 ) is the coupling along the rungs ( legs ) represented by thick ( thin ) dashed lines and @xmath33 are the spin operators acting on the site @xmath10 of the leg @xmath13 .
[ fig : singleladder ] ] in the present paper we focus on the case of spin-@xmath0 antiferromagnetic ladders weakly coupled to one another .
this means that the interladder coupling @xmath34 is much smaller than the intraladder couplings @xmath35 and @xmath36 , i.e. @xmath37 as we will show , the model accurately describes the magnetic properties of the compound bpcb .
its detailed description is given in sec .
[ sec : experimental ] .
the couplings have been experimentally determined to be@xcite @xmath38 and@xcite @xmath39 more details about the determination of the couplings for the compound bpcb are given in sec .
[ sec : experimental ] .
the physical properties of a single ladder are defined by the value of the dimensionless coupling @xmath40 in the limit @xmath41 ( therefore @xmath42 ) the rungs of the ladder are decoupled .
we denote this _ decoupled bond limit _ hereafter .
the four eigenstates of each decoupled rung are : the singlet state @xmath43 with the energy @xmath44
spin @xmath45 and @xmath25-projection of the spin @xmath46 , and three triplet states @xmath47 with @xmath48 @xmath49 , and energies @xmath50 , @xmath51 , @xmath52 , respectively .
the ground state is @xmath53 below the critical value of the magnetic field , @xmath54 , and @xmath55 above .
the dependence of the energies on the magnetic field is shown in fig .
[ fig : phasediagramm].a .
, @xmath56 , @xmath57 ( solid lines ) and singlet @xmath58 ( dashed line ) versus the applied magnetic field in the absence of an interrung coupling ( @xmath41 ) .
the dotted lines represent the limits of the triplets excitation band when @xmath59 .
( b ) phase diagram of weakly coupled spin ladders : crossovers ( dotted lines ) and phase transition ( solid line ) that only exists in the presence of an interladder coupling are sketched.[fig : phasediagramm ] ] a small but finite @xmath60 delocalizes triplets and creates bands of excitations with a bandwidth @xmath61 for each triplet branch .
this leads to three distinct phases in the ladder system depending on the magnetic field : * _ spin liquid phase _ , which is characterized by a spin - singlet ground state ( see sec .
[ sec : criticalfields ] ) and a gapped excitation spectrum ( see sec .
[ sec : spinliquidexcitations ] ) .
this phase appears for magnetic fields ranging from @xmath62 to @xmath63 * _ gapless phase , _ which is characterized by a gapless excitation spectrum .
it occurs between the critical fields @xmath64 and @xmath65 . the ground state magnetization per rung , @xmath66 , increases from @xmath62 to @xmath2 for @xmath31 running from @xmath64 to @xmath65 .
the low energy physics can be described by the ll theory ( see sec .
[ sec : luttinger_liquid ] ) .
* _ fully polarized phase , _ which is characterized by the fully polarized ground state and a gapped excitation spectrum .
this phase appears above @xmath67 . besides ladders the transition between ( i ) and
( ii ) can occur in several other gapped systems such as haldane @xmath68 chains or frustrated chains@xcite . in the gapless phase , the distance between the ground state and the bands @xmath69 and @xmath70 which is of the order of @xmath71 , is much larger than the width of the band @xmath72 , since @xmath73 . for small @xmath74 the ladder problem
can be reduced to a simpler spin chain problem .
the essence of the _ spin chain mapping_@xcite is to project out @xmath69 and @xmath75 bands from the hilbert space of the model .
the remaining states @xmath53 and @xmath76 are identified with the spin states @xmath77 the local spin operators @xmath33 can therefore be identified in the reduced hilbert space spanned by the states with the new effective spin-@xmath0 operators @xmath78 : @xmath79 the hamiltonian reduces to the hamiltonian of the spin-@xmath0 xxz heisenberg chain @xmath80 here the pseudo spin magnetization is @xmath81 , the magnetic field @xmath82 and the anisotropy parameter @xmath83 note that the spin chain mapping constitutes a part of a more general strong coupling expansion of the model , as discussed in the appendix [ sec : tjmodelmapping ] . for the compound
bpcb the parameter @xmath74 is rather small @xmath84 and the spin chain mapping gives the values of many observables decently well .
some important effects are , however , not captured by this approximation .
examples will be given in later sections .
let us now turn back to the more general hamiltonian and discuss the role of a weak interladder coupling @xmath8 ( couplings ordered as in eq . ) .
the spin liquid and fully polarized phases are almost unaffected by the presence of @xmath8 whenever the gap in the excitation spectrum is larger than @xmath8 ( see , e.g. , ref . for more details ) .
however , a new 3d antiferromagnetic order in the plane perpendicular to @xmath31 emerges in the gapless phase for @xmath85 the corresponding phase , called _ 3d - ordered _ , shows up at low enough temperatures @xmath86 in numerous experimental systems with reduced dimensionality and a gapless spectrum @xcite .
for the temperature @xmath87 the ladders decouple from each other and the system undergoes a deconfinement transition into a luttinger liquid regime ( which will be described in sec .
[ sec : luttinger_liquid ] ) .
for @xmath88 the rungs decouple from each other and the system becomes a ( quantum disordered ) paramagnet .
all the above mentioned phases are illustrated in fig .
[ fig : phasediagramm].b .
in this section , we present the methods used to study the ladder system and its mean - field extension to the case of weakly coupled ladders .
we first focus on the so called density matrix renormalization group ( dmrg ) or matrix product state ( mps ) methods .
these numerical methods allow us to investigate dynamical correlations at zero and finite temperature .
additionally we discuss the bethe ansatz used to obtain properties of the system after the spin chain mapping .
furthermore we introduce an analytical low energy description for the gapless phase , the luttinger liquid theory . this theory in combination with a numerical determination of its parameters ( see appendix [ sec : llappendix ] ) gives a quantitative description of the low energy physics .
finally , we treat the weak interladder coupling @xmath89 by a mean field approach , both analytically and numerically , and a quantum monte carlo ( qmc ) technique .
a numerical method used to determine static and dynamical quantities at zero and finite temperature of a quasi one - dimensional system is the dmrg .
this method was originally introduced by s.r .
white @xcite to study static properties of one dimensional systems .
since usually the dimension of the total hilbert space of a many - body quantum system is too large to be treated exactly , the main idea of the dmrg algorithm is to describe the important physics using a reduced effective space .
this reduced effective space is chosen optimally by using a variational principle .
the dmrg has been proven very successfully in many situations and has been generalized to compute dynamical properties of quantum systems using different approaches in frequency space @xcite .
recently the interest in this method even increased after a successful generalization to time - dependent phenomena and finite temperature situations @xcite .
the real - time calculations give an alternative route to determine dynamical properties of the system@xcite which we use in the following .
an overview of the method , its extensions and its successful applications to real - time and finite temperature can be found in refs . .
further details on the method and its technical aspects are given in the appendix [ sec : dmrg ] .
the spin-1/2 xxz chain which is obtained after the spin chain mapping of the system is exactly solvable : the so - called bethe ansatz technique gives explicit analytic expressions for its eigenfunctions and spectrum@xcite . to convert this information into a practical recipe of calculation of the correlation functions is a highly sophisticated problem .
however , a known solution to this problem ( ref . and
references therein ) incorporates involved analytics and numerics , the latter limiting the precision of the final results to about the same extent as to - date implementations of the dmrg method . calculation of the thermodynamic properties of the spin-1/2 xxz chain model by the bethe ansatz technique is a simpler , but still non - trivial task , requiring a solution of an infinite set of non - linear coupled integral equations@xcite .
the solution of such equations can be only found numerically , and already in the 1970s this was done with a high precision@xcite . later on , an alternative to the bethe ansatz , the quantum transfer matrix method ,
was used to get the thermodynamics of the xxz chain in a magnetic field@xcite . within this approach
the free energy of the system is expressed through the largest eigenvalue of the transfer matrix .
this largest eigenvalue is given by the solution of a set of non - linear equations ( eq .
( 66.a ) of ref . )
which are written in a form very suitable for solving them iteratively .
in the present paper we followed this route and got the results for the specific heat for all temperatures and various magnetic fields with very high precision , see sec .
[ sec : specificheat ] .
the luttinger liquid hamiltonian governs the dynamics of the free bosonic excitations with linear spectrum and can be written as@xcite @xmath90,\ ] ] where @xmath91 and @xmath92 are canonically commuting bosonic fields , @xmath93=i\pi\delta(x - y).$ ] the dimensionless parameter @xmath94 entering eq .
is customarily called the luttinger parameter , and @xmath95 is the propagation velocity of the bosonic excitations ( velocity of sound ) .
many 1d interacting quantum systems belong to the luttinger liquid ( ll ) universality class : the dynamics of their low - energy excitations is governed by the hamiltonian and the local operators are written through the free boson fields @xmath92 and @xmath91 ( the latter procedure if often called bosonization ) .
the spin-@xmath0 xxz heisenberg chain , eq .
, in the gapless phase is a well - known example of a model belonging to the ll universality class .
its local operators are expressed through the boson fields as follows@xcite : @xmath96\end{gathered}\ ] ] and @xmath97 here the continuous coordinate @xmath98 is given in units of the lattice spacing @xmath99 @xmath100 is the magnetization per site of the spin chain , and @xmath101 , @xmath102 and @xmath103 are coefficients which depend on the parameters of the model .
how to calculate @xmath104 @xmath105 @xmath101 , @xmath106 and @xmath103 is described in the appendix [ sec : llparameters ] .
the hamiltonian is the leading term in the strong coupling expansion ( the parameter @xmath107 see appendix [ sec : tjmodelmapping ] ) of the model .
local operators of the latter model are bosonized by combining eqs . , , and .
the analysis of the model suggests@xcite that the bosonization of the local spins can be performed for _ any _ values of @xmath36 and @xmath35 in the gapless regime .
we would like to stress that even for a small @xmath74 some parameters out of @xmath104 @xmath105 @xmath101 , @xmath106 and @xmath103 show significant numerical differences if calculated within the spin chain compared to the spin ladder .
we discuss this issue in the appendix [ sec : llparameters ] . up to now
, we have presented methods adapted to deal with one dimensional systems . in real compounds , an interladder coupling is typically present . as discussed in sec .
[ sec : weakinterladderc ] , in the incommensurate regime this interladder coupling @xmath89 ( cf .
eq . ) can lead to a new three dimensional order ( 3d - ordered phase in fig .
[ fig : phasediagramm].b ) at temperatures of the order of the coupling @xmath89 . in the case of bpcb
the interladder coupling is much smaller than the coupling inside the ladders , i.e. @xmath108 ( sec . [ sec : bpcb ] ) .
therefore , unless one is extremely close to @xmath109 or @xmath67 one can treat the interladder coupling with a standard mean - field approximation .
this approach incorporates all the fluctuations inside a ladder .
however , it overestimates the effect of @xmath89 by neglecting quantum fluctuations between different ladders .
such effects can partly be taken into account by a suitable change of the interladder coupling@xcite that will be discussed in sec .
[ sec : coupledladderproperties ] .
close to the critical fields the interladder coupling @xmath89 becomes larger than the effective energy of the one dimensional system .
this forces one to consider a three dimensional approach from the start and brings the physics of the system in the universality class of bose - einstein condensation @xcite . in the following
we consider that we are far enough ( i.e. by an energy of the order of @xmath89 ) away from the critical points so that we can use the mean - field approximation . since the single ladder correlation functions along the magnetic field direction ( @xmath25-axis ) decay faster than the staggered part of the ones in the perpendicular @xmath110-plane
( see eqs . and for the ll exponent @xmath94 of the ladder shown in fig .
[ fig : llparameter ] ) , the three dimensional order will first occur in this plane .
thus the dominant order parameter is the @xmath111 staggered magnetization perpendicular to the applied magnetic field .
the mean - field decoupling of the spin operators of neighboring ladders thus reads @xmath112 we have chosen the @xmath110-ordering to be along the @xmath113-axis and @xmath114 will be very small and therefore neglected .
this approximation applied on the interladder interaction part of the 3d hamiltonian @xmath115 ( eq . ) leads to @xmath116 here we assume that the coupling is dominated by @xmath117 neighboring ladders , where @xmath117 is the rung connectivity ( @xmath118 for the case of bpcb , cf .
[ fig : structure ] ) .
this mean - field hamiltonian corresponds to a single ladder in a site dependent magnetic field with a uniform component in the @xmath25-direction and a staggered component in the @xmath113-direction .
the ground state wave function of the hamiltonian must be determined fulfilling the self - consistency condition for @xmath119 and @xmath120 using numerical or analytical methods .
this amounts to minimize the ground state energy of some variational hamiltonian .
the order parameters @xmath119 and @xmath120 can be computed numerically by treating the mean - field hamiltonian @xmath121 self consistently with dmrg .
these parameters are evaluated recursively in the middle of the ladder ( to minimize the boundary effects ) starting with @xmath122 and @xmath123 .
an accuracy of @xmath124 on these quantities is quickly reached after a few recursive iterations ( typically @xmath125 ) of the dmrg keeping few hundred dmrg states and treating a system of length @xmath126 .
we verified by keeping as well the alternating part of the @xmath25-order parameter @xmath114 that this term is negligible ( @xmath127 ) .
using the low energy ll description of our ladder system ( see sec .
[ sec : luttinger_liquid ] ) , it is possible to treat the mean - field hamiltonian @xmath121 within the bosonization technique . introducing the ll operators and in @xmath121 and keeping only the most relevant terms leads to the hamiltonian@xcite @xmath128\\+\sqrt{a_x}n_cj'm_a^x\int dx\cos(\theta(x))\end{gathered}\ ] ] where we neglected the mean - field renormalization of @xmath31 in .
this hamiltonian differs from the standard ll hamiltonian @xmath129 by a cosine term corresponding to the @xmath113-staggered magnetic field in .
it is known as the sine - gordon hamiltonian@xcite .
the expectation values of the fields can be derived from integrability@xcite . in particular @xmath120
can be determined self - consistently . in order to take into account the detailled coupling structure of the bpcb compound shown in fig .
[ fig : structure ] , which is neglected in the mean - field approximation ( sec .
[ sec : mean - field ] ) , we employ a stochastic series expansion implementation of the qmc technique with directed loop updates @xcite provided with the alps libraries @xcite . nevertheless due to the strong anisotropy of the couplings , the temparatures at which the effects of the interladder coupling @xmath89 become visible are not reachable with this method .
the qmc results for the transition temperature of the 3d - ordered phase , @xmath86 , presented in ref . , sec .
[ sec : transitiontemperature ] and appendix [ sec : qmc ] are then computed with a @xmath89 of @xmath130 times bigger than that extracted in ref . and sec .
[ sec : transitiontemperature ] making the 3d effects numerically accessible .
we begin our analysis of the different phases of the coupled spin ladder system , fig .
[ fig : phasediagramm].b , by computing thermodynamic quantities , such as the magnetization , the rung state density and the specific heat .
in particular , we test the ll low energy prediction of the latter and evaluate the related crossover to the quantum critical regime .
furthermore we discuss the effect of the 3d interladder coupling computing the staggered magnetization in the 3d - ordered phase and its critical temperature .
we finally discuss the nmr relaxation rate in the gapless regime related to the low energy dynamics . in order to compare these physical quantities to the experiments ,
all of them are computed for the bpcb parameters ( see sec . [
sec : bpcb ] ) .
the zero temperature magnetization contains extremely useful information .
its behavior directly gives the critical values of the magnetic fields @xmath109 and @xmath67 at which the system enters and leaves the gapless regime , respectively ( fig . [
fig : phasediagramm].b ) . in fig .
[ fig : mz_zerot ] the dependence of the magnetization on the applied magnetic field is shown for a single ladder and for the weakly coupled ladders . at low magnetic field ,
@xmath131 , the system is in the gapped spin liquid regime with zero magnetization , and spin singlets on the rungs dominate the behavior of the system ] , see fig .
[ fig : tripletdensity ] . at @xmath132
, the zeemann interaction closes the spin gap to the rung triplet band @xmath76 ( fig .
[ fig : phasediagramm ] ) . above @xmath133
the triplet @xmath76 band starts to be populated leading to an increase of the magnetization with @xmath31 .
the lower critical field in a 13th order expansion@xcite in @xmath74 is @xmath134 for the bpcb parameters . at the same time
the singlet and the high energy triplet occupation decreases , fig .
[ fig : tripletdensity ] . for @xmath135 ( for the compound bpcb ) , the @xmath76 band
is completely filled and the other bands are depopulated .
the system becomes fully polarized ( @xmath136 ) and gapped .
the two critical fields , @xmath109 and @xmath67 , are closely related to the two ladder exchange couplings , @xmath36 and @xmath35 . as they are experimentally easily accessible , assuming that a ladder hamiltonian is an accurate description of the system , these critical fields can be used to determine the ladder couplings@xcite .
such a general behavior of the magnetization is seen for both the single ladder and the weakly coupled ladders in fig .
[ fig : mz_zerot ] . in particular
, the effect of a small coupling @xmath89 between the ladders is completely negligible in the central part of the curve . only in the vicinity of the critical fields , the single ladder and
the coupled ladders show a distinct behavior .
the single ladder behaves like an empty ( filled ) one - dimensional system of non - interacting fermions which leads to a square - root behavior @xmath137 close to the lower critical field and @xmath138 close to the upper critical field .
in contrast , in the system of weakly coupled ladders , a 3d - ordered phase appears at low enough temperatures in the gapless regime ( see sec .
[ sec : weakinterladderc ] and [ sec : mean - field ] ) .
the magnetization dependence close to the critical fields becomes linear , @xmath139 , and @xmath140 , respectively@xcite . in comparison with the single ladder ,
the critical fields given above are shifted by a value of the order of @xmath89 .
this behavior is in the universality class of the bose - einstein condensation and is well reproduced by the mean - field approximation shown in the insets of fig .
[ fig : mz_zerot ] close to the critical fields . for comparison , the magnetization of a single ladder in the spin chain mapping
is also plotted in fig .
[ fig : mz_zerot ] .
this approximation reproduces well the general behavior of the ladder magnetization discussed above .
however , note that for the exchange coupling constants considered here the lower critical field in this approximation is different from the ladder one .
the lower critical field is @xmath141 .
the upper critical field @xmath142 is the same as for the ladder . if we rescale @xmath143 and @xmath144 to match the critical fields @xmath109 and @xmath67 ( @xmath145 ) , the magnetization curve gets very close to the one calculated for a ladder
however , in contrast to the magnetization curve for the ladder , the corresponding curve in the spin chain mapping is symmetric with respect to its center at @xmath146 due to the absence of the high energy triplets . on the magnetic field @xmath31 at zero temperature for the single ladder with the bpcb couplings ( sec .
[ sec : bpcb ] ) ( dashed red line ) , the spin chain mapping ( dotted blue line ) rescaled to fit with the single ladder critical fields ( dash - dotted blue line ) , and for the weakly coupled ladders treated by the mean - field approximation ( solid black line ) .
the insets emphasize the different behavior of the magnetization curves for the single ( dashed red line ) and coupled ( solid black line ) ladders close to the critical fields which are indistiguishable in the main part of the figure .
the dotted lines in the insets correspond to the linear and square root like critical behaviour .
[ fig : mz_zerot ] ] at zero temperature for the single ladder with the bpcb couplings ( zero temperature dmrg calculations averaging on the central sites of the ladder ) .
the dash - dotted ( black ) lines correspond to the singlet density @xmath147 .
the triplet densities are represented by the solid ( red ) lines for @xmath148 , the dashed ( blue ) lines for @xmath149 and the dotted ( green ) lines for @xmath150 .
[ fig : tripletdensity ] ] the thermodynamics of the spin-@xmath0 ladders has been studied in the past@xcite .
we here summarize the main interesting features of the magnetization and the specific heat focussing on the crossover between the ll regime and the quantum critical region using the bpcb parameters ( sec .
[ sec : bpcb ] ) .
as the interladder exchange coupling @xmath89 is very small compared to the ladder exchange couplings @xmath35 and @xmath36 , it is reasonable to neglect it in this regime far from the 3d phase .
therefore we focus on a single ladder in the following .
we start the description of the temperature dependence of the magnetization in the two gapped regimes : the spin liquid phase and the fully polarized phase ( not shown , cf . refs . ) . for small magnetic fields
@xmath131 , the magnetization vanishes exponentially at zero temperature and after a maximum at intermediate temperatures it decreases to zero for large temperatures . for large magnetic fields @xmath151
the magnetization increases exponentially up to @xmath136 at low temperature and decreases monotonously in the limit of infinite temperature . in the gapless regime ,
the magnetization at low temperature has a non - trivial behavior that strongly depends on the applied magnetic field .
the temperature dependence of the magnetization computed with the t - dmrg ( see appendix [ sec : temperaturedmrg ] ) is shown in fig . [
fig : tmagnetization].a for different values of the magnetic field in the gapless phase , @xmath152 ( @xmath153 ) . in this regime new extrema appear in the magnetization at low temperature .
this behavior can be understood close to the critical fields where the ladder can be described by a one - dimensional fermion model with negligible interaction between fermions .
indeed , in this simplified picture@xcite and in more refined calculations@xcite the magnetization has an extremum where the temperature reaches the chemical potential , i.e. , at the temperature at which the energy of excitations starts to feel the curvature of the energy band .
this specific behavior is illustrated in fig .
[ fig : tmagnetization].a with the curve for @xmath154 ( @xmath155 ) .
the low temperature maximum moves to higher temperature for @xmath156 and goes over to the already discussed maximum for @xmath131 .
symmetrically with respect to @xmath157 , a low temperature minimum appears in the curve for @xmath158 ( @xmath159 ) .
this minimum slowly disappears for @xmath160 ( the curve for @xmath161 is close to that ) .
the location of the lowest extremum is a reasonable criterion to characterize the crossover temperature between the ll and the quantum critical regime@xcite , since the extremum occurs at temperatures of the order of the chemical potential . a plot of this crossover temperature versus the magnetic field is presented in fig .
[ fig : tmagnetization].c . following this criterium , the crossover has a continuous shape far from @xmath157 .
nevertheless , close to @xmath157 both extrema are close to each other ( since the maximum still exists for @xmath156 field at which the minimum appears ) . the criterium is thus not well defined .
it presents a discontinuity at @xmath157 which is obviously an artefact . in the vicinity of @xmath157
, we thus use another crossover criterium based on the specific heat ( see sec .
[ sec : specificheat ] ) that seems to give a more accurate description . in ref .
, both criteria have been applied on the magnetocaloric effect and specific heat measurements on the compound bpcb , and give a crossover temperature in agreement with the computed ones .
the temperature dependence of the magnetization of the spin chain mapping , fig .
[ fig : tmagnetization].b , exhibits a single low temperature maximum if @xmath162 ( minimum if @xmath163 ) .
the appearance of a single extremum and its convergence to @xmath164 when @xmath165 is due to the exact symmetry with respect to the magnetic field @xmath166 .
this approximation reproduces the main low energy features of the ladder but fails to describe the high energy behavior which strongly depends on the high energy triplets . , for ( a ) the ladder with the bpcb couplings , and ( b ) the spin chain mapping at different applied magnetic fields @xmath158 ( solid blue lines ) , @xmath161 ( dash - dotted green lines ) and @xmath154 ( dashed red lines )
the results were obtained using t - dmrg .
the stars at @xmath167 are the ground state magnetization per rung determined by zero temperature dmrg .
the triangles ( squares ) marks the low ( high ) energy extrema .
( c ) crossover temperature @xmath168of the ll to the quantum critical regime versus the applied magnetic field ( blue circles for the extremum in @xmath169 criterium and red crosses for the maximum in @xmath170 criterium)[fig : tmagnetization ] . ]
the specific heat has been investigated for similar parameters as the ones considered in this paper in the gapped and gapless regime in refs . .
here we concentrate on the detailed analysis in the gapless regime , in particular on the low temperature behavior and the determination of a crossover temperature from the first maximum .
we show in fig .
[ fig : cvst ] the typical temperature dependence of the specific heat for different values of the magnetic field . comparisons with actual experimental data@xcite for bpcb are excellent ( see fig . [
fig : cvst].b ) . for these comparisons
the theoretical data are computed with @xmath171 related to the experimental orientation of the sample with respect to the magnetic field@xcite ( see sec . [
sec : bpcb ] ) and rescaled by a factor @xmath172 in agreement with the global experimental uncertainties has to be applied on the theoretical specific heat ( per rung ) to convert to the experimental units . ] .
at low temperatures the specific heat has a contribution due to the gapless spinon excitations .
this results in a peak around @xmath173 .
this peak is most pronounced for the magnetic field values lying mid value between the two critical fields . at higher temperatures the contribution of the gapped triplet excitations leads to a second peak
whose position depends on the magnetic field .
to separate out the contribution from the low lying spinon excitations , we compare the specific heat of the ladder to the results obtained by the spin chain mapping in which we just keep the lowest two modes of the ladder ( see sec .
[ sec : spinchainmap ] and appendix [ sec : tjmodelmapping ] ) .
the resulting effective chain model is solved using bethe - ansatz and t - dmrg methods .
the agreement between these methods is excellent and the corresponding curves in fig .
[ fig : cvst ] can hardly be distinguished . however , a clear difference with the full spin ladder results is revealed .
while at low temperatures the curves are very close , the first peak in the spin chain mapping already lacks some weight , which stems from higher modes of the ladder . in fig .
[ fig : cvst_zoom ] the low temperature region is analyzed in more detail . at very low temperatures
the spinon modes of the ladder can be described by the ll theory ( see sec .
[ sec : luttinger_liquid ] ) which predicts a linear rise with temperature inversely proportional to the spinon velocity@xcite @xmath174 in fig .
[ fig : cvst_zoom ] we compare the results of the ll , the bethe ansatz and the dmrg results for the effective spin chain and the numerical dmrg results taking the full ladder into account .
the numerical results for the adaptive t - dmrg at finite temperature are extrapolated to zero temperature by connecting algebraically to zero temperature dmrg results ( see appendix [ sec : temperaturedmrg ] ) . a very good agreement between and numerics
is found for low temperatures .
however , at higher temperatures , the slope of the @xmath175 ll description slightly changes with respect to the curves calculated with other methods .
this change of slope reflects the fact that the curvature of the energy dispersion must be taken into account when computing the finite temperature specific heat , and this even when the temperature is quite small compared to the effective energy bandwidth of the system . the effective spin chain and the numerical results for the ladder agree for higher temperatures ( depending on the magnetic field ) , before the higher modes of the ladder cause deviations . as for the magnetization ( sec .
[ sec : finitetmagnetization ] ) , the location of the low temperature peak can be interpreted as the crossover of the ll to the quantum critical regime
. indeed , in a free fermion description which is accurate close to the critical fields , this peak appears at the temperature for which the excitations stem from the bottom of the energy band .
the corresponding temperature crossover is compared in fig .
[ fig : tmagnetization].c to the crossover temperature extracted from the first magnetization extremum ( sec .
[ sec : finitetmagnetization ] ) .
the two crossover criteria are complementary due to their domain of validity . the first specific heat maximum is well pronounced only in the center of the gapless phase .
in contrast in this regime the presence of two extrema close to each other in the magnetization renders the magnetization criterium very imprecise ( cf .
[ sec : finitetmagnetization ] ) . versus the temperature @xmath176 for different applied magnetic fields in the gapless regime .
( a ) full ladder t - dmrg calculations with the bpcb couplings for @xmath158 ( solid blue line ) , @xmath161 ( dash - dotted green line ) , and @xmath154 ( dashed red line ) .
spin chain mapping at @xmath161 solved by dmrg ( dashed black line ) and by bethe ansatz ( dotted black line ) .
note that the two lines are hardly distinguishable .
the triangles ( squares ) mark the low ( high ) energy maxima of the specific heat versus temperature .
the vertical dashed line marks the temperature @xmath177 below which the dmrg results are extrapolated ( see appendix [ sec : temperaturedmrg ] ) .
( b ) comparison between measurements on the compound bpcb from ref .
( dots ) and the t - dmrg calculations ( solid lines ) at ( b.1 ) @xmath158 , ( b.2 ) @xmath161 and ( b.3 ) @xmath154.[fig : cvst ] ] , for ( a ) @xmath178 , ( b ) @xmath179 , ( c ) @xmath180 . the t - dmrg calculations are in red thick lines for the ladder with the bpcb couplings ( black thick lines for the spin chain mapping ) .
the two curves can hardly be distinguished .
their low temperature polynomial extrapolation is plotted in thin lines below @xmath177 ( represented by a vertical dashed line ) .
the linear low temperature behavior of the ll is represented by dashed lines ( red for the ladder , black for the spin chain mapping ) .
the dashed yellow lines correspond to the bethe ansatz computation for the spin chain mapping .
[ fig : cvst_zoom ] ] as the spin - lattice relaxation in quantum spin systems is due to pure magnetic coupling between electronic and nuclear spins , the nmr spin - lattice relaxation rate @xmath181 is directly related to the local transverse correlation function@xcite @xmath182 defined as in eq .
@xmath183 where @xmath184 is assumed with the larmor frequency @xmath185 .
@xmath186 is the nuclear gyromagnetic ratio of the measured nucleus ( @xmath187 in ref . for bpcb ) and @xmath188 is the transverse hyperfine coupling constant .
assuming @xmath189 , the @xmath190 due to the spin dynamics can be computed in the gapless regime using the ll low energy description . following ref .
we introduce the ll correlation into eq . , and obtain @xmath191 according to the known ll parameters ( fig .
[ fig : llparameter ] ) the shape of @xmath192 plotted in fig .
[ fig : t1 ] at @xmath193 is strongly asymmetric with respect to the middle of the gapless phase .
the only free ( scaling ) parameter , @xmath194 , is deduced from the fit of eq . to the experimental data@xcite and agrees with other measurements . for comparison ,
the @xmath190 obtained in the spin chain mapping approximation is also plotted in fig .
[ fig : t1 ] . as for other physical quantities ,
this description fails to reproduce the non - symmetric shape . , at @xmath195 .
the solid red line is the bosonization determination using the ladder ll parameters for the bpcb couplings ( the dashed blue line uses the ll parameters of the spin chain mapping ) .
the black circles are the measurements from ref .
.[fig : t1 ] ] the interladder coupling @xmath89 induces a low temperature ordered phase ( the 3d - ordered phase in fig .
[ fig : phasediagramm].b ) . using the mean - field approximation presented in sec .
[ sec : mean - field ] we characterize the ordering and compute the critical temperature and the order parameter related to this phase . in order to compute the critical temperature of the 3d transition
, we follow ref . and treat the staggered part of the mean - field hamiltonian @xmath121 perturbatively using linear response .
the instability of the resulting mean - field susceptibility , due to the 3d transition , appears at @xmath86 when@xcite @xmath196 where @xmath197 is the transverse staggered retarded correlation function of an isolated single ladder system ( appendix [ sec : finiteluttingerliquidcorr ] ) .
this correlation can be computed analytically ( see eq . ) using the ll low energy description of the isolated ladder ( eq . ) in the gapless regime .
applying the condition to the ll correlation leads to the critical temperature @xmath198 introducing the computed ll parameters @xmath95 , @xmath94 and @xmath101 ( see fig .
[ fig : llparameter ] ) in this expression , we get the critical temperature@xcite as a function of the magnetic field .
this is shown in fig .
[ fig : criticaltemperature ] together with the experimental data .
this allows us to extract the _ mean - field _ interladder coupling @xmath199 for the experimental compound bpcb ( the _ only _ free ( scaling ) parameter in eq . ) .
the asymmetry of the ll parameters induces a strong asymmetry of @xmath86 with respect to the middle of the 3d phase .
as the mean - field approximation neglects the quantum fluctuations between the ladders , the critical temperature @xmath86 is overestimated for a given @xmath200 .
we thus performed a quantum monte carlo ( qmc ) determination of this quantity based on the same 3d lattice structure .
let us note that qmc simulations of the coupled spin ladder hamiltonian are possible , since the 3d lattice structure , fig .
[ fig : structure ] , is unfrustrated . in appendix
[ sec : qmc ] we present results on how to determine the critical temperatures for the 3d ordering transition using qmc
. this determination shows@xcite that the real critical temperature is well approximated by the mean - field approximation , but with a rescaling of the real interladder coupling @xmath201 with @xmath202 .
the rescaling factor @xmath203 is similar to the values obtained for other quasi one - dimensional antiferromagnets@xcite . ,
is plotted in solid red line for the ladder ll parameters of bpcb ( in dashed blue line for the ll parameters of the spin chain mapping ) .
the nmr measurements from ref . are represented by black circles and the neutron diffraction measurements from ref . by green dots .
[ fig : criticaltemperature ] ] the staggered order parameter in the 3d - ordered phase , @xmath120 , can be analytically determined at zero temperature using the mean - field approximation for the interladder coupling and the bosonization technique ( see sec .
[ sec : mean - field ] ) . as @xmath204 in the bosonization description and the expectation value@xcite of the operator
@xmath205 is @xmath206 for the sine - gordon hamiltonian @xmath207 with @xmath208^{\frac{8k } { 8k-1}}}{\left[\gamma \left(\frac{4k}{8k-1 } \right ) \gamma \left(\frac{16k-3}{16k-2 } \right)\right]^2},\ ] ] we can extract @xmath209 this can be evaluated in the 3d - ordered phase by introducing into the ll parameters @xmath95 , @xmath94 and @xmath101 from fig .
[ fig : llparameter ] . fig .
[ fig : orderparameter ] shows the order parameter versus the magnetic field determined analytically and numerically by dmrg ( see sec .
[ sec : numericalmean - field ] ) .
the two curves are almost indistinguishable and exhibit a strongly asymmetric camel - like shape@xcite with two maxima close to the critical fields .
the asymmetry of the curve is again due to the presence of the additional triplet states .
this asymmetry disappears in the spin chain mapping .
, in the 3d - ordered phase determined for the ground state .
( a ) its computation with the analytical bosonization technique for the ll parameters of the compound bpcb and @xmath210 is represented by the dash - dotted red line ( dashed blue line for the ll parameters of the spin chain mapping ) .
the dmrg result for @xmath211 is represented by black dots ( as a comparison the bosonization result for @xmath212 is plotted in solid red line ) .
note , that these two data are almost indistinguishable .
( b ) comparison between nmr measurements ( black cirles ) done at @xmath213 from ref . and scaled on the theoretical results for @xmath214 , neutron diffraction measurements on an absolute scale from ref . at @xmath215 ( @xmath216 ) ( red crosses ( black dots ) ) and the computation for @xmath210 ( blue line as in panel ( a ) ) .
recent neutron diffraction measurements as a function of temperature suggest that the data of ref . was taken at temperatures approximately @xmath217 higher than the nominal indicated temperature . [
fig : orderparameter ] ]
in the next paragraph , we discuss the possible excitations and the corresponding correlation functions created by the spin operators .
such dynamical correlations allow us to study the excitations of our system .
they are also directly related to many experimental measurements ( nmr , ins ... ) .
we first focus on the gapped spin liquid and then treat the gapless regime , for fields @xmath153 ( see fig .
[ fig : phasediagramm].b ) .
all correlations are computed using the t - dmrg at zero temperature for the single ladder ( appendix [ sec : timedmrg ] ) and are compared to analytical results when such results exist . in particular we check the overlap with the ll description at low energy and use a strong coupling expansion ( appendix [ sec : tjmodelmapping ] ) to qualitatively characterize the obtained spectra .
we start by a discussion of the correlations for the parameters of the compound bpcb ( see sec . [
sec : bpcb ] ) and then turn to the evolution of the spectra with the coupling ratio @xmath74 from the weak ( @xmath218 ) to strong coupling ( @xmath219 ) . in a ladder system
different types of correlations are possible .
we focus here on the quantities @xmath220 where @xmath221 are the symmetric ( @xmath222 ) and antisymmetric ( @xmath223 ) operators with rung momentum @xmath224 and parity @xmath225 respectively , @xmath226 , @xmath227 ( for a single ladder @xmath228 corresponds to the hamiltonian ) and the momentum @xmath229 is given in reciprocal lattice units @xmath230 .
the rung momentum @xmath231 is a good quantum number .
the dynamical correlations are directly related to ins measurements ( see sec . [ sec : ins ] ) .
they select different types of rung excitations ( as summarized in table [ tab : excitations ] ) . using the reflection and translation invariance of an infinite size system ( @xmath232 )
, we can rewrite the considered correlations in a more explicit form ( at zero temperature ) .
] , i.e. @xmath233 where @xmath234 denotes the ground state of @xmath228 with energy @xmath235 , @xmath236 and @xmath237 is the sum over all eigenstates @xmath238 of @xmath228 with energy @xmath239 .
the form of eq
. clearly shows that @xmath240 is non - zero if the operator @xmath241 can create an excitation @xmath238 of energy @xmath242 and momentum @xmath229 from the ground state .
the correlations @xmath243 are then direct probes of the excitations @xmath238 in the system .
@xmath244 since the experimentally relevant case ( compound bpcb ) corresponds to a relatively strong coupling situation ( @xmath245 , eq .
[ equ : couplingratio ] ) , we use the decoupled bond limit introduced in sec .
[ sec : spinchainmap ] to present the expected excitations @xmath246 , @xmath56 , @xmath57 or @xmath58 . in table
[ tab : excitations ] , we summarize the rung excitations created by all the operators @xmath241 and their properties . for example , the rung parity @xmath247 is changed by applying an operator with rung momentum @xmath248 and the @xmath25-magnetization is modified by @xmath249 by applying the operators @xmath250 respectively . using the decoupled bond limit in the spin liquid phase
, the excitations in the system can be pictured as the excitation of rung singlets to rung triplets . at zero magnetic field @xmath251 ,
the system is spin rotational symmetric and the different triplet excitations have the same energy @xmath252 .
it has been seen previously that in this system both single triplet excitations and two - triplet excitations play an important role @xcite .
we discuss these excitations in the following focusing on the ones that can be created by the symmetric @xmath253 and the antisymmetric @xmath254 correlations ( see fig . [
fig : spectrummz0 ] ) for the bpcb parameters .
note that these correlations are independent of the direction @xmath17 due to the spin rotation symmetry .
our results are in very good agreement to previous findings @xcite .
( a ) symmetric part @xmath255 with @xmath17 .
the dashed ( black ) line marks the @xmath256 position of the two - triplet bound state , eq .
( 61b ) in ref . .
the dash - dotted ( white ) lines correspond to the boundaries of the two - triplet continuum .
( b ) antisymmetric part @xmath257 .
the dashed ( black ) line corresponds to the predicted dispersion relation of a single triplet excitation ( eq . ( 8) in ref . )
.[fig : spectrummz0 ] ] at @xmath251 , the system is in a global spin singlet state@xcite ( @xmath258 ) .
the @xmath248 correlation couples then to states with an odd number of triplet excitations with rung parity @xmath259 and total spin @xmath260 ( see table [ tab : excitations ] ) . nevertheless , only single triplet excitations are numerically resolved .
their spectral weight is concentrated in a very sharp peak whose dispersion relation can be approximated using a _ strong coupling expansion _ in @xmath74 . up to first order it is simply given by a cosine dispersion@xcite , i.e. @xmath261 .
further corrections up to third order in @xmath74 have been determined in ref . .
[ fig : spectrummz0].b we compare the numerical results for the bpcb parameters to the expression up to third order in @xmath74 .
the strong coupling expansion describes very well the position of the numerically found excitations .
the structure of the @xmath262 correlation is more complex ( fig .
[ fig : spectrummz0].a ) . due to the rung parity @xmath263 of the operators @xmath264 ,
the excitations correspond to an even number of triplet excitations with total spin @xmath260 .
we focus here on the two - triplet excitations that can be resolved numerically .
these can be divided into a broad continuum and a very sharp triplet ( @xmath68 ) bound state of a pair of rung triplets .
since these excitations stem from the coupling to triplets already present in the ground state ( fig .
[ fig : tripletdensity ] ) , their amplitude for the considered bpcb parameters is considerably smaller than the weight of the single triplet excitations@xcite .
the dispersion relation of the bound states has been calculated using a linked cluster series expansion@xcite .
the first terms of the expansion have an inverse cosine form and the bound state only exists in an interval around @xmath111 ( cf . ref . ) .
the numerical results for the bpcb parameters agree very well with the analytic form of the dispersion ( fig . [
fig : spectrummz0].a ) .
the upper and lower limits of the continuum can be determined by considering the boundary of the two non - interacting triplet continuum .
they are numerically computed using the single triplet dispersion ( eq . ( 8) in ref . ) and shown in fig .
[ fig : spectrummz0].a .
they agree very well with the numerically found results .
the comparison with the known solutions serves as a check of the quality of our numerical results . a small applied magnetic field ( @xmath131 ) , at first order , only smoothly translates the excitations shown in fig .
[ fig : spectrummz0 ] by an energy @xmath265 due to the zeeman effect .
however , if the magnetic field exceeds @xmath109 , the system enters into the gapless regime with a continuum of excitations at low energy . for small values of @xmath74 most features of this low energy continuum are qualitatively well described by considering the lowest two modes of the ladder only . beside the low energy continuum ,
a complex structure of high energy excitations exist .
contrarily to the low energy sector this structure crucially depends on the high energy triplet modes . in the following ,
we give a simple picture for these excitations starting from the decoupled bond limit .
-correlation function ( 1 ) at @xmath266 ( @xmath267 ) , ( 2 ) at @xmath268 ( @xmath269 ) , and ( 3 ) at @xmath270 ( @xmath271 ) .
( a ) symmetric part @xmath272 without bragg peak at @xmath273 .
the dashed black lines correspond to the location of the slow divergence at the lower edge of the continuum predicted by the ll theory .
the dashed white curve corresponds to the predicted two - triplet bound state location .
( b ) antisymmetric part @xmath274 .
the dashed black lines correspond to the position of the high energy divergences or cusps predicted by the approximate mapping on the t - j model .
the vertical white dash - dotted lines mark the momenta of the minimum energy of the high energy continuum and the black cross is the energy of its lower edge@xcite at @xmath111.[fig : zzcorrelationmz ] ] -correlation function ( 1 ) at @xmath266 ( @xmath267 ) , ( 2 ) at @xmath268 ( @xmath269 ) , and ( 3 ) at @xmath270 ( @xmath271 ) .
( a ) symmetric part @xmath275 .
the vertical dash - dotted white lines mark the momenta of the minimum energy of the high energy continuum and the horizontal ones the frequency of its lower edge@xcite at @xmath276 .
the dashed white lines correspond to the position of the high energy divergences or cusps predicted by the approximate mapping on the t - j model .
the dotted white curve corresponds to the predicted two - triplet bound state location .
the high energy excitations at @xmath277 are hardly visible .
( b ) antisymmetric part @xmath278 .
the dashed and dash - dotted ( dotted ) white lines correspond to the location of the strong divergence ( cusp ) at the lower edge of the continuum predicted by the ll theory .
[ fig : pmcorrelationmz ] ] -correlation function ( 1 ) at @xmath266 ( @xmath267 ) , ( 2 ) at @xmath268 ( @xmath269 ) , and ( 3 ) at @xmath270 ( @xmath271 ) .
( a ) symmetric part @xmath279 .
the vertical dash - dotted white lines correspond to the momenta at which the minimum energy of the high energy continuum occurs and the horizontal line to the frequency of its lower edge@xcite at @xmath276 .
the dashed black curve corresponds to the predicted two - triplet bound state location .
( b ) antisymmetric part @xmath280 .
the dashed and dash - dotted ( dotted ) white lines correspond to the location of the strong divergence ( cusp ) at the lower edge of the continuum predicted by the ll theory.[fig : mpcorrelationmz ] ] the evolution of the spectra for the bpcb parameters with increasing magnetic field are presented in fig .
[ fig : zzcorrelationmz ] for @xmath281 , in fig .
[ fig : pmcorrelationmz ] for @xmath282 , and in fig .
[ fig : mpcorrelationmz ] for @xmath283 .
three different classes of excitations occur : * a continuum of excitations at low energy for @xmath284 and @xmath285 * single triplet excitations at higher energy with a clear substructure for @xmath286 , @xmath287 , and @xmath288 * excitations at higher energy for @xmath284 and @xmath287 and @xmath289 stemming from two - triplet excitations which have their main weight around @xmath290 .
in the following we summarize some of the characteristic features of these excitations , before we study them in more detail in secs .
[ sec : lowenergyexcitations ] to [ sec : singletripletll ] .
\(i ) the continuum at low energy which does not exist in the spin liquid is a characteristic signature of the gapless regime .
it stems from excitations within the low energy band which corresponds to the @xmath53 and @xmath55 states in the decoupled bond limit ( cf .
[ fig : phasediagramm].a and table [ tab : excitations ] ) : * excitations within the triplet @xmath55 mode * excitations between the singlet @xmath53 and the triplet @xmath55 mode .
this continuum is smoothly connected to the spin liquid spectrum in the case of @xmath291 .
it originates from the single triplet @xmath76 branch ( sec .
[ sec : singletripletspinliquid ] ) when the latter reaches the ground state energy due to the zeeman effect .
since two modes play the main role in the description of these low energy features , many of them can already be explained qualitatively by the spin chain mapping .
the excitations in the chain have been studied previously using a bethe ansatz description and exact diagonalization calculations in ref . .
more recently they were computed in ref .
due to recent progress in the bethe ansatz method .
in particular , the boundary of the spectrum at low energy is well described by this approach , since the ll velocity determining it is hardly influenced by the higher modes ( cf .
[ fig : llparameter ] ) .
however , a more quantitative description requires to take into account the higher modes of the system as well . in sec .
[ sec : lowenergyexcitations ] we compare in detail our results with the ll theory and the spin chain mapping pointing out their corresponding ranges of validity .
\(ii ) the single high energy triplet excitations form a continuum with a clear substructure . in the decoupled bond limit
, these excitations correspond to * single triplet excitations @xmath56 at energy @xmath292 * single triplet excitations @xmath56 at energy @xmath292 * single triplet excitations @xmath57 at energy @xmath293 .
many of the features of these continua can be understood by mapping the problem onto a mobile hole in a chain , as pointed out first in ref . .
we detail in sec .
[ sec : singletripletll ] and appendix [ sec : tjmodelmapping ] this mapping .
it opens the possibility to investigate the behavior of a single hole in a t - j like model using experiments in pure spin ladder compounds .
\(iii ) the high energy continuum , which has almost no weight close to the brillouin zone boundary ( @xmath276 ) , is related to two - triplet excitations of the spin liquid ( sec .
[ sec : zerofieldtwotriplets ] ) .
they are generated from high energy triplet components of the ground state .
their weight therefore vanishes for @xmath294 and the excitations correspond to * two - triplet excitations .
@xmath295 at energy @xmath292 * two - triplet excitations @xmath296 at energy @xmath297 * two - triplet excitations @xmath298 at energy @xmath299 . in this section
we concentrate on the low energy excitations of type ( i ) discussing first their support and then comparing their spectral weight to the ll prediction .
[ [ support - of - the - low - energy - excitations ] ] support of the low energy excitations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + , ( b ) the triplet band @xmath246 in the gapless phase for a given magnetization @xmath119.[fig : bandfilling ] ] the position of the soft modes in the low energy continuum can be directly obtained from the bosonization representation @xcite ( see appendix [ sec : luttingerliquidcorr ] ) .
they can also be understood in a simple picture which we outline in the following .
the distribution of the rung state population in the ground state depends on the magnetic field @xmath31 ( see fig . [
fig : tripletdensity ] ) . taking a fermionic point of view
, the magnetic field acts as a chemical potential that fixes the occupation of the singlet and triplet rung states . increasing the magnetic field reduces the number of singlets , whereas at the same time the number of triplets increases ( see sketch in fig .
[ fig : bandfilling ] ) .
the fermi level lies at the momenta @xmath300 for the singlet states and at the momenta @xmath301 for the triplet states . in this picture
the soft modes correspond to excitations at the fermi levels . for transitions
@xmath302 the transferred momenta of these zero energy excitations are @xmath303 .
in contrast the interspecies transitions @xmath304 allow the transfer of @xmath305 . therefore the positions of the soft modes in the longitudinal correlation @xmath284 which allows transitions within the triplet states shift from the boundaries of the brillouin zone inwards towards @xmath111 when @xmath119 increases ( fig .
[ fig : zzcorrelationmz].a ) .
in contrast the positions of the soft modes in the transverse correlations @xmath285 which allow transitions between the singlets and the triplets move with increasing magnetic field outwards ( figs .
[ fig : pmcorrelationmz].b and [ fig : mpcorrelationmz].b ) .
the top of these low energy continua are reached when the excitations reach the boundaries of the energy band . in particular , the maximum of the higher boundary lies at the momentum @xmath111 which is easily understood within the simple picture drawn above ( cf . fig .
[ fig : bandfilling ] ) .
a more detailed description of different parts of these low energy continua is given in ref . .
let us compare the above findings with the predictions of the ll theory for the dynamical correlations @xcite .
details on the ll description of the correlations are given in appendix [ sec : luttingerliquidcorr ] .
the ll theory predicts a linear momentum - frequency dependence of the lower continuum edges with a slope given by the ll velocity @xmath306 ( fig .
[ fig : llparameter ] ) .
the position of the soft modes are given by the ones outlined above ( see fig . [
fig : llcorrschema ] ) . the predicted support at low energy agrees very well with the numerical results ( fig . [ fig : zzcorrelationmz].a , [ fig : pmcorrelationmz].b , and [ fig : mpcorrelationmz].b ) .
of course when one reaches energies of order @xmath35 in the spectra one can not rely on the ll theory anymore .
this is true in particular for the upper limit of the spectra .
[ [ sec : strongcouplingllcomp ] ] spectral weight of the excitations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + let us now focus on the distribution of the spectral weight in the low energy continuum .
in particular we compare our numerical findings to the luttinger liquid description .
qualitatively , the ll theory predictions for the low energy spectra are well reproduced by the dmrg computations .
the luttinger liquid predicts typically an algebraic behavior of the correlations at the low energy boundaries which can be a divergence or a cusp .
* the luttinger liquid predicts peaks at the @xmath276 branches and a slow divergence at the lower edge of the incommensurate branches @xmath307 ( with exponent @xmath308 ) . in the numerical results ( fig .
[ fig : zzcorrelationmz].a ) a slight increase of the weight towards the lower edge of the incommensurate branches can be seen . * a strong divergence at the lower edge of the @xmath111 branch ( with exponent @xmath309 )
is obtained within the luttinger liquid description .
this is in good qualitative agreement with the strong increase of the spectral weight observed in the numerical data .
a more interesting behavior is found close to the momenta @xmath310 in the incommensurate branches . here
a strong divergence is predicted for momenta higher ( lower ) than the soft mode @xmath311 ( @xmath312 ) with exponent @xmath313 .
in contrast for momenta lower ( higher ) than the soft mode @xmath311 ( @xmath312 ) a cusp with exponent @xmath314 is expected . in the numerical results ( fig .
[ fig : pmcorrelationmz].b ) this very different behavior below and above the soft modes is evident .
the divergence and cusp correspond to a large and invisible weight , respectively .
* the same behavior as @xmath288 replacing @xmath315 can be observed in fig .
[ fig : mpcorrelationmz].b . and magnetization @xmath268 of the low energy spectrum ( a ) @xmath316 , and ( b ) @xmath317 .
the ( red ) circles and the ( black ) squares are the numerical results for the ladder and its spin chain mapping , respectively .
the dashed lines correspond to the ll predictions and the solid lines are the latter convolved with the same gaussian filter than the numerical data .
the dmrg frequency numerical limitation is of the order of the peak broadening of width @xmath318 ( see appendix [ sec : tdmrgcorrelation ] ) .
[ fig : llcomparisons ] ] to compare quantitatively the predictions of the ll to the numerical results we show in fig .
[ fig : llcomparisons ] different cuts of the correlations at fixed momentum @xmath111 and magnetization @xmath268 for the ladder and the spin chain mapping .
these plots show the dmrg results , the ll description , and the latter convolved with the gaussian filter .
the filter had been used in the numerical data to avoid effects due to the finite time - interval simulated ( see appendix [ sec : tdmrgcorrelation ] ) .
note that the amplitude of the ll results are inferred from the static correlation functions , such that the ll curve is fully determined and _ no fitting parameter _ is left .
therefore the convolved ll results can directly be compared to the numerical results . even though the presented numerical resolution might not be good enough to resolve the behavior close to the divergences ( cusps ) , interesting information as the arising differences between the spin chain mapping and the full ladder calculations can already be extracted . in fig .
[ fig : llcomparisons].a , we show a cut through the correlation at fixed momentum @xmath316 .
the convolved ll and the numerical results compare very well . the difference between the real ladder calculations and the spin chain mapping that neglect the effects of the higher triplet states
@xmath57 , @xmath56 is obvious . from the ll description point of view , the shift of the spin chain correlation compared to the real ladder curve comes mainly from the prefactor @xmath103 and the algebraic exponent which are clearly modified by the effects of the high energy triplets ( see fig . [
fig : llparameter ] ) . for the transverse correlations ,
the ll theory predicts a strong divergence ( with an exponent @xmath319 at the lower boundary of the continuum branch at @xmath111 . a cut through the low energy continuum @xmath317 is shown in fig .
[ fig : llcomparisons].b .
the convolved luttinger liquid reproduces well the numerical results . [ [ sec : weakcouplinghighenergy ] ] weak coupling description of the high energy excitations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + before looking in detail at the two kinds of high energy excitations presented in sec .
[ sec : excitationcharacterization ] we compare our computed high energy spectra with the weak coupling description ( @xmath320 ) . in this limit information on the spectrum
can be extracted from the bosonization description@xcite . in particular
one expects a power law singularity at the lower edge continuum with a minimal position at @xmath321 ( for @xmath286 ) , @xmath322 ( for @xmath323 ) and an energy @xmath31 at momentum @xmath111 ( for @xmath286 ) , @xmath273 ( for @xmath323 ) . except for @xmath289
in which the spectral weight is too low for a good visualization , our computed spectra reproduces well the predictions for the minimal positions even though the coupling strength considered is not in the weak coupling limit ( cf . figs . [
fig : zzcorrelationmz].b , [ fig : pmcorrelationmz].a and [ fig : mpcorrelationmz].a ) .
[ [ sec : singletripletll ] ] high energy single triplet excitations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the high energy single triplet continua originate from the transition of the low energy rung states @xmath53 and @xmath76 to the high energy triplets @xmath69 and @xmath75 .
the excitations coming from the singlets @xmath53 ( in @xmath286 and @xmath288 ) are already present in the spin liquid phase ( cf .
[ sec : singletripletspinliquid ] ) in which they have the shape of a sharp peak centered on the triplet dispersion .
the transition between the gapped spin liquid and the gapless regime is smooth and consists in a splitting and a broadening of the triplet branch that generates a broad continuum of new excitations .
contrarily to the latter the excitations coming from the low energy triplets @xmath76 ( in @xmath287 ) are not present in the spin liquid phase .
the corresponding spectral weight appears when @xmath133 .
an interpretation of the complex structure of these high energy continua can be obtained in terms of itinerant quantum chains . using a strong coupling expansion of the hamiltonian ( appendix [ sec : tjmodelmapping ] ) it is possible to map the high energy single triplet excitations @xmath69 to a single hole in a system populated by two types of particles with pseudo spin @xmath324 , @xmath325 ( with the sec .
[ sec : spinchainmap ] notation ) . in this picture
the effective hamiltonian of the @xmath36 energy sector is approximately equivalent to the half filled anisotropic 1d t - j model with one hole ( see appendix [ sec : sec_sector_jperp ] ) .
the effective hamiltonian is given by @xmath326 where @xmath327 is an energy shift and @xmath328 is the usual hopping term . here
@xmath329 ( @xmath330 ) is the creation ( annihilation ) operator of a fermion with pseudo spin @xmath331 at the site @xmath10 .
note that although we are dealing here with spin states , it is possible to faithfully represent the three states of each site s hilbert space ( @xmath53 , @xmath76 , @xmath69 ) using a fermion representation .
additionally to the usual terms of the t - j model a nearest neighbor interaction term between one of the spins and the hole arises @xmath332.\ ] ] here @xmath333 is the density operator of the hole on the site @xmath10 . in this language
the spectral weight of @xmath286 and @xmath287 corresponding to the single high energy triplet excitations is equivalent to the single particle spectral functions of the up - spin and down - spin particle respectively : @xmath334 here @xmath335 . for the standard t - j model ( for @xmath336 invariant xxx spin background and without the anisotropic term @xmath337 in eq .
) , these spectral functions have been studied in refs . .
the presence of singularities of the form @xmath338^{2x_\nu(q)-1}\ ] ] were found .
here @xmath339 is the @xmath69 triplet dispersion relation , @xmath340 the spinon momentum at the fermi level and @xmath341 the algebraic decay exponent at the singularity .
this exponent is not known in our case and depends on the magnetization @xmath119 and the momentum @xmath229 .
it generates a peak or a cusp at the energy @xmath342 .
the spinon momentum @xmath340 depends on the type of the rung state before excitation ( @xmath343 ) . for an excitation created from a singlet state @xmath344 ( for @xmath286 ) and from the triplet state @xmath345 ( for @xmath287 ) ( fig .
[ fig : bandfilling ] ) . at @xmath251 ,
a series expansion of @xmath339 can be performed ( eq . 8 in ref . ) . to extend it into the gapless phase ( @xmath153 ) , we approximate @xmath339 by shifting the value @xmath346 at @xmath251 by the zeeman shift , i.e. @xmath347 here we used the shift of the ground state energy per rung @xmath348 .
@xmath349 was determined by dmrg calculations ( fig .
[ fig : decay_exp].b for the bpcb parameters ) .
the resulting momentum - frequency positions @xmath342 of the high energy singularities ( cusps or divergencies ) are plotted on the spectrum fig .
[ fig : zzcorrelationmz].b and fig .
[ fig : pmcorrelationmz].a .
they agree remarkably well with the shape of the computed spectra in particular for small magnetic field . and
@xmath287 , some of these singularities correspond to the lower edge description in ref . and
discussed in sec .
[ sec : weakcouplinghighenergy ] . ] neglecting the additional interaction term @xmath337 the t - j model hamiltonian would lead to a symmetry of these excitations with respect to half magnetization .
however in the numerical spectra the effect of the interaction shows up in a clear asymmetry of these excitations ( figs .
[ fig : zzcorrelationmz].1.b and [ fig : pmcorrelationmz].3.a ) . in particular in the @xmath287 correlation some of the weight is seemingly detaching and pushed towards the upper boundary of the continuum ( fig . [
fig : pmcorrelationmz].3.a ) for large magnetization . a more detailed account of the spectra can be found in ref . .
a similar mapping can be performed for the single @xmath75 excitation .
in contrast to the @xmath36 sector in the @xmath350 sector not only the @xmath75 excitation occurs , but the effective hamiltonian mixes also @xmath69 triplets into the description . therefore the description by a single hole in a spin-@xmath0 chain breaks down and more local degrees of freedom are required .
this results in a more complex structure as seen in fig .
[ fig : pmcorrelationmz].1.b .
previously high - energy excitations in dimerized antiferromagnets have been described rather generally by a mapping to an x - ray edge singularity problem @xcite .
it is interesting though that in the present setup these excitations can be understood as t - j hole spectral functions , which display a much richer structure than anticipated .
[ [ sec : twotripletll ] ] high energy two - triplet excitations + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the two - triplet continua and bound states already discussed in the spin liquid phase ( cf .
[ sec : zerofieldtwotriplets ] ) are still visible in the gapless regime in the symmetric correlations ( @xmath284 and @xmath323 ) . at low magnetic field
the location of their maximal spectral weight can be approximated by the expression of the bound state dispersion at zero field ( eq . ( 61b ) in ref . )
shifted by the zeeman energy.b ) and the shift of the excited state . ] .
the two - triplet excitation location obtained in this way agrees to a good extent with the location found in the numerical calculations ( cf . figs .
[ fig : zzcorrelationmz].1.a , [ fig : pmcorrelationmz].1.a and [ fig : mpcorrelationmz].1.a ) . since these excitations are generated from the high energy triplet components in the ground state and these vanish with increasing magnetic field ( cf .
[ fig : tripletdensity ] ) , their residual spectral weight slowly disappears with increasing magnetization .
for all the excitation spectra presented above the intrachain coupling ratio of bpcb @xmath351 was taken .
for this chosen value of @xmath74 , a strong coupling approach gives a reasonable description of the physics . in this section
we discuss the evolution of the spectra from weak ( @xmath352 ) to strong coupling ( @xmath353 ) . to illustrate this behavior , we show in fig .
[ fig : weaktostrong ] the symmetric and antisymmetric parts of the correlations @xmath354 at @xmath266 for different coupling ratios @xmath355 . at @xmath352 ( fig .
[ fig : weaktostrong].1 ) , the chains forming the ladder correspond to two decoupled heisenberg chains . in this case the symmetric and antisymmetric correlations are identical @xmath356 and are equivalent to the correlation @xmath357 of the single chain@xcite with magnetization per spin @xmath358 .
a complex low energy continuum exists with zero energy branches@xcite at momenta @xmath359 similar to that discussed in sec .
[ sec : lowenergyexcitations ] .
in contrast , in the strong coupling limit ( @xmath360 ) ( fig .
[ fig : weaktostrong].5.b ) the symmetric correlations vanish and the antisymmetric part corresponds to the single chain correlation @xmath357 with anisotropy @xmath361 and magnetization per spin @xmath362 ( see the spin chain mapping in sec .
[ sec : spinchainmap ] ) .
the antisymmetric part consists of a low energy continuum with branches at momenta @xmath363 ( sec .
[ sec : lowenergyexcitations ] ) .
note , that a bosonization description of the low energy sectors of both extreme regimes can be formulated @xcite ( appendix [ sec : luttingerliquidcorr ] ) . in the following
we discuss the evolution between these two limits . in the antisymmetric correlation ( cf .
[ fig : weaktostrong].2 - 5.b ) a low energy continuum exists at all couplings with a zero energy excitation branch at @xmath111 .
these low energy excitations correspond mainly to the excitations with @xmath364 .
this has been pointed out for the weak coupling limit@xcite .
they become the transitions @xmath365 with the same quantum numbers in the decoupled bond limit . additionally the upper part of the excitation spectrum at weak coupling , which mainly corresponds to excitations with @xcite @xmath366 and @xmath367 splits from the lower part of the spectrum and moves to higher energy while increasing the coupling .
it evolves to a high energy excitation branch which corresponds in the decoupled bond limit to the @xmath368 transition , i.e. single triplet excitations of type ( ii ) ( sec .
[ sec : singletripletll ] ) approximately at and @xmath31 become very close , such that this position is in agreement with the ones previously discussed for the strong coupling limit . ] @xmath350 .
the properties of the zero energy excitation branch at @xmath111 show a smooth transition between the two limits@xcite .
for example the slope of the lower edge continuum which is determined by the ll velocity @xmath95 decreases smoothly from its value for the heisenberg chain to the lower value for the anisotropic spin chain with @xmath361 in the strong coupling limit . in contrast to this smooth change , the presence of a finite value of @xmath36 leads to the formation of a gap in the incommensurate low energy branches@xcite at @xmath369 . with increasing coupling strength @xmath36 new low energy branches at momenta @xmath370
become visible@xcite .
the weight of these gapless branches is very small for small coupling and increases with stronger coupling @xcite .
in contrast to the antisymmetric part , the symmetric part @xmath371 becomes gapped when the interladder coupling @xmath36 is turned on .
the lowest energy excitations remain close to the momenta @xmath369 in agreement with ref . .
they connect to the single triplet excitations of type ( ii ) ( sec .
[ sec : singletripletll ] ) which are approximately at an energy @xmath36 . while increasing @xmath74 the higher part of the spectrum starts to separate from the main part and evolves to a branch of high energy two - triplet excitations of type ( iii ) ( sec .
[ sec : twotripletll ] ) .
these are located at approximately @xmath372 .
our computed spectra for @xmath373 presented in fig .
[ fig : weaktostrong].2 - 4.a clearly show this behavior . in fig .
[ fig : weaktostrong].4.a the highest two - triplet excitations can not be seen anymore since their spectral weight is too low . up to now
we only discussed the excitations of a single spin ladder and neglected the weak interladder coupling @xmath89 usually present in real compounds .
deep inside of the spin liquid phase , the correlations for a single ladder are dominated by high energy single or multi triplet excitations as discussed in sec .
[ sec : spinliquidexcitations ] .
the presence of a small interladder coupling @xmath89 causes a dispersion in the interladder direction with an amplitude of order @xmath89 .
this effect can be evaluated for independent triplet excitations using a single mode approximation@xcite . however , for the compound bpcb the interladder coupling is so small that present day experiments do not resolve this small broadening@xcite .
in contrast in the gapless phase the effect of the interladder coupling can change considerably the excitations .
in particular below the transition temperature to the 3d - ordered phase , a bragg peak appears at @xmath111 in the transverse dynamical functions @xmath285 . as discussed in ref .
, this bragg peak is surrounded by gapless goldstone modes and it has been measured in the compound bpcb@xcite .
additional high energy modes are predicted to occur in the transverse @xmath285 and longitudinal @xmath284@xcite .
it would be interesting to compute the excitations using random phase approximation analogously to ref . in combination with the computed dynamical correlations for the single ladder in order to investigate the effect of a weak interladder coupling in more detail .
however , this goes beyond the scope of the present work and will be left for a future study .
in this section we summarize the results of experiments on the compound bpcb and compare them to the theoretical predictions .
first we introduce the structure of the bcpb compound
. then we focus on static and low energy results of the system .
finally we discuss the detection of excitations by ins@xcite measurements .
in particular we give a prediction of the ins cross section in a full range of energy and compare its low energy part with the measured spectra on the compound bpcb .
the compound @xmath1 , customarily called bpcb or ( hpip)@xmath3cubr@xmath4 , has been intensively investigated using different experimental methods such as nuclear magnetic resonance@xcite ( nmr ) , neutron diffraction@xcite ( nd ) , inelastic neutron scattering@xcite ( ins ) , calorimetry@xcite , magnetometry@xcite , magnetostriction@xcite , and electron spin resonance spectroscopy@xcite ( esr ) .
the magnetic properties of the compound are related to the unpaired highest energy orbital of the @xmath374 ions .
thus the corresponding spin structure ( fig .
[ fig : structure ] ) matches with the @xmath374 location@xcite .
the unpaired spins form two types of inequivalent ladders ( fig .
[ fig : structure ] ) along the @xmath375 axis ( @xmath375 , @xmath376 and @xmath377 are the unit cell vectors of bpcb ) .
the direction of the rung vectors of these ladders are @xmath378 in the primitive vector coordinates ( fig .
[ fig : structure].b ) .
as one can see from the projection of the spin structure onto the @xmath379-plane ( fig .
[ fig : structure].b ) , each rung has @xmath118 interladder neighboring spins .
atoms in the first ( second ) type of ladders are pictured by red ( blue ) spheres .
the @xmath36 , @xmath35 and @xmath89 coupling paths are represented in turquoise , pink , and green , respectively .
@xmath375 , @xmath376 and @xmath377 are the three unit cell vectors of the structure .
gray arrows are the rung vectors of the two types of ladders @xmath380 .
( a ) 3d structure .
( b ) projection of the 3d structure onto the @xmath381-plane .
( c ) projection of the 3d structure onto the @xmath382-plane .
( d ) projection of the 3d structure onto the @xmath383-plane .
[ fig : structure ] ] the bpcb structure has been identified as a good experimental realization of the system of weakly coupled spin-@xmath0 ladders@xcite described by the hamiltonian . as we will explain in the next section
, the interladder coupling @xmath89 has been experimentally determined to be@xcite @xmath384 the intraladder couplings from eq . were determined to be @xmath385 , @xmath386 with different experimental techniques and at different experimental conditions@xcite . in this paper
, we use the values @xmath387 recently a slight anisotropy of the order of @xmath388 of @xmath36 has been discovered by esr@xcite measurements .
this anisotropy could explain the small discrepancies between the couplings found in different experiments .
the magnetic field in tesla is related to @xmath31 replacing @xmath389 in eq . with @xmath390 being the bohr magneton and @xmath391 being the land factor of the unpaired copper electron spins .
the latter depends on the orientation of the sample with respect to the magnetic field . for the orientation chosen in the nmr measurements@xcite ,
it amounts to @xmath392 .
it can vary up to @xmath393 for other experimental setups@xcite .
many interesting thermodynamic measurements have been performed on bpcb .
we select in the following some of these experiments and compare them to the theoretical predictions .
the longitudinal magnetization that can be measured very precisely by nmr ( see ref . ) was shown to agree remarkably with the one computed using the weakly coupled ladder model ( fig .
[ fig : mz_zerot ] ) . unfortunately , the magnetization is not very sensitive to the underlying model ( sec .
[ sec : criticalfields ] ) .
thus it can not be used to select between various models , but once the model is fixed , it can be used to fix precisely the parameters given the high accuracy of the experimental data . in particular , the position of the critical fields are very sensitive to the values of the intraladder couplings ( sec . [
sec : criticalfields ] ) .
the couplings determined by this method are @xmath394 and @xmath395 .
a more selective test to distinguish between various models is provided by the specific heat .
this is due to the fact that the specific heat contains information on high energy excitations which are characteristic for the underlying model . as shown in fig .
[ fig : cvst ] the experimental data are remarkably described , up to an accuracy of a few percent , by a simple heisenberg ladder hamiltonian with the parameters extracted from the magnetization . in particular , not only the low temperature behavior is covered by the ladder description , but also the higher maxima .
this indicates that the ladder hamiltonian is an adequate description of the compound .
the small discrepancies between the specific heat data and the calculation which is essentially exact can have various sources . first of all , the substraction of the non - magnetic term in the experimental data can account for some of the deviations .
furthermore the interladder coupling can induce slight changes in the behavior of the specific heat .
finally deviations from the simple ladder hamiltonian can be present .
small anisotropy of the couplings can exist and indeed are necessary to interpret recent esr experiments @xcite .
other terms such as longer range exchanges or dzyaloshinskii - moryia ( dm ) terms might occur along the legs even if the latter is forbidden by symmetry along the dominant rung coupling .
clearly all these deviations from the heisenberg model can not be larger than a few percents .
they will not lead to any sizeable deviation for the luttinger parameters ( fig .
[ fig : llparameter ] ) in the one dimensional regime .
close to the critical points they can , however , play a more important role .
it would thus be interesting in subsequent studies to refine the model to take such deviations into account .
after having fixed the model and the intraladder couplings up to a few percents we use it to compute other experimentally accessible quantities such as the magnetostriction , thermal expansion , the nmr relaxation rate , the transition temperature to the ordered phase and its order parameter . in refs .
the magnetostriction and thermal expansion were compared to the theoretical results using the described ladder model .
a very good agreement was found in a broad range of temperature ( not shown here ) .
note that only a full ladder model allows a global quantitative description of the magnetostriction effect which provides an additional confirmation of the applicability of the model .
the quality of the determination of the model and its intraladder parameters becomes even more evident in the comparison of the nmr data for the relaxation rate @xmath181 with the theoretical results of the luttinger liquid theory as shown in fig .
[ fig : t1 ] .
one _ adjustable parameter is left , namely the hyperfine coupling constant ( see sec .
[ sec : relaxationtime ] ) .
this parameter allows for a global expansion of the theoretical curve , but not for a change of its shape .
the agreement between the theory and the experimental data is very good over the whole range of the magnetic field and only small deviations can be seen .
the compound bpcb thus allows to _ quantitatively _ test the luttinger liquid universality class .
even though the luttinger liquid description is restricted to low energies , in bpcb its range of validity is rather large .
indeed at high energy , its breakdown is approximately signaled by the first peak of the specific heat@xcite ( see sec . [
sec : specificheat ] ) . here
this has a maximum scale of about @xmath396 at midpoint between @xmath109 and @xmath67 ( see fig .
[ fig : tmagnetization].c ) . given the low ordering temperature which has a maximum at about @xmath397 this leaves a rather large luttinger regime for this compound .
taking now the coupling between ladders into account , one can induce a transition to a three - dimensional ordered phase .
the transition temperature is shown in fig .
[ fig : criticaltemperature ] .
experimentally it is determined by nmr@xcite and neutron diffraction measurements @xcite .
theoretically the ladders are described by luttinger liquid theory and their interladder coupling is treated in a mean - field approximation ( secs .
[ sec : luttinger_liquid ] and [ sec : mean - field ] ) . as shown in fig .
[ fig : criticaltemperature ] , the luttinger liquid theory provides a remarkable description of the transition to the transverse antiferromagnetic order at low temperatures .
the shape of @xmath398 is almost perfectly reproduced , in agreement with both the nmr@xcite and the nd data@xcite .
the comparison with the experiments determines the interladder coupling @xmath89 , the only adjustable parameter .
the simple mean - field approximation would give a value of @xmath399 .
as discussed in sec .
[ sec : transitiontemperature ] mean - field tends to underestimate the coupling and it should be corrected by an essentially field independent factor .
taking this into account we obtain a coupling of the order of @xmath400 .
the order parameter in the antiferromagnetic phase can also be observed by experiments .
it shows a very interesting shape . at a pure experimental level neutron diffraction and nmr
have some discrepancies as shown in fig .
[ fig : orderparameter ] .
these discrepancies can be attributed to the different temperatures at which the data has been taken , and a probable underestimate of the temperature in the neutron diffraction experiments@xcite .
indeed the order parameter close to the critical magnetic field @xmath67 is very sensitive to temperature , since the transition temperature drops steeply in this regime .
note that although the nmr allows clearly for a more precise measurement of the transverse staggered magnetization it can not give its absolute value .
thus the amplitude of the order parameter is fixed from the neutron diffraction measurement . even though a good agreement between the theoretical results and the experimental results is obtained ,
several questions concerning the deviations remain to be addressed .
first , the theoretical curve does not fully follow the shape of the experimental data . particularly at high fields
the experimental data shows a stronger decrease .
a simple explanation for this effect most likely comes from the fact that the calculation is performed at zero temperature , while the measurement is done at @xmath401 .
this is not a negligible temperature with respect to @xmath86 , in particular at magnetic fields close to @xmath67 .
extrapolation of the experimental data to zero temperature@xcite improves the agreement .
however , for a detailed comparison either lower temperature measurements or a calculation of the transverse staggered magnetization at finite temperature would be required .
both are quite difficult to perform and will clearly require further studies .
the second question comes from the amplitude of the staggered magnetization .
indeed the experimental data seem to be slightly above the theoretical curve , even if one uses the value @xmath402 for the interladder coupling .
this is surprising since one would expect that going beyond the mean - field approximation could only reduce the order parameter .
naively , one would thus need a larger coupling , perhaps of the order of @xmath403 to explain the amplitude of the order parameter .
this is a much larger value than the one extracted from the comparison of @xmath86 .
how to reconcile these two values remains open .
the various anisotropies and additional small perturbations in the ladder hamiltonian could resolve part of this discrepancy .
however , it seems unlikely that they result in a correction of @xmath89 by a factor of about 2 - 3 .
another origin might be the presence of some level of frustration present in the interladder coupling .
clearly more experimental and theoretical studies are needed on that point .
the inelastic neutron scattering ( ins ) technique is a direct probe for dynamical spin - spin correlation functions .
measurements have been performed on the compound bpcb in the spin liquid phase@xcite ( low magnetic field ) and in the gapless regime@xcite .
modeling the compound bpcb by two inequivalent uncoupled ladders oriented along the two rung vectors @xmath380 ( see fig .
[ fig : structure ] ) the magnetic ins cross section@xcite is given by the formula @xmath404 \\
\left.+2\left(1-\frac{{q^z}^2}{\mathbf{q}^2}\right ) \left[\cos(\mathbf{q}\cdot \mathbf{d}_i)s^{zz}_{21}+s^{zz}_{11}\right]\right\}.\end{gathered}\ ] ] here @xmath405 is the momentum transferred to the sample ( @xmath406 , @xmath407 are the incident , scattered neutron momenta ) and @xmath408 is the transferred energy ( @xmath409 , @xmath410 are the incident , scattered neutron energies ) .
the correlations @xmath411 are understood to be evaluated at the momentum @xmath412 and frequency @xmath413 , and are defined by @xmath414 ( @xmath222 sign if @xmath415 and @xmath223 sign if @xmath416 ) with @xmath417 defined at zero temperature in eq .
( [ equ : correlation1 ] ) and evaluated for a momentum @xmath418 along the @xmath375 unit cell vector ( momentum along the ladder direction ) . the magnetic form factor @xmath419 of the @xmath374 and the ratio @xmath420
are corrected in the experimental data .
the ins cross section ( [ equ : laddercross_section ] ) is directly related to a combination of different correlation functions @xmath243 with weights depending on the transferred momentum @xmath421 and the magnetic field orientation . in the model definition
( see sec .
[ sec : model ] ) , the magnetic field @xmath422 is pointing along the @xmath25 direction . aligning it to the @xmath376 unit cell vector and tuning @xmath421 in the @xmath423-plane ( @xmath424 , @xmath425 and @xmath426 are the reciprocal vectors of @xmath375 , @xmath376 and @xmath377 ) allows one to keep constant the prefactors in eq .
scanning the @xmath427-momenta with the condition @xmath428 to target the symmetric or antisymmetric part , respectively .
we focus here on the antisymmetric part for which the low energy spectra have already been studied experimentally and theoretically@xcite .
theoretically the focus so far lay on the description by the spin chain mapping .
we compute here the ins cross section ( [ equ : laddercross_section ] ) for the full ladder at @xmath429 using the correlations presented in sec .
[ sec : dynamicalcorrelation ] .
the results are shown in figs .
[ fig : theoreticalins1 ] and [ fig : theoreticalins2 ] and are compared to the results from the spin chain mapping . as expected from expression ( [ equ : laddercross_section ] ) , it contains the different excitations present in the spectra of @xmath430 and @xmath431 ( cf . figs .
[ fig : zzcorrelationmz].b , [ fig : pmcorrelationmz].b and [ fig : mpcorrelationmz].b ) : * the low energy continuum originates from the transversal correlations @xmath431 .
it is qualitatively well described by the spin chain mapping that presents a symmetry with respect to half magnetization . *
the continuum of excitations at energy @xmath252 comes from the longitudinal correlation @xmath430 and is not present in the spin chain mapping . *
the continuum of excitations at energy @xmath432 comes from the transversal correlation @xmath433 and is not present in the spin chain mapping .
( @xmath434 ) and @xmath418 at @xmath268 in ( a ) a ladder system and ( b ) the spin chain mapping .
the horizontal dashed lines correspond to the constant energy scans at @xmath435 shown in fig .
[ fig : experimentalins1].[fig : theoreticalins1 ] ] the main features of the low energy continuum ( a ) are well covered by the spin chain mapping@xcite
. however , slight differences between the low energy excitations in the spin ladder and the spin chain are still visible ( cf . also sec .
[ sec : lowenergyexcitations ] ) .
they can even be distinguished in the experimental data as shown in figs .
[ fig : experimentalins1 ] and [ fig : experimentalins2 ] where some cuts at fixed energy @xmath436 are plotted .
the ins measured intensity is directly compared to the theoretical cross section ( [ equ : laddercross_section ] ) computed for the ladder and the spin chain mapping at @xmath437 convolved with the instrumental resolution .
the amplitude is fixed by fitting one proportionality constant for all fields , energies , and wave vectors .
these scans at fixed energy present peaks when the lower edge of the continua ( related to the correlations @xmath431 ) is crossed ( see dashed white lines in figs .
[ fig : theoreticalins1 ] and [ fig : theoreticalins2 ] ) .
as one can see , the theoretical curves for the ladder and the spin chain both reproduce well the main features in the experimental data and only small differences are present : * the spectral weight intensity at @xmath268 and @xmath438 ( in fig .
[ fig : experimentalins1].b ) is slightly overestimated by the spin chain mapping . * the height of the two central peaks at @xmath439 and @xmath440 ( in fig .
[ fig : experimentalins2].c ) is underestimated by the spin chain mapping . whereas the low energy excitations ( a ) only showed a slight asymmetry with respect to the magnetization , a very different behavior can be seen in the high energy part ( b)-(c ) . indeed , the high energy part of the ins cross section ( fig .
[ fig : theoreticalins2 ] ) is very asymmetric with respect to half magnetization . as we discussed in sec .
[ sec : dynamicalcorrelation ] , these excitations are due to the high energy triplets that can be excited in @xmath430 and @xmath433 ( see fig . [
fig : zzcorrelationmz].b and fig .
[ fig : pmcorrelationmz].b ) and are totally neglected in the spin chain mapping .
their corresponding spectral weight is of the same order than the low energy spectra , and thus should be accessible experimentally .
it would be very interesting to have an experimental determination of this part of the spectrum , since as we have seen it contains characteristic information on the system itself and related to itinerant systems via the various mappings . of bpcb@xcite with the momentum @xmath441 in the rung direction ( @xmath442 ) at @xmath443 ( @xmath444 ) and @xmath445 after substraction of the zero - field background .
in each panel , fixed energy scans ( shown by white dashed lines in fig .
[ fig : theoreticalins1 ] ) are plotted : ( a ) @xmath440 , ( b ) @xmath438 .
the circles correspond to the experimental data .
the red ( black ) solid lines are the @xmath268 theoretical data for the ladder ( the spin chain mapping ) convolved with the instrumental resolution .
the shaded bands indicate the error bar in the experimental determination of a single proportionality constant valid for all fields , energies , and wave vectors .
the width of these areas combines the statistics of all our scans with uncertainties in the exact magnetization values at the chosen fields and in the convolution procedure .
[ fig : experimentalins1 ] ] of bpcb@xcite with a @xmath441 momentum in the rungs ( @xmath442 ) at @xmath445 after substraction of the zero - field background . in each panel , cuts at fixed energy ( shown by white dashed lines in fig .
[ fig : theoreticalins2 ] ) are plotted : ( a ) @xmath446 and @xmath447 , ( b ) @xmath446 and @xmath448 , ( c ) @xmath449 and @xmath447 , ( d ) @xmath449 and @xmath448 .
the circles correspond to the experimental data .
the solid red ( black ) curves are the theoretical data for the ladder ( the spin chain mapping ) convolved with the instrumental resolution .
the shaded bands indicate the error bar in the experimental determination of a single proportionality constant valid for all fields , energies , and wave vectors .
the width of these areas combines the statistics of all our scans with uncertainties in the exact magnetization values at the chosen fields and in the convolution procedure.[fig : experimentalins2 ] ]
in this paper we have looked at the thermodynamic and dynamical properties of weakly coupled spin ladders under a magnetic field .
this was done by a combination of analytical techniques , such as bethe ansatz , bosonization and luttinger liquid theory , and numerical techniques such as dmrg and quantum monte carlo . using this combination of techniques we were able to explore the physical properties in the three main regions of the phase diagram of such spin ladders under a magnetic field , namely :
a ) a gapped spin liquid at low fields ; b ) a massless phase at intermediate fields @xmath450 ; c ) a saturated phase at larger fields . in addition to the theoretical analysis we compared our findings to the experimental results on the compound bpcb ( @xmath1 ) which is an excellent realization of such ladder systems . for thermodynamics we computed the magnetization and specific heat of the system as a function of temperature and magnetic field .
the extension of the dmrg technique to finite temperature allows us to compute these quantities with an excellent accuracy . in the gapless
phase the low energy part of the specific heat agrees well with the prediction of the luttinger liquid theory , which is the low energy theory describing most of massless one dimensional systems . at higher temperatures
, the numerical solution is needed to capture the precise structure of the peaks in the specific heat , that reflect the presence of the excited states in the ladder .
comparison of the theoretical calculations with the measured magnetization and specific heat proves to be remarkable .
this good agreement confirms that the ladder model is indeed a faithful description of the compound bpcb .
it also gives direct access , via the maxima in magnetization and peaks in the specific heat to the approximate region of applicability of the luttinger liquid description .
for the low energy dynamics we used a combination of the numerical techniques to determine the luttinger liquid parameters and then the analytical description based on luttinger liquids to compute the dynamical spin - spin correlation functions . this allowed to extract the nmr relaxation rate and the one dimensional antiferromagnetic transverse susceptibility
. if the ladders are very weakly coupled , which is the case in the considered material , the divergence of the susceptibility leads to a three dimensional antiferromagnetic order at low temperatures in the direction transverse to the applied magnetic field .
we computed this transition temperature and the order parameter at zero temperature .
comparison with the measured experimental quantities both by nmr and neutron diffraction proved again to be remarkable .
this excellent agreement between theory and experiment for these quantities as a function of the magnetic field allows to _ quantitatively _ test the luttinger liquid theory .
indeed it shows that several different correlations are indeed totally described by the knowledge of the two luttinger liquid parameters ( and the amplitudes relating the microscopic operators to the field theory one ) .
this is something that could not really be tested previously since either the microscopic interactions were not known in details , leaving the luttinger parameters as adjustable parameters , or only one correlation function could be measured in a given experiment , not allowing to test for universality of the description .
we also gave a detailed description of the dynamical spin - spin correlations , for @xmath451 using the time dependent dmrg method , for a wide range of energies and all momenta .
the excitations reveal many important informations on the system and are well suited to characterize it .
in particular we show the interesting evolution of the excitations in the system with the magnetic field and with different coupling strengths .
quite interestingly the intermediate energy part can be related to the excitations of a t - j model and shows thereby features of itinerant systems .
we also showed that the dynamical correlations of the ladder posses characteristic high energy features that are clearly distinct from the corresponding spectrum for spin chains .
the numerical calculation is efficient for the high and intermediate energy part of the spectrum for which the luttinger liquid description can not be applied .
we showed that the two methods , numerics and ll have enough overlap , given the accuracy of our calculation so that we can have a full description of the dynamical properties at all energies .
this allowed us to use each of the method in the regime where it is efficient .
in particular , in the present study we did not push the numerical calculations to try to obtain the exact behavior at low energies , but focus on the high and intermediate energy regime .
we used the analytical description coupled to the numerical determination of the luttinger parameters to obtain a very accurate low energy description .
we made the connection between our results and several analytical predictions .
in particular at intermediate - low energy our calculation agrees with the luttinger liquid prediction of incommensurate points and behavior ( divergence , convergence ) of the correlations .
we compared our numerical results with existing ins data on the compound bpcb and found excellent agreement .
it is rewarding to note that the resolution of our dynamical calculation is , in energy and momentum , at the moment better than the one of the experiment .
the comparison between theory and experiment is thus essentially free of numerical error bars .
given the current resolution of the ins experiment it is difficult to distinguish in the low energy part of the spectrum the difference between the dynamical correlation of the true ladder and the one of an anisotropic spin @xmath0 system , which corresponds to the strong rung exchange limit .
more accurate experiments would be desirable in this respect .
an alternative route is to probe experimentally the high energy part of the spectrum , since these high energy excitations contain many characteristic features of the underlying model . on the conceptual side and also in connection with
the compound bpcb several points remain to be investigated .
an extension of the dynamical results to finite temperature would be desirable .
this could be used to study different effects such as the interesting shifts and damping of the triplet modes with temperature that have been observed in three dimensionnal gapped system@xcite .
a second important step would be to improve the description of the quasi one dimensional systems , by including in a mean - field way the effect of the other ladders in the numerical study .
this is something we already partly performed , but the extension to the dynamical quantities remains to be done .
this is specially important close to the quantum critical points @xmath109 and @xmath67 where the interladder coupling becomes crucial and the system undergoes a dimensional crossover between a one dimensional and a higher dimensional ( three dimensional typically ) behavior .
understanding such a crossover is a particularly challenging question since the system goes from a description for which an image of essentially free fermions applies ( in the one dimensional regime ) to one for which a description in terms of essentially free bosons ( the three dimensional regime ) applies .
one step further would be the extension of the investigation of the combined numerical - analytical methods we used here to other systems including ladder systems with certain asymmetries or in the presence of disorder .
we thank t. barthel , c. berthod , j .- s .
caux , v. v. cheianov , a. furusaki , u. schollwck , s. trebst , and s. white for useful discussions and specially j .- s .
caux for the data exchange .
we acknowledge support of the triangle de la physique , the anr famous , and of the swiss nsf , under manep and division ii .
t. giamarchi would like to thank the aspen center for physics , where part of this work was completed .
m. klanjek acknowledges support from the centre of excellence en@xmath452fist , dunajska 156 , si-1000 ljubljana , slovenia .
in this appendix we show how the spin ladder hamiltonian at strong coupling ( @xmath453 ) can be expressed in bosonic operators acting on single bonds introduced in ref . .
this representation allows a classification of the excitations due to their position in energy .
we first derive perturbatively an effective system based on this hilbert space organization by energy sectors .
we introduce the schrieffer - wolff transformation that maps the physical system to the effective , and approximate it using a strong coupling expansion .
thus , we evaluate the rung densities of the ground state in the spin liquid , and derive an effective theory for the gapless regime .
furthermore we evaluate the deviation of the ll parameters from the spin chain mapping .
the four - dimensional hilbert space on each rung @xmath10 is spanned by the states @xmath454 @xmath455 @xmath456 and @xmath70 ( cf .
eqs . and ) , obtained by applying the boson creation operators @xmath457 @xmath458 @xmath459 and @xmath460 to a vacuum state .
a hardcore boson constraint applies on each rung @xmath10 , i.e. @xmath461 where @xmath462 and @xmath463 , @xmath464 .
while the hamiltonian on the rung @xmath465 is quadratic in the boson operators @xmath466 -\frac34l,\ ] ] the chain hamiltonian @xmath467 is quartic , and its structure is quite complex .
the advantage of the boson representation reveals when considering the case of small @xmath74 . in that case
we perform a schrieffer - wolff transformation of the spin ladder hamiltonian @xmath468 the hermitian operator @xmath469 can be expanded in powers of @xmath74 @xmath470 thus @xmath471 can be written in orders of @xmath74 as @xmath472 where @xmath473 , \\ & h^{(2 ) } = i[a_2,h_\perp ] -\frac12 [ a_1,[a_1,h_\perp]]+i [ a_1,h_\parallel],\label{equ : h2_def}\end{aligned}\ ] ] etc . through this expansion , the unitary transform @xmath474 can be perturbatively determined computing the @xmath475 recursively in order to eliminate the transitions between the energy sectors of excitations in @xmath476 . since the first term @xmath477 in eq
. leads to a separation of excitations on the order of the energy scale @xmath36 ( cf .
[ fig : phasediagramm].a ) the decoupled bond limit provides the effective hilbert space that contributes to each energy sector .
the second term @xmath478 causes broadening of these bands on the order of @xmath35 and can induce a complex structure within the energy bands . to obtain the desired expansion up to the first order in @xmath74 we choose @xmath479 where h.c
. stands for the hermitian conjugation . in the spin liquid phase
, the decoupled bond limit provides the effective ground state @xmath480 which is related to the physical ground state by @xmath481 so the triplet density of the ground state @xmath482 ( with @xmath464 ) is given by @xmath483 using eq . , and
keeping only the non - vanishing corresponding terms in up to second order we get @xmath484 in the case of the compound bpcb ( see sec .
[ sec : bpcb ] ) this expansion gives @xmath485 , and due to the hardcore boson constraint ( eq . ) @xmath486 . even though we took into account only the first order term for @xmath469 in eq .
, this approximation of the triplet density differ from the direct numerical computations ( in fig . [
fig : tripletdensity ] ) of only @xmath487 .
the first order term @xmath488 of the effective hamiltonian is computed substituting into .
this leads to @xmath488 in the form @xmath489 where @xmath490 @xmath491 @xmath492 @xmath493 and @xmath494 here we regrouped the terms such that each @xmath495 acts on the corresponding energy sector @xmath496 , @xmath497 in the gapless regime .
note that in each sector @xmath498 such that to the considered order in @xmath74 the hamiltonian corresponds to the effective hamiltonian .
when focusing on the low energy sector , the @xmath53 and the @xmath55 modes dominate the behavior and we can assume a vanishing density of @xmath499 and @xmath500 triplets .
thus the hardcore boson constraint simplifies to @xmath501 and the rung hamiltonian to @xmath502 further the only contribution to the first order term in @xmath74 comes from @xmath503 . taking this into account we obtain from eq . for the hamiltonian in the lowest energy sector @xmath504 where @xmath465 is given by eq . and @xmath503 by eq . .
following ref . , we map the two low energy modes on the two states of a pseudo spin ( eq . ) and replace the boson operators @xmath505 and @xmath506 with the spin-@xmath0 operators @xmath507 the effective hamiltonian is the spin-@xmath0 xxz heisenberg hamiltonian , eq . .
the effective hilbert space of the @xmath36 energy sector corresponds to a single @xmath69 triplet excitation lying in a sea of singlets @xmath53 and triplets @xmath76 .
the effective hamiltonian up to first order in @xmath74 is given by @xmath508 the excitation @xmath69 can be interpreted as a single hole excitation in a spin chain formed by @xmath58 and @xmath246 .
each rung state of this sector is identified with @xmath509 in this picture the hamiltonian can be mapped onto the anisotropic t - j model @xmath510 where @xmath511 is an energy shift .
the hopping term @xmath512 stems from the term @xmath513 in eq . .
here @xmath329 ( @xmath330 ) is the creation ( annihilation ) operator of a fermion with pseudo spin @xmath514 at the site @xmath10 .
note that although we are dealing here with spin states , it is possible to faithfully represent the three states of each site s hilbert space ( @xmath53 , @xmath76 , @xmath69 ) using a fermion representation .
additionally a nearest neighbor density - density term between the up spin and the hole arises @xmath332.\ ] ] here @xmath333 is the density operator of the hole at site @xmath10 .
this term stems from the nearest - neighbour interaction between the @xmath55 triplets , i.e. the second term in equ . .
mapping this term onto a spin chain in the presence of a hole leads to an interaction term between the hole and the spin - up state .
note that this is in contrast to the usual mapping on a spin chain without a hole in which case it would only causes a shift in energy .
the second order term @xmath515 of the expansion , contains a huge amount of terms .
nevertheless considering the low energy sector , only the following terms @xmath516 are important . the first term in eq .
is a singlet density - density interaction which can be absorbed into the coupling of the xxz chain , and the second term is a conditionnal hopping@xcite . in order to study the effects of @xmath517 on the ll parameters @xmath95 and @xmath94 ( see fig . [
fig : llparameter ] ) , we first replace the boson operators with the spin-@xmath0 operators ( eq . ) .
so the hamiltonian in the low energy sector becomes @xmath518+\text{const}\end{gathered}\ ] ] where @xmath519 is the xxz heisenberg chain hamiltonian ( eq . ) with the corrected parameters @xmath520 up to the second order in @xmath74 .
for the bpcb parameters ( eq . [ equ : couplingratio2 ] ) @xmath521 instead of @xmath522 for the spin chain mapping ( first order approximation ) .
the ll parameters @xmath95 , @xmath94 and @xmath101 of @xmath519 with the anisotropy @xmath523 are computed , and we treat the three terms interaction ( conditional hopping ) approximating @xmath524 ( mean - field approximation ) . the remaining term is then bosonized using the expression for @xmath525 .
it leads to the corrected ll parameter @xmath526 and @xmath527 of the hamiltonian through the relations @xmath528 the corrected @xmath526 and @xmath527 are plotted in fig .
[ fig : llparameter ] and clearly show the asymmetric signature of the full ladder parameters induced by the conditional hopping term in .
note that the lack of convergence @xmath529 when @xmath530 is obviously an artefact of the mean - field approximation @xmath524 and the big sensitivity in the @xmath94 determination with the ratio of eq . .
in the following , we present several properties of the luttinger liquid providing a quantitative description of the low energy physics of the gapless regime in the spin ladder model and the spin chain mapping .
we start with the determination of the parameters that totally characterize the ll description .
then we summarize the spin - spin correlation functions deduced from the ll theory at zero and finite temperature . in this appendix
we detail the determination of the ll parameters @xmath95 , @xmath94 and the prefactors @xmath101 , @xmath102 and @xmath103 ( see eqs . , and ) .
the parameters @xmath94 , @xmath101 , @xmath102 and @xmath103 and their dependence on the magnetic field have been previously determined in refs . for different values of the couplings than considered here .
we obtain these parameters in two steps@xcite : * we determine the ratio @xmath531 from its relation to the static susceptibility , eq . , which is a quantity numerically easily accessible with dmrg or bethe ansatz ( for the spin chain only ) .
* the parameters @xmath94 and the prefactors @xmath101 , @xmath102 and @xmath103 are extracted by fitting numerical results for the static correlation functions obtained by dmrg with their analytical ll expression . for the spin chain ,
it is also possible to deduce the product @xmath532 from the magnetic stiffness computed by bethe ansatz@xcite .
we give the relations used for the spin chains only , since from these the relations for the spin ladders can be easily inferred using the following relations @xmath533 the ll theory predicts that the static susceptibility @xmath534 is related to the ratio @xmath531 through the relation@xcite @xmath535 we numerically compute the static susceptibility using dmrg or bethe ansatz ( for the spin chain only ) and infer the ratio @xmath531 with a negligible error . for the extraction of the parameters @xmath94 , @xmath101 , @xmath102 and @xmath103 ,
we fit numerical dmrg results for the correlation functions @xmath536 , @xmath537 and the local magnetization @xmath538 with their analytical expression for finite system size , eqs . ( 11 ) , ( 12 ) and ( 13 ) in ref . respectively . in the limit of infinite system size and
far from the boundaries these ll correlations simplify to the well known power law decay @xmath539}{|l - l'|^{2k+\frac{1}{2k}}}\ ] ] @xmath540}{|l - l'|^{2k}}-\frac{k}{2\pi^2|l - l'|^2}\ ] ] and the local magnetization becomes constant , @xmath541 . ,
@xmath94 and the prefactors of the spin operators @xmath101 , @xmath102 , @xmath103 versus the magnetization per rung @xmath119 computed for a spin ladder with the bpcb couplings ( red crosses ) and for the spin chain mapping ( blue stars ) .
the strong coupling expansion of @xmath95 and @xmath94 up to the second order in @xmath74 ( discussed in appendix [ sec : secondorder_ll ] ) is plotted in black dashed lines.[fig : llparameter ] ] we first fit the transverse correlation ( @xmath542-correlation @xmath536 ) to extract the parameters @xmath94 , @xmath101 , and @xmath102 .
then we use the previously extracted value for @xmath94 to fit the longitudinal correlation ( @xmath543-correlation @xmath537 ) and the magnetization , @xmath538 , which allow us to determine @xmath103 .
the values determined by both fits are very close and in fig .
[ fig : llparameter ] the average value of both is shown .
all the results presented in fig . [ fig : llparameter ] were done for @xmath544 and several hundred dmrg states after an average on the four sets of used data points in the fit @xmath545 , @xmath546 , @xmath547 , @xmath548 .
the error bars correspond to the maximum discrepancy of these four fits from the average .
we further checked that different system lengths lead to similar results .
the ll parameters of the ladder system for the bpcb couplings are presented in fig .
[ fig : llparameter ] as a function of the magnetization per rung .
additionally we show the parameters of the spin chain mapping ( computed for the spin chain hamiltonian ) for comparison .
when the ladder is just getting magnetized , or when the ladder is almost fully polarized , @xmath529 ( free fermion limit ) and @xmath549 ( because of the low density of triplons in the first case , and low density of singlets in the second case ) .
for the spin chain mapping , the reflection symmetry around @xmath268 arises from the symmetry under @xmath441 rotation around the @xmath113 or @xmath550axis of the spin chain .
this symmetry has no reason to be present in the original ladder model , and is an artefact of the strong coupling limit , when truncated to the lowest order term as shown in appendix [ sec : tjmodelmapping ] .
the values for the spin ladder with the compound bpcb parameters can deviate strongly from this symmetry .
the velocity @xmath95 and the prefactor @xmath102 remain very close to the values for the spin chain mapping .
in contrast , the prefactors @xmath103 , @xmath101 and the exponent @xmath94 deviate considerably and @xmath101 and @xmath94 become strongly asymmetric .
the origin of the asymmetry lies in the contribution of the higher triplet states @xcite , and can be understood using a strong coupling expansion of the hamiltonian up to the second order in @xmath74 ( see appendix [ sec : secondorder_ll ] ) .
this asymmetry has consequences for many experimentally relevant quantities and it was found to cause for example strong asymmetries in the 3d order parameter , its transition temperature and the nmr relaxation rate ( see fig . [ fig : orderparameter ] , fig .
[ fig : criticaltemperature ] and fig .
[ fig : t1 ] ) . in this appendix
we summarize previous results of the luttinger liquid description of the dynamical correlation functions@xcite at zero temperature .
note that the weak and strong coupling limits which have been studied separately @xcite can be connected @xcite .
the expression of the correlations in the whole regime is given by @xmath551 + \frac{k\omega}{u}\theta(\omega)\left[\delta(\omega - uq)+\delta(\omega+uq)\right]\end{gathered}\ ] ] @xmath552}\right)^{1-\eta_-}\left(\frac{2u}{\omega+u[q-\pi(1 - 2m^z)]}\right)^{1-\eta_+}\right.\\ \left.+\theta(\omega - u|q-\pi(1 + 2m^z)|)\left(\frac{2u}{\omega - u[q-\pi(1 + 2m^z)]}\right)^{1-\eta_+}\left(\frac{2u}{\omega+u[q-\pi(1 + 2m^z)]}\right)^{1-\eta_-}\right]\end{gathered}\ ] ] with the heaviside function @xmath553 and @xmath554 .
the correlation @xmath291 is obtained replacing @xmath555 in the @xmath288 expression eq . .
( color online ) : map of the momentum - frequency low energy correlations of the ll model where the white areas represent the continuum of excitations . in the striped areas no excitations are possible .
( a ) @xmath556 : the dash - dotted lines ( blue ) are the excitation peaks close to @xmath276 and the dashed lines ( red ) are the continuum lower boundary with edge exponent @xmath557 close to @xmath558 .
( b ) @xmath559 , ( c ) @xmath560 : the continuum lower boundary close to @xmath561 is represented by solid lines ( black ) ( edge exponent @xmath562 ) , dashed lines ( red ) ( edge exponent @xmath563 ) and dash - dotted lines ( blue ) ( edge exponent @xmath564 ) . ] .
the solid ( dashed ) lines are determined from the ladder ( spin chain mapping ) exponent @xmath94 in fig .
[ fig : llparameter ] .
the exponent @xmath557 of the @xmath284 correlations is shown in ( a.1 ) , and the exponent @xmath562 of the @xmath285 correlations at the @xmath111 branch in ( a.2 ) ( lower red curves ) .
the exponents @xmath563 ( upper black curves ) in ( a.2 ) and @xmath564 in ( a.3 ) correspond of both sides of the incommensurate branches of the @xmath285 ( see fig . [
fig : llcorrschema ] ) .
( b ) shift of the ground state energy per rung versus the magnetization @xmath565.[fig : decay_exp ] ] the expressions eq . and eq . exhibit the typical behavior of the frequency - momentum ll correlations : a continuum of low energy excitations exists with a linear dispersion with a slope given by the luttinger velocity @xmath566 .
the spectral weight at the lower boundary of the continuum displays an algebraic singularity with the exponents related to the luttinger parameter @xmath94 .
a summary of this behavior is sketched in fig .
[ fig : llcorrschema ] .
for the considered system the longitudinal correlation @xmath284 is predicted to diverge with the exponent @xmath557 at its lower edge . as shown in fig .
[ fig : decay_exp].a.1 the exponent of this divergence is very weak @xmath567 for the parameters of bcpb .
the transverse correlations @xmath285 exhibit a distinct behavior depending on the considered soft mode .
close to @xmath111 the weight diverges with an exponent given by @xmath562 .
this divergence is strong for the considered parameters ( @xmath319 in fig .
[ fig : decay_exp].a.2 ) .
in contrast at the soft mode @xmath568 a divergence ( cusp ) is predicted at the lower edge with the exponent @xmath569 in fig .
[ fig : decay_exp].a.2 ( @xmath570 in fig .
[ fig : decay_exp].a.3 ) .
the determination of the relaxation time @xmath190 ( see sec . [
sec : relaxationtime ] ) and of the transition temperature @xmath86 to the 3d - ordered phase ( see sec . [
sec : transitiontemperature ] ) requires the staggered transverse retarded correlation function @xmath571\right\rangle\ ] ] written in term of the spin chain mapping operators with @xmath572 the heaviside function . in the gapless regime , using the bosonization formalism , and taking into account only the most relevant term , we can compute it as described in ref .
for the ll hamiltonian : @xmath573 and its fourrier transform : @xmath574 where @xmath575 .
in this appendix , we describe first the time dependent dmrg method and its extension to finite temperature .
then we give the technical details related to the computation of the momentum - frequency correlations . the t - dmrg@xcite ( time dependent dmrg )
method is based on the principle of the original dmrg to choose an effective reduced hilbert space to describe the physics one is interested in .
instead of choosing only once the effective description for the evolution of the state , the t - dmrg adapts its effective description at each time - step .
the implementation of this idea can be performed using different time - evolution algorithms .
here we use the second order trotter - suzuki expansion for the time - evolution operator using a rung as a unit@xcite .
the errors arising in this method are the so - called truncation error and the trotter - suzuki error .
both are very well controlled ( see ref . for a detailed discussion ) .
the main idea of the finite temperature dmrg @xcite ( t - dmrg ) is to represent the density matrix of the physical state as a pure state in an artificially enlarged hilbert space .
the auxiliary system is constructed by simply doubling the physical system .
starting from the infinite temperature limit the finite temperature is reached evolving down in imaginary time .
the infinite temperature state in this auxiliary system corresponds to the totally mixed bell state @xmath576 , where @xmath577 is the state at the bond @xmath10 of the auxiliary system that has the same value @xmath578 on the two sites of the bond ( physical and its copy ) .
the sum @xmath579 is done on all these @xmath580 states @xmath581 .
we evolve the physical part of @xmath582 in imaginary time to obtain @xmath583 using the t - dmrg algorithm presented in appendix [ sec : timedmrg ] with imaginary time .
we renormalize this state at each step of the imaginary time evolution .
thus the expectation value of an operator @xmath584 acting in the physical system with respect to the normalized state @xmath585 is directly related to its thermodynamic average , i.e. @xmath586}{\mathrm{tr}[e^{-\beta h}]}=\langle\psi(\beta)|o |\psi(\beta)\rangle\ ] ] at the temperature @xmath587 .
we use this method to compute the average value of the local rung magnetization @xmath588 and energy per rung @xmath589 in the center of the system .
additionally we extract the specific heat @xmath590 by @xmath591 where @xmath592 is the imaginary time - step used in the t - dmrg .
to reach very low temperatures @xmath593 for the specific heat , we approximate the energy by its expansion in @xmath176 @xmath594 up to @xmath595 . the energy at zero temperature @xmath235 is determined by a standard @xmath451 dmrg method . since @xmath589 has a minimum at @xmath451 the linear term in the expansion does not exist .
the numbers @xmath596 ( @xmath597 ) are obtained fitting the expansion on the low @xmath176 values of the numerically computed @xmath589 .
typical system lengths used for the finite temperature calculations are @xmath598 ( @xmath599 for the spin chain mapping ) keeping a few hundred dmrg states and choosing a temperature step of @xmath600 ( @xmath601 for the spin chain mapping ) .
let us note that recently a new method has been developed to treat finite temperatures @xcite which is very promising to reach even lower temperatures . to obtain the desired frequency - momentum spin - spin correlations ( @xmath240 in ) ,
we first compute the correlations in space and time @xmath602 with @xmath603 , @xmath13 , and @xmath604 ( @xmath605 ) is the discrete time used .
these correlations are calculated by time - evolving the ground state @xmath606 and the excited state @xmath607 using the t - dmrg ( see appendix [ sec : timedmrg ] ) .
afterwards the overlap of @xmath608 and @xmath609 is evaluated to obtain the correlation function . in an infinite system reflection symmetry
would be fulfilled . to minimize the finite system corrections and to recover the reflection symmetry of the correlations , we average them @xmath610 we then compute the symmetric ( antisymmetric ) correlations ( upon leg exchange ) ( see sec . [ par : zerocorrelations ] ) @xmath611 with the rung momentum @xmath224 respectively .
finally , we perform a numerical fourrier transform due to the boundary effects and the limitation in the numerical precision , we compute the fourrier transform of @xmath612 only with its imaginary part that has no zero frequency component as proposed in ref . .
] @xmath613 for discrete momenta @xmath614 ( @xmath615 ) and frequencies @xmath413 are deduced from their value at positive time since @xmath616 , with @xmath617 , for translation invariant systems , and for correlations such as @xmath618 with @xmath226 . ] . the momentum @xmath229 has the reciprocal units of the interrung spacing @xmath619 ( @xmath620 is used if not mentioned otherwise ) . due to the finite time step @xmath621
, our computed @xmath240 are limited to the frequencies from @xmath622 to @xmath623 .
the finite calculation time @xmath624 induces artificial oscillations of frequency @xmath625 in @xmath240 . to eliminate these artefacts and reduce the effects of the finite system length
, we apply a filter to the time - space correlations before the numerical fourrier transform , i.e. @xmath626 we tried different functional forms for the filter @xmath627 ( cf .
as well ) . in the following the results are obtained by a gaussian filter @xmath628 if not stated otherwise .
as the effect of this filtering on the momentum - energy correlations consists to convolve them by a gaussian function @xmath629 , it minimizes the numerical artefacts but further reduces the momentum - frequency resolution .
typical values we used in the simulations are system lengths of up to @xmath630 sites while keeping a few hundreds dmrg states .
we limited the final time @xmath631 to be smaller than the time necessary for the excitations to reach the boundaries ( @xmath632 with @xmath95 the ll velocity in fig .
[ fig : llparameter ] ) in order to minimize the boundary effects .
the computations for the bpcb parameters , eq . , were typically done with a time step of @xmath633 up to @xmath634 ( but calculating the correlations only every second time step ) .
the momentum - frequency limitations are then @xmath635 and @xmath636 .
concerning the other couplings and the spin chain calculations , we used a time step @xmath637 up to @xmath638 ( also with the correlation evaluations every second time steps ) for a momentum - frequency precision @xmath639 and @xmath640 .
different techniques of extrapolation in time ( using linear prediction or fitting the long time evolution with a guess asymptotic form cf . refs . ) were recently used to improve the frequency resolution of the computed correlations .
nevertheless , as none of them can be apply systematically for our ladder system due to the presence of the high energy triplet excitations ( which result in very high frequency oscillations ) , we decided not to use them .
the three dimensional network of couplings of the coupled ladder hamiltonian eq .
[ equ : coupledladdershamiltonian ] and shown in fig .
[ fig : structure ] is not frustrated , and can therefore be simulated using quantum monte carlo algorithms .
we employ a stochastic series expansion implementation with directed loop updates @xcite provided with the alps libraries @xcite .
this numerical exact approach is complementary to mean - field approaches , because the later tend not to be quantitatively accurate due to the neglect of quantum fluctuations in the interladder coupling . in order to detect the transition temperature we measure the spin stiffness @xmath641 based on winding numbers in the three spatial directions . as pointed out in ref . , when plotting @xmath642 for different system sizes @xmath643 , the different curves cross at @xmath86 .
alternatively we measured the squared order parameter @xmath644 .
the quantity @xmath645 ( assuming the 3d xy universality class value of @xmath646 ) also crosses at @xmath86 when plotted for different @xmath643 . previously the 3d ordering temperatures of coupled spin ladders in a magnetic field have been determined using a specific feature of the magnetization @xmath647
@xcite . it turns out , that while the feature in @xmath647 indeed locates @xmath86 accurately for simple coupled dimer systems @xcite , the same does not hold for coupled ladders . in the ladder case one has to resort to the spin stiffness crossing or order parameter measurements to locate @xmath86 accurately .
when simulating the coupled ladder hamiltonian , a suitable finite size sample setup is required . due to the spatial anisotropies in the problem -
the ladder direction being singled out over the two transverse spatial directions - an appropriate aspect ratio needs to be kept @xcite . in our simulations we chose an aspect ratio of about 12 ,
i.e. the samples were 12 times longer along the ladder direction than in the transverse directions . in fig .
[ fig : qmc_tc ] we show simulation results for one particular set of couplings : the rung and leg couplings were set to @xmath648 and @xmath649 respectively , the @xmath391-factor was @xmath650 , the magnetic field amounted to @xmath651 and the interladder coupling @xmath89 was set to @xmath652 .
these couplings are identical to those used in ref . .
in the upper panel we show the rescaled spin stiffness @xmath653 for two different system sizes ( 768 versus 6144 spins ) .
one locates a crossing at @xmath654 for this observable . in the lower panel
we show the rescaled squared order parameter @xmath645 , which also exhibits a crossing at the same temperature .
these matching estimates for the critical temperature render us confident that we accurately locate the critical temperature .
based on this and subsequent simulations either with an identical @xmath89 but a higher magnetic field of @xmath655 , or the same field and a smaller @xmath656 , we are able to determine and verify the use of single rescaling factor @xmath657 relating the real and effective mean - field interladder coupling @xcite , as presented in sec . [
sec : transitiontemperature ] .
109ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop _ _ ,
( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , , ) @noop _ _ ( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.58.14401 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) in @noop _ _ , ( , , ) p. @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , , ) @noop _ _ ( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) , @noop * * , ( ) @noop _ _ ( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.82.054431 [ * * , ( ) ] @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevb.79.060406 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | we investigate weakly coupled spin-@xmath0 ladders in a magnetic field . the work is motivated by recent experiments on the compound @xmath1 ( bpcb ) .
we use a combination of numerical and analytical methods , in particular the density matrix renormalization group ( dmrg ) technique , to explore the phase diagram and the excitation spectra of such a system .
we give detailed results on the temperature dependence of the magnetization and the specific heat , and the magnetic field dependence of the nuclear magnetic resonance ( nmr ) relaxation rate of single ladders .
for coupled ladders , treating the weak interladder coupling within a mean - field or quantum monte carlo approach , we compute the transition temperature of triplet condensation and its corresponding antiferromagnetic order parameter .
existing experimental measurements are discussed and compared to our theoretical results .
furthermore we compute , using time dependent dmrg , the dynamical correlations of a single spin ladder .
our results allow to directly describe the inelastic neutron scattering cross section up to high energies .
we focus on the evolution of the spectra with the magnetic field and compare their behavior for different couplings .
the characteristic features of the spectra are interpreted using different analytical approaches such as the mapping onto a spin chain , a luttinger liquid ( ll ) or onto a t - j model . for values of parameters for which such measurements exist , we compare our results to inelastic neutron scattering experiments on the compound bpcb and find excellent agreement .
we make additional predictions for the high energy part of the spectrum that are potentially testable in future experiments . |
nuclear parton distribution functions ( npdfs ) are usually defined for each parton flavor in terms of nuclear modifications @xmath5 and the corresponding free proton pdf @xmath6 such that @xmath7 where the bjorken @xmath8 , with @xmath9 . in the collider frame , @xmath0 is simply the fraction of the nucleus momentum scaled by the factor @xmath10 . however , since these npdf sets are built `` on top '' of proton pdf sets , the tail @xmath11 is usually ignored .
phenomenological parametrizations for nuclear modifications , such as eks98 @xcite , eps09 @xcite , hkn07 @xcite , dszs @xcite , ncteq @xcite etc .
are largely based on deep inelastic scattering ( dis ) data , which is given as a function of @xmath12 , which is independent of the target mass .
the difference between @xmath13 and @xmath0 thus originates from the nuclear binding ( see refs . @xcite . ] @xmath14 where @xmath15 as an example , the nuclear binding energy @xmath16 mev for pb ( c ) .
in addition to the nuclear binding energy , the fraction of nucleus momentum carried by equivalent photons has to be taken into account in high energy heavy nuclear collisions .
the fraction of the nucleus momentum carried by the photons is found to be @xcite @xmath17 since the gluon npdfs are least constrained by the dis and dy data , presence of the photons in the momentum sum rule mostly affects the overall momentum carried by gluons .
the effect of the equivalent photon field can be taken into account by rescaling the gluons after eq.([xp ] ) has been applied to satisfy @xmath18 to apply the `` conventional '' npdfs given as a function of @xmath13 for the calculation of the nuclear effects in the ultra - relativistic heavy ion collisions where @xmath0 is used , one has to translate to the `` conventional '' npdfs , given as a function of @xmath13 , by taking into account the difference between @xmath13 and @xmath0 which from now on we explicitly call @xmath19 , as follows @xmath20 where the scaling factor for gluons , @xmath21 , is determined via eq.([rescale ] ) .
note that for a free proton , @xmath22 .
it follows from eq .
( 6 ) that while the fraction of the nucleus momentum carried by quarks is invariant with respect to the @xmath23 conversion , the amount of the nucleus momentum carried by gluons , @xmath24 , decreases by the factor of @xmath25 with the rescaling .
in this work we combine a small-@xmath0 theoretical model for the leading twist nuclear shadowing , the fgs model @xcite , with the phenomenological eks98/eps09 npdf set .
the fgs model is based on the generalization of the gribov - glauber multiple scattering formalism and qcd factorization theorems . using the picture of high energy scattering in the laboratory frame and the notion of cross section fluctuations of energetic projectiles ,
multiple interactions are modeled using the effective @xmath0-dependent and flavor - dependent rescattering cross section @xmath26 , which controls the strength of the resulting nuclear shadowing . in @xcite ,
based on the phenomenological analysis of cross section fluctuations in virtual photons , two models were suggested : model 1 ( here referred to as fgs1 ) and model 2 ( fgs2 ) corresponding to the upper and lower bounds on the predicted nuclear shadowing , respectively .
both of the models were built on top of cteq5 pdfs @xcite ( given as a function of @xmath13 ) , and we will use this set for the combined model as well . in this paper , we will work in lo . in the following the initial scale sea quark and gluon ( @xmath27 gev@xmath28 ) nuclear modifications for the region @xmath29 , where data practically do not constrain npdfs ,
are taken from the fgs1 and fgs2 parametrizations ; for @xmath30 , the nuclear modifications are taken from the eks98 parametrization @xcite ( which in this region is very similar to the newer set eps09 @xcite ) .
for the valence quarks , the nuclear modifications are taken from eks98 for the whole @xmath0-range .
the two parametrizations are combined by performing polynomial interpolation between them , and ( after being corrected for the difference in the argument according to eq .
( [ xp ] ) ) the gluons are rescaled as in eq .
( [ xshift ] ) .
the resulting scaling factor at @xmath27 gev@xmath28 for the gluons is @xmath31 ( @xmath32 ) for pb ( c ) nucleus .
( the change in the scaling factor is @xmath33 at @xmath34 gev@xmath28 , so in practice a uniform scaling factor can be used for any scale . )
consequently the amount of the momentum carried by the gluons decreases by 0.88% ( 0.14% ) for pb ( c ) .
the fraction of the momentum carried by the gluons for each model is listed in table 1 ( 41.80% for the proton in cteq5l ) .
[ momtab1e ] cccc & eks98 & fgs1+eks98 & fgs2+eks98 + & pb c & pb c & pb c + @xmath13 : & 43.98 42.61 & 44.20 42.58 &
44.30 42.62 + @xmath19 : & 43.58 42.54 & 43.81 42.52 & 43.91 42.56 + in fig .
[ gluon_ratio_p_play.eps ] we show how changing the definition of @xmath0 affects the eks98 gluon modification ratio @xmath35 at the initial scale @xmath27 gev@xmath28 and how this effect evolves up to the higher scale @xmath36 gev@xmath28 . in this work we use the qcdnum dglap evolution code @xcite . at the initial scale
the original eks98 gluon modification ratio @xmath37 ( solid line ) is first modified setting @xmath38 in eq .
( [ xshift ] ) ( dotted - dashed line ) . as a result ,
the gluon modification ratio is only essentially modified at the emc - region , @xmath39 , where the ( n)pdfs are decreasing rapidly . for the full conversion with @xmath40 ( dotted line ) , the gluon npdf
is naturally scaled down over the whole @xmath13-range . when evolved up to @xmath36 gev@xmath28 , the differences persist and spread towards smaller values of @xmath13 . using @xmath19 instead of
@xmath13 obviously affects gluons for the whole @xmath13-range due to the rescaling ( and at higher scales also the sea quarks via the dglap evolution ) , but for all the parton flavors the most prominent effect sets in at the emc - region , where the parton distribution functions change quickly .
as seen from above , the gluon ratio @xmath41 ( dashed line ) is @xmath42 for the whole @xmath13-range .
note in passing that even for the valence quark distributions , the experimental information on the emc effect in the region where the leading twist contribution dominates is very limited as the higher twist effects give a large ( dominant ? ) contribution to the @xmath43 scattering cross section for @xmath44 in the slac and jlab kinematics .
hence , to date , practically no data on the quark modification for @xmath44 in the scaling region is available for heavy nuclei such as , e.g. , lead . in fig .
[ gluon_ratio_d.eps ] we show the gluon ratio ( @xmath45 ) for the combined models fgs1+eks98 ( dotted - dashed ) and fgs2+eks98 ( dotted ) , together with the eks98 ( solid ) gluon modification . at the initial scale above @xmath46 ,
the difference between the models originates only from the different definition of @xmath0 ; below @xmath47 , the two fgs nuclear shadowing models span a region considerably smaller than the error band of eps09 gluons ( see @xcite for details ) .
when evolved up to @xmath34 gev@xmath28 , a relevant scale in the lhc kinematics , it is evident that processes which are sensitive to the emc - region are also sensitive to the proper definition of @xmath0 . at @xmath27 gev@xmath28 and for @xmath35 at @xmath27 gev@xmath28 and @xmath36 gev@xmath28 for eks98 .
, width=359 ] at @xmath27 , 100@xmath28 and 10000@xmath28 gev@xmath28 for eks98 , and the combination model fgs1(2)+eks98.,width=359 ]
in order to understand the sensitivity of the lhc kinematics to the emc effect , we study inclusive @xmath48 production in @xmath49 collisions , which schematically can be expressed as @xmath50 where the factorization and renormalization scales have been set equal ( see e.g. @xcite for the formulae and details ) . in this work ,
we choose @xmath51 ( the outgoing pion transverse momentum ) and @xmath52 ( partonic transverse momentum ) . in fig .
[ pb_eta_0.0_xbin.eps ] we show the lo invariant cross section @xmath53 for @xmath54 at @xmath55 and 100.0 gev as a function of @xmath56 ( the momentum fraction carried by the parton in pb , without conversion given in eq .
( [ xshift ] ) ) .
the results have been computed with the eps09 nuclear modifications @xcite , cteq6l pdfs @xcite and dss fragmentation functions @xcite .
working in lo the overall normalization of the spectra is not fixed , but the @xmath56-distribution and the relative normalization are not affected by this . the upper panel shows the mid - rapidity and the lower panel the forward rapidity ( at the lhc , the pb rapidity is positive ) @xmath48 production . for each @xmath57 value
studied , the mid - rapidity production peaks at an order of magnitude smaller values of @xmath56 than the forward rapidity results , and remains significant over a wider range of @xmath56 . in the forward direction
the pion production is concentrated on a rather narrow @xmath56-range , making it a more sensitive probe of nuclear effects .
in particular , at high-@xmath57 , the pions are produced exclusively from the emc - region , making them also sensitive to the definition of @xmath56 .
figures [ ppb_per_pp_mb_res.eps ] and [ ppb_per_pp_mb_res_eta_3.5.eps ] show the fgs1(2)+eks98 results for the minimum bias nuclear modification ratio , @xmath58 at the lhc at @xmath59 gev . for comparison ,
the eks98 results with the cteq5l pdfs are also shown . in the mid - rapidity ( fig .
[ ppb_per_pp_mb_res.eps ] ) the difference between the eks98 and fgs+eks98 results remain within a few-% . as can be seen from fig .
[ pb_eta_0.0_xbin.eps ] ( upper panel ) , up to @xmath60 gev , the pion spectra mostly originate from the gluon distribution dominated @xmath0-range below the emc - region . therefore the differences between the models are caused both by the different assumptions about shadowing ( see the lower panel where the small-@xmath57 part is shown on a logarithmic scale ) and the scaling of the gluon distribution with @xmath61 . of @xmath0 .
the difference between the models remain moderate even above @xmath62 gev . as seen from fig .
[ ppb_per_pp_mb_res_eta_3.5.eps ] , the situation is drastically different in the forward rapidity pion production . above @xmath63 gev ,
the two fgs - models are indistinguishable , but start to deviate from the eks98 result as @xmath57 increases .
this sizable effect is caused by the correction to the @xmath0-definition alone .
until now we have discussed the minimum bias results , where the impact parameter dependence of the nuclear effects has been spatially averaged . in the fgs model the transverse position @xmath64 dependence
is naturally built in as functions of @xmath65 since the nuclear shadowing is first calculated for fixed @xmath64 and next the integral over @xmath64 is taken . in the eks98s / eps09s model @xcite ,
the eks98 and eps09 npdf parametrizations were also assumed to have spatial dependence as a power series of @xmath65 . with the centrality classes modeled using the optical glauber model ,
the eps09s results were found to be consistent with the mid - rapidity phenix @xmath66 centrality systematics @xcite .
however , as already seen in fig .
[ ppb_per_pp_mb_res_eta_3.5.eps ] , the proper definition of @xmath0 has a major effect on the lhc predictions at forward rapidity . in fig .
[ ppb_per_pp.eps ] we applied the procedure described in @xcite to @xmath67 at @xmath68 for a selection of different centrality classes , and compared the eks98s results ( with cteq6l pdfs ) with and without the @xmath0-corrections .
irrespective of the centrality , the correction causes a clear , measurable effect . for an impact parameter dependent theoretically motivated npdf model it is also important to pay special attention to the emc region for another reason . in @xcite
the magnitude of the emc effect was shown to be linearly related to the short range correlations ( src ) scale factor measured from electron inclusive scattering at @xmath69 .
consequently the impact parameter dependence of the emc effect should be proportional to the local density @xcite and the emc effect thus should be strongest in the center of the nucleus .
we will address this issue in @xcite , where a full nlo impact parameter dependent npdf set will be released .
s. s. adler _ et al .
_ [ phenix collaboration ] , phys .
lett . * 98 * , 172302 ( 2007 ) [ nucl - ex/0610036 ]
. i. helenius , k. j. eskola , h. honkanen and c. a. salgado , jhep * 1207 * , 073 ( 2012 ) [ arxiv:1205.5359 [ hep - ph ] ] .
l. b. weinstein , e. piasetzky , d. w. higinbotham , j. gomez , o. hen and r. shneor , phys .
lett . * 106 * , 052301 ( 2011 ) [ arxiv:1009.5666 [ hep - ph ] ] . l. l. frankfurt and m. i. strikman , phys .
* 160 * , 235 ( 1988 ) . for the lhc at @xmath59 gev and @xmath71 . for comparison
, the eks98 grid result with cteq5l pdfs is also shown .
upper panel on linear scale , lower panel on logarithmic scale.,title="fig:",width=453 ] for the lhc at @xmath59 gev and @xmath71 . for comparison ,
the eks98 grid result with cteq5l pdfs is also shown .
upper panel on linear scale , lower panel on logarithmic scale.,title="fig:",width=453 ] for the lhc at @xmath59 gev and @xmath68 for the eks98 ( with cteq6l1 ) results computed with @xmath13 and @xmath19 .
the different centrality classed are computed as in @xcite .
the fgs1+eks98 and fgs2+eks98 curves are indistinguishable.,width=604 ] | the presence of nuclear medium and collective phenomena which involve several nucleons modify the parton distribution functions of nuclei ( npdfs ) compared to those of a free nucleon .
these modifications have been investigated by different groups using global analyses of high energy nuclear reaction world data resulting in modern npdf parametrizations with error estimates , such as eps09(s ) , hkn07 and nds .
these phenomenological npdf sets roughly agree within their uncertainty bands , but have antiquarks for large-@xmath0 and gluons for the whole @xmath0-range poorly constrained by the available data . in the kinematics accessible at the lhc
this has negative impact on the interpretation of the heavy - ion collision data , especially for the @xmath1 benchmarking runs .
the emc region is also sensitive to the proper definition of @xmath0 , where the nuclear binding effects have to be taken into account , and for heavy nuclei one also needs to take into account that a fraction of the nucleus momentum is carried by the equivalent photons which modifies the momentum sum rule .
we study how these effects affect the predictions for the nuclear modification ratios at the lhc kinematics using a model where we combine theoretical input for the leading twist nuclear shadowing ( the fgs model ) and the eks98s / eps09s npdf set where the spatial dependence is formulated as a power series of the nuclear thickness functions @xmath2 . +
* keywords : * _ nuclear parton distribution function ; lhc ; impact parameter ; emc region _ * * modeling nuclear parton distribution functions + @xmath3__the pennsylvania state university , 104 davey lab , university park , pa 16802 , usa _ _
+ @xmath4__national research center
kurchatov institute , petersburg nuclear physics institute ( pnpi ) , gatchina , 188300 , russia _ _ |
as the first force recognized among the four fundamental interactions , gravity is mysterious and fascinating because of its unique feature .
gravity is universal , which is well known from the epoch of newton s theory .
nothing could escape the control of gravity , from the falling of apple towards the earth , to the motion of moon in the sky . in einstein s general relativity , gravity is related to the geometry of curved spacetime .
this beautiful and elegant idea significantly influences our world view .
spacetime is curved by matter / energy , while the motion of object is along the `` straight '' line ( geodesic ) of the curved spacetime .
general relativity has passed all experimental tests up to now . however , there is intrinsic conflict between quantum theory and general relativity .
lots of efforts have been made to quantize gravity , but no success has been achieved yet .
gravity is extremely weak compared to the other fundamental forces , so it is usually ignored in micro - physics .
nonetheless , on the scale of universe , things are mostly controlled by gravity because it is long - range and has no screening effect .
one century has passed since einstein established general relativity , but only a few solutions to the field equation have been found , among which three solutions are most famous and useful . the most simple case is for static and spherical spacetime , and the solution was derived by schwarzschild just one month after einstein s field equation .
the schwarzschild solution indicates also the existence of black hole , where everything is doomed to fall towards the center after passing through the event horizon . consider a non - vancum case with ideal fluid as source
, the field equation could be transformed to tolman - oppenheimer - vokoff equation@xcite , which could be applied to the interior of pulsar - like compact stars .
based on the so - called cosmological copernicus principle , friedmann equation can be derived with the assumption of homogenous and isotropic universe , which sets the foundation of cosmology .
these three solutions of einstein s field equation represent the most frontier topics in modern astrophysics . at the late stage of stellar evolution ,
how does the core of massive star collapse to a black hole ? or equivalently , how is normal baryonic matter squeezed into the singularity ? what s the state of compressed baryonic matter ( cbm ) before collapsing into a black hole ?
we are focusing on these questions in this chapter . in the standard model of particle physics , there are totally six flavours of quarks . among them , three ( @xmath7 , @xmath8 and @xmath10 ) are light , with masses @xmath19 mev , while other three flavours ( @xmath20 , @xmath21 and @xmath22 ) , with mass @xmath23 mev , are too heavy to be excited in the nuclear energy scale , @xmath24 mev .
however , the ordinary matter in our world is built from @xmath7 and @xmath8 quarks only , and the numbers of these two flavours tend to be balanced in a stable nucleus .
it is then interesting to think philosophically about the fact that our baryonic matter is 2-flavour symmetric .
an explanation could be : micro - nuclei are too small to have 3-flavour symmetry , but bigger is different .
in fact , rational thinking about stable strangeness dates back to 1970s@xcite , in which bodmer speculated that so - called `` collapsed nuclei '' with strangeness could be energetically favored if baryon number @xmath25 , but without quantitative estimation of the minimal number @xmath26 .
bulk matter composed of almost _ free _ quarks ( @xmath7 , @xmath8 , and @xmath10 ) was then focused on@xcite , even for astrophysical manifestations@xcite .
cbm can manifest as a pulsar if the mass is not large enough to form a black hole , and strange quark matter could possibly exist in compact stars , either in the core of neutron star ( i.e. , mixed or hybrid stars@xcite ) or as the whole star ( strange quark star@xcite ) . although the asymptotic freedom is well recognized ,
one essential point is whether the color coupling between quarks is still perturbative in astrophysical cbm so that quarks are itinerant there . in case of non - perturbative coupling ,
the strong force there might render quarks grouped in so - called _ quark - clusters _ , forming a nucleus - like strange object@xcite with 3-flavour symmetry , when cbm is big enough that relativistic electrons are inside ( i.e. , @xmath27 ) .
anyway , we could simply call 3-flavour baryonic matter as _ strange matter _ , in which the constituent quarks could be either itinerant or localized .
why is big cbm strange ?
this is actually a conjecture to be extensively discussed in the next section , but it could be reasonable .
besides being meaningful for understanding the nature of sub - nucleon at a deeper level , strangeness would also have a consequence of the physics of super - dense matter .
the discovery of strangeness ( a general introduction to murray gell - mann and his strangeness could be found in the biography by johnson@xcite ) is known as a milestone in particle physics since our normal baryons are non - strange .
nonetheless , condensed matter with strangeness should be worth exploring as the energy scale @xmath28 at the nuclear and even supra - nuclear densities .
previously , bulk strange object ( suggested to be 3-flavour _ quark _ matter ) is speculated to be the absolutely stable ground state of strong - interaction matter , which is known as the bodmer - witten s conjecture.@xcite but we are discussing a general conjecture in next subsections , arguing that quarks might not be necessarily free in stable strange matter , and would still be hadron - like localized as in a nucleus if non - perturbative qcd effects are significant ( i.e. , @xmath29 ) and the repulsive core keeps to work in both cases of 2-flavour ( non - strange ) nuclear matter and 3-flavour ( strange ) matter . in this sense ,
protons and neutrons are of 2-flavour quark clusters , while strange matter could be condensed matter of quark clusters with strangeness ( i.e. , strange quark - clusters ) .
summarily , it is well known that micro - nuclei are non - strange , but macro - nuclei in the form of cbm could be strange .
therefore , astrophysical cbm and nucleus could be very similar , but only with a simple change from non - strange to strange : `` 2 '' @xmath30 `` 3 '' .
we are explaining two approaches to this strange quark - cluster matter state , bottom - up and top - down , respectively as following .
micro - nucleus is made up of protons and neutrons , and there is an observed tendency to have equal numbers of protons ( @xmath31 ) and neutrons ( @xmath32 ) . in liquid - drop model ,
the mass formula of a nucleus with atomic number @xmath33 consists of five terms , @xmath34 where the third term is for the symmetry energy , which vanishes with equal number of protons and neutrons .
this nuclear symmetry energy represents a symmetry between proton and neutron in the nucleon degree of freedom , and is actually that of @xmath7 and @xmath8 quarks in the quark degree@xcite .
the underlying physics of symmetry energy is not well understood yet . if the nucleons are treated as fermi gas , there is a term with the same form as symmetry energy in the formula of fermi energy , known as the kinetic term of nuclear symmetry energy .
but the interaction is not negligible , and the potential term of symmetry energy would be significant .
recent scattering experiments show that , because of short - range interactions , the neutron - proton pairs are nearly 20 times as prevalent as proton - proton ( and neutron - neutron by inference ) pairs,@xcite which hints that the potential term would dominate in the symmetry energy .
since the electric charges of @xmath7 and @xmath8 quarks is + 2/3 are -1/3 respectively , 2-flavour symmetric strong - interaction matter should be positively charged , and electrons are needed to keep electric neutrality .
the possibility of electrons inside a nucleus is negligible because the nuclear radius is much smaller than the compton wavelength @xmath35 , and the lepton degree of freedom would then be not significant for nucleus .
therefore , electrons contribute negligible energy for micro - nuclei , as the coupling constant of electromagnetic interaction ( @xmath36 ) is much less than that of strong interaction ( @xmath37 ) .
the kinematic motion of electrons is bound by electromagnetic interaction , so @xmath38 . from heisenberg s relation , @xmath39 , combining the above two equations , we have @xmath40 , and the interaction energy is order of @xmath41 .
however , bigger is different , and there might be 3-flavour symmetry in gigantic / macro - nuclei , as electrons are inside a gigantic nucleus . with the number of nucleons @xmath42 , the scale of macro - nuclei should be larger than the compton wavelength of electrons @xmath35 .
if 2-flavour symmetry keeps , macro - nucleus will become a huge thomson atom with electrons evenly distributed there .
though coulomb energy could not be significant , the fermi energy of electrons are not negligible , to be @xmath43
. however , the situation becomes different if strangeness is included : no electrons exist if the matter is composed by equal numbers of light quarks of @xmath7 , @xmath8 , and @xmath10 in chemical equilibrium . in this case , the 3-flavor symmetry , an analogy of the symmetry of @xmath7 and @xmath8 in nucleus , may results in a ground state of matter for gigantic nuclei .
certainly the mass difference between @xmath7/@xmath8 and @xmath10 quarks would also break the 3-flavour symmetry , but the interaction between quarks could lower the effect of mass differences and favor the restoration of 3-flavour symmetry . if macro - nuclei are almost 3-flavour symmetric , the contribution of electrons would be negligible , with @xmath44 and @xmath45 .
the new degree of freedom ( strangeness ) is also possible to be excited , according to an order - of - magnitude estimation from either heisenberg s relation ( localized quarks ) or fermi energy ( free quarks ) . for quarks localized with length scale @xmath46 , from heisenberg s uncertainty relation ,
the kinetic energy would be of @xmath47 , which has to be comparable to the color interaction energy of @xmath48 in order to have a bound state , with an assumption of coulomb - like strong interaction .
one can then have if quarks are dressed , @xmath49 as @xmath37 may well be close to and even greater than 1 at several times the nuclear density , the energy scale would be approaching and even larger than @xmath50400 mev .
a further calculation of fermi energy also gives , @xmath51 if quarks are considered moving non - relativistically , or @xmath52 if quarks are considered moving extremely relativistically .
so we have the energy scale @xmath53 by either heisenberg s relation or fermi energy , which could certainly be larger the mass difference ( @xmath54 ) between @xmath10 and @xmath55 quarks , however , for micro - nuclei where electron contributes also negligible energy , there could be 2-flavour ( rather than 3-flavour ) symmetry because @xmath10 quark mass is larger than @xmath55 quark masses .
we now understand that it is more economical to have 2-flavour micro - nuclei because of massive @xmath10-quark and negligible electron kinematic energy , whereas macro / gigantic - nuclei might be 3-flavour symmetric .
the 2-flavour micro - nucleus consists of @xmath7 and @xmath8 quarks grouped in nucleons , while the 3-flavour macro - nucleus is made up of @xmath7 , @xmath8 and @xmath10 quarks grouped in so - called strange quark - clusters .
such macro - nucleus with 3-flavour symmetry can be named as strange quark - cluster matter , or simply strange matter . besides this bottom - up scenario ( an approach from the hadronic state )
, we could also start from deconfined quark state with the inclusion of stronger and stronger interaction between quarks ( a top - down scenario ) .
the underlying theory of the elementary strong interaction is believed to be quantum chromodynamics ( qcd ) , a non - abelian _
su_(3 ) gauge theory . in qcd ,
the effective coupling between quarks decreases with energy ( the _ asymptotic freedom_)@xcite .
quark matter ( or quark - gluon plasma ) , the soup of deconfined quarks and gluons , is a direct consequence of asymptotic freedom when temperature or baryon density are extremely enough . hot quark matter could be reproduced in the experiments of relativistic heavy ion collisions .
ultra - high chemical potential is required to create cold quark matter , and it can only exist in rare astrophysical conditions , the compact stars . what kind of cold matter can we expect from qcd theory , in effective models , or even based on phenomenology ? this is a question too hard to answer because of ( i ) the non - perturbative effects of strong interaction between quarks at low energy scales and ( ii ) the many - body problem due to vast assemblies of interacting particles . a color - superconductivity ( csc ) state
is focused on in qcd - based models , as well as in phenomenological ones@xcite .
the ground state of extremely dense quark matter could certainly be that of an ideal fermi gas at an extremely high density .
nevertheless , it has been found that the highly degenerate fermi surface could be unstable against the formation of quark cooper pairs , which condense near the fermi surface due to the existence of color - attractive channels between the quarks .
a bcs - like color superconductivity , similar to electric superconductivity , has been formulated within perturbative qcd at ultra - high baryon densities .
it has been argued , based on qcd - like effective models , that color superconductivity could also occur even at the more realistic baryon densities of pulsar - like compact stars@xcite .
can the realistic stellar densities be high enough to justify the use of perturbative qcd ?
it is surely a challenge to calculate the coupling constant , @xmath56 , from first principles . nevertheless , there are some approaches to the non - perturbative effects of qcd , one of which uses the dyson - schwinger equations tried by fischer et al.,@xcite who formulated , @xmath57 where @xmath58 , @xmath59 , @xmath60 , @xmath61 , @xmath62 with @xmath63 the typical momentum in gev , and that @xmath56 freezes at @xmath64 . for our case of
assumed dense quark matter at @xmath65 , the chemical potential is @xmath66 gev , and then @xmath67 gev@xmath68 .
thus , it appears that the coupling in realistic dense quark matter should be greater than 2 , being close to 3 in the fischer s estimate presented in eq.([dse ] ) .
therefore , this surely means that a weakly coupling treatment could be dangerous for realistic cold quark matter ( the interaction energy @xmath69 mev could even be much larger than the fermi energy ) , i.e. , the non - perturbative effect in qcd should not be negligible if we try to know the real state of compact stars .
it is also worth noting that the dimensionless electromagnetic coupling constant ( i.e. , the fine - structure constant ) is @xmath70 , which makes qed tractable .
that is to say , a weakly coupling strength comparable with that of qed is possible in qcd only if the density is unbelievably and unrealistically high ( @xmath71 ! with @xmath72 @xmath73 the baryon density of nuclear matter ) .
quark - clusters may form in relatively low temperature quark matter due to the strong interaction ( i.e. , large @xmath56 ) , and the clusters could locate in periodic lattices ( normal solid ) when temperature becomes sufficiently low .
although it is hitherto impossible to know if quark - clusters could form in cold quark matter via calculation from first principles , there could be a few points that favor clustering .
_ experimentally _ , though quark matter is argued to be weakly coupled at high energy and thus deconfined , it is worth noting that , as revealed by the recent achievements in relativistic heavy ion collision experiments , the interaction between quarks in a fireball of quarks and gluons is still very strong ( i.e. the strongly coupled quark - gluon plasma , sqgp@xcite ) .
the strong coupling between quarks may naturally render quarks grouped in clusters , i.e. , a condensation in position space rather than in momentum space .
_ theoretically _ , the baryon - like particles in quarkyonic matter@xcite might be grouped further due to residual color interaction if the baryon density is not extremely high , and quark - clusters would form then at only a few nuclear density .
certainly , more elaborate research work is necessary .
for cold quark matter at @xmath74 density , the distance between quarks is @xmath50 fm @xmath11 the planck scale @xmath75 , so quarks and electrons can well be approximated as point - like particles .
if q@xmath76-like clusters are created in the quark matter@xcite , the distance between clusters are @xmath77 .
the length scale @xmath46 and color interaction energy of quark - clusters has been estimated by the uncertainty relation , assuming quarks are as dressed ( the constituent quark mass is @xmath78 mev ) and move non - relativistically in a cluster .
we have @xmath79 fm if @xmath80 , and the color interaction energy @xmath81 could be greater than the baryon chemical potential if @xmath82 .
the strong coupling could render quarks grouped in position space to form clusters , forming a nucleus - like strange object@xcite with 3-flavour symmetry , if it is big enough that relativistic electrons are inside ( i.e. , @xmath27 ) .
quark - clusters could be considered as classical particles in cold quark - cluster matter and would be in lattices at a lower temperature . in conclusion
, quark - clusters could emerge in cold dense matter because of the strong coupling between quarks .
the quark - clustering phase has high density and the strong interaction is still dominant , so it is different from the usual hadron phase , and on the other hand , the quark - clustering phase is also different from the conventional quark matter phase which is composed of relativistic and weakly interacting quarks . the quark - clustering phase could be considered as an intermediate state between hadron phase and free - quark phase , with deconfined quarks grouped into quark - clusters . in summary , there could be some similarities and differences between micro - nuclei and macro - nuclei , listed as following .
_ similarity _ 1 : both micro - nuclei and macro - nuclei are self - bound by the strong color interaction , in which quarks are localized in groups called generally as quark - clusters .
we are sure that there are two kinds of quark - clusters inside micro - nuclei , the proton ( with structure @xmath83 ) and neutron ( @xmath84 ) , but do nt know well the clusters in macro - nuclei due to the lack of detailed experiments related .
_ similarity _ 2 : since the strong interaction might not be very sensitive to flavour , the interaction between general quark - clusters should be similar to that of nucleon , which is found to be lennard - jones - like by both experiment and modeling . especially , one could then expect a hard core@xcite ( or repulsive core ) of the interaction potential between strange quark - clusters though no direct experiment now hints this existence . _ difference _ 1 : the most crucial difference is the change of flavour degree of freedom , from two ( @xmath7 and @xmath8 ) in micro - nuclei to three ( @xmath7 , @xmath8 and @xmath10 ) in macro / gigantic - nuclei .
we could thus have following different aspects derived .
_ difference _ 2 : the number of quarks in a quark - cluster is 3 for micro nuclei , but could be 6 , 9 , 12 , and even 18 for macro nuclei , since the interaction between @xmath85-particles could be attractive@xcite so that no positive pressure can support a gravitational star of @xmath85-cluster matter .
we therefore call proton / neutron as light quark - clusters , while the strange quark - cluster as heavy clusters because of ( 1 ) massive @xmath10-quark and ( 2 ) large number of quarks inside
. _ difference _ 3 : a micro nucleus could be considered as a quantum system so that one could apply quantum mean - field theory , whereas the heavy clusters in strange matter may be classical particles since the quantum wavelength of massive clusters might be even smaller than the mean distance between them .
_ difference _ 4 : the equation of state ( eos ) of strange matter would be stiffer@xcite than that of nuclear matter because the clusters in former should be non - relativistic but relativistic in latter .
the kinematic energy of a cluster in both micro- and macro - nuclei could be @xmath86 , which is much smaller than the rest mass ( generally @xmath87 ) of a strange quark - cluster .
_ difference _ 5 : condensed matter of strange quark - clusters could be in a solid state at low temperature much smaller than the interaction energy between clusters .
we could then expect solid pulsars@xcite in nature although an idea of solid nucleus was also addressed@xcite a long time ago .
the 3-flavour symmetry may hint the nature of strong interaction at low - energy scale .
let s tell a story of science fiction about flavour symmetry .
our protagonist is a fairy who is an expert in qcd at high energy scale ( i.e. , perturbative qcd ) but knows little about spectacular non - perturbative effects .
there is a conversation between the fairy and god about strong - interaction matter .
god : `` i know six flavours of quarks , but how many flavours could there exist in stable strong - interaction matter ? '' fairy : `` it depends ... how dense is the matter ?
(: the nuclear saturation density arises from the short - distance repulsive core , a consequence of non - perturbative qcd effect she may not know much . ) '' god : `` hum ... i am told that quark number density is about 0.48 @xmath73(@xmath74 ) there . ''
fairy : `` ah , in this energy scale of @xmath88 gev , there could only be light flavours ( i.e. , @xmath7 , @xmath8 and @xmath10 ) in a stable matter if quarks are free . ''
god : `` two flavours ( @xmath7 and @xmath8 ) or three flavours ? '' fairy : `` there could be two flavours of free quarks if strong - interaction matter is very small ( @xmath16 ) , but would be three flavours for bulk strong - interaction matter (: the bodmer - witten s conjecture ) . ''
god : `` small 2-flavour strong - interaction matter is very useful , and i can make life and mankind with huge numbers of the pieces . we can call them atoms . '' fairy : `` thanks , god !
i can also help mankind to have better life . ''
god : `` but ... are quarks really free there ? '' fairy : `` hum ... there could be clustered quarks in both two and three flavour (: a bodmer - witten s conjecture generalized ) cases if the interaction between quarks is so strong that quarks are grouped together .
you name a piece of small two flavour matter _ atom _
, what would we call a 3-flavour body in bulk ?
'' god : oh ... simply , a _
strange _ object because of strangeness .
cold strange matter with 3-flavour symmetry could be in a solid state because of ( 1 ) a relatively small quantum wave packet of quark - cluster ( wave length @xmath89 , where @xmath90 is the separation between quark - clusters ) and ( 2 ) low temperature @xmath91 ( @xmath92 is the interaction energy between quark - clusters )
. the packet scales @xmath93 for free quark - cluster , with @xmath94 the rest mass of a quark cluster , but could be much smaller if it is constrained in a potential with depth of @xmath95 mev ( 2 @xmath96 ) . a star made of strange matter would then be a solid star .
it is very fundamental to study static and spherically symmetric gravitational sources in general relativity , especially for the interior solutions .
the tov solution@xcite is only for perfect fluid .
however , for solid strange stars , since the local press could be _ anisotropic _ in elastic matter , the radial pressure gradient could be partially balanced by the tangential shear force although a general understanding of relativistic , elastic bodies has unfortunately not been achieved@xcite
. the origin of this local anisotropic force in solid quark stars could be from the development of elastic energy as a star ( i ) spins down ( its ellipticity decreases ) and ( ii ) cools ( it may shrink ) .
release of the elastic as well as the gravitational energies would be not negligible , and may have significant astrophysical implications .
the structure of solid quark stars can be numerically calculated as following . for the sack of simplicity , only spherically symmetric sources are dealt with , in order to make sense of possible astrophysical consequence of solid quark stars . by introducing respectively radial and tangential pressures , @xmath97 and @xmath98 , the stellar equilibrium equation of static anisotropic matter in newtonian gravity is@xcite : @xmath99 , where @xmath100 and @xmath101 denote mass density and the gravitational constant , respectively , and @xmath102 . however , in einstein s gravity , this equilibrium equation is modified to be@xcite , @xmath103 where @xmath104 is introduced . in case of isotropic pressure , @xmath105 , eq .
( [ p ] ) turns out to be the tov equation .
it is evident from eq .
( [ p ] ) that the radial pressure gradient , @xmath106 , decreases if @xmath107 , which may result in a higher maximum mass of compact stars .
one can also see that a sudden decrease of @xmath98 ( equivalently of elastic force ) in a star may cause substantial energy release , since the star s radius decreases and the absolute gravitational energy increases .
starquakes may result in a sudden change of @xmath108 , with an energy release of the gravitational energy as well as the tangential strain energy .
generally , it is evident that the differences of radius , gravitational energy , and moment of inertia increase proportionally to stellar mass and the parameter @xmath108 .
this means that an event should be more important for a bigger change of @xmath108 in a quark star with higher mass .
typical energy of @xmath109 is released during superflares of sgrs , and a giant starquake with @xmath110 could produce such a flare@xcite .
a sudden change of @xmath108 can also result in a jump of spin frequency , @xmath111 .
glitches with @xmath112 could occur for parameters of @xmath113 and @xmath114 .
it is suggestive that a giantflare may accompany a high - amplitude glitch .
how to create macro - nuclei ( even gigantic ) in the universe ? besides a collapse event where normal baryonic matter is intensely compressed by gravity , strange matter could also be produced after cosmic hadronization.@xcite strange matter may manifest itself as a variety of objects with a broad mass spectrum , including compact objects , cosmic rays and even dark matter . in 1932 , soon after chandrasekhar found a unique mass ( the mass limit of white dwarfs ) , landau speculated a state of matter , the density of which `` becomes so great that atomic nuclei come in close contact , forming one _ gigantic nucleus_''@xcite .
a star composed mostly of such matter is called a `` neutron '' star , and baade and zwicky even suggested in 1934 that neutron stars ( nss ) could be born after supernovae .
nss theoretically predicted were finally _ discovered _ when hewish and his collaborators detected radio pulsars in 1967@xcite .
more kinds of pulsar - like stars , such as x - ray pulsars and x - ray bursts in binary systems , were also discovered later , and all of them are suggested to be nss . in gigantic nucleus , protons and electrons combined to form the neutronic state , which involves weak equilibrium between protons and neutrons . however , the simple and beautiful idea proposed by landau and others had one flaw at least : nucleons ( neutrons and protons ) are in fact _ not _ structureless point - like particles although they were thought to be elementary particles in 1930s .
a success in the classification of hadrons discovered in cosmic rays and in accelerators leaded gell - mann to coin `` _ quark _ '' with fraction charges ( @xmath115 ) in mathematical description , rather than in reality@xcite .
all the six flavors of quarks ( @xmath116 ) have experimental evidence ( the evidence for the last one , top quark , was reported in 1995 ) .
is weak equilibrium among @xmath7 , @xmath8 and @xmath10 quarks possible , instead of simply that between @xmath7 and @xmath8 quarks ? at the late stage of stellar evolution , normal baryonic matter is intensely compressed by gravity in the core of massive star during supernova .
the fermi energy of electrons are significant in cbm , and it is very essential to cancel the electrons by weak interaction in order to make lower energy state .
there are two ways to kill electrons as shown in fig [ f1 ] : one is via _ neutronization _
, @xmath117 , where the fundamental degrees of freedom could be nucleons ; the other is through _ strangenization _ , where the degrees of freedom are quarks . while neutronization works for removing electrons , strangenization has both the advantages of minimizing the electron s contribution of kinetic energy and maximizing the flavour number , the later could be related to the flavour symmetry of strong - interaction matter .
these two ways to kill electrons are relevant to the nature of pulsar , to be neutron star or strange star , as summarized in fig .
[ f1 ] .
there are many speculations about the nature of pulsar due to unknown non - perturbative qcd at low energy . among different pulsar models , hadron star and hybrid / mixed star
are conventional neutron stars , while quark star and quark - cluster star are strange stars with light flavour symmetry . in hadron star model , quarks are confined in hadrons such as neutron / proton and hyperon , while a quark star is dominated by de - confined free quarks .
a hybrid / mixed star , with quark matter in its cores , is a mixture of hadronic and quark states .
however , a quark - cluster star , in which strong coupling causes individual quarks grouped in clusters , is neither a hadron star nor a quark star . as an analog of neutrons
, quark - clusters are bound states of several quarks , so to this point of view a quark - cluster star is more similar to a _
real _ giant nucleus of self - bound ( not that of landau ) , rather than a `` giant hadron '' which describes traditional quark stars . different models of pulsar s inner structure are illustrated in fig .
[ f2 ] . , width=377 ] it is shown in fig .
[ f2 ] that conventional neutron stars ( hadron star and hybrid / mixed star ) are gravity - bound , while strange stars ( strange quark star and strange quark - cluster star ) are self - bound on surface by strong force .
this feature difference is very useful to identify observationally . in the neutron star picture ,
the inner and outer cores and the crust keep chemical equilibrium at each boundary , so neutron star is bound by gravity .
the core should have a boundary and is in equilibrium with the ordinary matter because the star has a surface composed of ordinary matter .
there is , however , no clear observational evidence for a neutron star s surface , although most of authors still take it for granted that there should be ordinary matter on surface , and consequently a neutron star has different components from inner to outer parts .
being similar to traditional quark stars , quark - cluster stars have almost the same composition from the center to the surface , and the quark matter surface could be natural for understanding some different observations .
it is also worth noting that , although composed of quark - clusters , quark - cluster stars are self - bound by the residual interaction between quark - clusters .
this is different from but similar to the traditional mit bag scenario .
the interaction between quark - clusters could be strong enough to make condensed matter , and on the surface , the quark - clusters are just in the potential well of the interaction , leading to non - vanishing density but vanishing pressure .
observations of pulsar - like compact stars , including surface and global properties , could provide hints for the state of cbm , as discussed in the following .
_ dirfting subpulses . _
although pulsar - like stars have many different manifestations , they are populated by radio pulsars . among the magnetospheric emission models for pulsar radio radiative process , the user - friendly nature of ruderman - sutherland@xcite model is a virtue not shared by others , and clear drifting sub - pulses were first explained . in the seminal paper ,
a vacuum gap was suggested above the polar cap of a pulsar .
the sparks produced by the inner - gap breakdown result in the subpulses , and the observed drifting feature is caused by @xmath118 . however , that model can only work in strict conditions for conventional neutron stars : strong magnetic field and low temperature on surfaces of pulsars with @xmath119 , while calculations showed , unfortunately , that these conditions usually can not be satisfied there .
the above model encounters the so - called `` binding energy problem '' .
calculations have shown that the binding energy of fe at the neutron star surface is @xmath120 kev@xcite , which is not sufficient to reproduce the vacuum gap .
these problems might be alleviated within a partially screened inner gap model@xcite for nss with @xmath121 , but could be naturally solved for any @xmath122 in the bare strange ( quark - cluster ) star scenario .
the magnetospheric activity of bare quark - cluster star was investigated in quantitative details@xcite . since quarks on the surface
are confined by strong color interaction , the binding energy of quarks can be even considered as infinity compared to electromagnetic interaction . as for electrons on the surface
, on one hand the potential barrier of the vacuum gap prevents electrons from streaming into the magnetosphere , on the other hand the total energy of electrons on the fermi surface is none - zero .
therefore , the binding energy of electrons is determined by the difference between the height of the potential barrier in the vacuum gap and the total energy of electrons .
calculations have shown that the huge potential barrier built by the electric field in the vacuum gap above the polar cap can usually prevent electrons from streaming into the magnetosphere , unless the electric potential of a pulsar is sufficiently lower than that at the infinite interstellar medium . in the bare quark - cluster star model , both positively and negatively charged particles on the surface
are usually bound strongly enough to form a vacuum gap above its polar cap , and the drifting ( even bi - drifting ) subpulses can be understood naturally@xcite . _ x - ray spectral lines . _ in conventional neutron star (
ns)/crusted strange star models , an atmosphere exists above the surface of a central star .
many theoretical calculations , first developed by romani@xcite , predicted the existence of atomic features in the thermal x - ray emission of ns ( also for crusted strange star ) atmospheres , and advanced facilities of _ chandra _ and _ xmm - newton _ were then proposed to be constructed for detecting those lines .
one expects to know the chemical composition and magnetic field of the atmosphere through such observations , and eventually to constrain stellar mass and radius according to the redshift and pressure broadening of spectral lines .
however , unfortunately , none of the expected spectral features has been detected with certainty up to now , and this negative test may hint a fundamental weakness of the ns models . although conventional ns models can not be completely ruled out by only non - atomic thermal spectra since modified ns atmospheric models with very strong surface magnetic fields@xcite might reproduce a featureless spectrum too , a natural suggestion to understand the general observation could be that pulsars are actually bare strange ( quark or quark - cluster ) star@xcite , almost without atoms there on the surfaces .
more observations , however , did show absorption lines of psr - like stars , and the best absorption features were detected for the central compact object ( cco ) 1e 1207.4 - 5209 in the center of supernova remnant pks 1209 - 51/52 , at @xmath123 kev and @xmath124 kev@xcite .
although initially these features were thought to be due to atomic transitions of ionized helium in an atmosphere with a strong magnetic field , soon thereafter it was noted that these lines might be of electron - cyclotron origin , and 1e 1207 could be a bare strange star with surface field of @xmath125 g@xcite .
further observations of both spectra feature@xcite and precise timing@xcite favor the electron - cyclotron model of 1e 1207 .
but this simple single particle approximation might not be reliable due to high electron density in strange stars , and xu et al .
investigated the global motion of the electron seas on the magnetized surfaces@xcite .
it is found that hydrodynamic surface fluctuations of the electron sea would be greatly affected by the magnetic field , and an analysis shows that the seas may undergo hydrocyclotron oscillations whose eigen frequencies are given by @xmath126 $ ] , where @xmath127 and @xmath128 is the cyclotron frequency .
the fact that the absorption feature of 1e 1207.4 - 5209 at @xmath129 kev is not much stronger than that at @xmath130 kev could be understood in this hydrocyclotron oscillations model , because these two lines with @xmath46 and @xmath131 could have nearly equal intensity , while the strength of the first harmonic is much smaller than that of the fundamental in the electron - cyclotron model . besides the absorption in 1e 1207.4 - 5209
, the detected lines around @xmath132 kev in the burst spectrum of sgr 1806 - 20 and those in other dead pulsars ( e.g. , radio quiet compact objects ) would also be of hydrocyclotron origin@xcite .
_ planck - like continue spectra . _ the x - ray spectra from some sources ( e.g. , rx j1856 )
are well fitted by blackbody , especially with high - energy tails surprisingly close to wien s formula : decreasing exponentially ( @xmath133 ) . because there is an atmosphere above the surface of neutron star / crusted strange stars ,
the spectrum determined by the radiative transfer in atmosphere should differ substantially from planck - like one , depending on the chemical composition , magnetic field , etc.@xcite can the thermal spectrum of quark - cluster star be well described by planck s radiation law ? in bag models where quarks are nonlocal , one limitation is that bare strange stars are generally supposed to be poor radiators in thermal x - ray as a result of their high plasma frequency , @xmath134 mev . nonetheless , if quarks are localized to form quark - clusters in cold quark matter due to very strong interactions , a regular lattice of the clusters ( i.e. , similar to a classical _ solid _ state ) emerges as a consequence of the residual interaction between clusters@xcite . in this latter case ,
the metal - like solid quark matter would induce a metal - like radiative spectrum , with which the observed thermal x - ray data of rx j1856 can be fitted@xcite .
alternatively , other radiative mechanism in the electrosphere ( e.g. , electron bremsstrahlung in the strong electric field@xcite and even of negligible ions above the sharp surface ) may also reproduce a planck - like continue spectrum .
_ supernova and gamma - ray bursts .
_ it is well known that the radiation fireballs of gamma - ray bursts ( grbs ) and supernovae as a whole move towards the observer with a high lorentz factor.@xcite the bulk lorentz factor of the ultrarelativistic fireball of grbs is estimated to be order@xcite of @xmath135 .
for such an ultra - relativistic fireball , the total mass of baryons can not be too high , otherwise baryons would carry out too much energy of the central engine , the so - called `` baryon contamination '' . for conventional neutron stars as the central engine , the number of baryons loaded with the fireball is unlikely to be small , since neutron stars are gravity - confined and the luminosity of fireball is extremely high .
however , the baryon contamination problem can be solved naturally if the central compact objects are strange quark - cluster stars .
the bare and chromatically confined surface of quark - cluster stars separates baryonic matter from the photon and lepton dominated fireball . inside the star , baryons are in quark - cluster phase and can not escape due to strong color interaction , but @xmath136-pairs , photons and neutrino pairs can escape from the surface .
thus , the surface of quark - cluster stars automatically generates a low baryon condition for grbs as well as supernovae.@xcite it is still an unsolved problem to simulate supernovae successfully in the neutrino - driven explosion models of neutron stars .
nevertheless , in the quark - cluster star scenario , the _ bare _ quark surfaces could be essential for successful explosions of both core and accretion - induced collapses@xcite .
a nascent quark - cluster star born in the center of grb or supernova would radiate thermal emission due to its ultrahigh surface temperature@xcite , and the photon luminosity is not constrained by the eddington limit since the surface of quark - cluster stars could be bare and chromatically confined .
therefore , in this photon - driven scenario,@xcite the strong radiation pressure caused by enormous thermal emissions from quark - cluster stars might play an important role in promoting core - collapse supernovae .
calculations have shown that the radiation pressure due to such strong thermal emission can push the overlying mantle away through photon - electron scattering with energy as much as @xmath137 ergs .
such photon - driven mechanism in core - collapse supernovae by forming a quark - cluster star inside the collapsing core is promising to alleviate the current difficulty in core - collapse supernovae .
the recent discovery of highly super - luminous supernova asassn-15lh , with a total observed energy @xmath138 ergs,@xcite might also be understood in this regime if a very massive strange quark - cluster star , with mass smaller than but approaching @xmath0 , forms .
_ free or torque - induced precession .
_ rigid body precesses naturally when spinning , either freely or by torque , but fluid one can hardly .
the observation of possible precession or even free precession of b1821 - 11@xcite and others could suggest a global solid structure for pulsar - like stars .
low - mass quark stars with masses of @xmath139 and radii of a few kilometers are gravitationally force - free , and their surfaces could then be irregular ( i.e. , asteroid - like ) .
therefore , free or torque - induced precession may easily be excited and expected with larger amplitude in low - mass quark stars .
the masses of axps / sgrs ( anomalous x - ray pulsars / soft gamma - ray repeaters ) could be approaching the mass - limit ( @xmath140 ) in the aiq ( accretion - induced quake ) model@xcite ; these objects could then manifest no or weak precession as observed , though they are more likely than ccos / dtns ( eg . , rx j1856 ) to be surrounded by dust disks because of their higher masses ( thus stronger gravity ) . _
normal and slow glitches . _
a big disadvantage that one believes that pulsars are strange quark stars lies in the fact that the observation of pulsar glitches conflicts with the hypothesis of conventional quark stars in fluid states@xcite ( e.g. , in mit bag models ) .
that problem could be solved in a solid quark - cluster star model since a solid stellar object would inevitably result in star - quakes when strain energy develops to a critical value .
huge energy should be released , and thus large spin - change occurs , after a quake of a solid quark star .
star - quakes could then be a simple and intuitional mechanism for pulsars to have glitches frequently with large amplitudes . in the regime of solid quark star , by extending the model for normal glitches@xcite , one can also model pulsar s slow glitches@xcite not well understood in ns models .
in addition , both types of glitches without ( vela - like , type i ) and with ( axp / sgr - like , type ii ) x - ray enhancement could be naturally understood in the star - quake model of solid strange star,@xcite since the energy release during a type i ( for fast rotators ) and a type ii ( for slow rotators ) starquake are very different .
_ energy budget_. the substantial free energy released after star - quakes , both elastic and gravitational , would power some extreme events detected in axps / sgrs and during grbs .
besides persistent pulsed x - ray emission with luminosity well in excess of the spin - down power , axps / sgrs show occasional bursts ( associated possibly with glitches ) , even superflares with isotropic energy @xmath141 erg and initial peak luminosity @xmath142 times of the eddington one .
they are speculated to be magnetars , with the energy reservoir of magnetic fields @xmath143 g ( to be still a matter of debate about the origin since the dynamo action might not be so effective and the strong magnetic field could decay effectively ) , but failed predictions are challenges to the model.@xcite however , axps / sgrs could also be solid quark stars with surface magnetic fields similar to that of radio pulsars .
star - quakes are responsible to both bursts / flares and glitches in the latter scenario,@xcite and kinematic oscillation energy could effectively power the magnetospheric activity.@xcite the most conspicuous asteroseismic manifestion of solid phase of quark stars is their capability of sustaining torsional shear oscillations induced by sgr s starquake@xcite .
in addition , there are more and more authors who are trying to connect the grb central engines to sgrs flares , in order to understand different grb light - curves observed , especially the internal - plateau x - ray emission.@xcite _ mass and radius of compact star_. the eos of quark - cluster matter would be stiffer than that of nuclear matter , because ( 1 ) quark - cluster should be non - relativistic particle for its large mass , and ( 2 ) there could be strong short - distance repulsion between quark - clusters .
besides , both the problems of hyperon puzzle and quark - confinement do not exist in quark - cluster star .
stiff eos implies high maximum mass , while low mass is a direct consequence of self - bound surface .
it has been addressed that quark - cluster stars could have high maximum masses ( @xmath144 ) as well as very low masses ( @xmath145).@xcite later radio observations of psr j1614 - 2230 , a binary millisecond pulsar with a strong shapiro delay signature , imply that the pulsar mass is [email protected] @xmath147@xcite , which indicates a stiff eos for cbm .
another 2@xmath147 pulsar is also discovered afterwards@xcite .
it is conventionally thought that the state of dense matter softens and thus can not result in high maximum mass if pulsars are quark stars , and that the discovery of massive 2@xmath147 pulsar may make pulsars unlikely to be quark stars .
however , quark - cluster star could not be ruled out by massive pulsars , and the observations of pulsars with higher mass ( e.g. @xmath148 ) , would even be a strong support to quark - cluster star model , and give further constraints to the parameters .
the mass and radius of 4u 1746 - 37 could be constrained by pre ( photospheric radius expansion ) bursts , on the assumption that the touchdown flux corresponds to eddington luminosity and the obscure effect is included.@xcite it turns out that 4u 1746 - 37 could be a strange star with small radius
. there could be other observational hints of low - mass strange stars .
thermal radiation components from some psr - like stars are detected , the radii of which are usually much smaller than 10 km in blackbody models where one fits spectral data by planck spectrum,@xcite and pavlov and luna@xcite find no pulsations with periods longer than @xmath149 s in the cco of cas a , and constrain stellar radius and mass to be @xmath150 and @xmath151 in hydrogen ns atmosphere models .
two kinds of efforts are made toward an understanding of the fact in conventional ns models .
( 1 ) the emissivity of ns s surface is nt simply of blackbody or of hydrogen - like atmospheres .
the cco in cas a is suggested to covered by a carbon atmosphere@xcite .
however , the spectra from some sources ( e.g. , rx j1856 ) are still puzzling , being well fitted by blackbody .
( 2 ) the small emission areas would represent hot spots on ns s surfaces , i.e. , to fit the x - ray spectra with at least two blackbodies , but this has three points of weakness in ns models .
@xmath90 , about @xmath97 and @xmath152 .
no or very weak pulsation has been detected in some of thermal component - dominated sources ( e.g. , the cas a cco@xcite ) , and the inferred magnetic field from @xmath152 seems not to be consistent with the atmosphere models at least for rx j1856@xcite .
@xmath22 , fitting of thermal x - ray spectra ( e.g. , psr j1852 + 0040 ) with two blackbodies finds two small emitting radii ( significantly smaller than 10 km ) , which are not yet understood@xcite .
@xmath20 , the blackbody temperature of the entire surface of some psr - like stars are much lower than those predicted by the standard ns cooling models,@xcite even provided that hot spots exist . nevertheless , besides that two above , a _
natural _ idea could be that the detected small thermal regions ( _ if _ being global ) of ccos and others may reflect their small radii ( and thus low masses in quark - cluster star scenario).@xcite another low - mass strange ( quark - cluster ) star could be 4u 1700 + 24 . because of strangeness barrier existing above a quark - cluster surface , a strange star may be surrounded by a hot corona or an atmosphere , or even a crust for different accretion rates .
both the redshifted o viii ly-@xmath153 emission line ( only @xmath154 ) and the change in the blackbody radiation area ( with an inferred scale of @xmath155 m ) could naturally be understood if 4u 1700 + 24 is a low - mass quark - cluster star which exhibits weak wind accretion.@xcite additionally , the mass function via observing the g - type red giant company is only @xmath156,@xcite from which the derived mass of compact star should be much lower than @xmath157 unless there is geometrically fine - tuning ( inclination angle @xmath158 , see fig . [ f3 ] ) .
, width=302 ] all these three independent observations ( redshift , hot spot and mass function ) may point to the fact that 4u 1700 + 24 could be a low mass strange quark - cluster star .
future observations with more advanced facilities , such as fast and ska , could provide more observational hints for the nature of cbm .
pulsar mass measurement could help us find more massive pulsar , while measurement of the momentum of inertia may give information on the radius .
searching sub - millisecond pulsars could be an expected way to provide clear evidence for ( low - mass ) quark stars .
normal neutron stars can not spin with periods less than @xmath159 ( @xmath160 ) , as the rotation is limited by kepler frequency .
but low - mass bare strange stars has no such limitation on the spin period , which could be even less than 1 ms .
we need thus a much short sampling time , and would deal with then a huge amount of data in order to find a sub - millisecond pulsars . besides , the pulse profile of pulsar is helpful for the understanding of its magnetospheric activity .
strange quark - nuggets , in the form of cosmic rays , could be ejected during the birth of central compact star,@xcite or during collision of strange stars in a binary system spiraling towards each other due to loss of orbital energy via gravitational waves.@xcite a strangelet with mass per baryon @xmath161 mev ( i.e. , binding energy per baryon @xmath162 mev ) could be stable in cosmic rays , and would decay finally into nucleons when collision - induced decrease of baryon number make it unstable due to the increase of surface energy .
when a stable strangelet bombards the atmosphere of the earth , its fragmented nuggets may decay quickly into @xmath85-particles by strong interaction and further into nucleons by weak interaction .
what if a strange nugget made of quark clusters bombards the earth ? it is interesting and necessary to investigate . in the early universe ( @xmath163 ) ,
quark - gluon plasma condenses to form hadron gas during the qcd phase transition .
if the cosmological qcd transition is first - order , bubbles of hadron gas are nucleated and grow until they merge and fill up the whole universe . a separation of phases during the coexistence of the hadronic and the quark phases could gather a large number of baryons in strange nuggets.@xcite if quark clustering occurs , evaporation and boiling may be suppressed , and strange nuggets may survive and contribute to the dark matter today .
strange nuggets as cold quark matter may favor the formation of seed black holes in primordial halos , alleviating the current difficulty of quasars at redshift as high as @xmath164,@xcite and the small pulsar glitches detected may hint the role of strange nuggets.@xcite
although normal micro - nuclei are 2-flavour symmetric , we argue that 3-flavour symmetry would be restored in macro / gigantic - nuclei compressed by gravity during a supernova .
strange matter is conjectured to be condensed matter of 3-flavour quark - clusters , and future advanced facilities ( e.g. , fast , ska ) would provide clear evidence for strange stars .
strange nuggets manifested in the form of cosmic rays and even dark matter have significant astrophysical consequences , to be tested observationally
. * acknowledgements . *
this work is supported by the national basic research program of china ( no .
2012cb821801 ) and nnsfc ( no . 11225314 ) .
the fast fellowship is supported by the special funding for advanced users , budgeted and administrated by center for astronomical mega - science , chinese academy of sciences ( cas ) .
we would like to thank ms .
yong su for reading and checking 2.4 . | normal baryonic matter inside an evolved massive star can be intensely compressed by gravity after a supernova .
general relativity predicts formation of a black hole if the core material is compressed into a singularity , but the real state of such compressed baryonic matter ( cbm ) before an event horizon of black hole appears is not yet well understood because of the non - perturbative nature of the fundamental strong interaction . certainly , the rump left behind after a supernova explosion could manifest as a pulsar if its mass is less than the unknown maximum mass , @xmath0 .
it is conjectured that pulsar - like compact stars are made of strange matter ( i.e. , with 3-flavour symmetry ) , where quarks are still localized as in the case of nuclear matter . in principle , different manifestations of pulsar - like objects could be explained in the regime of this conjecture . besides compact stars
, strange matter could also be manifested in the form of cosmic rays and even dark matter .
the baryonic part of the universe is well - understood in the standard model of particle physics ( consolidated enormously by the discovery of higgs boson ) , where quark masses are key parameters to make a judgment on the quark - flavour degrees of freedom at a certain energy scale . unlike the leptons
, quarks could be described with mass parameters to be measured indirectly through their influence on hadronic properties since they are confined inside hadrons rather than free particles .
the masses of both up and down quarks are only a few mev while the strange quark is a little bit heavier , with an averaged mass of up and down quarks , @xmath1 mev , as well as the strange quark mass of @xmath2 mev obtained from lattice qcd ( quantum chromo - dynamics ) simulations.@xcite for nuclei or nuclear matter , the separation between quarks is @xmath3 fm , and the energy scale is then order of @xmath4 mev according to heisenberg s relation @xmath5 mev@xmath6fm .
one may then superficially understand why nuclei are of two ( i.e. , @xmath7 and @xmath8 ) flavours as these two flavours of quarks are the lightest .
however , because the nuclear energy scale is much larger than the mass differences between strange and up / down quarks , @xmath9 , why is the valence strangeness degree of freedom absolutely missing in stable nuclei ? we argue and explain in this paper that 3-flavour ( @xmath7 , @xmath8 and @xmath10 ) symmetry would be restored if the strong - interaction matter at low temperature is very big , with a length scale @xmath11 the electron compton wavelength @xmath12 .
we call this kind of matter as _ strange matter _ too , but it is worth noting that quarks are still localized with this definition ( in analogy to 2-flavour symmetric nuclei ) because the energy scale here ( larger than but still around @xmath13 ) is still much smaller than the perturbative scale of qcd dynamics , @xmath14 gev . we know that normal nuclei are relatively small , with length scale @xmath15 fm @xmath16 , and it is very difficult for us to gather up huge numbers ( @xmath17 ) of nuclei together because of the coulomb barrier between them in laboratory . then , where could one find a large nucleus with possible 3-flavour symmetry ( i.e. , strange matter ) ? such kind of strange matter can only be created through extremely astrophysical events . a good candidate of strange matter
could be the supernova - produced rump left behind after core - collapsing of an evolved massive star , where normal _
micro_-nuclei are intensely compressed by gravity to form a single _ gigantic _ nucleus ( also called as compressed baryonic matter , cbm ) , the prototype of which was speculated and discussed firstly by lev landau.@xcite the strange matter object could manifest the behaviors of pulsar - like compact stars if its mass is less than @xmath0 , the maximum mass being dependent on the equation of state of strange matter , but it could soon collapse further into a black hole if its mass @xmath18 .
we may then conclude that strange matter could be the state of gravity - controlled cbm before an event horizon comes out ( i.e. , a black hole forms ) .
this paper is organized as follows . in 1 , the gravity - compressed dense matter ( a particular form of cbm ) , a topic relevant to einstein s general relativity ,
is introduced in order to make sense of realistic cbm / strange matter in astrophysics .
we try to convince the reader that such kind of astrophysical cbm should be in a state of strange matter , which would be distinguished significantly from the previous version of strange quark matter , in 2 .
cold strange matter would be in a solid state due to strong color interaction there , but the solution of a solid star with sufficient rigidity is still a challenge in general relativity .
nevertheless , the structure of solid strange star is presented ( 3 ) in the very simple case for static and spherically symmetric objects .
different manifestations and astrophysical implications of strange matter are broadly discussed in 4 . finally ,
5 is a brief summary . |
the `` ribbon '' structures of solar flares have long been observed at chromospheric lines ( e.g , h@xmath0 at 656.3 nm and caii h at 396.8 nm ) .
a ribbon in one magnetic polarity region has its counterpart in the other magnetic polarity region , and both run parallel to the magnetic polarity inversion line lying between them .
the well - observed separation motion of ribbons is a direct mapping of energy release via magnetic reconnection in the corona reconnecting current sheet ( rcs ) @xcite , the rate of which can be evaluated as @xmath1 , where @xmath2 is the corresponding magnetic field component perpendicular to the surface ribbon element @xmath3 . under a simplified two - dimensional reconnection model @xcite , the electric potential drop , @xmath4 , along coronal separator with length element @xmath5 equals the reconnection rate @xmath6 . by considering that the coronal separator has the same length as flare ribbon , the electric field
can thus be expressed as @xmath7 , where u is the ribbon expansion velocity .
this enables the determination of the electric field at the rcs through observable quantities in the lower atmosphere , while large uncertainties may be induced in evaluating @xmath8 especially for ribbons with irregular shape .
alternatively , we infer the _ average _ electric field used in the present study as @xmath9 where @xmath10 is the length of ribbon , and the `` @xmath11 '' and `` @xmath12 '' subscript denote the corresponding physical quantity measured for the ribbons located in the positive and negative magnetic fields , respectively . in this way , the reconnection rate @xmath6 can be calculated essentially based on progression of ribbon intensity ( e.g. , * ? ? ?
* ; * ? ? ?
importantly , reconnecting magnetic fluxes from positive and negative magnetic fields should be identical in principle ; however , their actual measurements do not always yield a good balance ( e.g. , * ? ? ?
therefore , the average value of @xmath13 , as defined above , may be able to represent the overall strength of reconnection electric field at the rcs , which is regarded as conceptually the most straightforward mechanism to directly accelerate electrons to high energies @xcite .
a powerful diagnostic of accelerated energetic electrons produced by flares is the hard x - ray ( hxr ) emission , the spectrum of which often appears as a power - law distribution in photon energy [ @xmath14 ( e.g. , * ? ? ?
the hardness of x - ray emission , i.e. , the power - law spectral index @xmath15 , implies a characteristic in the energy distribution of the electron flux bombarding the dense target under the bremsstrahlung emission mechanism @xcite .
specifically , a harder x - ray spectrum ( with smaller spectral index ) indicates that more electrons are accelerated to higher energies . for this study
, x - ray spectrum is measured for entire flaring region on the solar disc , thus the derived spectral index reflects the overall hardness of x - ray emission ( thereafter referred as @xmath16 ) created by downward traveling accelerated electrons impinging on the chromosphere .
previous studies of temporal evolution of reconnection electric field @xmath13 and hxr spectral index @xmath15 in individual flares show that they are correlated and anti - correlated , respectively , with that of the hxr flux ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , which naturally suggests an anticorrelation relationship between @xmath13 and @xmath15 .
this is also hinted by particle simulations with prescribed electric and magnetic fields @xcite , which show that the electron energy spectrum hardens when the electric field increases . a direct comparison of @xmath13 and @xmath15 using observational data is thus needed to bridge the research of magnetic reconnection and electron acceleration , which is believed to have an intrinsic causal relationship .
in retrospect , only a single such attempt has been reported .
it was the 2005 may 13 m8.0 flare with ribbon - like hxr emissions @xcite , where spatially resolved @xmath13 and @xmath15 determined by tracing each position along ribbon and imaging spectroscopy for each pixel , respectively , can be compared @xcite .
as ribbon - like hxr sources are not often observed , in this letter , we take a different approach and investigate the relationship between @xmath17 and @xmath16 using a sample of 13 two - ribbon flares .
the chosen flares were all well observed on the disk with an orientation cosine factor of @xmath180.7 ( @xmath1946@xmath20 from the disc center ) , exhibiting clear motion of ribbon separation with nearly full coverage of x - ray observations from the _ reuven ramaty high energy solar spectroscopic imager _ ( _ rhessi _ ; * ? ? ?
ribbon observations at h@xmath0 and caii h spectral line were acquired by the global h@xmath0 network @xcite ( 10 events ) , the solar optical telescope ( sot ; * ? ? ?
* ) on _ hinode _ ( 2 events ) , and the usaf / optical solar patrol network ( ospan , formerly known as isoon ; * ? ? ?
* ) ( 1 event ) .
these images have a cadence of 0.52 minutes and a pixel resolution of @xmath180.11 . for the sake of consistency
, we use photospheric magnetic fields of the flaring region ( @xmath182 pixel@xmath21 ) measured with the michelson doppler imager ( mdi ; * ? ? ?
* ) on the _ solar and heliospheric observatory _
( _ soho _ ) for all the events .
lcccccccc ( 1 ) 2002 feb 18 & 09830 & s20 e16 & 0.9 & m1.0 & 21:15 & 2.3 @xmath22 0.7 & 6.84 & 0 + ( 2 ) 2002 feb 20 & 09830 & s18 w11 & 0.93 & m2.4 & 21:07 & 5.8 @xmath22 1.3 & 3.64 & 0 + ( 3 ) 2003 apr 23 & 10338 & n20 w22 & 0.87 & m2.0 & 15:56 & 3.2 @xmath22 0.8 & 5.1 & -1 + ( 4 ) 2003 may 27 & 10365 & s07 w17 & 0.95 & x1.3 & 23:07 & 7.7 @xmath22 1.8 & 2.7 & 0 + ( 5 ) 2003 oct 29 & 10486 & s15 w02 & 0.96 & x10.0 & 20:49 & 16.7 @xmath22 5.9 & 1.88 & 1 + ( 6 ) 2004 mar 30 & 10581 & s05 e01 & 0.99 & c2.0 & 23:08 & 1.6 @xmath22 0.4 & 6.94 & 0 + ( 7 ) 2004 nov 4 & 10696 & n08 e18 & 0.94 & m5.4 & 23:09 & 6.2 @xmath22 1.3 & 2.98 & 0 + ( 8) 2005 jan 15 & 10720 & n15 w05 & 0.96 & x2.6 & 23:02 & 9.2 @xmath22 2.1 & 2.46 & 0 + ( 9 ) 2005 may 13 & 10759 & n12 e12 & 0.96 & m8.0 & 16:57 & 5.7 @xmath22 1.4 & 3.21 & -1.5 + ( 10 ) 2005 may 26 & 10767 & s06 e13 & 0.97 & c8.6 & 21:39 & 2.1 @xmath22 0.5 & 6.06 & 1 + ( 11 ) 2006 jul 6 & 10898 & s11 w32 & 0.83 & m2.5 & 08:36 & 4.9 @xmath22 1.7 & 3.46 & -3 + ( 12 ) 2006 dec 14 & 10930 & s06 w46 & 0.69 & x1.5 & 22:15 & 9.2 @xmath22 2.1 & 4.0 & -1 + ( 13 ) 2007 jul 10 & 10963 & s07 e45 & 0.7 & c5.2 & 17:53 & 2.1 @xmath22 0.5 & 7.73 & 0 we take the 2005 january 15 x2.6 flare as an example to demonstrate our method of data reduction .
figure [ fig1]@xmath23 shows an h@xmath0 image near the flare peak time superimposed with x - ray sources .
this is a typical two - ribbon flare , with part of h@xmath0 ribbons occupied by hxr emitting sources ( 2550 kev ) and a lower energy x - ray source ( 1225 kev ) lying in - between .
we developed a standardized procedure to derive the reconnection electric field from ribbon motion .
modules of this procedure include accurate image alignment , appropriate image destretching for ground - based observation , ribbon tracking using the intensity - based binary masks method ( e.g. , * ? ? ?
* ) , ribbon length estimation , and registering h@xmath0 images on the corresponding magnetogram by multiple feature matching .
we consider uncertainties in this method that mainly stem from the following aspects : ( 1 ) the choice of the cutoff value that defines the brightened ribbons .
we first determine a central cutoff value at the evolving edge of the ribbon based on the difference images relative to a fixed preflare frame .
we then vary this central cutoff value within @xmath2220% and use 11 thresholds in total to characterize the brightened ribbons at each time interval , in order to have a better evaluation of the reconnection rate ; ( 2 ) magnetic field measurement .
mdi level 1.8 full - disk magnetograms are used , and a largest noise level of 30 g is taken into account . since mdi magnetogram could be affected by high energy emissions during flares ( e.g. , * ? ? ?
* ) , we use magnetograms just before the flares for calculation .
we note that it is known that mdi data could suffer from zeeman saturation , and the level 1.5 data could underestimate the flux density by a factor of @xmath180.640.69 @xcite .
however , in most of our studied events , the regions swept by flare ribbons are not obviously affected by zeeman saturation , and the values of level 1.8 calibrated field density are @xmath181.6 times those of the uncalibrated level 1.5 data .
therefore , we do not expect the inherent limitations in mdi measurement to significantly alter out results ; and ( 3 ) ribbon length estimation .
we measure the length of the brightened ribbons as defined above . while as the shape of flare ribbons may not be regular and actual flare process may take place in a more complicated three - dimensional structure ( see more discussion in [ discussion ] ) , we take an error of 30% in ribbon length measurement . moreover , caution has been taken when there is a complicated ribbon dynamics , only later stage of which exhibits classical separation motion and can be regarded as complying with the standard two - dimensional model ( e.g. , the 2003 october 29 x10 flare ; e.g. , @xcite ) .
thus in such cases ( indicated in table [ tbl1 ] ) , we extract results only from the clear ribbon separation phase .
figure [ fig1]@xmath24 depicts front of the x2.6 flare ribbons as color - coded lines using a central brightness cutoff value , which shows clearly the ribbon separation motion in opposite magnetic fields .
the computed @xmath17 ( _ red _ ) is plotted in figure [ fig1]@xmath25 .
its peak ( @xmath189.2 v cm@xmath21 ) is co - temporal with that of the hxr emission in 25 - 50 kev ( _ black _ ) , which is previously found in other events @xcite . for analysis of x - ray spectrum
, we used the object spectral executive ( ospex ) software package and applied the pileup correction , if necessary , in solarsoftware ( ssw ) .
we set the energy bin width to 1/3 kev , the standard finest binning , and choose the length of fitting time interval to be the same as the cadence of ribbon images centering on the time of each frame , except in the case that we have to avoid the time when _ rhessi _
attenuators change status .
we model the thermal plasma and high - energy emissions with an isothermal component and nonthermal power - law distributions , respectively . quite often ,
the nonthermal emission is best fitted using a double power - law function with two different spectral indices . in such a case , we take the spectral index reflecting the hardness of photons with medium energies ( @xmath182050 kev ) as the result for that event ( see discussion in [ discussion ] ) .
the uncertainty in the values of hxr spectral index may arise from the choice of the length of fitting interval and spectrum background , and is estimated to be @xmath185% .
the x - ray spectral fitting results of the x2.6 flare around its peak time is shown in figure [ fig1]@xmath27 , where @xmath16 reaches @xmath182.46 in @xmath182060 kev . the time profile of @xmath16 ( _ blue _ ) is plotted in figure [ fig1]@xmath25 , in comparison with those of @xmath17 and the 2550 kev photon flux . as usual , the minimum overall spectral index occurs when the hxr peaks , which is believed to be an intrinsic feature of electron acceleration in the flare impulsive phase ( e.g. , * ? ? ?
following similar procedure , we calculate reconnection electric field and hxr spectral index for other events , and give the results in table [ tbl1 ] .
the maximum @xmath17 and the minimum @xmath16 lie in the range of @xmath18120 v cm@xmath21 and @xmath1828 , respectively , which is generally comparable with what previously reported ( e.g. , * ? ? ?
* ; * ? ? ?
in a majority of events , the times when @xmath17 and @xmath16 reach the maximum and minimum values , respectively , agree within one minute .
figure [ fig2 ] ( _ top _ ) shows the scatter plot of these two quantities , with a value of logarithmic correlation coefficient ( cc ) computed to be @xmath120.94 .
that is to say , a larger electric field in the rcs corresponds to a harder x - ray spectrum ( with smaller @xmath15 ) and presumably to a harder electron precipitation spectrum , which is qualitatively consistent with the trend predicted by numerical simulations @xcite .
this high correlation relationship thus strongly support the hypothesis that direct acceleration by the electric field generated by magnetic reconnection may play an important role in producing energetic electrons in flares .
we note that x , m , and c - class flares appear to be three populations , which indicates that flares with larger magnitude are almost always associated with larger electric fields and harder hxr spectra .
this tendency is more clearly seen in figure [ fig2 ] ( _ bottom _ ) , where we make scatter plot of the maximum @xmath17 versus the peak flux in 18 soft x - ray measured by the _ geostationary operational environmental satellite _
( _ goes _ ; * ? ? ? * ) .
the logarithmic cc also has a high coefficient of 0.97 for all the data sets .
this further implies that larger electric field strength may be responsible for stronger electron acceleration and hence stronger high - energy emissions seen in flares with larger magnitude .
joining the research of magnetic reconnection and electron acceleration , we have determined the average electric field @xmath17 in the rcs and the overall hardness of x - ray emission @xmath16 from flare observations of 13 events .
we then find a strong correlation relationship between both the minimum @xmath16 and the flare magnitude corresponding to the maximum @xmath17 .
this is , to our knowledge , the clearest direct observational evidence obtained so far of electron acceleration by reconnection electric field in solar flares , which has long been explored but mostly by theoretical models @xcite .
we discuss the present results with related observation and simulation works .
the present research is closest to that of @xcite , who compared spatially resolved electric field and hxr spectral index for a single event . in their result , an anticorrelation relationship between @xmath13 and @xmath15 is also found , while for the @xmath13 with similar strength , the corresponding @xmath15 generally has smaller values than that in our study .
a possible explanation for this discrepancy is that different from the average @xmath13 and the overall @xmath15 herein presented , they explored the local electric field by tracing motion of every position along ribbon , and local spectral index in each pixel by applying an imaging spectroscopy technique .
a larger error could be induced in calculating such @xmath13 and @xmath15 , but it is more possible that different approach , i.e. , global or local , matters in studying the flare energy release . in another study , @xcite found that @xmath28 is anti - correlated with @xmath15 in two flares .
however , one of their event is close to limb ( with a cosine factor of @xmath180.49 ) , and they used ribbon observations in 1600 that forms in upper chromosphere and transition region .
more importantly , instead of @xmath28 , @xmath13 is believed to be the physical quantity directly associated with electron acceleration .
previous studies of reconnection electric field usually use the form , @xmath7 , while @xmath8 is not easy to be measured accurately ( cf .
we use another form , @xmath29 , where @xmath28 can be evaluated in detail using ribbon intensity ( e.g. , * ? ? ?
we must note that the ribbon length @xmath10 is an invariable in the standard two - dimensional model ; however , it usually involves rapidly during actual flares especially in the early phase .
although chromospheric ribbons may be regarded as an instantaneous response of energy release in the coronal current sheet through precipitation of accelerated electrons , recent observational studies have suggested the requirement of consideration of three - dimensional magnetic structures ( e.g. , * ? ? ?
* ; * ? ? ?
the deviation from the standard two - dimensional model is also implied by the asymmetry of flare geometry . for the studied events , the ratio of reconnection electric field and ribbon length measured in opposite magnetic field regions ,
i.e. , @xmath30 and @xmath31 , generally falls into the range of @xmath180.22 and 0.33 , respectively . without an applicable three - dimensional flare model , we thus use the equation in
[ introduction ] as an approximation , where we measure the length of ribbon when it fully developed and treat it as a constant .
this simplification may be justified , because most of our events do not show a significant change of ribbon length at times when the maximum @xmath17 and the minimum @xmath16 occur .
using the time - dependent change of ribbon length as a probe to the magnetic reconnection process is an area of on - going research ( e.g. , * ? ? ?
our observational results demonstrate that there exists a remarkable anticorrelation relationship between @xmath13 and @xmath15 .
this qualitatively agrees with model results @xcite , in that both indicate a hardening of electron energy distribution with increasing electric field strength .
quantitatively comparison is beyond the scope of this study , as further assumptions need to be made before numerical results of electron spectra can be compared with observed photon spectra .
an important issue is that in the simulation of @xcite , it is found that the accelerated electrons exhibit not only a power - law energy spectrum but also an exponential tail at the high energy end .
the latter indicates that diffuse shock acceleration may be at work .
therefore we take only the spectral indices of the hxrs with medium energies as the results , when the spectrum can not be fitted using a single power - law function .
this also implies that different acceleration mechanisms may all contribute to electron acceleration during flares @xcite .
nevertheless , the presence of such clear correlations of observational signatures provides strong support for electron acceleration models based on electric field generated at the rcs .
the authors thank the teams of ghn , _ hinode _ , ospan , _ rhessi _ , and _
soho _ for efforts in obtaining the data .
we thank the referee for helpful comments that improved the letter . c.
l. is indebted to na deng for valuable help in programming and discussion . c. l. and h. w. were supported by nsf grants atm 08 - 19662 and atm 07 - 45744 , and nasa grants nnx 08aq90 g and nnx 07ah78 g .
ghn is operated by the space weather research lab , new jersey institute of technology .
_ hinode _ is a japanese mission developed and launched by isas / jaxa , with naoj as domestic partner and nasa and stfc ( uk ) as international partner .
it is operated by these agencies in co - operation with esa and nsc ( norway ) .
ospan is a pi driven project by air force research laboratory space vehicles directorate ( rvbxs ) and the national solar observatory .
_ rhessi _ is a nasa small explorer . _
soho _ is a project of international cooperation between esa and nasa .
, p. h. , bogart , r. s. , bush , r. i. , hoeksema , j. t. , kosovichev , a. g. , schou , j. , rosenberg , w. , springer , l. , tarbell , t. d. , title , a. , wolfson , c. j. , zayer , i. , & mdi engineering team .
1995 , , 162 , 129 , m. , denker , c. , goode , p. r. , marquette , w. h. , varsik , j. , wang , h. , otruba , w. , freislich , h. , hanslmeier , a. , luo , g. , chen , d. , & zhang , q. 2000 , in esa spec .
463 , the solar cycle and terrestrial climate , solar and space weather , ed .
a. wilson , 617622 , s. , ichimoto , k. , katsukawa , y. , nagata , s. , otsubo , m. , shimizu , t. , suematsu , y. , nakagiri , m. , noguchi , m. , tarbell , t. , title , a. , shine , r. , rosenberg , w. , hoffmann , c. , jurcevich , b. , kushner , g. , levay , m. , lites , b. , elmore , d. , matsushita , t. , kawaguchi , n. , saito , h. , mikami , i. , hill , l. d. , & owens , j. k. 2008 , , 249 , 167 | magnetic reconnection is believed to be the prime mechanism to trigger solar flares and accelerate electrons up to energies of mev . in the classical two - dimensional reconnection model
, the separation motion of chromospheric ribbons manifests the successive reconnection that takes place higher up in the corona .
meanwhile , downward traveling energetic electrons bombard the dense chromosphere and create hard x - ray ( hxr ) emissions , which provide a valuable diagnostic of electron acceleration .
analyses of ribbon dynamics and hxr spectrum have been carried out separately . in this letter , we report a study of the comparison of reconnection electric field measured from ribbon motion and hardness ( spectral index ) of x - ray emission derived from x - ray spectrum .
our survey of the maximum average reconnection electric field and the minimum overall spectral index for 13 two - ribbon flares show that they are strongly anti - correlated .
the former is also strongly correlated with flare magnitude measured using the peak flux of soft x - ray emissions .
these provide strong support for electron acceleration models based on the electric field generated at reconnecting current sheet during flares . |
studying the stellar mass functions of quiescent and star - forming galaxies indicates that star - forming galaxies shut down their star - formation ( `` quenching '' ) with time and increase the number of quiescent galaxies @xcite .
consequently , the color bimodality of the galaxies exist in the local universe and persist up to the high redshifts ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite proposed distinct `` mass '' and `` environment '' processes or modes of quenching for central and satellite galaxies .
however , as we focus on central galaxies in this paper , we do not look at environmental quenching mechanisms . to explain this bimodality ,
several quenching mechanisms are proposed to prevent cooling of gas in / into galaxies .
these processes could e.g. , expel the gas via stellar feedback or active galactic nuclei ( agn ) feedback ( quasar - mode ; e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) or prevent cooling of the gas in halo and keep the gas hot against further accretion to the galaxies , e.g. , halo quenching @xcite , agn radio - mode feedback @xcite and gravitational heating @xcite .
morphological quenching @xcite which stabilize the disk against gravitational collapse of gas and preventing formation of giant molecular clouds is another recently suggested mechanism .
in addition , diverse morphologies of galaxies seen in the local universe have raised questions of how and when the galaxies assembled their stellar masses and structural components .
distinct properties of galaxies at high redshifts @xmath11 compare to the local universe , indicates that galaxies of various types have gone through complex processes .
star - forming galaxies in the local universe could have consist of bulges ( classical , pseudo- or peanut / boxy types ) , or bars in addition to the disk components .
processes that form these structural components of star - forming galaxies , are still matter of debate . many processes including accretion of cold gas and formation / migration of giant gas clumps , major mergers or secular evolutionary mechanisms such as bar - instability are proposed for the formation of the galaxies central over densities or bulges @xcite , ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for recent reviews . ) .
quiescent galaxies , on the other hand , observed to have more concentrated light / mass profiles ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) and have been grown in size since @xmath12 , in particular , the average size of the population of massive quiescent galaxies has grown by a factor of @xmath13 since redshift of @xmath14 to the present ( e.g. , * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ) , implying that these galaxies could have assembled their stellar masses via different mechanism at different epochs .
one possible channel could be the formation of compact spheroids ( or `` blue nuggets '' ) at high redshifts @xcite and their evolution at later times via other processes such as major / minor mergers ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . considering the increasing number density of quiescent galaxies by factor of 10 since @xmath0 from studying the stellar mass functions
, an alternative channel would be due to the progenitor bias , i.e , newly quenched galaxies increase the average size of the quiescent population @xcite .
an open question is how much quiescent galaxies grow individually in size ( e.g. , by merging ) and how much of the average size - increase of the population can be explained by progenitor bias .
therefore , to understand the growth of quiescent galaxies , one has to study the growth of star - forming galaxies .
the emergence of spheroidal - dominant morphology of quiescent galaxies while they migrate from the blue cloud to the red sequence encourages to investigate the correlation between structural parameters with the cessation of star - formation activity .
early studies by @xcite demonstrate that stellar mass surface density of galaxies are better correlated with the age of galaxies than stellar mass , in the local universe ( also * ? ? ?
* ) using sloan digital sky survey ( sdss ) .
the later works shows that galaxies with prominent bulges ( higher srsic index @xmath15 ) are more prone to be quenched @xcite , implying that there might be a link between the star - formation activity and the stellar mass distribution within galaxies .
@xcite extended this study to high redshifts and demonstrate the better correlation of the specific star - formation rate with surface density and inferred velocity dispersion than stellar mass up to @xmath16 ( also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite explored sdss galaxies and showed that central velocity dispersion is the best indicator for predicting galaxy colors than stellar mass , surface density and srsic index .
moreover , @xcite used central stellar mass surface density ( @xmath5 ) , which is the stellar mass in the central 1 kpc region of galaxies and found that this parameter shows less scattered relation with galaxies color than surface density , stellar mass and srsic index at intermediate redshift @xmath17 . following that
, @xcite demonstrate this for local galaxies and stated the necessity of bulge formation for galaxies to be quenched , but not sufficient .
in addition , they discuss that central velocity dispersion with 1 kpc is also correlated with @xmath5 and could be good predictor if measured robustly .
@xcite explored the formation of compact central density for central and satellite galaxies and their halo mass dependence , and argued on the fast quenching conditions related to central density compactness compare to the slow mechanism of halo for central galaxies .
nevertheless , results of these works suggest that determining distribution of stellar masses within galaxies , in particular central mass density , is important for better understanding of quenching mechanism . though the good correlation between central mass density and quenching does not prove the causal relation . @xcite
have shown that a mass - dependent quenching mechanism acting on star - forming galaxies whose size follow the observed size - mass evolution leads automatically to a high central density at the onset of quenching , without however any causal connection between the central density and quenching . in this paper , we study the stellar mass distributions of intermediate to massive galaxies from @xmath18 to @xmath1 , by deriving point - spread function ( psf ) corrected stellar mass profiles of these galaxies consistently at low and high redshifts , and further comparing the corresponding mass distribution parameters with star - formation rates of the galaxies .
many studies used the evolution of light profiles of galaxies to constrain the primary mechanism for assembling the observed structural properties of galaxies at a given stellar mass .
the optical - near infrared light could be a good representative of stellar masses , though the existence of color - gradients at low and high redshifts @xcite could affect estimating the true complex underlying stellar population within galaxies .
the origin of the color - gradient could be due to variation of the stellar population age , dust content and stellar metallicity .
consequently , as discussed by several authors ( e.g. , * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the mass - weighted profiles of galaxies are more robust than the light - weighted profiles , while comparing different galaxy populations .
hence , the constrains on the assembly history of various galaxies can be provided by studying their mass growth on the central and outer regions , at different epochs .
therefore , tracing and constraining the plausible evolutionary models of galaxies could benefits from studying the evolution of the stellar mass distributions within them .
several approaches are employed for deriving mass surface density of galaxies at low and high-@xmath19 . in recent works by @xcite and @xcite ,
the light profiles are converted to the mass profiles , without assuming mass - to - light ratio ( @xmath3 ) gradients .
some authors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) overcome this by exploiting the empirical relation between rest - frame color and @xmath3 , however , as pointed by these authors , the effects of age , metallicty and dust are not exactly the same and hence introduce scatter around the color-@xmath3 relation . to infer the variation of the stellar masses and stellar populations , @xcite and @xcite fitted stellar population models to the resolved 2-dimensional images of the galaxies .
thanks to the high - resolution , sensitive instruments on board of the _ hubble space telescope _ ( hst ) such as advanced camera for survey ( acs ) and wide field camera 3 ( wfc3 ) , this could be feasible out to high - redshifts ( @xmath20 ) . although , the lack of high - resolution middle infra - red images could slightly limited this method , the multi - wavelength observations will allow for fitting the stellar population models to a large sample of galaxies at these redshifts .
this approach is extended by @xcite , by converting 2-dimensional images to 1-dimension images and median - stacking sample of galaxies at each redshift bin , prior to the spectral energy distribution ( sed ) modeling .
the derived stellar surface density are then de - convolved to reduce the effect of psf .
recently , @xcite used observed 1-d profiles of galaxies to derive their mass - profiles . in this study , we exploit the sed fitting technique on one dimensional `` psf - corrected '' light profiles of individual galaxies .
instead of deriving mass density profile from the stacked images ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , we derived psf - corrected mass profiles of individual galaxies from all their available hst filters , taking into account the gradients of @xmath3 .
the layout of this paper is as following . in section 2
, we introduce the sample and data used in this work .
the methods for deriving the stellar mass density are described in section 3 .
we present the comparison of galaxies mass profiles and explore the mass profile parameters on star - formation activity in section 4 & 5 and discussed the results in section 6 .
we summarize our work on section 7 .
the cosmological parameters adopted in this work are @xmath21 = 0.3 , @xmath22 = 0.7 and @xmath23 @xmath24 @xmath25 @xmath26 .
+ color - color selection @xcite , analogues to high-@xmath19 uvj .
sources are color - coded according to their specific star formation rate ( ssfr ) from the mpa - jhu dr7 catalog @xcite.,scaledwidth=40.0% ]
for our analysis , we have used the catalog and imaging of the 3d - hst treasury program @xcite on the great observatories origins deep survey ( goods : * ? ? ? * ) , goods - northern and goods - southern fields ( goods - north and goods - south ) of the cosmic assembly near - ir deep extragalactic legacy survey ( candels ; * ? ? ? * ; * ? ? ? * ) .
the catalogs provided photometry , redshift ( spectroscopic and photomteric ) , the stellar mass and color of galaxies , in addition to the hst imaging of these fields , which are all available at the 3d - hst website .
we use sources with ` use_phot`@xmath27 ( photometric quality flag ) in the 3d - hst catalogs to select a reliable sample . in order to reduce the systematic effects in our methodology ( described in the next section ) for deriving consistent light profiles at different filters , we used psf - matched images at different filters ( smoothed to the @xmath28 image resolution ) . in the @xcite ,
the photometric redshifts are determind using @xmath29 code @xcite for objects without spectroscopic redshift and the stellar masses are measured using fast code @xcite .
the total area of goods - north and goods south fields are about 340 arcmin@xmath30 .
these two goods fields , have observations on seven hst filters in common , @xmath31 , @xmath32 , @xmath33 , @xmath34 , @xmath35 , @xmath36 , @xmath37 ( @xmath38,@xmath39 , @xmath40 , @xmath41 , @xmath42 , @xmath43 , @xmath28 , hereafter ) , providing sufficient wavelength coverage for deriving stellar population modeling from sed fitting , in comparison to other candels fields , which have less wavelengths range ( five filters ) . we have divided our sample into quiescent and star - forming galaxies using uvj color criteria @xcite , based on the @xmath44 and @xmath45 rest - frame colors . the eazy code is used to derive @xmath44 and @xmath45 colors .
our criteria for selecting quiescent galaxies at @xmath46 , is as following : @xmath47\ ] ] @xmath48\ ] ] @xmath49\ ] ] which is similar as the criteria defined by @xcite .
any sources falling outside of these criteria are defined as star - forming galaxies .
the combination of these equations allows separation of un - obscured and dusty star - forming galaxies from the quiescent ones while preventing the contamination . in figure [ fig1 ]
, we show the color criteria ( sold lines ) used in this study to separate star - forming and quiescent galaxies at three different redshift intervals between @xmath50 .
the sources are color - coded according to their total specific star formation rate from 3d - hst catalog using star - formation rate ( using combined uv+ir sfr ) derived by @xcite ( see also * ? ? ? * ) .
as can be seen , the clumps of quiescent galaxies clearly falling in a separate regions of the uvj diagrams .
our samples are then selected to have stellar masses @xmath51 . as shown by @xcite and @xcite , both star forming and quiescent galaxies are complete out to the redshift of 2 ( see also * ? ? ?
we have also used a sample of local galaxies using sdss data release 7 ( dr7 ) images @xcite .
we randomly select 1000 galaxies at @xmath52 , stellar mass matched with the total mass distributions of sdss .
the stellar masses are taken from max - planck - institute for astrophysics ( mpa)-johns hopkins university ( jhu ) sdss dr7 catalog @xcite and the star formation rate from @xcite . in order to divide the local sample into star - forming and quiescent
we use analogues method to the uvj method , using @xmath53 versus @xmath54 color - color criteria , as defined by @xcite : @xmath55 @xmath56 as shown in figure [ fig2 ] , quiescent galaxies occupy separate region from star - forming ones .
similar to figure [ fig1 ] , galaxies are color - coded according to their specific sfr .
we use the new york university value - added galaxy catalog ( nyu - vagc ; * ? ? ?
* ) for local galaxies color and absolute magnitudes .
note that using similar color - color selection criteria to separate quiescent and star - forming galaxies at low and high redshifts reduces biases may caused by different methods of star - formation rate measurements .
+ -band @xmath57 of @xcite and this work as a function of stellar masses .
the green points represent the median differences at each mass bin and 1-@xmath58 error bars .
there is no systematic error in our size measurements .
, scaledwidth=45.0% ]
in this section , we describe the method for deriving the galaxies surface brightness profiles in all available filters .
we first derive the de - convolved profiles of galaxies at different filters , are first derived by fitting psf convolved single srsic @xcite models to the observed two dimensional surface brightness images of galaxies .
we use galfit v3 @xcite to perform two - dimensional fitting while using @xmath28-band psfs ( taken from 3d - hst ) .
the procedure is similar to the @xcite .
we extract @xmath59 cutout images around each galaxy prior to the fit , and masked neighboring objects during the fitting , using mask map provided by 3d - hst .
we take initial guesses for position , ellipticity , effective radius @xmath60 , position angle , from the original catalog and initially assumed srsic index of 2 .
we compare sizes of @xmath28-band images of our galaxies for derived from our method to the ones measured by @xcite , in figure [ fig3 ] , which shows very good consistency between the results and depict the robustness of our size measurements . the procedure for deriving the surface brightness profiles at shorter wavelengths ( i.e. , @xmath38,@xmath39 , @xmath40 , @xmath41 , @xmath42 , @xmath43 ) , is similar to the @xmath28-band .
however , to avoid systematic errors in deriving color and stellar - mass profiles it is essential that the light profiles in different filters originate from a common center .
hence , following @xcite and @xcite , the position , ellipticity and position angle of the best - fit srsic model for @xmath28-band images were adopted as priors while fitting two - dimensional single srsic models to galaxy images at shorter wavelengths . in case , fitting did not converged at these shorter wavelengths , ( or for cases with large uncertainties of the results , i.e. , converging to the boundaries ) , we fixed the srsic index and @xmath60 to the best - fit of its closest longer filter .
we also note that the light profiles are all circularized to remove the effects of ellipticity .
any deviation from a single srsic model ( due to the variation of light distributions between different bands or sub - structural components ) is a matter of concern for deriving the true color and @xmath3 profiles .
the residual - corrected method @xcite could have help to mitigate the problem .
however , this method is sensitive to the noise / background and works best for massive objects . in this study
, we use a sample of galaxies in a wide range of stellar masses and , for the consistency , in our analysis we only use single srsic model profiles . in the appendix , we show that using residual - corrected method should not change the overall results of this study .
the total number of objects used in this study at high-@xmath19 is 2391 . about @xmath61 of sources from original catalog which have less @xmath62 bands detection ( hst filters ; in order to have sufficient wavelength coverage for deriving their mass profiles ) and those with large uncertainties in their derived light profile parameters ( including objects close to very bright sources ) are excluded from our analysis .
+ the procedure for deriving light profiles of the local galaxies ( @xmath1 ) is similar .
we use the @xmath63 images of sdss dr7 galaxies .
we use the @xmath64-band images as a reference ( due to having high s / n ) and derived the light profiles .
we have used synthesized psf images generated by the sdss photo pipelines at the position of each object and for different filters , to derive the bets - fit srsic profiles using galfit ( see * ? ? ?
* for more details ) .
in the following analysis , we use the circularized 1d profiles . the best - model light profiles at different filters derived for each galaxy ,
can be used to measure stellar mass , color and mass - to - light ratio profiles .
the stellar mass profile of galaxies were measured by finding the best spectral energy distribution ( sed ) model at each radius .
we used the fitting and assessment of synthesis template ( fast ) code @xcite and follow the @xcite , for finding the best sed models , i.e. , using the @xcite stellar population evolution models with exponential declining star formation history , assuming @xcite initial mass function ( imf ) and solar metallicity , @xcite dust attenuation law and allowing the @xmath65 to vary between 0 and 4 .
we fixed the redshift of galaxies to the ones provided by the 3d - hst catalog . in details
, we divided the light profiles into small bins of @xmath66 kpc out to 20 kpc and sed fitting each bin , separately .
we assigned the flux errors at each bin , which determined using empty aperture technique around each object , in a similar way described by @xcite , appropriate to the bin sizes .
however , at larger radii , the model fluxes are very low , hence it is crucial to increase the ( model ) signal - to - noise ( @xmath67 ) for robust sed fitting at each profile location .
we used an adaptive binning method ( similar to * ? ? ? * ) and created bins with sufficient @xmath67 ( @xmath68 ) in the @xmath69 color .
however , at larger radii , where model fluxes are very low , we used adaptive binning method to create meshes of 10 model signal - to - noise ( @xmath67 ) in the @xmath69 color , in order to increase the @xmath67 for robust sed fitting .
the stellar masses are then measured by sed - fitting at each radius , and hence creating the stellar mass profiles of galaxies . using @xmath69 color which straddles between , age , metallicity and dust sensitive 4000 and balmer break feature and provides better constrain on sed modelling .
we note our tests shows that the results are robust against using other colors such as @xmath70 color instead of @xmath69 or even using just @xmath71-band filter .
examples of the surface brightness profiles and stellar mass surface densities of six galaxies are shown in figure [ fig4 ] .
for each galaxy , the solid color lines in the upper panels , are best fit surface brightness profiles at different filters and the dashed lines are where the surface brightness profiles of galaxies reaches to the average surface brightness limit of their images .
the bottom panel depicts their stellar mass profiles . for deriving the mass - profiles for @xmath1 galaxies , we use isedfit code @xcite and applying similar @xcite imf to high-@xmath19 sample .
we follow @xcite to constrain star - formation history of local galaxies , i.e. , allowing random star burst in addition to the exponential declining star formation history .
similar to high-@xmath19 sample , we use @xmath72 color for adaptive binning light profiles prior to the sed fitting and deriving their mass profiles .
we note that using similar star - formation history and the same code for high redshift object will not alter the general results of this paper , as described in appendixes .
+ the next step is to measure the half - mass radii from the mass profiles for each object .
for that , we first interpolated between the radial mass bins and then extrapolated the mass density profiles out to @xmath73 kpc by fitting a single srsic model in the outer parts of mass profiles , and then integrating the mass profiles out to this radius to find the half - mass radii .
defining the radius at which the extrapolation should be started beyond that point , is not an straightforward task . in order to secure that the results were not strongly biased toward our guess of this connection point and the extrapolation process in general , we have measured the half mass radii for each galaxy by fitting about 1000 srsic profiles starting at different radii and finding the most frequent output results . in order to check the reliability of this method
, we applied it on a set of 3000 two - component srsic models of density profiles , while random noise was added to the profiles to resemble the measured mass profiles of this study .
the result of our simulation shows that our method is robust and half - mass radii can be recovered without any systematic over our studied stellar mass range ( bottom panel of figure [ fig5 ] ) .
the blue points in figure [ fig5 ] are the median of the relative error of the half - mass sizes with the error on the median and average scatter of @xmath74 . note that our test using simulated profiles shows that fitting a single srsic model to entire profiles cause large scatter on derived half - mass sizes , especially for objects with @xmath75 kpc .
the total stellar masses derived by this method are compared with the total masses from the 3d - hst catalog in the top panel of figure [ fig5 ] .
there is no systematic differences between total stellar masses derived for sources using their mass density profiles and the ones from the original catalog .
note that we only used seven hst filters ( at most ) for each galaxy for sed fitting comparing to so many available complementary filters used in the 3d - hst catalog , including rest - frame near - infrared data from other instruments / telescopes .
the differences in stellar mass estimates possibly arises from outshining effects or dusty regions @xcite .
note that for the rest of the paper , we have corrected the stellar mass profile ( and consequently their related parameters ) to the total masses derived from integrated star - formation history using @xcite which includes lived stellar masses , remnant and mass returned to interstellar medium ( ism ) , simply , the stellar mass is the mass of gas turned into stars .
the advantage of this stellar mass definition is that quiescent galaxies do not suffer from mass loss .
we have compared the half - mass radii and half - light radii of our star - forming and quiescent galaxies at different redshifts and in different stellar mass bins in figure 6 .
the effect of morphological @xmath76-correction has been minimized by using the half - light sizes of object in the closest wfc3/ir or acs band to the rest - frame @xmath775000 i.e. , using @xmath28 for sources at @xmath78 , @xmath42 for objects at @xmath79 and @xmath41 for galaxies at @xmath80 .
as shown in figure [ fig6 ] , at all studied redshifts half - mass sizes are always smaller than half - light radii for both star - forming and quiescent ones ( right and left panels , respectively ) .
the color symbols represent median ratio values for different stellar mass bins .
the quiescent galaxies are on average have half - mass radii about @xmath81 smaller than their half - light radii over the entire studied redshifts , slightly depends on the stellar masses .
massive quiescent ones ( @xmath82 ) are @xmath83 smaller size ratio compare to the lower mass bins with @xmath84 size ratios . using galaxies with
@xmath85 @xcite also found smaller half - mass sizes compare to the half - light sizes but with average of @xmath86 size ratios .
the ratio of sizes for star - forming galaxies , on the other hand , ( right panel of figure [ fig6 ] ) shows little dependence on redshift .
above @xmath87 the ratio is smaller that at lower redshifts , presumably light and stellar masses following more similar distributions , however , at lower redshifts , half - mass sizes are about @xmath88 smaller than their half - light sizes , again depending on stellar masses , with the most massive ones shows larger differences , indicating the stellar mass profiles are more concentrated compare to their light profiles .
as discussed earlier the stellar mass distributions of galaxies at different epoch could help to study their stellar mass assembly .
therefore , in this section , we first compare the stellar mass distributions of different types ( quiescent and star forming ) of galaxies at different redshifts , by means of their mass profiles and then we study how the mass distribution varies for each type by the changes of half - mass radii and surface densities at effective radii and central compactness of our samples .
we have split the sample of galaxies into four stellar mass bins for each type between @xmath89 and compare them at different redshifts .
+ the comparison of mass densities of galaxies at different epochs reveals how the distributions of stellar mass varies at fixed mass .
the median radial distribution of stellar mass densities of quiescent galaxies at different redshifts are shown in figure [ fig7 ] , for different mass bins ( decreasing from left to right panels ) .
starting from the highest stellar mass bin , i.e. , quiescent galaxies within @xmath90 , in the left panel , the significant changes with redshift can be seen in the outer parts of these galaxies .
the density profiles of these galaxies are higher in the outer regions of these galaxies at lower reshifts compared to their high-@xmath19 counterparts , meaning that these galaxies have more stellar masses at their larger radii .
clearly , it can also be seen from middle panels of figure [ fig7 ] that intermediate massive quiescent galaxies with stellar masses @xmath91 and @xmath92 have similar differences in their density profiles in the outer regions at different epochs . in general , at later time , quiescent galaxies at massive and intermediate mass ranges have more assembled stellar masses at larger radii , compared to their counterparts at high-@xmath19 .
accretion of mass via ( minor ) mergers or arrival of the newly quenched systems into red sequence , both are plausible scenarios for the differences @xcite .
this can be examined by studying the evolution of central and effective surface densities that we discuss in the following sections .
it is also needed to study the evolution of the mass profiles of the star - forming galaxies .
note that for the lowest mass bin of quiescent objects within @xmath93 ( right panel of figure [ fig7 ] ) , there is also a hint of having similar behavior , though their specific changes are small compared to the massive ones , and hence the way they changes looks slightly noisy .
we further test this by comparing their half - mass radii in the next section . for the star - forming galaxies , the changes of density profiles have somewhat different behavior .
figure [ fig8 ] presents the comparison of median radial density profiles at different epochs and at fixes mass bins .
the most massive star - forming galaxies with stellar masses @xmath90 , have similar radial distributions of stellar masses at different redshift bins ( left panel of figure [ fig8 ] ) with subtle changes at larger radii at later times .
this shows that at any epoch , the star - forming galaxies have very similar profiles , i.e. , they grow self - similar at all epochs as predicted in simulations @xcite
. however , star - forming galaxies with intermediate stellar masses @xmath94 , have different mass densities at inner and outer regions comparing to their counterparts at higher redshifts .
these galaxies have higher stellar mass densities within @xmath95 kpc at later times , indicating the existence of more prominent central densities ( can be expressed as bulges ) at lower redshifts .
meanwhile the stellar mass densities in the outer regions also changes . therefore , comparing low-@xmath19 and high-@xmath19 star - forming galaxies at intermediate masses , the mass densities varies at all radii , meaning that these objects have assembled stellar masses in both central and outer regions with time , with a different growth rate at the center and the outskirts .
hence it argues against the scenario of the formation of bulges prior to the disk at @xmath96 .
however , as we discuss below , central densities varies less at later time comparing to the mass profiles at larger radii , therefore , suggesting slightly different time scales for their changes .
perhaps , different mechanisms with different time scales could play separate roles for the evolution of mass profiles of star - forming galaxies , e.g. , accretion of gas into disk ( or maybe bulge ) and migration of stars into the central regions via mechanisms such as disk instabilities .
this needs to be addressed in more details .
nevertheless , studying the mass profiles delineates the existence of different mechanisms of stellar mass assembly for star - forming galaxies compared to quiescent ones which is reflected into their mass distributions .
+ the evolution of the half - mass radii ( @xmath97 ) of galaxies is an effective way of illustrating their structural evolution . in left panel of figure [ fig9 ] , the evolution of median half - mass radii of quiescent galaxies between @xmath98
is illustrated for different mass bins ( different colored symbols ) .
as expected , the half - mass radii of the quiescent galaxies increases with time in a similar manner to their half - light radii ( right panel of figure [ fig9 ] ) .
this is the case for galaxies at different mass bins .
we have parameterized the rate of changes as a function of @xmath99 .
the change of mass - weighted sizes for massive ones is as @xmath100 . for highest mass bin ( red circles ) ,
the half - mass sizes increase by a factor of @xmath101 from @xmath102 to @xmath103 , consistent with results from the simulation by @xcite .
the half - mass size evolution is marginally faster than the evolution of their light - weighted sizes with @xmath104 .
the pace of evolution is also similar for quiescent galaxies at lower mass bins , @xmath105 , @xmath106 , @xmath107 , for mass bins of @xmath108 , @xmath109 and @xmath110 , respectively .
this shows that at fixed masses , quiescent galaxies at low redshifts have extended mass density profiles compared to their counterparts at higher redshifts .
the rate of half - mass size evolution for star - forming galaxies is not as fast as quiescent ones .
the left - panel of figure [ fig10 ] , shows that the mass - weighted sizes of these galaxies evolve slowly for our studied mass ranges since @xmath0 .
the rate of evolution for massive star - forming ones is @xmath111 slightly slower than their half - light radii evolution with @xmath112 .
the rate of half - mass size evolution is very similar for intermediate mass ranges of @xmath113 and @xmath114 , with @xmath115 and @xmath116 , respectively .
however , this is much slower for the lowest mass bin of @xmath117 with @xmath118 , consistent with very subtle to no evolution .
the slow evolution of the half - mass sizes , in combination with the results presented in section 4.1 and figure [ fig8 ] , leads us to propose two evolutionary phase of @xmath97 in star - forming galaxies , ( 1 ) self - similar changes of mass - profiles at high-@xmath19 ( at least @xmath119 ) ( e.g. , * ? ? ?
* ) and ( 2 ) increasing the disk sizes once galaxies central regions are saturated at later times ( @xmath120 ) ( see e.g. , * ? ? ?
* ) . in detail
, these trends may actually be mass dependent .
it worth noting that a flattening of the m / l gradient is expected , in particular for massive star - forming galaxies , if the effects at short rest - frame wavelengths of dust and clumpy structures are not fully taken into account .
hence , the single - srsic models and the methodology described in section 3.1 could have introduced a bias affecting the rate of mass - weighted size evolution of the star - forming galaxies .
however , we argue that our results do not change significantly when accounting for residuals from poor best fits ( see appendix for more details ) .
we note that at fixed redshift , the half - light radii of massive star - forming galaxies are larger than their less massive ones . however
, the striking similarities between half - mass radii of these galaxies at different masses at fixed reshift bins , could either be a representative of the similarities between the shapes of their mass profiles ( having exponential shapes with just different normalizations , perhaps at high-@xmath19 ) or the faster increase of the central mass densities of massive ones compared to less massive galaxies , therefore reducing the half - mass sizes for massive ones .
overall , the changes in stellar mass - weighted radii of the star - forming galaxies is very gradual , compared to the quiescent ones at fixed mass .
this has been pointed by @xcite , that mass - size relation of late - type galaxies has very little evolution since redshift of @xmath121 ( see also * ? ? ?
concurrent formation of the bulge and growth of the disk could contribute to this gradual change of half - mass sizes .
based on @xcite , the specific star formation rate ( ssfr ) of galaxies within @xmath122 and @xmath123 are on average flat , which would cause very little mass - weighted size evolution .
we also note that the observed rate of ( light - weighted ) size growth of star - forming galaxies is slightly faster : @xmath124 ( but consistent within the errors in this study , see figure [ fig10 ] ) .
however , the exponent varies between different studies ; @xcite found observed growth of @xmath125 for the light - weighted sizes of star - forming galaxies ( see also * ? ? ?
@xcite show @xmath126 , @xcite observed @xmath127 and @xcite observed @xmath128 .
this discrepancy could first be originated from different sample selection at different redshift ranges , e.g. , @xcite studied disk - like sample ( morphological selection ) at @xmath129 while @xcite used uv - bright galaxies ( color - based selection ) at @xmath130 .
in addition , different mass - normalization of the @xmath57 ( i.e. , different value of @xmath131 in @xmath132 ) between different studies , might introduce some uncertainties . in this work ,
we selected star - forming galaxies based on the color - color selection method and used @xmath1 sample as a reference point for measuring the rate of ( half - mass and half - light ) size evolution . however , a larger sample of galaxies , in particular between @xmath1 and @xmath133 with similar resolution are required for a better estimation of the half - mass size growth rate of star - forming galaxies .
we should also mention that the accuracy of the m / l gradient is particularly important for measuring the rate of half - mass size evolution . to first order , correcting the light profiles using residual - corrected method shows that this evolution is real ( figure [ figa3 ] in the appendix , illustrates this for each stellar mass bin ) .
+ the evolution of mass profiles can also be studied by means of the evolution of surface and central mass densities of galaxies .
stellar mass densities within central 1 kpc region , @xmath5 , is used by some authors @xcite as a proxy for central densities .
in addition surface density @xmath6 is another parameters which can representative of the mass profiles evolution . in this paper , @xmath6 is defined to be the mass surface density at the half - mass radii ( @xmath97 ) .
moreover , we also use the traditional `` effective surface density '' , @xmath134 which is the effective surface density within half - mass radii ( @xmath135 ) . as galaxies half - mass sizes
increase towards lower redshift , it is expected that the surface densities @xmath6 decline correspondingly .
this is shown in the right panels of figure [ fig11 ] and [ fig12 ] , for quiescent and star - forming ones , respectively . at fixed mass ,
quiescent galaxies have higher surface densities in the past compared to their counterparts in the local universe .
this evolution scales as @xmath136 for massive quiescent galaxies and is similar for lower mass bins .
however , the changes of @xmath6 is less significant for massive star - forming galaxies with @xmath137 and is even slower for less massive ones ( see right panel of figure [ fig12 ] ) .
the significant changes in the @xmath6 for quiescent galaxies reflects the increases of stellar masses in the outer regions , as depicted in the figure [ fig7 ] , implying the inside - out growth of these galaxies . for massive star - forming ones ,
the slow evolution of @xmath6 indicates that most of the effect is due to the mass assembly in the outer regions of low-@xmath19 galaxies compared to their high-@xmath19 counterparts .
however , the marginal changes of @xmath6 for star - forming at intermediate masses is due to the increase of surface density at all radii , specifically in their central regions .
this can be further examined by the evolution of the central densities , i.e. @xmath5 . in the left panel of figure [ fig11 ] , the central mass density evolution of quiescent galaxies
is shown at fixed masses .
the central densities decline slightly and the rate of evolution is slow , @xmath138 for massive quiescent galaxies , from @xmath0 to @xmath1 .
this indicates that the central regions of the massive quiescent galaxies were slightly denser at higher redshifts compared to their counterparts at low-@xmath19 .
it is worth reminding that by stellar mass one means here the time integral of the sfr . using stellar masses of live and remnant stars will introduce more significant changes .
we caveat that this is not reflecting the true evolution of properties of individual galaxies .
overall , the slow decline of central densities might be described by arrival of the newly quenched systems ( with slightly lower @xmath5 ) .
@xcite model ( see their fig .
12 ) predicts @xmath139 for the evolution of the population of quiescent @xmath140 galaxies , which is in agreement with our findings , within the errors .
another scenario could be the possible decline in the velocity dispersion of galaxies as a results of dynamical frictions by minor mergers @xcite , though the contributions of mergers needs to be quantified .
the @xmath5 is expected to change very slightly in that case of the inside - out scenario @xcite .
we will address this in more details in section 6 .
for the star - forming ones , there is a weak trend of increasing @xmath5 with time up to @xmath141 , specifically for intermediate and less massive ones ( left panel of figure [ fig12 ] ) , @xmath5 increases from @xmath0 to @xmath133 .
the most massive star - forming galaxies have already reached to their maximum by @xmath142 and remain roughly constant for redshift range of @xmath143 . in general , assuming @xmath5 as a bulge density , star - forming galaxies at lower redshifts have more masses in their central regions ( massive bulges ) compare to their analogous at high-@xmath19 , at fixed mass .
therefore , the results above show that there should be physical processes of mass growth ( as well as re - distributions of stellar masses ) that increase both central and outer parts of star - forming galaxies with time .
as mentioned early , it is important to understand any relation between the stellar mass assembly of galaxies and the star - formation quenching .
therefore , comparing mass profile parameters of star - forming and quiescent galaxies at different epochs is an effective way for investigating the possible relation and physical mechanisms .
several studies have already shown that galaxies with higher surface densities have lower specific star formation rates and the relation hold up to high redshifs ( e.g. , * ? ? ?
* ; * ? ? ?
* ) hence quenching seems to be correlated to the structural parameters ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and reference therein ) .
these studies is followed by @xcite and @xcite to show the links between quenching and bulge mass density ( @xmath5 ) for low and intermediate redshifts .
@xcite also investigate these relations up to @xmath16 .
we revisit these correlations using psf - corrected mass profiles of our sample at fixed masses in different redshift bins @xmath98 measured in a consistent way . in figure [ fig13 ]
we compare the specific star - formation rate of galaxies in our sample ( @xmath144 ) with surface densities within 1 kpc ( @xmath5 ) , surface densities at half - mass radii of galaxies ( @xmath6 ) and effective surface densities within @xmath97 ( i.e. , @xmath134 ) .
the samples are divided into four redshift bins , starting from highest redshift ( @xmath145 ) in the left to the lowest redshift ( @xmath1 ) in the right panel . from top
to bottom row , the distributions are shown for ssfr versus @xmath5 , @xmath6 and @xmath134 , respectively , and objects are color - coded according to their total stellar masses .
as can be seen in this figure , galaxies with higher surface densities have lower ssfrs and this holds at all our studied redshifts .
however , the distribution is tighter for ssfr versus @xmath5 compared to the distributions between ssfr and @xmath6 or @xmath134 , i.e. , @xmath5 of galaxies has less scattered distributions below average ssfr of star - forming galaxies at each redshift compared to the other surface density parameters ( see also * ? ? ?
the distributions of galaxies on ssfr-@xmath5 plane is following a knee shape ( or l-shape as already seen in * ? ? ?
* ) ( top panels of figure [ fig13 ] ) .
this is also similar to the color - central density trend found by @xcite .
the `` l''-shape distributions of galaxies is originated by the two correlations , 1 ) star - formation rate - stellar mass relation ( sfr - m ) of star - forming galaxies known as main sequence relation ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) and 2 ) the correlation between @xmath5 and stellar masses @xcite . the correlation between central densities ( @xmath5 ) and stellar masses is shown to exist up to high-@xmath19 by several authors @xcite and also in simulations @xcite . it is also indicated that @xmath5-stellar mass correlation is tighter than the correlation between average surface density and total stellar masses ( see * ? ? ?
* ) , specifically for quiescent ones .
the tighter distribution between @xmath5 and stellar masses of quiescent galaxies ( low ssfrs ) in comparison to their @xmath6 , originated from the fact that @xmath5 of these galaxies do not change significantly , in contrast to @xmath6 which increased significantly .
therefore , in the top panels of figure [ fig13 ] , galaxies below central density thresholds , which depends on total masses and redshifts , e.g. , @xmath146 @xmath147 , for @xmath148 , retain relatively constant ssfr with slight negative slopes ( reflecting the main - sequence relation and the @xmath5-total mass relation ) .
galaxies with a critical central density threshold have left or about to leave the main sequence relation and the star - formation is ceased in these galaxies .
the distributions above @xmath5 thresholds remain less scattered by the tight correlation between @xmath5 and stellar mass of quiescent ones .
it is tempting to interpret the tighter @xmath5-ssfr correlations in terms of cause and effect between central density and quenching .
however , as @xmath149 also correlates with the total stellar mass , it will be difficult to say whether central density or mass drive the quenching .
therefore , the causal links are difficult to establish .
we shall return to this point in section 6 . to gain more insight on the relation between being quenched and central density , figure [ fig14 ] shows the distributions of galaxies over @xmath5 and the relative distances of galaxies to the main sequence , i.e. , @xmath150sfr plane , for each redshift separately .
we use the best - fit main sequence relation for galaxies at high-@xmath19 and low-@xmath19 from @xcite and @xcite , respectively to find the relative distances .
using @xmath150sfr cancels out the redshift dependence of the distributions on sfr .
objects in the top row panels of figure [ fig14 ] are color - coded according to their srsic indices ( form their light - profiles ) , emphasizing previous results that objects with large distance from the main sequence have higher srsic indices ( more concentrated ) ( e.g. , * ? ? ? * ) .
the middle panels of figure [ fig14 ] is the same as the top panels , but color - coded according to their half - mass sizes .
this reveals that at fixed sfr ( or @xmath150sfr ) galaxies with higher central densities ( @xmath5 ) are more compact , in other words , their stellar masses are more concentrated in their central regions . at fixed mass
, this is also valid ; if we split samples into four stellar mass bins ( figure [ fig15 ] ) , galaxies with higher central densities at fixed @xmath150sfr , have smaller half - mass sizes . as central densities increase the sfr declines for star - forming galaxies .
this leads to nominate @xmath5 thresholds to start quenching mechanism at least , up to @xmath0 .
the bottom panels of figure [ fig14 ] shows the same distributions but only for those classified as star - forming in our sample . at fixed central mass density ,
there is no indication that star - forming galaxies below and above main sequence are more ( or maybe less ) compact ( see also figure [ fig15 ] ) consistent with the results of recent simulations @xcite .
however , these plots point out that star - forming galaxies with higher central densities have smaller mass - weighted sizes with a negative tilt between @xmath150sfr and @xmath5 . if increasing the central mass density is correlated with quenching , then at fixed mass , galaxies which leave the main sequence should have similar mass profiles , particularly in their central regions . to investigate this , we first split the plane of @xmath5-@xmath150sfr into three main parallel regions , i.e.
, galaxies on the main sequence ( @xmath150sfr@xmath151 ) , galaxies in the green valley regions ( @xmath152sfr@xmath153 ) and quenched galaxies ( @xmath154sfr@xmath155 ) .
we compare the median mass profiles of galaxies at fixed masses of these three regions in figure [ fig16 ] and [ fig17 ] , for each redshift bins , separately ( left to right panels , from high-@xmath19 to low-@xmath19 , respectively ) . in these figures ,
red , green and blue lines represent the mass densities of quenched , green valley and main sequence galaxies , respectively .
the median half - mass sizes for each profile is mentioned in each panel accordingly .
as can be seen , at fixed masses , quenched galaxies have higher central densities and smaller half - mass sizes compare to the main - sequence ones .
the green valley objects ( recently quenched or in transition ) also have higher central density profiles compared to the main sequence ones , and their mass profiles are similar to the quenched ones , and this holds at all redshifts up to @xmath0 .
the half - mass size values of green valley galaxies are in between main - sequence and quenched ones .
we caveat that at @xmath1 , massive galaxies on the main - sequence and green valley ( with @xmath156 ) have similar mass density profiles indicating that these galaxies are potentially subject to quenching .
these results seem to indicate a high central density as a prerequisite for quenching .
however , @xcite have shown that there may be no direct causality link between central density and quenching , as both may be driven by a third factor , such as the galaxy mass .
still , we emphasize that the relatively empty region at low @xmath5 and low @xmath150ssfr implies that at least for our studied mass range , almost no galaxies can be quenched prior of reaching to a certain @xmath5 threshold , so there might always be a `` compaction '' phase prior to quenching .
if star - forming galaxies quench after reaching a required bulge mass density , then their mass profiles should become similar to that of the quenched galaxies at later times .
we will address this in next section , in addition to the mass and redshift dependence of @xmath5 thresholds .
the results of previous sections are tempting to indicate the necessity of bulge formation beforehand the secession of star - formation , but the causality links are difficult to establish .
it is possible that `` mass quenching '' process @xcite alone is able to establish a density - quenching correlation , simply because the first galaxies to ( mass ) quench are the most dense .
still , it remains important to first understand how galaxies build up their central regions and how the mass profiles of galaxies change once they start quenching by reaching to a certain central density limit .
secondly , what mechanism(s ) related to the bulge growth can play a role in preventing the gas to cool and for how long ?
we try to address these questions in the following .
in section 4 we have shown that at fixed mass , central mass concentration ( resembling bulges ) of star - forming galaxies increase with time by a factor of about @xmath157 between @xmath0 and @xmath133 ( depending on the total stellar mass ) .
star - forming galaxies can build and growth their bulges via different proposed mechanisms , including violent disk instabilities and mergers of giant clumps at high redshifts ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , weak disk instabilities a.k.a secular evolution @xcite and inflow of gas via mergers ( e.g. , * ? ?
* ; * ? ? ?
* ; * ? ? ?
these processes depends on the redshift , for instance due to the high gas fraction at high redshifts @xcite , formation of giant @xmath158 clumps by strong gravitational instabilities and hence coalescence of them to build bulges ( e.g. , * ? ? ?
* ; * ? ? ?
however , at low-@xmath19 weak instabilities of galaxies sub - components such as bars , can likely trigger formation / evolution of bulges ( see also recent reviews by * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and references therein . for simulated galaxies @xcite argued that oscillations about the main - sequence ridge are important , i.e. , when galaxies are above the main sequence , they form most of their stars in the bulge , while when then are below the main - sequence , they form most of stars in the disk .
the outcome is that the mass in the central and outer regions grow concurrently .
as shown in figure [ fig12 ] , there are some hints that lower mass star - forming galaxies form their bulges at later times than massive ones , concurrently with their disk .
star - forming galaxies with moderate - mass form their bulges about 2 - 3 gyr later , comparing to the massive ones .
therefore , viable mechanisms might also depend on total stellar mass of galaxies and hence the _ `` compaction '' _ phase ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) can be different processes at different redshift and mass .
in addition , we have also seen that at fixed mass , disks of star - forming galaxies grow with time in lockstep with bulges @xcite , hence indicating that the bulge formation mechanism(s ) , such as violent gas inflow induced by counter rotating streams and/or merging supported by violent disk instabilities presumably should be able to preserve the disks . note that in this work , we have compared similarly selected galaxies at fixed masses , therefore the true evolution of individual galaxies with different stellar masses needs to be understood .
nevertheless , bulges of star - forming galaxies grow with time up to a certain values ( stellar mass dependent ) and possibly retain their bulge masses roughly constant thereafter . )
quiescent galaxies at @xmath1 and trace the evolution of their central densities from @xmath0 to @xmath1 .
the central densities ( @xmath5 ) of these galaxies change as @xmath159 consistent with little to no evolution .
however , this method has limitations as many galaxies ending up at this massive end at @xmath1 ( see text).,scaledwidth=48.0% ] ( red symbols ) , compared to the prediction from eq . [ equation6 ] ( gray circles ) .
the green circles are again the observed surface densities , now using the same redshift binning as for the model.,scaledwidth=48.0% ] on the other hand , at fixed mass , quiescent galaxies at @xmath1 have slightly lower @xmath5 values compare to their high-@xmath19 counterparts ( see figure [ fig11 ] ) .
if quiescent galaxies grow inside - out , then it is expected that @xmath5 remain roughly constant , i.e. , galaxies central densities remain almost intact after they become quenched
. however , as pointed earlier the newly quenched galaxies could introduce biases and decrease the average @xmath5 values . to further examine this
, it is needed to trace back individual galaxy out to early times which is observerationally difficult .
statistically , it is proposed to use `` constant '' cumulative number density approach @xcite however , this method fails to consider different galaxy mergers and scatter in mass accretion histories ( see e.g. , * ? ? ?
@xcite takes into account these effects by using abundance matching method @xcite and cumulative number density evolution of massive galaxies to match progenitors and descendants .
if we use this later approach only for massive quiescent galaxies ( figure [ fig18 ] ) , we see little to no evolution of the central densities for quiescent galaxies ending up at @xmath160 quiescent galaxies at @xmath1 ( i.e. , @xmath161 ) . for that we have used @xcite ( @xmath1 ) and @xcite ( high-@xmath19 ) mass functions to estimate the cumulative number densities at different redshifts .
similar results also found by @xcite ( see also * ? ? ?
* ; * ? ? ?
however , this approach is still problematic as so many galaxies of different stellar masses can end up at @xmath1 with @xmath162 .
we use now a simple empirical model to interpret the observed evolution of the central density of quiescent galaxies ( @xmath163 ) .
for each pair of redshifts @xmath164 and @xmath165 ( @xmath166 ) , we write : @xmath167 where @xmath15 is the comoving number density of galaxies in the considered mass bin .
this assumes that the central density of quiescent galaxies at @xmath164 remain identical at @xmath165 and that the star - forming galaxies that quench between @xmath164 and @xmath165 maintain the central mass density they had at @xmath164 . in this way we model the effect of the so - called progenitor bias .
we have used the number density of quiescent and star - forming galaxies at different redshifts from @xcite and the result for our higher mass bin are shown in figure [ fig19 ] .
the predicted central density of quiescent galaxies is shown by gray circles and compared to the observational results as in figure [ fig11 ] ( red circles ) .
the green circles also shows the observed ones but with different redshift binning ( i.e. , binning similar to the ones used for the model ) .
any difference between the value calculated with this equation and the actual data could be ascribed to other effects , such as a structural evolution of individual galaxies . however , the predicted central densities agree quite well the observed ones , implying that the central regions of galaxies do not evolve much following quenching .
+ threshold values at fixed masses form high-@xmath19 to low-@xmath19 , using the median @xmath5 values of galaxies in green valley ( gray regions of figure [ fig15 ] ) . at fixed mass ,
the threshold is higher at earlier times.,scaledwidth=48.0% ] as shown in figure [ fig14 ] and [ fig15 ] , at fixed mass , the star formation of galaxies declines sharply around specific @xmath5 threshold .
this can suggest a `` knee - shaped '' or `` l - shaped '' evolutionary path on ssfr-@xmath5 or color-@xmath5 planes at all our studied redshifts and different mass bins ( as also suggested by * ? ? ?
the total stellar mass and central density of galaxies are correlated ( e.g. , * ? ? ?
in addition , simulations show that while the total stellar masses of star - forming galaxies increase , their central densities also grow ( via star - formation or redistribution of stellar masses ) till they saturate around specific central density ( see e.g. , * ? ? ? * ) .
therefore , at first sight this may suggest that the galaxies start to quench once they reach around a certain @xmath5 threshold .
this looks to be the same across all our studied redshifts , although the threshold values depend slightly on the redshifts ( higher values for high-@xmath19 galaxies ) .
this is consistent with early observational studies on galaxies mass profiles at low and intermediate redshifts @xcite , at higher redshifts @xcite and simulations @xcite ( see also * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . however , as mentioned earlier the links between cause and effects are difficult to establish .
we note that the advantage of this work compared to previous studies is to use psf - corrected mass - profiles and reducing the uncertainties due to @xmath168 and color - gradients in a consistent way from @xmath0 to @xmath169 .
to further understand this transition , we compare galaxy mass profiles . from figure [ fig12 ]
it can be seen that at @xmath170 star - forming galaxies ( with @xmath171 ) have already reached to their maximum central densities at fixed masses compared to their counterparts at high-@xmath19 .
therefore , if these galaxies will become quenched at later times , then their expected mass profiles should be comparable to the quiescent ones at @xmath1 with similar masses ( assuming galaxies will not experience major mergers which is less frequent at low redshifts ) .
in other words , assuming quiescent galaxies in the local universe once have been star - forming at high redshifts and they have gradually become quenched without any significant changes in their internal structures , then it is expected that their mass distributions become comparable to the local quenched galaxies at fixed masses ( inside - out quenching scenario , see also @xcite and @xcite ) .
we put this in a simple test and investigate it by using stellar population synthesis models of @xcite to predict their mass profiles at @xmath1 and compare them to local quiescent galaxies at fixed masses . for this purpose
, we used the best - fit parameters of the star - formation rate and the age derived from sed fitting , in addition to the stellar population model created with the same parameters . using the table of the age and stellar mass from the stellar population model , the expected stellar mass at an specific time ( here @xmath1 )
are estimated .
the comparison between predicted mass - profiles at @xmath1 of high-@xmath19 star - forming galaxies ( @xmath172 ) are shown in figure [ fig20 ] , divided into three mass bins , decreasing from left to right . for most massive bins ( left and middle panels ) ,
the expected mass profiles are very similar to local quiescent ones , at all radii . for lowest mass bin ( right panel )
, the predicted mass profile of star - forming ones is slightly extended and less compact as the quiescent galaxies at low-@xmath19 .
perhaps , additional process(es ) required to redistribute stellar masses or assemble masses in their central regions .
note that these less massive galaxies build their central densities later , hence using a lower redshift samples would help better for this test . in figure [ fig21 ]
, we show the same but for the star - forming galaxies at @xmath173 .
the @xmath1 expected profiles of most massive star - forming galaxies at @xmath173 is very similar to the quiescent ones in the local universe .
these massive star - forming galaxies have reached their relative maximum central densities by @xmath174 .
this shows indeed that stellar mass and central density correlate , hence it is difficult to disentangle whether it is mass or density ( or even the halo mass ) that control the quenching .
however , for intermediate mass bins ( middle and right panels ) , which have not reached to their maximum central densities by @xmath173 , their expected profiles differs from the low-@xmath19 quiescent ones .
they have a more extended and less concentrated mass profiles .
note that objects in the low mass bins might change their `` mass bin '' before they quench .
this test recalls that star - forming galaxies at this high-@xmath19 need to grow their central densities and reach to a specific @xmath5 threshold value , prior to their quenching .
the specific threshold of @xmath5 values depends on stellar mass and redshifts .
it can be defined approximately by the median values of galaxies in green valleys ( gray shaded region of figure [ fig15 ] ) for each stellar masses and is shown in figure [ fig22 ] .
the specific threshold values increase with redshift at fixed mass , implying higher central density threshold requirements for galaxies at high-@xmath19 to become quenched .
@xcite showed similar evolution for the effective surface density threshold and pointed the reason for this evolution could be due to the higher specific star - formation rate at higher redshift .
it worth noting that central density strongly correlated with central velocity dispersion ( @xmath175 ) , as derived by @xcite from virial theorem .
it has also pointed by @xcite that central velocity dispersion is better correlated with color of galaxies .
therefore , possibly @xmath175-ssfr might have similar or tighter correlation than @xmath5 .
this needs to be examined observationally for a large redshift range , though it is very expensive .
in this paper we have derived psf - corrected stellar mass profiles of 2391 mass - complete sample of @xmath176 galaxies up to @xmath0 from 3d - hst catalog and randomly selected of @xmath177 mass - matched @xmath176 galaxies at @xmath1 .
we examined the half - mass radii of star - forming and quiescent galaxies in small stellar mass bins .
we also examined the relation between stellar mass structural parameters and star - formation activity .
we find that : * half - mass radii of all types of galaxies are on average smaller that their half - light radii ( by @xmath178 ) with a weak dependence of stellar masses .
* at fixed mass , the average half - mass size of quiescent galaxies increase by a factor of 4 since @xmath0 in a similar pace to their half - light radii ( @xmath179 for massive quiescent galaxies ) .
mass - weighted sizes of star - forming galaxies change very little at fixed mass ( @xmath180 for massive star - forming galaxies ) .
* star - forming galaxies build up their central densities with cosmic time concurrently with their outer regions .
quiescent galaxies build up most of the stellar mass in their outskirts .
* stellar mass central density within 1 kpc of galaxies ( @xmath5 ) is better correlated with ssfr than effective surface density at all studied redshift . *
galaxies follow a `` knee - shaped '' path on the @xmath5-@xmath150ssfr plane at all masses and independent of redshift , i.e. , galaxies need to build central densities before quenching .
threshold @xmath5 depends on redshift and stellar mass .
* we emphasize that the correlation between @xmath5 and ssfr does not imply a causal link .
we thank the anonymous referee for helpful comments and suggestions .
this work is based on observations taken by the 3d - hst treasury program ( go 12177 and 12328 ) with the nasa / esa hst , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , university of cambridge , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
using difference star - formation history ( sfh ) at low and high redshift could introduce systematic effects on the derived parameters .
as discussed in the text , we have assumed exponentially declining sfh for high redshift sample .
however , at low redshift we assumed stochastic burst in addition to the declining sfh . assuming
additional burst could introduce systematic in total stellar masses @xcite . to examine
whether our results depend on these choices , we repeat analysis at high redshift assuming similar sfh at low-@xmath19 , using isedfit code .
we remeasure half - mass radii using the new stellar mass profiles and show their evolution at fixes masses for quiescent and star - forming galaxies in left and right panel of figure [ figa1 ] , respectively .
it can be seen that our results do not depends on assumption of different sfh .
the complexity of the light profiles of galaxies due to sub - structural components ( e.g. , inner disks , clumpy features , etc . ) could have introduced deviations of galaxies surface brightness profiles from the best - fit single - srsic profiles and consequently on the derived stellar mass densities . to test whether this could have affected the results of this work , we follow @xcite and apply the same technique to derive the corrected surface brightness profiles in different filters .
the de - convolved profiles of galaxies are derived by fitting psf convolved single - srsic models to the observed two dimensional surface brightness images of galaxies .
we then derived the residual profile of galaxies , taking into account the geometry and position of the best - fit model and added this to the best - fit de - convolved srsic model . where the uncertainties of the sky background dominate , we extrapolated the profiles at the larger radii , using a srsic model profile .
the stellar mass surface density was then derived from these corrected surface brightness profiles in the same way as described in section 3.2 .
the right and left panels of figure [ figa2 ] show the evolution with redshift of half - mass radii of star - forming and quiescent galaxies , respectively .
a comparison with the left panels of figures [ fig9 ] and [ fig10 ] shows that the rate of evolution of half - mass radii of both quiescent and star - forming galaxies are consistent with the results based on the best - fit single - srsic models .
this indicates that independent of the method for deriving the stellar mass profiles , the half - mass radii of quiescent galaxies at fixed masses increase with time .
similarly , for the star - forming galaxies , the rate of evolution of the half - mass at fixed masses is slower than for quiescent galaxies . in figure [ figa3 ] ,
we show in detail the evolution of half - mass radii of star - forming galaxies using residual - corrected method .
the grey points show the mass - weighted sizes for individual galaxies and the color symbols show the medians in each redshift bin .
the best - fit single srsic models are compared in 1d and 2d to the observed light profiles of few galaxies in figure [ figa4 ] .
we conclude that , with the current resolution of hst images , the results of this paper are robust when using the stellar mass profiles derived from single - srsic models . | we study the history from @xmath0 to @xmath1 of the stellar mass assembly of quiescent and star - forming galaxies in a spatially resolved fashion . for this purpose
we use multi - wavelength imaging data from the hubble space telescope ( hst ) over the goods fields and the sloan digital sky survey ( sdss ) for the local population .
we present the radial stellar mass surface density profiles of galaxies with @xmath2 , corrected for mass - to - light ratio ( @xmath3 ) variations , and derive the half - mass radius ( @xmath4 ) , central stellar mass surface density within 1 kpc ( @xmath5 ) and surface density at @xmath4 ( @xmath6 ) for star - forming and quiescent galaxies and study their evolution with redshift . at fixed stellar mass
, the _ half - mass sizes _ of quiescent galaxies increase from @xmath0 to @xmath1 by a factor of @xmath7 , whereas the half - mass sizes of star - forming galaxies increase only slightly , by a factor of @xmath8 .
the central densities @xmath5 of quiescent galaxies decline slightly ( by a factor of @xmath9 ) from @xmath0 to @xmath1 , while for star - forming galaxies @xmath5 increases with time , at fixed mass .
we show that the central density @xmath5 has a tighter correlation with specific star - formation rate ( ssfr ) than @xmath6 and for all masses and redshifts galaxies with higher central density are more prone to be quenched . reaching a high central density ( @xmath10 )
seems to be a prerequisite for the cessation of star formation , though a causal link between high @xmath5 and quenching is difficult to prove and their correlation can have a different origin . + |
Subsets and Splits