article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
ice crystals with a variety of shape , size and mass are present in clouds .
the properties of these crystals are markedly dependent on the temperature and other properties of the atmosphere@xcite .
a classification of ice crystals with a description of crystal shapes , size and mass can be found in @xcite .
certain atmospheric and cloud behaviors are characterized by parameters related to the ice particle dynamics for different shapes and sizes @xcite .
a precise estimation of ice crystals falling velocity is required to quantitatively determine their evolution in the atmosphere .
the knowledge of the falling velocity is necessary for the simulation of ice water paths and for the determination of cloud boundaries @xcite .
also it is used for the study of microphysical process in clouds and for climate modeling @xcite . the need for more accurate theoretical models to simulate clouds
has increased the requirement of more detailed measurements of relationships between the settling velocity , masses , and dimensions for a large spectrum of ice crystal types .
a precise determination of these relationships allow us to obtain accurate parameterizations of the settling velocity of cloud particles .
these parametrizations are essential to have an accurate simulation of cloud in general circulation models ( gcms ) of precipitation amount , cloud dissipation and cloud optical properties @xcite .
although there have been many proposals in literature , the ice crystal sedimentation in the atmosphere has not been completely characterized @xcite .
there are analytical solutions that precisely determine the sedimentation falling velocity for spheres particles . due to the large variety of shapes , sizes and masses of ice crystals , and the range of reynolds numbers involved in these problems , there is no precise analytical estimation to predict the falling velocity for shapes other than spheres .
many works in literature @xcite provide schemes to parameterize the ice crystal masses , shapes and size to predict the settling velocity .
ice particle terminal velocities are often calculated theoretically or experimentally by determining a relationship between the reynolds number ( @xmath1 ) , and the best ( or davis ) number , ( @xmath2 ) @xcite .
there are a number of experimental works in which the most important variables are measured .
terminal velocity , mass and size have been measured for various ice particle types .
these datasets are obtained from laboratory measurement and observations of real ice particles falling through the atmosphere .
some well known experimental datasets can be found in @xcite .
the proposals by @xcite have shown quite good approximations to the falling speed of ice particles for @xmath3 .
these proposals proved to be in good agreement with experimental data for a variety of particle types .
however @xcite showed that for viscous flow regimes ( @xmath4 ) these formulations overestimate the crystal falling velocity .
@xcite , using the approximation of @xcite with results from @xcite , give an estimate for the sedimentation rate of small ice crystals whose maximum dimension is smaller than @xmath5 m .
this estimate for columnar ice crystals is in agreement with most experimental data ( within 20% ) . a complete review of the main theoretical approximation that have been proposed and many experimental results can be found in @xcite ; also a lots of relevant references are presented in these works .
the sedimentation of an ice crystal in the atmosphere is a fluid mechanical problem that can be modeled as a rigid body moving immersed in a fluid flow .
this rigid body moves under the action of its own weight , buoyancy force and interacting with others crystals and with the fluid that surrounds it . given a characterization for the shape , size and mass density of the crystals , together with its atmospheric habitat
, it is possible to completely determine the dynamical behavior of the crystals by using some adequate computational fluid dynamic ( cfd ) method .
having an accurate numerical method to solve these problems allows us in turn to test and to improve the parametrization laws , and to compute the settling velocities for sizes , shapes and masses for which experimental data are not available .
also , the sensitivity of the problem to different parameters can be studied numerically . to the knowledge of the author ,
there are no numerical results that describe appropriately the dynamics of ice crystals in the atmosphere .
the lbm is a cfd method that proved to be successful to treat multiple problems involving both compressible and incompressible flows on simple and complex geometrical settings .
in particular the lbm provide a simple way for treating accurately the flow surrounding an immersed body , in arbitrary movement , with no regular geometry .
for a complete modern review of this topic see @xcite .
the behavior of particles in sedimentation have been analyzed using lbm in a variety of problems @xcite . in this paper
we use lbm to study the dynamical behavior of ice crystals with two different shapes : simple columns and six bullet policrystal ( a combination of columns or bullets ) .
in particular , the ice crystal settling velocity is obtained numerically for both shapes in a range of masses and sizes ( characteristic lengths of @xmath6 ) .
the lbm results for the fluid mechanical problems are obtained in a pure viscous regime ( @xmath4 ) .
this is the flow regime of the smallest particles falling in a cloud .
the accuracy in the lbm to treat this problem is evaluated by comparison with some well known experimental data in literature @xcite , and with theoretical proposal from @xcite and @xcite . in order to prove the correctness and accuracy of the proposed lbm algorithm , two fluid mechanical problems at low reynolds number
were tested .
the idea of these benchmarks is to test our algorithm in translational and rotational problems , within the @xmath1 regime that include the ice crystal sedimentation problems .
the paper is organized as follows : in section [ sec : lbm ] we present the basic equations of the lbm , introduce notation and some details about the boundary conditions methods , force evaluation , and grid refinement techniques . in section [ sec : lbm_benchmarks ] we evaluate the correctness of the proposed lbm algorithm to solve two benchmarks , _ sphere sedimentation in a square duct _
( section [ subsec : spheresedimentation ] ) and _ pure rotation of rigid bodies in couette flow _ ( section [ subsec : couetteflow ] ) , well known in the literature . in section [ sec : iceparticlesedimentation ] the sedimentation of ice crystals in the atmosphere is solved using lbm .
numerical results for _ columnar ice crystals _ and _ six bullet - rosette ice policrystals _ are showed in sections [ subsec : columnariceparticles ] and [ subsec : rosetteiceparticles ] respectively . in section [ sec : conclusionanddiscussion ] conclusions and discussions are presented .
in this section we present the basic equations of the lbm , introduce notation and the main concepts we use along the paper .
in addition to the lattice boltzmann equation that govern the physics of the bulk fluid ; one needs to prescribe a method to apply boundary conditions , to evaluate the fluid force on a body and to implement grid refinement where necessary . in the next sections we briefly review these topics .
the numerical results in this paper are obtained by solving the lattice boltzmann equation ( _ lbe _ ) @xcite , a particular phase - space and temporal discretization of the boltzmann equation ( _ be _ ) @xcite . the be governs the time evolution of the single - particle distribution function @xmath7 where @xmath8 and @xmath9 are the position and velocity of the particle in phase space .
the lbe is a discretized version of the be , where @xmath8 takes values on a uniform grid ( the lattice ) , and @xmath9 is not only discretized , but also restricted to a finite number @xmath10 ( the number of discrete velocities in the model ) of values @xcite . in an isothermal situation and in the absence of external forces , like gravity
, the lbe can be written as : @xmath11 here @xmath12 is the @xmath13-th component of the discretized distribution function @xmath14 at the lattice site @xmath15 time @xmath16 and discrete velocity @xmath17 .
the function @xmath18 is a discrete version of the equilibrium distribution function @xcite and @xmath19 is a linearized collision operator . in our simulations we use a simple relaxation time model ( _ srt _ ) , a simplified approximation of @xmath19 that follows from the bhatnagar , gross , and krook ( _ bgk _ )
@xcite approximation with relaxation time @xmath20 . in compressible - flow models
the lattice constant @xmath21 that separate two nearest neighbor nodes , and the time step @xmath22 are related with the speed of sound @xmath23 by @xmath24 . in incompressible - flow models ,
the same relation between @xmath25 and @xmath22 holds , but the constant @xmath26 is no longer related to the speed of sound .
the coordinates of a lattice node are denoted by @xmath27 , where the integer multi index @xmath28 ( or , @xmath29 in the two - dimensional case ) denotes a particular site in the lattice .
the macroscopic quantities such as the fluid mass density @xmath30 and velocity @xmath31 , are obtained , in boltzmann theory , as marginal distributions of @xmath32 and @xmath33 when integrating over @xmath9 . in lbm
this integrals are approximated by proper quadratures .
specific values of @xmath34 s and @xmath35 s , @xmath36 are made so that these quadratures give exact results for the @xmath9-moments of order 0 , 1 and 2 @xcite .
we have @xmath37 and @xmath38 in the simulations we present in this paper , we are interested in incompressible flow problems , where we modify eq .
[ eq : lbmmacroscopics_1 ] according to the quasi - incompressible approximation presented in @xcite . in this approximation @xmath39
is replaced by @xmath40 , a constant fluid mass density .
a single time step of the discrete evolution equation [ eq : lbe ] is frequently written as a two - stage process @xmath41 and @xmath42 the computation of @xmath43 on the whole lattice , eq .
[ eq : collision ] , is called the _ collision step _ , while the computation of @xmath44 at @xmath45 , eq .
[ eq : streaming ] , on the whole lattice is called _ streaming step .
_ we refer to the lattice boltzmann models with the standard notation @xmath46 , where @xmath47 is the number of space dimensions of the problem , and @xmath48 is the number of discrete velocities .
also we add an @xmath13 at the end to indicate that we use a quasi - incompressible approximation , @xmath49 .
the problems we are interested in are those in which rigid bodies move inside an unbounded fluid domain . because the impossibility to model an infinite fluid domain
, we have to restrict the problem to a finite computational fluid domain .
the size of the computational fluid domain has to be a compromise between minimizing the computational work
the smaller the size the better , and minimizing the undesirable effect of the boundary conditions the larger the domain the better . the computational fluid domain is a block of fluid bounded by regular borders .
the rigid bodies that move inside the domain are described by geometries as required .
the flow in the interior of the domain is computed by solving the lbe .
close to the boundaries a special treatment is used so that the flow obeys the physical boundary conditions . in the present work ,
we use both dirichlet and outflow open - boundary conditions ( convective boundary conditions or sommerfeld like ) . the correct imposition of the boundary conditions on arbitrary boundary geometries , like the boundary of rigid bodies , has been one of the main issues in lbm . to impose dirichlet boundary conditions , for the velocity or pressure , on a regular boundary which is coincident with the grid ,
we use the method proposed in @xcite .
these regular boundaries are adequately represented by linked lattice nodes . to impose dirichlet conditions for the velocity on boundaries of arbitrary shape we use the method proposed by @xcite .
we also use the method shown in @xcite and obtain no significant differences in the results . in order to reduce the length of the numerical flow domain
we apply a convective , or sommerfeld like , boundary conditions on the open boundaries of the domain .
these boundary conditions allow us to model a long or quasi - infinite physical domain with a reduced computational domain .
these type of conditions have been extensively applied in computational fluid mechanics .
we make a particular implementation of convective boundary condition in the lbm context .
more details about our implementation are given in section [ subsec : lbm_cbc ] .
the outflow open - boundary conditions allow us to represent a long or quasi - infinite physical domain by a finite computational domain . in many fluid mechanical problems on a large , quasi - infinite fluid domain ,
the region in which the velocity gradients , viscosity and inertial forces are significant is rather small .
it is a challenge to adequately represent such a fluid mechanical problem on a bounded , as small as possible , computational domain in such a way that the interesting physics is captured accurately and at the same time the computational cost is reduced significantly .
strictly speaking , one needs to apply convenient boundary conditions at the finitely close computational boundaries so that these conditions do not affect the physical behavior of the flow in the interior or ruin the intrinsic lbm accuracy .
there are different approaches in the literature to treat the outflow open - boundary conditions in the lbm context .
we can divide these approaches in at least two categories , the ones based on mesoscopic variables @xcite and the ones based on macroscopic variables @xcite .
the latter are generally extensions of boundary conditions extensively applied in classical methods ( finite difference ( _ fd _ ) , finite volume ( _ fv _ ) and finite element ( _ fe _ ) methods ) of computational fluid mechanics ( _ cfd _ ) to solve the navier - stokes ( _ ns _ ) equations .
we are mainly interested in non - stationary quasi - incompressible problems . in the lbm context
the convective boundary condition ( cbc ) proposed in @xcite to treat outflow open - boundaries has shown acceptable results in these kind of problems .
also , in @xcite it is shown that cbc gives better results than neumann boundary conditions ( _ nbc _ ) when using fv methods to solve the incompressible ns equation .
the nbc are based on macroscopic variables , and were also tested in lbm @xcite .
the results presented in @xcite show that cbc are a better option than nbc in non - stationary problems .
these works show that nbc introduce undesirable perturbations in the fluid domain , specially in non - stationary problems . in our numerical tests we use cbc to fix the velocity in the outflow open - boundaries .
to complete the cbc method we use some dirichlet boundary condition method to fix the unknown mesoscopic ( @xmath50 ) variables that have to satisfy the macroscopic cbc .
the cbc is defined as : @xmath51\ ] ] where @xmath52 is a velocity or pressure function , and @xmath53 is a reference velocity that have to be defined adequately .
@xmath54 is the boundary of the fluid domain @xmath55 , and @xmath56 is the outward normal vector in @xmath54 . in our numerical implementation cbc are applied on straight , grid - coincident boundaries only . to show the implementation of the cbc , we consider as an example a boundary with normal vector @xmath57 .
given the lbe solution at time @xmath16 [ eq : cbcgeneral ] gives , for @xmath58 @xmath59 where we have approximated the space derivative by a second order accurate , backward difference operator and the time derivative by a first order accurate , forward difference operator .
from equation [ eq : u_incbcborder ] the velocity @xmath60 that satisfy the cbc can be determined .
knowing the boundary velocity at @xmath45 some dirichlet boundary condition can be used to determine the corresponding mesoscopic variables .
we implement the cbc method coupled with the proposal of @xcite to impose the boundary velocity determined from equation [ eq : u_incbcborder ] .
equation [ eq : u_incbcborder ] is incomplete until we provide a prescription for the value of @xmath61 .
various criteria can be found in the literature to do this .
@xcite sets this quantity equal to the mean velocity on the outflow open convective boundary .
@xcite sets @xmath53 equal to the upstream free velocity @xmath62 with some additional conditions to ensure mass conservation in the fluid domain .
mass conservation problems are also observed in the case of @xcite .
in our implementation we set @xmath53 as a constant : the upstream mean velocity .
we have observed , by performing typical benchmark tests , that mass is acceptably conserved by the resulting scheme .
it is of crucial importance , in many applications that involve moving bodies surrounded by a fluid flow , to have a good method or algorithm to compute the flow force and torque acting on the bodies . by good we mean a method that is simple to apply , that is accurate and fast , so as not to spoil the efficiency of the flow computing method .
the accuracy in the determination of the force and torque acting on a moving body directly affects the body s movement . for a review of lbm methods that involve flow force evaluation on suspended particles we refer to section 6 of @xcite and references therein . the classical way to compute forces , and
so torque , on submerged bodies is via the computation and integration of the stress tensor on the surface of the body . in lbm
the stress tensor is a local variable , its computation and extrapolation from the lattice to the surface is computationally expensive , which ruins the efficiency of the lbm .
however , this method is widely used in lbm @xcite . an standard method to evaluate forces on submerged bodies in lbm is the _ momentum exchange _ ( _ me _ ) , introduced firstly by @xcite in lbm applications . the me algorithm is specifically designed and adapted to lbm
; it is therefore more efficient than stress integration from the computational point of view .
some improvements to ladd proposal have been introduced in @xcite , and different approaches to improve the methods in problems with moving bodies were made @xcite . in this work ,
force and torque are evaluated by using the methods presented in @xcite .
the motion of each body is determined by solving the newton s equations of motion with rotations formulated using quaternions @xcite .
the forces acting over bodies are given by the fluid flow forces , weight and buoyancy forces .
to integrate in time we use euler forward numerical scheme , which is first order accurate as the lbm method itself .
many problems in fluid mechanics are such that big gradients of the fluid variables appear only in regions which are small as compared with the whole computational domain .
to resolve well the space variations of the fluid variables one needs a grid size which is small enough . in lbm
a simple lattice is a cartesian grid of equispaced nodes .
the distance between two nearest neighbor nodes , the grid size , is @xmath63 for a real problem , the computational domain is covered by an arrangement of grids
. this arrangement can be as simple as a unique lattice or block grid with a single size @xmath25 , or a complex arrangement of grids with different grid sizes . in a problem with more or less uniform space variations all over , a single block grid that covers the whole computational domain may be suitable . in a problem where high space variations occur in a small region , a small grid size needs to be used in that region . but using this small grid size on the whole computational domain would be a waste of computational effort .
the right thing to do is to use an arrangement of grids with different grid sizes .
the methods to integrate various grid blocks with different grid sizes into a single computational domain are known as grid refinement methods . in lbm
there are at least two grid refinement methods : _ multi - grid method _ ( mg ) @xcite and _ multi - domain method _
( md ) ( or multi - block ) @xcite . in the mg method , a grid block with small grid size is always superimposed to a grid block with larger grid size .
several layers of grids can be superimposed in this way . in md method
the grids with different grid sizes overlap just in a selected set of lattice nodes .
this overlapping occurs only on a small region with two adjacent grid blocks of different grid sizes , see @xcite . in this work
we use md methods .
we select this method because it has better numerical performance and lesser memory requirement than mg method .
a disadvantage of the md method , though , is that its implementation is more complex than that of mg where some additional grids are used as interface to interchange data between different levels of grid size .
we use a priori refinement method .
this means that we chose the arrangement of refined grids in the domain before solving the fluid problem , and this arrangement is fixed in time .
this last characteristic becomes the main numerical performance limitation of our method . in non - stationary fluid problems the smallest scales to solve and
its position within the fluid domain are not easy to predict .
a priori refinement methods are not the most efficient option in non - stationary fluid problems , one frequently needs to implement _ adaptive refinement methods _ @xcite .
this adaptive methods are able to change the refined grids arrangement at run time .
in this section we show lbm results for two fluid mechanical problems well known in the literature . our purpose is to test the correctness of the proposed lbm algorithm to solve problems at low reynolds numbers as those in ice crystals sedimentation .
the rigid body equations of movement are independently solved for translations and rotations , they are only coupled through the fluid forces .
therefore we chose two particular problems , with well known results , to test pure translations and pure rotations .
we are interested in the problem of an sphere sedimenting inside a square section duct full of a viscous fluid .
we are particularly interested in the determination of the sphere s terminal falling velocity @xmath64 in figure [ fig : schematicfallingsphereandresults]@xmath65 we show a scheme of the problem configuration .
we compute @xmath66 for different values of the relation @xmath67 between the sphere s diameter @xmath68 and the length @xmath69 of the square cross section .
we are looking for the relation @xmath70 , as a function of @xmath67 , where @xmath71 is the sphere s terminal falling velocity in an unbounded fluid domain .
the ratio @xmath70 is commonly named _
wall correction factor_. the terminal falling velocity is the stationary velocity that a body ( sphere ) reaches in a sedimentation process ; the body and viscous forces are equal in magnitude in this stationary regime .
@xmath72 is obtained from a stokes flow regime defined by @xmath73 with @xmath74 , @xmath75 the sphere s material density , @xmath76 the fluid s density and @xmath77 the fluid s kinematic viscosity .
the value of @xmath71 given by equation [ eq : terminalvelocity ] is an adequate estimation if the fluid flow is consistent with stoke problem hypothesis @xcite .
the sphere sedimenting in a square section duct is a problem analyzed in the literature .
@xcite numerically determine the wall correction factor using the lbm , and @xcite experimentally analyze this problem .
@xcite evaluate the wall correction factor for different configurations and compare their results with the experimental results presented in @xcite .
@xcite use a quasi - incompressible lbm and some particular boundary conditions and forces evaluation method .
@xcite experimentally study the wall interference effects over the sphere terminal velocity .
they test the wall correction factor for different sphere diameters and duct cross section configuration . as a main result
they show a curve fitting to the obtained experimental data in those problems .
we will use this fitting equation to compare with our lbm results . in figure [
fig : schematicfallingsphereandresults]@xmath65 we show an schematic arrangement of grids refinement and the sphere placed in its initial position at @xmath78s .
the refinement region is placed from @xmath79 to @xmath80 , a bounded fluid region where will take place the rigid body displacement .
we use a computational domain of length @xmath81 to represent a quasi - infinite physical one .
we made several numerical test for different domain configurations .
we look for the best combination of length domains ( @xmath81,@xmath82 , and @xmath83 ) and sphere placement ( @xmath84 ) that produce acceptable results and minimize the computational cost .
we refer as acceptable results , those results uninfluenced by the computational domain length @xmath81 . from the numerical test we found that acceptable results require @xmath85 . in figure
[ fig : schematicfallingsphereandresults]@xmath86 we show the obtained lbm numerical results along with the polynomial fitting from the experimental data presented in @xcite .
we use in our computations @xmath87 m and @xmath88 .
the refinement region with length @xmath89 is placed at @xmath90 from the bottom wall .
the results showed in figure [ fig : schematicfallingsphereandresults]@xmath86 are obtained using the @xmath91 lattice boltzmann model with a domain discretization of @xmath92 for the fine grid .
dirichlet boundary conditions are set over all the domain s walls . in the horizontal top wall ( see figure [ fig : schematicfallingsphereandresults]@xmath65 ) the boundary velocity is set according to the convective boundary condition presented in section [ subsec : lbm_cbc ] . over the sphere surface we set the velocity boundary condition using methods for non regular boundaries as presented in @xcite .
we evaluate the wall correction factor for eight relations @xmath67 ranging from @xmath93 to @xmath94 , with diameters @xmath95 mm to @xmath96 mm respectively . the sphere is initially placed in the cross sectional center at @xmath97 from the bottom wall domain . as initial condition we set an equal and homogeneous velocity @xmath98 in the fluid flow domain .
the numerical tests consist in releasing a sphere in a particular way , as explained below , from its initial position @xmath84 with velocity @xmath98 at @xmath78s .
the initial velocity @xmath98 is approximately equal to and opposite in direction to the terminal falling velocity @xmath66 of the rigid body inside the square duct .
there are at least two main reasons to analyze the problem in a constant velocity frame . on the one hand
, being the rigid body at an approximately constant position , relative to the computational fluid domain , reduces the region of refinement and simplifies the refinement algorithm , since we can keep the refinement region fixed ( no dynamical refinement process is necessary ) . on the other hand , the constant velocity fluid domain with adequate boundary conditions allows us to reduce the length of the domain .
the rigid body is roughly static in the computational fluid domain length and we can fix the minimal appropriate relation @xmath99 so that the rigid body stays in the refined region .
we apply a particular initialization procedure so that the body stays in a prefixed gap ( refined region ) .
we initially impose a fictitious external force during some time interval so that the relative velocity between the rigid body and the fluid domain is approximately @xmath66 .
after this time interval the rigid body is left to move freely in the fluid domain . because the application of the artificial initial force there is a fictitious transient which we are not interested in . once the sphere reaches the stationary regime ( after @xmath100 ) , the terminal falling velocity @xmath66 is considered equal to the sphere stationary falling velocity .
is the length of the refined region starting at height @xmath82 as measured from the bottom domain wall .
b ) wall correction factor @xmath70 vs. @xmath67 for a sphere falling inside a square cross section duct .
the continuous line is the polynomial fitting of experimental data from @xcite.,title="fig : " ] is the length of the refined region starting at height @xmath82 as measured from the bottom domain wall .
b ) wall correction factor @xmath70 vs. @xmath67 for a sphere falling inside a square cross section duct .
the continuous line is the polynomial fitting of experimental data from @xcite.,title="fig : " ] the reynolds number ( @xmath1 ) in our tests are in the range @xmath101 , while the @xmath1 range in @xcite and @xcite are @xmath102 and @xmath103 respectively .
the results we have obtained , presented in figure [ fig : schematicfallingsphereandresults]@xmath86 , have an acceptable correlation with those experimentally obtained in @xcite .
our results are close to the experimentally adjusted curve in the whole @xmath67-range .
the largest differences between experimental and lbm results are observed for @xmath104 which is particularly noticed for spheres of small radii .
the reason for these differences can be that the discretization is not fine enough , since the domain discretization is kept constant for all @xmath67 relations .
our numerical results show a better approximation to the experimental results than those presented in @xcite for the same problem .
the pure translation assumption is justified by the angular displacement results .
if flow instabilities are present , like typically those from high reynolds number problems , this statement is not longer true . in this section we want to test our lbm algorithm for describing pure rotation of rigid bodies immersed in a fluid flow .
particularly , two rigid bodies ( 2d and 3d ) immersed in a couette flow are analyzed .
our main interest is to benchmark our lbm results with those presented by @xcite and @xcite .
we test two rigid body configurations : an ellipse and an ellipsoid as we show in the problem scheme in figure [ fig : esquemaellipsoid ] .
the rigid body geometry is defined by : @xmath105 where @xmath106 , @xmath107 and @xmath26 are the three principal semi - axis length of the rigid body , and @xmath108 are the cartesian coordinates in a frame fixed to the body . the ellipse description from equation [ eq : ellipsoidalequation ] is obtained for @xmath109 .
the rigid body spatial orientation is defined by the euler angles @xmath110 and @xmath111 following the @xmath112 rotational order .
: free rotational ellipsoid in a @xmath113 couette flow , the fluid domain is a rectangular prism with side length @xmath114 and @xmath115 .
we show the case @xmath116 .
scheme @xmath86 : free rotational ellipse in a @xmath117 couette flow in a rectangular domain with length @xmath69 and @xmath81 . ]
the fluid domain in the ellipsoidal @xmath113 test is a rectangular prism with @xmath69 , @xmath81 and @xmath115 side lengths .
the ellipse @xmath117 test has a rectangular fluid domain with @xmath69 and @xmath81 side lengths as shown in figure [ fig : esquemaellipsoid ] . the rigid body and the fluid domain centroids are initially coincident .
the fluid and rigid body mass density are set equal with no resultant buoyancy forces . in two parallel walls or edges
we impose a uniform velocity @xmath118 as shown in figure [ fig : esquemaellipsoid ] . on all the remaining walls or edges ,
periodic boundary conditions are applied . the dirichlet velocity boundary conditions in non regular geometries are imposed by @xcite method .
the rigid body dynamics in a pure rotational problem is strongly influenced by the reynolds number problem defined as @xcite : @xmath119 where @xmath120 is the shear ratio and @xmath106 is the major rigid body semi - axis .
some cautions should be taken when these results are compared with those in literature , because there are different definitions for @xmath121 .
the dynamic of a rigid body immersed in a couette flow has , as has been shown by @xcite , an analytical solution in the limit @xmath122 .
the rigid body rotational velocity @xmath123 about @xmath124 axis is determined from @xcite by : @xmath125 where the rotational angle @xmath111 is : @xmath126 this solution , valid to @xmath117 and @xmath113 ( ellipse and ellipsoid ) rotational problems in the limit @xmath127 help us validating our lbm numerical results , which we present in the next sections . in this test case we analyze the rotational behavior of an ellipse immersed in a couette flow . in figure
[ fig : esquemaellipsoid]@xmath86 we show a schematic configuration of the ellipse and the fluid domain we use .
we test the problem at two reynolds numbers @xmath128 and @xmath129 and evaluate the results convergence to the problem s analytical solution presented in @xcite .
the obtained results are presented in figure [ fig : resellipseellipsoid]@xmath65 along with the jeffery analytical solution ( equation [ eq : angvelanaliticajeffery ] ) . in fractions of shear ratio @xmath121 at low reynolds numbers .
figure @xmath65 : ellipse rotational angle @xmath130 as function of @xmath131 at @xmath128 and @xmath129 compared with the @xcite analytical solution at @xmath132 .
the problem configuration is showed schematically in figure [ fig : esquemaellipsoid]@xmath86 .
figure @xmath86 : ellipsoid rotational angle @xmath130 as function of @xmath131 at @xmath133 and @xmath134 compared with the @xcite analytical solution at @xmath132 .
the problem configuration is showed schematically in figure [ fig : esquemaellipsoid]@xmath65.,title="fig : " ] in fractions of shear ratio @xmath121 at low reynolds numbers .
figure @xmath65 : ellipse rotational angle @xmath130 as function of @xmath131 at @xmath128 and @xmath129 compared with the @xcite analytical solution at @xmath132 .
the problem configuration is showed schematically in figure [ fig : esquemaellipsoid]@xmath86 .
figure @xmath86 : ellipsoid rotational angle @xmath130 as function of @xmath131 at @xmath133 and @xmath134 compared with the @xcite analytical solution at @xmath132 .
the problem configuration is showed schematically in figure [ fig : esquemaellipsoid]@xmath65.,title="fig : " ] the presented results are obtained with parameters @xmath135 m and @xmath136 m , and rigid body geometry defined by @xmath137 m , and @xmath138 .
we impose a dirichlet velocity boundary condition at the walls , where we set @xmath139 m/s as we show in figure [ fig : esquemaellipsoid ] . to achieve the desired reynolds number we choose an appropriate kinematic viscosity @xmath77 . for the test at @xmath128 we set @xmath140 and a domain discretization of @xmath141 lattice grid points ; while for the test at @xmath129 we set @xmath142 with a domain discretization of @xmath143 .
the fluid flow problems are solved using a @xmath144 lbm with simple relaxation time ( srt ) .
the results presented in figure [ fig : resellipseellipsoid]@xmath65 show convergence to the analytical solution when @xmath132 .
it is possible to observe , from figure [ fig : resellipseellipsoid]@xmath65 , a highly sensitive dynamical rigid body behavior when the @xmath145 number increase . for the @xmath129
the results show considerable changes with respect to the results at @xmath128 .
in particular there is an increase in the rotational period and a decrease in the lower rotational velocity for @xmath129 in comparison with those at @xmath128 . in both problems
we observe a periodic dynamical behavior as expected .
the obtained results for the rotational behavior of an ellipse in couette flow are thus in agreement with those presented in @xcite which shows the correctness of our algorithm to solve 2d pure rotational problems .
we study here the rotational behavior of an ellipsoid immersed in a couette flow , and compare the results with the analytical solution of @xcite .
figure [ fig : esquemaellipsoid]@xmath65 shows a scheme of the geometry and fluid domain configuration for this problem .
our lbm results are showed in figure [ fig : resellipseellipsoid]@xmath86 , obtained for reynolds numbers @xmath146 and @xmath134 , along with the jeffery s analytical solution ( see equation [ eq : angvelanaliticajeffery ] ) .
the parameters in our computations are @xmath147 m ( see figure [ fig : esquemaellipsoid]@xmath65 ) .
the ellipsoidal geometry is defined in equation [ eq : ellipsoidalequation ] with @xmath148 m , @xmath149 , and the initial rigid body orientation is @xmath150 .
the fluid domain is discretized by @xmath151 grid lattice points .
we impose a constant velocity @xmath152 in the walls parallel to planes @xmath153 and @xmath154 as we shown in figure [ fig : esquemaellipsoid]@xmath65 .
the kinematic viscosity is set to get the desired reynolds numbers , and we use the @xmath155 simple relaxation time lattice boltzmann model .
we also test the problem using a @xmath91 lattice boltzmann model and we do not find appreciable differences with the results obtained using @xmath155 . from the results in figure [ fig : resellipseellipsoid]@xmath86 we can see that the lbm results converge to the @xcite analytical solution when @xmath132 .
our results for the rotational behavior of an ellipsoid in couette flow agree with those presented in @xcite ; which shows the correctness of our algorithm to solve 3d pure rotational problems .
in this section we study the main problem of this paper ; we solve and analyze the dynamics of ice crystals sedimentation in the atmosphere .
in particular , we are interested in evaluate the ice crystal settling velocity by using lbm for different ice crystal shapes in a range of size and mass .
the sedimentation of an ice crystal in the atmosphere is a fluid mechanical problem we model as follows .
the crystal is considered a rigid body that moves under the action of its own weight , the buoyancy force and interacting only with the fluid that surrounds it .
a simplifying assumption is adopted : no interactions between rigid bodies is considered .
we are only interested in the dynamics of isolated rigid bodies in the atmosphere .
this assumption is a good approximation to the movement of ice crystals in a cloud , since the concentration of ice particles in cirrus typically ranges between 50 and 500 @xmath156 , while the maximum ice particle concentrations in cumulonimbus clouds reached 300 @xmath156 @xcite .
the dynamics of ice crystals of two shapes is analyzed , simple columns and six bullet - rosette policrystals . in both cases
the settling velocity is determined for several characteristic lengths .
the obtained lbm results are compared with some well known experimental data in literature @xcite , as much as with the theoretical proposals from @xcite and @xcite . with the intention to be clear , in the next sections we present
separately the obtained lbm results for both geometrical configurations . in this section
we present the settling velocity @xmath157 results for sedimentation of columnar ice crystals in atmosphere obtained by using lbm .
we compare these results with those obtained by experimental and theoretical methods well known in the literature .
columnar ice crystals with quasi - hexagonal cross section and needle ice crystals , typically grown at temperatures between @xmath158 and @xmath159 , have also been observed below @xmath160 @xcite . in our simulations
the ice crystals are modeled by columns of hexagonal cross section , and the sedimentation studied in fluid flow regimes with @xmath161 .
this is approximately the flow regime where the smallest particle can fall in a cloud .
we adopt @xmath162 as the length reference to evaluate the reynolds number , where @xmath163 is the ice crystal length and @xmath106 is the semi - length of its cross section .
we perform numerical tests for a variety of aspect ratios @xmath164 $ ] .
the length of the columnar ice crystals analyzed are in the range @xmath165 .
the fluid properties are set as those at temperature @xmath166 with an atmospheric pressure @xmath167pa .
we have selected these fluid properties to simulate the atmospheric laboratory conditions used in @xcite .
the ice density for columns are set in the range reported by @xcite for @xmath166 .
it is also possible to obtain the ice crystal mass from relationships like those shape based proposed by @xcite . in the figures [ fig : rescolumnarexperimental_ar=1_2 ] and [ fig : rescolumnarexperimental_ar=2_3 ]
we show the settling velocity we obtained using lbm in comparison with the laboratory experimental results presented by @xcite as a function of crystals length . in the figure [ fig : rescolumnarexperimental_ar=1_2 ] ,
lbm and experimental results are presented for aspect ratios between 1 and 2 ; while in the figure [ fig : rescolumnarexperimental_ar=2_3 ] comparative results for aspect ratios between 2 and 3 are showed .
the presented lbm results were obtained for hexagonal columns in a variety of orientations and with different ice densities chosen from @xcite for @xmath166 . for columnar ice crystals as a function of their length @xmath163 .
the showed results correspond to @xmath168 $ ] .
velocity is expressed in centimeters per second @xmath169 , space in micrometers @xmath170 .
the lbm results are showed with squares ( @xmath171 ) .
the experimental results are represented with : triangles ( @xmath172 ) for data from @xcite , hollow circles ( @xmath173 ) for @xcite data , and with filled circles ( @xmath174 ) for @xcite data . ] for columnar ice crystals as a function of the length @xmath163 .
the showed results correspond to @xmath175 $ ] .
velocity is expressed in centimeters per second @xmath169 , space in micrometers @xmath170 .
the lbm results are showed with squares ( @xmath171 ) .
the experimental results are represented with : triangles ( @xmath172 ) for data from @xcite , and with filled circles ( @xmath174 ) for @xcite data . ] as can be observed in the figures [ fig : rescolumnarexperimental_ar=1_2 ] and [ fig : rescolumnarexperimental_ar=2_3 ] , the lbm results are in accordance with laboratory measurements .
the results from @xcite present some dispersion as expected for a set of experimental data . for the length range and
aspect ratio analyzed , all the numerical results are included in the data dispersion . in the figure
[ fig : results_inc ] the settling velocities obtained by lbm are shown in comparison with laboratory measurements presented in @xcite as a function of crystals capacitance @xmath176 .
this parameter depends on the particle geometry and is obtained in @xcite for different geometries . for hexagonal columns
the capacitance is : @xmath177 the numerical results are in the regime which allow us to compare with the measurements presented in @xcite , where they choose @xmath176 as the characteristic length to evaluate the reynolds number .
the reynolds number regime of the lbm results is @xmath178 if we take @xmath176 as the characteristic length . for columnar ice crystals as a function of the capacitance @xmath176 .
velocity is expressed in centimeters per second @xmath169 and distances in micrometers @xmath170 .
the lbm results are showed with squares ( @xmath171 ) .
the experimental data from @xcite are represented as : triangles @xmath172 , for aspect ratios between 1 and 2 , and diamonds @xmath179 for aspect ratios between 2 and 3 . ]
it is possible to observe from the figure [ fig : results_inc ] that the dispersion of the lbm results have a notable decrease when the capacitance is used as variable .
the same observation was pointed out by @xcite for the experimental results .
we can also observe from figure [ fig : results_inc ] that the results obtained with lbm are not unifomly distributed within the region containing experimental data .
a bias towards the low part in this region can be seen .
a bigger amount of computation would allow us to adjust a curve to the lbm results , though this is out of the scope of this work . in the figure [ fig : rescolumnariceparticle ]
we show the normalized falling velocity @xmath180 obtained by lbm in comparison with some well known theoretical and experimental results from the literature .
these results are the same we have presented for hexagonal columns but in a normalized way . the normalized velocity is computed as proposed by @xcite , this is , to obtain @xmath180 , the crystal settling velocity @xmath157 is divided by a sedimentation velocity @xmath181 for an `` equivalent sphere '' . here , equivalent sphere means a sphere with diameter @xmath162 and mass @xmath182 equal to the ice crystal mass @xmath183 .
@xcite , based in results from @xcite , propose an expression for @xmath157 using the stokes @xcite solution for an sphere in a viscous flow where the sphere radius is replaced by an effective hidrodynamic radius proportional to the capacitance @xmath176 . for hexagonal columns in function of ice crystals
aspect ratio @xmath184 .
the dashed line shows the normalized terminal velocity proposed in @xcite for a random orientation ice crystals .
the experimental results are presented like those in @xcite . ]
the dashed line in the figure [ fig : rescolumnariceparticle ] is the theoretical proposal presented by @xcite to the normalized falling velocity .
this proposal is formulated for columnar ice crystals with hexagonal cross section in random orientation .
the results labeled as mhkc in the figure [ fig : rescolumnariceparticle ] correspond to the proposals from @xcite in random orientation .
mhkc is a typical nomenclature in literature to reference this group of methods .
these are considered identical for @xmath185 @xcite .
also we select from the literature and show in figure [ fig : rescolumnariceparticle ] some experimental data presented by @xcite .
the lbm results at @xmath186 ( @xmath178 with @xmath176 as characteristic length ) in the figure [ fig : rescolumnariceparticle ] are close and slightly above to the @xcite theoretical proposal for hexagonal columns in random orientation .
as expected from the previous comparison with experimental data , the lbm results in the figure [ fig : rescolumnariceparticle ] are below the mhkc proposals .
the lbm results placed further above than @xcite proposal belong to ice crystals falling in vertical or approximately vertical orientation .
the lbm results are obtained using a computational fluid domain as we have shown in the figure [ fig : schematicfallingsphereandresults ] with a relation @xmath85 .
a @xmath91 lattice boltzmann model was used to obtain the numerical results .
using models with more velocities ( @xmath187 or @xmath188 ) , more accuracy can be obtained at a higher computational cost .
nevertheless , as shown in section [ sec : lbm_benchmarks ] , the @xmath91 model give acceptable ( precise enough ) results with a moderate computational cost .
we use a grid refinement scheme with three to five grid scales along the longitudinal axis of the fluid domain .
the boundary conditions in the fluid domain are assigned as : free slip on vertical walls ; dirichlet constant velocity on the bottom wall ; and convective boundary condition ( as presented in section [ subsec : lbm_cbc ] ) on the top wall .
the fluid dynamics was computed in a fluid domain that moves with constant velocity .
we use an initialization procedure as explained in section [ subsec : spheresedimentation ] .
therefore , there is a fictitious transient movement which we are not interested in .
the presence of near free slip walls may have a minor but not negligible influence in the numerical falling velocity .
these weak blockage effects are similar to those tested in section [ subsec : spheresedimentation ] .
for blockage ratios ( defined here as @xmath189 ) bigger than certain value , corrections should be applied to the results of numerical simulations .
we observe that for blockage ratios smaller than @xmath190 the influence of the walls is negligible .
this maximum acceptable blockage value was obtained by evaluating the interference effects on an sphere in sedimentation .
these interference effects are quantified by the relation between the lbm obtained settling velocity and that obtained from theoretical estimations in an unrestricted domain .
differences less than @xmath191 between these velocities were observed for @xmath192 .
the computed results shown in this paper were obtained with blockage ratios smaller than @xmath190 .
this configuration allow us to get @xmath180 by numerical evaluation of @xmath157 and with @xmath181 obtained from stokes equation . for particles in the analyzed reynolds number regime
we have not observed a preferential orientation .
the observed crystal behaviors are in accordance with the results presented by @xcite , and with the experimental observations in @xcite .
@xcite has pointed out some behaviors about orientation of falling bodies at low @xmath1 . from experiments and theoretical predictions
, he notes that needles will fall with indeterminate orientation in the limit @xmath193 .
also , at small but finite @xmath1 there will be an aligning torque tending to make its long axes horizontal . to prove these results at small but finite @xmath1 , a number of numerical test were made in the regime @xmath194 . for cylindrical particles in this regime
its long axes keep approximately horizontal or a few degrees around this orientation .
to check this preference orientation , we also made numerical tests with particles whose long axes are initially in a vertical orientation .
the result we observe is that the particles always turn its long axes to an approximately horizontal orientation .
this section presents the obtained falling velocity @xmath157 results for ice policrystals sedimentation .
the geometry type of the analyzed ice policrystals is the six bullet - rosette .
a schematic geometrical configuration is shown in the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath65 .
bullets and bullet - rosettes are policrystals found in a variety of configurations and orientations in cirrus cloud @xcite . @xcite presented policrystals habit percentages and a classification in shape and size in cirrus clouds as a function of temperature .
@xcite found that bullet rosettes were the predominant form above @xmath195 in a mid - latitude cirrus cloud .
@xcite noted that bullet rosettes were typically observed between @xmath195 and @xmath196 .
there are different bullet - rosette models , @xcite use 19-bullet rosette configuration in their study .
we are interested in policrystals with six hexagonal bullets placed at @xmath197 from each others .
these policrystal geometries are known as six bullet - rosette ( or 6-rosette ) in the literature .
the geometrical models used to represent the ice policrystals are built by six arms of length @xmath163 and diameter @xmath198 .
the characteristic length of the policrystals we analyze are in a range @xmath199 .
the arms are represented by hexagonal columns .
these arms are joined together from one of its ends and placed at @xmath197 from each others as shown the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath65 .
the policrystal aspect ratio is defined as relative to the columnar length , @xmath200 .
the analyzed models are a slight variation of those presented in @xcite .
a bullet density of @xmath201 is used as recommended in @xcite .
the fluid properties are as those at @xmath202 in an international standard atmosphere ( isa ) . :
schematic model used to represent the 6 bullet - rosette ice policrystal .
figure @xmath86 : normalized settling velocity @xmath180 as a function of the crystal aspect ratio @xmath184 .
the figure also shows the normalized terminal velocity proposed in @xcite for 6 bullet - rosette ice policrystal , and the proposal from @xcite for the settling velocity of bullet - rosette with @xmath203.,title="fig : " ] : schematic model used to represent the 6 bullet - rosette ice policrystal .
figure @xmath86 : normalized settling velocity @xmath180 as a function of the crystal aspect ratio @xmath184 .
the figure also shows the normalized terminal velocity proposed in @xcite for 6 bullet - rosette ice policrystal , and the proposal from @xcite for the settling velocity of bullet - rosette with @xmath203.,title="fig : " ] the obtained lbm results and the normalized velocity @xmath180 as proposed by @xcite , are shown in the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath86 .
the @xmath180 is obtained by dividing the six bullet - rosette settling velocity @xmath157 by the terminal falling velocity @xmath181 for an equivalent sphere .
where equivalent sphere means an sphere with diameter @xmath204 and mass @xmath182 equal to the ice policrystal mass @xmath183 . the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath86 shows the lbm results for aspect ratios between 1 and 2.5 .
the analyzed fluid mechanical problems are in a pure viscous regime at @xmath205 . to compare with our results ,
we show in dashed line a theoretical proposal by @xcite for six bullet - rosette policrystal .
also we show an estimation from @xcite for the falling velocity of bullet - rosette with @xmath203 . to the knowledge of the author ,
there are not available experimental measurement of falling velocity for the six - bullet ice policrystals .
as can be observed in the figure , the numerical results are close and slightly below to those from the @xcite proposal .
one of the reasons of the noticeable gap between lbm results and @xcite proposal may be the geometrical differences in the policrystal model .
it can be observed from the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath86 that the @xcite proposal overestimate the falling velocity for @xmath206 in comparison with @xcite and lbm results . the lbm normalized vertical velocity in the figure [ fig : res6bulletrosetteicepolicrystal ] @xmath86 were obtained as in section [ subsec : columnariceparticles ] for columnar ice crystals .
we present in this work a lattice - boltzmann method to determine the dynamics of ice crystals in the atmosphere . given a characterization for the shape , size and mass density of the crystals , together with its atmospheric habitat , it is possible to completely determine the dynamical behavior of the crystals .
the numerical method proposed provides good results for the sedimenting velocity for the geometries , sizes and range of reynolds number analyzed . for the hexagonal column crystals , the results obtained by lbm are completely inside the dispersion region of the experimental , laboratory , measurements .
when the capacitance ( eq . [ eq : columnarcapacitance ] , @xcite ) is used as variable , the dispersion of both experimental and lbm results decreases noticeable . in this case a small bias of the lbm results towards the lower end of the dispersion region can be observed . by direct comparison , we see that the lbm computed sedimenting velocity turns out to be a little higher than the @xcite proposal , and much lower than the proposal of mhkc ( proposals from @xcite ) .
this is observed for all the aspect ratios analyzed .
the bigger differences we find with the proposal of @xcite occur for columns with vertical or near vertical alignment . in the problem of the six bullet rosette crystals the lbm results turn out to be very close and slightly below to the proposal of @xcite .
the difference in the models used to represent the geometry of the crystals could be the reason for this slight difference . to the best of the author s knowledge , there are no reported experimental results for polycrystals of the size , shape and in the reynolds regime analyzed .
we want to emphasize that a great deal of problems , with different geometries and values of parameters ( like density , aspect ratio , etc . )
can be analyzed by using the lbm methods . in this way
one could get statistical characterizations , as those obtained via laboratory experiments , for new crystal shapes ; one could also study the sensitivity models parameters , etc .
the author wants to thank nesvit e. castellano , rodrigo e. brgesser and omar e. ortiz for useful discussions and contributions .
n. e. castellano and r. e. brgesser brought the author s attention to this subject .
j. p. giovacchini is a fellowship holder of conicet ( argentina ) .
this work was supported in part by grants 05-b454 of secyt , unc and piddef 35/12 ( ministry of defense , argentina ) . | lattice boltzmann method ( lbm ) is used to simulate and analyze the sedimentation of small ( @xmath0 ) ice particles in the atmosphere .
we are specially interested in evaluating the terminal falling velocity for two ice particle shapes : columnar ice crystals and six bullet - rosettes ice policrystal .
the main objective in this paper is to investigate the lbm suitability to solve ice crystal sedimentation problems , as well as to evaluate these numerical methods as a powerful numerical tool to solve these problems for arbitrary ice crystal shapes and sizes .
lbm results are presented in comparison with laboratory experimental results and theoretical proposals well known in the literature .
the numerical results show good agreement with experimental and theoretical results for both geometrical configurations . |
in our previous work on steering with josephson phase qubits @xcite we found two peculiar single - step controlled - not ( cnot ) implementations involving off - resonance qubits detuned by the amount smaller than the characteristic coupling constant .
does the detuning always have to be so small ? what happens if it gets larger ?
these questions seem to be important on several counts .
firstly , the majority of entangling gate designs proposed in the literature assume resonant qubits @xcite .
( however , see refs .
@xcite for notable exceptions . )
this is hardly surprising since in the rotating wave approximation ( rwa ) typically used to analyze superconducting qubits the resonant condition leads to relatively simple and easily solvable hamiltonians containing no local @xmath2 terms .
this works well when the system consists of only two qubits .
however , it is reasonable to expect that when thousands of such qubits are assembled into an integrated circuit , maintaining them on resonance will become a difficult task . how then can we be sure that an architecture involving detuned qubits is able to reliably generate universal gates , such as , for example , a cnot gate ?
secondly , some qubits may be fabricated with various defects preventing them from being tuned to resonance exactly .
would they still be usable ? and what if for some applications it becomes advantageous ( or even necessary ) to use qubits with intrinsically different level splittings ?
thirdly , we should also keep in mind that detuning is routinely used in actual experiments to `` decouple '' the qubits in order to perform local ( that is , nonentangling ) operations @xcite
. it would be useful if , after such decoupling is performed , we do not have to worry about bringing the qubits back to exact resonance when doing subsequent entangling operations .
additionally , from the purely theoretical viewpoint , decoupling provides a useful limit against which to check our calculations .
if at larger detuning the interaction is expected to lose its entangling properties , we must be able to predict a crossover into the regime when the gate fidelity ( of the cnot , in our case ) starts to deteriorate .
thus , the primary goal of this paper will be to establish the exact conditions under which such crossover occurs for previously proposed cnot implementations involving superconducting qubits .
the paper is organized as follows : in section [ sec : hamiltonian ] we introduce the hamiltonian for capacitively and inductively coupled josephson phase qubits .
since there is an infinite number of possible local drives and rotating frames to choose from , it will be necessary to limit our discussion to situations that are simple and yet powerful enough to provide some new and interesting physical insights .
our specific choices for rabi pulses and rotating frames will be made in eqs .
( [ eq : ih ] ) , ( [ eq : hathrwa ] ) , and ( [ eq : hrwa ] ) of that section . in section [ sec:2step ]
we generalize to finite detuning the familiar two - step cnot sequence involving a local @xmath3-pulse sandwiched between two entanling operations .
the exact bound on detuning that guarantees generation of the perfect ( in the rwa ) controlled - not logic gate will then be given in eq .
( [ eq : bound2 ] ) . in section [ sec:1step ] we numerically solve the single - step case .
we will see that in that case the restriction on detuning for perfect cnot generation is somewhat stronger than in the two - step case ( eq .
( [ eq : bound1 ] ) ) . in section [ sec : largedetuning ] we present optimal results for approximate cnot gates at detuning larger than maximally allowed .
we will see how the makhlin invariants and fidelity of the optimized gates deviate from their ideal cnot values .
we will also simulate the weyl chamber steering trajectories corresponding to such optimized gates .
we conclude in section [ sec : conclusion ] with a brief summary of our results .
in what follows , we will use the notation that is convenient for lie algebraic manipulations , & & x_k = _ k^x , y_k = _ k^y , z_k = _ k^z , ( k = 1,2 ) , + & & xx = _
2^x_1^x , yy = _
2^y_1^y , zz = _ 2^z_1^z , + & & xy= _ 2^x_1^y , yx = _ 2^y_1^x .
notice , that @xmath4=[xy , yx]=0 $ ] , and @xmath5 commutes with each of the @xmath6 , @xmath7 , @xmath8 , @xmath9 operators . why is this notation convenient ? consider the following transformation ( called `` going to a rotated frame '' ) on the lie algebra @xmath10 of the two - qubit system : [ eq : lierotation ] xx e^-z_1 xx e^z_1 = xx + xy .
this transformation can be nicely interpreted as a rotation of vector @xmath6 by an angle @xmath11 in the _ real _ vector space spanned by the generators of the group su(4 ) .
this is how the continuous group acts on its lie algebra .
mathematicians call it the adjoint representation .
the algebra plays the role of the representation space for its own group .
since we do work at the level of algebra , and not at the level of the hilbert space when discussing equivalence classes of gates , this is a very convenient notation .
it simplifies things .
also notice how naturally the periodicity of @xmath12 appears at this level of description .
now , if we were to write the same transformation in terms of the pauli matrices , we would have to remember to put in the imaginary unit @xmath13 and the factors of 1/2 in the exponents on the left hand side of eq .
( [ eq : lierotation ] ) , @xmath14 breaking its beautiful symmetry .
when restricted to the computational subspace , the hamiltonian for two coupled josephson phase qubits , one of which is driven by a resonant rf pulse , is given by [ eq : ih ] ih(t ) & = & -z_1 -(+ ) z_2 + & & + 2 _ 1(t)x_1 + 2(g yy + zz ) , where @xmath15 are the coupling constants , @xmath16 is the level splitting of the first qubit , @xmath17 is the detuning , and @xmath18 is the corresponding rabi frequency . here
we assume that @xmath19 , which differs from the condition @xmath20 adopted in ref .
for realistic systems , @xmath21 ghz , @xmath22 mhz . for capacitive coupling , @xmath23 ; for inductive coupling , @xmath24 . in the doubly rotating frame defined by @xmath25 , with @xmath26 ,
after averaging over fast oscillations , the system hamiltonian is time - independent , [ eq : hathrwa ] ih_rwa & = & -z_2 + _ 1 x_1 + ih + zz , with ih = g(xx+yy ) = i g 0 & 0 & 0 & 0 0 & 0 & 1 & 0 0 & 1 & 0 & 0 0 & 0 & 0 & 0 . alternatively , to perform a useful consistency check , we consider another rotating frame defined by @xmath27 , with @xmath28 . in this frame ,
the rwa hamiltonian is [ eq : hrwa ] ih_rwa(t ) & = & _ 1 x_1 + i ( t ) + zz , where now we have a slowly varying interaction term given by [ eq : timedependentinteraction ] i(t ) & = & g + & = & i g 0 & 0 & 0 & 0 0 & 0 & e^-it & 0 0 & e^+it & 0 & 0 0 & 0 & 0 & 0 .
the central block of this matrix has the form of a rotating drive for a spin-1/2 system for which the analytical solution is well known .
, the corresponding time evolution operator is given by @xmath29 .
( [ eq : timedependentinteraction ] ) corresponds to @xmath30 . ]
this observation will prove helpful for calculations in section [ sec:2step ] .
we will now show how these two rwa hamiltonians lead to locally equivalent cnot implementations .
the well - known two - step cnot implementation for _ resonant _ ( @xmath31 )
qubits has control sequence @xcite [ eq:2stepcnot ] cnot_(2 ) = e^i(/4 ) r_post r_pre , with @xmath32 , where u(t_(2 ) ) & = & e^- t_(2 ) ( g(xx+yy)+zz ) + & = & e^- ( /4g)zz 1 & 0 & 0 & 0 0 & 1/ & -i/ & 0 0 & -i/ & 1/ & 0 0 & 0 & 0 & 1 , and @xmath33 are some local rotations .
we will choose [ eq : rprepost0 ] r_post = e^-(/2)y_2 , r_pre = e^-(/2)z_2e^+(/2)(x_2+x_1 ) .
of particular importance to us is the entangling part @xmath34 that determines the local equivalence class of the full gate . in our case
the local class is controlled - not whose canonical representative in the computational basis is defined to be 1 & 0 & 0 & 0 0 & 1 & 0 & 0 0 & 0 & 0 & 1 0 & 0 & 1 & 0 , ( cnot ) = -1 .
we ask the following question : if @xmath35 ( _ detuned _ qubits ) , can we still use the sequence in eq .
( [ eq:2stepcnot ] ) to generate a cnot gate , possibly with _
different _ gate time and different pre- and post - rotations ?
the answer to this question turns out to be `` yes '' , provided @xmath36 is restricted in a certain way . setting @xmath37 in eqs .
( [ eq : hathrwa ] ) and ( [ eq : hrwa ] ) gives the time evolution operator u(t ) = e^- t zz e^it/2 & 0 & 0 & 0 0 & u & -iv & 0 0 & -iv & u^ * & 0 0 & 0 & 0 & e^-it/2 in frame 1 , and ( t ) = e^- t zz 1 & 0 & 0 & 0 0 & u e^-it/2 & -iv e^-it/2 & 0 0 & -ive^it/2 & u^*e^it/2 & 0 0 & 0 & 0 & 1 in frame 2 , where u = ( t ) + ( t ) , v = ( t ) . in both frames
the makhlin invariants @xcite of @xmath34 are g_1 & = & ( ) ^2 , and g_2 & = & ( 3 ^ 4 + 8 ^ 2g^2 + & & + 16g^4)/(^2 + 4g^2)^2 , + and are independent of the @xmath5 coupling .
this shows that for any @xmath38 the resulting gates are represented by the same point on the @xmath39-plane of the weyl chamber ( see @xcite for discussion ) .
since cnot class corresponds to @xmath40 , @xmath41 , we get [ eq : t2 ] t_(2 ) = , with the limit @xmath42 trivially recovered for vanishing detuning . for example , for @xmath23 , @xmath43
, we get @xmath44 .
the corresponding cnot gate is given by eq .
( [ eq:2stepcnot ] ) , where now & & u(t_(2 ) ) = 0.9180 + 0.3965i & 0 & 0 & 0 0 & 0.6124 + 0.3536i & - 0.7071i & 0 0 & - 0.7071i & 0.6124 - 0.3536i & 0 0 & 0 & 0 & 0.9180 - 0.3965i , + & & r_post = e^-(/2)y_2e^-(/2)(_2 z_2 + _ 1 z_1 ) , r_pre = e^-(/2)((1+_2)z_2 + _ 1 z_1)e^+(/2)(x_2+x_1 ) , with @xmath45 , @xmath46 in frame 1 , and & & u(t_(2 ) ) = 1 & 0 & 0 & 0 0 & 0.7024 + 0.0817i & -0.2804 - 0.6491i & 0 0 & 0.2804 - 0.6491i & 0.7024 - 0.0817i & 0 0 & 0 & 0 & 1 , + & & r_post = e^-(/2)y_2e^-(/2 ) z_2 , + & & r_pre = e^-(/2)(1+)z_2e^+(/2)(x_2+x_1 ) , with @xmath47 , @xmath48 in frame 2 . here , the pre- and post - rotations ( cf .
( [ eq : rprepost0 ] ) ) needed to generate the perfect in the rotating wave approximation canonical controlled - not gate have been found numerically .
notice that the two - step control sequence works only for [ eq : bound2 ] ||2 g .
here , for simplicity , we limit our discussion to the rwa hamiltonian given in eq .
( [ eq : hathrwa ] ) with @xmath23 ( capacitive coupling ) .
[ in this case , a single rabi term suffices to implement a cnot gate .
when @xmath49 , an additional local rabi drive @xmath50 must be applied to the second qubit . ] the cnot sequence @xcite is then [ eq:1stepcnot ]
cnot_(1 ) = e^i(5/4 ) r_post u(t_(1 ) ) r_pre , where @xmath51 , @xmath52 , @xmath53 , u ( t_(1 ) ) & = & e^- t_(1 ) [ _ 1 x_1 + g(xx+yy ) ] + & = & 1 & 0 & 0 & -i 0 & 1 & -i & 0 0 & -i & 1 & 0 -i & 0 & 0 & 1 , and [ eq : rprepost15 ] r_post = e^-(/2)y_2 , r_pre = e^-(/2)z_2e^+(/2)(x_2-x_1 ) . in order to generalize this single - step implementation to finite detuning
, we optimize the gate parameters @xmath54 and @xmath18 of [ eq : optimizedgate1 ] u ( t_(1 ) ) = e^- t_(1 ) [ -z_2 + _ 1 x_1 + g(xx+yy ) ] using nelder - mead simplex direct search with bound constraints for the minimum of the distance from the cnot class defined by [ eq : makhlinnorm ] d^2(_1 , t_(1 ) ) & : = & |g_1(_1 , t_(1))-g_1(cnot)|^2 + & & + |g_2(_1 , t_(1))-g_2(cnot)|^2 , with @xmath55 being the makhlin invariants of @xmath56 .
it is important to keep in mind that the distance function introduced in eq .
( [ eq : makhlinnorm ] ) is _ not _ a measure of the gate fidelity .
infinitely many gates all differing from each other by arbitrary local rotations may have the same value of @xmath57 . once the entangling part @xmath58 is found to have @xmath59 , it
can then be made into the canonical cnot gate by additional local rotations .
actual experiments motivate this choice of the distance function .
it is generally believed that doing local rotations is easy , but performing entanglement is difficult .
thus , if by using experimentally available interaction and the local controls we can somehow steer the system into the `` right '' equivalence class , then making the actual target gate will be relatively straightforward . performing the optimization we find that in order to generate the exact cnot in the single - step case the detuning must be restricted by [ eq : bound1 ] || g .
for example , for maximally allowed @xmath60 , we get @xmath61 , @xmath62 . the resulting gate is given by eq .
( [ eq:1stepcnot ] ) with & & u ( t_(1 ) ) = -0.2553 - 0.4300i & 0.4821 - 0.1324i & -0.4821 + 0.1324i & 0.5001i 0.4821 - 0.1324i & -0.0001 + 0.5001i & 0.5001i & 0.4821 + 0.1324i -0.4821 + 0.1324i & 0.5001i & -0.0001 - 0.5001i & 0.4821 + 0.1324i 0.5001i & 0.4821 + 0.1324i & 0.4821 + 0.1324i & -0.2553 + 0.4300i , + & & r_post = e^-(/2)y_2e^-(/2)(_2 z_2 + _ 1 z_1 ) , r_pre = e^-(/2)((1+_2)z_2 + _ 1 z_1 ) e^+(/2)(x_2-(1+_1 ) x_1 ) , where @xmath63 , @xmath64 , @xmath65 . to visualize how this gate is reached in the course of the unitary evolution we simulate the weyl chamber steering trajectory @xmath66 for the above mentioned values of the gate parameters .
the goal here is to establish a correspondence @xcite , u(t ) ~e^- c_1(t)xx - c_2(t)yy - c_3(t)zz , between the physical gate @xmath67 and the unphysical matrix exponential that formally resides in the same local equivalence class as @xmath68 .
the time - dependent vector @xmath69 then represents the dynamically generated class at every moment of system s evolution .
figures [ fig:1 ] and [ fig:2 ] show how the cnot class that has @xmath70 is generated in our single - step example with @xmath60 .
time dependence of @xmath71 ( top curve ) and @xmath72 ( bottom curve ) for single - step cnot implementation with capacitively coupled josephson phase qubits at maximal detuning , @xmath60 . here , @xmath73 at all times .
dashed curves represent the resonant case , @xmath31 . ]
single - step steering trajectory corresponding to fig .
[ fig:1 ] that generates cnot class at maximal detuning , @xmath60 .
the steering parameters are given in units of @xmath74 .
dashed curve is for @xmath31 . ] .
[ tab : table1 ] generation of ideal controlled - not logic ( @xmath75 , @xmath41 ) using capacitively coupled josephson phase qubits at finite detuning . in the two - step case , @xmath76 , @xmath77 , and @xmath78 , as given in eq .
( [ eq : t2 ] ) . in the single - step case , @xmath79 .
the optimal values of @xmath80 and @xmath81 have been found numerically using nelder - mead simplex direct search with bound constraints . [ cols="<,^,^,^",options="header " , ] time dependence of @xmath71 ( top curve ) and @xmath72 ( bottom curve ) for single - step implementation generating the class closest to cnot in terms of eq .
( [ eq : makhlinnorm ] ) at @xmath82 . here , @xmath73 at all times .
dashed curves represent the resonant case .
time is measured in units of @xmath83 .
the closest class is reached at @xmath84 . ]
single - step steering trajectory corresponding to fig .
[ fig:3 ] at detuning @xmath82 .
the steering parameters are given in units of @xmath74 .
dashed curve is for @xmath31 . ]
figures [ fig:3 ] and [ fig:4 ] show the simulated single - step steering trajectory generated by the rwa hamiltonian given in eqs .
( [ eq : hathrwa ] ) and ( [ eq : optimizedgate1 ] ) at @xmath82 . direct search for local pulses ( not shown ) results in the optimized gate , u_opt = 0.9866 & - 0.1122i & 0.0258i & 0.1158 -0.1122i & 0.9866 & -0.1158 & - 0.0258i -0.1186 & 0.0009i & 0.1122i & 0.9866 - 0.0009i & 0.1186 & 0.9866 & 0.1122i , whose intrinsic fidelity with respect to the canonical controlled - not gate is & & + & = & 0.9448 .
in summary , we have demonstrated that the cnot pulse sequences previously proposed for resonant josephson phase qubits may still be used at finite detuning . to achieve high fidelity of the resulting gate the value of the detuning during the entangling operations should not be greater than @xmath0 in the two - step implementation and @xmath1 in the single - step implementation , respectively .
j. majer , j. m. chow , j. m. gambetta , j. koch , b. r. johnson , j. a. schreier , l. frunzio , d. i. schuster , a. a. houck , a. wallraff , a. blais , m. h. devoret , s. m. girvin , and r. j. schoelkopf , nature * 449 * , 443 ( 2007 ) . | we establish theoretical bounds on qubit detuning for high fidelity controlled - not logic gate implementations with weakly coupled josephson phase qubits .
it is found that the value of qubit detuning during the entangling pulses must not exceed @xmath0 for two - step , and @xmath1 for single - step control sequences , where @xmath1 is the relevant coupling constant . |
the rich transport behavior in complex systems is an important research topic in statistical physics @xcite .
consider the mean square of a physical quantity ( such as position ) as a function of time @xmath0 for an ensemble of particles .
for normal diffusion , this mean quantity is proportional to @xmath0 ; whereas for anomalous diffusion , the mean quantity goes like @xmath1 ( @xmath2 ) , with subdiffusion referring to cases of @xmath3 and superdiffusion referring to cases of @xmath4 . in the classical domain , known examples of anomalous diffusion include brownian motion and heat conduction . in the quantum domain ,
wavepacket spreading in a periodic potential leads to ballistic transport in the mean square position ( @xmath5 ) .
by contrast , wavepacket spreading in a quasi - periodic potential often induces subdiffusion or superdiffusion @xcite .
the special class of diffusion with @xmath6 , which may be termed as `` superballistic transport '' , is however not as well - studied as other cases of anomalous transport behavior . in the classical domain ,
superballistic transport was observed for brownian particles @xcite . in the quantum case ,
a time - dependent random potential was demonstrated to cause superballistic transport using paraxial optical setting @xcite .
more related to this work , earlier superballistic transport was found in the dynamics of wavepacket spreading in a tight - binding lattice junction @xcite .
remarkably , such type of quantum superballistic transport was recently experimentally realized by use of optical wave packets in a designed hybrid photonic lattice setup @xcite .
the main objective of this work is to use a relatively simple model system to better understand the difference and connection between quantum and classical superballistic transport in purely hamiltonian systems . to our knowledge
, the model studied in this work represents the only dynamical model that can possess superballistic transport in both quantum and classical hamiltonian dynamics .
this will shed more light on various mechanisms of superballistic transport as well as on the general issue of quantum dynamics in classically chaotic systems .
further studies regarding the subtle correspondence between quantum and classical superballistic transport can be also motivated .
specifically , we consider a relativistic variant @xcite of the well - known kicked - rotor ( kr ) model @xcite and reveal the quantum and classical superballistic transport dynamics in the momentum space .
such a system can be also regarded as a periodically driven dirac system , and it should be of some experimental interest due to recent advances in the quantum simulation of dirac - like particles . in the massless case
in which the kinetic energy of a relativistic particle is a linear function of momentum , the relativistic kr variant was known as the `` maryland model '' , first investigated by grempel _ et .
@xcite , berry @xcite and simon @xcite to analytically understand the issue of anderson localization . for a dirac particle with a nonzero mass
, the bare dispersion relation now lies between linear and quadratic : for low momentum values the dispersion is almost quadratic and for very high momentum values the dispersion approaches a linear function . in effect
, this realizes a situation , now in the momentum space , in which two ( momentum ) sublattices with different dispersion relations ( and hence different nature of on - site potential ) are connected as a junction @xcite . as shown later , this indeed induces superballistic transport in both the classical and quantum dynamics .
interestingly , the detailed mechanism in the former is still markedly different from that in the latter . in particular , in the classical case , it is necessary to break the global kam curves in the classical phase space because the superballistic transport roots in an unusually complicated escape from a phase space regime of random motion to a simple ballistic structure .
by contrast , in the quantum case , breaking global kam curves are not essential for quantum superballistic transport to occur thanks to quantum tunneling through the kam curves .
this paper is organized as follows . in sec .
ii , we introduce our model and discuss its relation with the well - known maryland model . in sec .
iii we study quantum superballistic dynamics , followed by a parallel study of classical superballistic transport and the associated classical phase space structure in sec .
section v concludes this work .
consider a one - dimensional relativistic quantum kr @xcite : @xmath7 where all the variables are scaled and hence in dimensionless units . here
@xmath8 and @xmath9 are pauli matrices , @xmath10 represents the speed of light , @xmath11 represents the static mass energy , @xmath12 represents the strength of a delta - kicking field that is @xmath13-periodic in the coordinate @xmath14 ( @xmath15 is an integer ) and unity - periodic in time , @xmath16 , where @xmath17 is a dimensionless effective planck constant . in the case of a vanishing @xmath11
, the hamiltonian can be decoupled into two independent hamiltonians , each associated with one eigen - spinor of @xmath8 .
they are nothing but the so - called maryland model @xcite : @xmath18 early studies on this massless relativistic kr @xcite investigated the consequences of rational or irrational values of @xmath19 .
indeed , by mapping the maryland model onto a one - dimensional anderson model , it becomes clear that @xmath20 with an irrational @xmath19 should display anderson localization in the momentum space @xcite , whereas @xmath20 with a rational @xmath21 ( @xmath22 and @xmath23 integers ) should show ballistic transport , so long as the parameter @xmath15 in the kicking potential is an integer multiple of @xmath23 @xcite .
this condition is called a resonance condition .
when this resonance condition is not fulfilled , i.e. , @xmath24 , it can be shown that the time - evolving wavefunction of the system ( with rational @xmath19 ) repeats itself after every @xmath23 kicking periods , so neither anderson localization nor ballistic transport occurs .
roughly speaking , what is known in the standard quantum kr @xcite applies also here , concerning the importance of the arithmetic nature of @xmath19 for the dynamics of the maryland model , as well as the ballistic transport due to quantum resonances therein .
though in this work we will focus on cases with a nonzero @xmath11 , the above - mentioned results for the maryland model do guide us when it comes to choose interesting parameter regimes . for example , we shall pay special attention to whether or not @xmath19 is a rational value , and if the kicking potential is resonant or not .
as it turns out , the quantum dynamics is most interesting in cases with rational @xmath19 and under the resonance condition .
since our model is a periodically driven system , we write down its floquet operator in the @xmath14-representation , i.e. , the propagator associated with one - period time evolution : @xmath25 without loss of generality we set @xmath26 throughout , keeping in mind that if @xmath27 , we may just absorb it into other parameters @xmath11 and @xmath12 . due to this choice
, with periodic boundary conditions in @xmath14 , momentum can only take integer values .
the above expression of the floquet operator is a product of two exponentials .
the second factor @xmath28 $ ] comes from the kicking field , giving rise to the hopping between different momentum states ( different sites in the momentum space ) , whereas the first factor @xmath29 $ ] is responsible for generating momentum - dependent on - site phases .
in particular , the state at a site @xmath30 can be decomposed into local spin - up and spin - down components , each component acquiring a phase factor @xmath31 : @xmath32.\ ] ] clearly then , it is the periodic and alternative on - site phase accumulation @xmath31 and the hopping @xmath33 $ ] that determine the quantum dynamics .
this motivates us to also consider a slightly different floquet operator @xmath34 : @xmath35 for this spinless floquet operator , the two spin components are decoupled .
nevertheless , it still contains the same local phase accumulation given by @xmath31 and the same hopping term as in our original model described by @xmath36 .
as seen later , it does possess the essential properties of @xmath36 and as such our physical analysis can be reduced .
more importantly , the system @xmath34 does not have the spin degree of freedom , so its classical limit can be constructed with ease , with the classical hamiltonian given by @xmath37 indeed , the so - called classical relativistic kr map studied in the literature @xcite was based on such a spinless classical hamiltonian . in the following , we study the quantum dynamics using both @xmath36 and @xmath34 , and the classical dynamics based on @xmath38 .
the dynamics of a quantum relativistic kr was previously studied in ref .
@xcite for relatively short time scales . by extending to a longer time scale and choosing the right parameter regime ,
quantum superballistic transport is found for driven systems described by @xmath36 as well as its spinless version @xmath34 . to justify our choices of the system parameters we first examine cases with an irrational value of @xmath19 . in the maryland model
, an irrational @xmath19 leads to localization in the momentum space .
so it is interesting to first investigate how a nonzero mass @xmath11 changes this picture .
when @xmath19 is irrational , the previously defined phase factor @xmath31 in eq .
is in general a pseudo - random function of the momentum site .
this is different from the maryland model , in which @xmath39 and the corresponding @xmath31 would then reduce to a quasi - periodic function of momentum sites . in the light of the mapping from a kr system to the anderson localization model @xcite , this seems to indicate that a non - zero @xmath11 favors dynamical localization in the momentum space .
note also that , for very large values of @xmath30 , the relative importance of the @xmath11 term in @xmath31 will diminish , and then effectively a maryland model will be recovered and dynamical localization is still guaranteed @xcite .
thus , in the entire momentum space , a non - zero @xmath11 is expected to strengthen the dynamical localization , thus also wiping out any possibility of anomalous diffusion or superballistic transport .
results of our numerical simulations presented in fig .
[ fig : m_irra ] support our above view .
for both the full floquet operator @xmath36 and its spinless version @xmath34 , it is seen from fig . [ fig : m_irra ] that the momentum spread @xmath40 decreases as @xmath11 increases , i.e. , a larger @xmath11 enhances dynamical localization . as a function of number of kicks ( time @xmath0 ) .
solid lines represent results generated by the floquet operator @xmath36 defined in eq .
( [ floquetr ] ) , with the initial state given by spin - up and a gaussian wavepacket @xmath41 .
the dot - dashed lines represents the results generated by a spinless floquet operator @xmath34 defined in eq .
( [ floquetrn ] ) , with the initial wavepacket given by the gaussian @xmath41 . in both cases , @xmath42 , irrational @xmath43
, @xmath44 , @xmath45 and @xmath11 takes values of 0 , 6 and 10 .
( results of @xmath36 and @xmath34 with @xmath39 are identical . )
@xmath40 is seen to be strongly localized in all the examined examples .
here and in all other figures , all plotted quantities are in dimensionless units . ] with the same set of system parameters , the mean momentum spread under the evolution of the floquet operator @xmath36 is similar to that under the evolution of its spinless version @xmath34 .
this confirms that the common on - site phase accumulation function @xmath31 has captured the main features of dynamical localization .
it is also interesting to comment on the differences between these two cases .
that is , @xmath40 associated with @xmath36 is always larger than that associated with @xmath34 .
as such , the spin degree of freedom is seen to slightly weaken dynamical localization .
this is somewhat expected .
indeed , the spin degree of freedom introduces two channels for the dynamics and a multi - channel anderson model does increase the localization length @xcite .
we note in passing that the spin degree of freedom can even cause anderson transition in two - dimensional disordered systems @xcite . certainly , as @xmath11 increases , this spin effect should decrease because it becomes more costly in energy for the two spin channels to interact . finally , we mention the benchmark result in fig . [
fig : m_irra ] for the @xmath39 case that represents the maryland model .
the perfect revival of @xmath40 was predicted by berry @xcite . using our system parameters depicted in the caption of fig .
[ fig : m_irra ] , berry s result gives @xmath46 and our simulation agrees with this .
as we already discussed in the previous subsection , for large momentum values , the effect of a nonzero mass term @xmath11 will diminish and effectively the maryland model will re - emerge .
so if we were not to choose an on - resonance potential , then in regimes of large momentum quantum revivals should occur for rational @xmath19 , which is not of interest here .
in addition , we also observed that for an off - resonance kicking potential and for rational @xmath19 , a non - zero @xmath11 further suppresses the already bounded momentum spread . with these understandings ,
it is clear that we should step into the interesting situation where @xmath21 is rational and the kicking potential @xmath47 is on resonance , i.e , @xmath48 . for convenience
we choose @xmath49 .
reference @xcite computationally investigated exactly the same situation , but it was argued therein that the phase factor @xmath31 as a pseudo - random function of momentum should suffice to localize the momentum spread . as we show below , both qualitatively and quantitatively , this claim is correct only for low momentum values , and overall a much richer transport behavior can be found .
let us start with the maryland model for which @xmath39 .
then the momentum space is translational invariant with period @xmath23 . as such , states will in general spread ballistically (
the off - resonance case is an exception ) . in our case ,
@xmath50 and within each period , the on - site phase @xmath31 acquired by the system becomes a quasi - random function of @xmath30 , thus dynamical anderson localization or suppression of momentum spread is expected .
however , as momentum increases , the nonzero @xmath51 term in @xmath52\right)$ ] becomes less important as compared with @xmath53 .
for very large momentum , the @xmath51 term represents a very weak perturbation and hence the dynamics should resemble that of the maryland model .
the above qualitative analysis makes it clear that the overall dynamics depends on many factors . on a sublattice representing low momentum values from @xmath54 to @xmath55 ,
dynamical localization takes place and the system is effectively in a disordered regime . on a sublattice representing higher momentum values ,
ballistic transport is expected and the system is effectively in a periodic regime . whether or not a state is localized or delocalized now depends on where it is initially located , and on the size of the disordered regime as compared with the localization length .
for example , if an initial state is localized at the center of the disordered regime and if the localization length is much shorter than the disordered region , then the system may be trapped there for an extremely long time . on the other hand ,
if the initial state is already located close to the high - momentum sublattice ( closeness is with respect to the localization length ) , then as the kicking field induces population transfer between the disordered sublattice and the periodic sublattice , the system will be quickly delocalized . in this sense ,
our model , through its natural dispersion relation , realizes a lattice junction analogous to that considered in refs .
certainly , in our model here there is no sharp transition between the two qualitatively different sublattices , but the critical momentum value is expected to scale with @xmath56 .
vs time , for @xmath39 , @xmath57 , @xmath58 , @xmath59 and @xmath60 ( from top to bottom ) .
results with @xmath57 , @xmath58 , @xmath59 are for the spinless floquet operator @xmath34 defined in eq .
( [ floquetrn ] ) . for @xmath60 ,
results for both @xmath36 ( the one fitted by a power - law with @xmath61 ) and @xmath34 ( the most localized case ) are plotted .
( result with @xmath39 is applicable to both @xmath36 and @xmath34 . )
other system parameters are the same as in fig .
[ fig : m_irra ] , except for @xmath62 .
the two dashed straight lines represent a power - law fitting @xmath63 ( for a certain time window ) with @xmath61 or @xmath64 , indicating quantum superballistic transport with an exponent close to @xmath65 . ]
representative results from our numerical experiments are presented in fig .
[ fig : m_ra_log ] .
there it is seen that as @xmath11 increases , an initial state localized at the center of the disordered sublattice will be trapped for a longer period .
this is consistent with our understanding that an increasing @xmath11 leads to a longer disordered sublattice as well as a shorter localization length .
other numerical results ( not shown ) show that the values of @xmath19 , the kicking strength @xmath12 , and the potential parameter @xmath15 can all affect the duration during which an initial state is trapped in the disordered regime .
two of the computational examples shown in fig .
[ fig : m_ra_log ] also display quantum superballistic transport , i.e. @xmath63 with @xmath6 , where @xmath0 is the number of kicks .
this can now be explained using the idea from ref .
in particular , assuming that the length of the disordered momentum sublattice is larger than the corresponding localization length , then the disordered regime serves as a source to provide slow probability leakage into a periodic regime at an almost constant rate .
then we have @xmath66 here @xmath67 is the probability leaking rate from the disordered sublattice to the periodic sublattice , @xmath68 represents the contribution from the disordered regime , which is almost constant and can be neglected .
@xmath69 represents the probability of the initial state already placed in the periodic regime , which is 0 due to our choice of the initial state located at the center of the disordered regime .
@xmath70 characterizes the ballistic transport coefficient , which depends on many system parameters .
if we approximate @xmath67 by a constant @xmath71 , then from eq .
we approximately have @xmath72 as reflected by the two cases shown in fig .
[ fig : m_ra_log ] , namely , the case of @xmath60 for the full floquet operator @xmath36 and the case of @xmath59 for the spinless floquet operator @xmath34 defined in eq .
( [ floquetrn ] ) , @xmath63 , with @xmath73 or @xmath74 , for a very long time scale and covering a huge range of @xmath40 ( note the logarithmic scales used in the plot ) .
these two superballistic exponents are very close to @xmath65 , in agreement with the above theory .
for the same two cases , we have set @xmath75 and record the probabilities inside @xmath76 $ ] as a function of time .
this probability indeed decreases linearly with time .
this further confirms the physical mechanism behind the quantum superballistic transport seen here .
the case shown in fig .
[ fig : m_ra_log ] with @xmath57 displays ballistic transport as the case of @xmath39 .
this is so because for a small value of @xmath11 the initial state quickly experiences ballistic transport on the clean sublattice . for the intermediate case @xmath58
, the momentum spread does not show any clear power - law dependence . in this transitional case ,
the localization length and the size of disordered lattice are comparable and hence the leakage from the disordered sublattice to the periodic sublattice occurs no longer at an almost constant rate .
we stress that the shown cases represent but a few examples .
many similar results of quantum superballistic transport are obtained for both the full floquet operator @xmath36 involving two spin channels and for the spinless floquet operator @xmath34 . though the quantum superballistic transport here is explained in the same manner as in ref .
@xcite , we stress that the effective two - sublattice configuration is not artificially designed . rather , it emerges as a natural consequence of the relativistic dispersion relation with a nonzero mass .
in this subsection we will study the classical relativistic kr described by the hamiltonian in eq . .
to that end , it is necessary to examine the phase space structure , which can be generated from the relativistic standard map @xcite . in particular , the states right after the @xmath77-th and @xmath78-st kick are connected by the map @xmath79 in the quantum case , either an irrational @xmath19 or a non - resonant kicking potential causes localization . considering quantum - classical correspondence ,
this suggests bounded ( i.e. , localized in momentum ) invariant curves in the phase space .
results in fig .
[ fig : phasespaceallcases](b)-(d ) support this view .
defined in eq .
( [ spinlessh ] ) , for four different situations , namely , ( a ) rational @xmath19 and on - resonance kicking potential , ( b ) irrational @xmath19 with off - resonance kicking potential , ( c ) irrational @xmath19 with an almost on - resonance kicking potential ; and ( d ) irrational @xmath19 with an off - resonance kicking potential . @xmath59 and @xmath80 . the phase space invariant curves are seen to be unbounded in momentum in panel ( a ) , but bounded in panels ( b)-(d ) . ] to understand the phase space structure , we first recall the maryland model @xcite , which can approximately describe the dynamics for sufficiently large momentum values .
that is , if @xmath30 is large , then we again neglect the @xmath11 term in the hamiltonian .
then the mapping in eq .
( after dropping the @xmath11 term ) reduces to the mapping associated with the maryland model . for the maryland model , the following equations hold for either an irrational @xmath19 or a non - resonant kicking potential : @xmath81\right)\\ & \quad + p_0 , \end{split}\ ] ] where as introduced before , @xmath82 .
clearly then , for an irrational @xmath19 , the term @xmath83 can densely cover the entire @xmath14 domain @xmath84 $ ] . as such , as @xmath77 increases , the values of @xmath85 fill a complete sine curve in the phase space .
on the other hand , if @xmath19 is rational , @xmath86 becomes a fraction under the assumed off - resonance condition , then @xmath87 can only take discrete points in @xmath84 $ ] such that @xmath88 also takes a few isolated values @xcite . these observations for the maryland model can be used to directly explain the panel ( d ) in fig .
[ fig : phasespaceallcases ] . for panel ( b ) where @xmath19 is rational but
the kicking potential is off resonance , a nonzero @xmath11 also causes the dynamics to densely fill the entire @xmath14 domain , which constitutes an interesting difference from the maryland model .
however , as expected , after the same number of iterations , regimes with low momentum values can generate a complete phase invariant curve faster , and regimes with high momentum values may still have holes to be filled in . as to panel ( c ) where the product of @xmath89 is close to an integer ,
the oscillation amplitude in momentum get larger as the factor @xmath90 in the map in eq .
( [ p_theta_bounded ] ) can be a large number .
putting all these cases together , it is seen that in the three situations represented by panels ( b)-(d ) of fig .
[ fig : phasespaceallcases ] , kam invariant curves localized in momentum are the main characteristic of the classical phase space .
so now we are left with the last situation in which @xmath19 is rational and the kicking potential is on resonance .
the associated phase space structure is presented in panel ( a ) of fig .
[ fig : phasespaceallcases ] and in fig .
[ fig : kam_detail ] .
the phase space structure is remarkably complicated and interesting . to investigate this in detail
, we divide the phase space into four regimes , namely regimes i , ii , iii and iv for increasing absolute values of momentum .
let us take one example to look into the special phase space structure . as seen in fig .
[ fig : kam_detail ] for @xmath44 , @xmath59 , the four regimes have qualitatively different behavior .
in regime i where @xmath30 is small , local kam curves dominate . as @xmath30 increases ,
the feature of the phase space become chaotic in regime ii .
this is followed by global kam curves in regime iii .
these global kam curves are localized in momentum and bound the chaotic sea seen in regime ii . finally , in regime iv for quite large momentum , ballistic curves , which become more and more parallel to the momentum axis , are seen .
these curves are called ballistic curves because once a trajectory lands on such a structure its momentum variance will evolve ballistically . . here
system parameters are given by @xmath44 , @xmath59 , @xmath62 , and @xmath45 .
panel ( a ) depicts regimes with large momentum values , and panel ( b ) depicts regimes with low momentum values .
curves in regime i are local kam curves , kam curves in regime iii are global curves and hence localize momentum .
the black dots regime in ii are generated by a single initial condition and indicate a chaotic sea .
curves in regimes iv that are almost parallel to the momentum axis are ballistic structures . ]
the four regimes identified above may not always appear together .
their presence and borders are determined by @xmath11 and @xmath12 . to shed some light on this ,
one may consider @xmath38 , in the two opposite limits , i.e. , `` non - relativistic '' @xmath91 limit and `` ultra - relativistic '' limit @xmath92 .
in the first limit , @xmath38 becomes @xmath93 and in the second limit , @xmath38 assumes @xmath94 regimes i and ii can be understood via @xmath95 , while regime iv can be well understood by @xmath96 . in particular
, @xmath95 is the conventional kicked rotor , which makes the regular - to - chaotic transition as @xmath12 increases @xcite .
@xmath96 is much similar to the hamiltonian of the maryland model , which is known to produce ballistic trajectories in the momentum space @xcite .
indeed , it can be shown that the asymptotic ( in the large @xmath30 limit ) form of the ballistic trajectories are described by @xmath97 , and they are hence completely parallel to the momentum axis .
as is found from our computational studies , an increase in @xmath12 may destroy the kam curves in regimes i and iii , and then turns them to a ( possibly transient ) chaotic sea as well .
an increase in @xmath11 will generate more kam curves in the phase space .
the complexity of the phase space perhaps deserves more careful studies . for our purpose here
, we emphasize that for a sufficiently large @xmath12 , the phase space is mainly composed of a seemingly chaotic sea and ballistic trajectories .
it is however challenging to identify a clear boundary between trajectories eventually landing on the ballistic structure and those always doing random motion . to appreciate this complexity , in fig .
[ fig : phasesinglepoint ] we illustrate that once global curves are all broken , how an individual trajectory might eventually land on the ballistic structure after transient , but a long period of , chaotic " motion .
the whole process is like the following : after the system has wandered in the transient chaotic sea [ see panel ( b ) ] for a long time , the system finally reaches @xmath98 and keeps moving to the left
. then it reaches @xmath99 and continues to move towards the left .
it then passes the central transient chaotic sea " .
later the system has a chance to arrive @xmath100 , followed by @xmath101 .
the system eventually reach a ballistic curve @xmath102 and then keeps moving up in the momentum space . in brief
, it takes 3 stages for a trajectory launched from the regime illustrated in panel ( b ) of fig .
[ fig : phasesinglepoint ] to finally turn to ballistic motion .
first , it wanders highly randomly in a transient chaotic sea .
second , it moves alternatively along some smooth curves [ such as those shown in panel ( a ) and panel ( c ) in fig .
[ fig : phasesinglepoint ] ] , between which the system returns to the transient chaotic sea , but with the overall tendency towards curves of large momentum values .
lastly , the system evolves on a simple ballistic structure . .
the total evolution time is @xmath103 , by which time this trajectory becomes clearly ballistic .
the whole process is remarkably complicated .
panels ( a ) ( b ) and ( c ) zoom in some small portions of the shown phase space , with the momentum values in the middle-@xmath14 regime seen to be highly localized .
green dash - dotted arrow shows the moving direction of the trajectory along the curves " traced by the trajectory .
these curves shown here emerges after @xmath104 , and similar curves developed in earlier time with smaller @xmath30 values .
note that these curves are not kam invariant curves as the system will eventually leave them .
we define the motion along these curves as transient chaotic motion " . ] in our simulations , we always choose @xmath105 phase space points randomly sampled from the following gaussian distribution @xmath106 note that this gaussian distribution is analogous to the initial gaussian wavepacket we used in our quantum dynamics calculations .
we then evolve this ensemble of classical trajectories according to the relativistic kr map described by eq . and examine the ensemble averaged @xmath40 as a function of @xmath0 , i.e. , the number of iterations .
interestingly , the time dependence of @xmath40 is rich , and if we fit @xmath40 by the power - law @xmath107 for appropriate time windows , the exponent @xmath108 can be larger than 2 .
in fact , sometimes @xmath108 can be even larger than three or even four .
some examples are shown in fig .
[ fig : ksm10 ] . in one case ,
the superballistic transport exponent is found to be as large as @xmath109 , for a time window from @xmath110 to @xmath111 . on the one hand
this confirms that the classical dynamics may display superballistic transport , on the other hand it is necessary to better understand the underlying mechanism . by exploring many parameter choices
, it is found that breaking the global kam curves with an increasing ratio @xmath112 is a necessary condition .
this is already a clear difference from quantum superballistic transport .
for example , for the results shown in fig . 2 , @xmath113 , @xmath45 , @xmath59 , quantum superballistic transport occurs already for @xmath44 , but in the classical case shown in fig .
[ fig : ksm10 ] , superballistic transport is observed only when @xmath12 exceeds 1.4 .
thus , in the quantum case , a global kam invariant curve does not forbid the population leakage from the classically chaotic regime to the ballistic regime , an indication of quantum tunneling .
vs time ( the number of kicks ) .
the initial conditions are sampled from a gaussian distribution described by eq . with @xmath114 and @xmath115 , so that the initial phase space distribution is analogous to the quantum initial state used in fig .
[ fig : m_ra_log ] .
here @xmath113 , @xmath45 , @xmath59 , and from top to bottom , @xmath116 , and @xmath117 .
note that with other system parameters being the same , quantum superballistic transport already occurs for @xmath44 ( see fig .
[ fig : m_ra_log ] ) . here
the classical superballistic transport emerges until @xmath12 reaches 1.4 . ]
to further understand the numerical results , we find it necessary to also account for normal diffusion as the trajectories seek to land on ballistic trajectories from the chaotic sea .
the associated normal diffusion rate is assumed to be @xmath118 .
we further assume that the leakage rate from the chaotic sea to the ballistic structure is given by @xmath67 . then analogous to eq .
( [ psquare ] ) , we expect to have @xmath119 where @xmath120 , with @xmath121 being the fraction of trajectories confined in some local stable islands in regime i and @xmath122 being their average momentum spread ; @xmath123 , with @xmath124 representing the fraction of trajectories undergoing normal diffusion , with @xmath118 being the associated diffusion constant ; @xmath125 is the faction of trajectories initially placed on ballistic structures , with @xmath70 being the diffusion coefficient
. note that , unlike in the quantum case , at this point we do not first assume a constant probability leakage rate because , as seen below , the leakage involves different behavior at different time windows , and so @xmath67 can be rather complicated .
indeed , as seen from fig .
[ fig : phasesinglepoint ] , once the global kam curves are destroyed , the escape from a ( transient ) chaotic sea to a ballistic structure is extremely complicated : the boundary between them is hard to identify and different initial conditions sampled from an initial gaussian ensemble may need drastically different times to reach a ballistic structure .
the coexistence of normal diffusion , ballistic transport , and the potentially complicated leakage rate @xmath67 makes the time dependence of @xmath126 even more interesting than the quantum case .
let us roughly define a regime @xmath127 $ ] in the phase space , where trajectories are not doing ballistic motion . in connection with our observations made from fig .
[ fig : phasesinglepoint ] , we choose @xmath128 .
let @xmath129 be the occupation probability of this regime and @xmath130 be the occupation probability on ballistic structures .
then @xmath131 .
@xmath129 is plotted in fig .
[ fig : ksm10probleak_combine ] on either linear or logarithmic scales , for different time windows , for the value @xmath132 already studied in fig .
[ fig : ksm10 ] . the right inset of fig .
[ fig : ksm10probleak_combine ] indicates that initially most trajectories are trapped in the regime @xmath127 $ ] , until @xmath133 .
then @xmath129 starts to decrease more appreciably , with a time dependence not easy to fit [ see the first part of the curve in panel ( c ) ] .
after @xmath134 kicks however , the relation between @xmath129 and @xmath0 becomes much more evident , i.e. @xmath135 , which indicates an exponential decay .
the emergence of an exponential decay suggests that the ensemble has reached a certain steady configuration , as the escape probability now becomes proportional to the occupation probability itself .
as also shown by the bottom inset and by the main figure of fig .
[ fig : ksm10probleak_combine ] , the coefficient @xmath71 slightly changes with time .
expanding such an exponential decay to the first order , this escape would amount to an almost constant leakage rate of jumping onto phase space ballistic structures . as a result
, one would naively expect , like our analysis in the quantum part , a superballistic transport case with @xmath65 .
this prediction is certainly oversimplified as compared with our actual results shown in fig .
[ fig : ksm10 ] . vs time , where @xmath136 defined in the text represents the occupation probability in a non - ballistic regime . for the two insets ( a ) and ( b )
, @xmath129 vs time is plotted for two early stages , where it can be seen that initially most trajectories are trapped in the initial non - ballistic regime for @xmath137 .
then @xmath129 starts to decrease appreciably in a nonlinear fashion .
the inset ( c ) shows that the time dependence of @xmath138 is highly nonlinear before @xmath139 .
however , after @xmath140 , @xmath138 and @xmath0 display a linear relation .
the system parameters are the same as the case of @xmath132 in fig .
[ fig : ksm10 ] . ] to better digest the results shown in fig .
[ fig : ksm10 ] , we again focus on the case @xmath132 in connection with the time dependence of @xmath129 in fig .
[ fig : ksm10probleak_combine ] . in the very
beginning , @xmath129 remains almost a constant until @xmath141 kicks , so for this time period @xmath67 is essentially zero .
therefore initially only the first three terms in eq .
are non - zero , suggesting that the diffusion exponent of @xmath40 should be less than two .
this explains the actual numerical result during the early stage .
the plotted curve in fig . [ fig : ksm10 ] at early times also has an increasing slope .
this can be explained as follows . during the early stage , we have @xmath142 . when @xmath143 , the @xmath144 term starts to dominate
so @xmath40 is close to normal diffusion .
similarly , when @xmath145 , the @xmath146 term exceeds the first two , so we have a behavior close to ballistic transport for a quite long period until @xmath134 . on the other hand , from fig .
[ fig : ksm10probleak_combine ] , it is observed that since as early as @xmath133 , @xmath67 is already non - zero .
so the leakage to the ballistic regime is building up long before an exponential leakage is observed at @xmath147 .
this early - stage leakage to the ballistic regime starts to affect the time dependence of @xmath40 only until @xmath148 .
we conjecture that this is the reason why in fig .
[ fig : ksm10 ] a simple relation @xmath149 is not observed .
in addition , the lack of such a simple superballistic behavior with @xmath65 is also consistent with the apparent nonlinear time dependence shown in panel ( c ) of fig .
[ fig : ksm10probleak_combine ] before @xmath139 .
$ ] vs @xmath150 , an analysis motivated by a reset of the start time at @xmath139 ( see the text for the details ) .
the result here mainly displays a normal diffusion stage and a superballistic transport stage with an exponent close to @xmath65 .
the computational example presented here is the same one in fig .
[ fig : ksm10probleak_combine ] with @xmath132 .
the dashes lines represent power - law fitting . ] to confirm our qualitative analysis above , we now redefine the start time as the point when an exponential decay of @xmath129 can be clearly identified . again using the computational example shown in fig .
[ fig : ksm10probleak_combine ] , we now use @xmath140 as the start time to count change in the momentum spread .
that is , we now examine @xmath151 $ ] . because at @xmath152 , the population inside the regime @xmath127 $ ] is about 0.88
, we have @xmath153 $ ] .
then we have @xmath154 & \approx 0.88\ d_0 e^{-\gamma ( t - t_2 ) } ( t - t_2 ) \\ &
\quad + 0.12\ a ( t - t_2)^2 \\ & \quad + a \int^t_{t_2 } { r(t ' ) ( t - t')^2 { \mathrm{d}}t ' } , \end{split}\ ] ] where the first term account for the normal diffusion as the trajectories diffuse from a chaotic sea to eventually land on a ballistic structure , the second term describes the ballistic transport for those trajectories already outside the regime @xmath127 $ ] at the start time @xmath152 , and the last term describes the impact on the transport dynamics due to the population leakage from the regime @xmath127 $ ] , with @xmath155 \approx 0.88\ \gamma.\end{aligned}\ ] ] equation ( [ psquare2_ciii ] ) thus suggests that once we reset the start time at @xmath152 , there should be a normal diffusion stage , a transition stage due to the second term , followed by a superballistic transport period , i.e. , @xmath156 \sim ( t - t_2)^3 $ ] . in fig .
[ fig : k1_4m10_truncated ] we present a numerical log - log plot of @xmath151 $ ] vs @xmath150 , in very good agreement with our analysis . as a final note , the classical superballistic transport shown in fig .
[ fig : k1_4m10_truncated ] with the diffusion exponent @xmath157 lasts very long , but this behavior can not last forever . in the end ,
almost all trajectories from the initial ensemble will end up on ballistic structures and then purely ballistic transport will take over . returning to an early study @xcite of the classical relativistic kr under the resonance condition
, we have to disagree with some of their statements ; from our results we conclude that classical superballistic transport was not observed there because the investigation time scale there was too short .
in this work , we show that both quantum and classical superballistic transport can occur in a simple periodically driven system , namely , a relativistic kicked - rotor system with a nonzero mass term . to our knowledge , this appealing scenario has not been discussed before . compared with previous lattice - junction models for quantum superballistic transport ,
the superballistic transport in our model occurs in momentum space as a consequence of a natural divide imposed by the relativistic dispersion : regions with low momentum effectively have a quadratic ( bare ) dispersion relation ( hence effectively a quasi - random on - site potential ) and regions of high momentum effectively have a linear ( bare ) dispersion relation ( hence effectively a quasiperiodic potential ) .
remarkably , though found in the same dynamical system , the quantum superballistic and classical superballistic transport we have analyzed are observed in much different parameter regimes .
indeed , in the quantum case , the mechanism lies in the leakage of the quantum state from a regime of dynamical localization to a regime of ballistic transport .
this leakage can occur even when the underlying classical limit has global kam invariant curves separating the two regimes . in the classical case ,
it is necessary to break the global kam curves first to allow for leakage from a chaotic sea to ballistic trajectories . as a side result
, we find that this kind of leakage in the classical dynamics is unexpectedly complicated and further studies can be motivated . for example , strictly speaking , the random patterns shown in the panel ( b ) of fig .
[ fig : phasesinglepoint ] do not represent chaos ( chaos is defined as a positive lyapunov exponent in the asymptotic long - time limit , but this trajectory will eventually become ballistic and hence has a zero lyapunov exponent ) .
the detailed characteristics of this type of irregular trajectory eventually becoming a regular ballistic one deserve more attention .
the issue of quantum - classical correspondence concerning this type of trajectories is also of considerable interest for future studies .
the classical relativistic kicked rotor model may be realized by considering relativistic electrons moving in the field generated by a special electrostatic wavepacket @xcite . on the quantum side , a spinless version of the relativistic kicked rotor
may be also realized by considering a kicked tight - binding lattice whose on - site potential can be determined by the relativistic dispersion relation @xcite .
however , due to the large time scales involved , a direct observation of our numerical results reported here is unlikely . as such it should be interesting enough to explore the system more to identify other signatures of superballistic transport at shorter time scales .
99 b. d. hughes , random walks and random environments ( clarendon press , oxford , 1995 ) .
y. he , s. burov , r. metzler , and e. barkai , phys .
* 101 * , 058101 ( 2008 ) .
r. metzler and j. klafter , phys . rep . *
339 * , 1 ( 2000 ) .
j. p. bouchaud and a. georges , phys
* 195 * , 127 ( 1990 ) .
b. i. henry , t.a.m .
langlands , and p. straka , in complex physical , biophysical and econophysical systems , world scientific lecture notes in complex systems , edited by r. l. dewar and f. detering ( world scientific , singapore , 2010 ) , vol . 9 .
j. wang , i. guarneri , g. casati , and j. b. gong , * 107 * , 234104 ( 2011 ) ; h. l. wang , j. wang , i. guarneri , g. casati , and j. b. gong , * 88 * , 052919 ( 2013 ) .
s. abe and h. hiramoto , phys .
a * 36 * , 5349 ( 1987 ) ; h. hiramoto and s. abe , j. phys .
* 57 * , 230 ( 1988 ) ; * 57 * , 1365 ( 1988 ) .
f. pichon , phys .
. lett . * 76 * , 4372 ( 1996 ) .
r. ketzmerick , k. kruse , s. kraut , and t. geisel , phys .
lett . * 79 * , 1959 ( 1997 ) .
k. l and j .- d .
bao , phys .
e * 76 * , 061119 ( 2007 ) .
p. siegle , i. goychuk , p. talkner , and p. hnggi , phys .
e * 81 * , 011136 ( 2010 ) ; p.
siegle , i. goychuk , and p. hnggi , phys .
lett . * 105 * , 100602 ( 2010 ) ; p. siegle , i. goychuk , and p. hnggi , europhys . lett . *
93 * , 20002 ( 2011 ) .
l. levi , y. krivolapov , s. fishman and m. segev , nat .
* 8 * , 912 ( 2012 ) .
l. hufnagel , r. ketzmerick , t. kottos , and t. geisel , phys .
e 64 , 012301 ( 2001 ) .
z. j. zhang , p. q. tong , j. b. gong , and b. w. li , , 070603 ( 2012 ) . s. sttzer , t. kottos , a. tnnermann , s. nolte , d. n. christodoulides , and a. szameit , opt .
lett . * 38 * , 4675 ( 2013 ) .
a. a. chernikov , t. tl , g. vattay , and g. m. zaslavsky , phys .
a * 40 * , 4072 ( 1989 ) .
d. u. matrasulov , g. m. milibaeva , u. r. salomov , and b. sundaram , phys .
e * 72 * , 016213 ( 2005 ) .
g. casati and b. v. chirikov , quantum chaos : between order and disorder ( cambridge university press , new york , 1995 ) .
d. r. grempel , s. fishman , and r. e. prange , phys .
49 * , 833 ( 1982 ) .
r. e. prange , d. r. grempel , and s. fishman , phys .
b * 29 * , 6500 ( 1982 ) . m. v. berry , physica d * 10 * , 369 ( 1984 )
. b. simon , annals of physics * 159 * , 157 ( 1985 ) .
y. nomura , y. h. ichikawa , and w. horton , phys .
a * 45 * , 1103 ( 1992 ) .
s. fishman , d. r. grempel , and r. e. prange , phys .
* 49 * , 509 ( 1982 ) .
p. mello , p. pereyra , and n. kumar , ann .
phys . * 181 * , 290 ( 1988 ) .
s. n. evangelou and t. ziman , journal of physics c : solid state physics * 20 * , l235 ( 1987 ) .
d. n. sheng and z. y. weng , phys .
b * 54 * , r11070 ( 1996 ) .
j. b. gong and j. wang , phys .
e * 76 * , 036217 ( 2007 ) .
t. boness , s. bose , and t. s. monteiro , phys .
* 96 * , 187201 ( 2006 ) . | as an unusual type of anomalous diffusion behavior , superballistic transport is not well known but has been experimentally simulated recently .
quantum superballistic transport models to date are mainly based on connected sublattices which are constructed to have different properties . in this work ,
we show that both quantum and classical superballistic transport in the momentum space can occur in a simple periodically driven hamiltonian system , namely , a relativistic kicked - rotor system with a nonzero mass term .
the nonzero mass term essentially realizes a junction - like scenario : regimes with low or high momentum values have different dispersion relations and hence different transport properties .
it is further shown that the quantum and classical superballistic transport should occur under much different choices of the system parameters .
the results are of interest to studies of anomalous transport , quantum and classical chaos , and the issue of quantum - classical correspondence . |
unruh @xcite in 1981 suggested a theoretical method based on the fact that , considering the sound wave motion , a change of a subsonic runoff for supersonic flow forms an event horizon analogous to an event horizon of a gravitational black hole .
since then , the study of analog models of gravity @xcite has become an important field to investigate the hawking radiation as well as to improve the theoretical understanding of quantum gravity . for such analog models
there are many examples , so we highlight gravity wave @xcite , water @xcite , slow light @xcite , optical fiber @xcite and electromagnetic waveguide @xcite . specially in fluid systems , the propagation of perturbations of the fluid has been analyzed in many analog models of acoustic black holes , such as the models of superfluid helium ii @xcite , atomic bose - einstein condensates @xcite and one - dimensional fermi degenerate noninteracting gas @xcite that were elaborated to create a sonic black hole geometry in the laboratory .
more specifically , in a quantum liquid the quasiparticles are bosons ( phonons ) in @xmath0he and bosons and fermions in @xmath1he which move in the background of effective gauge and/or gravity simulated by the dynamics of collective modes .
these quasiparticles are analogs of elementary particles of low - energy effective quantum field theory .
in particular , phonons propagating in the inhomogeneous liquid can be described by the effective lagrangian mimicing the dynamics of an scalar field in the curved spacetime given by the effective acoustic metric where the free quasiparticles move along geodesics . in superfluid @xmath1he -
a , the effective quantum field theory contains chiral fermion quasiparticles where the collective bosonic modes interact with these ` elementary particles ' as gauge fields and gravity .
these advances are important for a better insight into the understanding of quantum gravity .
in addition , the study of a relativistic version of acoustic black holes was presented in @xcite .
furthermore , the acoustic black hole metrics obtained from a relativistic fluid in a noncommutative spacetime @xcite and lorentz violating abelian higgs model @xcite have been considered .
the thermodynamics of acoustic black holes in two dimensions was studied in @xcite .
the authors in @xcite studied acoustic black holes in the framework of neo - newtonian hydrodynamics and in @xcite the effect of neo - newtonian hydrodynamics on the superresonance phenomenon was analyzed . in 1959 , aharonov and bohm showed that when the wave function of a charged particle passing around a region with the magnetic flux , despite the magnetic field being negligible in the region through which the particle passes , experiences a phase shift as a result of the enclosed magnetic field @xcite .
the aharonov - bohm ( ab ) effect , that is essentially the scattering of charged particles , has been used to address several issues in planar physics and it was experimentally confirmed by tonomura @xcite for a review see @xcite .
the effect can also be simulated in quantum field theory as for example by using a nonrelativistic field theory for bosonic particles which interact with a chern - simons field @xcite .
more specifically , in @xcite was addressed the ab effect considering the noncommutative spacetime and in @xcite the effect was obtained due to violation of lorentz symmetry in quantum field theory .
furthermore , several other analogs of the ab effect were found in gravitation @xcite , fluid dynamics @xcite , optics @xcite and bose - einstein condensates @xcite . in
@xcite it was shown that the background flow velocity @xmath2 plays the role of the electromagnetic potential @xmath3 , and the integrated vorticity in the core @xmath4 plays the role of the magnetic flux @xmath5 . thus , surface waves on water crossing an irrotational ( bathtub ) vortex experience an analog of the ab effect .
also , the gravitational analog of the electromagnetic ab effect which is purely classical is related to the particles constrained to move in a region where the riemann curvature tensor vanishes . however , a gravitational effect arises from a region of nonzero curvature from which the particles are excluded .
an interesting system was investigated in @xcite , where it was shown that planar waves scattered by a draining bathtub vortex develops a modified ab effect that has a dependence on two dimensionless parameters related to the circulation and draining rates @xcite .
it has been found an inherent asymmetry even in the low - frequency regime which leads to novel interference patterns .
more recently in @xcite , were extended the analysis made in @xcite to a lorentz - violating and noncommutative background @xcite which allows to have persistence of phase shifts even if circulation and draining vanish .
one of the theoretical methods of investigating the underlying structure of the spacetime is studying the solution of field equations for the fermion fields , besides the bosonic ones , in a curved geometry .
the behavior of matter fields in the vicinity of black holes results in the better understanding of their properties .
fermion fields were analyzed in a kerr black hole , in the near horizon limit , as well as in the case of reissner - nordstrm black hole @xcite . in @xcite
non - zero dirac fermion modes in the spacetime of black hole cosmic string system was considered .
the authors studied the near - horizon behavior of fermion fields which results in superconductivity in the case of extremal charged dilaton black hole .
although the effective sonic black hole spacetime is not the one fermion fields would observe , but it can be used as a toy model and a mathematical tool to study and as a result understand better the dynamics of the massive and massless fermion fields in a gravitational rotating black hole as well as a vortex background and shed light on the underlying physics . in the present study
we consider the dynamics of a massive dirac spinor field in a curved spacetime and apply the acoustic black hole metric to obtain the differential cross section for scattered planar waves which leads to an analog ab effect . besides that , we compare the results with the massless scalar field case .
the paper is organized as follows . in sec .
[ ii ] we briefly introduce the acoustic black hole . in sec .
[ iii ] we compute the differential cross section due to the scattering of planar waves that leads to an analog ab effect . finally in sec .
[ sec : conc ] we present our final considerations .
the acoustic line element in polar coordinates is governed by @xmath6 where @xmath7 is the velocity of sound in the fluid and @xmath8 is the flow / fluid velocity .
we consider the flow with the velocity potential @xmath9 whose the flow / fluid velocity is given by @xmath10 thus , equation ( [ eq1 ] ) can be rewritten as follows @xmath11 where @xmath12 and @xmath13 are the constants of the circulation and draining rates of the fluid flow .
the radius of the ergosphere given by @xmath14 , and horizon , the coordinate singularity , given by @xmath15 are @xmath16 we set @xmath17 , and choose the following change of variables @xcite @xmath18 where @xmath19 , and which results in the following line element @xmath20 d\tilde{t}^{2}-\frac{dr^{2}}{f(r)}-r^2 d\tilde{\phi}^{2}+c\left(d\tilde{t } \
d\tilde{\phi}+d\tilde{\phi } \
d\tilde{t}\right ) .
\end{aligned}\ ] ] therefore , the metric can be written in the form @xmath21 with the inverse @xmath22
@xmath23 we know that this effective metric is acoustic which describes the propagation of sound waves .
it is trivial that the dirac field does not obey the acoustic metric . in the case it exists in the considered media
, it would have its own metric with its own speed of light " .
we are interested in the dynamics of the fermion field in this background as a toy model to investigate the physical properties in contrast with the scalar field case . in the following ,
we study the scattering of monochromatic planar waves of frequency @xmath24 for a massive fermion by the draining vortex , a process governed by two key quantities ; circulation and draining .
now let us consider a spinor filed in the background of the sonic black hole .
the dynamics of a massive spinor field in curved spacetime is described by the dirac equation @xmath25where @xmath26 , @xmath27 are the dirac matrices in curved spacetime and @xmath28 are the spin connections .
let us choose the following representation @xmath29 for the geometry at hand , using the relation @xmath30 , the gamma matrices take the form @xmath31 one can find the triad coordinate using @xmath32 as @xmath33 where @xmath34 the only nonzero christoffel symbols are [ cols= " < , < " , ] now , using @xmath35 one can find the spin connection components for the system as follows @xmath36 in the dirac equation there appears @xmath37 now , we substitute the following spinor field in the dirac equation @xmath38 where @xmath39 . replacing the corresponding parameters in dirac equation , we obtain @xmath40\psi_1+i\left[\sqrt{f } \
\partial_r+\frac{\left(j-1/2\right)}{r}+\left(\frac{d^2}{2r^3\sqrt{f}}+\frac{(1+\sqrt{f})}{2r}\right)\right]\psi_2=0 \ , \label{dirac - eqn1}\ ] ] @xmath41\psi_2+i\left[\sqrt{f } \
\partial_r-\frac{\left(j+1/2\right)}{r}+\left(\frac{d^2}{2r^3\sqrt{f}}+\frac{(1+\sqrt{f})}{2r}\right)\right]\psi_1=0 \ .
\label{dirac - eqn2}\ ] ] after decoupling the above equations for @xmath42 and @xmath43 and considering the change of variables as @xmath44 and @xmath45 , we get @xmath46\right]\chi_1=0 \ , \label{sec1}\end{aligned}\ ] ] @xmath47\right]\chi_2=0 \ , \label{sec2}\end{aligned}\ ] ] written as a power series in @xmath48 and in @xmath49 , with @xmath50 , @xmath51 , @xmath52 and @xmath53 @xmath54 in equations ( [ sec1 ] ) and ( [ sec2 ] ) , @xmath55 and @xmath56 are @xmath57,\ ] ] @xmath58.\ ] ] choosing @xmath59 , @xmath60 and @xmath61 , we have @xmath62 @xmath63 therefore for large @xmath64 , ignoring the terms @xmath65 $ ] and @xmath66 $ ] , one can obtain analytic solutions @xmath67 in the limit @xmath68 , these two solutions converge to @xmath69 it is not difficult to show numerically that these solutions besides satisfying the second order equations ( [ sec1 ] ) and ( [ sec2 ] ) , satisfy the first order equations ( [ dirac - eqn1 ] ) and ( [ dirac - eqn2 ] ) in the limit @xmath68 . comparing this result with the one in the minkowski space , we obtain the following approximate expressions for the phase shift for the upper and lower components of the spinor fields @xmath70 where @xmath71 . for @xmath72
, one can expand the above expressions .
thus , we obtain @xmath73}{4n}-\frac{\pi \alpha\left[-\alpha+\alpha^2+\beta+\gamma^2\right]}{4n^2}+\mathcal{o}[\frac{1}{n^3}]\right]\\ \delta_d & = \frac{m}{|m|}\left[-\frac{\pi \alpha}{2}+\frac{\pi \left[\alpha+\alpha^2+\beta+\gamma^2\right]}{4m}-\frac{\pi \alpha\left[\alpha+\alpha^2+\beta+\gamma^2\right]}{4m^2}+\mathcal{o}[\frac{1}{m^3}]\right ] \ .\end{aligned}\ ] ] the first term in both expressions , the dominant contribution , is exactly the same as the bosonic case considered in @xcite .
now , we can calculate the differential scattering cross section which is given by @xmath74 where @xmath75 for the @xmath76 when @xmath77 one needs to use the eqs .
( [ deltau]-[deltad ] ) .
therefore , the differential scattering coss section at long wavelengths , @xmath78 , for a massive fermion in the acoustic black hole background is given as follows @xmath79 \ .
\label{cross - section}\end{aligned}\ ] ] obviously , in the case @xmath80 the two expressions for @xmath81 and @xmath82 are the same .
in the absence of the circulation , the differential scattering coss section reduces to @xmath83 which is @xmath84 independent .
this is an expected result , since in the absence of the circulation the only term that contributes in the long wavelength regime to the differential cross section is @xmath77 , polar symmetric term .
the same happens for the bosonic case which is equal to @xmath85 . in the absence of the circulation ,
the result for the fermionic case is exactly the same as the bosonic one in the zero mass limit .
the above expression vanishes when the draining is also zero .
this is also an expected result , because without circulation and draining there is no source for the scattering and we should recover the minkowski result . expanding the expression for the differential scattering cross section in eq .
( [ cross - section ] ) for small angles results in @xmath86 \ , \\
\frac{d\sigma_{\alpha<\beta+\gamma^2}}{d\phi}&=\frac{2 \pi \alpha ^2 } { \lambda \phi ^2}-\frac{\pi \alpha \left(\sqrt{\alpha + \beta + \gamma ^2}+\sqrt{-\alpha + \beta + \gamma ^2}\right)}{\lambda \phi } + \mathcal{o}[\phi^ 0 ] \ .\end{aligned}\ ] ] therefore , for @xmath87 and small @xmath88 the dominant term in the differential scattering cross section is the first term in the above expression .
it means that for small angles the circulation plays more important role compared to the draining . using @xmath89 and @xmath90 , the first term in the above expression
can be expanded as @xmath91\right),\end{aligned}\ ] ] for small @xmath92 . the first term which is the result for massless fermion , matches exactly the one for the massless boson case .
now one can compare the results for the fermionic and bosonic case considering a massless fermion , @xmath93 . to compare these results , we need the same arbitrary energy scale for both cases .
we call this energy scale @xmath94 .
figures [ fig1]-[fig2 ] show the differential cross section for the massless boson and fermion for two cases @xmath95 and @xmath96 with @xmath97 , choosing @xmath98 and also the special case where there is no draining , in the absence of the black hole .
as can be seen , the result for the fermionic case tends to the bosonic one for small angles . besides that , the left graphs in the figs .
[ fig1]-[fig2 ] show that for the limit @xmath99 the differential cross section for the spinor field goes to a nonzero constant in contrast with the boson field . to see this
more closely let us expand eq .
( [ cross - section ] ) for small @xmath24 where @xmath100 , which is the case for the graphs in the figs .
[ fig1]-[fig2 ] , as @xmath101\end{aligned}\ ] ] as can be seen , in the limit @xmath99 , the scattering cross section is equal to @xmath102 for the fermion case which is only dependent on the circulation .
we think this happens due to the fact that the fermionic field has an intrinsic angular momentum which interacts with the angular momentum of the black hole originating from the circulation .
in this paper , we have studied the differential cross section of a massive spinor field in the background of the acoustic black hole spacetime , as a toy model for the gravitational rotating black hole , using the partial wave approach .
we have investigated the scattering of planar waves in a fermionic system by a background vortex as an analog for a rotating black hole .
because of the form of the spinor field which has upper and lower components , we have calculated the phase shifts for these components separately and then averaged them .
we have worked with three dimensionless parameters @xmath103 , @xmath104 and @xmath105 related to the circulation , fermion mass and draining , respectively .
we have obtained the differential cross section at long wavelengths and discussed the limiting cases including small angles and also in the absence of the circulation .
one could see that the dominant contribution in the small angle limit for the fermionic case is exactly the same as the bosonic one when the mass of the fermion is zero .
this dominant contribution comes from the circulation term .
therefore , for small angles the contribution of the circulation is more important than the draing one .
furthermore , considering the cross section in the absence of the circulation shows that the result is the same for the bosonic and fermionic cases when the fermion is massless . in this limit ,
the result is independent of the angle @xmath84 .
the reason for that is in the absence of the circulation the only term that contributes in the long wavelength regime to the differential cross section is a polar symmetric term .
finally , we have shown that in contrast with the bosonic case in the limit @xmath99 , considering @xmath100 , the scattering cross section goes to a nonzero constant equal to @xmath102 for the fermion case which is only dependent on the circulation .
the similar scattering behavior for both boson and fermion for small angles seems to be in accord with both being scattered between the horizon and the ergosphere radius . at this regime
both enjoys the penrose effect of gaining energy after the scattering ( the superresonance effect ) . as figs .
[ fig1]-[fig2 ] ( right panel ) show , this indeed agrees with the increasing of the differential cross section as the frequency increases .
this precisely happens as long as the frequency belongs the interval @xmath106 , where @xmath107 is the azimuthal mode number and @xmath108 is the angular velocity of the black hole @xcite . on the other hand , for large angles ,
both particles tend to be scattered outside the ergoregion and in turn they almost keep the scattering constant as figs .
[ fig1]-[fig2 ] ( left panel ) show .
this is particularly more accurate for fermions .
a detailed analysis of the superresonance effect for fermion fields should be addressed elsewhere . besides that
, in a future work , we plan to study a more realistic model considering the dynamics of the fermionic fields interacting with a black hole , gravitational or an analog in a condensed matter system .
we would like to thank cnpq and capes for partial financial support .
thanks pnpd / capes for the financial support .
g. volovik , _ the universe in a helium droplet _ , oxford university press , 2003 ; l. c. b. crispino , a. higuchi , and g. e. a. matsas , rev .
phys . * 80 * , 787 ( 2008 ) ; m. cadoni , s. mignemi , phys .
d * 72 * , 084012 ( 2005 ) ; m. cadoni , class .
quant . grav .
* 22 * , 409 ( 2005 ) .
l. c. garcia de andrade , phys .
d * 70 * , 64004 ( 2004 ) ; t. k. das , _ transonic black hole accretion as analogue system _ , [ arxiv:0411006 ] ; g. chapline , p.o .
mazur , _ superfluid picture for rotating space - times_,[arxiv:0407033 ] ; o. k. pashaev , j .- h . lee , theor . math
. phys . * 127 * , 779 ( 2001 ) ; s. e. perez bergliaffa , k. hibberd , m. stone , m. visser , physica d * 191 * , 121 ( 2004 ) ; s. w. kim , w.t .
kim , j.j .
oh , phys .
b * 608 * , 10 ( 2005 ) ; xian - hui ge , shao - feng wu , yunping wang , guo - hong yang , you - gen shen , int
d21 * , 1250038 ( 2012 ) ; c. barcelo , s. liberati , m. visser , living rev . rel . * 8 * , 12 ( 2005 ) ; e. berti , v. cardoso , j.p.s .
lemos , phys .
d * 70 * , 124006 ( 2004 ) ; v. cardoso , j.p.s .
lemos , s. yoshida , phys .
d * 70 * , 124032 ( 2004 ) ; peter a. horvathy , mikhail s. plyushchay , phys .
b * 595 * , 547 ( 2004 ) ; m. a. cuyubamba , class .
. grav . * 30 * , 195005 ( 2013 ) .
r. schtzhold and w. g. unruh , phys .
d * 66 * , 044019 ( 2002 ) .
g. rousseaux , c. mathis , p. massa , t. g. philbin , and u. leonhardt , new journal of physics * 10 * , 053015 ( 2008 ) .
u. leonhardt and p. piwnicki , phys .
rev . lett . * 84 * , 822 ( 2000 ) ; u. leonhardt , nature * 415 * , 406 ( 2002 ) ; w. g. unruh and r. schtzhold , phys .
d * 68 * , 024008 ( 2003 ) .
t. g. philbin , c. kuklewicz , s. robertson , s. hill , f. knig , and u. leonhardt , science * 319 * , 1367 ( 2008 ) .
r. schtzhold and w. g. unruh , phys .
* 95 * , 031301 ( 2005 ) .
m. novello , m. visser and g. volovik ( eds . ) , _ artificial black holes _ , world scientific , singapore , ( 2002 ) .
l. j. garay , j. r. anglin , j. i. cirac , and p. zoller , phys .
85 , 4643 ( 2000 ) .
ge , s .- j .
sin , jhep * 1006 * , 087 ( 2010 ) ; n. bilic , class .
* 16 * , 3953 ( 1999 ) ; s. fagnocchi , s. finazzi , s. liberati , m. kormos and a. trombettoni , new j. phys .
* 12 * , 095012 ( 2010 ) ; m. visser and c. molina - paris , new j. phys .
* 12 * , 095014 ( 2010 ) .
m. a. anacleto , m. gomes , a. j. da silva , and d. spehler , phys .
d * 70 * , 085005 ( 2004 ) , phys .
d * 70 * , 129905 ( 2004 ) ; phys .
d * 71 * , 107701 ( 2005 ) , m. a. anacleto , j. r. nascimento and a. yu .
petrov , phys .
b * 637 * , 344 ( 2006 ) .
l. h. ford , a. vilenkin , j. phys .
a * 14 * , 2353 ( 1981 ) ; v. b. bezerra , phys .
d * 35 * , 2031 ( 1987 ) ; v. b. bezerra , j. math .
* 30 * , 2895 ( 1989 ) ; g. a. marques , v. b. bezerra , mod .
a * 19 * , 49 ( 2004 ) ; v. b. bezerra , g. a. marques , and s. g. fernandes , j. math . phys . * 47 * , 072504 ( 2006 ) .
d. bazeia and r. menezes , phys .
d * 73 * , 065015 ( 2006 ) ; a. de souza dutra , m. hott , and f. a. barone , phys .
d * 74 * , 085030 ( 2006 ) ; d. bazeia , m. m. ferreira jr . , a. r. gomes and r. menezes , physica d * 239 * , 942 ( 2010 ) ; a. de souza dutra , r. a. c. correa , phys .
d * 83 * , 105007 ( 2011 ) ; p. r. s. carvalho , phys .
b * 726 * , 850 ( 2013 ) ; r. a. c. correa , roldo da rocha , and a. de souza dutra , ann .
phys . * 359 * , 198 ( 2015 ) ; r. a. c. correa , r. da rocha and a. d. s. dutra , phys .
d * 91 * , 125021 ( 2015 ) . | in this paper we consider the dynamics of a massive spinor field in the background of the acoustic black hole spacetime .
although this effective metric is acoustic and describes the propagation of sound waves , it can be considered as a toy model for the gravitational black hole . in this manner ,
we study the properties of the dynamics of the fermion field in this gravitational " rotating black hole as well as the vortex background .
we compute the differential cross section through the use of the partial wave approach and show that an effect similar to the gravitational aharonov - bohm effect occurs for the massive fermion field moving in this effective metric .
we discuss the limiting cases and compare the results with the massless scalar field case . |
run iia at the tevatron will deliver @xmath11 fb@xmath2 of integrated luminosity with peak luminosity nearly 1.0@xmath3 @xmath4sec@xmath2 by mid-2005 .
the d@xmath0 detector and trigger are performing very well , however aging of the inner silicon tracker and occupancy - related trigger rate issues will become areas of concern by the end of run iia .
the plans for run iib , beginning fall 2005 , are to achieve peak and integrated luminosities of 2.8@xmath3 @xmath4sec@xmath2 and 8 fb@xmath2 , respectively.@xcite during a tevatron shutdown after run iia , the d@xmath0 experiment will complete significant detector and trigger upgrades to deal with the consequences of such an intense beam environment.@xcite@xmath5@xcite of d@xmath0 s various subdetectors,@xcite the inner silicon microstrip tracker ( smt ) will be most severely impacted by the run iib integrated radiation dose .
the innermost of the four smt silicon layers will likely die from radiation damage at @xmath14 fb@xmath2.@xcite the outer central fiber tracker ( cft ) will experience a decrease in tracking efficiency due to high track multiplicities corresponding to run iib peak luminosities .
d@xmath0 employs a three - level trigger architecture . with a current output rate of 1.5 khz , level 1 ( l1 )
consists of fast custom electronics making trigger decisions every 396 ns on the signal output from each subdetector .
the level 2 ( l2 ) systems apply pattern recognition algorithms on the data using generic processors in custom hardware , achieving an 1 khz l2 accept rate .
finally , the level 3 ( l3 ) linux farm runs simplified reconstruction code to make selections from the l2 output for an event - to - tape rate of 50 hz . in run iib , d@xmath0 will see the average number of minimum bias interactions per beam crossing increase from 1 to 5 , and the output from the current l1 system would reach 30 khz .
all three trigger levels will need to be enhanced in order to handle the high occupancies and rates of run iib .
to compensate for the loss of silicon devices in the smt from radiation damage and other aging effects , a new radiation - hard inner silicon layer 0 ( l0 ) will be installed on the beam pipe at r = 1.6 cm.@xcite the additional l0 coverage will help to recover losses in tracking and b - tagging efficiency that result from dead regions , particularly in the inner layer of the run iia smt .
the improvement in impact parameter resolution with l0 compared to the run iia smt with and without its inner layer is shown in fig .
1 . the smaller resolution that results from including l0 hits near the interaction point corresponds to a 15% improvement over run iia b - tagging efficiency . the l0 design utilizes the new svx4 readout chip and a novel grounding scheme which provides excellent noise reduction by covering all carbon fiber support surfaces with a laminated ground mesh.@xcite@xmath5@xcite
the most ambitious trigger system upgrade is the replacement of the entire l1 calorimeter trigger ( l1cal ) .
the new l1cal uses digital filtering to improve resolution on the measurement of transverse energy and a sliding window algorithm to perform better clustering.@xcite these techniques will dramatically sharpen the trigger threshold turn - on curves and bring much of the current l2 rejection up to l1 .
the current central track trigger ( l1ctt ) employs fiber doublet logic to find cft track candidates which are then passed on to the muon and silicon trigger systems .
the upgraded l1ctt will use fiber singlets to reduce track candidate combinatorics and improve fake track rejection at high occupancies .
this will be accomplished with new digital front end boards that include larger fpga s capable of handling more complex singlet logic algorithms .
d@xmath0 will add a completely new trigger system to match calorimeter clusters to cft tracks ( l1caltrk ) .
this system will enhance triggers for higgs searches by improving fake electron rejection by a factor of two and fake tau rejection by ten .
it is a straightforward project which only involves modest modifications to the existing l1 muon - cft track matching system .
the d@xmath0 trigger includes a l2 silicon track trigger ( l2stt ) that finds tracks in the smt .
the l2stt will be expanded with additional silicon trigger boards identical to existing ones .
the additional boards will process the inputs from the new l0 silicon channels .
this extra information enhances pattern recognition crucial to the separation of background from signal topologies .
the l2 systems were designed to use alpha - based single board computers for pattern recognition .
these were recently replaced with more powerful l2@xmath6 processors , and for run iib d@xmath0 will install additional new l2@xmath6 s that can employ more complex algorithms for improved background rejection .
several upgrades to the data acquisition and online computing systems will increase d@xmath0 s capacity to record more high - quality data .
the most significant of these projects will be the addition of 96 linux nodes to the l3 computing farm .
the expansion will effectively double the l3 processing power which will confer the ability to efficiently process the more complex high luminosity run iib events .
in addition , the l2 accept rate will be raised to 1 khz .
the d@xmath0 upgrades for run iib are maintaining a schedule for installation during the tevatron shutdown in 2005 .
all electronics for the trigger upgrades are either in final prototype testing or ordered for production , and the l0 project is ahead of schedule for early production of silicon modules .
significant resources are now being brought to bear on the software and commissioning efforts required to ensure a quick return to physics quality data taking when run iib begins . | recent performance of fermilab s tevatron has exceeded this year s design goals and further accelerator upgrades are underway . the high - luminosity period which follows these improvements
is known as run iib .
the d@xmath0 experiment is in the midst of a comprehensive upgrade program designed to enable it to thrive with much higher data rate and occupancy .
extensive modifications of and additions to all levels of the trigger and the silicon tracker are in progress .
all upgrade projects are on schedule for installation in the 2005 shutdown . |
medical imaging modalities aim to combine high resolution with high contrast between healthy and unhealthy tissues .
optical tomography and electrical impedance tomography display such high contrasts but often suffer from poor resolution . ultrasound tomography and magnetic resonance imaging are high resolution modalities that sometimes suffer from low contrast .
the ultrasound modulation of electrical or optical properties of tissues and the combination of simultaneous electrical and magnetic resonance measurements both offer the possibility to combine high resolution with high contrast . for the acquisition of ultrasound - modulated measurements ,
we refer the reader to , e.g. , @xcite for works in the mathematical literature . for the acquisition of internal information on electrical conductivities by magnetic resonance imaging ,
we refer the reader to , e.g. , @xcite .
mathematically , we aim to reconstruct a scalar diffusion coefficient @xmath8 in an elliptic equation from knowledge of internal information of the form @xmath11 for @xmath2 and @xmath12 , where @xmath3 and @xmath13 are solutions of the elliptic problem with different boundary conditions ; see and below .
coupling impedance ( or diffusion ) with acoustic waves or magnetic resonance correspond to the cases @xmath6 and @xmath7 , respectively .
such information can be obtained from functionals of the form @xmath14 by standard polarization ( expressions of the form @xmath15 ) .
this problem was first solved in the two dimensional setting in @xcite in the case @xmath16 and @xmath6 .
the three dimensional setting was addressed in @xcite with @xmath17 and @xmath6 . in these papers ,
the elliptic equation is recast as a system of equations for quantities of the form @xmath18 using the elliptic equation and the fact that @xmath19 is curl free .
this strategy allows one to eliminate @xmath8 from the system of equations and solve for the vectors @xmath20 .
the stable reconstruction of @xmath8 is then straightforward .
the case @xmath6 in the setting of non - redundant measurements , i.e. , with @xmath21 and measurements of the form @xmath22 is considered in @xcite .
it is shown in that paper that the stable reconstruction of @xmath8 may not be possible from such non - redundant measurements .
this justifies the analysis of redundant measurements .
the objectives of this paper are twofold .
we first generalize the reconstruction of @xmath8 to the case of arbitrary space dimension @xmath23 and almost arbitrary @xmath2 . assuming that the vectors @xmath20 form a frame , we obtain a system of equations for the vectors @xmath20 that involves @xmath24 but no longer @xmath8
. the resulting system of equations may be seen as an overdetermined nonlinear system of equations . by appropriately choosing the boundary conditions used to construct the internal functionals @xmath25 , we obtain a global uniqueness and stability result for the reconstruction of the scalar quantity @xmath0 .
although several portions of the algorithm generalize to the reconstruction of anisotropic diffusion tensors , we restrict ourselves to the scalar case in this paper .
we also describe and investigate the compatibility conditions associated with such a redundant system .
the second objective of the paper is to present a system of elliptic equations for the solutions @xmath3 with constitutive parameters that depend on the measurements @xmath26 but not on the unknown diffusion coefficient @xmath8 .
we show that the system is uniquely solvable when a fredholm alternative holds .
we obtain existence and uniqueness results for the proposed system for all but a discrete number of values of the dimension @xmath23 and the coefficient @xmath2 .
both algorithms require boundary conditions for the elliptic solutions that ensure that @xmath23 of the vectors @xmath20 form a frame in @xmath27 at each point of the domain of interest .
whereas such a condition is easy to meet in two space dimensions , in dimensions three and higher , the only available technique that guarantees such an independence is based on using complex geometrical optics ( cgo ) solutions .
we generalize here the cgo construction of @xcite to the multi - dimensional setting and for almost all values of @xmath10 .
the inverse diffusion problems with internal functionals considered here are examples of hybrid inverse problem where two imaging modalities are combined to provide both high resolution and high contrast .
for recent works on the mathematics of hybrid inverse problems and their many applications in medical imaging , we refer the reader to the articles in the book @xcite and to the recent review paper @xcite .
the rest of the paper is structured as follows .
section [ sec : results ] presents the main results of the paper on the stable reconstruction of @xmath8 from available internal functionals .
the elimination of @xmath8 from the system of equations for the vectors @xmath20 and the corresponding differential calculus is explained in section [ sec : datatof ] .
the redundant system of equations for the vectors @xmath20 is addressed in section [ sec : ode ] while the system of linear equations for the solutions @xmath3 is given in section [ sec : elliptic ] .
finally , section [ sec : comp ] presents further reconstruction algorithms in the two dimensional case and analyzes the compatibility conditions satisfied by the redundant data and their potential use .
let @xmath28 be an open convex bounded domain of @xmath27 with @xmath29 . in the following ,
we address the reconstruction of the scalar conductivity ( or diffusion ) coefficient @xmath8 in the equation @xmath30 where @xmath31 , from knowledge of the interior functionals @xmath32 where @xmath2 is fixed and such that @xmath33 .
the derivation of the internal functionals in the case @xmath34 is detailed in @xcite as examples of synthesized focusing , in @xcite in a setup of temporal , physical , focusing , and in @xcite by considering thermoelastic effects .
the case @xmath7 with @xmath21 related to mreit and cdii ( current density impedance imaging ) is addressed in @xcite . following a similar approach to @xcite
, we first perform the change of unknown functions @xmath35 for every @xmath36 and define @xmath37 we also equip @xmath38 with its euclidean metric @xmath39 in the canonical basis @xmath40 . for a given vector field @xmath41 defined on @xmath28 ,
we define the corresponding one - form @xmath42 ( i.e. , by means of the _ flat operator _ ) . with this notation
, we obtain that the vector fields @xmath43 satisfy the system of equations @xmath44 where @xmath45 and @xmath46 denote the usual exterior product and exterior derivative , respectively .
the first equation stems directly from whereas the second one states that the one - form @xmath47 is exact , therefore closed , and hence @xmath48 .
when @xmath49 , equation is recast as : @xmath50 where in dimension @xmath51 , we define @xmath52 $ ] and @xmath53 , and in dimension @xmath54 , @xmath55 denotes the standard cross - product .
the available information becomes @xmath56 .
a crucial hypothesis for our reconstruction procedure is that the @xmath57 gradients have maximal rank in @xmath27 at every point @xmath58 .
this hypothesis can be formalized by the somewhat stronger statement : there exists a finite open covering @xmath59 of @xmath28 ( i.e. @xmath60 ) , an indexing function @xmath61\ni i\mapsto \tau(i ) = ( \tau(i)_1,\dots,\tau(i)_n)\in [ 1,m]^n$ ] and a positive constant @xmath62 such that @xmath63 this assumption is equivalent to imposing the following condition on the data @xmath64 where @xmath65 stands for the @xmath66 matrix of elements @xmath67 . while one can always find illuminations such that holds in two dimensions with @xmath68 and @xmath69 ( the most preferrable case ) by virtue of ( * ? ? ?
* theorem 4 ) , higher dimensions can be dealt with using complex geometrical optics solutions provided that @xmath8 has enough regularity , as the following lemma shows .
[ lem : cgo ] let @xmath70 and @xmath71 for some @xmath72 be bounded from below by a positive constant .
then * for @xmath23 even , there exists a non - empty open set @xmath73 of illuminations @xmath74 such that for any @xmath75 , the condition holds with @xmath76 for some constant @xmath77 . * for @xmath23 odd , there exists a non - empty set @xmath73 of illuminations @xmath78 such that for any @xmath75 there exists an open cover of @xmath28 of the form @xmath79 and a constant @xmath80 such that @xmath81 for @xmath82 and with @xmath83 .
the first step toward an inversion is to express the source term @xmath84 in terms of a local frame : [ lem : datatof ] let @xmath85 open where , up to renumbering solutions , we have @xmath86 then at every point @xmath87 and denoting @xmath88 , @xmath89 , the vector field @xmath90 is given by the following formulas @xmath91 where @xmath92 denotes the element @xmath93 of the matrix @xmath94 .
formula was first proved in @xcite in the two- and three - dimensional cases with @xmath95 and is here proved for general @xmath23 and @xmath2 such that @xmath96 .
this formula gives us a way to reconstruct @xmath84 locally from @xmath23 linearly independent solutions .
assuming condition , one can then reconstruct @xmath84 globally over @xmath28 . from lemma [ lem : datatof ] , one can follow two directions to reconstruct the conductivity , which we now describe in more detail in the next two paragraphs .
[ [ section ] ] the first approach consists in plugging equation back into the system - and obtain a closed system for the vectors @xmath43 .
we then show that the resulting system leads to a gradient system , which can then be solved for the vectors @xmath43 by ode integration .
once the vectors @xmath43 are reconstructed , one recovers @xmath8 from the knowledge of its value at a given point and the fact that @xmath97 is now known by equation .
this approach is a generalization of the results of @xcite to higher - dimensional settings and general @xmath98 such that @xmath96 , and leads to well - posed reconstructions as stated in the following : [ thm : odestab ] let @xmath99 be an open convex bounded domain , and let two sets of @xmath31 solutions of generate measurements @xmath100 whose components belong to @xmath101 , and who jointly satisfy condition with the same triple @xmath102 .
let also @xmath103 and @xmath104 and @xmath105 be given .
let @xmath8 and @xmath106 be the conductivities corresponding to the measurements @xmath107 and @xmath108 , respectively .
then we have the stability estimate : @xmath109 where @xmath110 is the error committed at the point @xmath111 : @xmath112 the solution for the vectors @xmath20 and then for @xmath113 requires the solution of full gradient equations of the form @xmath114 , where @xmath115 stands for either unknown .
these overdetermined pdes require compatibility conditions on @xmath116 if we wish to ensure that their solution does not depend on the path of integration .
theorem [ thm : odestab ] shows that the reconstruction is unique and stable with respect to the data once a fixed family of integration curves is chosen .
the compatibility conditions addressed in section [ sec : comp ] are shown to depend quadratically on the unknown frame .
it is therefore difficult to enforce them while solving for the frame @xmath117 . nonetheless ,
depending on the value of @xmath10 , they may lead to algebraic ( i.e. pointwise ) reconstructions of all or part of the unknown frame , and may also provide further conditions on the data @xmath26 .
such analyses are carried out in section [ sec : comp ] .
[ rem : rframe ] solving a system of equations for the unknown @xmath20 may not be efficient numerically .
let @xmath117 be the matrix whose columns are the @xmath23 linearly independent vectors @xmath43 at a given @xmath118
. then @xmath119 is known . by the gram - schmidt ( gs )
orthonormalization procedure or by setting @xmath120 , we can write an equation for an oriented orthonormal frame @xmath121 ; see section [ sec : rframe ] below .
this approach requires that we reconstruct @xmath122 scalar functions instead of the @xmath123 components of the vector fields @xmath124 .
the only additional constraint is that the transition matrix from @xmath117 to @xmath121 satisfies a certain stability property with respect to the data @xmath107 , see section [ sec : rframe ] for details .
[ rem : alphazero ] the case @xmath125 , corresponding to information of the form @xmath126 simplifies in the sense that the elimination of @xmath84 is not necessary .
indeed , we show in the next section that knowledge of @xmath127 and the constraints @xmath128 for @xmath129 uniquely determine the vectors @xmath43 provided they are known at one point @xmath111 .
once @xmath19 is known , the reconstruction of @xmath8 may proceed from using .
note that , alternatively , the equation may be seen as a transport equation for @xmath130 when @xmath131 once the vector field @xmath132 is known .
the stability properties of such a reconstruction are established in @xcite .
[ [ section-1 ] ] the second approach is novel and consists in injecting equation back into the initial conductivity equations and obtain a _ strongly coupled elliptic system _ of the form @xmath133 where the vector fields @xmath134 are known from the data and where the illuminations @xmath135 were prescribed in the first place . here and below
, we use the einstein convention of summation over repeated indices .
the vector fields @xmath134 satisfy stability conditions of the form @xmath136 whenever two data sets @xmath107 and @xmath108 jointly satisfy condition with the same triple @xmath137 . after
proving solvability of this system , one is able to reconstruct the functions @xmath3 and then to reconstruct @xmath8 as described below .
uniqueness and stability of the solution to with respect to the drift fields @xmath134 relies on the fact that @xmath138 is not an eigenvalue of the operator @xmath139 defined by @xmath140_i \bfe_i = \delta_d^{-1 } ( w_{ij}\cdot\nabla v_j)\ \bfe_i , \label{eq : pw}\end{aligned}\ ] ] where @xmath141 denotes the inverse of the dirichlet laplacian on @xmath28 , and where we have defined the space @xmath142^m$ ] , which makes @xmath143 hilbert once equipped with the norm @xmath144 when the coefficients @xmath134 are bounded , we show that the operator @xmath145 is compact and its operator norm satisfies the estimate @xmath146 ( see lemma [ lem : pw ] ) , where @xmath147 denotes the operator norm of @xmath148 . as a consequence ,
the system satisfies a fredholm alternative which will provide uniqueness and stability as stated in the next proposition , for all @xmath2 when @xmath51 , and for all @xmath10 but possibly a discrete set ( possibly converging to @xmath149 ) in the interval @xmath150 $ ] whenever @xmath70 .
[ prop : elliptic ] let vector fields @xmath151 belong to @xmath152 and such that @xmath153 is an eigenvalue of neither @xmath145 nor @xmath154 .
let @xmath155 be the unique solutions to with same illumination @xmath156 and respective drift terms @xmath157 , @xmath158 .
then we have that @xmath159 and satisfies the stability estimate @xmath160 in the case @xmath51 or ( @xmath125 with @xmath161 ) , we can recast as a coercive system in divergence form , the injectivity of which follows immediately .
these cases correspond to @xmath162 .
once the solutions @xmath3 are reconstructed , one may reconstruct @xmath8 using a formula of the form @xmath163 .
however , such a formula may not offer the best stability estimates .
another reconstruction strategy is deduced from , which can be recast locally as @xmath164 where @xmath165 denote the @xmath23 linearly independent gradients . as in the ode - based reconstruction procedure
, we can devise an ode - based algorithm to reconstruct @xmath8 locally from formula .
we then arrive at the following stability result .
[ thm : elliptic ] let the conditions of proposition [ prop : elliptic ] be satisfied .
then the corresponding @xmath166 satisfy the estimate @xmath167 note that a necessary and sufficient condition for the unique solvability of on a simply connected domain is that the exterior derivative of the right - hand side ( seen as a one - form ) vanish by an application of the poincar lemma .
the compatibility condition that arises here takes the form of a quadratic equation in the components of @xmath19 that is difficult to ensure as it depends on the unknowns .
defining geometric notation for now , let us first denote the euclidean orthonormal frame @xmath168 and @xmath169 . for @xmath170 , @xmath171 denotes the space of @xmath172 forms .
we recall the definition of the _ hodge star operator _
@xmath173 for @xmath170 , such that for any elementary @xmath174-form @xmath175 with @xmath176 , we have @xmath177 we recall the following useful identities , see e.g. , @xcite : @xmath178 we now prove lemma [ lem : datatof ] which is the cornerstone of our explicit reconstructions . because @xmath179 is a basis of @xmath27 at any point @xmath58 , a vector @xmath180 can be represented in this basis by the following representation ( @xmath118 is implicit here ) @xmath181 for @xmath182 , let us introduce the following @xmath183-forms : @xmath184 where the hat indicates an omission and @xmath185 is the signature of the permutation @xmath186 . at each @xmath87 , the vector @xmath187 obtained from @xmath188 by `` raising an index ''
can also be seen as the unique vector obtained by the riesz representation lemma that corresponds to the linear form @xmath189 such that for any @xmath190 , @xmath191 we now show that the vector fields @xmath192 satisfy a simple divergence equation .
we compute @xmath193 and using the identity @xmath194 , we deduce @xmath195 the decomposition of @xmath192 in the basis @xmath196 may be obtained by computing its dotproducts with @xmath197 .
indeed , for @xmath198 , there is an @xmath199 such that @xmath200 and we have @xmath201 by repetition of the term @xmath202 in the determinant . now
if @xmath203 , we have @xmath204 using formula , we deduce that @xmath192 admits the expression @xmath205 plugging this expression into equation , and using @xmath206 , we obtain @xmath207 finally using the representation for @xmath84 itself yields @xmath208 we can also recast the previous expression as follows @xmath209 = c_f \left [ \nabla\log d + ( ( \nabla h^{ij})\cdot s_i ) s_j \right ] , \label{eq : fnd2}\end{aligned}\ ] ] and the proof is complete .
we now give a proof of lemma [ lem : cgo ] , which guarantees the existence of illuminations that ensure condition and thus justifies the two global reconstruction approaches .
the cgo constructions , introduced in @xcite in this context , generalize those defined in @xcite .
since @xmath8 is bounded from above and below by positive constants , it suffices to study the case @xmath34 since we have for any @xmath210 , @xmath211 consider the problem @xmath212 on @xmath27 with @xmath0 extended in a continuous manner outside of @xmath28 and such that @xmath8 equals @xmath183 outside of a large ball .
the construction requires sufficient smoothness of @xmath8 in order to be valid .
let @xmath213 on @xmath27 .
we assume that @xmath214 , which holds if @xmath215 for some @xmath72 , i.e. , the original @xmath216 .
note that by sobolev imbedding , @xmath8 is of class @xmath217 while @xmath218 is of class @xmath219 . with the above hypotheses
, we can apply ( * ? ? ?
* corollary 3.2 ) which states the following .
let @xmath220 so that @xmath221 on @xmath27 .
let @xmath222 be of the form @xmath223 with @xmath224 , and @xmath225 .
thus , @xmath226 satisfies @xmath227 and @xmath228 is a harmonic complex plane wave ( hence the name of complex geometrical optics solutions ) .
now , it is shown in @xcite , following works in @xcite , that @xmath229 with @xmath230 and hence @xmath231 in @xmath27 .
we have used again the sobolev imbedding stating that functions in @xmath232 are of class @xmath233 for a bounded domain @xmath234 . taking gradients of the previous equation and rearranging terms , we obtain that @xmath235 because @xmath236 is bounded and @xmath237 in @xmath219 , the @xmath238-valued function @xmath239 satisfies @xmath240 independent of @xmath226 . moreover , the constant @xmath241 is in fact independent of @xmath8 provided that the norm of the latter is bounded by a uniform constant in @xmath242 .
both the real and imaginary parts of @xmath243 , denoted @xmath244 and @xmath245 , count as solutions of the free - space conductivity equation , thus @xmath246 and @xmath247 can serve as vectors @xmath20 .
more precisely , we have @xmath248 [ [ section-2 ] ] set @xmath249 , define @xmath250 for @xmath251 , and construct @xmath252 using , we obtain that @xmath253 where @xmath254 .
letting @xmath255 so large that @xmath256 and denoting + @xmath257 , we have @xmath258 .
we conclude after the next paragraph .
[ [ section-3 ] ] set @xmath259 , define @xmath250 for @xmath260 , and @xmath261 and construct @xmath252 using , we obtain that @xmath262 where @xmath263 for @xmath264 .
letting @xmath255 so large that @xmath265 and denoting @xmath266 , we have that @xmath267 where @xmath57 is a signed integer .
since the previous sets are open and a finite number of them covers @xmath28 ( because @xmath28 is bounded and @xmath255 is finite ) , we therefore have fulfilled the desired requirements of the construction . upon changing the sign of @xmath268 or @xmath269 on each of these sets if necessary
, we can assume that the determinants are all positive .
[ [ section-4 ] ] in each of the previous cases , let @xmath270 be the traces of the solutions defined above with @xmath271 . these illuminations generate solutions that satisfy the desired properties of maximal rank and positive determinants . by continuity arguments , any boundary conditions @xmath272 in an open set sufficiently close to @xmath273 will ensure that the maximum of the determinants stay bounded from below by @xmath80 .
this concludes the proof of the lemma .
in this section , we extend the results presented in @xcite to general dimension and for a more general class of measurements ( described by the coefficient @xmath10 ) .
we first need to introduce standard geometric notation , without which the derivations become quickly intractable .
we work on a convex set @xmath274 with the euclidean metric @xmath275 on @xmath27 .
following @xcite , we denote by @xmath276 the euclidean connection , i.e. the unique connection that is torsion - free , and compatible with the euclidean metric in the sense that @xmath277 for smooth vector fields @xmath278 . on zero- and one - forms
, this connection takes the expression : @xmath279 for given vector fields @xmath280 and @xmath281 .
an important identity for the sequel is the following characterization of the exterior derivative of a one - form @xmath282 @xmath283 ) , \label{eq : dcarac}\end{aligned}\ ] ] or equivalently in the euclidean metric , writing @xmath284 for some vector field @xmath285 , @xmath286 = \del_x ( z\cdot y ) - \del_y ( z\cdot x ) - dz^\flat(x , y ) , \label{eq : dcarac2 } \end{aligned}\ ] ] where the lie bracket ( commutator ) of @xmath28 and @xmath234 coincides with ( and thus may be `` defined '' here as ) @xmath287=\del_x y-\del_y x$ ] by virtue of the torsion - free property .
a _ frame _ refers to an oriented family @xmath288 of @xmath23 vector fields over @xmath289 such that for every @xmath87 , @xmath290 is a basis of @xmath291 . for a given frame @xmath292
, we define the _ christoffel symbols _ ( of the second kind ) with respect to this frame , by the relations @xmath293 the following very useful identity allows us to compute the christoffel symbols from inner products and lie brackets of a given frame ( see e.g. ( * ? ? ?
5.1 p. 69 ) ) : @xmath294 - z\cdot[y , x ] + x\cdot[z , y ] , \end{split } \label{eq : chriscarac}\end{aligned}\ ] ] where @xmath278 are smooth vector fields .
for a vector @xmath295 , we want to form the matrix of partial derivatives @xmath296 .
geometrically , gradients generalize to tensors via the total covariant derivative , which maps a vector field @xmath28 to a tensor of type @xmath297 defined by @xmath298 in a given frame @xmath292 , we may express @xmath299 in the basis @xmath300 of such tensors by writing @xmath301 and identifying the coefficients @xmath302 by writing @xmath303 and also @xmath304 equating the two , we obtain the representation @xmath305 the theory of the following sections proves that all partial derivatives of a frame ( given in ) are uniquely determined by inner products @xmath306 and by lie brackets , as indicates , or equivalently by exterior derivatives , as expresses . these derivations will be carried out first for the @xmath117 frame and second for the @xmath121 frame with values in the space of rotations @xmath307 .
we now study the properties of the @xmath117 frame .
@xmath117 is a frame provided that the determinant condition @xmath308 holds .
our objective in this section is to find an expression for @xmath309 that allows us to solve for @xmath20 by the method of characteristics .
we have seen in the preceding section that this involved calculating the lie brackets ( commutators ) of the vectors composing the frame . for @xmath310 , we have @xmath311 = h^{kl } ( [ s_i , s_j]\cdot s_k ) s_l .
\label{eq : lie1}\end{aligned}\ ] ] now using we write @xmath312 & = \del_{s_i } ( s_k\cdot s_j ) - \del_{s_j } ( s_k\cdot s_i ) - d s^\flat_k ( s_i , s_j ) \\ & = s_i\cdot\nabla h_{kj } - s_j\cdot\nabla h_{ki } - \alpha f^\flat\wedge s^\flat_k ( s_i , s_j ) \\ & = s_i\cdot\nabla h_{kj } - s_j\cdot\nabla h_{ki } + \alpha ( - h_{kj } f\cdot s_i + h_{ki } f\cdot s_j ) .
\end{aligned}\ ] ] plugging this into and using that @xmath313 , we obtain the lie brackets @xmath314 $ ] for @xmath315 : @xmath311 = h^{kl}[\nabla h_{jk } \cdot s_i - \nabla h_{ik}\cdot s_j ] s_l + \alpha ( ( f\cdot s_j ) s_i - ( f\cdot s_i ) s_j ) .
\label{eq : liesi}\end{aligned}\ ] ] returning to the computation of @xmath309 using , we combine with to arrive at @xmath316 - s_p\cdot[s_i , s_q ] + s_q\cdot[s_p , s_i ] \\ & = \nabla h_{iq}\cdot s_p + \nabla h_{ip}\cdot s_q - \nabla h_{pq}\cdot s_i + 2 \alpha ( h_{pq } ( f\cdot s_i ) - h_{qi}(f\cdot s_p)).\end{aligned}\ ] ] plugging this expression into ( expressed in the @xmath117 frame ) , and using @xmath317 , we obtain @xmath318 where we have used @xmath319 and have defined @xmath320 using formulas @xmath321 and @xmath322 for any smooth vector field @xmath180 , we obtain for @xmath323 @xmath324 using , we observe that @xmath309 is equal to a polynomial of degree at most three in the frame @xmath117 with coefficients involving the known inner products @xmath26 . for each @xmath325 , @xmath326 is nothing but @xmath327 , which can be obtained from . denoting @xmath328 $ ]
, we are then able to construct the system of equations @xmath329 where @xmath330 depends only on the data and @xmath331 is an @xmath332-index .
this redundant system can then be integrated along any curve ( where it becomes a system of ordinary differential equations with lipschitz right - hand sides ensuring uniqueness of the solution ) in order to solve for the matrix - valued function @xmath333 .
the above system involves a priori @xmath332 unknowns since the matrix @xmath117 does not necessarily have any useful symmetries .
however , we know the inner products @xmath334 , i.e. , a matrix of dimension @xmath335 .
we therefore hope to be able to find a closed - form system involving @xmath336 dimensions .
this is the dimension of the orthonormal @xmath121 frame .
we now provide the details of remark [ rem : rframe ] . from the frame @xmath117
, we build an oriented orthonormal frame @xmath337 $ ] ( or equivalently , an @xmath338-valued function ) from a matrix - valued function @xmath339 that satisfies the relations @xmath340 and @xmath341 at every @xmath87 , as well as a stability property of the form @xmath342 where @xmath343 depends only on the way we construct @xmath344 from @xmath107 .
@xmath344 can either be constructed by the gs procedure or by setting @xmath345 , the positive square root of @xmath94 .
the stability statement , first proved in the gs case for @xmath49 in @xcite , can be obtained for both gs and @xmath346 , see @xcite for proofs of these statements . the function @xmath347 satisfies everywhere @xmath348 and @xmath349 , hence @xmath121 is an @xmath338-valued function .
the column vectors of @xmath117 and @xmath121 transform according to : @xmath350 we also define for @xmath325 @xmath351 we are brief on the derivation of the gradient system for @xmath121 as it is very similar to that of the @xmath117 frame .
the system of equations - together with the transformation rules allow us to derive the following system of equations for the @xmath121 frame : @xmath352 from this system , we express @xmath84 in the @xmath121 frame as @xmath353 equation can also be derived directly from and the transformation rules .
then , using equation and formula , the lie brackets @xmath354 $ ] of the vectors take the form , for @xmath310 : @xmath355 = ( -v_{pj}\cdot r_i + v_{pi}\cdot r_j ) r_p + \alpha((f\cdot r_j)r_i - ( f\cdot r_i)r_j ) .
\label{eq : lieri}\end{aligned}\ ] ] from we deduce the christoffel symbols relative to the @xmath121 frame : @xmath356 finally , in the orthonormal case , the expression of the gradient reduces to @xmath357 , from which we deduce that @xmath358 as for the @xmath117 frame , the r.h.s .
of depends polynomially on @xmath121 and on the data .
this system can thus be solved for the vectors @xmath359 via ode integration along any curve in a connected domain and provided that we know the @xmath121 frame at one point . in practice , this system is less expensive to integrate than since the @xmath121 frame can be locally parameterized with @xmath360 scalar functions ( such as the euler angles ) whereas the @xmath117 frame requires @xmath332 scalar functions .
the proof of the stability theorem [ thm : odestab ] can be found in @xcite in dimension @xmath54 with @xmath95 ( although the proof would be identical in arbitrary dimension ) . in that paper ,
the theorem is proved using the system for the rotation matrix @xmath121 and thus requires the extra stability condition .
this condition is necessary only if we reconstruct @xmath8 via the @xmath121 frame .
the same stability result can be obtained without this requirement if we reconstruct @xmath8 via the @xmath117 frame directly . in the latter setting , the proof is quite similar to the one in @xcite with the further simplification that we do not need to change bases when switching subdomain @xmath361 .
the system of odes that one must solve based on the gradient system is well - posed since the function @xmath333 satisfies a priori the uniform bound @xmath362 and the right - hand side of is lipschitz in @xmath333 over the set @xmath363 as a polynomial of the components of @xmath333 , and using the fact that the polynomial @xmath330 are bounded ; see @xcite for additional details , which we do not reproduce here .
in this case , condition is satisfied with the partition @xmath364 . equation can be rewritten as @xmath365 rewriting the conductivity equation as @xmath366 and plugging into it yields the coupled elliptic system of equations @xmath367 where we have have defined @xmath368 from the last form of @xmath369 , we derive and since the denominators only involve @xmath370 which is bounded away from zero and the rest is polynomial in the @xmath26 s and their derivatives .
multiplying by @xmath371 and writing it in divergence form , one obtain the following equivalent formulation to in variational form : @xmath372 in the case where we have @xmath373 solutions , we can still define @xmath97 over the entire domain @xmath28 using a partition of unity that is subordinate to the open cover @xmath374 , call it @xmath375
. then we can define @xmath97 globally over @xmath28 by writing @xmath376 where the restrictions are constructed from the @xmath23 solutions of positive determinant on each @xmath361 .
each of these restrictions can still be written in the form @xmath377 with @xmath378 .
thus we can patch these formulas together into a globally defined @xmath379 plugging this expression into the conductivity equation yields the coupled elliptic system @xmath380 and one arrives at a system of the form by setting @xmath381 for every @xmath382 . in this case
, the stability inequalities and can be derived using the fact that @xmath383 and noticing that on each @xmath361 , @xmath134 is either zero or locally defined by ( and weighed by @xmath384 ) whose expression has been proved to be stable .
a similar argument holds for proving thanks to the fact that the partition of unity @xmath385 is the same for two data sets @xmath386 that jointly satisfy with the same triple @xmath102 .
let us assume a system of the form , where the vector fields @xmath134 belong to @xmath152 .
assuming the illumination @xmath156 to be in @xmath387^m$ ] , we use a lifting operator to define functions @xmath388^m$ ] of traces @xmath156 at @xmath389 . defining the unknown @xmath390 ,
we are now left with analyzing the solvability of the system @xmath391 as well as the stability of its solution with respect to the vector fields @xmath134 .
@xmath392 denotes the dual space of @xmath393 . as described in section [ sec : results ] , we apply the inverse of the dirichlet laplacian to and obtain the system of integral equations @xmath394 which can be recast in vector notation as @xmath395_i \bfe_i = \delta_d^{-1 } ( w_{ij}\cdot \nabla v_j ) \bfe_i , \qandq \bff : = \delta_d^{-1 } h_i\ \bfe_i .
\end{split } \label{eq : integralsys}\end{aligned}\ ] ] because @xmath141 is continuous in the functional setting @xmath396 ( see @xcite ) , it is clear that @xmath397 belongs to the space @xmath398 .
we now have the following : [ lem : pw ] assuming that the vector fields @xmath399 , the operator @xmath139 defined in is compact , and its norm satisfies @xmath400 as can be seen in @xcite for instance , the operator @xmath401 is bounded .
therefore , by the rellich compactness theorem , the operator @xmath402 is compact and of norm denoted by @xmath147 .
now @xmath403 is also compact since each of its components is the composition of the continuous operator @xmath404 with the compact operator @xmath402 .
moreover , for @xmath405 and every @xmath382 , we have the obvious bounds @xmath406 and thus @xmath407_i\|_{h_0 ^ 1}^2 \le \|\delta_d^{-1}\|^2 \|w\|_\infty^2 \|\bfv\|_\h^2 .
\end{aligned}\ ] ] summing over @xmath36 proves .
the proof is complete . as a consequence of lemma [ lem : pw ] and by virtue of standard compact operator theory ( e.g. ( * ? ? ?
* theorem 6 p 643 ) ) , we have the following facts : * @xmath408 is eigenvalue of @xmath145 , which corresponds to the case @xmath409 , a value for @xmath10 that we exclude from our analysis , * the remaining spectrum of @xmath410 is point spectrum and consists of at most a discrete sequence of values that is either finite or converges to zero .
finally , the operator @xmath411 satisfies a fredholm alternative .
therefore it suffices that @xmath412 in order to obtain uniqueness and stability of the solution of and therefore of the solution of as well .
the proof of proposition [ prop : elliptic ] makes these statements more precise .
let @xmath413 have their coefficients in @xmath152 and such that @xmath414 , and let @xmath415 solve the system with respective drift terms @xmath157 , @xmath158 and same illumination @xmath156 . by virtue of the fredholm alternative , the operators @xmath416 and @xmath417 are invertible with continuous inverses in @xmath418 . applying the inverse dirichlet laplacian to both systems
, we obtain the systems @xmath419 taking the difference of both systems , the resulting system reads @xmath420 the first difference in the right - hand side of may be bounded by @xmath421 where the constant @xmath241 depends on @xmath422 and @xmath423 . applying lemma [ lem : pw ] to the operator @xmath424 , the second difference in the right - hand side of may be bounded by @xmath425 combining the last two estimates with and the fact that @xmath416 is invertible with continuous inverse in @xmath418 , we arrive at @xmath426 for some constant @xmath427 . since @xmath428 , this concludes the proof .
we now conclude with the proof of theorem [ thm : elliptic ] .
we focus on the case @xmath429 and @xmath430 .
the proof for @xmath125 is identical up to small changes in notation .
let @xmath431 have their components in @xmath101 and jointly satisfy with the same triple @xmath102 .
then the families of vector fields @xmath157 and @xmath158 have their coefficients in @xmath152 and we further assume that @xmath432 .
let @xmath433 solve the system with respective drift terms @xmath157 and @xmath158 and same illumination @xmath156 , and let @xmath166 be the corresponding conductivities . without loss of generality , we work on one of the open sets @xmath434 and renumber the @xmath23 solutions whose gradients are linearly independent from @xmath183 to @xmath23 . the result will then hold provided that we have @xmath435 and thus @xmath436 for a given @xmath437 , and defining @xmath438 , we write , using equality @xmath439 similarly to the vector fields @xmath134 , the vector fields @xmath440 satisfy estimates of the form @xmath441 since @xmath107 is bounded and @xmath166 are assumed to be bounded from below by a constant @xmath442 , each of the @xmath19 is uniformly bounded by @xmath443 . taking @xmath444 norms over @xmath361 and using the triangle inequality in , we obtain @xmath445 which by virtue of proposition [ prop : elliptic ] and estimates yields an estimate of the form @xmath446 further , from the pointwise relations @xmath447 and similarly for @xmath106 , we write @xmath448 taking @xmath444 norms and using the triangle inequality , we obtain that @xmath449 which again yields an estimate of the form @xmath450 combining and , we arrive at @xmath451 for every @xmath434 .
this concludes the proof .
[ [ section-5 ] ] + in the case @xmath51 , ( * ? ? ?
* theorem 4 ) guarantees that we can pick @xmath161 . in so doing and using the form of the elliptic system together with the fact that @xmath452 for all @xmath2 , we arrive at the system @xmath453
the weak formulation of the corresponding problem with homogeneous dirichlet conditions involves the bilinear form @xmath454 since @xmath94 is uniformly elliptic over @xmath28 and @xmath455 , this bilinear form is coercive over @xmath398 as seen from the following calculation @xmath456 where @xmath457 stands for the largest eigenvalue of @xmath107 , for which we have , pointwise ( @xmath458 designates a unit eigenvector associated with @xmath457 ) @xmath459 and hence the estimate @xmath460 therefore by virtue of the lax - milgram theorem , the system admits a unique solution in @xmath398 .
in particular , this shows that @xmath153 is not an eigenvalue of @xmath145 in this case for any @xmath2 .
[ [ section-6 ] ] using the fact that the spectrum of @xmath461 is bounded in norm by @xmath461 and that @xmath145 is compact , the elliptic system admits a unique and stable solution , except for a discrete set of values @xmath462 $ ] possibly converging to zero . in terms of @xmath10 , this corresponds to almost all values of @xmath2 except a sequence @xmath463 taking values in the interval @xmath464 $ ] and possibly converging to @xmath149 .
[ [ section-7 ] ] in this case we have @xmath452 .
this implies that whenever one can ensure the positivity condition with only @xmath161 solutions ( e.g. in even dimension and using lemma [ lem : cgo ] ) , one can rewrite the system into the form with term @xmath465 , that is @xmath466 using the same arguments as in the two - dimensional case , this system is coercive and therefore ensures that @xmath467 is not an eigenvalue of @xmath145 .
[ [ section-8 ] ] as a conclusion of this discussion , the following statements hold : 1 . if @xmath51 , we have @xmath468 , 2 .
if @xmath70 , then @xmath469 consists of at most a sequence @xmath470 where @xmath471 belongs to @xmath464 $ ] and possibly converges to @xmath149 . in the case
@xmath161 , the value @xmath408 is excluded from the latter interval .
the ode - based reconstructions use the full redundancy of the data to construct an overdetermined system of equations for the vectors @xmath20 ( or @xmath359 ) and the vector @xmath97 .
the pde - based method defines a well - posed system of equations for the scalar quantities @xmath3 and an overdetermined system of equations for vector @xmath97 .
each of these overdetermined systems needs to satisfy compatibility conditions in order to admit a solution . in this section
, we aim to extract information from the over - determinacy of the system .
we first revisit the two - dimensional case and use the redundancy to extract explicit reconstruction algorithms in the setting @xmath472 .
we then consider the case of arbitrary dimension and show that the compatibility conditions that data must satisfy in order for the aforementioned systems to have solutions take the form of vanishing appropriately defined curvatures together with the cancellation of a given two - form .
these conditions generate quadratic functionals of the unknown vectors @xmath20 or @xmath359 whose pratical applicability is discussed below . in this section
, we revisit the two - dimensional case which was first solved in @xcite and generalize the approach to the case @xmath473 . in that approach ,
the reconstruction of @xmath474 requires the reconstruction of a function @xmath475 that characterizes the unknown information about the frames @xmath117 or @xmath121 .
we consider the @xmath476-valued @xmath477 frame and parameterize it as @xmath478 and @xmath479 , with @xmath480 $ ] . with the notation @xmath481 for @xmath482
, we recast equations and for @xmath2 as follows : @xmath483 we next derive an equation for @xmath484 , which by construction is nothing but @xmath485 $ ] .
we have : @xmath486 = \del_{r_2 } r_1 - \del_{r_1 } r_2 = -\gamma_{22}^1 r_2 + \gamma_{11}^2 r_1 , \label{eq : gradtheta0}\end{aligned}\ ] ] where the christoffel symbols @xmath487 and @xmath488 are given by @xmath489 = -\nabla\cdot r_2 \\ & = - ( ( 2\alpha-1)v_{22}-(1-\alpha)n ) \cdot r_2 - ( ( 2\alpha-1)v_{12}^s - v_{12}^a ) \cdot r_1 , \\
\gamma_{22}^1 & = \del_{r_2}r_2\cdot r_1 = -[\nabla , r_2 ] = -\nabla\cdot r_1 \\ & = - ( ( 2\alpha-1)v_{11}-(1-\alpha)n ) \cdot r_1 - ( ( 2\alpha-1)v_{12}^s + v_{12}^a ) \cdot r_2 .
\end{split } \label{eq : chris2d}\end{aligned}\ ] ] by orthonormality the other christoffel symbols are given by @xmath490 plugging the expressions into , we arrive at @xmath491 using the following identity @xmath492 equation may be recast as @xmath493 whose expression matches the one given in @xcite when @xmath34 . since @xmath84 is a function of @xmath494 , the above equation is then a non - linear pde whenever @xmath472 .
this is to be contrasted with the seemingly much nicer case @xmath495 , whose r.h.s .
is independent of @xmath494 .
a right - hand side independent of @xmath494 can , however , be obtained by taking divergence of both sides of since @xmath496 and @xmath497 .
the equation we obtain is @xmath498 this elliptic pde requires knowledge of @xmath494 at the domain s boundary .
assume that we know @xmath499 , @xmath500 for @xmath264 , and @xmath8 at the boundary .
in this setting , we find that @xmath501 with @xmath502 and @xmath503 the unit outgoing normal vector and its direct orthogonal vector , respectively . once @xmath494 is reconstructed , we know the r.h.s . of and solve for @xmath113 , either by integrating along a curve , or by taking the divergence of both sides of and solving a poisson equation provided that @xmath504 is known .
note that the inversion for @xmath494 and @xmath113 by means of the elliptic equations and `` divergence of '' with dirichlet conditions is _ unique _ and _ lipschitz - stable _ in @xmath505 w.r.t .
the data @xmath26 .
the details are left to the reader .
we now discuss the compatibility conditions for the gradient equations and , which admit a solution only if their respective r.h.s .
are curl - free .
such conditions lead to a better understanding of the range of the measurement operator and are necessary to ensure that reconstructions based on ode integrations do not depend on the choice of integration path .
the compatibility conditions for and are that @xmath506 and @xmath507 , respectively . for @xmath472 ,
these equations not only provide constraints on the redundant data , but in fact give us direct information about the unknown coefficients . in the two - dimensional case , they allow us to solve algebraically for @xmath508 , which in turn characterizes @xmath84 in terms of the data ( and therefore does not require the prior resolution of @xmath494 ) .
let us first simplify the expression of @xmath84 as follows : @xmath509 where the matrices @xmath510 and @xmath511 are reflexion matrices that can be expressed in the following manner : @xmath512 .
\end{split } \label{eq : reflexions}\end{aligned}\ ] ] as a result , we are able to express @xmath84 in a rather compact way @xmath513 note the property that @xmath514 .
now turning to the compatibility conditions proper , equations and are well - defined only if their curls are zero , which gives the following two scalar conditions @xmath515 now using the chain rule and @xmath516 , we have @xmath517 where we have defined @xmath518 similarly , the second compatibility equation can be recast as @xmath519 if @xmath472 , we thus see that the two compatibility equations imply the system @xmath520 \left [ \begin{array}{c } c_2 \\ s_2 \end{array } \right ] = \left [ \begin{array}{c } 0 \\ h \end{array } \right ] , \quad h:= \big ( \alpha-\frac{1}{2 } \big)^{-1 } \nabla\cdot ( jg ) - 2\big ( \alpha-\frac{1}{2 } \big ) |f_c|^2 , \label{eq : h}\end{aligned}\ ] ] which may be inverted as @xmath521 note that this solution makes sense only if the functions @xmath522 are such that @xmath523 , that is , if they satisfy the relation @xmath524 in this case , @xmath84 may be expressed as @xmath525 the right - hand - side is guaranteed to be curl - free by construction .
inserting into the last equation and using @xmath526 with @xmath527 and @xmath528 , we obtain the following _ explicit _ reconstruction formula : @xmath529 when @xmath95 , the equation @xmath506 _ depends solely on the data _ and reads @xmath530 the other compatibility equation @xmath531 is still of the form @xmath532 , with @xmath533 defined above , which by itself only gives us @xmath534 up to a sign , i.e. @xmath535 the above constraint provides partial answer about @xmath494 that may be used in practical reconstructions to mitigate the influence of noise in the data .
reconstructions based solely on these algebraic relations , however , seem to be less stable than the two approaches based on integration of gradient or laplace equations .
the two - dimensional case is special in that @xmath476 is both one dimensional and abelian .
this is not the case in higher dimensions , where parameterizations are much more complicated , even in three dimensions .
similar compatibility conditions arise in the @xmath23-dimensional case for gradient equations as an application of the poincar lemma : the `` curl '' ( or exterior derivative ) of the equation vanishes on both sides . for equations or , we have that @xmath536 which implies a system of @xmath537 scalar equations ( i.e. the cancellation of a 2-form in dimension @xmath23 ) . regarding the systems and , their complete integrability is equivalent to ensuring that the curvature tensor of the euclidean connection is identically zero when expressed in either frame @xmath117 or @xmath121 . indeed , according to (
* theorem 1 p30 ) , a system of the form @xmath538 is integrable if and only if the following conditions hold @xmath539 which is equivalent , after using the chain rule , to @xmath540 the last equation is nothing but the fact that the curvature @xmath541 of the euclidean metric is zero for every quadruple @xmath542 , where the curvature tensor @xmath543 is defined , for three vector fields @xmath278 , by @xmath544 } z. \label{eq : curvature}\end{aligned}\ ] ] @xmath543 is known to be a tensor , i.e. it is linear over smooth functions in all its arguments , and thus the above zero curvature conditions are strictly equivalent to the equations @xmath545 for @xmath546 , where @xmath292 is any frame , including @xmath117 and @xmath121 . as one
can see from ( * ? ? ?
10 and 12 pp 196 - 197 ) , this highly redundant set of @xmath547 scalar equations is equivalent to @xmath548 non - redundant equations which express the cancellation of the _ sectional curvatures _
@xmath549 for each frame , we thus have the two systems of @xmath548 equations and .
we now work with the @xmath121 frame because its christoffel symbols have nicer symmetry properties and show that both systems and may be recast as @xmath550 where the matrices @xmath551 depend only on the data ( we could obtain similar equations for the @xmath117 frame ) .
we first rewrite the christoffel symbols under the form @xmath552 the vector fields @xmath553 depend only on the data and have the antisymmetry properties @xmath554 and @xmath555 .
we now derive systems of the form for both systems and in the case of the @xmath121 frame .
starting from equation , the equation @xmath556 reads @xmath557 now using identity with vector fields @xmath558 , we have @xmath559 \\ & = \del_{r_i } ( v_{kj}^s\cdot r_k ) - \del_{r_j}(v_{ki}^s\cdot r_k ) \\ & \qquad + ( v_{kl}^s\cdot r_k ) ( ( w_{ij}^{lp}-w_{ji}^{lp})\cdot r_p ) \,\,=\,\,0.\end{aligned}\ ] ] decomposing the first term in the last r.h.s . as follows @xmath560 and
doing simlarly for the second term , we obtain the set of scalar equations @xmath561 this system can be written in the form , where the matrices @xmath562 depend only on the data .
given the symmetries of the christoffel symbols , one can show for an orthonormal frame that the zero sectional curvature equations can be recast as @xmath563 using expression of the christoffel symbols , the first term in the left - hand side of may be rewritten as @xmath564 proceeding similarly for the second term of the l.h.s . of and plugging expression into the r.h.s .
, we obtain the set of equations for @xmath310 @xmath565 ( @xmath566 are not being summed over but @xmath567 are ) .
this is also a quadratic system of the form with different matrices @xmath562 . * either give us compatibility conditions on the data ( that do not depend on the unknown frame ) , thus characterizing the range of the measurement operator , * or they may allow us to invert algebraically for the cosines and sines of the @xmath360 spherical angles that parameterize the @xmath338-valued @xmath121 frame .
this in turn may come at the price of other compatibility conditions that only depend on the data .
it remains an interesting , so far unresolved , question to find an algorithm that enforces the compatibility conditions as the system of odes is used to ensure that the reconstruction does not depend on the choice of integration paths . the system of elliptic equations for the scalar solutions @xmath3 is well - posed for almost all values of @xmath10 and @xmath23 .
once the solutions @xmath3 are obtained , it remains to solve the equation for @xmath568 .
the only remaining compatibility condition is therefore that the latter term indeed be a gradient . in a similar manner to what we just saw for the ode - based method , writing the condition @xmath556 yields @xmath336 equations of the type @xmath569 where the matrices @xmath562 depend only on the data @xmath26 .
99 h. ammari , e. bonnetier , y. capdeboscq , m. tanter and m. fink , siam j. appl .
, 68(*6 * ) ( 2008 ) , pp .
15571573 . g. alessandrini and v. nesi , arch .
, * 158 * ( 2001 ) , pp . 155171 .
g. bal , in `` inside out '' ( ed .
gunther uhlmann ) , cambridge university press ( 2012 ) .
height 2pt depth -1.6pt width 23pt , submitted .
g. bal , e. bonnetier , f. monard and f. triki , , inverse probl .
imaging , in press ( 2012 ) .
g. bal and k. ren , , inverse problems , * 27 * ( 2011 ) .
g. bal and g. uhlmann , , inverse problems , 26(*8 * ) ( 2010 ) , p. 085010 .
a. caldern , , seminar on numerical analysis and its applications to continuum physics , soc .
brasileira de matematica , rio de janeiro , ( 1980 ) , pp .
6573 . c. carathodory , `` calculus of variations and partial differential equations of the first order '' , 3ed . , ams chelsea , 1999 .
y. capdeboscq , j. fehrenbach , f. de gournay and o. kavian , , siam journal on imaging sciences , * 2 * ( 2009 ) , pp . 10031030 .
evans , `` partial differential equations '' , graduate studies in mathematics , vol . 19 , ams ( 1998 ) . b. gebauer and o. scherzer , , siam j. applied math
. , 69(*2 * ) ( 2009 ) , pp .
s. kim , o. kwon , j. k. seo , and j .-
yoon , , siam j. math .
, * 34 * ( 2002 ) , pp . 511526 .
p. kuchment and l. kunyansky , , inverse problems * 27 * ( 2011 ) , 055013 .
lee , `` riemannian manifolds , an introduction to curvature '' , graduate texts in mathematics , vol . 176 , springer ( 1997 ) . f. monard , `` taming unstable inverse problems
. mathematical routes toward high - resolution medical imaging modalities '' , ph.d .
thesis , columbia university in the city of new york ( 2012 ) .
a. nachman , a. tamasan , and a. timonov , , inverse problems , * 23 * ( 2007 ) , pp . 25512563 . height 2pt depth -1.6pt width 23pt , , inverse problems , * 25 * ( 2009 ) , p. 035014 .
height 2pt depth -1.6pt width 23pt , , contemporary mathematics , american mathematical society , in press ( 2012 ) .
o. scherzer , `` handbook of mathematical methods in imaging '' , springer verlag , new york ( 2011 ) .
m. spivak , `` a comprehensive introduction to differential geometry , vol . 2 '' , 2nd ed . , publish or perish ( 1990 ) .
j. sylvester and g. uhlmann , , ann . of math .
, 125(*1 * ) ( 1987 ) , pp .
. m. taylor , `` partial differential equations i , basic theory '' , springer new york ( 1996 ) . | this paper concerns the reconstruction of a scalar diffusion coefficient @xmath0 from redundant functionals of the form @xmath1 where @xmath2 and @xmath3 is a solution of the elliptic problem @xmath4 for @xmath5 .
the case @xmath6 is used to model measurements obtained from modulating a domain of interest by ultrasound and finds applications in ultrasound modulated electrical impedance tomography ( umeit ) , ultrasound modulated optical tomography ( umot ) as well as impedance acoustic computerized tomography ( impact ) .
the case @xmath7 finds applications in magnetic resonance electrical impedance tomography ( mreit ) .
we present two explicit reconstruction procedures of @xmath8 for appropriate choices of @xmath9 and of traces of @xmath3 at the boundary of a domain of interest .
the first procedure involves the solution of an over - determined system of ordinary differential equations and generalizes to the multi - dimensional case and to ( almost ) arbitrary values of @xmath10 the results obtained in two and three dimensions in @xcite and @xcite , respectively , in the case @xmath6 .
the second procedure consists of solving a system of linear elliptic equations , which we can prove admits a unique solution in specific situations .
franois monard guillaume bal ( communicated by the associate editor name ) |
after a few decades of observation by several projects , the nature of the uhecr is gradually being revealed .
based on the observational results , there is no doubt of the existence of uhecr up to @xmath1 ev @xcite .
this has led to consideration of a large number of production models .
most of the models assume that uhecrs are protons or photons .
several studies of anisotropy have been done to reveal the origin of the uhecr @xcite . around @xmath0 ev , a large scale anisotropy which is correlated to the galactic center was reported by agasa group @xcite .
this result , if confirmed , supports an interpretation that cosmic rays below @xmath0 ev are dominated by a galactic origin . above this energy ,
however , no significant large scale anisotropy has been detected by agasa @xcite .
on the other hand , agasa reported clusters in the distribution of arrival directions including 4 doublets and 2 triplets with energy above @xmath2 ev@xcite .
this results indicate that the uhecr come from extra - galactic point sources .
the cluster events reported by agasa are composed of particles with different energies .
a possible explanation would be neutral particles ( like neutrinos ) propagating in straight lines from the source .
another possible explanation would be charged particles with large rigidity , i.e. with small charge like protons .
if there is only the effect of the galactic magnetic field then protons of different energy can arrive at same direction @xcite . however , these two possibilities can not explained the three bumps in the energy spectrum of the cluster events between @xmath3 ev and @xmath4 ev @xcite reported by agasa .
this structure is not statistically significant .
however , it suggests to us the interesting possibility that uhecr might be nuclei as it will be discussed in this paper . in this scenario ,
cluster events would be formed by particles with different charge and energy but with the same rigidity .
the atomic number of the primary cosmic ray does not affect the energy determination of agasa which measures the number of charged particles on the ground . however , the capability of agasa to measure the composition of the primary cosmic rays is not very strong @xcite .
nevertheless , the structure of the energy distribution of the cluster events may be used to measure of the composition indirectly as discussed in this paper .
if uhecr come from point sources and their rigidity is large enough or their source close enough so the direction is not lost , a fraction of the particles emitted by the source will arrive within a small angle from the source position .
these particles are what we will call in this paper `` small deflection angle '' particles .
clusters will be formed when the number of particles expected within this angle in our experiment is larger than one .
the small deflection angle particles should represent the cluster events observed by agasa .
uhe protons should interact with cosmic microwave background radiation ( cmbr ) and lose energy by photo - pion production .
for this reason , @xmath1 ev protons can not propagate in intergalactic space beyond 50 mpc , leading to a cut off in the spectrum below @xmath1 ev , the so called gzk cut off . on the other hand
heavy nuclei will interact during its propagation with the intergalactic infrared background radiation(iibr ) .
unfortunately the iibr can not be measured directly so far because of the strong background of infrared photons of galactic origin .
the main component of the iibr is thought to be red - shifted stellar emission and reemission from interstellar dust .
the uhe nucleus should be disintegrated by the iibr and lose energy by the photo - disintegration interaction .
for this reason , it was believed that the attenuation length of uhe nuclei is shorter than that of protons in intergalactic space @xcite .
however , recent calculations which are based on empirical data show the iibr density is much lower than originally thought @xcite .
the lack of absorption in the energy spectrum of very high energy gamma rays from agns strongly supports this calculation . using this result ,
the attenuation length of the uhe nuclei in the intergalactic photon field was re - estimated @xcite and found to be longer than that of a proton ( a few 100 mpc for @xmath1 ev fe ) .
we simulate the propagation of uhe nuclei which are emitted from extra - galactic point sources based on a monte carlo approach . using this simulation , we found structures in the energy distribution of the uhe nuclei which are observed as cluster events on the earth . in section
[ sec : photons ] we briefly describe the interaction of nuclei with background photons . in section
[ sec : magnetic ] we describe the scattering of the nuclei by the intergalactic magnetic field . in section [ sec : energy ]
we show the result of this simulation .
we conclude in section [ sec : discussion ] and discuss the results briefly .
there are three processes which affect cosmic rays during intergalactic propagation : energy loss by interaction with cosmic background radiation ; deflection by magnetic fields ; and energy loss by adiabatic expansion of the universe .
the effect of the adiabatic expansion is negligibly small if propagation distances are less than 1 gpc .
there are four interaction processes of cosmic rays with intergalactic background radiation . the first process , compton interaction , is negligibly small for our purpose .
the second process is pair - production of e@xmath5e@xmath6 .
this interaction can occur if the energy of the background photon is greater than 1 mev in the rest system of the nucleus .
if the lorentz factor of the nucleus is larger than @xmath7 , it will efficiently interact with the cmbr ( figure [ fig : f2 ] ) . in this case
, the energy loss rate ( @xmath8 ) is proportional to the square of the charge of the nucleus .
the third process is photo production which mainly affects protons with energy greater than @xmath1 ev .
the interaction of heavier nuclei with the cmbr by this process is negligible for the energy range considered in this paper .
the fourth process is photo - disintegration .
the nucleus interacts with the background photon and disintegrates to a lighter nucleus . in this paper
, we calculate this process by a monte - carlo method according to @xcite .
these authors parametrized the total cross section @xmath9 as a function of photon energy in the rest frame of the nucleus .
then the probability of photo - disintegration per unit length @xmath10 is calculated by following equation : @xmath11 d \epsilon\ ] ] where @xmath12 and @xmath13 are the energies of the photon in the lab frame and in the nucleus rest frame respectively , @xmath14 is the differential number density of photons including the cmbr and the iibr @xcite , and @xmath15 is the lorentz factor of the nucleus .
the term inside the bracket corresponds to the angle - averaged cross - section for a photon of energy @xmath12 .
figure [ fig : f2 ] shows characteristic time @xmath16 for the energy loss due to photo - disintegration for a single - nucleon emission process as a function of energy .
@xmath16 is defined by following equation : @xmath17 the effect of the cmbr is maximized when the lorentz factor @xmath18 is around @xmath19 . in case of @xmath20 , this value corresponds to an energy of @xmath21 ev , yielding a value of @xmath22 s ( 2 mpc ) .
because the iibr density is much lower than the cmbr density , @xmath16 increases rapidly at lower energy .
@xmath16 of @xmath20 with @xmath1 ev is about @xmath23s ( 2.5gpc ) .
in general , one or a few nucleons and a single lighter nucleus are emitted by this interaction . the lorentz factor is conserved at each photo - disintegration interaction , though it is reduced by the pair - production .
finally , uhe particles will pile up at a lorentz factor around @xmath7 .
when @xmath24 is disintegrated , one proton and two @xmath25 are emitted .
therefore no nucleus is created with a mass number between 5 and 8 .
the interaction probability is calculated for every possible photo - disintegration process involving the emission of one or more nucleons for all nuclei lighter than @xmath20 .
based on these probabilities , the propagation of the nucleus is simulated by the monte - carlo method .
the effects of pair production , the adiabatic expansion , and photo - pion production for protons are taken into account analytically in this simulation .
figure [ fig : life_of_fe ] shows the life of an @xmath20 nucleus whose initial energy is @xmath26 ev .
the @xmath20 is disintegrated during its propagation and turns into a lighter nucleus with the emission of nucleons .
when the nucleus is @xmath27 , four protons and four neutrons are emitted . then a @xmath28 is created . in comparison with energy ,
the lorentz factor changes smoothly because photo - disintegration process does not change the lorentz factor .
this process makes a lighter nucleus and nucleons with same lorentz factor .
the effect of magnetic fields is described by the rigidity defined as the ratio of energy to charge ( @xmath29 ) .
particles with small rigidity are deflected by the magnetic field and can not be observed as a cluster .
the deflection of magnetic field increases the propagation time , and therefore particles emitted from distant sources may not reach to the observer .
we calculate the effect of the intergalactic magnetic field based on a monte carlo method . to simulate the scattering by the intergalactic magnetic field , we assume a kolmogorov spectrum for the random magnetic field according to the reference of @xcite . in this reference ,
authors divide space in a lattice of 250 kpc cubes .
the lattice is filled with a random magnetic field which follows the kolmogorov spectrum with three logarithmic scales .
three field vectors of random orientation are sampled at scales @xmath30 = 1000 , 500 , and 250 kpc with amplitudes proportional to @xmath31 .
the final magnetic field in each 250 kpc cube is vectorial sum of these three vectors .
the average magnitude of the magnetic field is assumed to be 1 ng . particles propagate in spiral trajectories until they leave the lattice .
figure [ fig : trace2 ] shows examples of the trajectories .
@xmath20 with energy of @xmath0 ev does not rapidly lose energy by the interaction with photons , and is trapped by the magnetic field inside 1 mpc cubes . in case of @xmath32 ev , @xmath20 is disintegrated by the photons rapidly and the products propagates close to a straight line in the initial direction of the @xmath20 .
in our simulation , the source energy spectrum @xmath33 is assumed to have a power law dependence @xmath34 and have a cut - off at the energy of @xmath35 ev , where @xmath36 is charge of each primary nucleus .
below 300 mev / nucleon , the cosmic nuclear composition can be divided to 4 types : @xmath25 , @xmath37 , @xmath38 , and @xmath20 .
nuclear abundance between @xmath25 and @xmath39 is relatively low as well as between @xmath40 and @xmath20 .
we treat the composition as 4 components , @xmath41 , @xmath42 , @xmath43 , and @xmath44 .
the fraction of these nuclear abundances in the uhe region are unknown .
therefore we assume equal fractions relative to @xmath25 at the source .
20000 particles are created for each nucleus .
figure [ fig:180_d_fe ] shows the variation of the energy distribution after propagation from a single source initiated by @xmath20 through the photon and magnetic field .
nuclei with lorentz factor above @xmath7 interact with the cmbr rapidly .
after 1 mpc propagation a small pile - up is induced below the corresponding energy and the maximum energy of the particle is reduced . since we assume
a cut - off energy in the source spectrum , protons are emitted with energies below @xmath45 ev where @xmath46 is mass number of the primary nucleus .
the number of low energy nuclei ( @xmath47ev ) increases up to 10 mpc from the source since these nuclei are trapped by the magnetic field .
these low energy nuclei can not propagate farther than few 10 mpc .
therefore , the energy distribution becomes steeper for distances below 10 mpc and flatter for larger distances .
no particles remain after propagation of a few 100 mpc .
figure [ fig : psrc_50mpc ] is same result as lower left panel of figure [ fig:180_d_fe ] .
the energy distributions at a distance of 50 mpc from the source are shown for different deflection angle ( angle between the arrival direction and the source position ) .
particles with the deflection angle smaller than 2.5 , 5 , 10 , 180 degrees are counted ( 180 degrees corresponds to the entire sample of particles included in figure [ fig:180_d_fe ] ) .
three bumps appear clearly in the distribution of particles with a small deflection angle .
the bump below @xmath48 ev is composed of protons which are emitted from higher energy nuclei in the vicinity of the observer who is located at 50 mpc distance from the source .
the bump around @xmath49 ev is composed of @xmath25 , and the bump above @xmath50 ev is composed of nuclei heavier than @xmath51 .
these bumps can be explained as follows : the rigidity is proportional to @xmath52 and @xmath53 , and @xmath36 is approximately proportional to its mass number @xmath46 while the lorentz factor is proportional to @xmath54 and @xmath53 .
this means that the lorentz factor of a particle is approximately proportional to the rigidity .
therefore , if the lorentz factor of the nucleus is large , the energy is reduced by the photo - disintegration .
if the lorentz factor is small , the particle is deflected by the magnetic field due to its small rigidity . however particles with lorentz factor of
about @xmath55 remain after applying the deflection angle cut . as a result
, the bumps appear in the energy distribution of the small deflection angle particles at positions dependent on the mass numbers .
the combination of the photodisintegration and the magnetic field acts as a `` filter '' for @xmath56 particles .
the energy distribution of the small deflection angle particles shows clearly the spectral features created by interactions with the intergalactic photon background and magnetic fields during their propagation .
if the sources are distributed uniformly in the universe , the expected energy distribution measured at the earth can be estimated by adding up linearly the energy distributions corresponding to different distances from the source . in figure [ fig:180_d_fe ]
the number of particles around @xmath1 ev coming from sources at 100 mpc is 1.5 orders of magnitude smaller than from sources at 100 kpc . when adding linearly the different distances , the number of particles from sources at 100 mpc is enhanced 3 orders of magnitude compared to sources at 100 kpc .
therefore , the contribution to the energy spectrum at earth is increasing with the distance to the source till a couple of hundred mpc .
on the other hand , most of the low energy particles emitted by distant sources will be trapped by the magnetic field and they will not reach the earth . however , the low energy secondaries resulting from the photodisintegration of high energy heavy nuclei close to earth will regenerate the low energy spectrum coming from distant sources .
therefore , even at low energies we will still be dominated by the contribution of distant sources .
the spectrum of small deflection angle particles will follow the same behavior , i.e. the contribution of sources at large distances to the spectrum exceed largely the one from small distant sources , even if the fraction of small deflected particles for closer sources is slightly larger .
it should be noted that the relation between cluster events and small deflection particles is qualitatively clear but difficult to estimate quantitatively .
current work is in progress for a better comparison with experimental data .
figure [ fig : spect_all ] shows the result of the expected energy distribution under the assumption that the sources are distributed uniformly in the universe .
the solid line shows the expected energy distribution of all the particles that reach the earth . the expected energy spectrum is steeper than the source spectrum . above @xmath57
ev , the energy spectrum becomes even steeper and the maximum energy is about @xmath58 ev .
the exact details of the upper end of the spectrum will depend on the source spectrum cut off . the dashed lines in figure [ fig : spect_all ] show the expected energy distribution of the small deflection angle particles .
this spectrum is much flatter than the source spectrum .
four bumps appear in this figure .
the bump just below @xmath48 ev corresponds to protons and strongly depends on the cut off energy at the source .
the bump around @xmath59 ev is composed of @xmath60 and @xmath25 .
every nucleus with a lorentz factor around @xmath55 contributes to the @xmath25 bump after propagation .
the bump between @xmath50 ev and @xmath1 ev correspond to nuclei with mass between @xmath51 and @xmath61 .
the bump above @xmath62 ev is composed of @xmath20 and other nuclei of similar atomic number .
this total spectral structure bears a striking resemblance to the result of agasa @xcite .
we have simulated propagation of uhe nuclei which are emitted from extra - galactic point sources using a monte carlo approach . using this simulation , structures in the energy distribution of the small deflection angle particles
have been predicted . in this simulation
, we assume that uhecr are nuclei which are emitted from extra - galactic point sources .
the nuclei interact with cosmic background radiation and lose energy during their propagation .
based on recent estimations of the iibr , it has been shown that uhe nucleus can propagate longer distances than a proton with same energy .
it is also expected that heavier nuclei are accelerated more efficiently to high energy than lighter nuclei or protons .
we have considered the effect of the intergalactic magnetic fields .
we use a kolmogorov spectrum for random magnetic fields that is 1 ng in average intensity with a 1 mpc correlation length .
the energy distribution of the particles with deflection angles smaller than 10 , 5 , and 2.5 degrees are investigated .
consequently , for this model no cosmic rays can be observed after propagation of a few hundred mpc .
lower rigidity particles are deflected by the magnetic field relatively rapidly .
particles with larger lorentz factor lose energy by photodisintegration . since the rigidity and the lorentz factor are proportional to the particle energy and inversely proportional to the mass number , particles which have common a lorentz factor , or similar rigidity , remain after the deflection angle cut .
the @xmath63 bump appears at an energy more than 4 times larger than the @xmath25 bump , and the @xmath20 bump appear around 6 times larger energy .
the rigidity gives a lower bound and energy loss gives an upper bound on the bumps .
the energy distribution of the clustered events observed by agasa can be explained by this model . in the simulation , we assume a specific intergalactic magnetic field distribution , cut - off energy and source composition of primary nuclei .
the galactic magnetic field is ignored in this simulation .
the events with certain deflection angle are selected and shows the effects due to the intergalactic photons and magnetic fields .
this effect depends on the strength and correlation of the intergalactic magnetic field .
this dependence gives an ambiguity to the analysis since the deflection angle cut is adjustable .
therefore , this angle should be treated carefully .
this result suggests that if we detect few hundred events above @xmath1 ev in future experiments , we should be able to extract the structure of the bumps in the energy distribution of cluster events .
the pierre auger observatory is quite sufficient for this purpose and may confirm the signature of nuclei in the spectrum in the near future .
we thank angela olinto , alan watson , james w. cronin , maximo ave , aaron chou , and medina tanco for useful discussions .
this research was supported by grant nsf phy-0114422 to the center for cosmological physics , university of chicago , and grant nsf phy0103717 .
an @xmath20 is emitted with an energy of @xmath26 ev .
the @xmath20 is disintegrated during its propagation and turn into lighter nucleus with the emission of nucleons .
these figures show the properties of the surviving nucleus as a function of propagation distance from the source .
the upper and lower left panels show the variation of the mass number and the energy of the surviving nucleus respectively .
these parameters change in a similar way .
the upper and lower right panels show the variation of the lorentz factor and the rigidity respectively . in comparison with the energy ,
the lorentz factor changes smoothly because the photo - disintegration process does not change the lorentz factor .
only the pair - production process affects this parameter .
therefore variations of the lorentz factor and the rigidity are smaller than that of the energy .
the fluctuation of the rigidity in lower right panel is caused by the variation of the ratio of mass number to charge(a / z ) .
, scaledwidth=100.0% ] in the intergalactic magnetic field . a @xmath20 emitted in the @xmath64 direction with an energy of @xmath65 ev from a source located on the origin in left panel .
a projection of the trajectory on x - z plane is shown . at this energy ,
the attenuation length is sufficiently large .
therefore the energy of the particle is almost conserved and the particle is scattered by the magnetic field significantly . the @xmath20 can not propagate a distance of more than 1 mpc from the source . in the right panel , a @xmath20 with energy of @xmath66 ev emitted in the same direction as the left panel . the @xmath20 and secondary nuclei
are indicated by the wide line and secondary nucleons ( proton and neutron ) are indicated by the narrow lines .
the @xmath20 interact with photons rapidly and is disintegrated to lighter nuclei with emission of protons and neutrons .
all of the particles are scattered by the magnetic field but deflection angle is comparatively small .
arrows in the panels indicate 0.5 mpc distance.,scaledwidth=100.0% ] after propagation over various distances from a single source .
the distances from the source are indicated in each panel .
the number of particles ( arbitrary scale ) are shown as a function of energy .
primary nuclei ( @xmath20 ) , secondary lighter nuclei , and protons are indicated by white , light shade and dark shade in the histograms respectively .
the differential source spectrum is assumed to be a power law , @xmath67 , with energy cut - off at @xmath68 ev ( upper left panel ) . at 100 mpc from the source , most of the @xmath20 disappear because of their low rigidity .
, scaledwidth=100.0% ] | we present a calculation of nuclei propagation with energies above @xmath0 ev in the intergalactic photon field .
the calculation is based on a monte carlo approach for the nucleus - photon interaction as well as the intergalactic magnetic field .
we then assume that the ultra - high energy cosmic rays ( uhecr ) are nuclei which are emitted from extra - galactic point sources .
four bumps are found in the energy spectrum of the uhecr which form clusters in the distribution of their arrival directions .
based on this calculation , the energy distribution of the clustered events is discussed .
ultra high energy cosmic rays , propagation |
einstein s field equations are adiabatic and reversible and hence can not provide an explanation for the origin of cosmological entropy.@xcite but matter constituents may be produced quantum mechanically in the framework of einstein s equations.@xcite the quantum fluctuations of the gravitational field near singularity , as well as the dynamics of any quantum field living on such a non stationary background , lead to matter- creating phenomena , which can in turn modify the behavior of the early universe.@xcite a detailed microscopic description of matter creation phenomena should be provided by a fully self - consistent quantum field theory on a curved space time.@xcite many attempts have been made to treat the matter creation process at a phenomenological macroscopic level.@xcite@xcite the basic idea is that matter creation contributes at the level of the einstein field equations as a negative pressure term .
prigogine et al . incorporated its effect in the classical einstein field equations.@xcite at the expense of the gravitational field , matter creation can occur only as an irreversible process .
the additional pressure term due to the creation depends on the creation rate .
the total entropy within a comoving volume is not a constant , due to the variation of number of particles .
this approach explained the average effect induced on the dynamics of an isotropic expanding universe . in this work we study the effect of adiabatic matter creation on brane world models .
we havent considered the details of the source of created matter .
recently tetradis has shown that absorption of energy flux from the bulk by the three brane leads to acceleration or deceleration of the universe.@xcite we have also derived a phantom field equivalent formulation . this paper is organised as follows . in section 2 we give a brief description of rsii brane world model . in the section 3 , thermodynamics of matter creation
is explained . in section 4
we describe the effect of matter creation on brane worlds in general , and on rsii in section 5 .
an equivalent phantom field formulation is given in section 6 . in sections 7 and 8
we give the adiabatic matter creation effects on dgp and gb brane models .
the idea that our 4-dimensional universe might be a 3-brane embedded in a higher dimensional space - time has attracted much attention over the last several years .
randall and sundrum ( rs ) models@xcite are the most important . according to the brane world scenario ,
the physical fields in our 4-dimensional space - time are confined to the three brane , while gravity can freely propagate in the bulk space - time .
we assume a friedmann metric for 4-dimensional spacetime embedded in the bulk 5- dimensional space with a cosmological constant @xmath0 . the friedmann equation on the rsii brane is,@xcite @xmath1 with @xmath2 where @xmath3 is the cosmological constant on the brane , @xmath4 is the brane tension , @xmath5 is energy density of ordinary cosmological matter on the brane , k is the curvature index , and the term with c ( a constant ) is the dark radiation term.@xcite cosmology constrains the amount of dark radiation to be at most 10 percent of the energy density in photons.@xcite in the following we will assume ads bulk spacetime , and @xmath6 , and @xmath7 in rsii model the important change in friedmann s equation compared to the usual four dimensional form is the appearance of a term proportional to @xmath8 .
it tells us that if the matter energy density is much larger than the brane tension , i.e. @xmath9 , the expansion rate h is proportional @xmath8 . only in the limit where the brane tension is much larger than the matter energy density
, the usual behavior @xmath10 is recovered .
the einstein s field equations with bianchi identities give , for homogeneous and isotropic universes in closed systems , @xmath11 where @xmath12 is the true thermodynamical pressure .
this equation is used to describe an adiabatic evolution of a closed system with a comoving volume @xmath13 .
but in the presence of matter creation , the analysis is in the context of open systems.@xcite in this case the number of particles @xmath14 in a given volume is not fixed . for adiabatic transformation , @xmath15 where @xmath16 and @xmath17 is the enthalpy per unit volume . in such a transformation ,
the heat received by the system is due to the change of the number of particles . in our cosmological context
, this change is due to the transfer of energy from gravitation to matter .
correspondingly the entropy change @xmath18 , for adiabatic transformation in open systems is , @xmath19 where @xmath20 is the chemical potential and @xmath21 .
therefore the only particle number variations admitted are such that @xmath22 the eq .
( [ eqn5 ] ) can be expressed as , @xmath23 @xmath24 where the over dot represents time derivative .
the creation of matter corresponds to a supplementary pressure @xmath25 , which is negative or zero depending on the presence or absence of particle production .
thus eq .
( [ eqn5 ] ) can be written as , @xmath26 where @xmath27 is the true thermodynamical pressure and @xmath28 this scenario is applied to cosmology as follows . in the case of an isotropic and homogeneous universe , we choose , for @xmath13 , the value @xmath29 where @xmath30 is the scale factor .
the above expression becomes , @xmath31 the conservation equation becomes ,
@xmath32 @xmath33 with @xmath34 the balance equation for particle number density is,@xcite @xmath35 with @xmath36 where @xmath37 is thermostatic pressure , @xmath38 is fluid 4-velocity , @xmath39 is particle number density , and @xmath40 is matter creation rate . for @xmath41 , there is particle production , and @xmath42 corresponds to annihilation .
the above - mentioned macroscopic phenomenological approach is not completely self consistent since it can not determine the expression for the rate of particle creation , which is an open degree of the scheme .
the existence of some kind of negative - pressure dark energy is usually assumed to explain the cosmic acceleration . vacuum energy or a cosmological constant is the simplest dark energy candidate .
it has been known that matter creation may lead to negative pressures.@xcite the modified friedmann equation on rsii brane world indicates that the universe evolves slowly in the very early times .
the process of matter creation may add to the accelerated expansion and thus can compensate to the effect of brane world at early times . a scenario with particle production driving cosmic acceleration
was proposed by zimdahl et al.@xcite using a phenomenological macroscopic description of matter creation,@xcite we investigate this scenario in the brane world context .
however , we do not consider that particle creation as the only source for cosmic acceleration .
it has been shown based on the supernova data that matter creation alone does not drive cosmic acceleration.@xcite our motivation is the following : the modified friedmann equation on rsii brane world shows that there is transition for the universe evolution between an early high energy regime characterized by the behavior @xmath43 and a late low energy regime characterized by the standard evolution @xmath44 .@xcite the matter creation process due to its negative pressure can be the cause of this transition .
our attempt is to consider the possible effects on dynamics of the early universe due the matter creation .
the field equations on the rs ii brane model are,@xcite @xmath45 @xmath46 where @xmath47 , and @xmath48 for perfect fluid . with particle creation , creation pressure , eq .
( [ eqn11 ] ) , is added to thermostatic pressure @xmath27 .
@xmath49 for simplicity we take @xmath50 . also , since the dark radiation term decays rapidly we can put @xmath51 , eq .
( [ eqn18 ] ) becomes , @xmath52 the energy conservation equation is unchanged in 5-d setup , which including particle creation , becomes , @xmath53 for perfect fluid , using eq .
( [ eqn19 ] ) we get , @xmath54 we take the linear dependence of matter creation rate on hubble parameter @xmath55
as,@xcite @xmath56 where the parameter @xmath57 is defined in the interval @xmath58 $ ] and assumed to be constant .
the observational value of @xmath59 ( @xmath60 ) can be translated to the value of @xmath61 .
the joint result of sne ia and grb indicate that @xmath61 is ( @xmath62 ) .
the grb data shows @xmath61 to be greater than @xmath63 ( ie.,@xmath64 ) , and the snia data alone shows ( @xmath65).@xcite this linear relation implies @xmath66 , at high energy densities .
( recently freaza et .
al . @xcite
have considered a general form for the matter creation in non - brane scenario as , @xmath67 , where @xmath68 and @xmath57 are dimensionless constants ) .
now the solution of eq .
( [ eqn22 ] ) is , @xmath69 substituting for @xmath8 , and for @xmath70 , eq .
( [ eqn17 ] ) becomes , @xmath71 defining @xmath72 , this becomes , @xmath73 solving we get @xmath74 using the boundary condition , @xmath75 , when @xmath76 , we find the solution for the scale factor as follows , @xmath77 the frw evolution laws are regained for @xmath78 , and @xmath79 , which corresponds to high energy densities .
thus we get , for radiation @xmath80 , and for dust @xmath81 . using the phenomenological limit , @xmath82 , @xcite the inequality
, @xmath79 gives @xmath83 .
this means that our model is valid for any time in the evolution of the universe .
the value of @xmath84 depends on @xmath61 . assuming that the rate of matter creation is insignificant during later stages of the evolution of the universe , the evolution of scale factor reduces to the frw relation .
the effect of matter creation may be translated into the dynamics of a minimally coupled scalar field . the effect of negative pressure due to the matter creation is equivalent to a suitable scalar field potential .
simple models of a super accelerated universe include a phantom field , i.e. , a minimally coupled scalar field with negative kinetic energy.@xcite@xcite it has also been proposed that early inflation and late time acceleration of the universe can be unified in a single theory based on a phantom field.@xcite phantom type of matter may also arise from a bulk viscous stress due to particle production.@xcite we consider particle production from a phantom field , and derive the corresponding potential .
the field equation including the phantom field , @xmath85 is , @xmath86 the rhs can be written as , @xmath87 , then the sum @xmath88 is conserved by the bianchi identity but the individual terms are not .
this was interpreted in the steady state theory@xcite as creation of matter by the creation field in order to maintain a constant matter density . the conservation equation , assuming friedmann metric , is
, @xmath89 we assume the total energy and pressure consist of matter and phantom field components as , @xmath90 @xmath91 @xmath92 to obtain a phenomenological classical description of the essentially quantum creation process , we assume that the creation pressure is proportional to the energy density and pressure of the scalar field as , @xmath93 now the eq .
( [ eqn33 ] ) becomes , @xmath94 the expression for creation pressure , from eq .
( [ eqn11 ] ) is , @xmath95 for perfect fluid matter , @xmath96 @xmath97 the solution is , @xmath98 where @xmath99 for phantom fields,@xcite the energy density , @xmath100 and pressure , @xmath101 where @xmath13 is the potential of phantom field .
the scalar field equations are , @xmath102 @xmath103 the solution is , @xmath104 using this solution and eqs .
( [ eqn24 ] ) , ( [ eqn25 ] ) and ( [ eqn26 ] ) in eq .
( [ eqn23 ] ) we get the potential as , @xmath105 the modified friedmann equation , eq .
( [ eqn1 ] ) , for very high energy densities becomes , @xmath106 using eq .
( [ eqn31 ] ) for @xmath107 , @xmath108 we can write this equation as , @xmath109 where @xmath110 defining , @xmath72 where @xmath111 , the above equation can be integrated as , @xmath112 this integral can be solved for a given value of q. with @xmath113 , ( the value which gives the frw law for the rsii case ) , and @xmath114 ( radiation ) , @xmath84 becomes @xmath115 , thus the solution of eq .
( [ eqn49 ] ) is , @xmath116 neglecting the second term in comparison with the first , and @xmath117 we get , @xmath118 substituting in eq .
( [ eqn43 ] ) and solving for @xmath119 we get , @xmath120 using eqs .
( [ eqn44 ] ) , ( [ eqn51 ] ) and ( [ eqn52 ] ) , we get the expression for the potential in terms of phantom field as , @xmath121 this potential of the phantom field gives the same evolution for the sale factor as produced by the adiabatic matter creation , in the high density regime . in this sense
the phantom field model is equivalent to the adiabatic matter creation model discussed previously in section 4 .
gauss - bonnet ( gb ) combination arises as the leading order for quantum corrections in the heterotic string effective action . in five dimensions it represents the unique combination of curvature invariants that leads to second - order field equations , linear in the second derivatives in the metric tensor , and it is ghost - free.@xcite@xcite the gb correction to the einstein - hilbert action will disrupt mildly the degeneracy between tensor and scalar perturbations.@xcite recently there is a renewed interest upon this scenario . here
we consider the effect of adiabatic matter creation in this scenario . in the rsii scenario with a gb term , the effective friedmann equation describing the motion of brane containing general perfect fluid
may be derived from a generalization of the birkhoff s theorem.@xcite the 5-dimensional bulk action for the gb braneworld scenario is given by,@xcite @xmath122 } \right ) } + s_{\partial m } + s_m \label{eqn54}\ ] ] where @xmath123 = \left ( { r^2 - 4\,r_{ab } \,r^{ab } + r_{abcd } r^{abcd } } \right ) $ ] is the gauss - bonnet term , @xmath124 with dimension @xmath125 is the gauss - bonnet coupling , @xmath126 is the bulk cosmological constant and @xmath127 determines the 5d planck scale .
the @xmath128 is the boundary term that is required to cancel normal derivatives of the metric tensor which arises when varying the action with respect to the metric , @xcite and @xmath129 is the action for matter confined on the brane . we assume a z2 symmetry across the brane .
we will consider the case that a perfect fluid matter source with density @xmath107 is confined to a brane with tension @xmath4 .
a new constant is defined , @xmath130 . since @xmath131
, there is an upper bound on @xmath132 , @xmath133 . in order
that the standard friedmann equation be recovered at sufficiently low energy scales ( @xmath134 ) , @xmath135 , where @xmath136 is the 4d planck scale . introducing a dimensionless variable @xmath137 , the modified friedmann equations are @xmath138 and , @xmath139 \label{eqn56}\ ] ] where @xmath140 is the brane tension . in the high energy approximation , @xmath141 ,
we can write it as , @xmath142 in order for that the 4d effective cosmological constant vanishes , the brane tension should satisfy @xmath143 .
we can see that @xmath144 for any @xmath145
. following a similar calculation as in rs ii model in section 5 , with a substitution @xmath146 in eq .
( [ eqn57 ] ) , we get @xmath147 the solution with the boundary condition , @xmath148 when @xmath149 , gives @xmath150 @xmath151 where @xmath152 and @xmath153 .
this evolution is for high energy densities @xmath154 , since we considered the modified friedmann equation for that regime .
the case of @xmath155 , which was considered in rsii model , gives for radiation and dust as @xmath156 and @xmath157 respectively .
these are modified evolution equations compared to standard cases without matter creation .
an alternative scenario to the rs ii brane model was proposed by dvali , gabadadze and porrati ( dgp ) .@xcite@xcite in this model the extra dimension is infinitely large .
the key feature is the presence of a four - dimensional curvature scalar in the action .
the dgp model predicts that 4d newtonian gravity on a brane world is regained at distances shorter than a given crossover scale @xmath158 ( high energy limit ) , whereas 5d effects become manifest above that scale ( low energy limit ) .
the interesting feature is that the effective cosmological constant on the brane can be extremely reduced in contrast to the case of the randall - sundrum model even if a bulk cosmological constant and a brane tension are not fine - tuned .
also , the model can explain late - time acceleration without having to invoke a cosmological constant or quintessential matter.@xcite here we consider a generalized dgp brane model of sahni and shtanov.@xcite the action of the generalized dgp model is,@xcite @xmath159 where , @xmath160,~~and,~~ \label{eqn62}\ ] ] @xmath161 \label{eqn63}\ ] ] here @xmath162 is the 5-dimensional gravitational constant , @xmath163 and @xmath164 are the 5-dimensional curvature scalar and the matter lagrangian in the bulk , respectively .
@xmath165 are the induced 4-dimensional coordinates on the brane , @xmath166 is the trace of extrinsic curvature on either side of the brane and @xmath167 is the effective 4-dimensional lagrangian , which is given by a generic functional of the brane metric @xmath168 and matter fields @xmath169 on the brane .
the brane lagrangian consists of the following terms , @xmath170 where @xmath171 is the brane tension .
the 5-d bulk space includes only a cosmological constant @xmath172 .
it is a generalized version of dvali - gabadadze - porrati model , which is obtained by setting @xmath173 as well as @xmath174 . from the field equation induced on the brane , the effective 4-dimensional cosmological constant on
the brane is , @xmath175 .
considering a friedmann - robertson - walker metric on the brane , the friedmann equation is found to be,@xcite @xmath176 \label{eqn65}\ ] ] where @xmath177 is either + 1 or -1 , and @xmath178 is given by , @xmath179^{\frac{1}{2 } } \label{eqn66}\ ] ] with @xmath180 , @xmath181 , and @xmath182 .
there are two mass scales , @xmath183 and @xmath184 , in this model . at high energy limit , @xmath185 , the modified friedmann equation eq .
( [ eqn65 ] ) reduces to @xmath186 for simplicity , the curvature term and dark radiation term can be omitted . then the eq .
( [ eqn67 ] ) can be further simplified to @xmath187 we will use this evolution law to investigate the effects of adiabatic matter creation .
we assume the expression , @xmath188 , for matter creation .
defining , @xmath189 , and inserting the solution , @xmath190 , of the conservation equation ( with modified creation pressure term ) in eq .
( [ eqn68 ] ) , we get @xmath191 defining @xmath192 , we get @xmath193 substituting this in the above equation , @xmath194 this equation can be written as , @xmath195 where @xmath196 and @xmath197 . the solution of eq .
( [ eqn72 ] ) is @xmath198 with the boundary condition , @xmath199 , when @xmath76 , the constant @xmath200 becomes @xmath201 .
therefore we get , @xmath202 where @xmath203 , @xmath204 and @xmath153 . for @xmath78 ( special solution in the case of rsii ) , and for @xmath205 , which corresponds to high energy densities ,
we get , @xmath206 @xmath207 these are new evolution equations different from frw laws . for @xmath208 ( later stage ) the rate of matter creation
is assumed to be insignificant , thus the evolution of scale factor is unaffected .
the adiabatic matter creation rate is found to affect the evolution of scale factor and energy density of the universe , in all the cases we have considered .
the modified friedmann equation on brane world shows that the universe evolves slowly in the very early times .
thus there is a transition of the universe evolution between an early high - energy regime and a late low energy regime .
the negative pressure due to the matter creation process can explain this transition .
for the rsii model the standard frw evolution law can be obtained for a matter creation rate @xmath113 .
this value is consistent with the observed data .
we have derived a scalar field equivalent formulation using a phantom field .
a potential is derived which would give the dynamics same as that produced by adiabatic matter creation . in the case of gauss - bonnet and
warped dgp models there are modifications in evolution of the universe with matter creation . for later stages of the evolution of the universe ,
the rate of matter creation is insignificant , thus the evolution of scale factor is unchanged .
pg thanks csir , india for a research fellowship , and dr.k.indulekha , spap , for suggestions .
we thank prof .
naresh dadhich , iucaa for suggestions , and iucaa , pune for the hospitality during this work .
00 i. prigogine , et al . , _ gen .
grav . _ * 21*,767 ( 1989 ) .
l. parker , _ phys .
* 21 * , 562 ( 1968 ) ; _ phys .
rev . _ * 183 * , 1057 ( 1969 ) .
r. brout , f. englert and e. gunzig , _ ann .
( ny ) * 115 * , 78 ( 1978 ) ; _ nucl .
* b170 * , 228 ( 1980 ) . g. montani , _ class .
grav . _ * 18 * , 193 - 203 ( 2001 ) .
rnd birrell and pcw davies , quantum fields in curved space ( cambridge university press , 1982 ) .
m. s. turner , _ phys .
rev . _ * d28 * , 1243 ( 1983 ) .
j. d. barrow , _ phys .
lett . _ * b180 * , 335 ( 1986 ) ; _ nucl .
phys . _ * b310 * , 743 ( 1988 ) . j. a. s. lima , r. portugal and i. waga , _ phys . rev . _ * d37 * , 2755 ( 1988 ) .
n. turok , _ phys .
lett . _ * 60 * , 549 ( 1988 ) .
j. a. s. lima and a. s. m. germano , _ phys
. lett . _ * a170 * , 373 ( 1992 ) .
r. abramo and j. a. s lima , _ class .
* 13 * , 2953 ( 1996 ) .
e. gunzig , r. maartens , _ class .
grav . _ * 15 * , 923 ( 1998 ) , [ asto - ph/ 9703137 ] .
j. gabriel , g. le denmat , _ phys .
* a200 * , 11 ( 1995 ) .
n. tetrdis , _ phys
_ * b569 * , 1 ( 2003 ) .
l. randall and r. sundrum , _ phys .
lett . _ * 83 * , 3370 ( 1999 ) .
l. randall and r. sundrum , _ phys .
lett . _ * 88 * , 4690 ( 1999 ) .
p. brax , c. van de bruck and a. davis , _ rept .
phys . _ * 67 * , 2183 ( 2004 ) , [ hep - th/ 0404011 ] d. langlois , _ prog .
suppl . _ * 148 * , 181 ( 2003 ) , [ hep - th/ 0209261 ] s. mukohyama , _ phys .
lett . _ * b473 * , 241 ( 1999 ) .
d. ida , _ jhep _ * 0009 * , 014 ( 2000 ) .
r. maartens , [ astro - ph/9609119 ] ya .
b. zeldovich , _ jetp lett . _
* 12 * , 307 ( 1970 ) ; b. l. hu , _ phys .
* 90a * , 375 ( 1982 )
. w. zimdahl , et al .
_ * d64 * , 063501 ( 2001 ) .
m. p. freaza , r. s. de souza and i. waga , _ phys .
rev . _ * d66 * , 103502 ( 2002 ) .
p. binetruy , c. deffayet and d. langlois , _ nucl .
phys . _ * b565 * , 269 ( 2000 ) , [ hep - th/ 9905012 ] .
b. gumjudpai , r. maartens and c. gordon , _ class .
_ * 20 * , 3295 ( 2003 ) , [ gr - qc/0304067 ] . j. a. s. lima and j. s. alcaniz , _ astr . and
ast . phys .
_ * 348 * , 1 ( 1999 ) .
y. qiang and t. zhang , [ astro - ph/0503123 ] .
r. maartens , reference frames and gravitomagnetism , ed .
j pascual - sanchez et al . ( world sci . , 2001 ) ,
p 93 - 119 , [ gr - qc/0101059 ] .
r. r. caldwell , _ phys .
_ * b545 * , 23 ( 2002 ) .
r. r. caldwell , m. kamionkowski and n. n weinberg , _ phys .
_ * 91 * , 071301 ( 2003 )
. s. m. carroll , m. hoffman and m. trodden , _ phys .
rev . _ * d68 * , 023509 ( 2004 ) .
s. nojiri and s. d. odintsov , _ phys .
lett . _ * b562 * , 147 ( 2003 ) . j. d. barrow , _ nucl . phys .
_ * b310 * , 743 ( 1988 ) .
f. hoyle , g. burbidge and j. v. narlikar , _ a different approach to cosmology : from a static universe through the big bang towards reality _
( cambridge university press , 2000 ) .
t. padmanabhan , _ curr .
* 88 * , 1057 ( 2005 ) , [ astro - ph/ 0411044 ] . s. nojiri , s d odintsov and s ogushi , phys.rev.d65 , 023521(2002 ) , [ hep - th 0108172 ] m. cvetic , s nojiri and and s d odintsov , nucl .
phys.b628 , 295 ( 2002 ) , [ hep - th 0112045 ] s. nojiri and s d odintsov , jhep 0007 , 049 , 2000 , [ hep - th 0006232 ] .
s nojiri , s d odintsov and k. ostrin , phys . rev .
d63,084016 ( 2001 ) , [ hep - th 0009059 ] .
c. charmousis and j. dufaux , _ class .
* 19 * , 4671 ( 2002 ) , [ hep - th/ 0202107 ] .
s. c. davis , _ phys.rev . _ * d67 * , 024030 ( 2003 ) , [ hep - th/ 0208205 ] .
p. binetruy , et al .
_ * b544 * , 183 ( 2002 ) , [ hep - th/ 0206089 ] .
b. zweibach , _ phys.lett . _ * b156 * , 315 ( 1985 )
. j. f. dufaux , et al .
_ * d70 * , 083525 ( 2004).[hep - th/ 0404161 ]
. j. e. lidsey and n. j. nunes , _ phys .
rev . _ * d67 * , 103510 ( 2003 ) , [ astro - ph/ 0303168 ] .
r. c. myers , _ phys . rev .
_ * d36 * , 392 ( 1987 ) .
g. dvali , g. gabadadze and m. porrati , _ phys .
* b484 * , 112 ( 2000 ) ; g. dvali and g. gabadadze , _ phys .
rev . _ * d63 * , 065007 ( 2001 ) ; g. dvali , g. gabadadze , m. shifman , _ phys . rev . _
* d67 * , 044020 ( 2003 ) .
k. maeda , s. mizuno and t. torii , _ phys .
rev . _ * d68 * , 024033 ( 2003 ) , [ gr - qc/ 0303039 ] .
varun sahni and yuri shtanov , jcap , 0311 , 014 ( 2003 ) , [ astro - ph/0202346 ] . | we have treated the adiabatic matter creation process in various three - brane models by applying thermodynamics of open systems . the matter creation rate is found to affect the evolution of scale factor and energy density of the universe .
we find modification at early stages of cosmic dynamics . in gb and
rs brane worlds , by chosing appropriate parameters we obtain standard scenario , while the warped dgp model has different solutions . during later stages , since the matter creation is negligible the evolution reduces to frw expansion , in rs and gb models . |
measurements of the linear x - ray polarization in solar flares can provide essential information needed to identify the processes responsible for the acceleration of particles and the emission of radiation .
for photon energies in the hard x - ray region , polarization is produced either through electron bremsstrahlung or compton scattering .
the degree of polarization is usually a complex function of the strength and topology of the magnetic field .
thermal distributions of electron velocities result in small polarization values produced mainly by the conduction - driven anisotropy of the electrons in the primary source @xcite with some contribution from photons backscattered in the photosphere .
higher polarization levels are expected from non - thermal anisotropic distributions of electrons that are accelerated in well ordered magnetic fields .
in addition , for such cases the observed polarization degree is related to the photon directivity ( i.e. the anisotropy of the emitted radiation ) , which depends both on the electron beaming details and on the viewing angle .
starting in the late seventies , several non - thermal models of x - ray emission from solar flares were developed ( see and references therein ) .
generally the authors assume a uniform magnetic field perpendicular to the solar surface and electrons being accelerated towards the chromosphere (; ; ; ; ) .
these electrons , spiraling downwards along the magnetic field lines , produce bremsstrahlung radiation in collisions with hot plasma .
the polarization in the emitted x - rays is a function of the energy spectra of the electrons , their pitch angles ( angle between their velocity vector and the magnetic field ) , and the column density distribution of the ambient plasma .
studying the spectral characteristics of the detected x - rays provides information about the electron energies , while the polarization is a very sensitive tool to sample the electrons pitch angle distribution . both for very small pitch angles , corresponding to high electron beaming , and for very large ones
, the polarization degree can be equally high .
the two cases can be distinguished by their polarization direction : parallel and perpendicular to the magnetic field line , respectively .
the predicted polarization values can reach up to 60% at energies above 50 kev , being even higher at low energies @xcite . after introducing more realistic pitch angle distributions and taking into account photon backscattering processes in the photosphere
, the expected polarization is reduced down to 20% or 30% (; ) .
several more complex non - thermal models have been developed as well . in one of them
the magnetic field structure was defined as a semicircular loop anchored in the chromosphere @xcite .
this approach allowed studying the x - ray emissions separately from different parts of the loop .
the highest polarization could be produced at the top ( up to 85% ) , while the photons observed from the foot - points ( in the region of the dense chromosphere ) would be polarized to the level of around 20% . in general , lower energies are predicted to yield stronger polarization signals (; ; ) , although the strength of this relation can vary depending on the model . in some recent theories
an opposite trend has also been reported @xcite .
experimental verification is usually difficult because the low - energy part of the spectrum is strongly contaminated by a non - polarized thermal emission . in all the cases the observed value of flare polarization is strongly dependent on the viewing angle .
the highest polarization values are expected for large angles of view , when the line of sight is perpendicular to the magnetic field line .
thus , most theories predict higher polarization for flares located near the solar limb .
similar behavior is also expected for the directivity of the flare emission : the intensity of the emitted radiation should depend on the angle of view (; ) .
constraints related to the model assumptions favor two possible directions of the polarization vector : either parallel to the plane defined by the magnetic field lines and the line of sight or perpendicular to it (; ; ) .
contrary to the intense theoretical work , only a few polarization measurements have been conducted in hard x - rays . in this energy range
, the commonly used technique is based on compton scattering @xcite .
first attempts were done at energies around 15 kev by , , and using polarimeter instruments on board of several intercosmos satellites . although initial results @xcite showed a linear polarization of around 40% @xmath1 20% , the later studies of different flares (; ) found their polarization degrees always compatible with zero . moreover , the data suffered from limited photon statistics and systematic errors related with the detector calibration . in measurements of solar flare polarization with the reuven ramaty high energy spectroscopic imager ( rhessi ) , compton scattering can occur in a specially installed beryllium scatterer .
this method can only be used at low energies ( 20100 kev ) .
its details , together with rhessi s polarimetric features , are described in and .
the latest measurements at energies below 100 kev have been performed with the spr - n instrument on board of the coronas - f satellite by . from a sample of 25 solar flares , these authors determined upper limits on the polarization degree in the range from 8 to 40% ( @xmath2 ) .
only for the single case of the 29 october 2003 flare they found a significant polarization degree , which increases from about 50% at energies @xmath3 , up to more than 70% for the energy channel @xmath4 .
recently published , the only two measurements of solar flare polarization at high energies ( 0.21 mev ) , show 1@xmath5 values of 21% @xmath1 10% and -11% @xmath1 5% for one flare close to the limb and another near the sun center , respectively @xcite . in their studies ,
the authors applied the polarimetric capabilities of the rhessi satellite in the coincidence mode , i.e. without using the be - scatterer . in this paper , we present results in the energy range from 100 kev to 350 kev , obtained for seven solar flares ( x and m classes ) also selected from the rhessi instrument database .
we used a method based on the scattering of photons from detector to detector of rhessi that has previously been applied for polarization studies of gamma ray bursts (; ; ) . in [ sec :
method ] we explain how rhessi can be used as a compton polarimeter to study linear x - ray polarization at energies @xmath6 100 kev . the flares selected for analysis , and the criteria used for their selections is detailed in [ sec : flselec ] .
monte carlo simulations were performed to calculate the response of the instrument , and their results are discussed in [ sec : simulation ] .
the final polarization results are described and compared with previous measurements in [ sec : results ] .
the interpretation of the results is done in [ sec : interpretation ] by comparing with theoretical predictions .
finally , a brief summary of the conclusions of our work is given in
[ sec : conclusions ] .
the reuven ramaty high energy spectroscopic imager @xcite was designed to observe solar flares from 3 kev up to 17 mev .
it is capable of making spatially , spectrally and temporally resolved images of the sun @xcite using the rotation modulation principle (; ) .
the spacecraft is rotating with a period t @xmath7 4 seconds . the rhessi angular position in the solar coordinate system
is calculated using the satellite roll angle , which is continuously monitored by the spacecraft aspect systems (; ) .
the rhessi spectrometer @xcite consists of 9 cooled germanium detectors which are split in a thin front and a thick rear segment .
low - energy photons are mostly stopped in the front segments while high - energy photons can pass through and reach the rear segments .
the energy resolution is in the order of a few kev .
the arrangement of the 9 detectors in the spectrometer is sketched in figure [ fig : spectrometer ] .
each photon recorded by rhessi is characterized by its arrival time , deposited energy , detector number and segment .
these parameters are stored in the rhessi event list .
the time resolution of rhessi is equal to one binary microsecond ( 1b@xmath8s = @xmath9 sec ) .
polarization measurements are possible using photons that are compton - scattered from one detector into another one , making a signal in both of them . for such events ,
the effective area is very small , as most of the photons are either completely absorbed ( in photopeak reactions ) or scattered off the spectrometer @xcite .
therefore , only a small percentage of all photons observed by rhessi produces more than one entry in the event list .
the fact that rhessi rotates allows reducing the systematic errors in the polarization analysis .
the compton scattering probability of photons on free electrons is given by the cross section @xcite : @xmath10 where @xmath11 is the classical electron radius , @xmath12 is the angle between photon infall and outfall direction , and @xmath13 is the angle between the infall polarization and the outfall direction .
@xmath14 and @xmath15 are the energies of the photon before and after the scattering , respectively .
the cross section ( @xmath5 ) is minimal if @xmath13 = 0@xmath16 and maximal if @xmath13 = 90@xmath16 .
the dependence of the cross section on @xmath13 is most pronounced if @xmath12 = 90@xmath16 . as the rhessi axis is pointing to the sun ,
the detectors plane is perpendicular to the photon infall direction , and therefore @xmath17 for detector - to - detector scattering . in order to extract compton scattered photons from the rhessi event list ,
several cuts were applied ( see table [ tb : cuts ] ) .
.selection of cuts introduced to extract compton scattered photons from the event list . [ cols="<,<",options="header " , ] from the @xmath18 values of table [ tb : chi2 ] we can reject the model 5 from , with a 90% of confidence .
for the rest of the models the @xmath18 are very close to unity , preventing us from distinguishing between them . within the error bars ,
they all agree with the experimental data equally well .
the same is also valid for the 0% polarization hypothesis .
further refinement would require a polarimeter able to deliver data with error bars on the level of 12 percent .
measurements of hard x - ray polarization have been performed for six x - class and one m - class flares from the rhessi database .
our flare sample was identified after applying selection criteria to the signal strength , background levels and level of contamination by charged particles in rhessi detectors . the selected energy range , from 100 to 350 kev , connects the old (; ; ) and new @xcite results at low energies with the only measurement reported at a high energy band @xcite .
we found values for the polarization degree in the range between 2% and 54% , with statistical errors from 10% to 26% at the @xmath0 level .
the polarization angles are distributed between @xmath19 and @xmath20 in heliocentric coordinates .
they do not show any preferential orientation of the polarization , neither parallel nor perpendicular , with respect to the radial direction defined by the position of the flare in the sun .
in addition , no significant dependency between the orientation of the polarization and the distance of the flares to the sun center was found .
the polarization orientation with respect to the line that joins the two major foot - points of the flare was also studied . for this purpose ,
images of the flares were constructed with rhessi at different energies .
the emission above @xmath2130 kev was found to be produced around the foot - points .
however , no correlation between polarization direction and foot - points orientation was found .
similarly , no relationship between solar flare intensity and polarization degree could be observed .
the polarization degree from the 23 july 2002 flare measured by at high energies is in agreement with our results at the @xmath22 level .
their conclusion about the orientation of polarization respect to the radial direction passing trough the flare position ( perpendicular to it for flares in the limb and parallel for flares close to center ) can not be confirmed by our measurements .
our angles are distributed between @xmath23 and @xmath24 independently on the flare location .
comparison with theoretical predictions is more complex , as the direction of the polarization is expected by the theory to change around 300 kev @xcite .
again , more observations with better accuracy are needed . regarding the low energy measurements , our data are in good agreement with the recent results from the spr - n instruments on board of the coronas - f satellite @xcite , typically within @xmath25 .
their large sample of 25 solar flares reveals low polarization degrees , providing @xmath2 upper limits from 8 to 40% .
unfortunately , no information is given about the polarization direction .
in the 29 october 2003 flare , for which the coronas - f team claims a polarization of 70% , rhessi measurements were contaminated by charged particles .
theoretical predictions of the non - thermal bremsstrahlung emission provide polarization levels of the order of 20% .
however , depending on the assumptions used by different authors , the expected polarization can differ not only by its value , but also by its orientation ( see , for example , the calculations at 100 kev from and ) .
as the statistical uncertainties provided by our instrument are in the same order as the model predictions , equally good agreement is found for any of them as well as for the case of 0% polarization . in order to distinguish between different models , polarimetry measurements of at least 2% accuracy
are needed .
only model number 5 from , predicting polarization up to 85% , could be rejected by our rhessi data . in this model
, the magnetic field has the same strength at the top and the bottom of the loop , and the electrons spiral at pitch angles close to @xmath26 .
rhessi has made the first steps towards the understanding of the polarization phenomena in solar flares above 100 kev , where non - thermal emission dominates .
measurements with accuracy better than 10 to 20% were , however , hardly possible .
this is due to its small effective area and high levels of flare - induced background .
continuation of such studies will require a dedicated polarimeter that must solve the problems inherent to the rhessi design .
emphasis should be put on increasing the effective area and improving the background rejection capabilities . in particular
, a better time resolution will reduce the number of accidental coincidences , and the optimization of the detector dimensions will improve the detection efficiency for the compton scattering . | the degree of linear polarization in solar flares has not yet been precisely determined despite multiple attempts to measure it with different missions .
the high energy range in particular has very rarely been explored , due to its greater instrumental difficulties .
we approached the subject using the reuven ramaty high energy spectroscopic imager ( rhessi ) satellite to study 6 x - class and 1 m - class flares in the energy range between 100 kev and 350 kev . using rhessi as a polarimeter requires the application of strict cuts to the event list in order to extract those photons that are compton scattered between two detectors .
our measurements show polarization values between 2% and 54% , with errors ranging from 10% to 26% in @xmath0 level . in view of the large uncertainties in both the magnitude and direction of the polarization vector
, the results can only reject source models with extreme properties . |
asymptotically anti - de sitter ( ads ) spacetimes continue to attract a great deal of attention , primarily because of the recently conjectured ads conformal field theory ( cft ) correspondence principle .
this principle posits a relationship between supergravity or string theory in bulk ads spacetimes and conformal field theories on the boundary .
it offers the tantalizing possibility that a full quantum theory of gravity could be described by a well understood cft - yang - mills theory ( albeit with the asymptotic ads criterion ) .
observables in the quantum gravity theory could be computed using quantum field - theoretic methods , and vice versa .
although there is no proof of the conjecture , a dictionary is emerging that translates between different quantities in the bulk gravity theory and its counterparts on the boundary , including the partition functions and correlation functions of each .
a set of quantities of particular interest are those associated with gravitational thermodynamics .
it has long been known that a physical entropy @xmath2 and temperature @xmath3 can be ascribed to a given black hole configuration , where these quantities are respectively proportional to the area and surface gravity of the event horizon(s ) @xcite .
other black hole properties , such as energy , angular momentum and conserved charge can also be given a thermodynamic interpretation @xcite .
more recently it has been shown that entropy can be associated with a broader and qualitatively different object called a misner string , which is the gravitational analogue of a dirac string @xcite . in general
the problem of defining and calculating gravitational thermodynamic quantities remains a lively subject of interest . because they typically diverge for both asymptotically flat and asymptotically ads spacetimes , a common approach toward evaluating them
has been to carry out all computations relative to some other spacetime that is regarded as the ground state for the class of spacetimes of interest .
this is done by taking the original action @xmath4 for gravity coupled to matter fields and subtracting from it a reference action @xmath5 , which is a functional of the induced metric @xmath6 on the boundary @xmath7 . conserved and/or thermodynamic quantities are then computed relative to this boundary , which can then be taken to ( spatial ) infinity if desired .
this approach has been widely successful in providing a description of gravitational thermodynamics in regions of both finite and infinite spatial extent @xcite .
unfortunately it suffers from several drawbacks .
the choice of reference spacetime is not always unique @xcite , nor is it always possible to embed a boundary with a given induced metric into the reference background .
indeed , for kerr spacetimes , this latter problem forms a serious obstruction towards calculating the subtraction energy , and calculations have only been performed in the slow - rotating regime @xcite .
an extension of this approach , which addresses these difficulties was recently developed based on the conjectured ads / cft correspondence @xcite .
since quantum field theories in general contain counterterms , it is natural from the ads / cft viewpoint to append a boundary term @xmath8 to the action that depends on the intrinsic geometry of the ( timelike ) boundary at large spatial distances .
this requirement , along with general covariance , implies that these terms be functionals of curvature invariants of the induced metric and have no dependence on the extrinsic curvature of the boundary .
an algorithmic procedure @xcite exists for constructing @xmath8 for asymptotically ads spacetimes , and so its determination is unique .
the addition of @xmath8 will not affect the bulk equations of motion , thereby eliminating the need to embed the given geometry in a reference spacetime . hence thermodynamic and conserved quantities
can now be calculated intrinsically for any given spacetime .
the efficacy of this approach has been demonstrated in a broad range of examples , all of them in the spatially infinite limit , where the ads / cft correspondence applies @xcite .
however this limit is an idealization , and can never be satisfied in reality .
the purpose of this paper is to investigate the effects of including @xmath8 for quasilocal gravitational thermodynamics for situations in which the region enclosed by @xmath7 is spatially finite .
there are several reasons for considering this .
although the ads / cft correspondence is conjectured to be valid only at infinity , from a gravitational viewpoint the inclusion of additional boundary functionals is not uniquely dependent upon this correspondence and could in fact be carried out even if the conjecture is ultimately found to be invalid .
it is , therefore of interest to explore the implications of @xmath8 outside of its original conceptual framework .
furthermore , the inclusion of @xmath8 eliminates the embedding problem from consideration whether or not the spatially infinite limit is taken , and it is of interest to see what its impact is on quasilocal thermodynamics . since @xmath8 is a functional of the induced metric
, it will in general depend upon parameters in the solution related to conserved quantities at infinity , and could conceivably alter the results obtained for quasilocal thermodynamics relative to a fixed background .
of course @xmath8 itself is no longer unique , since one could add any functional of curvature invariants that vanished at infinity in a given dimension .
however the expression for @xmath8 obtained by the algorithm given in ref .
@xcite is the minimal one required to eliminate divergences , and therefore merits study .
we specifically consider in this paper an investigation of the thermodynamic properties of the class of kerr and kerr ads black holes in the context of spatially finite boundary conditions , with the boundary action supplemented by @xmath8 . with their lower degree of symmetry relative to spherically symmetric black holes , these spacetimes allow for a more detailed study of the consequences of including @xmath8 for spatially finite boundaries .
although the inclusion of @xmath8 was originally proposed for asymptotically ads spacetimes , it has been demonstrated that it can be extended to asymptotically flat spacetimes @xcite under certain topological constraints @xcite .
we consider the implications of this extension for kerr black holes in spatially finite regions .
the outline of our paper is as follows .
we review the basic formalism in sec .
[ genfor ] . in the next two sections we compute the quasilocal energy , mass , and angular momentum for the kerr and kerr - anti - de sitter black holes respectively .
these quantities must be computed numerically for general values of the parameters , and we present our results graphically . we perform a simple stability analysis for each case , and compare this to previous analyses carried out at infinity @xcite .
we finish our paper with some concluding remarks .
the action is the sum of three kinds of terms .
the first two are a volume ( or bulk ) term @xmath9 \label{e3}\ ] ] and a boundary term @xmath10 chosen so that the variational principle is well defined .
the euclidean manifold @xmath11 has the metric @xmath12 , covariant derivative @xmath13 , and time coordinate @xmath14 which foliates @xmath11 into non - singular hypersurfaces @xmath15 with unit normal @xmath16 over a real line interval @xmath17 .
@xmath18 is the trace of the extrinsic curvature @xmath19 of any boundary(ies ) @xmath20 of the manifold @xmath11 , with the induced metric(s ) @xmath21 ; the manifold can have internal boundary components as well as a boundary at infinity , although only the latter will be needed in what follows . here
@xmath22 is the matter lagrangian and @xmath23 the cosmological constant , which , in what follows , will be taken to be either zero or negative .
in general @xmath24 and @xmath25 are both divergent when evaluated on solutions , as is the hamiltonian , and other associated thermodynamic quantities @xcite . rather than eliminating these divergences by incorporating a reference term in the spacetime @xcite , a new term @xmath26
is added to the action which is a functional only of boundary curvature invariants .
although there are a very large number of possible invariants one could add in a given dimension , only a finite number of them can eliminate the divergences at spatial infinity , which arise from @xmath24 and @xmath25 . for an asymptotically ads spacetime , these can be determined by an algorithmic procedure @xcite .
quantities such as energy , entropy , etc .
, can then be understood as intrinsically defined for a given spacetime , as opposed to being defined relative to some abstract ( and non - unique ) background , although this latter option is still available . in four dimensions
the algorithm yields @xmath27 where @xmath28 and the coefficients have been chosen @xcite so that the total action @xmath29 is finite .
we shall study the implications of including the terms in ( [ e5 ] ) for kerr - ads black holes in spatially finite regions .
although other counterterms ( of higher mass dimension ) may be added to @xmath8 , they will make no contribution to the evaluation of the action or hamiltonian due to the rate at which they decrease toward spatial infinity , and we shall not consider them in our analysis here . a generalization of the prescription ( [ e5 ] ) to spacetimes that are not asymptotically ads is @xcite @xmath30 which is equivalent to eq .
( [ e5 ] ) for small @xmath31 ( as well as fixed @xmath31 and large mean boundary radius ) , and which ( formally ) has a well defined limit for vanishing cosmological constant ( @xmath32 ) .
a similar formula , in which @xmath33 is replaced with the ricci scalar of a large nearly spherical 2-surface embedded in @xmath34 was proposed by lau @xcite .
we shall consider the implications of using eq .
( [ e7 ] ) for kerr black holes in spatially finite regions in the @xmath35 limit . as in the asymptotically ads case , although there are terms of higher mass dimension that could be included , we shall not consider these , since eq .
( [ e7 ] ) is sufficient to remove divergences at spatial infinity for this case . an algorithmic procedure , developedfor asymptotically flat cases yields eq .
( [ e7 ] ) , although there are topological restrictions on its applicability @xcite .
a thorough discussion of the quasilocal formalism has been given elsewhere @xcite and so we only briefly recapitulate it here .
the action can be written as a linear combination of a volume term eq .
( [ e3 ] ) , a boundary term eq .
( [ e4 ] ) and a counterterm which we shall take to be eq .
( [ e7 ] ) in either the small-@xmath31 or @xmath32 limits : @xmath36 under the variation of the metric @xmath37 one obtains : @xmath38^{\mu \nu } \delta g_{\mu \nu } + \int_{\partial { \cal m}}d^{3}x(p^{ij}+q^{ij})]% \delta \gamma _ { ij } \label{e9}\ ] ] where @xmath39 and latin letters are used as indices for tensors on hypersurfaces . decomposing the metric @xmath40 on the timelike boundary @xmath41 which connects the initial and final hypersurfaces @xmath42 yields after some manipulation @xmath43 where the coefficients of the varied fields are @xmath44 and we see that the counterterm variation @xmath45 supplants terms which come from a reference action in the original formulation of the quasilocal technique .
the standard assumption @xcite for the reference action is that it is a linear functional of the lapse @xmath46 and shift @xmath47 so that its contributions to @xmath48 and @xmath49 depend only on the induced two - metric @xmath50 , whose determinant we denote by @xmath51 .
this assumption does not hold for the counterterm prescription ( [ e7 ] ) .
since @xmath52 is the time rate of change of the action , @xmath48 is identified with the energy density on the surface @xmath53 , and the total quasilocal energy for the system is therefore @xmath54 and this quantity can be meaningfully associated with the thermodynamic energy of the system @xcite.using similar reasoning , the quantities @xmath49 and @xmath55 are respectively referred to as the momentum surface density and the spatial stress .
when there is a killing vector field @xmath56 on the boundary @xmath57 , an associated conserved charge is defined by @xmath58 provided there is no matter stress energy in the neighborhood of @xmath59 ( this assumption can be dropped , allowing one to compute the time rate of change of @xmath60 , if desired @xcite ) . when this holds , the value of @xmath60 is independent of the particular hypersurface @xmath61 , a property not shared by the energy @xmath0 . for boundaries with timelike ( @xmath62 ) and rotational ( @xmath63 ) killing vector fields ( which encompass all the metrics we consider in this paper )
we obtain @xmath64 provided the surface @xmath61 contains the orbits of @xmath65 .
these quantities are respectively the conserved mass and angular momentum of the system enclosed by the boundary .
note that they will both be dependent upon the location of the boundary @xmath61 in the spacetime , although each is independent of the particular choice of foliation @xmath61 within the surface @xmath59 . in the context of the ads
/ cft correspondence the limit in which the boundary @xmath61 becomes infinite is taken , and the counterterm prescription @xcite ensures that the conserved charges ( [ e18 ] ) are finite .
no embedding of the surface @xmath59 into a reference spacetime is required .
this is of particular advantage for the class of kerr spacetimes , in which it is not possible to embed an arbitrary two - dimensional boundary surface into a flat ( or constant - curvature ) spacetime @xcite .
the counterterm ( [ e7 ] ) includes the minimal number of boundary curvature invariants for which this holds in both the large-@xmath31 and @xmath66 limits .
we shall consider the extension of eq .
( [ e7 ] ) away from the @xmath67 limit in what follows . the class of metrics we shall henceforth consider are the kerr - ads family of solutions , whose general form is @xmath68^{2 } \nonumber\end{aligned}\ ] ] in @xmath69 dimensions , where @xmath70 and where @xmath71 .
the metric has two horizons located at @xmath72 , provided the parameter @xmath73 is sufficiently large relative to the other parameters , specifically @xmath74 in the limit @xmath32 , @xmath75 , and @xmath76 . for @xmath77 the metric ( [ kadsmet ] )
is that of pure ads spacetime ( or flat spacetime if @xmath32 ) , and for @xmath78 the metric is that of schwarzchild - ads spacetime which has zero angular momentum .
hence we expect the parameters @xmath79 and @xmath73 to be associated with the mass and angular momentum of the spacetime , respectively .
this can be confirmed using conformal completion methods @xcite , for which we find that the total mass @xmath80 and angular momentum @xmath81 are given by @xcite @xmath82 a result corroborated by counterterm methods @xcite .
we turn now to consider the extension of ( [ e7 ] ) away from the @xmath83 limit .
here we take @xmath32 , and so the action can be written as @xmath84 we write , for convenience , @xmath85 where @xmath86 and the energy of the system can be written as : @xmath87 where @xmath61 is a 2-dimensional surface defined by setting the radial coordinate to a constant value @xmath88 , and @xmath89 @xmath90 @xmath91 are given by @xmath92}{% \{(1+a^{2}z^{2})[(1+a^{2}-2m)a^{2}z^{2}+1+a^{2}+2ma^{2}]\}^{1/2 } } \nonumber \\ \varepsilon _ { 2 } & = & \frac{2}{n\sqrt{\sigma } } q_{2}^{ij}u_{i}u_{j\mbox { } } \label{km8 } \\ & = & \frac{1}{2}r\{(a^{4}-2m)a^{4}z^{4}+(2a^{2}-7ma^{2}+2 - 4ma^{4}-7m)a^{2}z^{2 } \nonumber \\ & & + 1+a^{2}+3ma^{2}+a^{4}m\ } \nonumber \\ & & \times \{(1 - 2ma^{2}z^{2}+a^{2}z^{2})[(1+a^{2}-2m)a^{2}z^{2}+1 + 2ma^{2}+a^{2}](1+a^{2}z^{2})^{3}\}^{-1/2 } \nonumber \\ \ \varepsilon _ { 3 } & = & \frac{2}{n\sqrt{\sigma } } q_{3}^{ij}u_{i}u_{j } \label{km9 } \\ & = & \frac{1}{2}r\{(1 - 2m^{2})(1+a^{2}-2m)a^{8}z^{10 } \nonumber \\ & & + ( 1 - 2m)(12m^{2}-4a^{2}m^{2}-24m-6ma^{2}+9 + 9a^{2})a^{6}z^{8 } \nonumber \\ & & + \left [ ( 2m-1)^{3}a^{4}-\left ( 36m^{3}-42m^{2}-7m+14\right ) a^{2}+(2m-1)\left ( 2m^{2}+15m-15\right ) \right ] a^{4}z^{6 } \nonumber \\ & & + \left [ -3(2m-1)(m-1)a^{4}-\left ( 8m^{3}+36m^{2}-25m-8\right ) a^{2}-8m^{2}-12m+11\right ] a^{2}z^{4 } \nonumber \\ & & + ( 3 - 8a^{4}m^{3}-3a^{4}+3m+15a^{2}m+4a^{4}m^{2}+6a^{2}m^{2})z^{2 } \nonumber \\ & & -1-m - a^{2}-2a^{2}m^{2}-3a^{2}m\ } \nonumber \\ & & \times \{(1 - 2ma^{2}z^{2}+a^{2}z^{2})^{5}[(1+a^{2}-2m)a^{2}z^{2}+1 + 2ma^{2}+a^{2}](1+a^{2}z^{2})^{3}\}^{-1/2 } \nonumber\end{aligned}\ ] ] where @xmath93 and @xmath94 and @xmath95 . for arbitrary values of @xmath96and @xmath97
, the term @xmath98 can be integrated analytically in terms of elliptic functions .
although the terms @xmath99and @xmath100 can not be analytically integrated , a term - by - term expansion of @xmath101 in powers of @xmath102 indicates that it is zero , a fact that can also be verified numerically . for small angular momentum @xmath102
, @xmath0 can be integrated easily : @xmath103+o(a^{4})]\}. \label{km10}\ ] ] in the asymptotic limit @xmath104the energy approaches the arnowitt - deser - misner ( adm ) energy @xmath97 , and for the case of @xmath105 one gets the thermodynamic energy of a schwarzschild black hole , which is equivalent to its mass .
the expression ( [ km10 ] ) was obtained by martinez @xcite using the quasilocal method with a flat reference spacetime in which the constant-@xmath106 boundary was embedded in the small-@xmath102 limit .
however even in this limit the embedding was technically quite complicated .
we find it interesting that the counterterm prescription ( [ km1 ] ) , extended to finite values of fixed @xmath107 gives exactly the same answer to this order in @xmath102 .
we next consider the total energy @xmath0 in ( [ km6 ] ) for @xmath108 where @xmath109 is the radius of the outer horizon . in this case
we can exactly compute @xmath0 in terms of elliptic functions : @xmath110 \label{kmexact}\ ] ] where @xmath111 , @xmath112 , and @xmath113 are the definitions of the elliptic functions @xmath114 and @xmath115 . in fig .. [ figure1 ] we plot the @xmath116-dependence of the energy at the horizon . in the slow - rotation approximation
it has been shown that @xmath117 , \label{km10a}\ ] ] where @xmath118 is the irreducible mass of the kerr black hole @xcite , the maximal amount of energy which can be extracted from a black hole via the penrose process .
it was conjectured in ref .
@xcite that the relation @xmath119 is an exact relation , valid for any value of @xmath102 .
we see using the counterterm prescription that this conjecture is not correct and that ( [ km10a ] ) is only an approximation which breaks down for @xmath120 , as shown in figs .
[ figure1 ] and [ figure2 ] .
we next consider evaluation of the energy for fixed arbitrary values of @xmath106 .
the calculation of @xmath121 must be carried out numerically , and in a manner which yields a well defined integral .
this means that the parameters of the integral must be such that the location of the quasilocal surface must be outside of the event horizon and the induced ricci scalar @xmath122 must be positive .
these criteria are satisfied provided that @xmath102 and @xmath97 are such that : @xmath123 we plot the dependence of @xmath124 in figs .
[ figure3 ] and [ figure 4 ] below , consistent with these constraints .
as one can see from fig .
[ figure3 ] , for small values of @xmath97 ( @xmath125 ) the energy is a weakly varying function of @xmath126 but as @xmath97 increases the @xmath102-dependence of energy becomes more relevant .
for example the average value of @xmath127 is approximately @xmath128 and @xmath129 for @xmath130 and @xmath131 respectively .
of course these values of @xmath95 depend on the value of the radius @xmath106 which is chosen for the quasilocal surface .
now we consider the quasilocal conserved charges of the kerr metric .
the kerr metric has two killing vectors and therefore there exist two conserved quantities .
the first one , which is due to the timelike killing vector @xmath132 , is associated with the conserved mass , @xmath133 of the system expressed as : @xmath134 where @xmath46 is the lapse function .
for the case of @xmath108 , we can explicitly compute the conserved mass : @xmath135 which should be compared with the expression for the quasilocal energy . in the limit of slow rotation it vanishes , a strikingly different feature from that of the energy as given in eq .
( [ km10a ] ) , which in this limit equals @xmath136 .
as @xmath137 we find that @xmath138 as expected .
it is clear from figs .
[ figure3 ] and [ figure 4 ] that the energy for a fixed value of @xmath102 is an increasing function of @xmath139 this is not the case for the conserved mass @xmath140 in ( [ km12 ] ) , which for fixed @xmath141 has a maximum value for @xmath142 , with the bound being saturated for the schwarzchild solution when @xmath143 .
we plot in figs .
[ figure 5 ] and [ figure6 ] the behavior of the conserved charge @xmath144 for fixed @xmath102 and @xmath97 respectively .
we find that the value of @xmath11 due to the @xmath145 third term in eq .
( [ km12 ] ) is very small compared to that of the first two terms .
the second non - trivial conserved charge is the @xmath146 component of the angular momentum , @xmath147 which can be expressed as : @xmath148 where @xmath149 is killing vector @xmath150 remarkably we find that the counter terms @xmath151 , while non - zero , give a vanishing contribution upon integration in ( [ km14 ] ) .
hence the total angular momentum is due entirely to the first term , which is : @xmath152 valid for any quasilocal surface of fixed @xmath88 .
we turn next to thermodynamic considerations of the kerr black hole .
the interior of the quasilocal surface can be regarded as a thermodynamic system which is co - rotating with an exterior heat bath whose angular velocity is @xmath153 @xcite , where @xmath154 is the angular velocity of the event horizon .
zero angular momentum observors ( zamo s ) at a given position outside of the black hole measure a proper inverse temperture @xmath155 and a proper angular velocity @xmath156 .
however quasilocal thermodynamic quantities are generally given by surface integrals over the boundary data rather than by products of constant surface data .
in particular , it is not possible to choose the quasilocal surface to simultaneously be both isothermal and a surface of constant @xmath157 and neither is true for the quasilocal surface we have chosen ( one at fixed @xmath88 ) .
furthermore it is by now well recognized that the entropy of a black hole depends only on the geometry of its horizon in the classical approximation , and is conversely independent of the asymptotic behavior of the gravitational field or of the presence of external matter fields . for stationary black holes ,
the entropy is one - quarter of the horizon area , yielding @xmath158 in this case .
we therefore shall regard the energy @xmath0 in eq .
( [ km6 ] ) as the thermodynamic internal energy for the kerr black hole within the boundary @xmath88 , view it as a function of @xmath106 , the entropy @xmath2 , and the angular momentum @xmath1 , and define the temperature and the angular velocity from their ensemble derivatives , integrated over the quasilocal boundary data . for the temperature
we obtain @xmath159 where @xmath160 is the surface gravity at the outer horizon .
the derivative @xmath161 is obtained by taking the derivatives of the terms in eqs .
( [ km7 ] [km9 ] ) above and then performing the integration with respect to @xmath93 .
it is straightforward to show that ( @xmath162 as @xmath137 , yielding the usual relationship between temperature and surface gravity .
however unlike the non - rotating case , it does not give the tolman redshift factor because a surface of fixed @xmath88 is not a surface of constant redshift ( it is not an isotherm ) .
now we investigate the @xmath102 and @xmath97 dependence of the temperature of a kerr black hole .
the @xmath102 dependence of the temperature is shown in fig .
[ figure7 ] .
for @xmath125 the temperature does not depend on @xmath102 very much ( except for @xmath163where @xmath164 as @xmath165 ) , but as @xmath97 increases ( @xmath166 ) the @xmath102-dependence of @xmath167 becomes more relevant .
for the case of @xmath168 as @xmath169 then @xmath167 goes to infinity and it goes to zero as @xmath102 goes to @xmath97 .
the @xmath97-dependence of @xmath167is shown in fig .
[ figure8 ] .
the @xmath102-dependence of @xmath170 can be seen from fig .
[ figure9 ] .
the temperature for a fixed value of @xmath1 does not increase monotically .
for example from fig .
[ figure9 ] for @xmath171 it can be seen that @xmath172 is zero at @xmath173 , and @xmath174 these two points correspond to the maximum and minimum of @xmath175 respectively .
@xmath175 for the schwarzschild case with @xmath105 has its minimum at @xmath176 for small values of @xmath102 , the situation is similar ; at a fixed value of the temperature @xmath167 below @xmath177 there will be only one black hole solution of small @xmath97 , in equilibrium with thermal radiation inside the box of characteristic radius @xmath88 , which will not be far from the extremal solution . at @xmath167 increases above @xmath177 two more possible solutions
appear , one of slightly larger @xmath97 and one of much larger @xmath97 . as @xmath167 increases above @xmath178 the two smaller @xmath97 coalesce , leaving only the largest @xmath97 solution .
as @xmath102 increases the region for the three distinct black hole solutions becomes vanishingly small , and for sufficiently large @xmath179 there will be only one black hole solution of large @xmath180 .
the heat capacity at constant surface area , @xmath181 is defined by : @xmath182 where the energy is expressed as a function of @xmath167 and @xmath183 expressing the energy and temperature as functions of @xmath184 and @xmath107 then the heat capacity can be written as @xmath185 for small values of @xmath97 ( @xmath186 ) , the heat capacity is not positive for all the allowed values of @xmath187 for example as one can see from fig .
[ fig13 ] for @xmath188 the heat capacity is positive only for @xmath189 .
indeed , for each value of @xmath97 @xmath190 there exists an @xmath102 for which the heat capacity becomes infinite . for @xmath191
the heat capacity is negative for a schwarzschild black hole .
however for a kerr black hole we see from fig .
[ figure 11 ] there are some values of @xmath102 for which the heat capacity is positive .
there is a phase of stable small black holes and a phase of stable large black holes , separated by a regime of intermediate unstable black holes , as illustrated in fig .
[ figure9 ] .
this fact was noted long ago by davies , who gave the relation @xmath192 for the values of @xmath193 at which the heat capacity becomes infinite @xcite in the limit @xmath137 . here since we use the quasilocal energy at @xmath88 instead of the adm mass parameters the relationship is considerably more complicated , and can only be obtained numerically .
however the qualitative features remain the same .
the @xmath102-dependence of the heat capacity for larger values of @xmath97 can be seen from figs .
[ fig13 ] and [ figure11 ] . for sufficiently large @xmath97 ,
the heat capacity is positive for all allowed values of @xmath102 , and there is a single phase consisting of a large black hole . expressing the energy as a function of @xmath107 @xmath2 and @xmath194
the angular velocity can be written as @xmath195 where @xmath196 is the angular velocity at the event horizon .
as @xmath197 , @xmath198 and thermodynamic chemical potential conjugate to the angular momentum approaches @xmath196 .
the @xmath102 dependence of @xmath199 is illustrated in fig .
[ figure13 ] .
as expected , @xmath199 decreases with decreasing @xmath102 .
however the derivative @xmath200 is not positive for all values of @xmath102 once @xmath97 becomes sufficiently large , or ( alternatively ) @xmath106 becomes sufficiently small .
this suggests that some kind of superradiance will take place for sufficiently small quasilocal boundaries . the numerical computation required to extrapolate
the curve for @xmath201 is more intense than the computational power we have available ; however we have explicitly checked that @xmath199 vanishes for small @xmath102 in this case .
we shall not pursue this issue further .
for the kerr - ads metric the action can be written as a linear combination of a volume term , a boundary term and the counterterm ( [ e7 ] ) in the small-@xmath31 limit : @xmath202 under variation of the metric , @xmath37 one obtains : @xmath203\delta g_{ij}+[\int_{\partial { \cal m}}\sqrt{-\gamma } % ( p^{ij}+q^{ij})]\delta \gamma _ { ij}\ } , \label{ka2}\ ] ] where now @xmath204 therefore the energy of the system can be written as : @xmath205 where @xmath206 and @xmath207 are given by @xmath208}{% \{(l^{2}+a^{2}l^{2}-a^{2}-2ml^{2}-a^{4})a^{2}z^{2}+a^{2}l^{2}-a^{2}-a^{4}+2a^{2}ml^{2}+l^{2}\}^{1/2}% } \ \ \ \ \ \ \ \
\nonumber\end{aligned}\ ] ] @xmath209\}^{1/2 } $ % \end{tabular } \label{ka6a}\ ] ] with @xmath94 , @xmath95 and @xmath210 . for @xmath105
we recover the quasilocal energy of a schwarzchild - ads black hole @xcite .
following our previous approach , we first consider special cases in which the integral in eq .
( [ ka4 ] ) can be evaluated . for @xmath108 , where @xmath211 is the radius of the horizon , we find that the total energy can be explicitly computed , yielding @xmath212 for small values of angular momentum , @xmath0 can be integrated easily .
then the total energy for small @xmath102 is : @xmath213+o(a^{4})\ } \label{ka7}\ ] ] when @xmath143 the energy is given by @xmath214 which differs from the expression obtained using background subtraction methods @xcite . for general values of the parameters ,
however , we can not analytically integrate ( [ ka4 ] ) exactly , and so we resort to numerical integration . to do this entails several restrictions on the parameters .
first , the kerr - ads solution is defined only for @xmath215 . furthermore , since the quasilocal surface must be outside of the horizon , @xmath216 must be real and less than unity .
note that there is no longer a restriction on the sign of the induced ricci scalar .
these criteria yield this fact that if one choose two of the three parameters @xmath97 , @xmath217 and @xmath102 , then the allowed values of the third parameter should obeys the following equations : @xmath218 where @xmath219 @xmath220 and @xmath221 given by ( [ kadscrit ] ) . for @xmath222 and @xmath223
there is no upper bound for @xmath224 and for @xmath225 the lower bound for @xmath102 is zero@xmath226 in the next four figures we plot the quasilocal energy as a function of @xmath102 and @xmath97 .
the @xmath102-dependence of the energy is given in figs .
[ figure 14 ] and [ figure15 ] .
we see that for small values of @xmath217 ( @xmath227 ) the energy is a slowly increasing function of @xmath187 for larger values of @xmath228 as in the case of kerr metric , for @xmath229 the energy decreases as @xmath102 increases .
figures [ figure 16 ] and [ figure 17 ] show the @xmath97-dependence and @xmath217-dependence of the energy . for large values of @xmath217
the energy is asymptotic to a linear function of @xmath217 @xmath230 this is due to the fact that the counterterm used for kerr - ads is proportional to @xmath217 for large values of @xmath217 . as in the case of kerr metric
we consider the quasilocal conserved charges of the kerr - ads metric which are the mass and @xmath146 component of angular momentum of the system defined in eqns .
( [ e19],[e20 ] ) .
the @xmath146 component of angular momentum due to the counterterm is zero and one can evaluate the total angular momentum as @xmath231 which is therefore valid for any quasilocal surface of fixed @xmath88 , and yields the angular momentum of the kerr metric as @xmath31 goes to infinity .
the total mass of the system , @xmath11 , is somewhat more complicated to evaluate . for @xmath108
it is @xmath232 where we have used the normalized killing vector @xmath233 @xmath234 instead of @xmath235 in eq .
( [ e19 ] ) so that the conserved quantities at infinity generate the @xmath236 algebra , in agreement with the conventions of ref .
the mass @xmath11 as @xmath237 goes to infinity then becomes @xmath238 which differs from that of eq.([kadsmj ] ) by this normalization factor .
the @xmath102 and @xmath97-dependence of @xmath11 are illustrated in figs .
[ figure 18 ] and [ figure 19 ] .
adopting the same stance as for the kerr black hole , we express the energy as a function of @xmath107 @xmath1 and @xmath2 . noting that the entropy is @xmath239 , the temperature can be written as @xmath240 where @xmath160 is the surface gravity at the event horizon @xmath241 and @xmath242 .
as for the kerr case , the temperature must be computed numerically .
we find ( fig . [ figure 20 ] ) that for small values of @xmath97 the temperature is a weakly decreasing function of @xmath102 , and ( similar to the kerr case ) it decreases rapidly as @xmath102 approaches its maximum allowed value . as @xmath97 increases , the @xmath102-dependence of @xmath167 becomes more relevant .
the @xmath217-dependence of the temperature can be seen from fig .
[ figure 21 ] . note that the value of @xmath167 due to the counterterm is very small compared to its value due to the boundary term . in contradistinction to the energy , we find that @xmath167 is an initially increasing function of @xmath217 for small @xmath97 . for l@xmath243e values of @xmath217 ( @xmath244 ) the change in @xmath167 due to
the boundary term is negligible and @xmath167 increases as @xmath217 increases .
the @xmath97-dependence of @xmath167 is plotted in [ figure 22 ] . in figure [ fig23 ]
we plot the @xmath245dependence of @xmath246 again this figure is in consistent with that of kerr case , and shows that for @xmath247 , @xmath175 has a maximum at @xmath248 and a minimum at @xmath249 we find for small values of @xmath102 a single stable phase consisting of a small black hole . as the temperature increases ,
a second stable phase emerges consisting of a large black hole , along with an intermediate - mass unstable phase .
this latter phase coalesces with the small black hole at an even larger temperature , beyond which there exists only the single stable large black hole phase .
as @xmath102 increases the small and unstable phases are present for continually smaller temperature ranges .
they eventually vanish for @xmath102 sufficiently large , leaving only the single large black hole phase . the heat capacity at constant surface area @xmath250 can be evaluated numerically , and the results , plotted as functions of @xmath102 and @xmath97 , are shown in figs .
[ figure 24 ] and [ figure 25 ] .
the stability analysis for these black holes is in qualitative agreement with that of caldarelli et .
@xcite , who recently performed a thorough analysis of the thermodynamic properties of kerr - newman ads black holes in the @xmath137 limit .
however they computed a generalization of the smarr formula for the ads case in terms of @xmath251 , whereas we are considering the quasilocal energy @xmath0 in ( [ ka4 ] ) as a function of @xmath106 , @xmath2 , and @xmath1 .
as a consequence the temperature they derive does not redshift to zero in the large-@xmath106 limit , unlike the expression in ( [ ka10 ] ) which does , although the explicit form differs from that of the tolman factor due to the quasilocal surface not being an isotherm when @xmath252 . the angular velocity can be written as : @xmath253 where @xmath254 is the angular velocity of the einstein universe : @xmath255 with @xmath256 the angular velocity of the hole at event horizon , and @xmath257 is the angular velocity at infinity . as @xmath137 , @xmath258 , and the results of ref .
@xcite are recovered .
the @xmath102 dependence of the angular velocity can be seen from [ figure 26 ] . as in the kerr case
, we find that @xmath259 is not positive once @xmath97 is sufficiently large . again , although the numerical computation required to extrapolate the curve for @xmath260 is more intense than the computational power we have available , we have checked explicitly that @xmath199 vanishes for small @xmath102 in this case .
we can analyze the ( 2 + 1)-dimensional axially symmetric black hole solution of einstein equation with a negative cosmological constant @xmath261 obtained by banados et.al . using methods similar to those given above . written in stationary
coordinates the btz metric is @xcite @xmath262^{2 } , \label{btz1}\ ] ] with @xmath263 the action becomes @xmath264 where we drop the counterterm corresponding to the induced ricci scalar since it makes no contribution at infinity in ( 2 + 1 ) dimensions . the analogue of eq .
( [ ka3 ] ) is @xmath265 from which it is easy to show that the energy is @xmath266 this expression is identical to the one given in ref .
@xcite , in which the reference spacetime was taken to be that of a zero - mass black hole .
we see here that the counterterm prescription at arbitrary @xmath237 is equivalent to this choice of spacetime background , and so all of the results of ref .
@xcite will hold .
our investigations of the boundary - induced counterterm prescription ( [ e7 ] ) at finite distances has allowed us to obtain a number of interesting results , which we recapitulate here .
we have been able to compute for the first time the energy , conserved mass , and angular momentum quasilocally for arbitrary values of the parameters of the kerr and kerr - ads solutions , apart from the mild reality restrictions ( [ kerrconds ] ) and ( [ kadscond ] ) .
these quasilocal quantities are intrinsically calculable at fixed @xmath88 without any reference to a background spacetime , a marked improvement over the methods given in ref .
@xcite , which were necessarily restricted to small values of the rotation parameter .
we recovered these results , and were further able to show that the conjecture that @xmath119 ( where @xmath267 is the irreducible mass of the kerr black hole ) was not correct , but instead was valid only for small @xmath102 .
we also found that quasilocal angular momentum equals its value at infinity at any fixed value @xmath88 .
the remarkable result holds because the counterterms integrate to zero , and not because they identically vanish .
we found that the entropy @xmath2 and angular momentum @xmath1 were not given by surface integrals over quasilocal boundary data , but rather were constants dependent on the parameters @xmath268 ( and @xmath31 ) of the black hole .
we therefore chose to regard the interior of the quasilocal surface as a thermodynamic system whose energy @xmath0 was a function of @xmath2 and @xmath1 .
a stability analysis yielded results that were in qualitative agreement with previous investigations carried out for either the non - rotating case @xcite , or at infinity @xcite .
this stability analysis is , of course , local , and phase transitions to other spacetimes of lower free energy might exist . for the btz black hole the boundary countermterm turned out to be formally identical to a contribution from a reference spacetime of zero mass @xcite , and so the thermodynamics was identical in the two cases
. it would be interesting to repeat these investigations for a different choice of quasilocal boundary , such as an isotherm .
the counterterm ( [ e7 ] ) reduces to the two cases given in eqs .
( [ km1 ] ) and ( [ ka1 ] ) the large-@xmath31 and @xmath66 limits respectively .
these two regimes are quite distinct except when the boundary curvature scalar @xmath269 , at which point one might expect that other boundary curvature invariants besides the minimal number we have considered here would be relevant .
however we have not found it necessary to include such terms . indeed
, boundary counterterms evidently have a robust degree of applicability , going well beyond that originally proposed from their origins in the ads / cft correspondence conjecture . in some ways
this is not too surprising , since one need have no _ a - priori _ knowledge of the conjecture to introduce such terms .
however we have found the somewhat unexpected result that inclusion of the minimal number of intrinsic boundary terms needed to make quantities well defined at infinity is sufficient to compute the thermodynamics of both asymptotically flat and asymptotically ads black holes for virtually all possible values of the rotation parameter and cosmological constant that leave the quasilocal boundary well defined .
this goes well beyond what is possible with background subtraction methods , and perhaps suggests that a deeper physical meaning be ascribed to such terms .
m.h.d . is grateful for the hospitality of the department . of physics of waterloo university , where this work was performed , and for partial support from the research council of shiraz university of iran .
we would like to thank i. booth for interesting correspondence .
this work was supported by the natural sciences and engineering research council of canada . _ _ | we consider the quasilocal thermodynamics of rotating black holes in asymptotically flat and asymptotically anti de sitter ( ads ) spacetimes . using the minimal number of intrinsic boundary counterterms inspired by the ads / conformal field theory correspondence , we find that we are able to carry out an analysis of the thermodynamics of these black holes for virtually all possible values of the rotation parameter and cosmological constant that leave the quasilocal boundary well defined , going well beyond what is possible with background subtraction methods .
specifically , we compute the quasilocal energy @xmath0 and angular momentum @xmath1 for arbitrary values of the rotation , mass and cosmological constant parameters for the ( 3 + 1)- dimensional kerr , kerr - ads black holes and ( 2 + 1)- dimensional banados teittelboim and zanelli ( btz ) black hole . we perform a quasilocal stability analysis and find phase behavior that is commensurate with previous analyses carried out at infinity .
watphys - th01/02 * quasilocal thermodynamics of kerr and kerr - anti - de sitter spacetimes * m.h .
dehghani and r. b. mann + department of physics , university of waterloo , + waterloo , ontario n2l 3g1 , canada + pacs numbers : 04.60.-m,04.70.-s,04.70.dy,11.25.db + + |
the explosion mechanism of core - collapse supernovae ( ccsne ) is still unresolved despite persistent efforts by many researchers over several decades .
recent state - of - art simulations show that the neutrino heating mechanism assisted by hydrodynamical instabilities revives the accretion shock , i.e. , the explosion results in .
however , estimated explosion energies are smaller than canonical value of order @xmath0 erg ( e.g. , * ? ? ?
* ; * ? ? ?
meanwhile , effects of magnetic field on the explosion dynamics has been studied well for the decade .
numerical simulations assuming a strong poloidal magnetic field ( typically @xmath1@xmath2 g at the pre - collapse phase ) , and rapid rotation in most cases , show that magnetic force assists in driving the energetic explosion ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the magnetic fields assumed in these simulations are so strong that the conservation of magnetic flux during collapse results in the field strength of @xmath3 g for the proto - neutron star surface .
this is comparable to inferred surface magnetic fields of magnetar candidates ( see * ? ? ?
* for review of magnetars ) .
at present , however , the strength of the magnetic field at the pre - collapse stage and the origin of strong magnetic field of magnetar are very uncertain .
stellar evolution simulations by @xcite , which implement tayler spruit dynamo model , show that the pre - collapse strength of the poloidal magnetic field in 15 @xmath4 star is only @xmath5 g. meanwhile , recent observations report that some ob stars in the main sequence stage possess an @xmath6 kg surface magnetic field ( e.g. , * ? ? ?
* ) , which corresponds to magnetar - class magnetic flux .
@xcite carried out a population synthesis calculation from main sequence stars to neutron stars , assuming the magnetic flux is conserved in the post - main - sequence evolution ( fossil field hypothesis ) .
their result implies that @xmath7% of ob stars have magnetar - class magnetic flux , while the majority have @xmath812 orders of magnitude weaker one .
even when the pre - collapse magnetic flux is weak , corresponding to the magnetic field of @xmath9 g for the proto - neutron star surface , mhd instabilities may amplify it to magnetar - class strength .
so far , there have been a small number of works focusing on this issue .
@xcite argued that convective dynamo in proto - neutron stars generates a magnetar - class magnetic field .
simulations performed by @xcite shows that standing accretion shock instability amplifies the magnetic field around the proto - neutron star surface from @xmath10 g to @xmath11 g. for rapidly rotating progenitors , magnetorotational instability ( mri , * ? ? ?
* ) may greatly amplify the magnetic field @xcite .
local box simulations characterizing a post - bounce core show that an initially weak magnetic field , @xmath10@xmath2 g inside the proto - neutron star , exponentially grows due to mri @xcite .
km , compared to the iron core , the global gradients of physical quantities are taken into account , by which they refer these simulations semi - global . ] in simulations of mri from a sub - magnetar - class seed magnetic field , @xmath9 g , a quite fine grid size compared to the scale of the iron core , e.g. , @xmath12 km , is required : the wavelength of maximum growing mode around the surface of the proto - neutron star is @xmath13 where @xmath14 is alfvn velocity . in order to attain such a high resolution , local simulations are adopted in previous works @xcite .
however , in local simulations , it is problematic to prepare a suitable background state : although an initially stationary background is usually used in local simulations , post - bounce cores are dynamical
. additionally , local simulations are incapable of studying the global dynamics . to relieve these problems , global simulations are desirable . in this letter , we report the first global simulations of mri in ccsne from a seed magnetic field of sub - magnetar - class flux . to ease high computational cost
demanded in global simulations , we carried out computations in axisymmetry and in limited radial range , 50500 km .
particular attention is paid to , ( i ) whether mri amplify the magnetic field to magnetar - class strength even in the dynamical background of ccsne , and ( ii ) whether the amplified magnetic field is strong enough to affect the dynamics .
in the simulations , we solve the following ideal mhd equations , employing a time - explicit eulerian mhd code , _ yamazakura _ @xcite : @xmath15 = -\rho(\nabla\phi)\cdot\mbox{\boldmath$v$}\label{eq.mhd.eng } , \\ & & \frac{\partial\mbox{\boldmath $ b$}}{\partial t}= \nabla\times\left(\mbox{\boldmath$v$}\times\mbox{\boldmath$b$}\right ) \label{eq.mhd.far},\end{aligned}\]]in which notations of the physical variables follow custom . here , @xmath16 is newtonian mono - pole gravitational potential .
computations are done with polar coordinates in two dimension , assuming axisymmetry and equatorial symmetry .
we adopt a tabulated nuclear equation of state produced by @xcite .
neutrinos are not dealt with in our simulations .
the electron fraction , which is required to obtain the pressure of a fluid element from the eos table , is given by a prescription suggested by @xcite .
we should note that the liebendrfer s prescription is only valid until bounce .
@xcite discussed that adopting this prescription in the post - bounce phase may lead to underestimation of pressure by upto 20% . before simulating mri
, we first follow the collapse of a 15 @xmath4 progenitor star ( s. e. woosley 1995 , private communication ) for the central region of 4000 km radius , until @xmath17 ms after bounce ( basic run ) .
after the central density reaches @xmath1 g @xmath18 , the core is covered with @xmath19 numerical grids , where the spatial resolution is 0.423 km .
the core is assumed to be rapidly rotating with the initial angular velocity profile of @xmath20 where @xmath21 km and @xmath22 rad s@xmath23 , corresponding to a millisecond proto - neutron star after collapse .
the same dipole - like magnetic field configuration as one employed by @xcite is initially assumed with the typical field strength of @xmath24 g around the pole inside a radius of 1000 km , with which the strength of @xmath25 g is obtained for the proto - neutron star surface
. the initial rotational energy and magnetic energy divided by the gravitational binding energy are @xmath26 and @xmath27 , respectively . to conduct high resolution global simulations to capture the growth of mri in the core , we restrict the numerical domain to a radial range of @xmath28km@xmath29 ( mri run ) .
the initial condition of mri run is set by mapping the data of basic run onto the above domain , at 5 ms after bounce . at that time , the density at the inner - most grids is @xmath30@xmath31 g @xmath18 .
the inner and outer radial boundary conditions for mri run are given by the data of basic run , except for @xmath32 .
the boundary values of @xmath32 are given so that the divergence free condition of the magnetic field is satisfied .
the grid spacing is such that the radial and angular grid sizes are same , viz .
@xmath33 , at the inner and outer most cells .
we perform mri runs with four different grid resolutions , where the grid sizes of the inner most cells , @xmath34 , ( and the numbers of grids , @xmath35 , ) are 12.5 m ( @xmath36 ) , 25 m ( @xmath37 ) , 50 m ( @xmath38 ) , and 100 m ( @xmath39 ) .
note that with above parameters , typical mri growth wavelength around the inner boundary is several 100 m at the beginning of mri runs .
we stop each mri run , before the shock surface reaches the outer boundary .
in each mri run , we found a clear exponential growth of the poloidal magnetic energy during @xmath4018 ms , where @xmath41 ms corresponds to the beginning of mri run ( see left panel of figure [ t - b ] ) . the growth timescale is larger for higher grid resolution , but almost converges to @xmath42
ms at mri run with @xmath43 m. afterward , the poloidal magnetic energy gradually increases , and roughly saturates around @xmath44 ms , in which the saturation energy is larger for higher grid resolution
. the evolution of toroidal magnetic energy does not show clear exponential growth in each mri run .
although the growth rate is initially larger compared with that of the poloidal magnetic energy , it becomes smaller after the exponential growth of the poloidal magnetic energy sets in . at saturation , in mri run with @xmath45 m ,
the poloidal magnetic energy is twice as large as the toroidal magnetic energy .
we also follow the variation of the poloidal magnetic field structure , for mri run with @xmath45 m. around @xmath40 ms , when the exponential growth begins , magnetic field lines around the pole in the vicinity of the inner boundary start to bend . during the exponential growth phase ,
the magnetic field lines in this region are further stretched , and filaments of strong magnetic field , @xmath11 g , appear ( see panel ( a ) of figure [ multi2d ] for @xmath46 ms ) .
after the exponential growth ceases , the topology of magnetic field lines becomes rather tangled , which implies that the flow becomes turbulent ( see panel ( b ) for @xmath47 ms ) .
subsequently , filaments of strong magnetic field and the turbulent flow pattern , which first appear around the pole , also emerge in other regions .
this phase corresponds to the duration of the gradual increase of the poloidal magnetic energy , @xmath4850 ms , observed in figure [ t - b ] .
after the poloidal magnetic energy saturates ( @xmath44 ms ) , a considerable fraction inside a radius of 150 km is dominated by turbulent flow , where the strength of the magnetic field reaches @xmath11@xmath49 g in filamentous flux tubes ( see panel ( c ) for @xmath50 ms ) . the bent magnetic field lines , the exponential growth of the poloidal magnetic field , the fact the growth rate is larger for higher gird resolution , and the turbulence after the exponential growth , all invoke the occurrence of mri in our simulations .
as seen in panel ( e ) of figure [ multi2d ] , the number of grids covering the maximum growing wavelength in mri run with @xmath45 m is more than 20 in most areas , although it is less than five in some limited locations in the vicinity of the pole and equator .
thus , our highest - resolution mri run seems almost capable to fairly capture the linear growth of mri . according to @xcite , the growth timescale obtained by the linear theory is @xmath51 where @xmath52 in panel ( f ) of figure [ multi2d ] , the distribution of @xmath53 for mri run with @xmath54 m at 4 ms is depicted .
it is found that , except for the very vicinity of the pole , where the grid resolution is not enough , the growth timescale is shortest , @xmath5510 ms , around the pole in the vicinity of the inner boundary .
this is consistent with the fact that magnetic field amplification first occurs there .
additionally , the numerically estimated growth timescale , @xmath42 ms ( figure [ t - b ] ) , approximately coincides with the above theoretical value .
we also estimated @xmath53 without the effect of the gradient of entropy and the lepton fraction , by setting @xmath56 in equation ( [ tmri ] ) , but found insignificant difference from the original one , which indicates that convection does not much contribute to amplify the magnetic field .
with all the above facts , we conclude that amplification of poloidal magnetic field found in our simulations is mainly due to mri .
since axisymmetry is assumed in our simulations , the toroidal component of the magnetic field is not directly amplified by mri .
the growth of the toroidal magnetic energy , which is slower than that of the poloidal magnetic energy , seems due to winding of magnetic field lines by differential rotation . in local simulations performed by @xcite ,
the pattern of coherent channel flows appears during the exponential growth phase . in our simulations , on the other hand , such a pattern is never observed .
although chanel - like configurations are locally observed in the exponential growth phase ( see panel ( d ) of figure [ multi2d ] ) , they are incoherent , and soon disrupted in timescale of milliseconds .
such immediate disruption of channel flows seems to be caused by dynamical flow motions in the background , which is difficult to be taken into account in local simulations . in left panel of figure [ spect ] ,
the energy spectra of the poloidal magnetic energy are plotted for mri run with @xmath45 m at different evolutionary phases .
each spectrum is derived for radial wave numbers , @xmath57 , and for a region @xmath58 and @xmath59 . at @xmath60
ms , the spectrum shows the dominance of a large scale , @xmath61 km , just reflecting the structure of the background magnetic field . during the linear growth phase , @xmath4018 ms , smaller scale structures , @xmath62 km , grow fast compared with the larger ones .
consistently , we found that the maximum growing wavelength of mri around the pole is generally @xmath6310 km during this phase . at the end of the simulation ( saturation phase ) , the spectrum shows almost flat distribution for @xmath64 km , and steep decay for @xmath651 km . in the range between them ,
a slope close to @xmath66 is observed , which seems to correspond to the inertial region of kologomorov s theory .
such spectrum affirms that the flow is turbulent .
also , it indicates that large scale components of the magnetic field , comparable to the size of the proto - neutron star , are produced in our simulations .
since the maximum growing wavelength of mri is generally smaller than this scale , as mentioned above , the large scale components are likely to be a result of inverse cascade .
we also compare the average spectra of the poloidal magnetic energy for @xmath6770 ms among mri runs ( right panel of fig [ spect ] ) .
it is found that in a lower resolution run , the spectrum tends to turn the steep decay at larger scale , which seems to be due to a larger numerical diffusion .
it is likely that this causes smaller spectral energy for a low resolution run over whole scales , and smaller saturation level observed in figure [ t - b ] .
figure [ ibeta ] shows that magnetic pressure reaches @xmath68% of matter pressure in some locations , which implies that the magnetic field mildly affects the dynamics .
note , however , that we did not obtain the convergence of the saturation magnetic field , and a larger saturation level is attained for a higher resolution run ( figure [ t - b ] ) .
this suggest that the simulation with a high - enough resolution might result in a larger impact of the magnetic field on the dynamics .
we have carried out the first global simulations of mri in ccsne from a sub - magnetar - class seed magnetic field , @xmath25 g around the surface of the proto - neutron star , and rapid rotation , in which two - dimensional ideal mhd equations are solved with assumptions of axisymmetry and equatorial symmetry . as a result of computations , we found that mri greatly amplifies the magnetic field to magnetar - class strength , @xmath69@xmath49 g , and that the magnetic field at saturation phase is dominated by large scale structure .
although , the impact of the magnetic field on the dynamics found to be mild , we do not obtain the convergence of the saturation magnetic field strength , which is larger for a higher resolution run .
a simulation with a higher resolution is necessary to assess the actual impact .
we also found that the evolution of flow pattern in our global simulations are quite different from those appear in local simulations by @xcite . since mri and turbulence are intrinsically non - axisymmetric phenomena , three - dimensional simulations are , in fact , necessary to lead to conclusive results .
note , however , that even the world s best computers may only capable to simulate global mri from a magnetar - class seed magnetic field in three dimension .
@xcite found that the main difference between two - dimensional and three dimensional simulations is that the former results in a larger saturation magnetic field compared with the latter .
they argued that the smaller saturation level in three dimension is caused by the disruption of coherent chanel flows before they grow well due to non - axisymmetric parasitic instabilities .
such difference might not appear in global simulations , since our simulations show that coherent chanel flows do not develop even in two dimension .
while the present simulations set the position of the inner boundary at @xmath70 km , we found in basic run that most rotational energy is reserved inside a radius of 50 km .
it is expected that a simulation with the inner boundary located at a smaller radius would result in a larger saturation magnetic field .
what we obtained in this study may be a lower limit .
although , in our simulations , a sub - magnetar - class magnetic field is amplified to magnetar - class strength due to mri , we should be cautious in concluding that mri can be the origin of magnetar fields . before that , further investigations may be required on the sustainability of a large scale , strong magnetic field until , e.g. , the formation of the neutron star .
additionally , it is worth investigating whether magnetar - class strength is also attained by mri from a weaker seed magnetic field , e.g. , @xmath71 g around the proto - neutron star surface .
note that previous local simulations of mri , including ones in the context of accretion disks , show that a weaker seed magnetic field results in a lower saturation level ( e.g. , * ? ? ?
* ; * ? ? ?
the dependence of the saturation magnetic field on the rotational velocity , which also has been discussed in many local simulations ( e.g. , * ? ? ?
* ) , is another issue to be studied in future works .
h.s . is grateful to kenta kiuchi for useful advise on mpi parallelization .
numerical computations in this work were carried out at the yukawa institute computer facility .
this work is supported by a grant - in - aid for scientific research from the ministry of education , culture , sports , science and technology , japan ( 20105004 , 21840050 , 24244036 . ) | we performed the first numerical simulations of magnetorotational instability from a sub - magnetar - class seed magnetic field in core collapse supernovae . as a result of axisymmetric ideal mhd simulations
, we found that the magnetic field is greatly amplified to magnetar - class strength . in saturation phase ,
a substantial part of the core is dominated by turbulence , and the magnetic field possesses dominant large scale components , comparable to the size of the proto - neutron star .
a pattern of coherent chanel flows , which generally appears during exponential growth phase in previous local simulations , is not observed in our global simulations . while the approximate convergence in the exponential growth rate is attained by increasing spatial resolution , that of the saturation magnetic field is not achieved due to still large numerical diffusion .
although the effect of magnetic field on the dynamics is found to be mild , a simulation with a high - enough resolution might result in a larger impact . |
it is well known that for a cross - section data set , the ordinary least squares ( ols ) estimator is typically consistent for the coefficient vector of the best linear approximation to the conditional mean , even if the conditional mean is not necessarily linear @xcite .
such a `` robust '' nature of ols is one of the reasons why ols is popular in empirical studies ( * ? ? ?
* chapter 3 ) .
suppose now that a panel data set is available .
in such a case , we are typically interested in estimating the partial effects of the observed explanatory variables , @xmath5 , on the conditional mean of the dependent variable , @xmath6 , conditional on @xmath5 and the unobservable individual effect @xmath7 , where @xmath8 denotes the index for individuals and @xmath9 denotes the index for time . for that purpose ,
a popular strategy is to model the conditional mean @xmath10 $ ] as additive in both @xmath7 and @xmath5 , and linear in @xmath5 .
the focus is then on estimating the coefficient vector of @xmath5 .
the fixed effects ( fe ) estimator , which is identical to the ols estimator treating the individual effects as parameters to be estimated , is often used to estimate the coefficient vector .
while for models without strict exogeneity , such as dynamic panel data models , the fe estimator is generally inconsistent when @xmath0 ( the number of individuals ) goes to infinity and @xmath1 ( the number of time periods ) is fixed because of the incidental parameters problem @xcite , it is still a fundamental estimator .
in particular , when @xmath0 and @xmath1 jointly go to infinity , the fe estimator becomes consistent .
furthermore , after the incidental parameters bias is properly corrected , the fe estimator is known to have a centered limiting normal distribution provided that @xmath11 , which restricts @xmath1 to be mildly large but allows @xmath1 to be small relative to @xmath0 .
such asymptotic properties of the fe estimator under large @xmath0 and @xmath1 asymptotics have been extensively studied in the econometrics literature , especially for panel autoregressive ( ar ) models , partly motivated by the fact that panel data sets with mildly large @xmath1 have become available in empirical studies .
the previous discussion presumes that the model is correctly specified , that is , the conditional mean @xmath10 $ ] is truly additive in @xmath7 and @xmath5 , and linear in @xmath5 .
the goals of this paper are twofold .
the first objective is to study the asymptotic properties of the fe estimator under possible model misspecification when both @xmath0 and @xmath1 are large .
the asymptotics used is the joint asymptotics where @xmath0 and @xmath1 jointly go to infinity ( more precisely , we index @xmath1 by @xmath0 and let @xmath12 as @xmath13 ) .
assume that @xmath14 is independent and identically distributed ( i.i.d . ) , and for each @xmath15 , conditional on @xmath7 , @xmath16 is stationary and weakly dependent .
suppose that @xmath17 $ ] may not be additive in @xmath7 and @xmath5 , nor linear in @xmath5 . under this setting , we show that the probability limit of the fe estimator is identical to the coefficient vector on @xmath5 of the best _ partial _ linear approximation to @xmath18 $ ] , which gives some rational to use the fe estimator ( and its variant ) even when the model is possibly misspecified .
we regard this probability limit as a pseudo - true parameter ( see section [ sec : interpretation ] for the discussion on interpretation or plausibility of this probability limit ; especially if @xmath18 $ ] is indeed additive in @xmath7 and @xmath5 , and linear in @xmath19 , then the pseudo - true parameter coincides with the `` true '' coefficient on @xmath19 in @xmath20 $ ] ) .
we then establish the asymptotic distributional properties of the fe estimator around the pseudo - true parameter when @xmath0 and @xmath1 jointly go to infinity .
we demonstrate that , as in the correct specification case , the fe estimator suffers from the incidental parameters bias of which the top order is @xmath21 .
moreover , we show that , after the incidental parameters bias is completely removed , the rate of convergence of the fe estimator depends on the degree of model misspecification and is either @xmath3 or @xmath4 .
the second goal of the paper is to establish asymptotically valid inference on the ( pseudo - true ) parameter vector .
since the fe estimator has the bias of order @xmath21 , the first step is to reduce the bias to @xmath22 . for that purpose , one can use existing general - purpose bias reduction methods proposed in the recent nonlinear panel data literature .
for example , one can use the half - panel jackknife ( hpj ) proposed by @xcite .
we refer to @xcite and @xcite for alternative approaches on bias correction for fixed effects estimators in panel data models .
after the bias is properly reduced , the fe estimator has the centered limiting normal distribution provided that @xmath11 or @xmath23 depending on the degree of model misspecification .
we are then interested in estimating the covariance matrix or quantiles of the centered limiting normal distribution . to this end , we study the asymptotic properties of the clustered covariance matrix ( ccm ) estimator @xcite and the cross section bootstrap @xcite under the prescribed setting . we show that the ccm estimator ( with an appropriate estimator of the parameter vector ) is consistent in a suitable sense and can be used to make asymptotically valid inference on the parameter vector provided that @xmath11 or @xmath23 .
this shows that inference using the ccm estimator is `` robust '' to model misspecification .
also the cross section bootstrap can consistently estimate the centered limiting distribution of the fe estimator without any knowledge on the degree of model misspecification and hence is robust to model misspecification , and moreover , interestingly , _ without any growth restriction on @xmath1_. the second feature of the cross section bootstrap is notable and shows ( in a sense ) that the incidental parameters bias does not appear in the bootstrap distribution .
allowing for potential model misspecification is of importance in practice . in particular , this paper is of practical importance because it provides an interpretation for the fe estimator under potential misspecification , and additionally , it proposes methods for inference in linear panel data model with large @xmath0 and @xmath1 that are robust to model misspecification .
however , the study of estimation and inference for linear panel data models that are robust to model misspecification is scarce .
an exception is @xcite where he considered the lag order misspecification of panel ar models and established the asymptotic properties of the fe estimator under possible misspecification of the lag order .
however , his focus is on the incidental parameters bias and he did not study the inference problem on the pseudo - true parameter .
moreover , @xcite did not cover a general form of model misspecification .
this paper fills this void .
furthermore , the asymptotic properties of the ccm estimator and the cross section bootstrap when model misspecification and the incidental parameters bias ( in the coefficient estimate ) are present have not been studied in a systematic form and hence is under - developed .
@xcite investigated the asymptotic properties of the ccm estimator when @xmath0 and @xmath1 are large but did not allow the case where the incidental parameters bias appears , nor did he cover model misspecification .
@xcite studied the asymptotic properties of the cross section bootstrap when @xmath0 and @xmath1 are large , but ruled out the case where the incidental parameters bias appears , nor did he cover model misspecification as well .
hence we believe that this paper is the first one that establishes a rigorous theoretical ground on the use of the ccm estimator and the cross section bootstrap when model misspecification and the incidental parameters bias are present .
it is important to notice that , even without model misspecification , these asymptotic properties of the ccm and cross section bootstrap when the incidental parameters bias is present are new .
we conduct monte carlo simulations to evaluate the finite sample performance of the estimators and inference methods under misspecification .
we are particularly interested in the empirical coverage of the 95% nominal confidence interval .
the empirical coverage probability using the ccm and cross - section bootstrap , especially the cross section bootstrap applied to pivotal statistics , is good .
we also apply the procedures discussed in this paper to a model of unemployment dynamics at the u.s .
state level .
the results generate speed of adjustment of the unemployment rate towards the state specific equilibrium of about 17% .
in addition , the analysis of estimates indicates that increments in economic growth are associated with smaller unemployment rates .
the organization of this paper is as follows . in section [ sec : interpretation ] , we discuss the interpretation of fe estimator under misspecification . in section [ sec : asymptotics ] , we present the theoretical results on the asymptotic properties of the fe estimator under misspecification . in section [ sec :
inference ] , we presents the results on the inference methods . in section [ sec : mc ] , we report a monte carlo study to assess the finite sample performance of the estimators and inference methods , together with a simple application to a real data .
section [ sec : conclusion ] concludes .
we place all the technical proofs to the appendix .
we also include additional theoretical and simulation results in the appendix .
* notation * : for a generic vector @xmath24 , let @xmath25 denote the @xmath26-th element of @xmath24
. for a generic matrix @xmath27 , let @xmath28 denote its @xmath29-th element .
for a generic vector @xmath30 with index @xmath31 , @xmath32 .
let @xmath33 denote the euclidean norm .
for any matrix @xmath27 , let @xmath34 denote the operator norm of @xmath27 .
for any symmetric matrix @xmath27 , let @xmath35 denote the minimum eigenvalue of @xmath27 .
we also use the notation @xmath36 for a generic vector @xmath24 .
* note on asymptotics * : in what follows , we consider the asymptotic framework in which @xmath12 as @xmath13 , so that if we write @xmath13 , it automatically means that @xmath37 .
the limit is always taken as @xmath13 .
this asymptotic scheme is used to capture the situation where @xmath0 and @xmath1 are both large .
in this section we clarify the probability limit , which we will regard as a pseudo - true parameter , of the fe estimator under the joint asymptotics and discuss interpretation ( or plausibility ) of the pseudo - true parameter .
the discussion is to some extent parallel to the linear regression case but there is a subtle difference due to the appearance of individual effects .
suppose that we have a panel data set @xmath38 , where @xmath7 is an unobservable individual - specific random variable taking values in an abstract ( polish ) space , @xmath6 is a scalar dependent variable and @xmath5 is a vector of @xmath39 explanatory variables .
typically , the random variable @xmath7 , which is constant over time , represents an individual characteristic such as ability or firm s managerial quality which we would include in the analysis if it were observable ( see * ? ? ?
* chapter 10 ) .
assume that @xmath40 is i.i.d .
, and for each @xmath41 , conditional on @xmath7 , @xmath42 is a realization of a stationary weakly dependent process . here
the marginal distribution of @xmath43 is invariant with respect to @xmath31 .
typically , we are interested in estimating the partial effects of @xmath5 on the conditional mean @xmath18 $ ] with keeping @xmath7 fixed .
a `` standard '' linear panel data model assumes that the conditional mean is of the form @xmath44 with unknown function @xmath45 and vector @xmath46 , and redefines @xmath7 by @xmath47 since , in any case , @xmath7 is unobservable and modeling a functional form for the individual effect is virtually meaningless ( see * ? ? ?
* chapter 5 ) . in this paper
, the `` correct '' specification refers to that the conditional mean @xmath10 $ ] is written in the form @xmath48 , and `` model misspecification '' signifies any violation of this condition . for instance
, this can happen if there are omitted variables or if nonlinearity occurs in the model .
we discuss more details below ( see examples 1 and 2 below for concrete examples ) .
the fe estimator defined by @xmath49 is consistent for the coefficient vector on @xmath5 as @xmath0 goes to infinity and @xmath1 is fixed if the specification is correct ( for the moment , assuming that @xmath50 exists ) and additionally the strict exogoneity assumption @xmath51 = { \mathbb{e}}[y_{it } \mid { \bm{x}}_{it},c_{i}]$ ] is met .
if the strict exogoneity assumption is violated , then the fe estimator is not fixed-@xmath1 consistent , but as @xmath37 with @xmath0 , the fe estimator becomes consistent for the coefficient vector on @xmath5 provided that the specification is correct .
suppose now that the specification is not correct , i.e. , @xmath18 $ ] may not be written in the form @xmath44 , and consider the probability limit of the fe estimator when @xmath0 and @xmath1 jointly go to infinity .
proposition [ prop1 ] ahead shows that , subject to some technical conditions , we have , as @xmath13 and @xmath12 , @xmath52^{-1 } { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{it } { \widetilde}{y}_{it } ] = : { \bm{\beta}}_{0 } , \label{ptrue}\ ] ] where @xmath53 $ ] and @xmath54 $ ] ( for a moment , assume that some moments exist ) . to gain some insight
, we provide a heuristic derivation of this probability limit under the sequential asymptotics where @xmath37 first and then @xmath13 . by definition ,
we have @xmath55 since @xmath42 is weakly dependent conditional on @xmath7 , as @xmath37 first , we have @xmath56 , \ \bar{{\widetilde}{{\bm{x}}}}_{i } \stackrel{{\mathbb{p}}}{\to } \bm{0 } , \ t^{-1 } \sum_{t=1}^{t } { \widetilde}{{\bm{x}}}_{it } { \widetilde}{y}_{it } \stackrel{{\mathbb{p}}}{\to } { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{i1 } { \widetilde}{y}_{i1 } \mid c_{i}]$ ] and @xmath57 , so that @xmath58 \}^{-1 } \ { n^{-1 } \sum_{i=1}^{n } { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{i1 } { \widetilde}{y}_{i1 } \mid c_{i } ] \}$ ] . by the law of large numbers
, the right side converges in probability to @xmath59 as @xmath13 . in what follows ,
we discuss an interpretation of @xmath59 defined in ( [ ptrue ] ) .
a direct interpretation is that @xmath59 is the coefficient vector of the best linear approximation to @xmath60 $ ] , but this interpretation does not explain the connection with the primal object of estimating the partial effects of @xmath5 on @xmath18 $ ] .
however , the link emerges from the following discussion .
let @xmath61 be the best partial linear predictor of @xmath6 on @xmath62 , i.e. , @xmath63 = \min_{g \in l_{2}(c_{1 } ) , \bm{b } \in \mathbb{r}^{p } } { \mathbb{e } } [ ( y_{it } - g(c_{i})-{\bm{x}}_{it}'\bm{b})^{2}],\ ] ] where @xmath64
< \infty \}$ ] . by a simple calculation ,
the explicit solution @xmath65 is given by @xmath66 - { \mathbb{e}}[{\bm{x}}_{it } \mid c_{i}]'\bm{b}_{0 } , \ \bm{b}_{0 } = { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{it}{\widetilde}{{\bm{x}}}_{it}']^{-1}{\mathbb{e } } [ { \widetilde}{{\bm{x}}}_{it } { \widetilde}{y}_{it } ] = { \bm{\beta}}_{0},\ ] ] and hence @xmath59 is the coefficient vector on @xmath5 of the best partial linear predictor .
moreover , it is not difficult to see that @xmath61 is indeed the best partial linear approximation to @xmath18 $ ] , i.e. , @xmath67 - g_{0}(c_{i } ) - { \bm{x}}_{it}'\bm{b}_{0})^{2 } ] = \min_{g \in l_{2}(c_{1 } ) , \bm{b } \in \mathbb{r}^{p } } { \mathbb{e } } [ ( { \mathbb{e}}[y_{it } \mid { \bm{x}}_{it},c_{i } ] - g(c_{i})-{\bm{x}}_{it}'\bm{b})^{2 } ] .
\label{papprox}\ ] ] therefore , the vector @xmath59 defined by ( [ ptrue ] ) is identical to the coefficient vector on @xmath5 of the best partial linear approximation to @xmath18 $ ] in ( [ papprox ] ) . just as the coefficient vector of the best linear approximation to the conditional mean is a parameter of interest in the cross section case , as discussed in chapter 3 of @xcite , the vector @xmath59 here can be regarded as a plausible parameter of interest in the panel data case .
hence , in this paper , we consider @xmath59 to be a parameter of interest and treat @xmath59 as a pseudo - true parameter .
clearly if @xmath17 $ ] is indeed additive in @xmath7 and @xmath5 , and linear in @xmath5 , then @xmath59 coincides with the `` true '' coefficient on @xmath5 in @xmath17 $ ] . as in the cross section case , letting the approximation error denote @xmath68 , we have a regression form @xmath69 = 0 , \ { \mathbb{e } } [ { \bm{x}}_{it } \epsilon_{it } ] = 0
. \label{regression}\ ] ] importantly , the `` error term '' @xmath70 here may not satisfy the conditional mean restriction @xmath71 = 0 $ ] due to possible model misspecification . in ( [ regression ] )
, there are two scenarios on violation of the conditional mean restriction @xmath71 = 0 $ ] .
one is the case where @xmath71 \neq 0 $ ] with positive probability but @xmath72 = \bm{0}$ ] a.s .
the other is the case where @xmath72 \neq \bm{0}$ ] with positive probability . depending on these two cases ,
the asymptotic properties of the fe estimator _ do _ change drastically ( see section 3 ) .
generally , both cases can happen .
we give three simple examples to fix the idea . _
example 1_. panel ar model with misspecified lag order .
suppose that the true data generating process follows a panel ar(2 ) model @xmath73 = 0,\ ] ] where @xmath74 is such that @xmath75 and @xmath76 .
conditional on @xmath7 , @xmath77 is stationary and typically weakly dependent . by a simple calculation
, we have @xmath78 = c_{i}/(1-\phi_{1}-\phi_{2})$ ] . letting @xmath79 $ ] , we have @xmath80 . hence @xmath81 is independent of @xmath7 .
suppose now that we incorrectly fit a panel ar(1 ) model .
note that in this case , we have @xmath82 = c_{i } + \phi_{1 } y_{i , t-1 } + \phi_{2 } { \mathbb{e } } [ y_{i , t-2 } \mid y_{i , t-1},c_{i } ]
. \end{aligned}\ ] ] hence @xmath83 $ ] is generally a nonlinear function of @xmath7 and @xmath84 except for the cases where @xmath85 or the distribution of @xmath86 is normal .
here the solution of the equation @xmath87 = 0 $ ] is the first order autocorrelation coefficient of @xmath88 , i.e. , @xmath89 . letting @xmath90
, we have @xmath91 , i.e. , @xmath92 . by the independence of @xmath88 from @xmath7
, we have @xmath93 = { \mathbb{e } } [ { \widetilde}{y}_{i , t-1 } \epsilon_{it } ] = 0 $ ] a.s .
@xcite studied such panel ar models with misspecified lag order in detail .
this paper covers lag order misspecification as a special case . _
example 2_. static panel with a mismeasured regressor .
suppose the true data generating process is as following @xmath94 = 0.\ ] ] in addition , @xmath95 = 0.\ ] ] suppose that @xmath96 and @xmath97 are independent , and we incorrectly fit a model using @xmath98 instead of @xmath99 . here
we have @xmath100= x_{it}^ { * } + v_{it } - { \mathbb{e } } [ x_{it}^ { * } \mid c_{i } ] = { \widetilde}{x}_{it}^ { * } + v_{it},\ ] ] where @xmath101 $ ] , and @xmath53 = \phi { \widetilde}{x}_{it}^ { * } + u_{it}$ ] .
hence the solution to the equation @xmath102 = 0 $ ] is given by @xmath103^{-1 } { \mathbb{e } } [ { \widetilde}{y}_{it } { \widetilde}{x}_{it } ] = \frac{{\mathbb{e } } [ ( { \widetilde}{x}_{it}^{*})^{2}]}{{\mathbb{e } } [ ( { \widetilde}{x}_{it}^{*})^{2 } ] + { \mathbb{e } } [ v_{it}^{2 } ] } \phi.\ ] ] finally , by @xmath104 , we have @xmath105 = { \mathbb{e } } [ ( { \widetilde}{x}_{it}^{*})^{2 } \mid c_{i } ] \phi - ( { \mathbb{e } } [ ( { \widetilde}{x}_{it}^{*})^{2 } \mid c_{i } ] + { \mathbb{e } } [ v_{it}^{2 } ] ) \beta_{0}$ ] , which is generally non - zero .
example 3_. random coefficients ar model .
suppose that the true data generating process follows the following random coefficients ar(1 ) model : @xmath106 in this case , @xmath107 = c_{i } y_{i , t-1}$ ] .
it is routine to verify that @xmath108 suppose that we incorrectly fit a panel ar(1 ) model . here
= y_{it}$ ] and the solution of the equation @xmath109 = 0 $ ] is given by @xmath110}{{\mathbb{e } } [ y_{i , t-1}^{2 } ] } = \frac{{\mathbb{e } } [ c_{i}/(1-c_{i}^{2})]}{{\mathbb{e } } [ 1/(1-c_{i}^{2})]}.\ ] ] here @xmath111 , and @xmath112 = { \mathbb{e } } [ c_{i } y_{i , t-1}^{2 } - \beta_{0 } y_{i , t-1}^{2 } \mid c_{i } ] = \frac{c_{i}-\beta_{0}}{1-c_{i}^{2}},\ ] ] which is non - zero a.s .
if @xmath7 obeys a continuous distribution .
the results in this paper could be interpreted as corresponding to the pseudo - likelihood model . under the additional assumptions of independence and normality ,
the resulting ( conditional ) maximum likelihood estimator ( mle ) of @xmath113 ( given @xmath114 ) is identical to the fe estimator .
thus , the fe estimator defined in ( [ fe ] ) can be viewed as a pseudo - mle .
@xcite made an interesting observation about the pseudo - true parameter when the link function is misspecified for the generalized linear model ( see their corollary 1 ) .
that is , the pseudo - true parameter is proportional to the true one up to nonzero scalar .
however , their setting is significantly different from ours ; first of all in corollary 1 they assumed that the true model satisfies a generalized linear model but only the link function is misspecified , and only cross section data are available . in our case basically
no `` model '' is assumed ( and hence the `` true parameter '' is not well - defined in general ) , so that their result does not extend to our setting . in this paper
we focus on the fe estimator .
this is because the fe estimator is widely used in practice and , as we have shown , the probability limit of the fe estimator under misspecification admits a natural and plausible interpretation , parallel to the linear regression case .
there could be alternative estimators ; for example , we could consider the average of the individual - wise ols estimators , i.e. , let @xmath115 denote the ols estimator obtained by regressing @xmath6 on @xmath116 with each fixed @xmath8 , and consider the estimator @xmath117 . however , this estimator does not share the interpretation that the fe estimator possesses . in some cases
the probability limit of @xmath118 happens to be identical to @xmath59 , but not in general . to keep tight focus , we only consider the fe estimator in what follows .
in this section , we study the asymptotic properties of the fe estimator under possible model misspecification ( i.e. , @xmath10 $ ] is not assumed to be additive in @xmath7 and @xmath5 , nor linear in @xmath5 ) .
we make the following regularity conditions .
( a1 ) : : @xmath119 , where @xmath120 is a polish space .
@xmath14 is i.i.d . , and
for each @xmath15 , conditional on @xmath7 , @xmath16 is a stationary @xmath121-mixing process with mixing coefficients @xmath122 .
assume that there exists a sequence of constants @xmath123 such that @xmath124 a.s .
for all @xmath125 , and @xmath126 for some @xmath127 .
( a2 ) : : define @xmath79 $ ] and @xmath128 $ ] ( assume that @xmath129 $ ] and @xmath130 $ ] exist ) .
there exists a constant @xmath131 such that @xmath132 \leq m$ ] a.s . and @xmath133 \leq m$ ] a.s . , where @xmath127 is given in ( a1 ) .
( a3 ) : : define the matrix @xmath134 $ ] .
assume that the matrix @xmath27 is nonsingular . in condition ( a1 )
, @xmath7 is an unobservable , individual - specific random variable allowed to be dependent with @xmath5 in an arbitrary form .
condition ( a1 ) assumes that the observations are independent in the cross section dimension , but allows for dependence in the time dimension conditional on individual effects .
we refer to section 2.6 of @xcite for some basic properties of mixing processes .
condition ( a1 ) is similar to condition 1 of @xcite , and allows for a flexible time series dependence .
the mixing condition is also similar to assumption 1 ( iii ) of @xcite , while she allowed for cross section dependence .
the mixing condition is used only to bound covariances and moments of sums of random variables , and not crucial for the central limit theorem .
therefore , in principle , it could be replaced by assuming directly such bounds .
we assume the mixing condition to make the paper clear .
note that because of stationarity assumption made in ( a1 ) , the marginal distribution of @xmath43 is invariant with respect to @xmath31 , i.e. , @xmath135 .
, is drawn from the stationary distribution , conditional on @xmath7 . ]
the stationary assumption rules out time trends , but is needed to well - define the pseudo - true parameter @xmath136 , and maintained in this paper .
extensions to non - stationary cases will need different analysis and are not covered in this paper .
condition ( a2 ) is a moment condition . as usual
, there is a trade - off between the mixing condition and the moment condition .
condition ( a2 ) implies that @xmath137 \leq m'$ ] a.s .
for some constant @xmath138 . condition ( a3 ) is a standard full rank condition .
now we introduce some notation .
recall the fe estimator : @xmath139 where @xmath140 and @xmath141 . under conditions ( a1)-(a3 ) , the matrix @xmath142 on the right side is nonsingular with probability approaching one ( see lemma [ lem2 ] in appendix [ appendix a ] ) .
recall that @xmath59 is defined by ( [ ptrue ] ) and @xmath70 is defined by @xmath143 ( see the previous section ) .
define @xmath144 , \
d_{t } = \sum_{| k | \leq t-1 } \left ( 1 - \frac{|k|}{t } \right ) { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{1,1 } { \widetilde}{{\bm{x}}}_{1,1+k } ' ] .\ ] ] here and in what follows , for the notational convenience , terms like @xmath145 $ ] for @xmath146 are understood as @xmath147 $ ] , i.e. , @xmath148 : = { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{1,1-k } \epsilon_{1,1 } ] , \ k < 0.\ ] ] we shall obey the same convention to other such terms .
for example , @xmath149 : = { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{1,1-k } { \widetilde}{{\bm{x}}}_{1,1 } ' ] $ ] for @xmath146 . we first note that under conditions ( a1)-(a3 ) , both @xmath150 and @xmath151 are well behaved in the following sense .
[ lem1 ] under conditions ( a1)-(a3 ) , we have @xmath152 \| < \infty$ ] and @xmath153 \|_{\operatorname{op } } < \infty$ ] . all the technical proofs for section 2 are gathered in appendix [ appendix a ] .
we now state the asymptotic properties of the fe estimator .
define @xmath154 = \bm{0}$ a.s .
n , \ & \text{otherwise } , \end{cases } \qquad \text{and } \qquad \sigma_{nt } = { \mathbb{e}}\left [ \left ( \frac{1}{nt } \sum_{i=1}^{n } \sum_{t=1}^{t } { \widetilde}{{\bm{x}}}_{it } \epsilon_{it } \right ) ^{\otimes 2 } \right],\ ] ] where recall that @xmath36 .
[ prop1 ] suppose that conditions ( a1)-(a3 ) are satisfied .
letting @xmath12 as @xmath13 , we have : @xmath155 .
\label{bahadur}\end{aligned}\ ] ] therefore , we have : @xmath156 \stackrel{d}{\to } n(\bm{0},a^{-1 } \sigma a^{-1}),\ ] ] where @xmath157 ( the limit on the right side exists ) .
we stress that proposition [ prop1 ] holds without any specific growth condition on @xmath1 . also note that this proposition implies that @xmath158 as long as @xmath12
we discuss some implications of proposition [ prop1 ] .
first , proposition [ prop1 ] shows that the fe estimator has the bias term of the form @xmath159 which , following the literature , we call the `` incidental parameters bias ''
. there are two sources that contribute to the incidental parameters bias .
the main source is conditional correlation between @xmath160 and @xmath70 for @xmath161 conditional on @xmath7 , which arises from using @xmath162 instead of @xmath163 $ ] in @xmath164 .
another source , which only contributes to higher order terms , is conditional correlation between @xmath160 and @xmath5 for @xmath161 conditional on @xmath7 , which arises from using @xmath162 instead of @xmath163 $ ] in @xmath142 .
proposition [ prop1 ] makes explicit the incidental parameters bias of any order , which appears to be new even in the correct specification case ( but under the current set of assumptions ) .
the expansion ( [ bahadur ] ) is important in investigating the asymptotic properties of the cross section bootstrap in section 4 .
second , proposition [ prop1 ] shows that , aside from the incidental parameters bias , the rate of convergence of the fe estimator depends on the degree of model misspecification , i.e. , after the incidental parameters bias is completely removed , the fe estimator is @xmath165-consistent if @xmath166 = \bm{0}$ ] a.s . and
@xmath167-consistent otherwise .
in fact , the rate of convergence of the fe estimator ( after the incidental parameters bias is removed ) depends on the order of the covariance matrix of the term @xmath168 ) + \frac{1}{n } \sum_{i=1}^{n } { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{i1 } \epsilon_{i1 } \mid c_{i}].\ ] ] by this decomposition , @xmath169 ) \right \}^{\otimes 2 } \right ] + { \mathbb{e}}\left [ \left ( \frac{1}{n } \sum_{i=1}^{n } { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{i1 } \epsilon_{i1 } \mid c_{i } ] \right ) ^{\otimes 2 } \right ] \\ & = \frac{1}{nt } { \mathbb{e}}\left [ \left \ { \frac{1}{\sqrt{t } } \sum_{t=1}^{t }
( { \widetilde}{{\bm{x}}}_{1 t } \epsilon_{1 t } - { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{i1 } \epsilon_{i1 } \mid c_{1 } ] ) \right \}^{\otimes 2 } \right ] + \frac{1}{n } { \mathbb{e}}[{\mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1 } \epsilon_{1,1 } \mid c_{1}]^{\otimes 2}].\end{aligned}\ ] ] since @xmath16 is weakly dependent conditional on @xmath7 , the first term is @xmath170 . on the other hand ,
the second term is zero if @xmath171 = \bm{0}$ ] a.s . , but @xmath172 otherwise , which shows that @xmath173 if @xmath171 = \bm{0}$ ] a.s . and
@xmath174 otherwise .
intuitively , unless @xmath171 = \bm{0}$ ] a.s .
, @xmath175 is unconditionally equicorrelated across @xmath9 , so that @xmath176 has the slow rate @xmath177 ( if @xmath171 = \bm{0}$ ] a.s . , by condition ( a1 ) , the covariance between @xmath178 and @xmath179 converges to zero sufficiently fast as @xmath180 , so @xmath176 has the faster rate @xmath181 ) . in some cases , at least theoretically , it may happen that @xmath171 \neq \bm{0}$ ] with positive probability but for some nonzero @xmath182 , @xmath183 = 0 $ ] a.s . , which corresponds to the case where the matrix @xmath184^{\otimes 2}]$ ] is nonzero but degenerate .
in such a case , we have the expansion @xmath185 \notag \\ & = \bm{r}'a^{-1 } \left \ { \frac{1}{nt } \sum_{i=1}^{n } \sum_{t=1}^{t } ( { \widetilde}{{\bm{x}}}_{it } \epsilon_{it } - { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{i1 } \epsilon_{i1 } \mid c_{i } ] ) \right \ } + o_{{\mathbb{p}}}[n^{-1/2}\max \
{ n^{-1/2 } , t^{-1 } \ } ] , \label{expansion2}\end{aligned}\ ] ] where the leading term of the right side multiplied by @xmath165 is asymptotically normal with mean zero and variance @xmath186 < \infty$ ] .
this shows that , after subtracting the incidental parameters bias , the fe estimator may have different rates of convergence within its linear combinations .
moreover , the remainder term in the expansion ( [ expansion2 ] ) has the constant term of order @xmath177 , so that the extra bias term of order @xmath177 appears in such a case .
, it is shown that the remainder term in the expansion ( [ expansion2 ] ) is in fact further expanded as @xmath187 a^{-1}{\mathbb{e } } [ { \widetilde}{{\bm{x}}}_{1,1}\epsilon_{1,1 } ' \mid c_{1 } ] ] + o_{{\mathbb{p } } } [ n^{-1/2 } \max \ { n^{-1},t^{-1 } \}]$ ] . ] by this , if additionally @xmath188 goes to some positive constant , the limiting normal distribution of the right side on ( [ expansion2 ] ) multiplied by @xmath165 has a nonzero bias in the mean of which the size is proportional to @xmath189 .
however , such a case seems to be rather exceptional and we mainly focus on the case where the matrix @xmath190 is always nonsingular ( i.e. @xmath190 is nonsingular in either case of @xmath171 = \bm{0}$ ] a.s . or not ) .
it is possible to have an alternative expression of the bias term of order @xmath21 .
[ cor1 ] suppose that conditions ( a1)-(a3 ) are satisfied .
then we have : @xmath191 + o(t^{-1 } ) = : b+o(t^{-1}).\ ] ] in particular , the bias term of order @xmath21 in the expansion ( [ bahadur ] ) is rewritten as @xmath192 .
finally , we provide some comments on the relation to the previous work .
proposition [ prop1 ] is a nontrivial extension of theorem 2 of @xcite , in which he established the asymptotic properties of the fe estimator for panel ar models with exogenous variables allowing for lag order misspecification .
proposition [ prop1 ] allows for a more general form of model misspecification , including lag order misspecification as a special case , and exhausts the incidental parameters bias of any order .
[ rem : hansen ] proposition [ prop1 ] is related to @xcite .
@xcite considered a model @xmath193 with @xmath194=0 $ ] for all @xmath195 or @xmath196 = \bm{0}$ ] , and showed that the ols estimator is @xmath167-consistent if there is no condition on time series dependence and @xmath165-consistent if a mixing condition is satisfied for time series dependence .
what matters for the rate of convergence of the ols estimator in @xcite is the order of the covariance matrix of the term @xmath197 , which is @xmath198 in the `` no mixing '' case and @xmath170 in the mixing case .
while there is a similarity , proposition [ prop1 ] is not nested to his results in several aspects .
first , in proposition [ prop1 ] , the rate of convergence of the fe estimator ( after the incidental parameters bias is removed ) depends on the degree of model misspecification ( i.e. , @xmath199 = \bm{0}$ ] a.s . or not ) , rather than the assumption on time series dependence .
second , while his model covers panel data models with individual effects by considering @xmath200 to be transformed variables @xmath201 , the mixing assumption is not satisfied for the transformed variables as he admitted in footnote 3 , so that his theorem 3 does not apply to models with individual effects .
additionally , his assumption 3 essentially requires that @xmath202 = \bm{0}$ ] for all @xmath203 under our setting ( if we think of @xmath200 as transformed variables @xmath201 ) , so that his results do not cover the case where the incidental parameters bias appears . on the other hand , while we exclusively assume that @xmath16 is mixing conditional on @xmath7 , @xcite covered the case where no such mixing condition is satisfied .
therefore , the two papers are complementary in nature .
@xcite obtained a general incidental parameters bias formula for nonlinear panel data models , allowing for potential model misspecification , when @xmath204 is going to some constant .
their general result could be applied to the present setting in the case where @xmath171=\bm 0 $ ] .
however , the stochastic expansion ( [ bahadur ] ) is not derived in @xcite ; ( [ bahadur ] ) exhausts the incidental parameters bias up to infinite order , and covers the case where @xmath171 \neq \bm 0 $ ] .
this expansion is also the key for studying the properties of cross section bootstrap .
by proposition [ prop1 ] , the fe estimator has the bias of order @xmath21 . in many econometric applications ,
@xmath1 is typically smaller than @xmath0 , so that the normal approximation neglecting the bias may not be accurate in either case of @xmath72 = \bm{0}$ ] or @xmath205 \neq \bm{0}$ ] .
therefore , the first step to make inference on @xmath59 is to remove the bias of order @xmath21 and reduce the order of the bias to @xmath206 . under model misspecification ,
bias reduction methods that depend on specific models ( such as panel ar models ) may not work properly .
instead , we can use general - purpose bias reduction methods proposed in the recent nonlinear panel data literature .
for example , the half - panel jackknife ( hpj ) proposed by @xcite is able to remove the bias of order @xmath21 even under model misspecification .
suppose that @xmath1 is even .
let @xmath207 and @xmath208 .
for @xmath209 , construct the fe estimator @xmath210 based on the split sample @xmath211 .
then the hpj estimator is defined by @xmath212 . using the expansion ( [ bahadur ] ) and corollary [ cor1 ] , as @xmath12
, we have the expansion @xmath213 , \label{bahadurh}\ ] ] so that the bias is reduced to @xmath22 in either case .
therefore , we have @xmath214 provided that @xmath215 .
@xcite proposed other automatic bias reduction methods applicable to general nonlinear panel data models .
their bias reduction methods are basically applicable to the model misspecification case .
is large . ] alternatively , a direct approach to bias correction is to analytically estimate the bias term . in this case
, we typically estimate the first order bias term @xmath216 by using the technique of hac covariance matrix estimation ( see * ? ? ?
moreover , another alternative approach is to use bias reducing priors on individual effects ( see , for example , * ? ? ? * and references therein ) .
see also @xcite for a review on bias correction for fixed effects estimators in nonlinear panel data models . by the previous discussion ,
a bias corrected estimator @xmath217 typically has the expansion @xmath218 . \label{expansion3}\ ] ] given this expansion , provided that @xmath215 , the distribution of @xmath217 can be approximated by @xmath219 . statistical inference on @xmath59
can be implemented by using this normal approximation . as usual , since the matrices @xmath27 and @xmath176 are unknown , we have to replace them by suitable estimators .
a natural estimator of @xmath27 is @xmath142 defined in ( [ fe2 ] ) , which is in fact consistent ( i.e. , @xmath220 ) as long as @xmath12 ( see lemma [ lem2 ] in appendix [ appendix a ] ) .
we thus focus on the problem of estimating the matrix @xmath176 .
we consider the estimator suggested by @xcite : @xmath221 where @xmath222 and @xmath217 is a suitable estimator of @xmath59 .
proposition [ prop2 ] establishes the rate of convergence of @xmath223 .
all the technical proofs of this section are gathered in appendix [ appendix b ] .
[ prop2 ] suppose that conditions ( a1)-(a3 ) are satisfied .
let @xmath217 be any estimator of @xmath59 such that @xmath224 $ ] .
letting @xmath12 as @xmath13 , we have : @xmath225.\ ] ] in particular , as long as @xmath12 , we have @xmath226 . in proposition [ prop2 ] , @xmath217 can be the fe estimator . actually , using a bias corrected estimator instead does not change the rate of convergence of @xmath223 . however , in the case where @xmath1 is relatively small , the fe estimator can be severely biased , which may affect the finite sample performance of @xmath223 .
thus , it is generally recommended to use a bias corrected estimator of @xmath59 in the construction of @xmath223 .
the results of @xcite are not directly applicable to the asymptotic properties of @xmath223 described in proposition [ prop2 ] .
this is because the individual dummies that control for fixed effects are not included in the model of @xcite ( see also previous remark [ rem : hansen ] ) .
thus , in contrast with @xcite , @xmath227 is not asymptotically normal with mean zero unless @xmath228 .
the main reason is that @xmath223 has a bias of order @xmath229 due to using @xmath162 instead of @xmath230 $ ] .
this bias appears even if we could use @xmath70 in place of @xmath231 . however , for inference purposes , the rate of convergence given in proposition [ prop2 ] is sufficient , and we do not consider the bias correction to @xmath223 .
assume now that @xmath190 is nonsingular in either case of @xmath72 = \bm{0}$ ] a.s . or not .
consider testing the null hypothesis @xmath232 , where @xmath233 is a @xmath234 matrix with rank @xmath235 ( @xmath236 ) , and @xmath237 is a constant vector .
suppose that we have a bias corrected estimator @xmath217 having the expansion ( [ expansion3 ] ) . then under the null hypothesis ,
the @xmath9-type ( for @xmath238 ) and wald - type statistics @xmath239^{-1}(r{\widetilde}{{\bm{\beta } } } - \bm{r } ) , \label{wald}\ ] ] converge in distribution to @xmath240 and @xmath241 , respectively , provided that @xmath215 .
for example , for @xmath238 , provided that @xmath215 , @xmath242 where the second equality is due to the fact that @xmath243 and @xmath244 .
the same machinery applies to the wald - type statistic .
importantly , in the construction of @xmath9-type or wald - type statistics , we do not need any knowledge on the degree of model misspecification . the resulting estimator @xmath245 of the covariance matrix @xmath246 is often called the clustered covariance matrix ( ccm ) estimator in the literature .
the ccm estimator is popular in empirical studies .
the appealing point of the ccm estimator , as discussed in @xcite , is the fact that it is free from any user - chosen parameter such as a bandwidth .
the previous discussion shows that inference using a suitable bias corrected estimator and the ccm estimator is `` robust '' to model misspecification .
bootstrap is generally used as a way to estimate the distribution of a statistic ( see * ? ? ? * for a general reference on bootstrap ) . for panel data , how to implement bootstrap
is not necessarily apparent .
see @xcite for some possibilities in bootstrap resamplings for panel data .
we here study , among them , the cross section bootstrap .
let @xmath247 with @xmath248 .
the cross section bootstrap randomly draws @xmath249 from @xmath250 with replacement .
the bootstrap fe estimator @xmath251 is defined by @xmath50 in ( [ fe ] ) with @xmath252 replaced by @xmath249 . by this definition
, we can express @xmath251 as the following form : @xmath253 where @xmath254 is the number of times that @xmath255 is `` redrawn '' from @xmath250 .
the vector @xmath256 is independent of @xmath257 and multinomially distributed with parameters @xmath0 and ( probabilities ) @xmath258 .
we need to prepare some notation and terminology .
let @xmath259 denote the probability measure with respect to @xmath260 .
let @xmath261 $ ] denote the expectation under @xmath259 . given a vector valued statistic @xmath262 depending on both @xmath263 and @xmath260 , and a deterministic sequence @xmath264 , we write `` @xmath265 in probability '' if for every @xmath266 and @xmath127 , @xmath267 as @xmath13 , and `` @xmath268 in probability '' if for every @xmath127 and @xmath269 , there exists a constant @xmath270 such that @xmath271 for all @xmath272 ( recall that @xmath273 ) .
we are now in position to state the main result of this section .
[ prop3 ] suppose that conditions ( a1)-(a3 ) are satisfied .
letting @xmath12 as @xmath13 , we have : @xmath274 , \label{bahadurb}\ ] ] in probability .
therefore , provided that @xmath190 is nonsingular , we have : @xmath275 where the inequalities are interpreted coordinatewise .
interestingly , proposition [ prop3 ] shows that despite the fact that the original fe estimator has the incidental parameters bias of which the top order is @xmath21 , as shown in proposition [ prop1 ] , the bootstrap distribution made by applying the cross section bootstrap to the fe estimator does not have the incidental parameters bias . as a consequence , the bootstrap distribution approaching the centered normal distribution holds without any specific growth condition on @xmath1 .
in fact , this is not surprising .
the main source of the incidental parameters bias comes from the term @xmath276 , which is linear in the cross section dimension .
the bootstrap analogue of this term is thus @xmath277 , so that the difference of these terms has mean zero with respect to @xmath259 .
the same machinery applies to the term @xmath278 , so that the incidental parameters bias is completely removed in the bootstrap distribution .
the previous discussion also has the following implication : the cross section bootstrap can not be used as a way to correct the incidental parameter bias .
recall that in the cross section case , the bootstrap can be used to correct the second order bias coming from the quadratic term ; here the incidental parameters bias comes from the terms linear in the cross section dimension , so that the cross section bootstrap does not work as a way to correct the bias .
proposition [ prop3 ] shows that , for @xmath279 fixed and @xmath280 , @xmath281 we have @xmath282 where @xmath283 is the distribution function of the standard normal distribution ( recall that @xmath25 is the @xmath26-th element of a vector @xmath24 and @xmath28 denotes the @xmath29-element of a matrix @xmath27 ) .
suppose that we have a bias corrected estimator @xmath217 having the expansion ( [ expansion3 ] ) . then by a standard argument , we can deduce that @xmath284 provided that @xmath215 .
note that @xmath285 can be computed with any precision by using simulation .
moreover , the computation of @xmath285 does not require any knowledge on the speed of @xmath286 , and in this sense the cross section bootstrap is robust to model misspecification .
an analogous result holds for the hpj estimator ( see section 4.1 ) .
[ cor4 ] suppose that conditions ( a1)-(a3 ) are satisfied .
let @xmath287 denote the hpj estimator based on the bootstrap sample @xmath288 .
letting @xmath12 as @xmath13 , we have : @xmath289,\ ] ] in probability .
therefore , provided that @xmath190 is nonsingular , we have : @xmath290 this corollary follows directly from the definition of the hpj estimator and proposition [ prop3 ] , and hence we omit the proof . basically , the conclusion of corollary [ cor4 ] holds for other reasonable bias corrected estimators .
we do not attempt to encompass generality in this direction .
analogous results hold for pivotal statistics .
because of the space limitation , we push the formal results on pivotal statistics to the appendix ( appendix [ appendix ] )
. the higher order properties of the cross section bootstrap will be very complicated in this setting and we do not attempt to study them here .
however , it is of interest to quantify the order of the convergence in , say , ( [ bootb ] ) in appendix [ appendix ] , which is left to future research .
there are some earlier works on the bootstrap for panel data .
@xcite called the cross section bootstrap in this paper the `` block bootstrap '' and studied its numerical properties by using simulations , but did not study its theoretical properties .
@xcite developed the asymptotic properties of the cross section bootstrap when the strict exogeneity is met , hence excluding the possibility that the incidental parameters bias appears .
@xcite studies the asymptotic properties of the moving block bootstrap for panel data , which resamples the data in the time series dimension and hence is different from the cross section bootstrap .
importantly , while @xcite allowed for cross section dependence which we exclude here , she assumed that the number of time periods , @xmath1 , is sufficiently large ( typically @xmath291 ) so that the incidental parameters bias does not appear .
lastly , @xcite proposed to use the cross section bootstrap for inference for nonlinear panel data models such as panel probit models , but did not give any theoretical result .
the asymptotic properties of the cross section bootstrap were largely unknown when the incidental parameters bias appears , even without model misspecification , and the results in this section contribute to filling this void and give useful suggestions to empirical studies . in ( [ bfe ] ) ,
the weights @xmath260 are multinomially distributed .
it is possible to consider other weights .
a perhaps simplest variation is to draw independent weights @xmath292 from a common distribution with mean @xmath293 and variance @xmath294 , which corresponds to the _ weighted bootstrap _ ( see , for example , * ? ? ?
let @xmath295 denote ( [ bfe ] ) with @xmath254 replaced by these independent weights @xmath296 .
then the conclusion of proposition [ prop3 ] holds with @xmath251 replaced by @xmath295 and @xmath297 replaced by @xmath298 .
since the proof is completely analogous , we omit the details for brevity .
so far , we have discussed the distributional properties of the cross section bootstrap . given proposition [ prop3 ] ,
it is natural to estimate the asymptotic covariance matrix @xmath299 by the conditional covariance matrix of @xmath300 .
however , since convergence in distribution does not imply moment convergence , proposition [ prop3 ] does not guarantee that @xmath301 \to a^{-1}\sigma a^{-1}$ ] in probability .
see @xcite for some examples of inconsistency of bootstrap variance estimators . in linear regression models with pure cross section or time series data
, @xcite discussed bootstrap - based covariance matrix estimation . in @xcite
, they made a modification to the bootstrap least square estimator , to guarantee the bootstrap estimator to satisfy a uniform integrability condition .
their modification is a sort of `` trimming '' to the second moment matrix . by doing so , they established the consistency of the bootstrap covariance matrix estimator . in the present panel data case , while their modification is straightforward to adapt to the fe estimator , it is not clear whether it actually works at the proof level . in proposition
[ prop3 ] , what we have done are : ( i ) to implement higher order stochastic expansions to the fe estimator to exhaust the incidental parameters bias of any order ; ( ii ) to implement ( i ) to the bootstrap analogue ; ( iii ) to eliminate the incidental parameters bias by taking the difference . in step ( ii )
, we need expansions of @xmath302 , or more precisely @xmath303 .
simply bounding @xmath304 from above , as @xcite did in step 3 of the proof of their theorem 1 , will leave the bias in the bootstrap distribution and cause a problem in establishing uniform integrability .
we leave this as an open problem .
in this section , numerical examples to illustrate the methods discussed in this paper are provided .
both monte carlo experiments and a real data analysis are presented .
all the numerical experiments were performed on the statistical software r @xcite .
computer programs to replicate the numerical analyses are available from the authors .
we use several different designs of simulation experiments to assess the finite sample performance of the estimates and inference procedures discussed in the previous sections . in the first design , as a benchmark , we analyze the estimates and inference procedures under correct specification . the true data generating process ( dgp ) in the first design follows a panel ar(1 ) model and we ( correctly ) fit panel ar(1 ) .
the next four models are designed to study the estimates and inference procedures under misspecification .. ] in the second design , the true dgp follows a panel ar(2 ) model ( see example 1 in section [ sec : interpretation ] ) ; in the third design we extend the true ar(2 ) dgp and include two lags of exogenous regressors ; and in the last design , the true dgp follows a random coefficient ar(1 ) model ( see example 2 in section [ sec : interpretation ] ) . in each of these cases ,
we incorrectly fit a panel ar(1 ) model and estimate the slope parameter .
note that the first three cases correspond to `` @xmath305 = \bm{0}$ ] '' case and the last case corresponds to `` @xmath305 \neq \bm{0}$ ] '' case .
the following sample sizes are considered : @xmath306 and @xmath307 .
the number of monte carlo repetitions is 2,000 .
we consider four different estimates : the fe , gmm @xcite , hpj estimates , and the bias corrected estimate proposed by ( * ? ? ?
* equation ( 6 ) ) ( hk ) .
more precisely , the gmm estimate we compare is the one - step gmm estimate formally defined in equation ( 8) in @xcite .
we also investigate the small sample properties of the inference procedures , paying particular attention to the empirical coverage probability .
the nominal coverage is 95% .
note that we are trying to construct a 95% confidence interval for the pseudo - true parameter .
we consider and compare the following options for inference on the pseudo - true parameter : [ cols="^,^,^",options="header " , ] in the hk option , we use a simple consistent estimate of the asymptotic variance based on the formula in @xcite . here
csb refers to `` cross section bootstrap '' .
the number of bootstrap repetitions in each case is 1,000 .
note that the bias and variance formula in @xcite are not valid under the misspecified settings below ( nevertheless the hk estimate is consistent for the pseudo - true parameter when @xmath0 and @xmath1 jointly go to infinity as the difference between the fe and hk estimates are @xmath2 ) , hence it is natural to expect the hk option does not perform well in those cases ( as it is not designed for covering model misspecification ) .
also it is expected that the fe estimate suffers from the incidental parameters bias and hence the fe - ccm option will not work well .
the gmm estimate is formally not known to be consistent for the pseudo - true parameter here , but the result of @xcite suggests that it is the case ( and hence comparison with the gmm estimate makes some sense ) .
first and then @xmath37 , so his result is not directly transferred to our case .
it is of interest to study the asymptotic properties of the gmm estimate under misspecification when @xmath0 and @xmath1 jointly go to infinity , which is left to future research .
] however , it is expected that the gmm option will not perform well as it is not designed for covering model misspecification . the last
four options are expected to work reasonably well at least when @xmath1 is moderately large .
the precise description of the hpj - hpjpb option is the following : in the hpj - hpjpb , we use the hpj estimate as the center , and apply the cross section bootstrap to the @xmath9-statistic . the @xmath9-statistic here is constructed by using the hpj estimate together with the ccm estimate , and the initial estimate in construction of the ccm estimate is the hpj estimate . in the first example , the true data generating process ( dgp ) is a panel ar(1 ) model : @xmath308 where @xmath309 ( @xmath9 distribution with @xmath310 degrees of freedom ) , @xmath311 , and @xmath312 . in generating @xmath6 we set @xmath313 and discard the first 500 observations , using the observations @xmath314 through @xmath1 for estimation . in this case
, we correctly fit panel ar(1 ) and there is no misspecification in the model . the results are collected in table [ table.mc.0 ] .
* table [ table.mc.0 ] * : the fe estimate has large bias . the hk and gmm estimates are biased when @xmath1 and @xmath0 are small , but the bias decreases as @xmath1 and @xmath0 are large , respectively . the hpj estimate is approximately unbiased . regarding inference , the empirical coverage of fe - ccm is close to zero , likely due to the large bias in the fe estimate .
hk is under coverage .
gmm , as expected , has a good coverage property in this case , especially for large @xmath0 .
it is important to notice that the robust inference procedures , especially hpj - hpjb and hpj - hpjpb , also have good coverage under no model misspecification . in the next example
, the true dgp follows a panel ar(2 ) model : @xmath315 where @xmath309 and @xmath311 .
two cases for the parameters @xmath316 and @xmath317 are considered : @xmath318 and @xmath319 . despite that the true dgp is a panel ar(2 ) model , suppose that we incorrectly fit a panel ar(1 ) model and estimate the slope parameter .
when @xmath318 , the pseudo - true parameter is @xmath320 , and when @xmath319 , @xmath321 .
the results for these two cases are presented in tables [ table.mc.1 ] and [ table.mc.2 ] , respectively .
* table [ table.mc.1 ] * : the fe and gmm estimates are severely biased .
the hk estimate is also biased as it is not designed for handling the case where misspecification is present .
the hpj estimate is able to reduce the bias substantially .
the bias is small even for modest @xmath1 , such as @xmath322 and @xmath323 .
regarding the standard deviations , the hpj estimate has slight variance inflation relative to the fe estimate in the finite sample ( which is also observed in @xcite in a different context of estimation of nonlinear panel data models such as panel probit models ) . as for the empirical coverage , the fe - ccm , hk , and gmm options perform poorly due to the fact that the fe , hk , and gmm estimates are largely biased .
the other options , namely , hpj - ccm , hpj - feb , hpj - hpjb and hpj - hpjpb , perform reasonably well , but the hpj - hpjpb option , as expected , seems to be the best .
the coverage of hpj - hpjpb is about 92% for @xmath324 and @xmath323 , and close to the nominal @xmath325 .
* table [ table.mc.2 ] * : in this case , the incidental parameters bias is small and all the options perform relatively well .
however , regarding coverage , as @xmath0 grows , the performance of the fe - ccm , hk , and gmm options deteriorates since the ratio between the bias and the standard deviation becomes larger in each case . on the other hand , the other options , hpj - ccm , hpj - feb , hpj - hpjb and hpj - hpjpb , perform well regardless of the combination of @xmath326 .
* table [ table.mc.5 ] * : to extend the ar(2 ) example , we include exogenous regressors in the true dgp . in this case
, the true dgp is as following : @xmath327 where @xmath309 , @xmath311 , and @xmath328 . finally , the parameters @xmath329 and @xmath330 . in this case
, we incorrectly fit a panel model with regressors @xmath331 and report estimates of the slope parameter of the autoregressive term .
this pseudo - true parameter value on the autoregressive term is @xmath332 .
the results for this case are presented in tables [ table.mc.5 ] .
the results in table [ table.mc.5 ] show evidence that the proposed methods are effective in finite sample .
the bias of the hpj estimator is small , on the other hand the bias of other estimators are large . regarding inference , given the bias in the fe , hk and gmm ,
their respective coverage rates are poor .
but , the empirical coverage of hpj - hpjb and hpj - hpjpb are close to the nominal . in the fourth example
, the true dgp is @xmath333 where @xmath334 , @xmath335 .
this model appears in example 3 in section [ sec : interpretation ] .
as before , we incorrectly fit a panel ar(1 ) model and estimate the slope parameter .
the value of the pseudo - true parameter is @xmath336 .
the simulation results for this case are presented in table [ table.mc.4 ] .
* table [ table.mc.4 ] * : as in the previous cases , the fe estimate is largely biased and fe - ccm performs poorly due to the bias .
note that the decreasing speed of the standard deviation for the fe estimate as @xmath1 grows is relatively slow , which would reflect the fact that the convergence rate of the fe estimate ( without the bias part ) under this dgp is @xmath4 and not @xmath3 ( the asymptotic variance of the fe estimate consists of the part decreasing like @xmath198 and also the part decreasing like @xmath170 , so even in this case , it is not surprising that the standard deviation of the fe estimate in the finite sample slowly decreases as @xmath1 increases ) .
the hpj estimate is able to largely remove the bias , nevertheless there is slight variance inflation in the finite sample . in this example , the hk and gmm estimates are able to reduce the bias , to some extent , for large time series .
however , the empirical coverage of the hk and gmm options is still poor .
lastly , among the inference procedures , hpj - hpjpb works particularly well . in this section
, we apply the procedures discussed in the previous sections to a model of unemployment dynamics at the u.s .
state level . @xcite and @xcite studied this subject using a dynamic panel data model .
in particular , @xcite modeled the current unemployment rate @xmath337 as a function of both lagged unemployment rate and economic growth rate @xmath338 .
in addition , to capture state specific effects , the model includes state individual intercepts @xmath339 .
the model can be written as follows : @xmath340 or equivalently @xmath341 where @xmath342 and @xmath343 is an innovation term .
the model described in equation ( [ eq.app2 ] ) shows that changes in unemployment rate are determined by two observable components .
the first is an adjustment of the unemployment rate towards a `` natural '' or `` equilibrium '' rate of unemployment , @xmath344 .
this rate of unemployment equilibrium is allowed to vary across states . moreover ,
the speed of adjustment of the unemployment rate towards the state specific equilibrium is equal to @xmath345 .
similarly , the second factor determining changes in unemployment rate is a deviation of the economic growth rate around a constant equilibrium .
the data for the unemployment rate are taken from the u.s .
bureau of labor statistics for the 19762010 period .
data for the state product are per capita personal income ( thousands of dollars ) from the u.s .
bureau of economic analysis deflated by annual implicit price deflator .
the economic growth rate is taken to be the relative growth of the state product .
data are available for all 50 u.s .
states and washington d.c .
we have a panel data set of 51 subjects over 35 years ( @xmath346 and @xmath347 ) .
we consider and compare several different inference procedures : the fe estimate with its associated ccm estimate ( fe - ccm ) ; and the hpj estimate with inference using its associated ccm estimate and pivotal cross - section bootstrap , which we denote by hpj - ccm and hpj - hpjpb , respectively . for comparison , we also report the results for the one - step gmm , and the two - stage least squares ( tsls ) @xcite estimates . and
@xmath348 as instruments . ]
we present 90% and 95% confidence intervals in all cases .
note here that the hpj - ccm and hpj - hpjpb options are misspecification robust , so they provide meaningful inference even when misspecification is present .
the results for point estimates and confidence intervals are collected in table [ t.app ] panel a for 1976 - 2010 .
the hpj estimate of @xmath349 ( columns hpj - ccm and hpj - hpjpb ) is 0.830 , which implies that the speed of adjustment is approximately 17% per year .
the fe estimate is 0.790 , implying a speed of convergence around 21% , and the gmm estimate is 0.80 with speed of approximately 20% . finally , the tsls estimate for @xmath349 is smaller than other estimates , and the speed of adjustment is larger , close to 67% . regarding confidence intervals for @xmath349 , fe - ccm , hpj - ccm , hpj - hpjpb and gmm have confidence intervals with similar length , while the confidence intervals of tsls are substantially larger than the other options .
the results for hpj - ccm and hpj - hpjpb are very similar .
now we move our attention to the economic growth rate variable .
the hpj estimate of @xmath350 is @xmath351 , which in absolute value is slightly smaller than the fe - ccm and gmm estimates but larger than the tsls estimate .
the confidence intervals of hpj - ccm and hpj - hpjpb are similar .
the fe estimate of @xmath350 is @xmath352 .
this estimate is accompanied with relatively narrow confidence intervals and zero is not included in the intervals . however , the tsls estimate of @xmath350 @xmath353 , and zero is inside both the 90% and 95% confidence intervals in the tsls option . for robustness purposes we use different subsamples to estimate the model .
we consider two subsamples : ( i ) 19762001 ; and ( ii ) 19761991 .
the results are , respectively , collected in panels b and c of table [ t.app ] . in the first robustness exercise we drop the last 9 years of observations and consider a subsample of 26 years , from 1976 to 2001 .
these results are displayed in panel b. they show point estimates for both @xmath349 and @xmath350 close to those in panel a , although slightly larger in absolute value .
confidence intervals are also similar to those in panel a.
lastly , we consider an even smaller subsample with 19 years , from 1976 to 1991 .
the results are presented in table [ t.app ] panel c. except for tsls , the point estimates of @xmath349 are smaller than those in the full sample case and the confidence intervals shift to the left .
in particular , the fe and gmm estimates of @xmath349 decrease substantially , from 0.790 and 0.800 in the full sample case to 0.676 and 0.669 , respectively .
the hpj estimate also decrease to 0.721 from 0.830 in the full sample case , but not so largely as gmm . moreover , fe - ccm , hpj - ccm and hpj - hpj - hpjpb have substantially narrower confidence intervals than tsls . regarding the results on @xmath350 , the point estimates are not larger in absolute value than in the full sample case ( except for tsls ) with wider confidence intervals than those using the full sample .
this paper has considered fixed effects ( fe ) ( or within group ) estimation for linear panel data models under possible model misspecification , where the conditional mean @xmath354 $ ] may not be additive in @xmath7 and @xmath5 , nor linear in @xmath5 , when both the number of individuals , @xmath0 , and the number of time periods , @xmath1 , are large .
we make several contributions to the literature .
first , we have shown that the probability limit of the fe estimator is identical to the coefficient vector on @xmath5 of the best partial linear approximation to @xmath354 $ ] which we regard as the pseudo - true parameter .
moreover , we have established the asymptotic distributional properties of the fe estimator around the pseudo - true parameter when @xmath0 and @xmath1 jointly go to infinity , and shown that after subtracting the incidental parameters bias , the rate of convergence of the fe estimate depends on the degree of model misspecification and is either @xmath3 or @xmath4 .
secondly , we have developed asymptotically valid inference on the pseudo - true parameter vector .
we have established the asymptotic properties of the clustered covariance matrix estimator and the cross section bootstrap when both model misspecification and the incidental parameters bias ( in the coefficient estimate ) are present .
finally , we have conducted monte carlo simulations and evaluated the finite sample performance of the fe and its bias corrected estimators , and several inference methods and confirmed that the cross section bootstrap to pivotal statistics works particularly well .
these inference methods were applied to a study of the unemployment dynamics in the u.s . state level .
the authors would like to express their appreciation to ivan fernandez - val , kazuhiko hayakawa , seojeong lee , ryo okui , and participants in the 2013 econometric society north america summer meeting , and the ny camp econometrics viii , for useful comments and discussions regarding this paper .
we also would like to thank the editor , the associate editor , and three anonymous referees for their careful reading and comments to improve the manuscript .
99 alvarez , j. and arellano , m. ( 2003 ) .
the time series and cross section asymptotics of dynamic panel data estimators .
_ econometrica _ * 71 * 1121 - 1159 .
an , h.z . and huang , f.c .
the geometric ergodicity of nonlinear autoregressive models .
_ statistica sinica _ * 6 * 943 - 956 .
anderson , t.w . and hsiao , c. ( 1982 ) : formulation and estimation of dynamic models using panel data .
_ journal of econometrics _ * 18 * 47 - 82
. angrist , j. d. and pischke , j .- s .
_ mostly harmless econometrics : an empiricist s companion_. princeton university press .
arellano , m. ( 1987 ) .
computing robust standard errors for within - groups estimators . _
oxford bulletin of economics and statistics _ * 49 * 431 - 434 .
arellano , m. and bond , s. ( 1991 ) .
some tests of specification for panel data : monte carlo evidence and an application to employment equations .
_ review of economic studies _ * 58 * 277 - 279 .
arellano , m. and bonhomme , s. ( 2009 ) .
robust priors in nonlinear panel data models .
_ econometrica _ * 77 * 489 - 536 .
arellano , m. and bover , o. ( 1995 ) .
another look at the instrumental variable estimation of error - component models .
_ journal of econometrics _ * 68 * 29 - 51 .
arellano , m. and hahn , j. ( 2006 ) . a likelihood - based approximate solution to the incidental parameter problem in dynamic nonlinear models with multiple effects .
arellano , m. and hahn , j. ( 2007 ) .
understanding bias in nonlinear panel models : some recent development . in : _ advanced in economics and econometrics _ , vol .
iii . econometric society , ed . by r.w .
blundell , w.k .
newey , and p. torsten , pp .
381 - 409 , cambridge university press .
baglan , d. ( 2010 ) .
efficient estimation of a partially linear dynamic panel data model with fixed effects : application to unemployment dynamic in the u.s .
bertrand , m. , duflo , e. and mullainathan , s. ( 2004 ) .
how much should we trust difference in differences estimates ?
_ quarterly journal of economics _ * 119 * 249 - 275 .
bun , m.j.g . and carree , m.a .
bias - corrected estimation in dynamic panel data models
. _ journal of business and economic statistics _ * 23 * 200 - 210 .
bun , m.j.g . and kiviet , j.f .
the effects of dynamic feedbacks on ls and mm estimator accuracy in panel data models .
_ journal of econometrics _ * 132 * 409 - 444 .
cheng , g. and huang , j. ( 2010 ) .
bootstrap consistency for general semiparametric @xmath355-estimation .
_ annals of statistics _ * 38 * 2884 - 2915 .
davidov , y. a. ( 1968 ) .
convergence of distributions generated by stationary stochastic processes .
_ theory of probability and its applications _ * 13 * 691 - 696 .
dhaene , g. and jochmans , k. ( 2009 ) . split - panel jackknife estimation of fixed - effect models .
dhaene , g. and jochmans , k. ( 2010 ) .
an adjusted profile likelihood for non - stationary panel data models with fixed effects .
preprint . fan , j. and yao , q. ( 2003 ) .
_ nonlinear time series_. springer .
gonalves , s. ( 2011 ) . the moving bootstrap for panel linear regression models with individual fixed effects .
_ econometric theory _ * 27 * 1048 - 1082 .
gonalves , s. and white , h. ( 2005 ) .
bootstrap standard error estimates for linear regression . _ journal of the american statistical association _ * 100 * 970 - 979 .
hahn , j. and kuersteiner , g. ( 2002 ) .
asymptotically unbiased inference for a dynamic panel model with fixed effects when both @xmath0 and @xmath1 are large .
_ econometrica _ * 70 * 1639 - 1657 .
hahn , j. and kuersteiner , g. ( 2011 ) .
bias reduction for dynamic nonlinear panel models with fixed effects .
_ econometric theory _ * 27 * 1152 - 1191 .
hansen , c.b .
( 2007a ) .
generalized least squares inference in multilevel models with serial correlation and fixed effects .
_ journal of econometrics _ * 140 * 670 - 94 .
hansen , c.b .
( 2007b ) .
asymptotic properties of a robust variance matrix estimator for panel data when @xmath1 is large .
_ journal of econometrics _ * 141 * 597 - 620 .
horowitz , j.l .
( 2001 ) . _
the bootstrap_. in : _ handbook of econometrics _ , vol .
5 , ed . by j.j . heckman and e.e .
leamer , pp .
3159 - 3228 , elsevier .
kapetanios , g. ( 2008 ) . a bootstrap procedure for panel data sets with many cross - sectional units .
_ econometrics journal _ * 11 * 375 - 395 .
kiviet , j.f .
( 1995 ) . on bias , inconsistency , and efficiency of various estimators in dynamic panel data models .
_ journal of econometrics _ * 68 * 53 - 78 .
lancaster , t. ( 2000 ) .
the incidental parameter problem since 1948 .
_ journal of econometrics _ * 95 * 391 - 413 .
lee , y. ( 2012 ) .
bias in dynamic panel models under time series misspecification
. _ journal of econometrics _ * 169 * 54 - 60 .
lu , w. , goldberg , y. and fine , j.p .
( 2012 ) . on the robustness of the adaptive lasso to model misspecification .
_ biometrika _ * 99 * 717 - 731 .
ma , s. and kosorok , m. ( 2005 ) .
robust semiparametric @xmath355-estimation and the weighted bootstrap . _ journal of multivariate analysis _ * 96 * 190 - 217 .
neyman , j , and scott , e. ( 1948 ) .
consistent estimates based on partially consistent observations .
_ econometrica _ * 16 * 1 - 31 .
nickell , s. ( 1981 ) .
biases in dynamic models with fixed effects .
_ econometrica _ * 49 * 1417 - 1425 .
okui , r. ( 2008 ) .
panel ar(1 ) estimators under misspecification . _
economics letters _ * 101 * 210 - 213 .
okui , r. ( 2010 ) .
asymptotically unbiased estimation of autocovariances and autocorrelations with long panel data .
_ econometric theory _ * 26 * 1263 - 1304 .
ozaki , t. ( 1985 ) .
non - linear time series models and dynamical systems . in : _
handbook of statistics 5 _ ( edited by e. j. hannan , p. r. krishnaiah and m. m. rao ) , elsevier science .
phillips , p.c.b . and
sul , d. ( 2007 ) .
bias in dynamic panel estimation with fixed effects , incidental trends and cross section dependence .
_ journal of econometrics _ * 137 * 162 - 188 .
r development core team .
r : a language and environment for statistical computing .
r foundation for statistical computing .
shao , j. ( 1992 ) .
bootstrap variance estimators with truncation . _ statistics and probability letters _ * 15 * 95 - 101 .
white , h. ( 1980 ) .
using least - squares to approximate unknown regression functions . _ international economic review _ * 21 * 149 - 170 .
white , h. ( 1982 ) . maximum likelihood estimation of misspecified models .
_ econometrica _ * 50 * 1 - 25 .
wooldridge , j.m .
_ econometric analysis of cross section and panel data_. mit press .
yokoyama , r. ( 1980 ) .
moment bounds for stationary mixing sequences .
_ z. wahrscheinlichkeitstheorie verw .
gebiete _ * 52 * 45 - 57 .
in this appendix , we consider bootstrapping pivotal statistics . we keep the notation used in section [ sec : inference ] .
we first consider the bootstrap version of the ccm estimator ( here and in what follows , `` bootstrap '' means `` cross section bootstrap '' ) .
let us write @xmath356 then the bootstrap ccm estimator is defined by @xmath357 where @xmath358 and @xmath359 is the bootstrap version of a suitable estimator of @xmath59 ( we formally assume that @xmath359 can be written as a statistic of @xmath252 and @xmath260 ) , for example , @xmath360 or @xmath361 .
note that @xmath362 where @xmath363 .
then we have the following proposition . [ prop4 ]
suppose that conditions ( a1)-(a3 ) are satisfied .
let @xmath359 be such that @xmath364 $ ] in probability .
letting @xmath12 as @xmath13 , we have : @xmath365,\ ] ] in probability .
in particular , as long as @xmath12 , we have @xmath366 in probability .
proposition [ prop4 ] establishes the rate of convergence of @xmath367 .
this result is parallel to that in proposition [ prop2 ] .
note that the condition that @xmath364 $ ] in probability is satisfied with @xmath360 or @xmath368 .
assume now that @xmath190 is nonsingular in either case of @xmath72 = \bm{0}$ ] a.s . or not .
consider , for the sake of simplicity , testing the null hypothesis @xmath369 , where @xmath279 is fixed , and consider the @xmath9-statistic and its bootstrap version based on either the fe or hpj estimate : @xmath370 where @xmath371 is the @xmath372 vector such that @xmath373 and @xmath374 for @xmath375 . here because @xmath376 ( see lemma [ lem3 ] ) and @xmath377 in probability , we can deduce that under conditions ( a1)-(a3 ) , @xmath378 this holds as long as @xmath12 and does not require any specific growth restriction on @xmath1 .
moreover , when @xmath379 , under conditions ( a1)-(a3 ) , we have @xmath380 provided that @xmath215 .
in order to shed more light on the performance of the proposed methods , figure [ fig1 ] presets the bias and rmse of the fe and hpj estimators from a fixed cross - section when varying the time series @xmath1 . in these simulations , we only considered the ar(2 ) model given in the second example with @xmath318 .
we fix the cross - section dimension at @xmath381 .
the left panel shows the bias of the estimators as a function of @xmath1 .
the results show that the hpj estimator has a small bias for small @xmath1 , but the bias disappears even for relatively small @xmath1 . on the contrary , the bias in the fe is large , and it remains relatively substantial even for large time series .
the right panel displays the rmse for the estimators .
it shows a good performance of the hpj estimator for moderate time dimensions .
increases for @xmath381 . _ [ fig1 ] ] here we consider the case where the true dgp is @xmath382 where @xmath334 , @xmath311 , @xmath383 and @xmath384 . this model is a panel - data version of exponential ar ( expar ) models ( see * ? ? ?
note that by ( * ? ? ?
* example 3.2 ) , for each @xmath15 , the process @xmath385 ( which is independent of @xmath7 ) is geometrically ergodic , so that condition ( a1 ) is satisfied . as in the previous case ,
we incorrectly fit a panel ar(1 ) model and estimate the slope parameter .
the value of the pseudo - true parameter is @xmath386 .
the simulation results for this case are presented in table [ table.mc.3 ] .
* table [ table.mc.3 ] * : the fe estimate is severely biased and the fe - ccm option performs poorly because of the presence of large bias . on the other hand , the hpj estimate is able to largely remove the bias at the cost of slight variance inflation relative to the fe estimate in the finite sample . in this case , the hk and gmm estimates are able to reduce the bias to some extent , although gmm still presenting larger bias .
the empirical coverage of the hk option is reasonably well ( which is partly due to the fact that the panel expar model is `` close '' to the panel ar(1 ) model ) .
gmm is under coverage .
in addition , the empirical coverage of the hk and gmm options worsen as @xmath0 becomes large . among the other options , hpj - hpjpb works particularly well .
we shall recall here the notational convention . for a generic vector @xmath24
, @xmath25 denotes the @xmath26-th element of @xmath24 , and for a generic matrix @xmath27 , @xmath28 denotes the @xmath29-th element of @xmath27 .
moreover , for a sequence @xmath387 indexed by @xmath31 , we write @xmath388 . in this section , we introduce some inequalities for @xmath121-mixing processes , which will be used in the proofs below .
let @xmath389 denote a stationary process taking values in some polish space @xmath120 , and let @xmath390 denote its @xmath121-mixing coefficients .
[ thma1 ] let @xmath391 denote the @xmath392-field generated by @xmath393 ( @xmath394 ) .
pick any integer @xmath125 .
let @xmath395 and @xmath396 be real - valued random variables measurable with respect to @xmath397 and @xmath398 , respectively . if @xmath399 < \infty$ ] and @xmath400 <
\infty$ ] for some @xmath401 and @xmath402 such that @xmath403 , then we have @xmath404 - { \mathbb{e } } [ \xi ] { \mathbb{e } } [ \eta ] | \leq 12
( { \mathbb{e } } [ | \xi |^{q } ] ) ^{q^{-1}}({\mathbb{e } } [ | \eta |^{r } ] ) ^{r^{-1 } } \alpha(k)^{1-q^{-1}-r^{-1}}.\ ] ] to illustrate an application of davidov s inequality , suppose that @xmath405 , and assume that @xmath406 < \infty$ ] and @xmath407 for some @xmath408 .
then by davidov s inequality , we have @xmath409 with @xmath410^{2/q } \sum_{k=0}^{\infty } \alpha(k)^{1 - 2/q}$ ] . for bounding higher order moments ,
we make use of yokoyama s ( 1980 ) theorem 3 .
[ thma2 ] suppose that @xmath405 .
assume that @xmath411 = 0 $ ] and for some constants @xmath127 and @xmath412 , @xmath413 < \infty$ ] .
if @xmath414 , then there exists a constant @xmath415 independent of @xmath1 such that @xmath416 \leq c t^{r/2}.\ ] ] [ lem2 ] suppose that conditions ( a1)-(a3 ) are satisfied . as @xmath13 ( and automatically @xmath12 ) , we have : ( i ) @xmath142 is nonsingular with probability approaching one , and + @xmath422 ; ( ii ) @xmath423 = t^{-1 } b_{t } = o(t^{-1})$ ] ; ( iii ) @xmath424 = o_{{\mathbb{p } } } ( n^{-1/2 } t^{-1 } ) $ ] ; ( iv ) @xmath425 ) \stackrel{d}{\to } n(\bm{0},v_{1})$ ] , where @xmath426)({\widetilde}{{\bm{x}}}_{1,1+k}\epsilon_{1,1+k}-{\mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1+k}\epsilon_{1,1+k } \mid c_{1}])']$ ] ( the right side is absolutely convergent in @xmath427 ) . * part ( i ) * : recall that @xmath428 $ ] .
observe that @xmath429 $ ] and @xmath430 ) \\
& \quad + \frac{1}{n } \sum_{i=1}^{n } ( { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{i1 } { \widetilde}{{\bm{x}}}_{i1 } ' \mid c_{i } ] - { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1 } { \widetilde}{{\bm{x}}}_{1,1 } ' ] ) - \frac{1}{n } \sum_{i=1}^{n } ( \bar{{\widetilde}{{\bm{x}}}}_{i}\bar{{\widetilde}{{\bm{x}}}}_{i } ' - { \mathbb{e } } [ \bar{{\widetilde}{{\bm{x}}}}_{1}\bar{{\widetilde}{{\bm{x}}}}_{1 } ' ] ) \\ & = : { \widehat}{d}_{1 } + { \widehat}{d}_{2 } - { \widehat}{d}_{3}.\end{aligned}\ ] ] fix any @xmath431 .
we wish to show that @xmath432 , @xmath433 and @xmath434 .
we make use of theorems [ thma1 ] and [ thma2 ] .
put @xmath435 $ ] . then @xmath436 and @xmath437 = n^{-1}t^{-2 } { \mathbb{e } } [ ( \sum_{t=1}^{t } \xi_{1 t } ) ^{2 } ] = n^{-1 } t^{-2 } { \mathbb{e}}[{\mathbb{e}}[(\sum_{t=1}^{t } \xi_{1 t } ) ^{2 } \mid c_{1}]].\ ] ] using theorem [ thma1 ] to bound @xmath438 $ ] , we have @xmath439 , which implies that @xmath432 .
the fact that @xmath433 is deduced from a direct evaluation of the variance .
we next show that @xmath440 . by the cross section independence and the cauchy - schwarz inequality
, we have @xmath441 = \frac{1}{n^{2 } } \sum_{i=1}^{n } \operatorname{var}(\bar{{\widetilde}{x}}_{i}^{a } \bar{{\widetilde}{x}}_{i}^{b } ) = n^{-1 } \operatorname{var}(\bar{{\widetilde}{x}}_{1}^{a } \bar{{\widetilde}{x}}_{1}^{b } ) \leq n^{-1 } ( { \mathbb{e } } [ { \mathbb{e } } [ ( \bar{{\widetilde}{x}}_{1}^{a})^{4 } \mid c_{1}]])^{1/2 } ( { \mathbb{e } } [ { \mathbb{e } } [ ( \bar{{\widetilde}{x}}_{1}^{b})^{4 } \mid c_{1 } ] ] ) ^{1/2}.\ ] ] using theorem [ thma2 ] to bound @xmath442 $ ] and @xmath443 $ ] , we have @xmath444 = o(n^{-1}t^{-2})$ ] , which implies that @xmath445 .
therefore , we have @xmath446 . by condition ( a3 )
, there exists a constant @xmath447 such that @xmath448 . by lemma [ lem1 ] , @xmath449 , which implies that @xmath142 is nonsingular with probability approaching one .
we wish to obtain the expansion of @xmath142 . by condition ( a3 ) and
lemma [ lem1 ] , @xmath450 is nonsingular for large @xmath0 , and @xmath451 . put @xmath452 so that @xmath453 . since @xmath454
, applying the taylor expansion to the term @xmath455 , we have @xmath456 since @xmath457 , applying the taylor expansion to the term @xmath458 , we have @xmath459 where the right side is absolutely convergent in @xmath427 for large @xmath0 . * part ( iii ) * : fix any @xmath279 .
it suffices to show that @xmath460 . by the cross section independence and the cauchy - schwarz inequality ,
we have @xmath461 = n^{-1 } { \mathbb{e } } [ ( \bar{{\widetilde}{x}}_{1}^{a } \bar{\epsilon}_{1})^{2 } ] \leq n^{-1 } ( { \mathbb{e } } [ { \mathbb{e } } [ ( \bar{{\widetilde}{x}}_{1}^{a})^{4 } \mid c_{1}]])^{1/2 } ( { \mathbb{e } } [ { \mathbb{e } } [ ( \bar{\epsilon}_{1})^{4 } \mid c_{1 } ] ] ) ^{1/2}.\ ] ] using theorem [ thma2 ] to bound @xmath442 $ ] and @xmath462 $ ] , we have @xmath463 . *
part ( iv ) * : the fact that @xmath464 is absolutely convergent is deduced from theorem [ thma1 ] .
we wish to show the asymptotic normality . by the cramr - wald device
, it suffices to show that for any fixed @xmath182 , @xmath465 ) \stackrel{d}{\to } n(0,\bm{r}'v_{1}\bm{r})$ ] .
define @xmath466)$ ] .
then , @xmath467 ) = \frac{1}{\sqrt{n } } \sum_{i=1}^{n } u_{ni}.\ ] ] since @xmath468 are i.i.d .
, we can apply the lyapunov central limit theorem to the right sum .
since @xmath469 , it suffices to show that @xmath470 = o(n^{1/2})$ ] . here , using theorem [ thma2 ] to bound @xmath471 $ ] , we have @xmath470 = { \mathbb{e } } [ { \mathbb{e } } [ | u_{n1 } |^{3 } \mid c_{1 } ] ] = o(1 ) = o(n^{1/2})$ ]
. therefore , we obtain the desired result .
the expansion ( [ bahadur ] ) follows from lemma [ lem2 ] ( note that we use the fact that @xmath472 ) .
suppose that @xmath171 = \bm{0}$ ] a.s .
then @xmath473 $ ] , so that by lemma [ lem2 ] ( iv ) , @xmath474 .
suppose now that @xmath171 \neq \bm{0}$ ] with positive probability .
then we have @xmath475 + \sqrt{n } ( { \widehat}{s}_{1}-{\mathbb{e}}[{\widehat}{s } \mid \ { c_{i } \}_{i=1}^{n } ] ) \\ & = \sqrt{n } { \mathbb{e}}[{\widehat}{s}_{1 } \mid \ { c_{i } \}_{i=1}^{n } ] + o_{{\mathbb{p}}}(t^{-1/2 } ) \\ & \stackrel{d}{\to } n(\bm{0},v_{2}),\end{aligned}\ ] ] where @xmath476^{\otimes 2}]$ ] .
the asymptotic normality follows from the fact that @xmath477 if @xmath171 = \bm{0}$ ] a.s . and @xmath478 otherwise .
by lemma [ lem1 ] , we have @xmath479 + o(t^{-1})$ ] .
observe that @xmath480 \| & = \sum_{|k| \geq t } |k|^{-1 } \cdot |k|\| { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1 } \epsilon_{1,1+k } ] \| \\ & \leq \frac{1}{t } \sum_{|k| \geq t } |k| \| { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1 } \epsilon_{1,1+k } ] \| \\ & \leq \frac{1}{t } \sum_{k=-\infty}^{\infty } |k| \| { \mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1 } \epsilon_{1,1+k } ] \| = o(t^{-1}).\end{aligned}\ ] ] therefore , we have @xmath481 , which implies the desired result . since @xmath482
, we have @xmath483 where @xmath484 by definition
, we have @xmath485 by theorems [ thma1 ] and [ thma2 ] , we can show that @xmath486 \leq c nd_{nt}^{-1}$ ] , @xmath487 \leq ( { \mathbb{e } } [ \| \bar{{\widetilde}{{\bm{x}}}}_{i } \|^{4}])^{1/2 } ( { \mathbb{e } } [ \bar{\epsilon}_{i}^{4}])^{1/2 } \leq c t^{-2}$ ] and @xmath488 \leq c$ ] for some constant @xmath489 .
therefore , we have @xmath490.\ ] ] in what follows , we wish to show that @xmath491 . fix any @xmath431 . by definition , we have @xmath492 ) ( { \widetilde}{x}_{it}^{b } \epsilon_{it } - { \mathbb{e } } [ { \widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad + \frac{1}{n^{2 } } \sum_{i=1}^{n } { \mathbb{e}}[{\widetilde}{x}_{i1}^{a } \epsilon_{i1 } \mid c_{i } ] \left \ { \frac{1}{t } \sum_{t=1}^{t } ( { \widetilde}{x}_{it}^{b } \epsilon_{it } -{\mathbb{e}}[{\widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad + \frac{1}{n^{2 } } \sum_{i=1}^{n } { \mathbb{e}}[{\widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] \left
\ { \frac{1}{t } \sum_{t=1}^{t } ( { \widetilde}{x}_{it}^{a } \epsilon_{it}-{\mathbb{e}}[{\widetilde}{x}_{i1}^{a } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad - \frac{1}{n^{2 } } \sum_{i=1}^{n } { \mathbb{e}}[{\widetilde}{x}_{i1}^{a}\epsilon_{i1 } \mid c_{i } ] { \mathbb{e}}[{\widetilde}{x}_{i1}^{b}\epsilon_{i1 } \mid c_{i } ] \\ & = : ( i ) + ( ii ) + ( iii ) - ( iv ) . \end{aligned}\ ] ] it suffices to show that the variance of each term in ( i)-(iv ) is @xmath493 . letting @xmath494 $ ] and @xmath495 $ ] , we have @xmath496 \\ & = \frac{1}{n^{3}t^{2 } } \left ( \frac{1}{t^{2 } } \sum_{s=1}^{t } \sum_{t=1}^{t } \sum_{u=1}^{t }
\sum_{v=1}^{t } { \mathbb{e}}[\xi_{1s}\eta_{1t}\xi_{1u}\eta_{1v } ] \right ) .\end{aligned}\ ] ] using the same argument as in the proof of ( * ? ? ?
* theorem 1 ) , we can show that the parenthesis on the right side is @xmath497.| \leq c t^{2}$ ] for some constant @xmath415 .
when @xmath498 , so that @xmath499 , the assertion directly follows from the proof of ( * ? ?
* theorem 1 ) with @xmath500 .
the proof for the @xmath501 case is almost the same as that for the @xmath498 case . ]
therefore , we have @xmath502 . if @xmath171
= \bm{0}$ ] a.s .
, ( ii)-(iv ) are zeros a.s .
, so that @xmath503 .
suppose now that @xmath171 \neq \bm{0}$ ] with positive probability . clearly , @xmath504 .
likewise , we have @xmath505^{2 } { \mathbb{e}}\left [ \left ( \frac{1}{\sqrt{t } } \sum_{t=1}^{t } \eta_{1 t } \right ) ^{2 } \mid c_{1 } \right ] \right ] .\ ] ] by theorem [ thma1 ] , we have @xmath506 \leq c$ ] a.s .
for some constant @xmath489 , so that @xmath507 .
similarly , we have @xmath508 .
therefore , we have @xmath509 . by lemma [ lemb1 ]
, it suffices to evaluate the remainder term in the expansion ( [ bahadurb ] ) unconditionally since it is translated to the evaluation under the conditional probability without changing rates .
[ lem3 ] suppose that conditions ( a1)-(a3 ) are satisfied .
as @xmath13 ( and automatically @xmath12 ) , we have : ( i ) @xmath302 is nonsingular with probability approaching one , and @xmath515 ; ( ii ) @xmath516 = t^{-1}b_{t } = o(t^{-1})$ ] ; ( iii ) @xmath517 = o_{{\mathbb{p } } } ( n^{-1/2 } t^{-1 } ) $ ] . in the proof of lemma [ lem3 ] ,
we use some elementary properties of multinomial distributions . recall that @xmath256 is multinomially distributed with parameters @xmath0 and ( probabilities ) @xmath258 .
then @xmath518 = 1 $ ] and , for any fixed @xmath519 , @xmath520 = \sum_{i=1}^{n } a_{i}^{2 } - n(n^{-1}\sum_{i=1}^{n } a_{i})^{2}$ ] ( which can be directly deduced from the fact that @xmath260 are nonparametric bootstrap weights ) . * part ( i ) * : by lemma [ lem2 ] ( i ) , it suffices to show that @xmath521 .
decompose @xmath302 as @xmath522 similarly , decompose @xmath142 as @xmath523 we wish to show that @xmath524 and @xmath525 , which implies the desired result .
fix any @xmath431 .
observe that @xmath526 & \leq \frac{1}{n^{2}t^{2 } } \sum_{i=1}^{n } \left ( \sum_{t=1}^{t } { \widetilde}{x}_{it}^{a } { \widetilde}{x}_{it}^{b } \right ) ^{2 } \\ & \leq \frac{2}{n^{2}t^{2 } } \sum_{i=1}^{n } \left ( \sum_{t=1}^{t } { \widetilde}{x}_{it}^{a } { \widetilde}{x}_{it}^{b } - { \mathbb{e}}[{\widetilde}{x}_{i1}^{a } { \widetilde}{x}_{i1}^{b } \mid c_{i } ] \right ) ^{2 } + \frac{2}{n^{2 } } \sum_{i=1}^{n } { \mathbb{e } } [ { \widetilde}{x}_{i1}^{a } { \widetilde}{x}_{i1}^{b } \mid c_{i}]^{2}.\end{aligned}\ ] ] by the proof of lemma [ lem1 ] ( i ) , the expectation of the first term is @xmath170 , while the expectation of the second term is @xmath527 .
therefore , we have @xmath528 .
it remains to show that @xmath529 .
observe that @xmath530 \leq \frac{1}{n^{2 } } \sum_{i=1}^{n } ( \bar{{\widetilde}{x}}^{a}_{i } \bar{{\widetilde}{x}}^{b}_{i})^{2}.\ ] ] by the proof of lemma [ lem2 ] ( i ) , the expectation of the right side is @xmath531 , which implies the desired result . *
part ( iii ) * : since @xmath533 = { \mathbb{e } } [ { \widehat}{s}_{2 } ] $ ] , we have @xmath534 & = { \widehat}{s}_{2}^ { * } -{\widehat}{s}_{2 } + { \widehat}{s}_{2 } - { \mathbb{e } } [ { \widehat}{s}_{2 } ] \\ & = \frac{1}{n } \sum_{i=1}^{n } ( w_{ni}-1 ) \bar{{\widetilde}{{\bm{x}}}}_{i } \bar{\epsilon}_{i } + ( { \widehat}{s}_{2 } - { \mathbb{e } } [ { \widehat}{s}_{2 } ] ) .\end{aligned}\ ] ] by lemma [ lem2 ] ( iii ) , the second term is @xmath535 . it remains to show that the first term is @xmath535 .
fix any @xmath279 . observe that @xmath536 \leq \frac{1}{n^{2}}\sum_{i=1}^{n } ( \bar{{\widetilde}{x}}^{a}_{i } \bar{\epsilon}_{i})^{2}.\ ] ] by the proof of lemma [ lem2 ] ( iii ) , the expectation of the right side is @xmath537 , which implies the desired result .
[ lem4 ] under conditions ( a1)-(a3 ) , we have : @xmath538 ) \leq { \bm{x}}\ } - { \mathbb{p}}\ { n(\bm{0 } , v_{1 } ) \leq { \bm{x}}\ } | \stackrel{{\mathbb{p}}}{\to } 0,\ ] ] where @xmath426)({\widetilde}{{\bm{x}}}_{1,1+k}\epsilon_{1,1+k}-{\mathbb{e}}[{\widetilde}{{\bm{x}}}_{1,1+k}\epsilon_{1,1+k } \mid c_{1}])']$ ] , provided that @xmath464 is nonsingular .
* step 2 * : ( second assertion of ( [ event ] ) ) fix any @xmath431 .
we wish to show that @xmath544 .
observe first that @xmath545 \to v_{1}^{ab}$ ] .
thus , it suffices to show that @xmath546 as @xmath13 . by the cross section independence , @xmath547 .
\label{moment}\ ] ] as in the proof of proposition [ prop2 ] , we can show that @xmath548 = o(1)$ ] , so that the right side on ( [ moment ] ) is @xmath198 , which implies the desired result .
* step 3 * : ( third assertion of ( [ event ] ) ) from the proof of lemma [ lem2 ] part ( @xmath549 ) above , one can show that @xmath550 = o(1 ) = o(n^{1/2})$ ] for any @xmath279 , which in turn implies the desired result .
we are now in position to prove the lemma .
define @xmath551 in such a way that @xmath552 if @xmath553 for some @xmath554 for @xmath41 .
letting @xmath555 , we have @xmath556 ) = n^{-1/2 } \sum_{i=1}^{n}{\widetilde}{\bm{u}}_{ni}^{*}$ ] .
observe that @xmath557 are i.i.d .
with mean zero and covariance matrix @xmath558 conditional on @xmath257 .
we use the following fact : let @xmath559 be a sequence of random variables and let @xmath560 be a constant ; if for any subsequence @xmath561 of @xmath562 there exists a further subsequence @xmath563 such that @xmath564 almost surely , then @xmath565 . recall that @xmath273 is indexed by @xmath0 . take any subsequence @xmath561 of @xmath562
. then there exists a further subsequence @xmath563 such that along with the subsequence @xmath563 , ( [ event ] ) holds almost surely .
this means that for almost every realization of @xmath566 , along with the subsequence @xmath567 , @xmath568 and the lyapunov condition is satisfied for @xmath569 .
therefore , along with the subsequence @xmath563 , conditional on @xmath257 , @xmath556 ) \stackrel{d}{\to } n(\bm{0},v_{1})$ ] for almost every realization of @xmath566 . by the above fact
, we obtain the desired result .
we wish to verify that @xmath570 .
observe that @xmath571 ) \notag \\ & \quad + \frac{1}{n } \sum_{i=1}^{n } ( w_{ni}-1 ) { \mathbb{e } } [ { \widetilde}{{\bm{x}}}_{i1 } \epsilon_{i1 } \mid c_{i } ] .
\label{expansion}\end{aligned}\ ] ] here the first term is @xmath572 and the second term is zero if @xmath573 = \bm{0}$ ] and is @xmath574 otherwise . hence we conclude that @xmath570 . for ( [ bootd ] )
, suppose first that @xmath171 = \bm{0}$ ] .
then the second term in ( [ expansion ] ) vanishes , and the assertion ( [ bootd ] ) follows from lemma [ lem4 ] .
suppose that @xmath171 \neq \bm{0}$ ] with positive probability .
then the second term dominates the first term in ( [ expansion ] ) , it is routine to verify that @xmath576 \leq { \bm{x}}\ } - { \mathbb{p}}\ { n(\bm{0 } , v_{2 } ) \leq { \bm{x}}\ } | \stackrel{{\mathbb{p}}}{\to } 0,\ ] ] where @xmath577^{\otimes 2 } ] $ ] .
this completes the proof .
the proof is similar to that of proposition [ prop2 ] . since @xmath578 , we have @xmath579 where @xmath580 by the proof of proposition [ prop2 ] , together with the assumption that @xmath581 $ ] , and lemma [ lemb1 ] , we have @xmath582,\ ] ] in probability .
let @xmath583 by the proof of proposition [ prop2 ] , we have @xmath584 , and hence we only need to show that @xmath585 in probability .
fix any @xmath431 , and observe that @xmath586 ) ( { \widetilde}{x}_{it}^{b } \epsilon_{it } - { \mathbb{e } } [ { \widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad + \frac{1}{n^{2 } } \sum_{i=1}^{n } ( w_{ni}-1){\mathbb{e}}[{\widetilde}{x}_{i1}^{a } \epsilon_{i1 } \mid c_{i } ] \left \ { \frac{1}{t } \sum_{t=1}^{t } ( { \widetilde}{x}_{it}^{b } \epsilon_{it } -{\mathbb{e}}[{\widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad + \frac{1}{n^{2 } } \sum_{i=1}^{n } ( w_{ni}-1){\mathbb{e}}[{\widetilde}{x}_{i1}^{b } \epsilon_{i1 } \mid c_{i } ] \left \ { \frac{1}{t } \sum_{t=1}^{t } ( { \widetilde}{x}_{it}^{a } \epsilon_{it}-{\mathbb{e}}[{\widetilde}{x}_{i1}^{a } \epsilon_{i1 } \mid c_{i } ] ) \right \ } \\ & \quad - \frac{1}{n^{2 } } \sum_{i=1}^{n } ( w_{ni}-1){\mathbb{e}}[{\widetilde}{x}_{i1}^{a}\epsilon_{i1 } \mid c_{i } ] { \mathbb{e}}[{\widetilde}{x}_{i1}^{b}\epsilon_{i1 } \mid c_{i } ] \\ & = : ( i ) + ( ii ) + ( iii ) - ( iv ) . \end{aligned}\ ] ] by the proof of proposition [ prop2 ] , we can deduce that @xmath587 = { \mathbb{e } } [ { \mathbb{e}}_{w } [ ( i)^2 ] ] = o(n^{-3}t^{-2})$ ] , and when @xmath171 \neq \bm{0}$ ] with positive probability , @xmath588 = o(n^{-3}t^{-1 } ) , { \mathbb{e } } [ ( iii)^{2 } ] = o(n^{-3}t^{-1})$ ] and @xmath589 = o(n^{-3})$ ] .
therefore , we obtain the desired assertion .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.1829 & -0.0482 & -0.0898 & -0.0023 & 0.0005 & 0.6610 & 0.7580 & 0.7870 & 0.7420 & 0.9145 & 0.8995 + & & ( 0.0366 ) & ( 0.0397 ) & ( 0.0717 ) & ( 0.0605 ) & & & & & & & + & & [ 1.0337 ] & [ 1.2365 ] & [ 1.0423 ] & [ 1.5825 ] & & & & & & & + & 16 & -0.1351 & -0.0311 & -0.0756 & 0.0073 & 0.0025 & 0.7550 & 0.6910 & 0.7845 & 0.7430 & 0.9255 & 0.8970 + & & ( 0.0299 ) & ( 0.0318 ) & ( 0.0497 ) & ( 0.0470 ) & & & & & & & + & & [ 1.0455 ] & [ 1.2048 ] & [ 1.0134 ] & [ 1.5672 ] & & & & & & & + & 20 & -0.1064 & -0.0218 & -0.0649 & 0.0087 & 0.0070 & 0.8115 & 0.6350 & 0.7910 & 0.7515 & 0.9305 & 0.8980 + & & ( 0.0257 ) & ( 0.0270 ) & ( 0.0390 ) & ( 0.0387 ) & & & & & & & + & & [ 1.0458 ] & [ 1.1849 ] & [ 1.0103 ] & [ 1.5475 ] & & & & & & & + & 24 & -0.0870 & -0.0156 & -0.0581 & 0.0091 & 0.0230 & 0.8485 & 0.5815 & 0.7840 & 0.7575 & 0.9210 & 0.8965 + & & ( 0.0226 ) & ( 0.0236 ) & ( 0.0330 ) & ( 0.0329 ) & & & & & & & + & & [ 1.0453 ] & [ 1.1663 ] & [ 1.0355 ] & [ 1.5169 ] & & & & & & & + 100 & 12 & -0.1814 & -0.0465 & -0.0471 & -0.0008 & 0.0000 & 0.4685 & 0.8500 & 0.7865 & 0.7490 & 0.9310 & 0.9250 + & & ( 0.0256 ) & ( 0.0278 ) & ( 0.0521 ) & ( 0.0423 ) & & & & & & & + & & [ 1.0095 ] & [ 1.2262 ] & [ 1.0330 ] & [ 1.5654 ] & & & & & & & + & 16 & -0.1342 & -0.0301 & -0.0417 & 0.0069 & 0.0000 & 0.6340 & 0.8255 & 0.7575 & 0.7385 & 0.9375 & 0.9105 + & & ( 0.0214 ) & ( 0.0228 ) & ( 0.0357 ) & ( 0.0343 ) & & & & & & & + & & [ 1.0412 ] & [ 1.2222 ] & [ 0.9890 ] & [ 1.6165 ] & & & & & & & + & 20 & -0.1046 & -0.0199 & -0.0361 & 0.0098 & 0.0000 & 0.7430 & 0.7730 & 0.7640 & 0.7440 & 0.9265 & 0.9075 + & & ( 0.0180 ) & ( 0.0189 ) & ( 0.0289 ) & ( 0.0277 ) & & & & & & & + & & [ 1.0192 ] & [ 1.1760 ] & [ 1.0191 ] & [ 1.5633 ] & & & & & & & + & 24 & -0.0863 & -0.0148 & -0.0335 & 0.0094 & 0.0000 & 0.7870 & 0.7240 & 0.7705 & 0.7525 & 0.9190 & 0.9050 + & & ( 0.0157 ) & ( 0.0163 ) & ( 0.0233 ) & ( 0.0232 ) & & & & & & & + & & [ 1.0145 ] & [ 1.1422 ] & [ 0.9861 ] & [ 1.5131 ] & & & & & & & + 200 & 12 & -0.1815 & -0.0466 & -0.0258 & -0.0007 & 0.0000 & 0.2195 & 0.9000 & 0.7875 & 0.7515 & 0.9340 & 0.9330 + & & ( 0.0185 ) & ( 0.0201 ) & ( 0.0359 ) & ( 0.0303 ) & & & & & & & + & & [ 1.0203 ] & [ 1.2507 ] & [ 0.9874 ] & [ 1.5811 ] & & & & & & & + & 16 & -0.1329 & -0.0287 & -0.0219 & 0.0075 & 0.0000 & 0.4295 & 0.8880 & 0.7775 & 0.7555 & 0.9295 & 0.9130 + & & ( 0.0143 ) & ( 0.0152 ) & ( 0.0254 ) & ( 0.0232 ) & & & & & & & + & & [ 0.9776 ] & [ 1.1576 ] & [ 0.9765 ] & [ 1.5514 ] & & & & & & & + & 20 & -0.1045 & -0.0198 & -0.0189 & 0.0095 & 0.0000 & 0.5930 & 0.8510 & 0.7440 & 0.7300 & 0.9085 & 0.8995 + & & ( 0.0128 ) & ( 0.0134 ) & ( 0.0209 ) & ( 0.0194 ) & & & & & & & + & & [ 1.0191 ] & [ 1.1802 ] & [ 1.0226 ] & [ 1.5466 ] & & & & & & & + & 24 & -0.0858 & -0.0144 & -0.0177 & 0.0088 & 0.0000 & 0.6890 & 0.8250 & 0.7560 & 0.7445 & 0.9150 & 0.9025 + & & ( 0.0108 ) & ( 0.0113 ) & ( 0.0174 ) & ( 0.0160 ) & & & & & & & + & & [ 0.9808 ] & [ 1.1181 ] & [ 1.0198 ] & [ 1.4731 ] & & & & & & & + notes : monte carlo experiments based on 2,000 repetitions .
the standard deviations are inside parenthesis . inside brackets
are the ratio of the averages of the standard errors to the simulation of the standard deviations .
the dgp is @xmath590 with true parameter @xmath312 .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.3534 & -0.2437 & -0.5201 & -0.0943 & 0.0000 & 0.0025 & 0.0000 & 0.6000 & 0.5615 & 0.7230 & 0.7960 + & & ( 0.0587 ) & ( 0.0636 ) & ( 0.1213 ) & ( 0.0862 ) & & & & & & & + & & [ 1.0592 ] & [ 1.6464 ] & [ 1.4145 ] & [ 1.4177 ] & & & & & & & + & 16 & -0.2719 & -0.1845 & -0.4207 & -0.0500 & 0.0005 & 0.0070 & 0.0000 & 0.7440 & 0.7180 & 0.8525 & 0.8770 + & & ( 0.0492 ) & ( 0.0523 ) & ( 0.0977 ) & ( 0.0704 ) & & & & & & & + & & [ 1.0309 ] & [ 1.6146 ] & [ 1.4062 ] & [ 1.3855 ] & & & & & & & + & 20 & -0.2205 & -0.1480 & -0.3500 & -0.0285 & 0.0000 & 0.0150 & 0.0000 & 0.7810 & 0.7660 & 0.8800 & 0.8910 + & & ( 0.0444 ) & ( 0.0467 ) & ( 0.0811 ) & ( 0.0633 ) & & & & & & & + & & [ 1.0570 ] & [ 1.6543 ] & [ 1.4037 ] & [ 1.4390 ] & & & & & & & + & 24 & -0.1852 & -0.1234 & -0.2984 & -0.0162 & 0.0005 & 0.0235 & 0.0000 & 0.8245 & 0.8145 & 0.9175 & 0.9175 + & & ( 0.0377 ) & ( 0.0393 ) & ( 0.0663 ) & ( 0.0536 ) & & & & & & & + & & [ 0.9921 ] & [ 1.5584 ] & [ 1.3460 ] & [ 1.3665 ] & & & & & & & + 100 & 12 & -0.3531 & -0.2433 & -0.5241 & -0.0951 & 0.0000 & 0.0000 & 0.0000 & 0.4230 & 0.3945 & 0.5985 & 0.6535 + & & ( 0.0408 ) & ( 0.0442 ) & ( 0.0922 ) & ( 0.0598 ) & & & & & & & + & & [ 1.0240 ] & [ 1.6167 ] & [ 1.4195 ] & [ 1.3820 ] & & & & & & & + & 16 & -0.2716 & -0.1842 & -0.4260 & -0.0500 & 0.0000 & 0.0000 & 0.0000 & 0.6545 & 0.6350 & 0.7920 & 0.8200 + & & ( 0.0343 ) & ( 0.0364 ) & ( 0.0743 ) & ( 0.0497 ) & & & & & & & + & & [ 0.9987 ] & [ 1.5902 ] & [ 1.4024 ] & [ 1.3721 ] & & & & & & & + & 20 & -0.2191 & -0.1466 & -0.3556 & -0.0278 & 0.0000 & 0.0000 & 0.0000 & 0.7340 & 0.7220 & 0.8645 & 0.8770 + & & ( 0.0316 ) & ( 0.0332 ) & ( 0.0647 ) & ( 0.0457 ) & & & & & & & + & & [ 1.0415 ] & [ 1.6647 ] & [ 1.4574 ] & [ 1.4550 ] & & & & & & & + & 24 & -0.1839 & -0.1220 & -0.3067 & -0.0152 & 0.0000 & 0.0010 & 0.0000 & 0.7995 & 0.7970 & 0.9060 & 0.9125 + & & ( 0.0280 ) & ( 0.0292 ) & ( 0.0554 ) & ( 0.0391 ) & & & & & & & + & & [ 1.0268 ] & [ 1.6383 ] & [ 1.4494 ] & [ 1.4042 ] & & & & & & & + 200 & 12 & -0.3518 & -0.2420 & -0.5212 & -0.0929 & 0.0000 & 0.0000 & 0.0000 & 0.2200 & 0.1990 & 0.3795 & 0.4220 + & & ( 0.0282 ) & ( 0.0305 ) & ( 0.0652 ) & ( 0.0414 ) & & & & & & & + & & [ 0.9874 ] & [ 1.5791 ] & [ 1.3727 ] & [ 1.3457 ] & & & & & & & + & 16 & -0.2714 & -0.1840 & -0.4277 & -0.0494 & 0.0000 & 0.0000 & 0.0000 & 0.5080 & 0.4880 & 0.6705 & 0.7050 + & & ( 0.0246 ) & ( 0.0262 ) & ( 0.0560 ) & ( 0.0361 ) & & & & & & & + & & [ 1.0091 ] & [ 1.6156 ] & [ 1.4378 ] & [ 1.4059 ] & & & & & & & + & 20 & -0.2187 & -0.1461 & -0.3603 & -0.0268 & 0.0000 & 0.0000 & 0.0000 & 0.6890 & 0.6790 & 0.8340 & 0.8515 + & & ( 0.0219 ) & ( 0.0230 ) & ( 0.0481 ) & ( 0.0308 ) & & & & & & & + & & [ 1.0143 ] & [ 1.6281 ] & [ 1.4588 ] & [ 1.3852 ] & & & & & & & + & 24 & -0.1837 & -0.1218 & -0.3116 & -0.0152 & 0.0000 & 0.0000 & 0.0000 & 0.7710 & 0.7670 & 0.8935 & 0.9020 + & & ( 0.0198 ) & ( 0.0207 ) & ( 0.0406 ) & ( 0.0278 ) & & & & & & & + & & [ 1.0164 ] & [ 1.6395 ] & [ 1.4264 ] & [ 1.4035 ] & & & & & & & + notes : monte carlo experiments based on 2,000 repetitions .
the standard deviations are inside parenthesis . inside brackets
are the ratio of the averages of standard errors to simulation of the standard deviations .
the dgp is @xmath591 with true parameter @xmath318 .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.0201 & 0.0374 & 0.0364 & 0.0065 & 0.8555 & 0.9180 & 0.9775 & 0.9170 & 0.9040 & 0.9245 & 0.9350 + & & ( 0.0268 ) & ( 0.0290 ) & ( 0.0307 ) & ( 0.0290 ) & & & & & & & + & & [ 1.0289 ] & [ 0.7480 ] & [ 0.6187 ] & [ 1.0853 ] & & & & & & & + & 16 & -0.0151 & 0.0283 & 0.0255 & 0.0038 & 0.8830 & 0.9435 & 0.9900 & 0.9230 & 0.9170 & 0.9300 & 0.9435 + & & ( 0.0226 ) & ( 0.0240 ) & ( 0.0250 ) & ( 0.0240 ) & & & & & & & + & & [ 1.0096 ] & [ 0.7137 ] & [ 0.5954 ] & [ 1.0522 ] & & & & & & & + & 20 & -0.0107 & 0.0242 & 0.0199 & 0.0045 & 0.9040 & 0.9515 & 0.9900 & 0.9225 & 0.9105 & 0.9290 & 0.9350 + & & ( 0.0209 ) & ( 0.0219 ) & ( 0.0227 ) & ( 0.0217 ) & & & & & & & + & & [ 1.0543 ] & [ 0.7275 ] & [ 0.6117 ] & [ 1.0772 ] & & & & & & & + & 24 & -0.0081 & 0.0212 & 0.0159 & 0.0046 & 0.9100 & 0.9650 & 0.9950 & 0.9390 & 0.9320 & 0.9415 & 0.9530 + & & ( 0.0182 ) & ( 0.0190 ) & ( 0.0197 ) & ( 0.0187 ) & & & & & & & + & & [ 1.0066 ] & [ 0.6881 ] & [ 0.5893 ] & [ 1.0189 ] & & & & & & & + 100 & 12 & -0.0206 & 0.0369 & 0.0421 & 0.0057 & 0.7925 & 0.8005 & 0.8960 & 0.9145 & 0.9005 & 0.9230 & 0.9330 + & & ( 0.0192 ) & ( 0.0208 ) & ( 0.0221 ) & ( 0.0206 ) & & & & & & & + & & [ 1.0344 ] & [ 0.7598 ] & [ 0.6246 ] & [ 1.0830 ] & & & & & & & + & 16 & -0.0147 & 0.0288 & 0.0328 & 0.0043 & 0.8350 & 0.8530 & 0.9310 & 0.9295 & 0.9165 & 0.9335 & 0.9355 + & & ( 0.0162 ) & ( 0.0172 ) & ( 0.0181 ) & ( 0.0170 ) & & & & & & & + & & [ 1.0184 ] & [ 0.7236 ] & [ 0.6017 ] & [ 1.0508 ] & & & & & & & + & 20 & -0.0110 & 0.0239 & 0.0269 & 0.0042 & 0.8675 & 0.8875 & 0.9475 & 0.9315 & 0.9255 & 0.9325 & 0.9340 + & & ( 0.0144 ) & ( 0.0152 ) & ( 0.0158 ) & ( 0.0149 ) & & & & & & & + & & [ 1.0156 ] & [ 0.7113 ] & [ 0.5954 ] & [ 1.0344 ] & & & & & & & + & 24 & -0.0086 & 0.0206 & 0.0227 & 0.0040 & 0.8915 & 0.9065 & 0.9675 & 0.9320 & 0.9280 & 0.9340 & 0.9385 + & & ( 0.0129 ) & ( 0.0134 ) & ( 0.0139 ) & ( 0.0134 ) & & & & & & & + & & [ 0.9987 ] & [ 0.6898 ] & [ 0.5782 ] & [ 1.0225 ] & & & & & & & + 200 & 12 & -0.0210 & 0.0364 & 0.0446 & 0.0053 & 0.6475 & 0.5475 & 0.6295 & 0.9185 & 0.9105 & 0.9325 & 0.9315 + & & ( 0.0133 ) & ( 0.0144 ) & ( 0.0153 ) & ( 0.0141 ) & & & & & & & + & & [ 1.0078 ] & [ 0.7440 ] & [ 0.6046 ] & [ 1.0382 ] & & & & & & & + & 16 & -0.0144 & 0.0291 & 0.0365 & 0.0046 & 0.7530 & 0.6240 & 0.6705 & 0.9160 & 0.9065 & 0.9215 & 0.9205 + & & ( 0.0113 ) & ( 0.0120 ) & ( 0.0127 ) & ( 0.0119 ) & & & & & & & + & & [ 1.0002 ] & [ 0.7146 ] & [ 0.5927 ] & [ 1.0359 ] & & & & & & & + & 20 & -0.0106 & 0.0243 & 0.0308 & 0.0046 & 0.8190 & 0.6935 & 0.7225 & 0.9205 & 0.9090 & 0.9205 & 0.9230 + & & ( 0.0101 ) & ( 0.0106 ) & ( 0.0109 ) & ( 0.0105 ) & & & & & & & + & & [ 1.0011 ] & [ 0.7024 ] & [ 0.5797 ] & [ 1.0266 ] & & & & & & & + & 24 & -0.0084 & 0.0208 & 0.0267 & 0.0043 & 0.8475 & 0.7465 & 0.7570 & 0.9140 & 0.9090 & 0.9120 & 0.9195 + & & ( 0.0093 ) & ( 0.0097 ) & ( 0.0100 ) & ( 0.0096 ) & & & & & & & + & & [ 1.0103 ] & [ 0.7022 ] & [ 0.5852 ] & [ 1.0275 ] & & & & & & & + notes : monte carlo experiments based on 2,000 repetitions .
the standard deviations are inside parenthesis . inside brackets
are the ratio of the averages of the standard errors to the simulation of the standard deviations .
the dgp is @xmath591 with true parameter @xmath319 .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.3161 & -0.1476 & -0.3222 & -0.0649 & 0.0000 & 0.0765 & 0.1395 & 0.6760 & 0.6480 & 0.8020 & 0.9470 + & & ( 0.0518 ) & ( 0.0557 ) & ( 0.1365 ) & ( 0.0767 ) & & & & & & & + & & [ 1.4428 ] & [ 1.5008 ] & [ 1.3745 ] & [ 1.4312 ] & & & & & & & + & 16 & -0.2415 & -0.1005 & -0.2524 & -0.0329 & 0.0000 & 0.1600 & 0.0910 & 0.7705 & 0.7580 & 0.8775 & 0.9650 + & & ( 0.0422 ) & ( 0.0437 ) & ( 0.0961 ) & ( 0.0617 ) & & & & & & & + & & [ 1.4489 ] & [ 1.4195 ] & [ 1.3400 ] & [ 1.4107 ] & & & & & & & + & 20 & -0.1931 & -0.0703 & -0.2074 & -0.0154 & 0.0000 & 0.3180 & 0.0610 & 0.8210 & 0.8025 & 0.9100 & 0.9695 + & & ( 0.0366 ) & ( 0.0373 ) & ( 0.0748 ) & ( 0.0532 ) & & & & & & & + & & [ 1.4951 ] & [ 1.3995 ] & [ 1.3368 ] & [ 1.4228 ] & & & & & & & + & 24 & -0.1632 & -0.0531 & -0.1791 & -0.0112 & 0.0000 & 0.4380 & 0.0395 & 0.8015 & 0.7965 & 0.9030 & 0.9690 + & & ( 0.0336 ) & ( 0.0341 ) & ( 0.0601 ) & ( 0.0480 ) & & & & & & & + & & [ 1.5805 ] & [ 1.4346 ] & [ 1.3171 ] & [ 1.4633 ] & & & & & & & + 100 & 12 & -0.3138 & -0.1455 & -0.2227 & -0.0620 & 0.0000 & 0.0040 & 0.2885 & 0.5865 & 0.5690 & 0.7205 & 0.9215 + & & ( 0.0366 ) & ( 0.0392 ) & ( 0.1084 ) & ( 0.0548 ) & & & & & & & + & & [ 1.4146 ] & [ 1.4938 ] & [ 1.2980 ] & [ 1.4406 ] & & & & & & & + & 16 & -0.2394 & -0.0984 & -0.1682 & -0.0309 & 0.0000 & 0.0295 & 0.2305 & 0.7245 & 0.7175 & 0.8530 & 0.9600 + & & ( 0.0302 ) & ( 0.0313 ) & ( 0.0776 ) & ( 0.0449 ) & & & & & & & + & & [ 1.4477 ] & [ 1.4367 ] & [ 1.3110 ] & [ 1.4475 ] & & & & & & & + & 20 & -0.1932 & -0.0703 & -0.1388 & -0.0172 & 0.0000 & 0.0970 & 0.1855 & 0.7945 & 0.7835 & 0.9020 & 0.9690 + & & ( 0.0264 ) & ( 0.0269 ) & ( 0.0585 ) & ( 0.0382 ) & & & & & & & + & & [ 1.5053 ] & [ 1.4262 ] & [ 1.2883 ] & [ 1.4448 ] & & & & & & & + & 24 & -0.1607 & -0.0505 & -0.1175 & -0.0078 & 0.0000 & 0.2265 & 0.1635 & 0.8030 & 0.7950 & 0.9215 & 0.9775 + & & ( 0.0236 ) & ( 0.0237 ) & ( 0.0478 ) & ( 0.0340 ) & & & & & & & + & & [ 1.5410 ] & [ 1.4149 ] & [ 1.2828 ] & [ 1.4585 ] & & & & & & & + 200 & 12 & -0.3128 & -0.1447 & -0.1126 & -0.0604 & 0.0000 & 0.0000 & 0.6185 & 0.4285 & 0.4050 & 0.6100 & 0.8805 + & & ( 0.0253 ) & ( 0.0268 ) & ( 0.0828 ) & ( 0.0381 ) & & & & & & & + & & [ 1.3624 ] & [ 1.4475 ] & [ 1.1826 ] & [ 1.4054 ] & & & & & & & + & 16 & -0.2387 & -0.0979 & -0.0749 & -0.0299 & 0.0000 & 0.0000 & 0.6560 & 0.6550 & 0.6505 & 0.8255 & 0.9515 + & & ( 0.0214 ) & ( 0.0220 ) & ( 0.0570 ) & ( 0.0308 ) & & & & & & & + & & [ 1.4342 ] & [ 1.4296 ] & [ 1.1797 ] & [ 1.4002 ] & & & & & & & + & 20 & -0.1923 & -0.0693 & -0.0577 & -0.0159 & 0.0000 & 0.0095 & 0.6360 & 0.7580 & 0.7485 & 0.8855 & 0.9675 + & & ( 0.0188 ) & ( 0.0193 ) & ( 0.0444 ) & ( 0.0271 ) & & & & & & & + & & [ 1.5034 ] & [ 1.4465 ] & [ 1.2020 ] & [ 1.4456 ] & & & & & & & + & 24 & -0.1607 & -0.0505 & -0.0444 & -0.0087 & 0.0000 & 0.0475 & 0.6585 & 0.7950 & 0.7915 & 0.9175 & 0.9765 + & & ( 0.0167 ) & ( 0.0169 ) & ( 0.0362 ) & ( 0.0236 ) & & & & & & & + & & [ 1.5366 ] & [ 1.4254 ] & [ 1.2127 ] & [ 1.4285 ] & & & & & & & + notes : monte carlo based on 2,000 repetitions .
standard deviations are inside parenthesis .
inside brackets are ratios of averages of the standard errors to simulation of the standard deviations .
dgp is @xmath592 with true parameter @xmath318 and @xmath330 .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.2076 & -0.0949 & -0.1267 & -0.0411 & 0.0750 & 0.3740 & 0.4210 & 0.7225 & 0.7250 & 0.8375 & 0.8570 + & & ( 0.0609 ) & ( 0.0660 ) & ( 0.0795 ) & ( 0.0844 ) & & & & & & & + & & [ 1.0806 ] & [ 1.7323 ] & [ 1.3986 ] & [ 1.5547 ] & & & & & & & + & 16 & -0.1591 & -0.0715 & -0.1087 & -0.0241 & 0.1720 & 0.4460 & 0.3815 & 0.7935 & 0.8090 & 0.8780 & 0.8990 + & & ( 0.0541 ) & ( 0.0574 ) & ( 0.0654 ) & ( 0.0733 ) & & & & & & & + & & [ 1.0101 ] & [ 1.7774 ] & [ 1.4647 ] & [ 1.4632 ] & & & & & & & + & 20 & -0.1290 & -0.0575 & -0.0956 & -0.0158 & 0.3155 & 0.4660 & 0.3550 & 0.7945 & 0.8205 & 0.8885 & 0.9090 + & & ( 0.0541 ) & ( 0.0568 ) & ( 0.0622 ) & ( 0.0712 ) & & & & & & & + & & [ 1.0530 ] & [ 1.9933 ] & [ 1.6559 ] & [ 1.4994 ] & & & & & & & + & 24 & -0.1099 & -0.0494 & -0.0888 & -0.0125 & 0.4295 & 0.4930 & 0.3420 & 0.8235 & 0.8365 & 0.8975 & 0.9180 + & & ( 0.0530 ) & ( 0.0552 ) & ( 0.0580 ) & ( 0.0660 ) & & & & & & & + & & [ 1.0572 ] & [ 2.1466 ] & [ 1.7639 ] & [ 1.4307 ] & & & & & & & + 100 & 12 & -0.2060 & -0.0932 & -0.1075 & -0.0379 & 0.0015 & 0.1715 & 0.3070 & 0.7095 & 0.7280 & 0.8430 & 0.8615 + & & ( 0.0421 ) & ( 0.0456 ) & ( 0.0565 ) & ( 0.0590 ) & & & & & & & + & & [ 1.0136 ] & [ 1.6894 ] & [ 1.3783 ] & [ 1.4919 ] & & & & & & & + & 16 & -0.1583 & -0.0707 & -0.0917 & -0.0233 & 0.0320 & 0.2700 & 0.2690 & 0.7730 & 0.8055 & 0.8805 & 0.8985 + & & ( 0.0395 ) & ( 0.0420 ) & ( 0.0493 ) & ( 0.0527 ) & & & & & & & + & & [ 1.0111 ] & [ 1.8367 ] & [ 1.5291 ] & [ 1.4531 ] & & & & & & & + & 20 & -0.1276 & -0.0559 & -0.0789 & -0.0135 & 0.0950 & 0.3310 & 0.2770 & 0.8085 & 0.8370 & 0.9005 & 0.9135 + & & ( 0.0380 ) & ( 0.0399 ) & ( 0.0447 ) & ( 0.0494 ) & & & & & & & + & & [ 1.0051 ] & [ 1.9788 ] & [ 1.6517 ] & [ 1.4253 ] & & & & & & & + & 24 & -0.1079 & -0.0474 & -0.0724 & -0.0093 & 0.1680 & 0.3700 & 0.2615 & 0.8340 & 0.8635 & 0.9195 & 0.9290 + & & ( 0.0366 ) & ( 0.0382 ) & ( 0.0407 ) & ( 0.0470 ) & & & & & & & + & & [ 1.0008 ] & [ 2.0981 ] & [ 1.7193 ] & [ 1.4012 ] & & & & & & & + 200 & 12 & -0.2051 & -0.0922 & -0.0955 & -0.0374 & 0.0000 & 0.0365 & 0.1610 & 0.6450 & 0.6750 & 0.8145 & 0.8370 + & & ( 0.0291 ) & ( 0.0315 ) & ( 0.0394 ) & ( 0.0411 ) & & & & & & & + & & [ 0.9832 ] & [ 1.6532 ] & [ 1.3494 ] & [ 1.4615 ] & & & & & & & + & 16 & -0.1575 & -0.0699 & -0.0817 & -0.0217 & 0.0005 & 0.0995 & 0.1415 & 0.7370 & 0.7710 & 0.8685 & 0.8815 + & & ( 0.0287 ) & ( 0.0305 ) & ( 0.0355 ) & ( 0.0391 ) & & & & & & & + & & [ 1.0241 ] & [ 1.8885 ] & [ 1.5463 ] & [ 1.5043 ] & & & & & & & + & 20 & -0.1258 & -0.0541 & -0.0693 & -0.0114 & 0.0085 & 0.1780 & 0.1640 & 0.7760 & 0.8080 & 0.8935 & 0.9050 + & & ( 0.0283 ) & ( 0.0297 ) & ( 0.0332 ) & ( 0.0371 ) & & & & & & & + & & [ 1.0488 ] & [ 2.0891 ] & [ 1.7197 ] & [ 1.5009 ] & & & & & & & + & 24 & -0.1071 & -0.0465 & -0.0630 & -0.0089 & 0.0160 & 0.2255 & 0.1655 & 0.8175 & 0.8455 & 0.9155 & 0.9310 + & & ( 0.0265 ) & ( 0.0276 ) & ( 0.0303 ) & ( 0.0336 ) & & & & & & & + & & [ 1.0112 ] & [ 2.1444 ] & [ 1.7938 ] & [ 1.4004 ] & & & & & & & + notes : monte carlo experiments based on 2,000 repetitions .
the standard deviations are inside parenthesis . inside brackets
are the ratio of the averages of the standard errors to the simulation of the standard deviations .
the dgp is @xmath593 .
+ @xmath0 & @xmath1 & fe & hk & gmm & hpj & fe - ccm & hk & gmm & hpj - ccm & hpj - feb & hpj - hpjb & hpj - hpjpb + & & & + 50 & 12 & -0.1750 & -0.0538 & -0.1056 & -0.0063 & 0.0540 & 0.6400 & 0.6525 & 0.7695 & 0.7835 & 0.9005 & 0.8955 + & & ( 0.0489 ) & ( 0.0530 ) & ( 0.0765 ) & ( 0.0710 ) & & & & & & & + & & [ 1.0552 ] & [ 1.4596 ] & [ 1.1175 ] & [ 1.6163 ] & & & & & & & + & 16 & -0.1299 & -0.0361 & -0.0888 & 0.0002 & 0.1415 & 0.6750 & 0.5820 & 0.7875 & 0.8020 & 0.9120 & 0.9100 + & & ( 0.0424 ) & ( 0.0451 ) & ( 0.0594 ) & ( 0.0588 ) & & & & & & & + & & [ 1.0391 ] & [ 1.4749 ] & [ 1.1636 ] & [ 1.5574 ] & & & & & & & + & 20 & -0.1050 & -0.0288 & -0.0789 & 0.0016 & 0.2165 & 0.7045 & 0.5190 & 0.8040 & 0.8230 & 0.9200 & 0.9270 + & & ( 0.0384 ) & ( 0.0404 ) & ( 0.0496 ) & ( 0.0513 ) & & & & & & & + & & [ 1.0296 ] & [ 1.5015 ] & [ 1.2033 ] & [ 1.5004 ] & & & & & & & + & 24 & -0.0857 & -0.0213 & -0.0707 & 0.0031 & 0.3425 & 0.7200 & 0.4760 & 0.7940 & 0.8205 & 0.9160 & 0.9190 + & & ( 0.0372 ) & ( 0.0388 ) & ( 0.0444 ) & ( 0.0480 ) & & & & & & & + & & [ 1.0691 ] & [ 1.6044 ] & [ 1.2667 ] & [ 1.5075 ] & & & & & & & + 100 & 12 & -0.1721 & -0.0506 & -0.0746 & -0.0039 & 0.0015 & 0.5060 & 0.6745 & 0.7635 & 0.7785 & 0.9165 & 0.9205 + & & ( 0.0344 ) & ( 0.0373 ) & ( 0.0563 ) & ( 0.0510 ) & & & & & & & + & & [ 1.0268 ] & [ 1.4532 ] & [ 1.1218 ] & [ 1.6227 ] & & & & & & & + & 16 & -0.1297 & -0.0359 & -0.0654 & 0.0002 & 0.0095 & 0.5795 & 0.5835 & 0.7915 & 0.8205 & 0.9265 & 0.9240 + & & ( 0.0302 ) & ( 0.0321 ) & ( 0.0446 ) & ( 0.0420 ) & & & & & & & + & & [ 1.0174 ] & [ 1.4837 ] & [ 1.1899 ] & [ 1.5444 ] & & & & & & & + & 20 & -0.1022 & -0.0258 & -0.0572 & 0.0042 & 0.0390 & 0.6515 & 0.5285 & 0.7970 & 0.8275 & 0.9350 & 0.9330 + & & ( 0.0270 ) & ( 0.0283 ) & ( 0.0368 ) & ( 0.0367 ) & & & & & & & + & & [ 0.9925 ] & [ 1.4936 ] & [ 1.2130 ] & [ 1.4868 ] & & & & & & & + & 24 & -0.0846 & -0.0202 & -0.0506 & 0.0039 & 0.1075 & 0.6570 & 0.5060 & 0.8100 & 0.8370 & 0.9265 & 0.9320 + & & ( 0.0261 ) & ( 0.0272 ) & ( 0.0333 ) & ( 0.0337 ) & & & & & & & + & & [ 1.0329 ] & [ 1.5900 ] & [ 1.2918 ] & [ 1.4699 ] & & & & & & & + 200 & 12 & -0.1721 & -0.0506 & -0.0591 & -0.0052 & 0.0000 & 0.2680 & 0.6205 & 0.7835 & 0.8065 & 0.9270 & 0.9300 + & & ( 0.0241 ) & ( 0.0261 ) & ( 0.0418 ) & ( 0.0352 ) & & & & & & & + & & [ 1.0084 ] & [ 1.4413 ] & [ 1.1601 ] & [ 1.5806 ] & & & & & & & + & 16 & -0.1276 & -0.0337 & -0.0487 & 0.0024 & 0.0000 & 0.4325 & 0.5700 & 0.8000 & 0.8345 & 0.9390 & 0.9395 + & & ( 0.0209 ) & ( 0.0222 ) & ( 0.0307 ) & ( 0.0297 ) & & & & & & & + & & [ 0.9847 ] & [ 1.4549 ] & [ 1.1398 ] & [ 1.5331 ] & & & & & & & + & 20 & -0.1011 & -0.0247 & -0.0436 & 0.0043 & 0.0005 & 0.5260 & 0.4985 & 0.8175 & 0.8500 & 0.9405 & 0.9405 + & & ( 0.0195 ) & ( 0.0205 ) & ( 0.0270 ) & ( 0.0260 ) & & & & & & & + & & [ 1.0041 ] & [ 1.5272 ] & [ 1.2305 ] & [ 1.4792 ] & & & & & & & + & 24 & -0.0843 & -0.0199 & -0.0402 & 0.0042 & 0.0050 & 0.5590 & 0.4365 & 0.8135 & 0.8445 & 0.9390 & 0.9435 + & & ( 0.0186 ) & ( 0.0194 ) & ( 0.0241 ) & ( 0.0238 ) & & & & & & & + & & [ 1.0302 ] & [ 1.6044 ] & [ 1.2930 ] & [ 1.4541 ] & & & & & & & + notes : monte carlo based on 2,000 repetitions .
the standard deviations are inside parenthesis .
inside brackets are ratios of averages of the standard errors to simulation of the standard deviations .
the dgp is @xmath594 , with true parameter @xmath383 and @xmath384 .
+ [ t.app ] + + & fe - ccm & hpj - ccm & hpj - hpjpb & gmm & tsls + @xmath595 & @xmath596&@xmath597&@xmath597&@xmath598&@xmath599 + @xmath325 ci & @xmath600 $ ] & @xmath601 $ ] & @xmath602 $ ] & @xmath603 $ ] & @xmath604 $ ] + @xmath605 ci & @xmath606 $ ] & @xmath607 $ ] & @xmath608 $ ] & @xmath609 $ ] & @xmath610 $ ] + @xmath611 & @xmath612&@xmath613&@xmath613&@xmath614&@xmath615 + @xmath325 ci & @xmath616 $ ] & @xmath617 $ ] & @xmath618 $ ] & @xmath619 $ ] & @xmath620 $ ] + @xmath605 ci & @xmath621 $ ] & @xmath622 $ ] & @xmath623 $ ] & @xmath624 $ ] & @xmath625 $ ] + + & fe - ccm & hpj - ccm & hpj - hpjpb & gmm & tsls + @xmath595 & @xmath626&@xmath627&@xmath627&@xmath628&@xmath629 + @xmath325 ci & @xmath630 $ ] & @xmath631 $ ] & @xmath632 $ ] & @xmath633 $ ] & @xmath634 $ ] + @xmath605 ci & @xmath635 $ ] & @xmath636 $ ] & @xmath637 $ ] & @xmath638 $ ] & @xmath639 $ ] + @xmath611 & @xmath640&@xmath641&@xmath641&@xmath642&@xmath643 + @xmath325 ci & @xmath644 $ ] & @xmath645 $ ] & @xmath646 $ ] & @xmath647 $ ] & @xmath648 $ ] + @xmath605 ci & @xmath649 $ ] & @xmath650 $ ] & @xmath651 $ ] & @xmath652 $ ] & @xmath653 $ ] + + & fe - ccm & hpj - ccm & hpj - hpjpb & gmm & tsls + @xmath595 & @xmath654&@xmath655&@xmath655&@xmath656&@xmath657 + @xmath325 ci & @xmath658 $ ] & @xmath659 $ ] & @xmath660 $ ] & @xmath661 $ ] & @xmath662 $ ] + @xmath605 ci & @xmath663 $ ] & @xmath664 $ ] & @xmath665 $ ] & @xmath666 $ ] & @xmath667 $ ] + @xmath611 & @xmath668&@xmath669&@xmath669&@xmath670&@xmath671 + @xmath325 ci & @xmath672 $ ] & @xmath673 $ ] & @xmath674 $ ] & @xmath675 $ ] & @xmath676 $ ] + @xmath605 ci & @xmath677 $ ] & @xmath678 $ ] & @xmath679 $ ] & @xmath680 $ ] & @xmath681 $ ] + | this paper considers fixed effects ( fe ) estimation for linear panel data models under possible model misspecification when both the number of individuals , @xmath0 , and the number of time periods , @xmath1 , are large .
we first clarify the probability limit of the fe estimator and argue that this probability limit can be regarded as a pseudo - true parameter .
we then establish the asymptotic distributional properties of the fe estimator around the pseudo - true parameter when @xmath0 and @xmath1 jointly go to infinity .
notably , we show that the fe estimator suffers from the incidental parameters bias of which the top order is @xmath2 , and even after the incidental parameters bias is completely removed , the rate of convergence of the fe estimator depends on the degree of model misspecification and is either @xmath3 or @xmath4 .
second , we establish asymptotically valid inference on the ( pseudo - true ) parameter . specifically , we derive the asymptotic properties of the clustered covariance matrix ( ccm ) estimator and the cross section bootstrap , and show that they are robust to model misspecification .
this establishes a rigorous theoretical ground for the use of the ccm estimator and the cross section bootstrap when model misspecification and the incidental parameters bias ( in the coefficient estimate ) are present .
we conduct monte carlo simulations to evaluate the finite sample performance of the estimators and inference methods , together with a simple application to the unemployment dynamics in the u.s . |
it was found recently that photoconductivity can sometimes be greatly enhanced by doping derivatives of poly(phenylene vinylene ) , ppv,@xcite poly(thiophene)@xcite and the @xmath4-conjugated polymer poly(methylphenylsilylene)@xcite with the fullerene molecule c@xmath5 .
this led to construction of photovoltaic cells containing a polymer - fullerene heterojunction.@xcite however these devices had poor efficiency because few of the photogenerated carriers were able to reach the electrodes . recently a great increase in efficiency was obtained by the use of interpenetrating donor ( d ) and acceptor ( a ) networks so that d - a junctions are distributed throughout the device.@xcite in one embodiment d is meh - ppv and a is cn - ppv.@xcite because of the cyano side groups on the ppv chains , the electron affinity of cn - ppv is greater than that of meh - ppv while the ionization potential of meh - ppv is less than that of cn - ppv .
therefore electrons in the excitons photoexcited on meh - ppv chains will tend to move onto cn - ppv chains while holes in the excitons on cn - ppv chains tend to move onto meh - ppv .
recently , rice and gartstein@xcite developed a microscopic theory of the photogeneration process to calculate the rate of photoinduced charge transfer due to exciton dissociation between a polymer chain and an adjacent acceptor molecule .
an essential feature of this theory is that the transfer occurs with energy being conserved , _
i.e. _ , no phonons are involved .
they simplified their theory by neglecting the effects of lattice relaxation , _
i.e. _ , polaron formation .
rice and gartstein suggested that their theory could be generalized to the case of the acceptor being another chain rather than a molecule . in this paper
we carry out such a generalization to study the dissociation of an exciton photoexcited on either the donor or acceptor chain into a free hole in the donor polymer s valence band and an electron in the conduction band of the acceptor polymer .
the theory is applied to an meh - ppv : cn - ppv pair .
we have extended it to investigate the dependence of the dissociation rate on the distance between d and a.
we start our discussion from a model hamiltonian for two coupled conjugated polymer chains of infinite length , @xmath6 .
@xmath7 describes the intrachain and charge - transfer ( ct ) electron - hole ( e - h ) pair excitations of the two polymer chains considered independent . by the term intrachain
e - h excitations we refer to the excitations of one chain without regard to the presence of the other polymer chain , while by ct e - h excitation we mean the excitation of a charge on one polymer chain in the presence of a charge of opposite sign on the adjacent chain .
such excitations have also been called interchain excitons , indirect excitons and polaron pairs .
@xmath7 includes no coupling between the two chains except for the long - range coulomb interaction between electron and hole .
@xmath8 and @xmath9 represent the coupling between the two chains , taken to be weak perturbations that cause the interchain charge transfer .
the former describes single - particle hopping and the latter a process in which a charge fluctuation in one species scatters a charge of the other species from one polymer chain to the other.@xcite it is convenient to assume that each polymer chain has e - h symmetry and measure energies from the center of its gap . because the gap centers of the two polymers need not coincide , we will take our zero of energy at the gap center of the second polymer , that of the first polymer differing from this by the energy @xmath10 .
@xmath7 may then be written @xmath11-\sum_{mn\sigma s}u(n - m)b_{n\sigma}^\dagger b_{n\sigma}a_{ms}^\dagger a_{ms}\nonumber\\ & & \mbox{}+\sum_{n\sigma}[{\bar \alpha}_0 ( c_{n\sigma}^\dagger c_{n\sigma}+d_{n\sigma}^\dagger d_{n\sigma})-{\bar t } ( c_{n+1\sigma}^\dagger c_{n\sigma } + \mbox{h.c.})\nonumber\\ & & \hspace{0.5cm}\mbox{}-{\bar t}(d_{n+1\sigma}^\dagger d_{n\sigma } + \mbox{h.c.})]-\sum_{mn\sigma s}{\bar u}(n - m ) d_{n\sigma}^\dagger d_{n\sigma}c_{ms}^\dagger c_{ms}\nonumber\\ \label{h0 } & & \mbox{}-\sum_{mn\sigma s}v_1(n - m ) d_{n\sigma}^\dagger d_{n\sigma}a_{ms}^\dagger a_{ms } -\sum_{mn\sigma s}v_2(n - m ) b_{n\sigma}^\dagger b_{n\sigma}c_{ms}^\dagger c_{ms}\;,\end{aligned}\ ] ] where @xmath12 ( @xmath13 ) and @xmath14 ( @xmath15 ) create an electron in the conduction band and a hole in the valence band , respectively , located at the @xmath16th monomer with spin @xmath4 in the first ( second ) chain . @xmath17 ( @xmath18 ) denotes the center of each band measured from midgap .
the upper ( lower ) sign before @xmath19 refers to the case where the midgap is above ( below ) the zero of energy . from the first and third terms , one can easily get the single electron and single hole energy spectra : @xmath20 and @xmath21 of the first chain and @xmath22 of the second chain . accordingly in these expressions @xmath23 ( @xmath24 ) is the band width .
@xmath25 ( @xmath26 ) represents the length of a monomer of the first ( second ) polymer chain . from the excitation spectrum
, one can easily get the energy gaps of e - h excitation , _ ie .
_ , the minimum energy required to create an e - h pair in one polymer chain , @xmath27 of the first chain and @xmath28 of the second chain .
it is then straightforward to see that one should assume @xmath29 and @xmath30 in order to have @xmath31 ( @xmath32 ) @xmath33 .
the second and fourth terms in eq .
( [ h0 ] ) represent the intrachain coulomb interaction with @xmath34 for the second chain and the corresponding expression for the first chain with @xmath35 and @xmath36 replaced by @xmath37 and @xmath38 .
the last two terms of eq .
( [ h0 ] ) denote the interchain coulomb interactions . for simplicity , we assume @xmath39 equals @xmath40 , with @xmath41 where @xmath42 and @xmath43 stands for the perpendicular distance between the two chains.@xcite @xmath44 and @xmath45 are dielectric constants along and perpendicular to the polymer chains , respectively . the hamiltonian @xmath46 represents the perturbation leading to the single - particle interchain hopping process with @xmath47 and @xmath48 being electronic hopping integrals . in general ,
@xmath47 and @xmath48 are not necessarily the same .
it is reasonable to assume they fall off exponentially according to@xcite @xmath49 the charge - fluctuation transfer process is described by @xmath50 where the first term corresponds to an electron being scattered to the other chain by a local fluctuation of the hole density in the vicinity of site @xmath51 , and in the second term the roles of electron and hole are reversed .
we further assume that the matrix element @xmath52 , because , like @xmath39 and @xmath40 , it depends on the coulomb attraction between the electron and hole on separate chains . it is noted that @xmath53 is also related to the electron hopping integrals and would vanish if @xmath47 ( @xmath54 .
therefore , we further assume @xmath55 .
however , we stress here that @xmath56 ( @xmath57 ) may also have additional dependence on @xmath58 .
again , @xmath56 is not necessarily equal to @xmath57 . in the following discussion of charge transfer
, we always assume that the exciton is photogenerated on the second chain
. the exciton can dissociate either with transfer of an electron or a hole to the first chain as shown schematically in fig .
it can also be deduced from fig . 1 that only for the midgap of the first chain above that of the second chain is it possible to have hole transfer ( ht ) without electron transfer , while only for the midgap of the first chain below that of the second chain is it possible to have electron transfer ( et ) without hole transfer . additionally , for both ht and et ,
in order that the other carrier not hop to the other chain as well , it is necessary to satisfy the additional condition @xmath59 .
we first discuss in detail the case of hole transfer . with
the intrachain and interchain coulomb interaction included , the intrachain e - h excitation spectrum of the second polymer chain consists of an intrachain exciton with energy @xmath60 and a continuum band of width @xmath61 above the energy gap @xmath62 .
the ht e - h excitation spectrum consists of an interchain ( ht ) exciton with energy @xmath63 and a continuum band of width @xmath64 above the gap @xmath65 .
( see fig . 1 ) .
energy conservation requires that ct due to an intrachain exciton in the second chain giving up a hole to the first chain ( ht ) can only take place when @xmath66 the decay rate of the exciton can be written from the fermi golden rule : @xmath67 in this equation , @xmath68 ( @xmath69 , 2 ) is the matrix element describing the transition from the initial intrachain exciton state @xmath70 to the final ct e - h excitation states .
the intrachain spectrum can be acquired as follows .
we first construct a two - particle wave function of a singlet e - h pair excitation on the second polymer chain as @xmath71 with @xmath72 denoting the ground state .
@xmath73 is assumed to be real and is normalized according to @xmath74 .
from @xmath75 , we obtain for the lowest excitonic bound state with energy @xmath60,@xcite @xmath76 similarly we can construct a two - particle ht wavefunction as @xmath77 @xmath78 is given by the equation @xmath79\phi_{n - m}= -(t+\bar t)(\phi_{n - m+1}+\phi_{n - m-1})-v(n - m)\phi_{n - m}\;,\ ] ] with the normalization condition @xmath80 , where @xmath81 is the number of monomers on the chain .
( here we have assumed @xmath82 to be real ) .
it is noted that in the above we only retain the solutions with zero cm wave vector , indicated by a superscript @xmath83 on @xmath84 and @xmath85 , as is required for momentum conservation .
because we assume that free carriers are generated by the dissociation , we only need to get the scattering states of these equations in order to calculate @xmath86 .
the decay rate of an exciton in the second chain by transferring a hole to the first chain may now be written as @xmath87 with @xmath69 standing for the process of single - particle hopping and @xmath88 for the charge fluctuation process .
we take @xmath89 and @xmath90 .
the dimensionless rate @xmath91 , which differs from that defined by rice and gartstein , is defined by @xmath92 where @xmath93
we apply our theory to an meh - ppv : cn - ppv pair .
we consider first an exciton photoinduced on the cn - ppv chain .
this exciton is dissociated by transferring a hole to meh - ppv . to this end
, we take the second chain as cn - ppv and the first one as meh - ppv .
the values of the various unknown parameters in eq .
( [ eq1 ] ) ( written with unbarred quantities for the first chain ) must be chosen to give the correct gap and exciton binding energy . for meh - ppv
the optical absorption edge is at 2.1 ev.@xcite the single particle energy gap of meh - ppv has been measured as 2.45 ev.@xcite the exciton binding energy is the difference of these two numbers,@xcite thus 0.35 ev .
the intrachain coulomb potential coefficient @xmath38 or @xmath36 is of the order of the coulomb attraction between an electron and a hole separated by one monomer ( @xmath94 ) , which is @xmath95 ev.@xcite @xmath96 ev . choosing @xmath97 ev , @xmath98 ev,@xcite and @xmath99 ev , we obtain from eq .
( [ eq1 ] ) @xmath100 ev , the exciton creation energy @xmath101 ev and its binding energy @xmath102 ev . for cn - ppv the optical absorption edge is at @xmath103 ev.@xcite it is reasonable to assume that the exciton binding energy is similar to that of ppv and meh - ppv . with @xmath104
ev , @xmath105 ev , and @xmath106 ev , we obtain @xmath107 ev , the exciton creation energy @xmath108 ev and @xmath109 ev . to obtain numerical results for ht we further take @xmath110,@xcite @xmath111 , @xmath112 ev , the energy difference between gap centers of meh - ppv and cn - ppv @xmath113 ev,@xcite and @xmath114 @xmath115.@xcite for the numerical calculations , we employ a chain of @xmath116 unit cells for each of the polymers and use periodic boundary conditions .
we then numerically solve eq .
( [ eqb ] ) for the ct band eigenfunctions @xmath117 and their eigenvalues @xmath118 and substitute them and the numerical solution of eq .
( [ eq1 ] ) into eq .
( [ gamma ] ) to get @xmath91 . in evaluating @xmath91 ,
we replace the @xmath119-function in eq .
( [ gamma ] ) by a gaussian with variance @xmath120 ev .
the resulting @xmath91 ( @xmath69 , 2 ) are plotted as solid curves as a function of @xmath58 .
since @xmath58 is proportional to the square of the distance between the chains , the transfer rate falls off very rapidly with increasing @xmath58 .
we estimate the minimum chain separation as @xmath121 , giving the minimum @xmath122 .
an increase in interchain distance by a factor 2 is seen to decrease the transfer rate @xmath123 by almost 4 orders of magnitude , and @xmath124 even more .
rewriting the first inequality in eq .
( [ ht ] ) we obtain the threshold value of @xmath19 , @xmath125 for ht in this case eq .
( [ cc ] ) gives @xmath126 ev .
it is noteworthy that @xmath123 and @xmath124 peak at the threshold value of @xmath19 .
this was also found to be the case for the calculation of charge transfer to a dopant molecule.@xcite the appearance in fig .
2 of finite @xmath127 values below the threshold is due to the replacement of the @xmath119-function in eq .
( [ gamma ] ) by a gaussian .
it is possible to change @xmath19 by changing one or both of the polymers or by application of an electric field perpendicular to the chains . for an electric field of @xmath128
v / cm , for example , if the meh - ppv and cn - ppv chains were separated by 4 , the change in @xmath19 due to the field would be @xmath129 ev . thus @xmath19 for ht in meh - ppv : cn - ppv would be either 0.3 or 0.7 ev depending on the direction of the electric field . as can be seen in figs .
3(a ) and 3(b ) , for @xmath123 this could be a change by a factor 2 .
for @xmath124 the change is smaller because @xmath124 decreases less rapidly with increase in @xmath19 .
thus the dissociation rate may be substantially increased or decreased by high electric fields .
having @xmath123 , we can get an approximate value , or at least an upper limit , for @xmath130 using the average of @xmath131 values for meh - ppv and cn - ppv , the individual values having been determined from the energy shift between exciton and excimer emission.@xcite that average is @xmath132 ev . for a spacing of 6 between meh - ppv and cn - ppv @xmath133 , leading to @xmath134 for @xmath113 ev and @xmath135 ps . for 4 , @xmath136 leading to @xmath137 fs for @xmath113 ev .
thus the dissociation process we are calculating is quite rapid .
consider now the case of an exciton photoinduced in meh - ppv dissociating into an e - h pair by transferring an electron to cn - ppv .
for this calculation we take the second chain to be meh - ppv , thus interchanging @xmath32 , @xmath18 , @xmath138 , @xmath35 and @xmath36 with @xmath31 , @xmath17 , @xmath139 , @xmath37 and @xmath38 respectively .
equations ( [ ht ] ) and ( [ eqb ] ) derived for the ht case are unchanged . numerical solution of the equations , as described earlier , leads to @xmath91 for et shown as the dashed curves in fig . 2 and 3 . the threshold value of @xmath19 for et determined from eq .
( [ cc ] ) is 0.4 ev rather than 0.2 ev because of the interchange of @xmath31 and @xmath32 ; the exciton is generated on the chain with the smaller gap in this case .
apart from the difference in threshold , @xmath91 for et varies with @xmath19 and @xmath58 in similar fashion to the case for ht .
the decay rate @xmath140 is obtained from eq .
( [ decay ] ) with @xmath141 , @xmath142 . since @xmath143 and @xmath123 is about twice as large for et when @xmath113 ev , the dissociation rate @xmath144 is about twice as large for et as for ht and the @xmath130 values 1/2 as large . in summary
we have calculated the dissociation rate of an exciton on a donor or acceptor chain into a free electron on a and a free hole on d by a process that conserves energy , _
i.e. _ , no phonons are involved .
also , the process is applicable to chains that are sufficiently long that broadening of their molecular orbital levels is at least as large as the spacing of the unperturbed discrete levels.@xcite for dissociation to be possible in the absence of an electric field whichever chain the exciton is generated on , the midgap of d must lie above that of a by at least the exciton binding energy plus @xmath145 the absolute value of the difference in the energy gaps . for the difference in midgaps greater than this amount the rate of dissociation decreases rapidly . the difference in midgap energy levels of the pair cn - ppv :
meh - ppv is close to satisfying the condition for maximum transfer rate .
the transfer rate falls off so rapidly as the distance between d and a increases that essentially transfer can take place only between neighbors separated by perhaps 4 to 6 perpendicular to the chain direction . for a cn - ppv : meh - ppv pair separated by 4 we calculate a hole transfer time @xmath3 fs , the electron transfer time being shorter by a factor 2 . as a consequence ,
since it has been demonstrated that there is phase separation of the two polymers on a scale of 10 - 100 nm,@xcite and the exciton hopping time in ppv has been estimated as 0.4 ps,@xcite the dissociation rate will likely be determined by the time it takes the exciton to diffuse to a location where the required donor or acceptor is within @xmath121 .
mailing address .
et al . _ ,
science * 258 * ( 1992 ) 1474 ; b. kraabel _ et al .
_ , chem phys . lett .
* 213 * ( 1993 ) 389 .
s. morita _
et al . _ , solid state commun . * 82 * ( 1992 ) 249 .
kepler and p.a .
cahill , appl .
* 63 * ( 1993 ) 1552 .
et al . _ ,
* 62 * ( 1993 ) 585 .
s. morita _
et al . _ , jpn .
* 32 * ( 1993 ) l873 .
s. morita _
* 256 * ( 1994 ) 839 .
sariciftci and a.j .
heeger , int .
b * 8 * ( 1994 ) 237 .
et al . _ ,
nature * 376 * ( 1995 ) 498 .
g. yu , j. gao , j.c .
hummelen , f. wudl , and a.j .
heeger , j. appl .
phys . * 78 * ( 1995 ) 4510 .
rice and yu.n .
gartstein , phys .
b * 53 * ( 1996 ) 10764 . l.d .
landau and e.m .
lifshitz , _ electrodynamics of continuous media _ ( addison - wesley , reading , ma 1960 ) , p. 62 .
mizes and e.m .
conwell , mol .
* 256 * ( 1994 ) 519 .
g. yu , c. zhang , and a.j .
heeger , appl .
* 64 * ( 1994 ) 1540 .
campbell , t.w .
hagler , d.l .
smith , and j.p .
ferraris , phys .
* 76 * ( 1996 ) 1900 .
conwell , synth . met .
, in press .
samuel , g. rumbles , and c.j .
collison , phys .
b * 52 * ( 1995 ) r11573 .
we choose the value found in solution , rather than film , because the film value is lowered by excimer formation .
j. fink _ et al .
_ , in `` _ electronic properties of conjugated polymers _ '' , ed . h. kuzmany , m. mehring , and s. roth , springer series in solid - state sciences , vol . * 76 * , ( springer , berlin , 1987 ) p. 79 .
m.w . wu and e.m .
conwell , phys .
b * 56 * , rapid commun . , in press .
u. rauscher , h. bssler , d.d.c .
bradley , and m. hennecke , phys .
b * 42 * ( 1990 ) 9830 . | in a pair of coupled donor - acceptor conjugated polymer chains , it is possible for an exciton photoexcited on either polymer to decay into a hole in the donor polymer s valence band and an electron in the conduction band of the acceptor polymer .
we calculate the corresponding exciton decay rate and its dependence on inter - polymer distance . for a pair of derivatives of poly(phenylene vinylene ) , ppv , specifically poly[2-methoxy , 5-(2@xmath0-ethyl - hexyloxy)-1 , 4 ppv ] , meh - ppv , and poly(2,5-hexyloxy @xmath1-phenylene cyanovinylene ) , cn - ppv , at a separation of 6 the characteristic decay time is @xmath2 ps , whereas at 4 it is @xmath3 fs . |
the launch of the chandra observatory ( weisskopf 2000 ) in 1999 july has provided the first opportunity for detailed , spatially - resolved x - ray spectroscopy of clusters of galaxies at moderate to high redshifts . the advanced ccd imaging spectrometer ( acis ) on chandra permits direct , simultaneous measurements of the x - ray gas temperature and density profiles in clusters and , via the hydrostatic assumption , the mass distributions , spanning scales from @xmath6 kpc in cluster cores out to @xmath7 mpc at the detection limits . in this paper , we present the first results from chandra observations of rxj1347.5 - 1145 , the most x - ray luminous galaxy cluster known ( schindler 1995 ) .
this cluster has been the subject of several previous x - ray , optical and sunyaev - zeldovich ( sz ) effect studies ( schindler 1995 , 1997 ; fischer & tyson 1997 , allen 1998 ; komatsu 1999 , 2001 ; pointecouteau 1999 , 2001 ) . here , we present the first measurements of the x - ray temperature structure within the cluster and place tight constraints on the total mass distribution .
we compare our results with those from strong and weak lensing analyses .
a comparison of the chandra and sz results , and a determination of the hubble constant from the combined data sets , is presented elsewhere ( schmidt & allen 2002 , in preparation ) . except where stated otherwise ,
the cosmological parameters @xmath8=50 , @xmath9 and @xmath10 are assumed . at the redshift of rxj1347.5 - 1145 ( @xmath11 ) ,
an angular scale of 1 arcsec corresponds to physical size of 6.81 kpc in this cosmology .
the chandra observations of rxj1347.5 - 1145 were carried out using the acis on 2000 march 05 and 2000 april 29 .
the target was observed in the back - illuminated s3 detector and positioned near the centre of node-1 on ccd 7 .
the light curves for both observations were of high quality with no strong background flares .
the net good exposure times were 8.9 and 10.0 ks , respectively , giving a total good exposure time of 18.9 ks .
the focal plane temperature at the time of both observations was -120c .
we have used the ciao software and the level-2 events files provided by the standard chandra pipeline processing for our analysis . only those x - ray events with grade classifications of 0,2,3,4 and 6
were included in our final cleaned data sets .
the raw @xmath12 kev image of the central @xmath13 arcmin@xmath14 region of the cluster , from the combined 18.9 ks data set , is shown in fig . [
fig : im1](a ) .
the pixel size is @xmath15 arcsec@xmath14 , corresponding to @xmath16 raw detector pixels .
[ fig : im1](b ) shows an adaptively smoothed contour plot of the same data , using the smoothing algorithm of ebeling , white & rangarajan ( 2002 ) .
the images reveal a number of notable features .
firstly , we see a sharp central surface brightness peak at a position 13h47m30.63 -11d45m09.3s ( j2000 . ) , in good agreement with the optical centroid for the dominant cluster galaxy , 13h47m30.5 -11d45m09s ( j2000 . ; schindler 1995 ) .
secondly , we identify a region of enhanced emission @xmath17 arcsec to the southeast of the x - ray peak ( see below ) .
thirdly , on large scales ( @xmath18 arcsec ) we detect an extension of the x - ray isophotes to the south . fig .
[ fig : im2 ] shows the data for the central regions of the cluster on a finer @xmath19 arcsec@xmath14 scale ( corresponding to @xmath20 raw detector pixels ) . from this figure
it appears that the region of enhanced emission to the southeast of the x - ray peak has a roughly circular morphology with a flattened northwest edge . beyond this edge , at a radius @xmath6 arcsec , the x - ray subclump appears to be separated from the main cluster core by a valley of reduced emission .
the spectral analysis in section 4.2 shows that the x - ray subclump has a hotter temperature than the gas at the same radius in other directions from the x - ray peak and probably contains shocked gas resulting from a recent subcluster merger event .
komatsu ( 2001 ) report the detection of a region of enhanced sz effect at a position coincident with the x - ray subclump .
the enhanced sz effect is consistent with the higher temperature and density for this region measured from the chandra data .
no strong x - ray point sources are detected in the chandra s3 field .
( the brightest point source has a flux of @xmath21 in the @xmath22 kev band ; gandhi , in preparation . )
the azimuthally - averaged , @xmath12 kev x - ray surface brightness profile for rxj1347.5 - 1145 for position angles of @xmath23 degrees ( i.e. excluding the southeast quadrant ) and for the disturbed , southeast quadrant are shown in figs .
[ fig : surbri](a , b ) , respectively .
the profiles have been flat - fielded and background subtracted using an on - chip region free from cluster emission .
the bin - size is 2 detector pixels ( 0.984 arcsec ) .
we see that the data for position angles of @xmath23 degrees appear regular and , for @xmath24 kpc , can be approximated ( @xmath25 for 72 degrees of freedom ) by a @xmath26-model ( jones & forman 1984 ) of the form @xmath27^{1/2 - 3\beta}$ ] with a core radius @xmath28 kpc and a slope parameter @xmath29 ( @xmath30 errors ) .
between radii of @xmath31 mpc we find evidence for a steepening of the surface brightness profile with increasing radius : for @xmath32 mpc the data can be described ( @xmath33 for 131 degrees of freedom ) by a simple broken power - law model , with a break at a radius of @xmath34 kpc , and slopes in the regions internal and external to the break radius of @xmath35 and @xmath36 , respectively .
a similar result was obtained for abell 2390 by allen , ettori & fabian ( 2001b ) , although in that case the slope interior to the break radius was flatter . the data for the southeast quadrant containing the x - ray subclump reveal a clear excess of emission relative to other directions .
enhanced emission is detected between radii of @xmath37 kpc , with a maximum enhancement of a factor @xmath38 at a radius @xmath39 kpc .
for our spectral analysis , we have separated the southeast quadrant containing the x - ray subclump from the rest of the cluster .
the data for position angles of @xmath23 ( the relaxed part of the cluster ) were divided into the annular regions detailed in table 1 .
kpc ( 18 - 58 pixels ) , where the emission enhancement in fig . [
fig : im2 ] is most obvious . ]
a single spectrum was extracted for each region in 1024 pulse invariant ( pi ) channels .
the spectra were re - grouped to contain a minimum of 20 counts per pi channel thereby allowing @xmath40 statistics to be used .
background spectra , appropriate for the regions studied , were extracted from the acis - s3 blank - field data sets provided by maxim markevitch ( which are available from the chandra x - ray center ) .
separate photon - weighted response matrices and effective area files were constructed for each region using the available calibration and response files appropriate for the -120c focal plane temperature .
pixel@xmath14 sub - region of the s3 chip , a spectral response ( .rmf ) and an auxiliary response ( .arf ) matrix were created using the ciao tools @xmath41 and @xmath42 , respectively . for each of the annular regions studied , the number of source counts in each @xmath43 pixel@xmath14 sub - region was determined .
the individual .rmf and .arf files were then combined using the ftools programs @xmath44 and @xmath45 ( using a script provided by roderick johnstone ) to form a counts - weighted spectral response and auxiliary response matrix appropriate for each annulus . ]
the data from both chandra observations were modelled simultaneously using the xspec code ( version 11.01 ; arnaud 1996 ) .
we have limited our analysis to the @xmath22 kev energy band , over which the calibration of the back - illuminated ccd detectors is currently best understood .
the spectra have been modelled using the mekal plasma emission code of kaastra & mewe ( 1993 ; incorporating the fe l calculations of liedhal , osterheld & goldstein 1995 ) and the photoelectric absorption models of balucinska - church & mccammon ( 1992 ) .
we have fitted each annular spectrum using a simple , single - temperature model with the absorbing column density fixed at the nominal galactic value ( @xmath46 ; dickey & lockman 1990 ) .
the free parameters in this model were the temperature ( @xmath47 ) , metallicity ( @xmath48 ; measured relative to the solar photospheric values of anders & grevesse 1989 , with the various elements assumed to be present in their solar ratios ) and emission measure .
we note that including the absorbing column density as an additional free parameter did not result in significant improvements in the goodness of fit .
.the results from the analysis of the annular spectra ( covering position angles @xmath23 degrees ) .
temperatures ( @xmath47 ) are in kev and metallicities ( @xmath48 ) in solar units .
the absorbing column density has been fixed at the nominal galactic value of @xmath49 .
the total @xmath40 values and number of degrees of freedom ( dof ) in the fits are listed in column 4 .
error bars are the @xmath50 ( @xmath51 ) confidence limits on a single interesting parameter . [
cols="^,^,^,^,^ " , ] having adjusted the parameters for the second mass component to produce the northern arc , we then examined whether the resulting mass model can explain the other arcs observed by sahu ( 1998 ) .
unfortunately , there are no measured redshifts for the other arcs .
( sahu 1998 note a possible lower limit for arc 4 of @xmath52 based on the non - detection of [ oii ] in their ground - based spectra , although they caution that the arc may simply be a galaxy without a strong emission - line feature ) .
the details of our model predictions are shown in the right panel of fig .
[ rxjmodel ] .
the results on the predicted arc redshifts and the assumed sizes of the sources are listed in table [ arcredshifts ] .
the predicted redshifts are cosmology dependent and would be modified slightly if , for example , an accelerating universe with a different hubble constant were assumed . for the mass model used here , the predicted redshifts should be accurate to @xmath53 , with the exception of arc 4 which is determined more precisely .
we have modelled the lensed galaxies using circular gaussian brightness distributions with the full - width - at half - maximum ( fwhm ) values listed in table [ arcredshifts ] .
the fwhms for the arcs were adjusted so that their 30 per cent brightness contours provide a good fit to the stis data .
the effects of using different source sizes is illustrated by the different brightness contours in fig .
[ rxjmodel ] .
although we include arc 2 in our modelling , we caution that this source appears quite red in the colour image published by fischer & tyson ( 1997 ) and that it is possible that the source is simply an edge - on spiral galaxy .
the predicted redshift for arc 2 is lower than @xmath54 , since otherwise the lens model would predict a much stronger distortion .
arcs 3 , 4 and 5 are consistent with being at the same redshift .
the projected separation of the sources producing arcs 4 and 5 is 43.4 @xmath55 kpc .
although arcs 4 and 5 have a similar colour in the data of fischer & tyson ( 1997 ) , it is not possible to explain them as images of the same source using our mass model . in exploring different redshifts for arc 3
, we found that for redshifts higher than @xmath56 it is possible to produce a counter image on the other side of the central galaxy .
this alerted us to the presence of a blue object ( identified from the colour image of fischer & tyson 1997 ) to the south - east of the merging subclump ( see the arrow in the left panel of fig .
[ rxjmodel ] ) .
however , both the appearance in the stis image and our lensing analysis suggest that these objects are not images of the same source .
we note that circular nfw models for one or both of the main mass components can not explain the observed arc shapes and orientations well . in particular , the orientation of arc 4 reflects the orientation of the critical curve ( fig .
[ rxjmodel ] ) in its vicinity .
a small @xmath57 difference between the orientation of the observed and predicted arc 4 remains in our best - fit model , which points to a small imprecision of the lensing model .
the grey circle in fig . [ fig : weak ] shows the azimuthally - averaged , surface mass density contrast at @xmath58 kpc ( centred on the main mass component ) for the best - fit strong lensing model .
note that the strong lensing results should lie above the x - ray results at the same radii , since the chandra analysis excludes the regions of the cluster affected by the second mass subclump .
the surface mass density contrast rises sharply between the innermost weak lensing point and the strong lensing result , suggesting that the second mass component is centrally concentrated and does not make a major contribution to the total mass beyond the strong lensing regime .
in conclusion we find that a simple two - component mass model , with ellipticities and orientations for the mass components matching those of the dominant cluster galaxies , provides a reasonable description for the overall arc geometry in rxj1347.5 - 1145 .
such a model is consistent with the chandra x - ray and weak lensing results .
the results on the electron density and cooling time as a function of radius , determined from the image deprojection analysis using the best - fit nfw mass model , are shown in fig .
[ fig : deproj ] . within the central 500 kpc radius ,
the electron density profile can be parameterized ( @xmath59 for 34 degrees of freedom ) by a @xmath26-profile with a core radius , @xmath60 kpc , @xmath61 and a central density , @xmath62 @xmath63 ( @xmath50 errors ) .
we measure a central cooling time ( _ i.e. _ the mean cooling time within the central 1.97 arcsec ( @xmath64 kpc ) bin of @xmath65 yr ( uncertainties are the 10 and 90 percentile values from 100 monte carlo simulations ) . the x - ray gas - to - total - mass ratio as a function of radius , @xmath66 , determined from the chandra data is shown in fig .
[ fig : baryon ] .
we find that the best - fit @xmath67 value rises rapidly with increasing radius within the central @xmath68 kpc radius and then remains approximately constant , or rises slowly , out to the limits of the data at @xmath69 mpc where we measure @xmath70 .
following the usual arguments , which assume that the properties of clusters provide a fair sample of those of the universe as a whole ( white & frenk 1991 ; white 1993 ; white & fabian 1995 ; evrard 1997 ; ettori & fabian 1999 ) , we may use our results on the x - ray gas mass fraction to estimate the total matter density in the universe , @xmath71 . assuming that the luminous baryonic mass in galaxies in rxj1347 - 1145 is @xmath72 of the x - ray gas mass ( white 1993 ; fukugita , hogan & peebles 1998 ) and neglecting other possible sources of baryonic dark matter in the cluster
, one can show that @xmath73 , where @xmath74 is mean baryon density in the universe . for @xmath75 ( omeara 2001 ) we obtain @xmath76 . for a @xmath77cdm cosmology with @xmath78
, our results for rxj1347.5 - 1145 would imply @xmath79 . under the assumption that the x - ray gas in the core of rxj1347.5 - 1145 is in a steady - state cooling flow
, we can parameterize the luminosity profile in terms of an equivalent mass deposition rate ( white 1997 ) .
the results of this calculation are shown in fig .
[ fig : mdot ] .
we see that the integrated mass deposition rate ( mdr ) rises with increasing radius within the central 100 kpc , to a maximum value of @xmath80 .
we have examined the spectrum for the central 100 kpc ( radius ) region using a model appropriate for a cooling flow with distributed mass deposition .
( we use the model of nulsen 1998 and assume that the integrated mdr within radius @xmath81 , @xmath82 . ) the normalization of the cooling - flow component is parameterized in terms of a mass deposition rate , @xmath83 .
the cooling flow was also allowed to be absorbed by an intrinsic column density , @xmath84 , of cold gas ( with solar metallicity ) located at the redshift of the cluster .
both @xmath83 and @xmath85 were free parameters in the fits .
accounting for projection effects , we find that the @xmath22 kev spectrum for the central 100 kpc radius region is consistent with a cooling flow of @xmath86 , intrinsically absorbed by a column density of @xmath87 , with a mean mass - weighted flow temperature of @xmath88 kev . fitting the mass deposition profile in fig .
[ fig : mdot ] with a broken power - law model , we determine a break radius of @xmath89 kpc , and slopes internal and external to the break radius of @xmath90 and @xmath91 , respectively ( @xmath50 errors ) .
allen ( 2001a ) argue that such breaks may mark the outer edges of the present - day cooling flows in clusters .
the predicted mdr for the central @xmath92 kpc radius , where a steady cooling flow with gas cooling to zero degrees might be expected to occur , is @xmath93 .
the cooling time of the x - ray gas at the break radius is @xmath94 yr .
as discussed in section 4.2 , the region of enhanced x - ray emission ( and enhanced sz effect ; komatsu 2001 ) to the southeast of the main x - ray peak is significantly hotter ( @xmath95 kev ) than at the same distance in other directions ( @xmath96 kev ) .
following markevitch ( 1999 ) , we can model the properties of this region as a one - dimensional shock . in this case
, the relative velocity , @xmath97 , of the infalling subcluster and the shock compression factor , @xmath98 , can be estimated from the pre - shock ( @xmath99 ) and post - shock ( @xmath100 ) temperatures . for @xmath101 kev and @xmath102 kev
, we obtain @xmath103 and @xmath104 .
if we instead adopt @xmath105 kev , a temperature consistent with the predicted mass of the merging subcluster from the strong lensing analysis discussed in section 6.2 ( using the mass - temperature relation of allen 2001c ) , then for @xmath102 kev we obtain @xmath106 and @xmath107 .
we note that the roughly spherical appearance of the shocked gas suggests that this material has expanded following the initial shock .
since the material will cool as it expands , it is likely that the initial post - shock temperature exceeded 18 kev and , for @xmath105 , that the relative collision velocity exceeded 5000 .
such values are broadly consistent with the peak x - ray surface brightness enhancement of a factor @xmath38 observed at a radius of @xmath108 kpc .
the second brightest galaxy in the cluster lies approximately 18 arcsec to the east and 2 arcsec to the south of the dominant cluster galaxy .
this galaxy has an extended , diffuse halo and appears to lie at the centre of a significant sub - concentration of galaxies within the cluster ( fig .
[ rxjmodel]a ) .
these properties are consistent with it having been the dominant galaxy of the recently merged subcluster .
the region of shocked x - ray gas lies approximately 11 arcsec to the south and 6.5 arcsec to the west of the second brightest galaxy .
assuming that the dark matter and galaxies associated with the subcluster have moved on ahead of the shock , it seems plausible that the subcluster travelled in from a pa of @xmath109 degrees .
this is consistent with the extension of the x - ray emission observed on large scales to the south ( fig .
[ fig : im1]b ) and the weak lensing mass map presented by fischer & tyson ( 1997 ) . given the separation of the shocked gas and second brightest galaxy , we estimate that the shock probably occurred a few @xmath110 years ago .
this timescale is consistent with the spatial extent of the shocked gas , assuming that it has expanded at the sound speed , @xmath111 .
note also that this timescale is comparable to the time required for the electrons and ions to reach equipartition , so it is possible that the electron and ion populations in the shocked gas have slightly different temperatures . despite the evidence for shocked gas in the southeast quadrant , the chandra data for the rest of the cluster place rxj1347.5 - 1145 on the @xmath112 and @xmath113 relations for relaxed , massive clusters discussed by allen , schmidt & fabian ( 2001 ) .
as discussed in section 4.4 , the shocked gas produces only a small rise in the mean , emission - weighted temperature measured within the central 0.74 mpc radius .
the excess bolometric luminosity of the southeast quadrant between radii of 60 - 195 kpc ( where the shock is most apparent ) is @xmath114 , which corresponds to only @xmath115 per cent of the total cluster luminosity ( section 4.4 ) .
it therefore appears that the overall temperature and luminosity of the cluster have not been boosted substantially by the merger event .
this is supported by the consistent x - ray and weak lensing mass results discussed in section 6.1 .
ricker & sarazin ( 2001 ) and ritchie & thomas ( 2002 ) discuss simulations of mergers between clusters with similar central densities and argue that such mergers can lead to significant short - term increases in the overall x - ray temperatures and luminosities of clusters during the periods of closest approach .
these effects are most pronounced and long - lived in clusters with relatively low central densities ( @xmath116 few @xmath117@xmath63 ; ritchie & thomas 2002 ) . the fact that large boosts in the overall temperature and luminosity are not observed in rxj1347.5 - 1145 may be related to the very high central density in the main cluster core ( @xmath118@xmath63 ; fig .
[ fig : deproj]a ) and could indicate that the recently - merged subcluster had a relatively low central gas density and interacted only weakly with the main cluster core .
recently , cohen & kneib ( 2002 ) have published spectroscopic redshift measurements for 47 cluster members for which they determine a velocity dispersion of @xmath119 .
inspection of their fig .
2 suggests that a significant fraction of the galaxies with measured redshifts may have been associated with the recently merged subcluster .
this could explain the relatively low velocity dispersion , in comparison to the x - ray and lensing results .
the agreement between the chandra x - ray and gravitational lensing mass measurements for rxj1347.5 - 1145 reinforces the results and conclusions drawn from previous studies of other relaxed lensing clusters abell 2390 ( allen 2001b ) and abell 1835 ( schmidt 2001 ) ; see also the results for ms1358.4 + 6245 by arabadjis ( 2002 ) .
the close agreement of the independent x - ray and lensing masses indicates that the mass measurements are robust and limits the contributions from non - thermal sources of pressure support in the x - ray gas , such as bulk and/or turbulent motions and magnetic fields , in the relaxed regions of the cluster .
the main conclusions from this work may be summarized as follows : 0.2 cm \(i ) we have reported chandra observations of rxj1347.5 - 1145 , the most x - ray luminous cluster of galaxies known .
we have identified a region of shocked gas ( enhanced x - ray brightness and temperature ) to the southeast of the main x - ray peak , at a position consistent with the region of enhanced sz effect reported by komatsu ( 2001 ) .
the shocked gas probably results from recent subcluster merger activity .
the merger appears not to have boosted the overall luminosity and temperature of the cluster substantially .
\(ii ) excluding the data for the southeast quadrant , we have measured the density , temperature and mass profiles for the cluster
. the mass profile can be parameterized by an nfw model with a scale radius @xmath120 mpc and a concentration parameter , @xmath121 ( 68 per cent confidence limits ) .
the normalization of the mass profile may also be expressed in terms of an effective velocity dispersion , @xmath122 .
\(iii ) the best - fit chandra mass model for rxj1347.5 - 1145 is in good agreement with independent measurements from weak gravitational lensing studies ( fischer & tyson 1997 ) . the observed strong lensing configuration in the cluster core can also be explained with the introduction of an additional mass clump centred on the second brightest galaxy , which is likely to have been the dominant galaxy of the recently merged subcluster .
we thank r. johnstone and s. ettori for supplying a number of scripts used in the x - ray reduction / analysis and philippe fischer for communicating his weak lensing results in a machine readable form .
swa and acf acknowledge the support of the royal society .
allen s.w . , 1998 , mnras , 296 , 392 allen s.w .
, fabian a.c . , 1998 , mnras , 297 , l57 allen s.w . , ettori s. , fabian a.c . ,
2001b , mnras , 324 , 877 allen s.w .
, schmidt r.w .. , fabian a.c . , 2001c , mnras , 328 , l37 allen s.w . , fabian a.c .
, johnstone r.m . ,
nulsen p.e.j . ,
arnaud k.a . , 2001a , mnras , 322 , 589 anders e. , grevesse n. , 1989 , geochemica et cosmochimica acta 53 , 197 arabadjis j.s .
, bautz m.w . ,
garmire g.p . , 2002 ,
apj , in press ( astro - ph/0109141 ) arnaud , k.a . , 1996 , in astronomical data analysis software and systems v , eds .
jacoby g. and barnes j. , asp conf .
series volume 101 , p17 balucinska - church m. , mccammon d. , 1992 , apj , 400 , 699 cohen j.g .
, kneib j .-
, 2002 , apj , submitted ( astro - ph/0111294 ) dickey j.m . , lockman f.j . , 1990 , ara&a , 28 , 215 ebeling h. , rangarajan f.v.n . , white d.a . ,
2002 , mnras , submitted ettori s. , fabian a.c .
, 1999 , mnras , 305 , 834 ettori s. , allen s.w .
, fabian a.c . , 2001 ,
mnras , 322 , 187 evrard a.e . , 1997 , mnras , 292 , 289 fischer p. , tyson j.a .
, 1997 , aj , 114 , 14 fukugita m. , hogan c.j .
, peebles p.j.e . , 1998 , apj , 503 , 518 jones c. , forman w. , 1984 , apj , 276 , 38 kaastra j.s . , mewe r. , 1993 , legacy , 3 , 16 , heasarc , nasa komatsu e. , kitayama t. , suto y. , hattori m. , kawabe r. , matsuo h. , schindler s. , kohji y. , 1999 , apj , 516 , 1 komatsu e. , 2001 , pasj , 53 , 57 kriss g.a . ,
cioffi d.f .
, canizares c.r . , 1983 ,
apj , 272 , 439 liedhal d.a . ,
osterheld a.l . ,
goldstein w.h . , 1995 , apj , 438 , l115 markevitch m. , sarazin c.l .
, vikhlinin a. , 1999 , apj , 521 , 526 moore b. , quinn t. , governato f. , stadel j. , lake g. , 1999 , mnras , 310 , 1147 navarro j.f . , frenk c.s . ,
white s.d.m .
, 1997,apj , 490 , 493 nulsen p.e.j . , 1998 , mnras , 297 , 1109 omeara j.m . , tytler d. , kirkman d. , suzuki n. , prochaska j.x . , lubin d. , wolfe a.m. , 2001 , apj , 552 , 718 pointecouteau e. , giard m. , benoit a. , dsert f.x . , aghanim n. , coron n. , lamarre j.m . ,
delabrouille j. , 1999 , apjl , 519 , 115 pointecouteau e. , giard m. , benoit a. , dsert f.x .
, bernard j.p . , coron n. , lamarre j.m . , 2001 , apj , 552 , 42 ricker p.m. , sarazin c.l . , 2001 , apj , 561 , 621 ritchie b.w . , thomas p.a . , 2002 , mnras , 329 , 675 sahu k.c . , 1998 , apjl , 492 , 125 schindler s. , 1995 , a&a , 299l , 9 schindler s. , hattori m. , neumann d.m .
, bhringer h. , 1997 , a&a , 317 , 646 schmidt r.w .
, allen s.w .
, fabian a.c
. , 2001 , mnras , 327 , 1057 schneider p. , ehlers j. , falco e.e . , 1992 ,
_ gravitational lensing _ , springer verlag , berlin weisskopf m.c . , tananbaum h.d . ,
van speybroeck l.p .
, odell s.l . , 2000 , spie 4012 , 1 ( astro - ph/0004127 ) white d.a .
, fabian a.c . , 1995 , mnras , 273 , 72 white d.a . , jones c. , forman w. , 1997 , mnras , 292 , 419 white s.d.m . ,
frenk c.s . , 1991 , apj , 379 , 52 white s.d.m . , efstathiou g. , frenk c.s . , 1993 , mnras , 262 , 1023 | we present chandra observations of rxj1347.5 - 1145 , the most x - ray luminous cluster of galaxies known .
we report the discovery of a region of relatively hot , bright x - ray emission , located approximately 20 arcsec to the southeast of the main x - ray peak , at a position consistent with the region of enhanced sunyaev - zeldovich effect reported by komatsu ( 2001 ) .
we suggest that this region contains shocked gas resulting from a recent subcluster merger event . excluding the data for the southeast quadrant
, the cluster appears relatively relaxed .
the x - ray gas temperature rises from @xmath0 kev within the central @xmath1 kpc radius to a mean value of @xmath2 kev between @xmath3 mpc .
the mass profile for the relaxed regions of the cluster , determined under the assumption of hydrostatic equilibrium , can be parameterized by a navarro , frenk & white ( 1997 ) model with a scale radius @xmath4 mpc and a concentration parameter @xmath5 .
the best - fit chandra mass model is in good agreement with independent measurements from weak gravitational lensing studies .
strong lensing data for the central regions of the cluster can be also explained by the introduction of an additional mass clump centred on the second brightest galaxy .
we argue that this galaxy is likely to have been the dominant galaxy of the recently merged subcluster . #
1to 0pt#1 = " 2218 # 1 # 1 to size # 1 to size = cmr10 scaled2 = cmbx10 scaled2 = cmti10 scaled2 ps .
galaxies : clusters : individual : rxj1347.5 - 1145 gravitational lensing
x - rays : galaxies cooling flows intergalactic medium |
inflation generates the source of the large - scale perturbation that is needed to explain the inhomogeneity and the structure of the universe @xcite . in the original scenario of the inflationary universe , the adiabatic mode of the inflaton fluctuation sources the curvature perturbation when the perturbation leaves the horizon .
however , if there are many light scalar fields during inflation , inflation may create isocurvature perturbations for these fields , and those isocurvature perturbations may cause significant creation of the curvature perturbation after the horizon exit . the generation of the curvature perturbation at the end of inflation @xcite is basically caused by ( 1 ) modulation of the coupling constants or ( 2 ) two - field hybrid inflation in which both inflaton fields are coupled to the waterfall field .
one thing that is common in the previous models of the `` end of inflation '' scenario is that * the extra light field has an explicit coupling to the inflaton sector * @xcite .
one might think that this requirement is essential and quite obvious ; however _ we are considering in this paper the removal of this basic requirement .
_ in our model the additional fields are not coupled to the inflaton sector , but it changes the definition of the adiabatic field . as a result ,
the inflation end is not identical to the uniform density hypersurfaces .
first consider a model in which the inflaton @xmath0 and the waterfall field @xmath1 have the hybrid - type potential given by @xmath2 this defines the `` inflaton sector '' of the model .
suppose that inflation starts with @xmath3 , and the end of inflation is defined by @xmath4 , where the waterfall begins .
the critical point @xmath5 is given by @xmath6 the number of e - foldings is given by @xmath7 here @xmath8 is the slow - roll parameter defined by @xmath9 where @xmath10 is the hubble parameter during inflation and the subscript means the derivative with respect to the field .
in the single - field inflation model we always find the trivial coincidence between the uniform density hypersurfaces and the end of inflation . in that case
, one can not find the perturbation @xmath11 created at the end @xcite . here
@xmath11 measures the discrepancy between the uniform density hypersurfaces and the end of inflation .
going back to the past model @xcite , generation of the perturbation at the end of inflation is considered for a two - field model . in that case
the additional light field is coupled to the waterfall field and that causes @xmath12 at the end .
namely , if one introduces another inflaton @xmath13 that has the same interaction as the primary field @xmath14 , the end of inflation depends on @xmath13 as ; @xmath15 then , suppose that @xmath13 is much lighter than @xmath14 , the entropy perturbation @xmath16 creates the perturbation @xmath17 at the end .
this leads to the perturbation of the number of e - foldings at the end of inflation , which is given by @xmath18_e \ne 0.\ ] ] this is the usual scenario of @xmath12 . in this paper
, we consider a similar mechanism of generating curvature perturbation at the end of inflation , _ but in contrast to the usual scenario , the additional fields are decoupled from the inflaton sector_. keeping the inflaton @xmath0 in the potential given by eq.([base - pot ] ) , we introduce * free * light fields ( @xmath19 ) , which have no explicit interaction with the waterfall field @xmath1 .
although these fields ( @xmath19 ) are not coupled to the `` inflaton sector '' , they must participate in the definition of the adiabatic inflaton field ; @xmath20 , which is mandatory .
a typical situation is shown in fig.[fig : plusone ] . )
are added to the conventional hybrid - type potential .
@xmath21 is the entropy perturbation , and @xmath22 measures the velocity ratio between @xmath0 and the others .
note that @xmath23 is assumed in the original hybrid inflation , but the assumption is not quite obvious when there are many light scalar fields in the model , even if they are decoupled from the hybrid - type potential . ] in this paper we consider the perturbation caused by the entropy perturbation @xmath21 .
the angled trajectory ( @xmath24 ) causes significant creation of the curvature perturbation at the end .
one might be skeptical about the generation of the curvature perturbation at the end , if it is realized just by adding free scalar fields to the model . to illustrate what happens in this model
, we show the simplest calculation , in which one `` standard '' inflaton @xmath0 and one additional light field @xmath25 have the same quadratic potential with the same mass @xmath26 .
therefore , the potential is given by @xmath27 this assumption makes the trajectory straight , and removes ambiguities related to the possible non - trivial evolution of the perturbations during inflation .
( see fig.[fig2nd ] . )
[ fig2nd ] , which appears along the uniform density hypersurfaces ( dotted circle).,title="fig : " ] defining the adiabatic field @xmath28 , we find @xmath29 in this model , the end of inflation defined by the constant @xmath4 does not coincide with the uniform density hypersurfaces defined by constant @xmath30 .
we find the end is given by @xmath31 which is perturbed by @xmath32 .
therefore , although @xmath5 is not perturbed in this model , the entropy perturbation ( @xmath33 ) causes @xmath12 at the end of inflation . the inhomogeneous end of inflation caused by the entropy perturbation is thus given by @xmath34\phi_c\delta \theta + \frac{1}{2}\left [ \frac{1+\cos^2\theta}{\sin^3 \theta } \right ] \phi_c\left(\delta \theta\right)^2+ .. ,\nonumber\\ & \simeq & -\frac{\cos\theta \phi_c}{\sin^2 \theta}\left(\frac{\delta s}{\sigma}\right ) _ * + \frac{1}{2}\frac{(1+\cos^2\theta)\phi_c}{\sin^3 \theta}\left(\frac{\delta s}{\sigma } \right)_*^2\nonumber\\ & = & -\frac{1}{\tan\theta}\frac{\sigma_e}{\sigma_*}\delta s _ * + \frac{1+\cos^2\theta}{2\sin^2 \theta}\frac{\sigma_e}{\sigma_*^2 } \left(\delta s_*\right)^2,\end{aligned}\ ] ] where the subscripts `` @xmath35 '' and `` @xmath36 '' denote the value at the end of inflation and at the epoch when the perturbation leaves the horizon , respectively . in this simplest example , the relation @xmath37_*$ ] is exact and @xmath38 is constant after the horizon exit .
note that in this formalism @xmath39 is gaussian but @xmath40 is not always a gaussian perturbation .
the curvature perturbation generated at the end of inflation is thus given by @xmath41_e\nonumber\\ & \simeq & \frac{1}{\eta_e}\frac{1}{\tan\theta } \left(\frac{\delta s}{\sigma}\right ) _ * -\frac{1+\cos^2\theta}{2\eta_e\sin^2 \theta } \left(\frac{\delta s}{\sigma } \right)_*^2.\end{aligned}\ ] ] since the perturbation of the inflation generated at the horizon exit is given by @xmath42_*,\ ] ] we find the ratio between the `` initial '' and `` at the end '' perturbations ; @xmath43 where the ratio is for the first order perturbations and @xmath44 is used for the calculation .
we can see that @xmath11 dominates ( i.e , @xmath45 is realized ) when @xmath46 .
the spectral index @xmath47 for the perturbation @xmath11 is @xcite @xmath48 where @xmath49 and @xmath50 are defined at the horizon exit . _
the spectral index is obviously different from the standard hybrid inflation scenario .
_ using the second order perturbation , we can estimate the non - gaussianity parameter @xmath51 where @xmath52 and @xmath53 suggest @xmath54 . here
the definition of @xmath55 is @xmath56 where a subscript @xmath57 denotes the derivative with respect to the @xmath57-th field , and `` @xmath58 '' includes loop corrections that are usually negligible .
since @xmath55 is obviously small for the quadratic potential , in which @xmath59 is mandatory , we need to consider non - quadratic potential for the enhancement . one way to enhance @xmath55 is to consider the effective potential that is dominated by the higher polynomial ; @xmath60 this potential may allow @xmath61 , which makes it possible to find @xmath45 and @xmath62 at the same time .
if there are @xmath63 light scalar fields ( @xmath19 , @xmath64 ) in the inflationary universe , a naive statistical expectation is @xmath65 . to avoid the significant creation of @xmath11 at the end , these scalar fields must be settled in their minima before the end of inflation , or should have very flat potential compared with the inflaton , so that @xmath66 is a plausible approximation during inflation .
if there are too many fields in the universe , n - flation @xcite may start before the onset of hybrid inflation . then hybrid inflation may begin when @xmath67 starts dominating the universe .
above we considered the simplest set - up for the model , in which only one field ( adiabatic field ) appears in the effective potential during inflation , while the end is caused by the coupling between @xmath0 and @xmath1
. however , in more general cases there may be many fields that may have positive / negative mass terms or higher polynomials . even in that case
, the first order perturbation requires very simple parameters . from fig.[fig : general ] , we can see that for @xmath68 the perturbation at the end of inflation is given by @xmath69 where we assume @xmath70 . here
@xmath22 is defined by @xmath71 and the adiabatic field @xmath30 is @xmath72 more general definitions of the adiabatic field ( probably including non - canonical kinetic terms @xcite ) would be interesting , but they are not the topic in this paper . .
the perturbed trajectory is approximately parallel when @xmath73 . ]
the curvature perturbation generated at the end of inflation is thus given by @xmath74\times \frac{1}{\tan\theta_e}.\end{aligned}\ ] ] comparing the result with the curvature perturbation generated at the horizon exit ( @xmath75 ) , we find that @xmath11 is enhanced by the factor @xmath76 . if one compares this result with the `` standard '' calculation @xmath77 _ * \simeq \left[\frac{1}{\sqrt{8\pi^2\epsilon_\phi}}\frac{h}{m_p}\right],\ ] ] where the slow - roll parameter @xmath78 is defined by @xmath79 , one finds @xmath80 for the simplest ( two field , equal mass ) model we find @xmath81 which leads to @xmath82 or @xmath83 the minimum ratio @xmath84 is obtained when @xmath85 , where both perturbations contribute ( i.e , @xmath86 ) .
obviously , the additional light field changes the spectrum .
in the inflationary universe we may expect many light scalar fields @xmath19 . since the adiabatic field during inflation
must be defined using all the fields that are moving during inflation , the mismatch between the end of hybrid inflation ( usually defined by @xmath0 that is coupled to the waterfall field ) and the uniform density hypersurfaces ( always defined by the adiabatic field ) causes creation of @xmath87 at the end of inflation . using the conventional @xmath87 formalism
, we showed that the fields that are decoupled from the inflaton sector can play significant role in creating the curvature perturbation .
the signs of the light fields may appear in the spectral index and/or in the non - gaussianity , which can be used to discriminate the origin of the perturbation .
the precise calculation of these parameters is quite difficult when there are many fields during inflation , but our modest prediction is that the deviation from the standard prediction may indicate the presence of light scalar fields in the early universe . in this paper hybrid inflation
is simplified assuming that it ends with an instantaneous waterfall starting at the critical instability point @xmath5 .
this simplification is not valid if inflation can continue for more than 60 e - folds during the waterfall @xcite .
however , the model that spends more than 60 e - folds during the waterfall should correspond to hilltop inflation ( with probably some modulation caused by the interaction with @xmath0 @xcite ) that should be discriminated from the conventional hybrid inflation scenario .
we discussed the physics related to the evolution before the waterfall using the simplification of the instantaneous waterfall .
our simplification is valid only when the end of inflation mechanism @xcite is valid for the hybrid inflation model .
we wish to thank k.shima for encouragement , and our colleagues at nagoya university and lancaster university for their kind hospitality
. 1 d. h. lyth and a. r. liddle , _ the primordial density perturbation _ , cambridge university press , 2009 .
f. bernardeau , l. kofman and j. -p .
uzan , phys .
d * 70 * , 083004 ( 2004 ) [ astro - ph/0403315 ] ; d. h. lyth , jcap * 0511 * , 006 ( 2005 ) [ astro - ph/0510443 ] ; m. p. salem , phys .
d * 72 * , 123516 ( 2005 ) [ astro - ph/0511146 ] .
m. sasaki , prog .
* 120 * , 159 ( 2008 ) [ arxiv:0805.0974 [ astro - ph ] ] ; a. naruko and m. sasaki , prog .
phys . * 121 * , 193 ( 2009 ) [ arxiv:0807.0180 [ astro - ph ] ] . t. matsuda , jcap * 0906 * , 002 ( 2009 ) [ arxiv:0905.0308 [ astro-ph.co ] ] ; t. matsuda , class
. quant .
* 26 * , 145016 ( 2009 ) [ arxiv:0906.0643 [ hep - th ] ] ; k. dimopoulos , k. kohri , d. h. lyth and t. matsuda , arxiv:1110.2951 [ astro-ph.co ] ; k. dimopoulos , k. kohri and t. matsuda , arxiv:1201.6037 [ hep - ph ] . a. r. liddle , a. mazumdar , f. e. schunck , phys .
rev . * d58 * , 061301 ( 1998 ) .
[ astro - ph/9804177 ] ; s. dimopoulos , s. kachru , j. mcgreevy , j. g. wacker , jcap * 0808 * , 003 ( 2008 ) .
[ hep - th/0507205 ] .
t. matsuda , phys .
b * 682 * , 163 ( 2009 ) [ arxiv:0906.2525 [ hep - th ] ] ; t. matsuda , jhep * 0810 * , 089 ( 2008 ) [ arxiv:0810.3291 [ hep - ph ] ] .
s. clesse , phys .
d * 83 * , 063518 ( 2011 ) [ arxiv:1006.4522 [ gr - qc ] ] .
t. matsuda , phys .
b * 665 * , 338 ( 2008 ) [ arxiv:0801.2648 [ hep - ph ] ] ; t. matsuda , jcap * 0805 * , 022 ( 2008 ) [ arxiv:0804.3268 [ hep - th ] ] . | we show that the free light scalar fields that may exist in the inflationary universe can change the predictions of the hybrid inflation model .
possible signatures are discussed , which can be used to discriminate the sources of the spectrum . |
magnetohydrodynamic ( mhd ) turbulence is the most promising candidate for angular momentum transport in astrophysical disk systems .
nonlinear behaviors of the magnetorotational instability ( mri ) are actively investigated as a driving mechanism of mhd turbulence over the last decades ( balbus & hawley 1991 , 1998 ) .
the central issue in mri research is its nonlinear properties , in particular , the saturation amplitude of the instability .
although the key processes governing the nonlinear saturation are scoped by global and local numerical studies , it is not fully explained yet ( hawley & balbus 1992 ; hawley et al .
1995 , 1996 ; brandenburg et al . 1995 ; matsumoto & tajima 1995 ; stone et al . 1996
; hawley 2000 ; machida et al .
2000 ; arlt & rdiger 2001 ; balbus 2003 ) .
recently , lesur & ogilvie ( 2008 ) argue that , in the shearing box context with zero - net vertical flux , mhd turbulence is sustained through nonlinear classical dynamo activity once the mri is operated .
it would be necessary and significative to study the saturation process from the microscopic viewpoint of the physical sustaining mechanism for mhd turbulence as they have done .
ohmic dissipation is one of the crucial processes that determine the saturation amplitude of the mri .
linear growth rate of the mri can be reduced significantly because of the suppression by ohmic dissipation .
two- and three - dimensional local shearing box simulations ( sano et al . 1998 , 2004 ; sano & stone 2002 ) have shown that physical properties of the saturated turbulence depend on the lundquist number @xmath11 , where @xmath1 is the alfvn velocity , @xmath12 is the angular velocity , and @xmath13 is the magnetic diffusivity ( see also fleming et al .
2000 ; ziegler & rdiger 2001 ; liu et al .
particularly , when @xmath14 , the size of the saturated stress decreases with the decrease of @xmath15 ( sano & stone 2002 ; pessah et al .
it is also pointed out that magnetic reconnection plays an important role in the energy dissipation of mri driven turbulence ( sano & inutsuka 2001 ) .
numerical issues are one of the main reasons why the saturation physics of the mri is remained to be understood .
fromang & papaloizou ( 2007 ) demonstrate the efficiency of angular momentum transport decreases linearly with the grid spacing as the resolution increases .
although it is very difficult to distinguish between the numerical and physical factors working as the saturation mechanism ( king et al .
2007 ; silvers 2007 ) , current researches of the mri pay much attention to the numerical factors with the greatest care .
pessah et al .
( 2007 ) derive a scaling law of the saturated stress from past wide variety of numerical results by analytically eliminating the numerical factors such as box size and resolutions .
more straightforward way for decontaminating the numerical factors is to bring explicitly the physical diffusivities much larger than the numerical one into the computational study .
lesur & longaretti ( 2007 ) have performed first systematic study of the mri in the presence of both viscous and magnetic dissipations , which are larger than the numerical diffusivities . for the cases with nonzero net flux of the vertical field ,
the transport property in the saturated state depends on the magnetic prandtl number @xmath16 , where @xmath17 is the kinematic viscosity .
fromang et al . (
2007 ) have reported that in zero magnetic flux cases the turbulent activity is an increasing function of the magnetic prandtl number @xmath18 .
linear behaviors of the mri in the presence of both the viscosity and resistivity are analytically studied by pessah & chan ( 2008 ) comprehensively .
the magnetic prandtl number @xmath18 takes a wide range of values in astrophysical disk systems . in protoplanetary disks surrounding young stellar objects ,
the magnetic prandtl number is much smaller than unity because of their low ionization degree ( nakano 1984 ; umebayashi & nakano 1988 ; sano et al . 2000 ) . in accretion disks of compact x - ray sources and active galactic nuclei , the magnetic prandtl number ranges from @xmath19 to @xmath20 depending on the distance from the central object ( balbus & henri 2008 ) . in collapsar disks which is known as the central engine of gamma - ray bursts ( woosley 1993 )
, the physical state with high magnetic prandtl number of @xmath21 is expected to be realized in their evolutionary stage as a result of the large neutrino viscosity ( masada et al .
therefore , more systematic and deeper study on the mri in the presence of both the viscosity and resistivity is quite important for understanding the accretion process triggered by the mri in various disk systems
. one important unsettled matter , in these situations , is the role of the kinematic viscosity at the nonlinear stage of the mri . in general , the viscosity as well as the magnetic resistivity can suppress the growth of the mri .
however the dependence of nonlinear outcome on the prandtl number indicates that the role of the viscosity in mri turbulence could be different from that of the resistivity .
then , the main purpose of this paper is to reveal nonlinear features of the mri in viscous accretion disks .
as is described in what follows , linear growth of the mri is characterized by the reynolds number @xmath0 in the viscous fluid , and by the lundquist number @xmath22 in the resistive fluid .
focusing on these two non - dimensional parameters , we clarify the difference in nonlinear behaviors of the mri between the viscous and resistive systems .
our paper is organized as follows . in 2 , linear features of the mri in the viscous fluid are presented . in 3 ,
nonlinear behavior of the mri is investigated by two - dimensional mhd simulations taking account of the viscous terms .
the differences between the effect of the viscosity and resistivity are also clarified in
finally we make an physical explanation for our nonlinear results with the help of the linear dispersion relation . applying our results to double diffusive systems ,
we predict a condition for sustaining active mri turbulence in the presence of both the viscosity and resistivity in 4 .
first , we provide the linear features of the mri in a viscous accretion disk threaded by a uniform vertical field @xmath23 . plane - wave perturbation theory , with wkb spatial and temporal dependence @xmath24 , gives a local axisymmetric dispersion equation for the mri , @xmath25\tilde{\gamma } ^2 & & \nonumber \\
+ \frac{2}{r_{\rm mri}}\tilde{k}_z^4\tilde{\gamma } + ( \tilde{k}_z^4 - 2q\tilde{k}_z^2 ) = 0 & & \ ; , \label{eq1 } \end{aligned}\ ] ] ( menou et al .
2006 ; masada et al . 2007 ;
pessah & chan 2008 ) , where @xmath26 is the growth rate normalized by angular velocity @xmath3 , and @xmath27 is the vertical wavenumber normalized by @xmath28 . the epicyclic frequency , normalized by the angular velocity , can be expressed as @xmath29^{1/2 } $ ] by using the shear parameter @xmath30 .
the reynolds number for the mri is defined as @xmath31 . here
the characteristic velocity and length are @xmath32 and @xmath33 , respectively . in this paper , we focus on the mri in keplerian disks where the epicyclic frequency is equal to the angular velocity ( @xmath34 ) .
the dispersion equation ( [ eq1 ] ) is characterized by the reynolds number @xmath4 .
the linear growth rate of the mri is shown as a function of the vertical wavenumber for the cases @xmath35 , @xmath36 , @xmath37 and @xmath38 in figure [ fig1 ] .
when the kinematic viscosity is negligible ( @xmath39 ) , the dispersion relation is identical to that of the ideal mhd case .
if the reynolds number is less than unity , the growth of the mri is suppressed and the maximum growth rate is reduced significantly . the most unstable wavenumber decreases with decreasing @xmath4 , because the viscous damping becomes more efficient for shorter wavelength perturbations .
however , it is interesting that the critical wavenumber for the instability , @xmath40 , remains unchanged in spite of the size of @xmath4 . figures [ fig2]a and [ fig2]b show the growth rate and wavenumber of the fastest growing mode as a function of the reynolds number .
these figures indicate that @xmath41 and @xmath42 are proportional to @xmath43 when @xmath44 ( pessah & chan 2008 ) . in the regime of @xmath44 ,
the dispersion equation ( [ eq1 ] ) can be simplified and reduced to @xmath45\ g^2 - 2q k^2 = 0 \ ; , \label{eq2}\ ] ] where @xmath46 and @xmath47 . the fastest growing wavelength and the maximum growth rate
are obtained analytically from this equation ; @xmath48^{1/4}\ r_{\rm mri}^{1/2 } \ ; , \label{eq3 } \\ \tilde{\gamma}_{\rm max } & \equiv & \left ( \frac{\gamma_{\rm max}}{\omega } \right ) = \left [ \frac{q^2}{2(2-q ) } \right]^{1/4 }
\ r_{\rm mri}^{1/2}\;. \label { eq4 } \end{aligned}\ ] ] qualitative features of the fastest growing mode can be explained as follows : keplerian shear flow is a key ingredient of the unstable growth of the mri .
the viscous dissipation affects the growth of the mri when the damping rate is comparable to the shear rate for a perturbation
. then the viscous damping rate of the unstable mode would be balanced with the shear rate in the viscous regime , @xmath49 .
this relation gives the most unstable wavenumber @xmath50 , and thus @xmath51 .
since the maximum growth rate of the mri is equal to the alfvn frequency of the fastest growing mode , it corresponds to @xmath52 , or @xmath53 .
this is why the normalized wavenumber and growth rate of the fastest growing mode are proportional to the @xmath43 in the presence of large viscous dissipation .
note that the @xmath4-dependence of the fastest growing mode is slightly different from that on the lundquist number ( @xmath22 ) in the presence of ohmic dissipation ( sano & miyama 1999 ) , where @xmath13 is the magnetic diffusivity .
based on the linear analysis , the resistivity can suppress the mri more efficiently compared to the viscosity . in the resistive regime
, the ohmic dissipation can suppress the mri when the dissipation time of the unstable mode is comparable to the alfvn time , @xmath54 , thus the most unstable wavelength is given by @xmath55 .
the growth rate of the fastest growing mode is given by @xmath56 .
then we can obtain the relations @xmath57 and @xmath58 .
it is stressed that the critical wavenumber @xmath59 in the resistive regime is also proportional to the lundquist number , @xmath60 .
unstable wavelength and growth rate of the mri for ideal mhd , resistive , and viscous cases are summarized in table [ table1 ] .
for elucidating the nonlinear features of the mri , viscous mhd equations are solved with a finite - differencing code which was developed by sano et al .
the hydrodynamic module of our scheme is based on the second - order godunov scheme ( van leer 1979 ) , which consists of lagrangian and remap steps .
the riemann solver is modified for accounting the effect of tangential magnetic fields .
the field evolution is calculated with the consistent moc - ct method ( clarke 1996 ) .
the energy equation is solved in the conservative form and the viscous terms are calculated in the lagrangian step .
the advantages of our scheme are its robustness for strong shocks and the satisfaction of the divergence - free constraint of magnetic fields ( evans & hawley 1988 ; stone & norman 1992 ) .
we use a local shearing box model ( hawley et al .
1995 ) . in this approximation ,
equations of viscous mhd are written in a local cartesian frame of reference corotating with the disk at the angular velocity @xmath12 corresponding to a fiducial radius @xmath61 .
then the coordinates are presented as @xmath62 , @xmath63 , and @xmath64 .
the fundamental equations are written in terms of these coordinates within a small region surrounding the fiducial radius , in @xmath65 , @xmath66 @xmath67 where @xmath68 \ ; , \label{eq12 } \\ \phi_{ij } & = & \xi ( \nabla\cdot { \boldmath v})^2 + \frac{\nu}{2 } \left ( \frac{\partial v_i}{\partial x_j } + \frac{\partial v_j}{\partial x_i } \right)^2 \;. \label{eq13 } \end{aligned}\ ] ] where @xmath69 is the specific internal energy , @xmath70 , and @xmath71 is the bulk viscosity . in the following ,
the effect of the bulk viscosity is neglected ( @xmath72 ) .
the term @xmath73 in the momentum equation is the tidal expansion of the effective potential .
assuming the ideal gas , the pressure is given by @xmath74 . the spatially uniform kinematic viscosity and a constant ratio of specific heats ( @xmath75 ) are considered .
since we focus on the local properties of the instability , we employ a numerical grid representing a small section of the disk interior for a local disk model . adopting the keplerian shear flow as unperturbed state , the azimuthal velocity is given by @xmath76 in the frame corotating with the velocity @xmath77 .
the initial field geometry is a weak uniform field in the vertical direction @xmath78 .
the radial force balance at the initial state is thus realized between the coriolis force and the tidal force .
two - dimensional calculation is performed in the radial - vertical plane with a volume bounded by @xmath79 , where @xmath80 is the scale height of the disk .
we use a uniform grid of @xmath81 zones .
a periodic boundary condition is applied in the vertical direction . for the radial boundary condition
, we adopt a sheared periodic boundary condition ( hawley et al .
1995 ) . in this model , the vertical component of gravity can be ignored . except for the shear velocity ,
the physical quantities are thus assumed to be spatially uniform ; @xmath82 and @xmath83 where @xmath84 and @xmath85 are constant values .
we choose normalizations with @xmath86 , @xmath87 , @xmath88 , and @xmath89 .
initial perturbations are introduced as spatially uncorrelated velocity and adiabatic pressure fluctuations .
these fluctuations have a zero mean value with a maximum amplitude of @xmath90 and @xmath91 .
our local disk model is characterized by non - dimensional parameters , which are the plasma beta of the initial field strength @xmath92 and the initial reynolds number @xmath4 . in this paper , we show the results focusing on the effects of the reynolds number . in what follows ,
we fix the initial field strength as @xmath93 in all our models . in the case
@xmath44 , the fastest growing wavelength of the local disk system @xmath94 can be defined from equation ( [ eq3 ] ) , @xmath95 where @xmath96 is the fastest growing wavelength expected by the linear analysis for viscous mhd case . to capture the most unstable mode in the computational domain , @xmath97
, the reynolds number must be @xmath98 .
we investigate the growth rate at the early linear phase in order to confirm the accuracy of our numerical scheme .
all the simulations begin with random perturbation of very small amplitude so that any growing modes should be well described by a linear analysis during the first few orbits of the evolution .
the time history of each mode is followed through a two - dimensional fourier decomposition carried out at frequent time intervals .
we define the fourier coefficient @xmath99 of the radial velocity @xmath100 as @xmath101 here we select the modes with @xmath102 because they are the most unstable .
the numerical growth rates in early linear phase are plotted over the analytic dispersion relation in figure [ fig1 ] .
the growth rate at @xmath103 , @xmath104 , @xmath105 , and @xmath106 are shown as representative cases , which are obviously reproducing the analytic results for all the cases .
thus our numerical scheme has ability to simulate correctly the unstable growth of perturbations excited by the mri in the presence of the viscous dissipation .
the efficiency of angular momentum transport is given by the @xmath107-@xmath108 component of the stress tensor , @xmath109 where @xmath110 and @xmath111 are maxwell and reynolds stresses , respectively .
this is related to the @xmath112 parameter of shakura & sunyaev ( 1973 ) by @xmath113 .
the maxwell stress is proportional to the magnetic energy and usually dominates over the reynolds stress in mri driven turbulence . to demonstrate the nonlinear features of the mri , the time - evolutions of volume - averaged reynolds and maxwell stresses , @xmath114 and @xmath115 , are depicted in figure [ fig3 ] for the cases with different reynolds numbers @xmath116 , @xmath117 , @xmath36 , and @xmath37 .
the single bracket indicates a volume average of physical quantities .
the horizontal axis is the time normalized by the rotation time @xmath118 .
the kinematic viscosity in those models is equivalent to the @xmath112 parameter of the size @xmath119 .
the linear growth rate decreases as the reynolds number decreases .
after the linear growth of the mri , a two - channel flow appears for the ideal mhd cases ( hawley & balbus 1992 ) .
the two - channel flow is an axisymmetric mri mode whose vertical wavelength fits the vertical box size .
this linearly unstable mode is also an exact solution of nonlinear mhd equations , so that the magnetic field can be amplified exponentially even at the nonlinear regime ( goodman & xu 1994 ) .
similar behavior is found in all the viscous models even though the reynolds number is much smaller than unity .
the magnetic energy continues growing and is not saturated even at the nonlinear regime .
the maxwell stress increases until the end of calculations for all the models .
the reynolds stress , on the other hand , approaches a constant value at the nonlinear stage .
the time evolution of @xmath120 , which is the sum of @xmath121 and @xmath122 , is shown in figure [ fig4]a .
the nonlinear behavior of the mri in viscous fluid is quite different from that in the models taking account of the magnetic diffusivity ( sano et al .
1998 , 2004 ; fleming et al .
2000 ; sano & inutsuka 2001 ) . for the purpose of comparison , figure
[ fig4]b shows @xmath120 in the resistive mhd runs with different lundquist numbers @xmath123 .
the initial conditions are the same as the viscous models except for the dissipation terms .
when the dissipation processes can be negligible ( @xmath124 and @xmath125 ) , the evolution is quite similar to that of ideal mhd .
the viscous dissipation can not damp mri driven turbulence in the nonlinear regime even though @xmath44 . in contrast , the mri saturates and mhd turbulence dies away at the nonlinear regime for the case with @xmath126 .
the total energy within the shearing box is defined as @xmath127 $ ] , where @xmath128 is the tidal expansion of the effective potential ( hawley et al . 1995 ) . using the evolution equations for viscous mhd system [ eqs .
( [ eq7])([eq13 ] ) ] , the time derivative of the total energy gives @xmath129 \nonumber \\ & \simeq & q\omega l_x \left ( \int_x { \rm d}a\ w_{xy } + \nu q \omega \int_x { \rm d}a\ \rho \right ) \ ; , \label{eq17}\end{aligned}\ ] ] where d@xmath130 is the surface element and the integral is taken over either of the radial boundaries .
we can derive the last term of above equation by assuming that the radial gradient of the perturbed azimuthal velocity is negligible , that is @xmath131 .
the energy injection rate through the radial boundary is proportional to the kinematic viscous stress as well as the turbulent stress @xmath132 at the boundary . using the volume - averaged values instead of the surface - averaged ones at the radial boundary ,
the volume average of the total energy changing rate @xmath133 is given by @xmath134 where @xmath135 is the energy injection rate caused by the turbulent stress ( sano & inutsuka 2001 ) , and @xmath136 is done by the kinematic viscous stress . the second term is always positive . when the turbulent stress is positive , the total energy of the system must increase .
the source of the input energy is the background shear motion . in realistic disk systems ,
positive stresses lead to inward mass accretion , bringing a loss of gravitational energy .
the gain in total energy in the shearing box represents this energy release .
the energy budget in our simulations can be satisfying the relation ( [ eq18 ] ) , because our numerical scheme solves the energy equation in terms of the total energy .
figure [ fig5 ] shows the time evolution of @xmath133 , @xmath137 and @xmath138 for the case with @xmath116 .
the vertical axis is given in the unit of @xmath139 , where @xmath140 is the initial thermal energy . during the early linear phase until about 15 orbits , the kinematic viscous stress takes a major role in the energy injection .
in contrast , the turbulent stress becomes predominant after the mri grows sufficiently .
the sum of these contributions is exactly identical to the energy gain of the system in our simulations .
the role of dissipation processes in the energy conversion is investigated in this subsection .
the injected energy @xmath141 should be balanced with the increase of the sum of @xmath142 , @xmath143 , and @xmath144 .
figure [ fig6]a depicts the time evolution of @xmath145 , @xmath146 , and @xmath147 for the case with @xmath116 .
the viscous heating rate @xmath148 is also shown in this figure , which is defined as @xmath149 where @xmath150 is the same definition that is used in the energy equation ( [ eq13 ] ) .
the sum of the components perfectly coincides with the energy gain of the system .
using the time- and volume - averaged values , which are indicated by double brackets , the gain rates of each energy component are @xmath151 , @xmath152 , and @xmath153 . here
the time average is taken at @xmath154 .
this indicates that a large portion of the injected energy is converted into the magnetic energy of the system .
because the system is almost incompressible , the dominant heating mechanism in our simulations should be the kinematic viscous heating .
however the thermal energy gain is much larger than the viscous heating rate ( see fig . [
the ratio of these components is evaluated as @xmath155 in the range @xmath154 .
this fraction becomes smaller and smaller for the cases with larger reynolds number .
the rest of the heating is caused by the numerical magnetic dissipation .
then our results suggest that the magnetic dissipation is the dominant mechanism of heating at the nonlinear phase of the mri and might play an essential role for the suppression of mri driven turbulence .
actually in the resistive mhd cases , the heating of the system is controlled by the joule heating almost completely ( sano & inutsuka 2001 ) .
the energy changing rates in a very resistive model with the lundquist number @xmath156 is demonstrated in figure [ fig6]b .
this is a typical case in which the channel flow is disrupted and mhd turbulence is damped at the nonlinear stage ( see fig .
[ fig4]a ) .
the ratio of the total energy gain and the thermal energy gain is @xmath157 in the range @xmath158 .
we find that the joule heating takes a major role in the thermal energy gain of the system , that is @xmath159 . in the resistive system ,
the magnetic energy amplified by the mri is transformed into the thermal energy via joule heating or magnetic reconnection throughout the nonlinear evolution .
finally , the time- and volume - averaged @xmath160 at the nonlinear stage is depicted as a function of the initial reynolds number @xmath4 and lundquist number @xmath123 in figure [ fig7 ] .
we take the average of @xmath161 over 5 orbits just after the time when the ratio @xmath162 begins to rise and the nonlinear evolution is started .
diamond - shape shows the results in the viscous fluid and cross - shape is that in the resistive one .
note that these are not the saturated values .
upward arrow over - plotted on the symbols denotes that the value is the lower limit , because @xmath163 is still increasing with time .
the downward arrow stands for decaying models and thus the stress is the upper limit
. the stress at the nonlinear stage are almost the same when the diffusion is weak ( @xmath124 or @xmath164 ) .
however , a huge difference can be seen in the highly diffusive regime . in the presence of the ohmic dissipation
, the stress rapidly decreases with decreasing @xmath123 .
for the models with the kinematic viscosity , on the other hand , it increases with the decrease of @xmath4 .
the origin of this difference between viscous and ohmic dissipative systems is discussed later in 4.1 .
the inverse correlation between @xmath165 and @xmath4 for the cases with large viscosity could be originated from stable growth of a channel flow .
this is because the large viscosity can suppress the growth of any other modes than the two - channel flow , and thus the channel mode can evolve up to highly nonlinear amplitude .
the viscosity may enhance the saturation amplitude of the mri .
however , our results are restricted in two - dimensional simulations .
the nonlinear evolution of the mri in three - dimension must be quite different , because the channel flow is known to be unstable to the nonaxisymmetric parasitic instability ( goodman & xu 1994 ) .
we are planning to perform three - dimensional study of the mri for verifying these nonlinear properties in the viscous accretion disks .
to give a physical explanation for the nonlinear behavior of the axisymmetric mri , we focus on the critical wavelength obtained from the linear theory in this section
. the diagrams of figures [ fig8]a and [ fig8]b indicate the critical and the fastest growing wavelengths of the mri as a function of the lundquist number @xmath123 and the reynolds number @xmath4 , respectively ( see also table [ table1 ] ) .
note that the vertical axes are normalized by @xmath166 in figure [ fig8]a and by @xmath167 in figure [ fig8]b .
shaded area denotes the linearly unstable regions for the mri . assuming fixed diffusivities , @xmath123 and @xmath4 increase as the instability grows because they are proportional to the squared alfvn velocity .
then the horizontal axis in figure [ fig8 ] can be regarded as the time direction in terms of the evolution of the mri .
first , we consider the resistive case shown by figure [ fig8]a . for the case of @xmath126 ,
the critical wavelength is described as @xmath168 . at the nonlinear stage of the two - dimensional mri
, mhd turbulence decays and it saturates only when @xmath126 ( see figs . [ fig4]b and [ fig7 ] ) .
this behavior can be interpreted schematically using the @xmath169-@xmath123 diagram ( sano & miyama 1999 ) . as the mri grows and amplifies the magnetic field , the critical wavelength shifts to the shorter length - scale .
then , many smaller scale fluctuations can become unstable .
those structures enhance the efficiency of ohmic dissipation in the turbulent state . in other words
, the system evolves toward a more dissipative state , and could be saturated at a critical point around @xmath170 , at which the critical wavelength reaches the shortest value and the ohmic dissipation is the most efficient .
in this way , mhd turbulence can decay if @xmath126 .
when @xmath125 , on the other hand , the critical wavelength is given by @xmath171 .
two - dimensional calculations of the mri suggests that the unstable growth can not saturate if @xmath125 . the nonlinear behavior in these cases can be explained by figure [ fig8]a analogously as follows : at the linear evolutionary stage , the critical wavelength in the radial direction becomes longer due to the exponential growth of the radial field component , while the vertical field grows slower than exponentially .
the wavevectors of the unstable modes thus become parallel to the vertical axis .
this channel flow mode is the exact solution of the nonlinear mhd equations ( goodman & xu 1994 ) .
as the field strength becomes larger , the critical wavelength shifts to the longer one .
the dissipation can be much less effective , and thus the channel solution continues to grow without saturation .
next , let us consider the viscous case shown in figure [ fig8]b . in the viscous fluid
, the critical wavelength is given by @xmath171 despite the size of the reynolds number ( see figure [ fig1 ] ) . even if @xmath172 is much smaller than unity , the critical wavelength thus shifts to larger scale as the instability grows .
then the system always evolves toward a less dissipative state and is not saturated .
this interpretation is consistent with our numerical results shown in figure [ fig7 ] .
these results indicates that the saturation process of the mri would be changed dramatically at the critical point at which the critical wavelength switches from the decreasing function of the field strength to the increasing one .
the differences in the nonlinear behavior of the mri between the viscous and resistive systems would be originated from whether the critical point exists or not . in this subsection
, we apply the discussion above to the doubly diffusive system that includes both the viscosity and resistivity .
the saturation behavior of the axisymmetric mri can be anticipated by the dependence of the critical wavelength on the field strength derived from the linear theory . with the same framework used in 2 , a local axisymmetric dispersion equation of the mri in the presence of both viscous and ohmic dissipations is given by @xmath173 where @xmath174 ( menou et al .
2006 ; masada et al . 2007 ;
lesur & longaretti 2007 ; pessah & chan 2008 ) .
this equation is , as expected , characterized by @xmath175 and @xmath123 .
here we focus on the system with a constant magnetic prandtl number ( @xmath176 ) .
figure [ fig9 ] demonstrates the critical wavelength of the mri as a function of @xmath123 for various values of @xmath18 obtained by solving the dispersion equation ( 18 ) .
the @xmath123-dependence of the critical wavelength varies with the size of @xmath18 .
when @xmath8 , the linear growth of the mri is independent of the magnetic prandtl number , and the critical wavelength is almost identical to the pure resistive case ( @xmath177 ) .
however , if the viscosity effects is added sufficiently , then the critical wavelength is enlarged by the suppression due to the viscosity in the middle range of @xmath123 . for the cases of @xmath7 , the critical wavelength around @xmath170
is given by @xmath178 ( pessah & chan 2008 ) .
although the critical wavelength shifts to longer length - scales as the magnetic prandtl number increases , the critical wavelength has a minimum value for all the cases .
this implies that the mri turbulence could be suppressed if the lundquist number is less than a critical value .
this diagram suggests that the critical lundquist number @xmath179 depends on @xmath18 . the critical lundquist number @xmath179 is plotted as a function of @xmath18 in figure [ fig10]a . in the regime of @xmath7 , it is proportional to the square root of the magnetic prandtl number , that is @xmath180 .
in contrast , it remains to be constant in the range @xmath8 . since the critical reynolds number is given by @xmath181
, we can also obtain the relation between @xmath18 and @xmath182 , and which is depicted in figure [ fig10]b .
nonlinear growth of the mri can be expected in the parameter region above this critical curve .
the magnetic prandtl number is proportional to @xmath183 in the regime of @xmath7 and @xmath184 when @xmath8 .
this curve is reminiscent of the critical curve for mhd turbulence sketched from nonlinear simulations of mri ( fromang et al .
the reynolds number @xmath4 in their models are at most a few tens in fromang et al .
( 2007 ) is different from @xmath4 in this paper .
the relation between these two is @xmath185 , where @xmath186 is the maxwell stress normalized by the ( initial ) pressure . ] , and the critical magnetic prandtl number is around unity .
thus our prediction is roughly consistent with nonlinear results even quantitatively .
note that the critical curve shown by figure [ fig10 ] is obtained by using only the features of the linear dispersion relation of mri .
this implies that the linear growth of the mri is very important even at the nonlinear saturated phase to sustain mhd turbulence .
the discussion in this paper is based only on two - dimensional simulations of the mri . for the understanding
the saturation mechanism of the mri , it is quite important to perform the systematic three - dimensional analysis of the mri in the presence of multiple diffusivities .
furthermore , the assumption of the local shearing box could affect the nonlinear evolution of the viscous mri .
the necessary ingredients of the unstable growth of the mri are the velocity shear and the magnetic field . in the numerical setting of the local shearing box ,
the velocity profile of the background shear flow can not disappear by the role of the kinematic viscosity , but is imposed by the boundary conditions .
it would be very interesting to use global disk models to investigate the mri in highly viscous disks .
these are our next tasks .
axisymmetric mri in viscous accretion disks is investigated by linear and nonlinear analyses .
a local shearing box threaded by a uniform vertical magnetic field is used for our nonlinear simulations .
the nonlinear results of the viscous mri are compared with the resistive case focusing on two non - dimensional parameters , the reynolds number for the mri @xmath4 and the lundquist number for the mri @xmath123 .
our main findings are summarized as follows .
\1 . in axisymmetric two - dimensional simulations
, the mri continues growing regardless of the size of the reynolds number .
when @xmath187 , the stress in its nonlinear stage is inversely correlated with @xmath4 , and thus can be larger than that in the ideal mhd run . in the highly resistive fluid ,
on the other hand , the growth of the mri is saturated and mhd turbulence dies away . when the lundquist number is less than unity , the saturated stress decreases dramatically with decreasing @xmath188 . \2 . at the nonlinear stage of the mri
, a large portion of the injected energy is converted to the magnetic energy in the viscous system ( @xmath189 ) .
the thermal energy gain is much larger than the viscous heating rate .
in contrast , for the case of the resistive system , the thermal energy is converted from the magnetic energy through the joule heating .
this difference in the energy dissipation efficiency may affect the saturation process of the mri .
nonlinear behavior of the mri in the single diffusive system can be understood with the help of the local dispersion relation .
the key characteristic is the dependence of the critical wavelength on the reynolds number @xmath4 and the lundquist number @xmath190 . applying this interpretation to the doubly diffusive system with both the viscous and ohmic dissipations , a condition for sustaining mri driven turbulence
is obtained as a function of @xmath4 and @xmath123 .
we thank neal turner for his careful reading of the manuscript .
we also thank jim stone , sebastien fromang , and shu - ichiro inutsuka for useful discussions . y.m
thanks ronald taam and kazunari shibata for helpful and encouraging comments on our paper .
a part of our simulations were carried out on vpp5000 at the national astronomical observatory of japan and sx8 at the institute of laser engineering , osaka university .
we thank the anonymous referee for useful comments .
arlt , r. , & rdiger , g. 2001 , , 374 , 1035 balbus , s. a. , & hawley , j. f. 1991 , , 376 , 214 balbus , s. a. , & hawley , j. f. 1998 , reviews of modern physics , 70 , 1 balbus , s. a. 2003 , , 41 , 555 balbus , s. a. , & henri , p. 2008
, , 674 , 408 brandenburg , a. , nordlund , a. , stein , r. f. , & torkelsson , u. 1995 , , 446 , 741 clarke , d. a. 1996 , , 457 , 291 evans , c. r. , & hawley , j. f. 1988 , , 332 , 659 fleming , t. p. , stone , j. m. , & hawley , j. f. 2000 , , 530 , 464 fromang , s. , & papaloizou , j. 2007 , , 476 , 1113 fromang , s. , papaloizou , j. , lesur , g. , & heinemann , t. 2007 , , 476 , 1123 goodman , j. , & xu , g. 1994 , , 432 , 213 hawley , j. f. , & balbus , s. a. 1992 , , 400 , 595 hawley , j. f. , gammie , c. f. , & balbus , s. a. 1995 , , 440 , 742 hawley , j. f. , gammie , c. f. , & balbus , s. a. 1996 , , 464 , 690 hawley , j. f. 2000 , , 528 , 462 king , a. r. , pringle , j. e. , & livio , m. 2007 , , 376 , 1740 lesur , g. , & longaretti , p .- y .
2007 , , 378 , 1471 lesur , g. , & ogilvie , g. i. 2008 , arxiv e - prints , 807 , arxiv:0807.1703 liu , w. , goodman , j. , & ji , h. 2006 , , 643 , 306 machida , m. , hayashi , m. r. , & matsumoto , r. 2000 , , 532 , l67 masada , y. , sano , t. , & shibata , k. 2007 , , 655 , 447 masada , y. , kawanaka , n. , sano , t. , & shibata , k. 2007 , , 663 , 437 matsumoto , r. , & tajima , t. 1995 , , 445 , 767 menou , k. , balbus , s. a. , & spruit , h. c. 2004 , , 607 , 564 nakano , t. 1984 , fundamentals of cosmic physics , 9 , 139 pessah , m. e. , chan , c .- k .
, & psaltis , d. 2007 , , 668 , l51 pessah , m. e. , & chan , c .- k . 2008 , arxiv e - prints , 801 , arxiv:0801.4570 sano , t. , inutsuka , s. , & miyama , s. m. 1998 , , 506 , l57 sano , t. , & miyama , s. m. 1999 , , 515 , 776 sano , t. , miyama , s. m. , umebayashi , t. , & nakano , t. 2000 , , 543 , 486 sano , t. , & inutsuka , s .- i . 2001 , , 561 , l179 sano , t. , & stone , j . m. 2002 , , 577 , 534 sano , t. , inutsuka , s. , turner , n. j. , & stone , j. m. 2004 , , 605 , 321 shakura , n. i. , & sunyaev , r. a. 1973 , , 24 , 337 silvers , l. j. 2007 , arxiv e - prints , 712 , arxiv:0712.4343 stone , j. m. , & norman , m. l. 1992 , , 80 , 791 stone , j. m. , hawley , j. f. , gammie , c. f. , & balbus , s. a. 1996 , , 463 , 656 umebayashi , t. , & nakano , t. 1988 , progress of theoretical physics supplement , 96 , 151 van leer , b. 1979 , journal of computational physics , 32 , 101 woosley , s. e. 1993 , , 405 , 273 ziegler , u. , & rdiger , g. 2001 , , 378 , 668 | axisymmetric magnetorotational instability ( mri ) in viscous accretion disks is investigated by linear analysis and two - dimensional nonlinear simulations .
the linear growth of the viscous mri is characterized by the reynolds number defined as @xmath0 , where @xmath1 is the alfvn velocity , @xmath2 is the kinematic viscosity , and @xmath3 is the angular velocity of the disk . although the linear growth rate is suppressed considerably as the reynolds number decreases , the nonlinear behavior is found to be almost independent of @xmath4 . at the nonlinear evolutionary stage , a two - channel flow continues growing and the maxwell stress increases until the end of calculations even though the reynolds number is much smaller than unity .
a large portion of the injected energy to the system is converted to the magnetic energy .
the gain rate of the thermal energy , on the other hand , is found to be much larger than the viscous heating rate .
nonlinear behavior of the mri in the viscous regime and its difference from that in the highly resistive regime can be explained schematically by using the characteristics of the linear dispersion relation . applying our results to the case with both the viscosity and resistivity
, it is anticipated that the critical value of the lundquist number @xmath5 for active turbulence depends on the magnetic prandtl number @xmath6 in the regime of @xmath7 and remains constant when @xmath8 , where @xmath9 and @xmath10 is the magnetic diffusivity . |
in the real - world applications such as e - commerce and online ad , a user s activity is not frequent , and item properties change dynamically over time . in a context of item recommendation ,
such scenario is referred to as _ persistent cold - start_. for instance , booking.com @xcite shows an example of users rare activity and mixed personas ( _ * user - side * persistent cold - start _ ) , and rakuten gora @xcite demonstrates price fluctuation and short life - span of packaged items ( _ * item - side * persistent cold - start _ ) .
most importantly , classical recommendation techniques have some drawbacks under the persistent cold - start setting .
in fact , matrix factorization ( mf ) @xcite is one of the most typical and promising techniques , but mf only holds latent vectors for every user / item ids ; there is no way to make meaningful recommendation under an unforeseen condition .
by contrast , context - aware recommender systems have been recently studied in order to profile more essential users preferences with auxiliary features . in particular , factorization machines ( fms ) @xcite are alternative effective factorization models which enable us to make context - aware recommendation with flexible feature representation . since the persistent cold - start problem commonly occurs in real applications , high feasibility of fms
is attractive compared to more specific methods developed by industrial researchers .
however , the captured context by fms is still static , and thus fm is incomplete in terms of robustness against persistently cold - starting data . in order to adjust the model parameters according to variation of context , this paper extends fms into online algorithms .
it should be noticed that the persistent cold - start problem is closely related to _ concept drift _
, a phenomenon of `` the relation between the input data and the target variable changes over time '' @xcite in data streams . on the user - side ,
since users interests may be changed , systems must recommend different items even for the same user .
meanwhile , due to the instability of item trends and properties , one item can be preferred by totally different users at a different point in time .
past studies proved that online algorithms are effective to the unpredictable phenomena @xcite , so the author assumes that incremental update of fms yields better accuracy compared to the static recommenders . in practice , online recommender systems behave as illustrated in fig . [
fig : stream - recommender ] .
list to a user , ( 2 ) interact with an item ( e.g. click , buy , rate ) , and ( 3 ) update parameters based on the interaction . ] more concretely , the author generalizes prior work in incremental mf ( imf ) @xcite , an online extension of a mf - based item recommender , to incremental fms ( ifms ) . here
, the framework specially equips the following properties : * * positive - only feedback * : model update is fast thanks to an one - pass online learning scheme . * * incremental adaptive regularization * : regularization parameters are automatically adjusted on - the - fly .
this section introduces the different factorization techniques which compose the proposed framework .
vinagre et al .
@xcite proposed an efficient imf algorithm for item recommendation , which is achieved by solving mf with a unique target value @xmath0 over the stochastic gradient descent ( sgd ) optimization .
moreover , they evaluated the method in a _ test - then - learn _ scheme as outlined in alg .
[ alg : prequential ] . for each pair of a user @xmath1 and an item @xmath2 ,
evaluation is first launched before updating the parameters .
data stream or finite set of positive events @xmath3 , size of recommendation list @xmath4 , window size @xmath5 @xmath6 @xmath7 if @xmath8 , otherwise @xmath9 @xmath10 avg .
recall@@xmath4 for latest @xmath5 samples more specifically , imf incrementally factorizes a binary matrix @xmath11 into @xmath12 and @xmath13 as : * learn @xmath14 and @xmath15 as the standard mf * score items by @xmath16 , and + recommend @xmath4 closest items to @xmath17 * @xmath18 + @xmath19 where @xmath20 are respectively a user , item latent vector .
the vectors are updated with a regularization parameter @xmath21 and a learning rate @xmath22 .
fms @xcite have been recently developed as a general predictor .
for an input vector @xmath23 , let us first imagine a linear model parameterized by @xmath24@xmath25 .
in addition , by incorporating interactions of the @xmath26 input variables , the linear model is extended to fms as : @xmath27 where @xmath28 is a rank-@xmath29 matrix which has @xmath30 .
arbitrary feature representation @xmath31 ( e.g. concatenation of one - hot vectors for several categorical variables ) work well with fms , and mf is actually a subset of the predictor .
at the beginning , the general ifm is proposed in sec . [
sec : fms - model ] and [ sec : fms - reg ] .
next , sec .
[ sec : ifms - pos ] optimizes it for positive - only - feedback - based online item recommendation .
this paper focuses on running fms in an incremental fashion .
generally , learning fm requires a set of parameters @xmath32 and a loss function @xmath33 , and the parameters can be optimized by sgd .
specifically , for a set of samples @xmath34 , the parameters of fm are updated as alg .
[ alg : fms - model ] . for simplicity
, @xmath33 is written as @xmath35 .
@xmath34 , learning rate @xmath22 , regularization parameters @xmath36 @xmath37 @xmath38 @xmath39 notice that of imf in sec .
[ sec : imf ] is sgd update for single sample , so the basic idea of this paper is that we replace the step with in alg .
[ alg : fms - model ] . as a result , ifm which can be evaluated in the _ test - then - learn _ scheme is derived without loss of generality .
rendle @xcite proposed an adaptive regularization scheme for fms .
as shown in alg .
[ alg : fms - reg ] , the technique adjusts the regularization parameters by using a sample @xmath40 in a validation set @xmath41 , an extra set of samples which is different from @xmath34 . using @xmath40
sampled from @xmath41 @xmath42 @xmath43 @xmath44 normally , alg .
[ alg : fms - reg ] is launched after in alg .
[ alg : fms - model ] .
however , when we consider incremental adaptive regularization , there is a difficulty that the validation set @xmath41 will be gradually outdated in a streaming environment . a key idea to conquer
the problem is that a newly observed sample @xmath45 is handled as a _ pseudo _
validation sample , so the regularization parameters are adjusted before updating @xmath46 as demonstrated in fig . [
fig : adapt - then - learn ] .
this section considers a particular combination of an output @xmath47 and a loss function @xmath35 to utilize ifms for online item recommendation . as the author explained in sec .
[ sec : imf ] , imf actually solves mf with @xmath0 .
similarly to the approach , let us again consider the unique output for a set of positive events @xmath3 . as a consequence , for a sample @xmath48 , our loss function is defined as : @xmath49 .
eventually , for arbitrary design of a feature vector @xmath31 , the proposed ifm - based item recommender which is feasible in a streaming environment can be described in the _ test - then - learn _ framework as : * learn @xmath50 , @xmath51 and @xmath52 as the standard fms * predict @xmath53 for every items , + and recommend @xmath4 closest items to @xmath17 * @xmath54 with @xmath55 importantly , since the number of users and items on real - world online applications is not constant , our systems must incorporate new users and items into a current model somehow . for example , imf handles a new user ( item ) as an additional row of @xmath14 ( @xmath15 ) ( i.e. @xmath56 for a new user , @xmath57 for a new item obtained from gaussian ) .
hence , ifms also take the simple approach that a zero and random vector are respectively inserted into @xmath51 and @xmath52 as the initial parameters of new features .
[ fig : new - feature ] depicts detection and insertion of new features in a stream of input vectors .
it is notable that , beyond new users and items , adding new contextual variables is also possible in the middle of data streams .
is increased by new users , items and/or contexts , initial values fill the corresponding parameters . ] in terms of computational complexity , ifms compute the interaction term @xmath58 in @xmath59 by letting the number of nonzero elements in @xmath31 be @xmath60 .
in fact , this complexity is efficient enough due to sparsity of @xmath31 , but running time will be relatively long compared to the single vector updating of imf . therefore , there is trade - off between running time and context - awareness in practice .
in the experiments , the _ test - then - learn _ procedure described in sec .
[ sec : imf ] is launched for time - stamped unstable datasets . as shown in fig .
[ fig : sample - split ] , the samples are separated similarly to what matuszyk et al . @xcite did . at the beginning
, a batch train / test step is executed for the first 20% training and following 10% validation samples .
next , the 10% samples are just used for one - pass model updating , and the incremental evaluation step is finally performed for the remaining 70% .
[ fig : sample - split ] additionally , the 70% samples are also evaluated by mean percentile rank ( mpr ) in order to assess an ordered list of items obtained from .
the metric is calculated based on percentile rank of the correct items in the ordered lists ; that is , @xmath61 indicates the best result that a recommender always gives the highest rank to a correct item , and @xmath62 is opposite .
in contrast to recall@@xmath4/@xmath5 which evaluates top-@xmath4 items , mpr can measure users overall satisfaction from all items .
following methods were employed as the competitors : * * static mf * : the traditional mf , * * imf * : a fast extension of mf introduced in sec .
[ sec : imf ] , * * static fms * : the parameters are not updated in the incremental stage similarly to * static mf*. the author implemented all of the methods in python 3.5.1 , and the code was run in a typical personal computer with the 2.7 ghz intel core i7 cpu and 4 gb ram .
datasets used in the experiments were binarized version of movielens 100k ( ml100k ) and synthetic click data ; the former is a real - world example of user - side volatility , and the latter shows item - side instability .
table [ tab : data ] summarizes statistics of the data .
each pair of a method and a dataset was tested five times with different initial parameters .
.statistics of the datasets . [ cols= " <
, > , > , > , > , > " , ] [ tab : result ] * ml100k . *
since we focus on item recommendation , ml100k was binarized by extracting 5-starred rating events .
an input vector of ifms was designed as : @xmath63{\strut user id}}},}_{1/|u| } \underbrace{\textrm{\colorbox{myorange!50}{\makebox[6em]{\strut demographics}}},}_{3/23 } \underbrace{\textrm{\colorbox{mypink!50}{\makebox[4em]{\strut movie id}}},}_{1/|i| } \\ & & \underbrace{\textrm{\colorbox{mypink!50}{\makebox[3em]{\strut genre}}},}_{18/18 } \underbrace{\textrm{\colorbox{mypink!50}{\makebox[7em]{\strut last rated genre}}},}_{18/18 } \underbrace{\textrm{\colorbox{mypurple!50}{\makebox[2em]{\strut day}}},}_{1/7 } \underbrace{\textrm{\colorbox{mypurple!50}{\makebox[6em]{\strut last rated day}}}}_{1/7 } \ ) .
\label{eq : vector_ml}\end{aligned}\ ] ] `` demographics '' includes user s occupation ( 1/21 ) , sex ( 1/1 ) and age ( 1/1 ) , and `` day '' is day of week in the timestamps . note that the numbers below the underbraces indicate \{_max .
# of nonzero dimensions_}/\ { _ # of total dimensions_}. most users on ml100k report positive events just for the initial rating activity , and they do not rate any more . even if some users continuously rate movies , intuition tells us that their interests change over time .
thus , ml100k is a real - world example of user - side persistent cold - start .
[ fig : result_ml100k ] shows the best recall behaviors on ml100k obtained from the five _ test - then - learn _ trials .
ml100k is rich in both user and item features , and we also considered time - related contexts . consequently ,
static fms and ifms respectively outperformed static mf and imf , especially in terms of mpr .
in addition , it is clear that online recommenders ( imf and ifms ) worked effectively compared to the static counterparts .
[ fig : result_ml100k ] * synthetic . * synthetic click data was generated as an example of item - side persistent cold - start , based on a rule - based procedure demonstrated in @xcite .
in particular , our generator first produced 0.5 million impressions of five ad variants , and additional half million impressions were also generated after updating a rule for the most popular ad . as a result , from the one million impressions ,
3,570 synthetic clicks were observed as positive events with erratic trend .
here , @xmath31 for the synthetic data was : @xmath64{\strut age}}},}_{1/1 } \underbrace{\textrm{\colorbox{myorange!50}{\makebox[2em]{\strut sex}}},}_{1/1 } \underbrace{\textrm{\colorbox{myorange!50}{\makebox[5em]{\strut geo ( state)}}},}_{1/50 } \underbrace{\textrm{\colorbox{mypink!50}{\makebox[3em]{\strut ad id}}},}_{1/|i| } \underbrace{\textrm{\colorbox{mypink!50}{\makebox[4em]{\strut category}}}}_{1/3 } \ ) .
\label{eq : vector_ad}\end{aligned}\ ] ] the best result on the synthetic data is illustrated in fig .
[ fig : result_synthetic ] .
all methods were easily fit to the batch training samples due to the simplicity of data , but the difference can be observed after the most popular ad was changed .
while recall of the static methods declined significantly , imf and ifms evidently adapted to the variation as expected .
moreover , mpr and recall of ifms were even better than imf .
it should be noted that , even though static mf and fms demonstrated the similar recall behavior , fms showed reasonably lower mpr compared to mf .
finally , table [ tab : result ] summarizes the results of 5 trials .
the positive effect of context - awareness and online model updating was validated in terms of accuracy . on the other hand , running time of fms was more than 100 times slower than mf .
ifms updated the parameters in a millisecond range , and recommendation for a user was done at least 30 milliseconds .
ifm thus seems to satisfy practical time requirements for now , but higher - dimensional and denser input vectors may lead worse results in the future . as the author mentioned in sec .
[ sec : ifms - pos ] , the fact proved trade - off between context - awareness and efficiency . overall , the proposed online item recommender based on ifms worked well as generalization of imf . in case that an organization develops production recommender systems
, our highly feasible framework can be an easy - to - implement baseline .
furthermore , availability of many third - party libraries ( e.g. @xcite ) is an important advantage of fms .
this paper has proposed an ifm - based context - aware online item recommender .
experimental results have demonstrated not only feasibility and effectiveness of the technique but also a new challenge in computational efficiency . | real - world item recommenders commonly suffer from a persistent cold - start problem which is caused by dynamically changing users and items . in order to overcome the problem ,
several context - aware recommendation techniques have been recently proposed . in terms of both feasibility and performance , factorization machine ( fm )
is one of the most promising methods as generalization of the conventional matrix factorization techniques .
however , since online algorithms are suitable for dynamic data , the static fms are still inadequate .
thus , this paper proposes incremental fms ( ifms ) , a general online factorization framework , and specially extends ifms into an online item recommender .
the proposed framework can be a promising baseline for further development of the production recommender systems .
evaluation is done empirically both on synthetic and real - world unstable datasets .
< ccs2012 > < concept > < concept_id>10002951.10003317.10003347.10003350</concept_id > < concept_desc > information systems recommender systems</concept_desc > < concept_significance>500</concept_significance >
< /concept > < concept > < concept_id>10002951.10002952.10002953.10010820.10003208</concept_id > < concept_desc > information systems data streams</concept_desc > < concept_significance>300</concept_significance >
< /concept > < concept >
< concept_id>10010147.10010257.10010293.10010309</concept_id > < concept_desc > computing methodologies factorization methods</concept_desc > < concept_significance>300</concept_significance > < /concept >
< /ccs2012 > |
the ejection of particles by radiation pressure has been considered many times as a possible mechanism for producing relativistic outflows from very luminous radiation sources , such as active galactic nuclei ( agns ) or galactic compact objects . in the case of an optically thin plasma , the particle dynamics
is dominated by the intense photon field and a flow results from the work done by the radiation force on the material .
in most previous studies it was assumed that the particles scatter photons with a cross section independent of frequency and front - to - back symmetric along the incident photon direction , such as true in the thomson limit ( noerdlinger 1974 ; odell 1981 ; phinney 1982 ; kovner 1984 ) .
direct momentum input to the fluid at rest results when outward - directed photons are removed from the anisotropic radiation field , and then scattered away .
no net momentum is carried off by the scattered photons and the material is accelerated radially .
a particle moving outward near the source , however , encounters many photons nearly at right angles to its motion , more so as @xmath2 because of aberration effects , and suffers collisional drag by them .
only mildly relativistic flows can be produced by an eddington - limited extended source , the radiation drag actually preventing acceleration to very high particle energies ( noerdlinger 1974 ) . in the context of agns , the role of compton drag in the dynamics of relativistic jets has been the subject of many studies ( e.g. abramowicz & piran 1980 ; sikora & wilson 1981 ; phinney 1987 ; melia & knigl 1989 ; sikora 1996 ) . in the past few years
, the _ compton gamma - ray observatory _ mission has revealed the existence of some classes of objects , such as the egret ( fichtel 1994 ) , osse ( mcnaron - brown 1995 ) , and comptel ( bloemen 1995 ) blazars , together with some @xmath0-ray pulsars ( ulmer 1994 ) , where the hard x - ray and @xmath0-ray fluxes are a significant fraction or completely dominate the overall radiation energy budget .
this is similar to what sometimes observed by the sigma experiment onboard of _ granat _ in the case of galactic black hole candidates ( mandrou 1994 ) .
the large compactnesses inferred from these data naturally lead to theoretical models in which compton scattering in the klein - nishina ( kn ) limit ( as opposed to thomson scattering ) , together with photon - photon pair production may play a role in determining the thermal state and dynamics of the source .
such process may also be relevant for studies of the effects of gamma - ray bursts on their gaseous environment ( their `` afterglows '' ) .
in this paper we study the dynamics of a _ cold _ plasma embedded in the hard " photon field of an external source .
we adopt a test - particle approach , which is valid only insofar gas pressure gradients can be neglected compared to the radiative force and the material is optically thin .
we consider the model problem of a fully ionized plasma where gravity is overwhelmed by radiation pressure and the distribution of particle momenta remains one - dimensional .
we ignore stimulated scattering , photon absorption , and photon - electron pair production .
free - free absorption is unimportant compared to thomson scattering if @xmath3 ( @xmath4 ) , where @xmath5 is the electron density .
photo - electron pair production , @xmath6 ( threshold @xmath7 , where @xmath8 is the photon energy in the electron rest - frame ) is a higher order process in the fine - structure constant , important only at large energies . ( the pair production cross section is @xmath9 of the kn scattering cross section at @xmath10 , e.g. svensson 1982 . ) in an accompanying paper ( thompson & madau 1999 , hereafter paper ii ) we show , however , how pair creation induced by collisions between hard photons and soft side - scattered photon may increase the acceleration rate by decreasing the inertia per particle , and by increasing the mean momentum deposited per scattering . the plan of this paper is as follows . in
2 we discuss the basic theory of radiative acceleration . in
3 we present sample solutions for extended and impulsive radiation sources in the thomson regime .
compton drag by side - scattered radiation is shown to be more effective in limiting the bulk motion than is the finite inertia of the accelerating medium . the random energy imparted to the plasma by @xmath0-rays
can be converted into bulk motion if the hot particles execute many larmor orbits before cooling .
the equation of motion in the kn limit is derived and numerically integrated in 4 for a monoenergetic spectrum , under the assumption that the outflowing particles maintain a one - dimensional distribution of momenta , i.e. that the plasma remains ` cold ' even in the presence of recoil .
blumenthal ( 1974 ) used the kn cross section to calculate the mean force due to compton scattering on electrons with arbitrary velocity .
an approximate solution to the problem of the radiative deceleration of a relativistic jet in the kn regime has been recently given by luo & protheroe ( 1999 ) .
consider an optically thin plasma exposed to a photon source .
the incident and scattered photon momenta ( in units of @xmath11 ) are denoted by @xmath12 and @xmath13 in the ( unprimed ) lab frame .
let @xmath14 be the specific intensity of the incident radiation in the direction @xmath15 , @xmath16 . in the simplest case of a leptonic plasma composed entirely of electrons and positrons ,
the bulk momentum is obtained directly from the normalized phase - space density @xmath17 , @xmath18 here , @xmath19 as usual .
this equation is easily generalized to include a hadronic component .
the mean rate of momentum transfer per particle is then given by @xmath20f({\bf p})d^3p .
\label{eq : eqp}\ ] ] here , @xmath21 is the mean mass per scattering charge , the angular integral is over the solid angle subtended by the source , and @xmath22 is the invariant differential cross section .
when the number of scattering charges is dominated by pairs , we have assumed implicitly a magnetic field strong enough to couple the pairs with the hadronic component of the plasma . the minimum flux density needed to enforce that coupling is estimated in [ seccompton ] .
one must then describe the accelerating particles as a fluid , with @xmath23 in a cold plasma composed of hydrogen ( of density @xmath24 ) and @xmath25 pairs ( of density @xmath26 ) , @xmath27 this expression neglects the inertia of the neutralizing electron component , but not of the pairs . the klein - nishina cross section for scattering of unpolarized radiation is most simply expressed in the ( primed ) electron rest - frame as @xmath28\sin \chi'd\chi ' d\phi ' , \label{eq : eqkn}\ ] ] e.g. rybicki & lightman ( 1979 ) .
the energy of the scattered photon is @xmath29 in this notation , the scattering takes place through a polar angle @xmath30 and an azimuthal angle @xmath31 .
measuring the angles @xmath32 and @xmath33 of the incident and scattered photon with respect to the electron velocity @xmath34 , we can write the following kinematic relations @xmath35 since the scattering cross section is independent of the azimuth @xmath31 , we must integrate equation ( [ eq : eqp ] ) over two angles , @xmath30 and @xmath36 .
the evaluation of the integral in equation ( [ eq : eqp ] ) will be made by assuming that the radiation field is symmetric around the @xmath37-axis , and the electrons ( and positrons ) will maintain a one - dimensional distribution of momenta along @xmath37 , @xmath38 .
note that multiple scatterings can be safely neglected , as the singly scattered photons are beamed into a cone of half - angle @xmath39 along the electron direction of motion .
further scatterings are therefore suppressed by the @xmath40 factor in the momentum transfer equation .
single side - scattered photons will , however , be absorbed via @xmath41 if the compactness of the @xmath0-ray source is high at @xmath42 .
we defer a detailed discussion of the effects of pair creation to paper ii .
to provide some physical insight and a framework to interpret our numerical results in the relativistic limit , we shall first discuss the acceleration of a test charge in the thomson scattering approximation , @xmath43 . in this regime , @xmath44 by equation ( [ eq : eqxp ] ) .
thus , in the electron rest - frame , photons of unchanged energies are scattered into an angular distribution that is forward - backward symmetric along the incident photon direction .
a particle accelerated by an external photon source will , in general , follow a curved trajectory .
the problem simplifies when the source is cylindrically symmetric . from equation ( [ eq : eqp ] ) , a parcel of matter moving along the axis of symmetry with speed @xmath45 is accelerated at the rate @xmath46.\label{eq : eqt}\ ] ] the first term in brackets represents the collimated flux of redshifted photons which accelerate particles radially , while the second term represents the dragging effect of the isotropic component of the radiation field . , where @xmath47 is the radiative flux at radius @xmath48 .
] the right - hand side of equation ( [ eq : eqt ] ) vanishes for @xmath49 .
particles are accelerated away from the source as long as @xmath50 .
if @xmath51 , the force reverses , being now directed inward .
this saturation of the lorentz factor is due to the aberration into the forward hemisphere of blueshifted photons , as seen in the electron rest - frame .
a particle will follow the @xmath52 equilibrium trajectory ( zero - inertia limit ) until , as it gets farther from the source , its acceleration is effectively limited by the large doppler shift to the red and by the @xmath53 decline of the radiation flux .
consider a sphere of radius @xmath54 and uniform brightness @xmath55 .
close to the extended source , the equilibrium lorentz factor is @xmath56 no matter how high the intensity @xmath55 , or whether the inertia is dominated by hadrons or leptons . in the point - source limit ( @xmath57 )
the photons stream radially . for a steady flow ,
equation ( [ eq : eqt ] ) reduces to @xmath58 the parameter @xmath59 is the dimensionless compactness , rescaled by the inertia per scattering charge , @xmath60 .
it may be expressed as @xmath61 for a source of mass @xmath62 , where @xmath63 is the eddington luminosity and @xmath64 the schwarzschild radius . for a motion starting at radius @xmath48 ,
the lorentz factor attained at infinity can be obtained by solving the algebraic equation @xmath65 which , in the limit @xmath66 , yields @xmath67^{1/3}=$ ] @xmath68^{1/3}$ ] .
effective acceleration of an @xmath69 plasma requires a compactness that is larger by a factor @xmath70 than for a pure pair plasma , as a conseguence of the greater inertia per unit cross section . matching this solution with expression ( [ eq : gameq ] )
yields the asymptotic bulk lorentz factor @xmath71 ( noerdlinger 1974 ) .
the cross - over radius beyond which the non - radial component of the radiation field causes negligible drag is @xmath72 .
larger values of @xmath73 can only be obtained if the particles are injected with relativistic bulk velocities at @xmath74 .
bulk motion starting with @xmath75 at a distance @xmath76 from the source is quickly decelerated to @xmath77 , and the excess kinetic energy converted into a collimated beam of upscattered photons .
infinite , keplerian accretion disks around black holes provide an example of non - uniform radiating sources where compton drag is most severe .
the surface flux distribution as a function of equatorial radius r is given by ( shakura and sunyaev 1973 ) @xmath78 , \label{eq : eqflux}\ ] ] where @xmath62 is now the black hole mass , and @xmath79 is the accretion rate .
the peak flux , @xmath80 @xmath81 , is reached at @xmath82 .
one can understand the essential features of the accretion disk model by approximating the flux distribution as @xmath83 in this case the equilibrium lorentz factor increases only as @xmath84 ( compared to a linear increase for a spherical source ) . in the point - source limit , one derives @xmath85 , where @xmath59 is now the nominal disk compactness , @xmath86 @xmath87 .
again , by matching the two limiting solutions , the asymptotic bulk lorentz factor is found to scale as @xmath88 ( kovner 1984 ; phinney 1987 ) .
the exact solutions for a spherical uniform source and a keplerian accretion disk are shown in figure 1 .
the latter is a much less efficient accelerator than the former because of the large photon drag associated with a radiation source extending to infinity .
the radiative force acting on an external medium induces a flow that is strongly time - dependent .
one important effect of acceleration to bulk lorentz factor @xmath1 is the bunching of the accelerated material into a shell of thickness @xmath89 .
let us now solve for the density of the accelerated material .
the equation of continuity of the hadronic component is @xmath90 material initially at radius @xmath91 reaches a radius @xmath92 at time @xmath93 ( speed @xmath94 ) . here , @xmath95 .
the total duration of the acceleration is @xmath96 eliminating @xmath91 from these two equations and taking @xmath97 at constant time @xmath93 , one derives @xmath98 , and we have @xmath99 this integrates to @xmath100 where @xmath101 is the proton density in the undisturbed ambient medium .
there is a corresponding amplification of the non - radial component of a magnetic field entrained in the flow , @xmath102 gamma - ray sources such as blazars and gamma - ray bursts are impulsive .
it is instructive , therefore , to consider a photon source that maintains a constant luminosity for a time @xmath103 .
this source can be visualized as a uniform shell of radius @xmath104 , thickness @xmath105 , and total flux @xmath106 .
we focus here on the effects of a radiation pulse and take the photons to move radially before scattering .
the compactness within the shell can be expressed directly in terms of the flux of @xmath0-rays , @xmath107 and thence in terms of a characteristic photon optical depth , multiplied by the frequency @xmath108 at which the spectrum peaks . ]
@xmath109 such that @xmath110 .
material overtaken by the photon shell at radius @xmath48 , and accelerated from rest , develops a large bulk lorentz factor when @xmath111 . integrating equation ( [ eq : eqps ] ) in this limit yields @xmath112 at a radius @xmath113 . as the difference between the particle speed and the speed of light decreases as @xmath114 , the accelerated matter will surf the photon shell over an extended range of radius , @xmath115 .
the maximum lorentz factor is therefore @xmath116 when the shell is thin enough that the acceleration length is much less than the radius , @xmath117 ( see figure 2 ) .
only a limited amount of material can be accelerated to this value . equating the kinetic energy within a volume @xmath118 with the energy of
the photon shell implies a very high pre - existing optical depth to compton scattering , @xmath119 . since
our calculation of @xmath120 is valid only for small @xmath121 , the ambient density at radius @xmath48 is bounded above by @xmath122 note the strong dependence on @xmath21 .
we will show is paper ii that , even if the ambient medium is composed of a hadronic plasma initially ( as is expected in most cosmological gamma - ray burst models ) , the mean mass per scattering charge will be rapidly reduced to @xmath123 by pair creation .
sidescattered photons will provide additional drag that reduces @xmath1 below the value given in equation ( [ eq : gams ] ) .
as the photon shell propagates ahead of an accelerating parcel of matter , it generates an isotropic radiation field of photon density @xmath124 where @xmath125 is the density of scattering charges in the ambient medium at rest , and @xmath126 is the lengthscale over which isotropic scattering occurs ( before the scattering medium gets accelerated to relativistic velocities ) .
the net effect is to introduce a second , negative term into equation ( [ eq : eqps ] ) , which becomes @xmath127 at large @xmath1 .
acceleration up to lorentz factor @xmath120 is then possible only if the ambient density is lower than @xmath128 this bound is stronger by a factor @xmath129 than equation ( [ eq : rhom ] ) : _ in the thomson scattering regime , the bulk lorentz factor is actually limited by side - scattered radiation_. this conclusion changes when the photon source is very hard , because side - scattered photons are immediately consumed by pair creation ( paper ii ) .
a compact source of @xmath0-rays always tends , in this sense , to maintain a radial photon distribution .
it is well known that a hot ( relativistic ) plasma in an anisotropic radiation field feels a much larger radiation pressure than a cold ( non - relativistic ) gas of the same inertial mass ( odell 1981 ) , and tends to drive itself away from the radiation source with momentum derived largely from the anisotropic loss of its own internal energy .
this `` compton rocket '' is , however , ineffective at generating bulk relativistic motion from internal random kinetic energy , as it is always accompanied by catastrophic compton cooling ( phinney 1982 ) . nonetheless , by integrating the momentum equation in the limiting case of a cold plasma we are actually underestimating its bulk lorentz factor
: the gas will be continuously heated by the hard radiation field itself , hot particles will radiate most of their energy in the direction of the radiation source , and part of the energy of relativistic random motion will be converted into bulk motion .
we term this supplementary source of momentum the ` compton afterburn ' .
the afterburn can be largely neglected in the case of pure kn scattering without pair creation ( 4.1 ) .
( in paper ii we will consider a situation in which relativistic @xmath130 pairs are injected into an anisotropic photon beam by collisions between side - scattered photons and the main beam thereby inducing a net force on the plasma as they compton cool . ) nonetheless , we discuss it here for completeness .
the afterburn is strongest when the radiative force on the relativistic , cooling particles is coupled to the rest of the fluid through a background magnetic field . above a critical flux density which we estimate , this allows the colder particles to dominate the inertia of the flow , and the momentum gained by compton scattering to be more effectively absorbed rather than being radiated away .
the afterburn is easiest to analyze when the period of a cyclotron orbit is much shorter than the compton cooling time .
then the linear momentum imparted to one charge is effectively shared with the others before the charge loses its excess energy in the bulk frame .
we also assume that the acceleration time is long compared with the compton cooling time , so that the proportion of energetic particles is small .
this is the case when the energetic charges are continually regenerated , e.g. , by pair creation .
finally , we assume that the magnetic field is non - radial ( as is appropriate to the rest frame of a plasma that has been accelerated to relativistic speed ) .
we divide the velocity of the charge into components @xmath131 and @xmath132 perpendicular and parallel to @xmath133 , and denote by @xmath134 the angle between the cyclotron motion and the radial direction .
the charge is immersed in a collimated photon beam of energy flux @xmath135 in the boosted frame of the bulk flow .
it feels a radial force ( odell 1981 ) @xmath136.\ ] ] the net radial momentum imparted to the cooling charge is obtained by averaging over a cyclotron orbit , @xmath137 , @xmath138 at the same time , the net energy lost to compton cooling is @xmath139 when the energetic charges are created with @xmath140 ( e.g. , pair creation between a hard radial photon and a soft side - scattered photon ) , the cooling rate is @xmath141 and the radial momentum gained as the particle cools is @xmath142 to leading order in @xmath143 , @xmath144 in this case , the source of energy for the charge is external to the accelerating medium .
the charge carries an excess lab - frame momentum @xmath145 , which is supplemented by ( [ eq : prad ] ) to yield a total @xmath146 ( in the limit @xmath147 ) .
these expressions change slightly when the energetic charges are distributed isotropically , and reduce to those found by odell ( 1981 ) .
the rate of compton cooling is @xmath148 @xmath149 @xmath150 , as usual , and @xmath151 the _ net _ radial momentum acquired is @xmath152 only @xmath153 of the value obtained for charges that are created moving parallel to the photon beam .
these calculations assume that a high energy particle will execute many cyclotron orbits before compton cooling .
let us check when this assumption is valid .
the rest frame magnetic flux density is related to the initial ( non - radial ) field before acceleration by @xmath154 ( eq . [ eq : bnew ] ) .
each fresh ( radially - moving ) particle has a rest - frame energy @xmath155 , and so it takes a time @xmath156 to complete one radian of a larmor orbit .
the compton cooling time is @xmath157 the energy flux in a thin photon shell is related to the characteristic optical depth @xmath158 ( eq . [ eq : tauc ] ) through @xmath159 , and so the ratio of these two timescales is @xmath160 the compression of the accelerating medium has a related consequence : the mean energy @xmath161 of the scattering charges in the bulk frame increases due to adiabatic heating , at a rate @xmath162 in the lab frame .
a balance between compressional heating and compton cooling results in an equilibrium energy @xmath161 that is only mildly relativistic . as a result
, we will assume that the heating process has isotropized the momenta of the pairs in the bulk frame : @xmath163 the reference ( lab - frame ) cooling time is @xmath164 .
the acceleration rate is decreased by a factor @xmath165 from the value calculated above for a cold plasma : @xmath166 > from equation ( [ eq : eqps ] ) , we have @xmath167 at @xmath168 . even in the case of a pair - loaded plasma ( @xmath169 )
, one sees that energetic particles cool faster than the bulk flow accelerates .
the same conclusion holds for scattering in the klein - nishina regime ( 4 ) , but not when pair creation drives the acceleration ( paper ii ) .
when @xmath170 recoil can no longer be neglected . in the general case , the integral over the scattering angle @xmath30 of the compton force on particles moving along the * r*-axis
, @xmath171 can be performed analitically by making the variable change @xmath172 , and integrating the resulting polynomial in @xmath173 .
the mean rate of momentum transfer can then be written as @xmath174i_xdxd\omega , \label{eq : eqpkn}\ ] ] where @xmath175 can be evaluated in closed form , @xmath176 \label{eq : int}\ ] ] ( cf .
blumenthal 1974 ) .
this function has limiting behavior @xmath177\ ] ] for @xmath178 , and @xmath179 for @xmath180 .
the first term in the last expansion , when substituted back into equation ( [ eq : eqpkn ] ) in the limit @xmath180 , reproduces the standard thomson results of equation ( [ eq : eqt ] ) ( for @xmath181 ) .
plasma at rest relative to a point - source of mass @xmath62 and monoenergetic flux @xmath47 is , from equation ( [ eq : eqpkn ] ) , @xmath182 .
the eddington luminosity for which this force just balances the gravitational attraction at all radii @xmath48 is then given by @xmath183,\ ] ] where the term in parenthesis ( equal to 2.4 , 5.9 , and 11.8 for @xmath184 and 10 , respectively ) is the relativistic correction factor to the classical expression in the thomson limit ( blumenthal 1974 ) . ]
figure 3 shows the integrand function in equation ( [ eq : eqpkn ] ) , @xmath185\}$ ] , versus incident photon angle for different photon energies and particle valocities .
the aberration effect of outward - directed photons into the forward hemisphere ( as observed in the electron rest - frame ) is clearly seen , together with the reduced radiative force as @xmath0-ray photons are preferentially scattered in the forward direction . in the thomson limit ,
photons at angles @xmath186 with respect to the direction of motion are seen in the rest - frame of the electron as blueshifted and inward directed , and work to decelerate the flow ( @xmath187 ) .
the same criterion does not apply in klein - nishina , as blueshifted , aberrated photons experience a decreased relativistic cross section and are scattered less efficiently . when @xmath188 , for example , the rate of momentum transfer becomes negative at angles @xmath189 and 0.13 for photons with energy @xmath184 and 10 , respectively .
the complexity of the kn cross section foils analytic calculations of the particle trajectories .
we have numerically integrated equation ( [ eq : eqpkn ] ) with the runge - kutta method assuming a monoenergetic photon spectrum .
the equilibrium lorentz factor as a function of distance from a radiating sphere and a disk is shown in figure 4 for different energies of the incoming photons . while in the thomson limit the radiation drag imposes a linear relation between @xmath190 and distance @xmath48 from a spherical source , in klein - nishina the importance of this braking force
is reduced : the zero - inertia limit is reached then at higher particle velocities .
the net momentum transfer in each interaction with a @xmath0-ray photon is significantly lower , however , and it is actually more difficult ( i.e. larger compactnesses are needed ) for an electron to approach the zero - inertia limit ( figure 4 ) . in figure 5 the asymptotic lorentz factor at infinity is depicted as a function of source compactness for different incoming photon energies .
a comparison with the thomson limit shows that @xmath0-ray photons with @xmath191 ( say ) are significantly more efficient at accelerating particles only for compactness @xmath192 , and more so in the extended disk case ( i.e. when the isotropic radiation field is larger ) than for a spherical source . in the general case , the detailed dynamics of an
the plasma will depend on the energy spectrum of the incoming radiation ( paper ii ) .
the kn reduction of the photon drag can be easily quantified in the extreme case of a jet moving with relativistic speed towards a point - source of @xmath0-rays . from equation ( [ eq : eqpkn ] ) we derive @xmath193,\ ] ] where @xmath194 .
the term in parenthesis , equal to 0.094 , 0.016 , and 0.006 for @xmath195 and @xmath184 and 10 , respectively , is the relativistic correction to the braking force in scattering thomson .
together with momentum , energy will be transferred from the radiation field to the gas at a rate @xmath196 @xmath197i_xdxd\omega .
\label{eq : ecomp}\ ] ] equations ( [ eq : eqpkn ] ) and ( [ eq : ecomp ] ) assume that all the particles in a small volume element of gas ( that moves with bulk lorentz factor @xmath1 ) are at rest with respect to that volume element .
because of recoil in each collision , however , the plasma will be compton heated ( hence @xmath198 in eq . [ eq : ecomp ] ) to relativistic temperatures , and in general these equations should be integrated over the appropriate distribution function .. for a power - law spectrum with energy indices @xmath199 and @xmath200 below and above the break frequency @xmath201 ( characteristic of gamma - ray bursts ) , one derives @xmath202 and @xmath203k . ]
the integral can can solved numerically for a momentum distribution that is isotropic in the electron rest - frame ( e.g. a relativistic maxwellian ) provided the relaxation time of the plasma is short compared with the dynamical time . in the case of a cold plasma at rest in an _ isotropic _ photon bath ,
the compton heating rate per unit volume can be written , from equation ( [ eq : ecomp ] ) with @xmath204 , @xmath205 xi_xdx , \label{eq : hcomp}\ ] ] where the term @xmath206 ( equal to 0.28 , 0.04 , and 0.008 for @xmath184 and 10 , respectively ) is the kn reduction factor to the classical thomson formula .
it has been recently pointed out that compton heating of electrons by hard x - ray background photons may provide a significant energy source for the intergalactic medium ( madau & efstathiou 1999 ) .
it is crucial in the cosmological context to use equation ( [ eq : hcomp ] ) rather the thomson limit , as relativistic corrections become increasingly more important at early times when the peak in spectral power of the x - ray background ( observed today at @xmath207 kev ) is blueshifted to higher energies .
it is also worth noting that the efficiency of the compton afterburn should be further diminished in the kn regime , as the decreased relativistic cross section reduces the anisotropy ( between highly scattering electrons moving towards the source and those , less scattering , moving away ) that ultimately drives the effect ( phinney 1982 ; see also renaud & henry 1998 ) .
the afterburn may be a supplementary source of momentum if the energetic particles are injected by an external energy source , and if they execute many larmor orbits while cooling ( [ seccompton ] ) .
however , in the present context ( in which the particles are assumed cold before scattering ) only a small increase in the bulk momentum will result from the afterburn effect .
even if we suppose that _ all _ the energy transferred from the radiation field to the gas is actually converted into bulk motion , and integrate equation ( [ eq : ecomp ] ) with @xmath208 , we get asymptotic lorentz factors that are at most ten percent higher than those computed from the momentum equation .
we have benefited from stimulating discussions with g. blumenthal , g. ghisellini , and m. rees .
support for this work was provided by nsf through grant phy94 - 07194 ( p. m. ) , and by the sloan foundation ( c. t. ) . | we consider the radiative acceleration to relativistic bulk velocities of a cold , optically thin plasma which is exposed to an external source of @xmath0-rays .
the flow is driven by radiative momentum input to the gas , the accelerating force being due to compton scattering in the relativistic klein - nishina limit .
the bulk lorentz factor of the plasma , @xmath1 , derived as a function of distance from the radiating source , is compared with the corresponding result in the thomson limit . depending on the geometry and spectrum of the radiation field , we find that particles are accelerated to the asymptotic lorentz factor at infinity much more rapidly in the relativistic regime ; and the radiation drag is reduced as blueshifted , aberrated photons experience a decreased relativistic cross section and scatter preferentially in the forward direction .
the random energy imparted to the plasma by @xmath0-rays can be converted into bulk motion if the hot particles execute many larmor orbits before cooling .
this ` compton afterburn ' may be a supplementary source of momentum if energetic leptons are injected by pair creation , but can be neglected in the case of pure klein - nishina scattering .
compton drag by side - scattered radiation is shown to be more important in limiting the bulk lorentz factor than the finite inertia of the accelerating medium .
the processes discussed here may be relevant to a variety of astrophysical situations where luminous compact sources of hard x- and @xmath0-ray photons are observed , including active galactic nuclei , galactic black hole candidates , and gamma - ray bursts .
= cmmib10 # 1to 0pt#1 |
colored models have been introduced by gurau in the context of group field theory@xcite .
these models admit a perturbative expansion dominated by spherical topologies@xcite and are at the roots of the first renormalizable group field theory@xcite . here , we follow closely the formalism introduced by gurau in the study of the schwinger - dyson equations@xcite . in dimension @xmath0 , the basic dynamical variables of a colored tensor model are two complex conjugate rank @xmath0 tensors @xmath1 and @xmath2 , with @xmath3 . to any bipartite graph with black and white vertices of valence @xmath0 and a proper coloring with @xmath0 colors ,
we associate a @xmath4 invariant denoted @xmath5 constructed as follows . to any white ( resp .
black ) vertex we associate a tensor @xmath6 ( resp .
@xmath2 ) and contract the @xmath7 indices of the tensors following the edges of color @xmath8 in @xmath9 .
for instance , @xmath10{coloredgraph.pdf } } \rightarrow \sum_{i_{a},\,j_{b},\,k_{c } } \overline{m}_{i_{1}i_{2}i_{3}}\overline{m}_{j_{1}j_{2}j_{3}}\overline{m}_{k_{1}k_{2}k_{3 } } m_{i_{1}k_{2}j_{3}}m_{j_{1}i_{2}k_{3}}m_{k_{1}j_{2}i_{3}}\ ] ] the main object of interest is the generating function @xmath11=\log \int[dmd\overline{m}]\,\exp\bigg\{\sum_{\gamma}\frac{\lambda_{\gamma}}{c_{\gamma}}\mbox{tr}_{\gamma}(m,\overline{m})\bigg\},\label{generating}\ ] ] where @xmath12 $ ] is the lebesgue measure on the real and imaginary part of the tensors and @xmath13 is the cardinal of the automorphism group of @xmath9 .
the integral @xmath14 $ ] is defined as a formal power series as follows .
we single out the contribution of the dipole graph @xmath15{dipole.pdf } } \rightarrow \sum_{i_{a}}\overline{m}_{i_{1},i_{2},i_{3}}m_{i_{1}i_{2}i_{3}}\ ] ] whose exponential defines a gauian measure on tensors .
then , we use wick s theorem to expand @xmath14 $ ] as a formal power series in the coefficients @xmath16 , where the number of vertices @xmath9 is strictly greater than 2 . from a geometric perspective
, each graph @xmath9 represents a triangulation of a space of dimension @xmath0 , with @xmath0-simplexes on its vertices glued together along their boundary @xmath17-simplexes following the edges of @xmath9 .
the feynman graphs appearing in the expansion of @xmath14 $ ] are associated with cellular decompositions of @xmath18-dimensional spaces .
in order to derive the schwinger - dyson equations for tensor models @xcite , it is convenient to define the contraction of a pair of vertices @xmath19 of different colors in a non necessarily connected , bipartite colored graph @xmath9 of degree @xmath0
. it is the graph @xmath20 obtained by removing the vertices @xmath21 and @xmath22 and attaching the resulting free half - edges of identical colors . for example , in @xmath23 , -0.8 cm [ cols="^,^,^ " , ] note that these operations play a important role in the proof of the renormalizabilty of some colored models@xcite . from a geometrical perspective
, this equation can be depicted as . in this equation ,
the effective action is a sum over connected feynman graph of the tensor model .
therefore , it may be understood as a sum over connected @xmath18-dimensional triangulations , whose boundary may be disconnected .
then , the evolution of such a triangulation results either from an identification of two @xmath0-simplexes in the boundary of the same connected @xmath18-dimensional triangulation or in different @xmath18-dimensional triangulations .
group field theories@xcite are particular quantum field theories defined on several copies of a group @xmath24 which may be @xmath25 , @xmath26 or @xmath27 . in the sequel ,
we restrict our attention to @xmath25 .
their perturbative yield a sum over @xmath18-dimensional triangulations weighted by the corresponding spin foam amplitudes appearing in quantum gravity@xcite .
the graphs @xmath9 are at the basis of the loop quantum gravity boundary hilbert spaces @xmath28 of gauge invariant functions @xmath29 with @xmath30 and @xmath31 source and target of the edge @xmath32 .
the group field theory generating function analogous to reads @xmath33=\log\int[d\phi ] \exp\big\{\sum_{\gamma}\int \!\!dh_{e}dg_{v , e}\\\ , \frac{\psi_{\gamma}(h_{e})}{c_{\gamma } } \prod_{e}\delta\big(h^{}_{e}\,g_{s(e),e}^{-1}g^{}_{t(e),e}\big)\prod_{v}\phi(g_{v , e_{1}},\cdots , g_{v , e_{d}})\big\}\end{gathered}\ ] ] a convenient basis of @xmath34 is provided by spin networks @xmath35 constructed by associating a spin to every edge and to every vertex an intertwiner between all the spin associated to the edges incident to that vertex .
therefore , all the previous formalism can be extended mutatis mutandis to group field theory replacing graphs by spin networks .
the author warmly thanks vincent rivasseau for his invitation to this conference and yongge ma for his hospitality at beijing normal university .
r. gurau , nucl .
b * 865 * ( 2012 ) 133 _ the schwinger dyson equations and the algebra of constraints of random tensor models at all orders _
http://arxiv.org/abs/arxiv:1203.4965[arxiv:1203.4965 [ hep - th ] ] a. connes and d. kreimer , _ renormalization in quantum field theory and the riemann - hilbert problem .
1 . the hopf algebra structure of graphs and the main theorem _ commun
. math .
* 210 * ( 2000 ) 249 http://arxiv.org/abs/hep-th/9912092[[hep-th/9912092 ] ] | in this talk , we elaborate on the operation of graph contraction introduced by gurau in his study of the schwinger - dyson equations . after a brief review of colored tensor models , we identify the lie algebra appearing in the schwinger - dyson equations as a lie algebra associated to a hopf algebra of the connes - kreimer type .
then , we show how this operation also leads to an analogue of the wilsonian flow for the effective action .
finally , we sketch how this formalism may be adapted to group field theories . |
the study of topological insulators have become a ubiquitous study in condensed matter physics .
these materials have captivated considerable attention of researchers @xcite .
they involve materials such as bi@xmath0se@xmath1 or bi@xmath0te@xmath1 .
one of the special properties of these materials is that they have an electronic bulk structure with a finite gap separating the conduction band from the valence band , but their edges ( for two - dimensional ( 2d ) tis ) or surfaces ( for 3d tis ) , however , have gapless states , which are protected by trs .
provided that the system remains time - reversal invariant , these gapless states are robust .
the bulk hamiltonian near the energy gap closing points ( dirac points ) replicates a 2d dirac hamiltonian , which describes the surface states .
interesting electronic transport properties emerge when trs is explicitly broken .
there are different ways to induce an explicit trs breaking interactions .
they can be induced by applying a perpendicular magnetic field or by depositing a ferromagnet on the surface of a ti .
these trs breaking interactions lift the degeneracy of the dirac points and introduce a finite gap on the surface of a ti .
the upshot is that one observes interesting electronic transport , which include a half quantized conductivity , majorana bound states , etc.@xcite .
however , the quantum spin hall ( qsh ) phase @xcite is obtained by preserving trs by the inclusion of an intrinsic spin - orbit coupling ( soc ) .
this phase is reminiscent of two copies of quantum anomalous hall ( qah ) phases , where each copy breaks trs and the total system remains time - reversal invariant .
it has been shown that qsh effect can persist even when trs is explicitly broken @xcite .
these interesting observations have paved the way for numerous investigations and the fates of qsh effect in complicated systems . in the qsh phase , there are two pairs of gapless edge state modes counter - propagating in the vicinity of the bulk gap , which can be captured in the tight binding limit of the low - energy hamiltonian .
the importance of tight binding model in topological insulators was first elucidated by bernevig , hughes , and zhang bhz @xcite on a square lattice .
following the same line of study , huichao li _
@xcite coined an alternative tight binding model for ti thin film on a square lattice . in recent years
, topological insulators have also been investigated on a honeycomb lattice , @xcite , kagome lattice @xcite lieb and perovskite lattices @xcite , and pyrochlore lattice @xcite .
the purpose of the paper is to provide another lattice geometry in which the physics of qsh effect and topological quantum phase transition are manifested .
we investigate a tight binding model on the porphyrin lattice @xcite , in which trs breaking through the zeeman effect requires the nearest neighbour sites to have a complex amplitude , reminiscent of the haldane model @xcite ; see fig . .
by incorporating an additional phase ( gauge ) into the complex amplitudes we show that in a special limit and with a specific gauge choice , the pristine model reduces to a time - reversal invariant hamiltonian , which in turn reduces to the usual ti thin film model in the continuum limit .
we also show that this model possesses different chern numbers in different regimes of the parameter space just like in real systems .
we further explore this model by incorporating spin degrees of freedom and explicit trs breaking couplings , such as an off - diagonal coupling ( electrostatic potential ) , @xmath2 , a magnetic field , @xmath3 , and a rashba soc , @xmath4 .
we explicitly map out different phase diagrams at the gap closing points .
lattice regularizations of the low - energy hamiltonian is an expedient way for studying the topological properties and the edge states of a ti in the entire brillouin zone . in this paper
, we study the tight binding model of porphyrin thin films .
@xcite the tight binding model can be generally written as @xmath5 lattice hamiltonian of this form is usually considered as a toy model used to understand the physics of qsh effect in real materials.@xcite here , @xmath6 creates ( annihilates ) an electron with spin @xmath7 at site @xmath8 ; @xmath9 denotes sublattice @xmath10 and @xmath11 respectively ; and @xmath12 ; where @xmath13 and @xmath14 represents the components of sublattice ( layer ) pseudospins .
@xmath15 for @xmath16 respectively , _
i.e. , _ if the electron makes a right or left diagonal turn to get to the second bond .
the coordinates @xmath17 are defined in fig . .
@xmath18 for @xmath19 ; where @xmath20 are complex variables .
the summations over @xmath21 and @xmath22 denote nearest and next nearest neighbours on a porphyrin square sublattice respectively ; see fig . .
in the absence of the phase @xmath23 and the spin degrees of freedom , the first three terms in eq .
correspond to the model proposed by joel yuen - zhou , _
et al.,_@xcite in the context of frenkel excitons ( quasiparticles with no net charge ) .
a special property of this tight binding model is that the nearest neighbour hopping sites in eq . have complex amplitudes , which are explained in ref .
the motivation for the phase factor ( gauge ) @xmath23 is rather technical .
it does not , however , change the complex nature of the hopping amplitudes .
in particular , the total phase in a unit cell vanishes , reminiscent of the haldane model .
@xcite the reason for this phase factor will become apparent in the subsequent sections , where we show that different choices of @xmath23 lead to different surface hamiltonians , perhaps not different physics .
the second and the third terms represent the diagonal couplings between the upper and the lower surface states of the thin film . in addition , we have introduced other couplings in order to enhance the topological properties of this thin film .
the fourth term is an electrostatic potential which couples ( off - diagonal ) the upper and lower surface states ; the fifth term can be identified as the rashba - like soc in the thin film , and the last term is a staggered sublattice potential , which corresponds to a magnetic field in the thin film .
next , it is expedient to fourier transform into momentum space , _ i.e. _ , @xmath24 .
we obtain h=_k ^ ( ) ( ) ( ) , where @xmath25 is the nambu operators for the thin film .
@xmath26\bold{i}_\tau\sigma_x -\bigg[\bar{\rho_1}\cos\lb\frac{k_x+k_y}{2}-{\phi}\rb+\bar{\rho_2}\cos\lb\frac{k_x - k_y}{2}+{\phi}\rb\bigg]\bold{i}_\tau\sigma_y \nonumber\\&+[{t_m}-t_{\perp}\lb
\cos k_x+\cos k_y\rb ] \tau_z \sigma_z+ 2t_{\parallel}\lb \cos k_x \cos k_y \rb\tau_x \sigma_y + 2t_{r}\cos k_x \cos k_y \lb \tau_y \sigma_x -\tau_x \sigma_y\rb + h\bold{i}_\tau\sigma_z ; \label{fullti}\end{aligned}\ ] ] where @xmath27 , @xmath28 ; and @xmath29 , @xmath30 ; @xmath31 and @xmath32 denote real and imaginary parts . the pauli matrices @xmath33 act on the real spin space , while @xmath34 act on the upper and the lower surfaces of the thin film , with spin - up on the upper surface and spin - down on the lower one and @xmath35 is a @xmath36 identity matrix . and @xmath11 .
nearest neighbour sites are along the diagonals with coordinates @xmath37 and @xmath38 ; where @xmath39 and @xmath40 .
next - nearest neighbour sites are along the horizontal and vertical directions with coordinates @xmath41 and @xmath42 respectively . , width=220 ]
in order the see how the physics of qsh effect is manifested on the porphyrin lattice , we consider a special limit of eq . .
this limit corresponds to the first three terms in eq .
( _ i.e. , _ all the terms involving the big and small square brackets ) , and we call it the _ pristine model_. the first two terms ( big square brackets ) in eq
. yield different functions depending on the choice of @xmath23 .
the gauge choice @xmath43 amounts to the hamiltonian studied in ref .
we consider an alternative gauge choice of the pristine hamiltonian with @xmath44 . in this gauge ,
the hamiltonian decouples into two trs copies , _ i.e.,_@xmath45 : @xmath46 , where the time reversal operator is given by @xmath47 and @xmath48 is a complex conjugation .
next , we take the limit @xmath49 and @xmath50 , which corresponds to @xmath51 .
the upshot of these limits is that the hamiltonian simplifies as : @xmath52\tau_z \sigma_z.\end{aligned}\ ] ] expanding eq . near the @xmath53 point @xmath54
, we see that the resulting low - energy hamiltonian is analogous to that of a 3d ti thin film .
@xcite separates different phases with trivial and non - trivial chern numbers .
, width=220 ] and @xmath55 , the qsh phase @xmath56 , and the insulating phase @xmath57 .
, width=288 ] as the pristine hamiltonian , eq . , has a different structure from previously studied model on the square lattice,@xcite we will show how topological effects are manifested in eq . just like in the previously studied models . as the system possesses trs , the corresponding eigenvalues are two - fold degenerate .
the topological phase transition points are located at the momenta where the energy eigenvalues vanish .
there are four semi - metallic states at the time reversal invariant momenta ( trim ) points : @xmath58 and @xmath59 for @xmath60 respectively ( see fig . ) and @xmath61 and @xmath62 for @xmath63 respectively .
we focus on the points @xmath64 and @xmath65 , with nontrivial phase boundaries .
the phase boundaries separate regions with qsh phase and ordinary insulator phase .
the spin chern number characterizing these different phases is given by @xcite @xmath66 the berry curvature @xmath67 has the form : @xmath68 where the summation is over all occupied valence bands in the first brillouin zone below the bulk energy gap , and @xmath69 ; @xmath70 define the velocity operators . direction , with @xmath71 and @xmath72 .
@xmath73 @xmath74 ; @xmath75 @xmath76 ; @xmath77 @xmath78 .
, width=288 ] the chern numbers can be computed separately .
the appropriate formula is given by @xcite _ _ z = dk_xdk_yn _ , [ hall ] where @xmath79 is an antisymmetric tensor ; @xmath80 are the components of a unit vector with , @xmath81 for this model , all the six terms in eq .
contribute equally .
the numerical computation of the spin chern numbers in the entire brillouin zone yields @xmath82 independent of @xmath83 .
the trivial chern number corresponds to an ordinary insulator phase .
the chern number is ill - defined at the gap closing points ( semi - metallic phase ) .
the gap further reopens , which corresponds to a qsh phase with a nontrivial chern number .
the phase diagram is depicted in fig . , where the phase boundary @xmath84 separates different phases with trivial and non - trivial chern numbers .
the nontrivial chern numbers in the qsh phase indicate that there are a pair of counter - propagating edge states in the vicinity of the bulk gap .
we have solved for the edge states by diagonalizing the hamiltonian in a cylindrical geometry , with periodic boundary conditions along the @xmath85 direction and open conditions along the @xmath86 direction @xcite . in fig . , we show the plot of the bulk energy spectrum in the three phases shown in eq . .
the edge states clearly reveal the topological properties of this system .
it is apparent that in the qsh regime @xmath73 , there are two pairs of edge states modes with opposite pseudospins ( which are degenerate ) that propagate in the vicinity of the bulk gap ; these states are gapless , @xmath87 at @xmath88 , corresponding to the absolute value of the chern number @xcite .
the bulk gap closes at @xmath75 corresponding to the phase transition point . in the insulating regime ,
the gap reopens and the edge states participate in the bulk gap separation @xmath77 , with no counter - propagating edge states modes in the vicinity of the bulk gap .
in this section we study the topological phases of the full hamiltonian in eq . .
henceforth we work with the gauge @xcite @xmath44 and take @xmath51 , which corresponds to @xmath49 and @xmath50 .
we further take @xmath83 as the energy unit . the full hamiltonian in eq . does not possess any analytical diagonalization .
we find that at trim points @xmath64 and @xmath65 the topological phase boundary is given by @xmath89 \bigg [ 1-\lb\frac{2 \bar{t}_r } { { t}^{s}}\rb^2\bigg ] } , \label{crit}\end{aligned}\ ] ] as a function of @xmath3 for fixed value of @xmath90 , @xmath91 .
the red dash line is for @xmath92 and the blue dash line is for @xmath93 .
there are two gapped regions separated by a gap closing point in the qsh regime @xmath74 and the ordinary insulating regime @xmath94 of the pristine model .
, width=307 ] which is determined by the condition that the conduction and valence bands touch each other as in the case of graphene @xcite . here , @xmath95 ; @xmath96 , and @xmath97 at @xmath64 and @xmath65 respectively .
the phase boundary at the other trim points @xmath98 and @xmath99 corresponds to @xmath100 in eq . .
note that the topological phase boundary diverges at the phase boundary of the pristine model @xmath60 .
now we consider the qsh regime @xmath101 and the ordinary insulating regime @xmath94 of the pristine model , with fixed value of @xmath2 , and several values of @xmath4 .
we find that the bulk gap as a function of @xmath3 at @xmath64 initially increases at @xmath102 , and gradually decreases as @xmath3 increases from zero , and then vanishes at the topological phase boundary @xmath103 .
the gap further reopens for @xmath104 ; see fig . .
this is not different from what is observed in other lattice geometries such as the honeycomb lattice , @xcite , the kagome lattice @xcite , the lieb and perovskite lattices @xcite , and the pyrochlore lattice @xcite .
we have solved for the edge states to substantiate the topological properties of these phases . in fig . , we plot the edge states along the @xmath105 direction . in the ordinary insulating regime ( @xmath106 ) of the pristine model with @xmath107 @xmath108 , we see that there is an ordinary insulating phase with no edge state modes at @xmath109 .
the gap closes at the phase boundary @xmath110 which indicates a phase transition . as seen in fig .
, at @xmath111 the magnetic field dominates , we see that the gap reopens and there are a pair of counter - propagating edge state modes in the vicinity of the bulk gap , which indicates a qah phase .
we find that no qsh phase appears in this regime .
on the other hand , in the qsh regime ( @xmath112 ) of the pristine model with @xmath113 @xmath114 , there are two pairs of counter - propagating edge state modes in the vicinity of the bulk gap at @xmath109 , where the red and blue colours represent the edge state modes with opposite pseudospins .
again the gap closes at the phase boundary @xmath110 which indicates a phase transition . at @xmath111 ,
the magnetic field once again dominates and the gap reopens with a pair of counter - propagating edge state modes in the vicinity of the bulk gap , indicating a qah phase . on the contrary , no ordinary insulator phase appears in this regime .
the phase diagram in this regime is shown in fig . .
it is interesting to note that a qah phase appears in each regime of the pristine model .
we have checked that the continuum hamiltonian also corroborates our results .
thus , the porphyrin lattice provides another lattice geometry to study qsh effect and topological quantum phase transition . and @xmath115 with @xmath112 .
the red dot represents the point where the boundary vanishes .
, width=220 ]
the porphyrin lattice represents another lattice geometry where the physics of quantum spin hall ( qsh ) effect and topological quantum phase transition are manifested . in this paper , we have explicitly shown how these physics can be constructed from a tight binding toy model .
we showed that the porphyrin lattice tight binding model captures the topological quantum phase transitions observed in other lattice geometries .
in particular , the transitions between qsh phase and ordinary insulator phase , as well as quantum anomalous hall ( qah ) phase .
we also showed that the complex hopping amplitudes and the gauge degree of freedom are the distinguishing features of the porphyrin lattice tight binding model as was first introduced in ref .
the author is indebted to tami pereg - barnea for pointing out a crucial error in a related model studied by the author .
research at perimeter institute is supported by the government of canada through industry canada and by the province of ontario through the ministry of research and innovation .
99 joel yuen - zhou , _
et al . _ , nature materials * 13 * , 1026 ( 2014 ) . f. d. m. haldane , , 2015 , 1988
. huichao li , _ et al .
_ , , 165104 , 2010 .
huichao li , l. sheng , and d. y. xing , , 045118 , 2012 ; , 035310 , 2012 .
zyuzin , m.d .
hook , a.a .
burkov , , 245428 ( 2011 ) .
hai - zhou lu , _ et al .
_ , , 115407 ( 2010 ) .
shan , h .- z .
lu , and s .- q .
shen , new j. phys .
* 12 * , 043048 , 2010 .
bernevig , t.l .
hughes , and s .- c .
zhang , science * 314 * , 1757 ( 2006 ) .
markus koenig , _ et al .
_ , j. phys .
. jpn . * 77 * , 031007 ( 2008 ) zhong xu , yunyou yang , l. sheng , and d y xing , j. phys . : condens .
matter 24 , 185504 , 2012 .
guo and m. franz ; , 113102 , 2009 .
c. weeks and m. franz ; , 085310 , 2010 .
guo and m. franz ; , 206805 , 2009 .
d. j. thouless , m. kohmoto , m. p. nightingale , and m. den nijs , , 405 , 1982 ; mahito kohmoto , annals of physics * 160 * , 343 , 1985 .
klinkhamer , g.e .
volovik , int . j. mod
* a20 * 2795 , ( 2005 ) ; grigory e. volovik , the universe in a helium droplet , oxford university press , 2003 .
yasuhiro hatsugai , , 3697 , 1993 . | we study the physics of quantum spin hall ( qsh ) effect and topological quantum phase transition on the porphyrin lattice .
we show that in a special limit the pristine model on this lattice reduces to the usual topological insulator ( ti ) thin film model . by incorporating explicit time reversal symmetry ( trs ) breaking interactions
, we explore the topological properties of this model and show that it exhibits an ordinary insulator phase , a quantum anomalous hall ( qah ) phase , and a quantum spin hall ( qsh ) phase . |
entanglement , one of the most striking features of quantum mechanics , has been considered as a key resource of quantum information processing @xcite . in recent years
, the manipulation of quantum entanglement for the system of the cavity quantum electrodynamics(qed ) has been extensively investigated@xcite .
the cavity qed , where atoms interacting with quantized electromagnetic fields inside a cavity , can be used to create the entanglement between atoms in cavities and establish quantum communications between different optical cavities .
it has been shown that the entanglement can arise in the interaction of a atom with a cavity field in a thermal state @xcite and two atoms with a cavity field @xcite . in ref.@xcite ,
solano _ et al_. have shown that multipartite entanglement can be generated by putting several two - level atoms in a cavity of high quality factor . with the help of a strong classical driving fields ,
the schrdinger cat state and other entangled states can be produced .
on the other hand , many efforts have been devoted to the study of the evolution of the joint system formed by two qubits@xcite . particularly , yu and eberly @xcite pointed out that the single - qubit dynamics and the global dynamics of an entangled two - qubit system subjected to independent environments may be rather different .
for example , for a single - qubit system subjected to an environment , the local coherence decays asymptotically .
however , the entanglement of an entangled two - qubit system may disappear within a finite time during the dynamics evolution .
the nonsmooth finite - time disappearance of entanglement is called `` entanglement sudden death '' ( esd ) .
recently , it is reported that the esd phenomenon has been observed in a quantum optics experiment @xcite . in the present paper
, we propose a scheme to enhance the entanglement of a quantum system consisting of two - level atoms within cavities by applying and controlling classical driving fields .
the influence of the classical field and the purity of the initial state on the esd is also studied .
first , we study the entanglement of the two - level atom and the field by employing logarithmic negativity @xcite .
then , we consider a quantum system consisting of two noninteracting atoms each locally interacting with its own vacuum field .
the two atoms , which are initially prepared in extended werner - like ( ewl ) states @xcite , are driven by two independent classical fields additionally .
it is shown that the esd may appear in this system and the esd time can be retarded by increasing the purity of the initial state of the two atoms .
in addition , the amount of the entanglement of the two atoms can be significantly increased by applying classical fields .
we find that the time of entanglement sudden death can be controlled by the classical driving fields .
it is interesting to point out that the entanglement sudden death phenomenon disappear if the classical driving fields are strong enough .
the present paper is organized as follows . in section
ii , we propose a scheme to improve the entanglement of a two - level atom interacting with vacuum or thermal field by applying a classical driving field . in section iii , we study the influence of the classical driving fields on the esd when the atoms are initially prepared in ewl states .
a conclusion is given in section iv .
now , we consider a system consisting of a two - level atom inside a single mode cavity .
the atom is driven by a classical field additionally .
the hamiltonian of the system can be described by @xcite @xmath0 where @xmath1 , @xmath2 and @xmath3 are the frequency of the cavity , atom and classical field , respectively .
the operators @xmath4 and @xmath5 are defined by @xmath6 , @xmath7 , and @xmath8 where @xmath9 and @xmath10 are the excited and ground states of the atom . here ,
@xmath11 and @xmath12 are the annihilation and creation operators of the cavity ; g and @xmath13 are the coupling constants of the interactions of the atom with the cavity and with the classical driving field , respectively .
note that we have set @xmath14 throughout this paper . in the rotating reference frame
the hamiltonian of the system is transformed to the hamiltonian @xmath15 under a unitary transformation @xmath16 @xmath17 with @xmath18 and @xmath19 . using the method similar to that used in ref.@xcite , diagonalizing the hamiltonian @xmath20 , and neglecting the terms which do not conserve energies ( rotating wave approximation )
, we can recast the hamiltonian @xmath15 as follows @xmath21,\end{aligned}\ ] ] with @xmath22 . here
@xmath23 stands for hermitian conjugation .
the hamiltonian ( 4 ) can be diagonalized by a final unitary transformation @xmath24 with @xmath25}$ ] . then
, we can rewrite the hamiltonian of the system @xmath26,\end{aligned}\ ] ] where @xmath27 and @xmath28 .
it is worth noting that the unitary transformations @xmath29 and @xmath24 are both local unitary transformations .
as we known the entanglement of a quantum system does not change under local unitary transformations @xcite .
thus , the entanglement of the system considered here will not be changed by applying the local unitary transformations @xmath29 and @xmath24 .
we study the entanglement of the system by employing the logarithmic negativity . for a bipartite system described by density matrix @xmath30 ,
the logarithmic negativity is @xmath31 where @xmath32 is the partial transpose of @xmath30 , @xmath33 stands for the trace norm of @xmath32 and n is negativity which is defined by @xcite @xmath34 which is just the absolute value of the sum of the negative eigenvalues of @xmath32 .
the logarithmic negativity has been proven to be an operational good entanglement measure@xcite .
we first assume that the atom is initially prepared in the state @xmath35 and the field is in the fock state @xmath36 .
then , we can find that the state vector at time t is @xmath37 , \nonumber\\ \beta_n(t)&=&-ig\cos^2{\frac{\theta}{2}}\sqrt{n}e^{-i\delta_2t/2}\sin{(\omega_{n-1}t)}/\omega_{n-1},\nonumber\\ \delta_2&=&\sqrt{(\omega_0-\omega_c)^2 + 4\lambda^2}+\omega_c-\omega , \nonumber\\ \omega_n&=&\sqrt{\frac{\delta_2 ^ 2}{4}+(n+1)(g\cos^2{\frac{\theta}{2}})^2},\nonumber\\ inserting the state vector into eq.(7 ) leads to the logarithmic negativity of the state @xmath38 @xmath39.\end{aligned}\ ] ] next , we consider the entanglement dynamics of the system when the atom and the field are initially prepared in the state @xmath40 and the thermal state , respectively .
the initial density matrix of the system is @xmath41 where @xmath42 is the mean photon number at the inverse temperature @xmath43 and @xmath1 is the frequency of the optical mode .
a straightforward calculation shows that the density matrix at time t is @xmath44\}\nonumber\\\end{aligned}\ ] ] which leads to the logarithmic negativity of the above density matrix @xmath45},\nonumber\\ \xi_n^-&=&\frac{1}{2}\{p_{n-1}|\beta_n|^2+p_{n+1}|\alpha_{n+1}|^2\nonumber\\ & & -[(p_{n-1}|\beta_n|^2-p_{n+1}|\alpha_{n+1}|^2)^2\nonumber\\ & & + 4|p_n\alpha_n\beta^*_{n+1}|^2]^{1/2}\}.\end{aligned}\ ] ] the logarithmic negativity is plotted as a function of time t in fig.1 that when the field is initially prepared in the vacuum state @xmath46 or in the fock state @xmath47 .
for the sake of simplicity , we set @xmath48 in the numerical calculation .
we can see clearly from fig.1 , the atom - field entanglement can be increased by applying the classical driving field . in fig.2 , the logarithmic negativity @xmath49 is plotted as a function of t with @xmath50 for @xmath51(dotted line ) and @xmath52(solid line ) .
the mean photon number @xmath53 for the upper panel and the lower panel are 0.1 and 0.3 , respectively .
it is not difficult to find that the entanglement of the two - atom system decreases with the mean photon number @xmath53 .
in other words , the amount of entanglement decreases with the increase of the temperature of the thermal field .
however , we can increase the amount of entanglement by controlling the classical field .
is plotted as a function of t with @xmath50 for @xmath51(dotted line ) and @xmath52(solid line ) .
upper panel : the mean photon number of the thermal field is 0.1 , i.e. , @xmath54 . lower panel : the mean photon number of the thermal field is 0.3 , i.e. , @xmath55 . ]
in the previous section , we have investigated the entanglement dynamics of a two - level atom interacting with a cavity field .
recently , bellomo _
et al_. @xcite investigated the influence of the markovian and non - markovian effects on the dynamics of two - qubit entanglement . in this section
, we consider a quantum system consisting of two noninteracting atoms each locally interacting with its own vacuum field .
the two atoms , which are initially prepared in ewl states , are driven by two independent classical fields additionally .
our purpose here is to investigate the influence of the classical driving field and the purity of the initial states on the entanglement dynamics of the system . here , we study the entanglement dynamics of two independent atoms each locally interacting with a vacuum field by using the procedure of ref.@xcite . in addition
, each atom is driven by a classical field . in the following ,
we find the reduced density matrix of the two - atom system . in the basis @xmath56
, the initial state of the atom interacting with the vacuum field can be cast in the following form @xmath57 where @xmath40 and @xmath58 are defined by eq.(8 ) . using the state vector in eq.(8 ) and tracing over the freedom of the field
, we obtain the single - atom density matrix evolution @xmath59 where @xmath60\rho_{++}(0),\nonumber\\ \rho_{+-}(t)&=&\alpha_0(t)\rho_{+-}(0)=\rho^*_{-+}(t).\end{aligned}\ ] ] by making use of the above equations , we can find the dynamics of the two atoms via a purely algebraic way . note that this procedure is applicable to arbitrary initial states of the whole system . in the basis @xmath61\rho_{11}(0)+|\alpha_0(t)|^2\rho_{22}(0),\nonumber\\ \rho_{33}(t)&=&|\alpha_0(t)|^2[1-|\alpha_0(t)|^2]\rho_{11}(0)+|\alpha_0(t)|^2\rho_{33}(0),\nonumber\\ \rho_{44}(t)&=&[1-|\alpha_0(t)|^2]^2\rho_{11}(0)+\rho_{44}(0)\nonumber\\ & & + [ 1-|\alpha_0(t)|^2 ] [ \rho_{22}(0)+\rho_{33}(0)],\nonumber\\ \rho_{12}(t)&=&\alpha_0(t)|\alpha_0(t)|^2\rho_{12}(0),\nonumber\\ \rho_{13}(t)&=&\alpha_0(t)|\alpha_0(t)|^2\rho_{13}(0),\nonumber\\ \rho_{14}(t)&=&\alpha_0(t)^2\rho_{14}(0),\nonumber\\ \rho_{23}(t)&=&|\alpha_0(t)|^2\rho_{23}(0),\nonumber\\ \rho_{24}(t)&=&\alpha_0(t)[1-|\alpha_0(t)|^2]\rho_{13}(0)+\alpha_0(t)\rho_{24}(0),\nonumber\\ \rho_{34}(t)&=&\alpha_0(t)[1-|\alpha_0(t)|^2]\rho_{12}(0)+\alpha_0(t)\rho_{34}(0),\end{aligned}\ ] ] with @xmath62
. we would like to point out that the above procedure allows us to obtain the reduced density matrix of the two - atom system for any initial state .
we assume the initial states of the two - atom system are the extended werner - like states defined by @xmath63 is a @xmath64 identity matrix , @xmath65 and @xmath66 are complex numbers with @xmath67 .
it is noted that the ewl states belong to the class of the ` x ' states .
explicitly , if the density matrix of a quantum state is of the form @xmath68 then it belongs to the class of the x states .
the ewl states have the following advantages .
first , we can easily find that the density matrix at arbitrary time t is still x structure under the single atom evolution determined by the hamiltonian @xmath69 ( in the basis @xmath70 and @xmath65 .
if @xmath71 the ewl states reduce to the bell - like states @xmath72 and @xmath73 . in the case of @xmath74
the ewl states become the bell states while in the case of @xmath75 they are the maximally mixed states .
note that we assume that the initial states of the system are the ewl states .
the x structure of the ewl states is maintained during the evolution of the two - atom system .
the explicit analytical expression of the logarithmic negativity for the two - atom system can be obtained as @xmath76\}\nonumber\\ & & + \max\{0,\frac{1}{2}[\sqrt{(\rho_{11}-\rho_{44})^2 + 4|\rho_{23}|^2}\nonumber\\ & & -\rho_{11}-\rho_{44}]\}.\end{aligned}\ ] ] we plot the logarithmic negativity as a function of time for different values of the purity of the initial states r and the strength of the classical field @xmath13 . in what follows
, we denote the entanglement of the two - atom system which is initially prepared in @xmath77 ( @xmath78 by @xmath79 ( @xmath80 ) .
is plotted as a function of @xmath13 and @xmath3 with @xmath81 .
lower panel : the logarithmic negativity @xmath80 is plotted as a function of @xmath13 and @xmath3 with @xmath81 . ] the logarithmic negativity @xmath79 and @xmath80 are plotted as a function of @xmath13 and @xmath3 with latexmath:[$\omega=\omega_0=1 , panel of fig.3 , respectively . from fig.3
, one can clearly see that if the parameters @xmath3 and @xmath13 are small , then the logarithmic negativity @xmath83 and @xmath84 are zero at time @xmath85 .
the situation is different if the parameters of the classical driving fields increase .
when the values of @xmath3 and @xmath13 are lager enough the two atoms become entangled .
the behavior of @xmath83 and @xmath84 are different when the classical fields are applied .
for example , the area for @xmath86 is small than that of @xmath87 which means that the entanglement of the two - atom system is more sensitive with the classical driving fields if they are initially prepared in @xmath88 .
this feature can be see more clearly in fig.4 and fig.5 .
it also interesting to note that the entanglement of the two - atom system can be significantly increased by controlling the classical driving fields both for the initial states @xmath88 and @xmath89 .
is plotted as a function of t with @xmath90 for @xmath51 ( dotted line ) , @xmath91 ( dashed line ) , and @xmath92 ( solid line ) .
lower panel : the logarithmic negativity @xmath80 is plotted as a function of t with @xmath90 for @xmath51 ( dotted line ) , @xmath91 ( dashed line ) , and @xmath92 ( solid line ) . ] in fig.4 , we plot the logarithmic negativity of the two - atom system as a function of time t for several values of @xmath3 and @xmath13 .
the upper panel of fig.4 shows that the esd occurs without the classical field ( see the dotted line in this panel ) .
after finite dark periods , the entanglement @xmath79 revivals completely . the entanglement can be increased by applying the classical field(see the dashed and solid lines of the panel ) .
it is interesting to point out that the esd disappears when the classical field is strong enough .
comparing the upper panel with the lower panel of fig.4 , we find that the behavior of entanglement @xmath79 and @xmath80 are different since the esd appears in @xmath80 with @xmath91 while there is no esd for @xmath79 with the same values of @xmath3 and @xmath13 .
besides , the entanglement dark periods of @xmath80 is longer than that of @xmath79 when all of the other correspongding parameters are the same . in other words , it takes more time for @xmath80 to revive the initial entanglement .
the parameters @xmath65 and @xmath66 in fig.4 are chosen to be @xmath93 and @xmath94 , respectively .
it is worth noting that different choices of @xmath65 and @xmath66 do not give entanglement dynamics qualitatively different from the case treated here .
another aspect of interest is how the entanglement dynamics is influenced by the purity of the initial states . in order to show it intuitively
, we plot the logarithmic negativity of the two - atom system as a function of t for different values of r in fig.5 , fixing the parameters @xmath65 and @xmath66 . from fig.5
, one can see that the entanglement dark periods increase with the decrease of the purity of the initial states .
the classical driving field again can increase the amount of the entanglement of the system .
is plotted as a function of t with @xmath95 for @xmath96 ( dotted line ) , @xmath97 ( dashed line ) , and @xmath71 ( solid line ) .
lower panel : the logarithmic negativity @xmath80 is plotted as a function of t with @xmath95 for @xmath96 ( dotted line ) , @xmath97 ( dashed line ) , and @xmath71 ( solid line ) . ]
in the present paper , we have considered a quantum system consisting of one two - level atom interacting with a single mode field .
the atom is driven by a classical field additionally .
the entanglement dynamics is described by the logarithmic negativity .
we first investigated the entanglement dynamics of the atom - field system .
the field is initially prepared in the fock state or the thermal state .
we find that the atom - field entanglement can be increased by applying the classical driving field .
then , we consider a quantum system consisting of two noninteracting atoms each locally interacting with its own vacuum field .
the two atoms , which are driven by two independent classical fields , are initially prepared in ewl states , i.e. , @xmath98 and @xmath99 .
we find that classical driving fields can increase the amount of entanglement of the two - atom system .
the behavior of entanglement @xmath83 and @xmath84 are different when the classical fields are applied .
the entanglement of the two - atom system is more sensitive with the classical driving fields if they are initially prepared in @xmath88 .
it is worth noting that the time of entanglement sudden death can be controlled by the classical driving fields .
specially , the entanglement sudden death phenomenon will disappear if the classical driving fields are strong enough .
it is interesting to verify the scheme of our paper in cavity qed experimentally .
this project was supported by the national natural science foundation of china ( grant no.10774131 ) and the national key project for fundamental research of china ( grant no .
2006cb921403 ) . | we investigate the entanglement dynamics of a quantum system consisting of two - level atoms interacting with vacuum or thermal fields with classical driving fields .
we find that the entanglement of the system can be improved by adjusting the classical driving field .
the influence of the classical field and the purity of the initial state on the entanglement sudden death is also studied .
it is shown that the time of entanglement sudden death can be controlled by the classical driving fields .
particularly , the entanglement sudden death phenomenon will disappear if the classical driving fields are strong enough . |
the detection of an extremely bright prompt optical flash accompanying grb 990123 @xcite highlighted the possibility of using grb afterglows as backlighting sources to probe the high - redshift universe and the reinonization era @xcite .
this promise was finally fulfilled with the discovery of grb 050904 by the _ swift _
satellite @xcite , lying at a redshift of @xmath0 @xcite .
not only was a bright nir afterglow discovered for this grb @xcite , a prompt flash contemporaneous to the @xmath1-ray emission was seen by the robotic 25 cm tarot telescope @xcite .
had the prompt optical flash of grb 990123 been at a redshift of @xmath2 and had it been unextinguished by any dust , it would have peaked at an apparent magnitude km s@xmath3 mpc@xmath3 , @xmath4 , @xmath5 @xcite . for @xmath0 ,
this leads to a distance modulus of @xmath6 mag . ]
this corresponds to an absolute @xmath8-band magnitude @xmath9 ( @xmath10 , assuming the @xmath11 color for a g2v star from * ? ? ?
* ) , making this the most luminous optical source ever detected at that time . in this paper
, we compile all available data on the afterglow of grb 050904 and construct a synthetic light curve for a hypothetical perfectly ionized universe showing no lyman dropout up to the redshift of the grb .
we then derive light curve parameters and use the spectral energy distribution ( sed ) of the afterglow to extrapolate the afterglow light curve into the @xmath8 band assuming @xmath2 .
we find that the prompt flash of grb 050904 was the most luminous optical / nir transient ever detected , even exceeding the @xmath8-band luminosity of grb 990123 at peak by more than one magnitude .
lccccccccccc 1 & 59.2 & 17 & @xmath12 & @xmath13 & @xmath14 & & @xmath15 & & @xmath16 & + 2 & 59.7 & 17 & @xmath17 & & @xmath18 & @xmath19 & & @xmath20 & & @xmath21 + h06 & & & & @xmath22 & @xmath23 & & @xmath24 & & & + t05 & & & & & @xmath25 & @xmath26 & & @xmath27 & & + [ lc ]
the optical / nir afterglow of grb 050904 was observed by different telescopes over almost five decades in time and more than six decades in flux density . due to its very high redshift
, it was not detected at wavelengths shorter than the @xmath28 band .
we compiled all data from the following papers : @xcite , @xcite , @xcite , @xcite , @xcite and @xcite .
all data were corrected for galactic extinction assuming @xmath29 @xcite , and then transformed into ab magnitudes @xcite . to transform the magnitudes into flux densities , we used the zero points given in @xcite to transform their data and standard zero points for all other data .
data in @xcite , @xcite and @xmath30 data from @xcite are already ab magnitudes .
the highest data density is available in the @xmath31 band .
therefore , we started with the @xmath31 band to construct a synthetic light curve by achromatically shifting the other bands to the @xmath31-band light curve .
this method is justified , as no systematic color evolution was found in the data set . for the normal forward shock evolution of the afterglow ( once the cooling frequency @xmath32 has passed the optical / nir bands ) such achromacy is expected .
while the early steep - to - shallow transition at @xmath33 days may be chromatic , we only have @xmath31 band data at this time anyway . in the following
, we will always assume an achromatic evolution of the afterglow .
we fitted the @xmath31-band data with a broken power - law ( for more details on the fitting procedure , see * ? ? ?
* henceforth z06 ) . as @xcite have noted , the early @xmath31-band data before 0.3 days are brighter than the extrapolation of later data to these times , indicating the steeper decline from what could be a reverse shock flash . initially , we excluded these early data from our fit .
we used the derived light curve parameters from the fit to the @xmath31-band light curve as a reference light curve and fitted the @xmath34-band data with this reference curve .
we thus found a color @xmath35 .
assuming @xmath36 , we shifted the @xmath34-band data to the @xmath31 zero point , merging the light curves .
we then fitted the combined @xmath37 light curve again and shifted the @xmath34-band data points in small increments until @xmath38 is minimized , thus deriving the true @xmath35 color and the light curve parameters of the joint light curve which spans a longer time range ( 0.3 to 12 days in the observer frame ) and gives a better constraint on the post - jet break ( after 2.6 days ) decay slope @xmath39 ( it is @xmath40 ) .
the @xmath41-band data ( where we make no distinction between @xmath41 , @xmath42 and @xmath43 ) and the @xmath44-band data were added in the same way .
these additional shifts were not larger than 0.03 mag .
we transformed the hst nicmos f160w ( ab ) data point from @xcite to @xmath34 ( ab ) by using the conv_ab value from the @xmath45 package @xcite , @xmath46 .
as the final hst data point is corrected for host contribution , we also corrected the other data points in the same way . from the synthetic host spectrum presented in @xcite , we derived for the host galaxy a @xmath31-band flux density of 0.055 @xmath47jy , which transforms into @xmath48 .
we assumed conservative errors of 0.3 mag .
the effect of host subtraction is very small , at most 0.03 mag , since the last afterglow data point at 5 days is still 4 magnitudes brighter than the host galaxy .
[ j ] the @xmath49 composite light curve consists of an early steep decay @xcite , which goes over into a typical afterglow decay , followed by a further break , which is identified as a jet break @xcite .
we fitted the two breaks separately , using the equation from z06 ( fixing the host flux to zero as the data have been corrected for host contribution ) .
for the first fit , we excluded all data beyond 3 days from the fit . the break smoothness parameter @xmath50 ( cf .
z06 ) had to be fixed to @xmath51 ( a negative value , as the break is from steep to shallow ) .
the jet break was fitted with @xmath50 left as a variable in the fit . for this fit , data earlier than 0.3 days were excluded .
we thus derived three decay constants , which we label @xmath52 ( the steep early decay ) , @xmath53 and @xmath39 ( the typical pre- and post - jet break decay ) .
we then transformed all data from ab back to vega magnitudes using standard zero points .
the @xmath30 data @xcite are already affected by ly@xmath54 damping . while we used the composite reference light curve to find the @xmath55 color , we did not implement the @xmath30 data in the derivation of the light curve parameters .
we also did not implement the subluminous @xmath56 measurement from @xcite .
@xcite give the tarot measurements both as flux density at 9500 and as @xmath28 magnitudes .
we used the flux density and transformed it to @xmath30 magnitudes . comparing these with the @xmath28 magnitudes of @xcite
, we find @xmath57 mag .
this is in very good agreement with the @xmath58 color of 2.48 mag derived from late afterglow data at 1.2 days , where @xmath59 is the vlt fors2 @xmath60-band filter ( for details on the @xmath61 and @xmath59 filters , see * ? ? ?
the @xmath55 color derived at late times was then used to shift all @xmath30 data points to the @xmath31 zero point ( as always , we assume achromacy , i.e. , a constant spectral slope ) .
the result is a final composite light curve which includes all data from the early prompt emission to the late hst detection . in the following ,
when we speak of the @xmath31-band light curve , we always refer to this composite light curve .
the results of the light curve fits using the @xmath49 light curve are given in table [ lc ] , and the complete composite light curve is shown in fig .
as expected , @xmath53 derived from the two different fits ( cf . [ sec22 ] ) agrees within errors .
there is some scatter in the light curve data which leads to the high values of @xmath38 .
this is either due to additional measurement uncertainties or small deviations from a power - law decay , as it has been found that the x - ray afterglow was extremely variable @xcite .
the values we derived are in full agreement with those of other authors ( table [ lc ] ) .
these light curve parameters are typical for afterglows ( see z06 for a compilation of light curve parameters of all pre-_swift _ afterglows ) . the light curve steepening @xmath62 is high but also not extraordinary ( z06 , their fig .
we note that this is one of the few light curves that allow the smoothness of the jet break to be fitted as a free parameter , and the result is in full agreement with the potential relationship between @xmath53 and @xmath50 found by z06 ( their fig . 8) . for our fit ,
the extrapolation to very early times meets the earliest tarot detection at 0.002 days , similar to what @xcite find .
however , the following plateau phase around 0.001 days and the flare are much brighter than the extrapolated light curve ( 1 and almost 3 magnitudes , respectively ) .
this implies that the decay after the optical flare must be steep .
taking into account the upper magnitude limit at 0.09 days @xcite , we find @xmath63 .
such a steep decay implies that the flare could be either due to reverse shock emission or internal shocks from late central engine activity @xcite .
since this is even steeper than @xmath52 , an additional break must exist , but no data are available during this time span . in particular , we find that the afterglow peaks at @xmath64 , which is among the brightest nir afterglows ever detected and is a first hint at its extreme luminosity .
[ spec ] clcc @xmath65 & none & @xmath66 & @xmath67 + @xmath65 & mw & @xmath68 & @xmath69 + @xmath65 & lmc & @xmath70 & @xmath71 + @xmath65 & smc & @xmath72 & @xmath73 + h06 & none & @xmath74 & @xmath67 + t05 & none & @xmath75 & @xmath67 + t05 & none & @xmath76 & @xmath67 + p06 & none & @xmath77 & @xmath67 [ sed ]
[ z ] to construct the sed , we used the colors @xmath35 , @xmath78 and @xmath79 that were derived in [ sec21 ] ( @xmath44 , @xmath31 , @xmath34 and @xmath41 data are not affected by lyman @xmath54 damping ) .
the sed is fitted as described in k06 .
first , we find an observed spectral slope @xmath80 .
then the sed was fitted with an additional dust extinction curve , using dust models for the milky way , the large and the small magellanic cloud @xcite .
this gives us the extinction in the @xmath81 band in the host galaxy frame @xmath82 and an intrinsic spectral slope @xmath83 , corrected for the host extinction . to extend the sed further into the optical bands ( in the observer frame ) , we used reported @xmath84 upper limits .
by shifting the composite light curve to these limits , we produced additional @xmath85 , @xmath86 and @xmath87 colors which are lower limits on these colors . instead of a flat or downward curved sed , we actually found a slight upward curvature , mainly due to the @xmath44-band data ( which have large errors ) being slightly too bright ( fig .
2 , this can also be seen in figs . 2 and 3 of * ? ? ?
this results in a fit with smc dust , which is typically favored for grb host galaxies ( k06 , and references therein ) , finding a negative extinction , which is unphysical .
the best result is actually obtained for milky way dust , even though this does not imply that the host galaxy of grb 050904 has dust resembling that of the milky way . while smc dust is characterized by strong far - ultraviolet extinction , mw dust has a much flatter extinction curve at these wavelengths .
indeed , @xcite have found dust with flat fuv extinction in a quasar at almost the same redshift as grb 050904 , and @xcite find that the optical and x - ray afterglow of grb 020405 can be explained by dust with similar properties , although this grb lies at a much closer redshift .
the characteristic 2175 bump of mw dust is not visible due to the very low extinction value we derive .
we thus conclude that the extinction in the host galaxy is very low and the dust is dissimilar to smc - type dust . in the following , we will use the values derived from the mw dust fit , deeming this as a conservative approach .
the results of the sed fitting are given in table [ sed ] .
once again , our fits are in good agreement with those of other authors , except for @xcite , who ascribe the low value they derive to possible short term fluctuations . @xcite and @xcite also note that while source frame dust extinction may be present , it must be low .
the hydrogen column density along the line of sight in the host galaxy derived via optical spectroscopy is large ( log @xmath88 , * ? ? ?
coupled with the very low extinction , this implies a dust - to - gas ratio that is even lower than for the smc .
the result is similar to grb 990123 , and in this case also , the prompt flash may have burned dust along the line of sight ( cf .
fig . 6 in k06 , and references therein ) .
knowledge of the light curve parameters and the intrinsic spectral slope allows us to derive conclusions on the nature of the circum - burst environment via closure relations @xcite .
we find that using @xmath53 and @xmath83 , a model is strongly favored where the cooling break frequency @xmath32 lies redward of the source frame ultraviolet after 0.4 days .
this is substantiated by the spectral slope of the late x - ray afterglow , which is identical to that in the ultraviolet within errors .
@xcite find @xmath89 , @xcite find @xmath90 .
furthermore , @xcite used their radio detections of the afterglow of grb 050904 to perform broadband modeling and arrived at the same conclusion . in this case , one can not draw any conclusion on whether the grb environment is of constant density ( ism ) or is shaped by a progenitor wind .
the observed very soft jet break ( @xmath91 , table [ lc ] ) may be indicative of a wind environment @xcite , whereas @xcite use an ism model .
finally , @xcite propose that the afterglow passes through a termination shock somewhere before 0.5 days in the observer frame , switching from an early wind medium to a late ism medium .
given the location of @xmath32 redward of the optical / nir bands , an electron spectral index of @xmath92 is found , which is typical for grb afterglows ( e.g. , * ? ? ?
* and references therein ) .
the standard fireball model predicts @xmath93 @xcite , consistent with the @xmath39 we find within @xmath94 ( table [ lc ] ) .
in the observer frame @xmath95 band , the afterglow was unobservable due to lyman damping .
typically , grb afterglows have the highest data density in the @xmath95 band , and all afterglows that were examined in k06 were @xmath95-band afterglows . because of the synchrotron nature of the afterglow radiation ( described by a power - law , @xmath96 , with @xmath83 being the spectral slope ) , and assuming @xmath97const .
, we can construct the @xmath95-band light curve from the @xmath31-band light curve .
the spectral slope we derived from the observer frame nir data can be extrapolated into the observer frame optical , allowing us to find a @xmath85 color without the damping influence of lyman @xmath54 .
for this purpose , we used the observed spectral slope that has not yet been corrected for host galaxy extinction , @xmath80 .
since the host galaxy extinction is very low , the spectral curvature is negligible and we can use a simple power - law for the extrapolation .
this power - law has an error from the fitting process , which we added in quadrature to the measurement errors of the data points . in order to place the afterglow of grb 050904 in the context of the afterglow sample discussed in k06
, we shifted the afterglow to a redshift @xmath2 . using the derived intrinsic spectral slope @xmath83 , the source frame extinction @xmath82 ( both from the mw dust fit ) and the redshift , we computed the magnitude shift denoted @xmath98 in k06 .
this shift takes into account the extinction correction , the redshift difference and cosmological @xmath99-corrections @xcite when moving the source from its original redshift to @xmath2 .
the magnitude shift creates a further error ( from @xmath100 and @xmath101 ) which we again added in quadrature to the errors of the data .
we find @xmath102 mag , which is higher than any value found in k06 . applying this correction to the @xmath95-band afterglow
, we finally arrive at our main result : had grb 050904 lain at @xmath2 and been unaffected by any line - of - sight extinction ( including lyman @xmath54 damping ) , the optical flare detected by tarot would have peaked at @xmath103 , almost visible to the naked eye .
this is more than one magnitude brighter than the reverse shock peak of grb 990123 ( the flux density is higher by a factor of 2.8 ) , making the afterglow of grb 050904 the most luminous optical transient ever detected .
@xcite , neglecting @xmath99-correction , find a peak flux density of 1080 mjy ( in the @xmath81 band ) for the afterglow of grb 990123 and 1300 mjy ( in the @xmath28 band ) for the afterglow of grb 050904 after scaling it to the redshift of grb 990123 , but do not remark upon this further except to call the two events comparable . for @xmath104 ,
the absolute @xmath95-band magnitude is @xmath105 ( @xmath106 ) , making the afterglow in the @xmath95 band alone between 60 and 600 times brighter than the intrinsic bolometric luminosity of the gravitationally lensed
hyperluminous bal quasar apm 08279 + 5255 , the most luminous persistent source known @xcite .
the complete composite @xmath107band light curve is shown in comparison with other afterglows in fig .
it is immediately apparent that the optical afterglow of grb 050904 was much brighter than any other afterglow over almost two decades in time . at the peak of the prompt flash , it was six magnitudes brighter than grb 021004 , which is the brightest afterglow at late times , and another six magnitudes brighter than grb 041006 , one of the faintest grbs in the sample .
this span is much larger than at one day after the grb , where the afterglow luminosities tend to cluster ( k06 ; * ? ? ?
* ; * ? ? ?
* ) . at later times ,
though , it resembles other afterglows . at one and four days after the burst ( fig .
3 ) , we find @xmath108 and @xmath109 , respectively , which transforms to absolute magnitudes @xmath110 and @xmath111 , respectively ( cf . k06 , their figs . 9 and 10 ) .
this ranks it among the brightest afterglows at one day , but it is not the brightest , as it has also been noted by @xcite . at four days after the burst , it lies close to the center of the luminosity distribution . since no multi - color data for the prompt emission phase are available , we shifted the early optical emission achromatically , assuming no spectral changes between this phase and the late afterglow phase ( where no color evolution is evident ) .
the assumption of achromacy for the first 100 seconds after a grb is surely an oversimplification .
strong spectral changes are expected within this time @xcite .
however , does this imply that the true luminosity of the flare was actually lower than what we have derived ?
the spectral slope of the late optical / nir afterglow , consistent with the cooling break lying at longer wavelengths than the optical , is very steep , @xmath112 . during the prompt emission
, the forward shock afterglow is expected to be in the fast cooling phase ( * ? ? ? * and references therein ) .
if the medium surrounding the grb progenitor is of constant density ( ism model ) , the optical band lies between the cooling frequency @xmath32 ( at longer wavelengths ) and the peak frequency @xmath113 , the expected spectral slope is @xmath114 . in case
even the cooling frequency has not yet passed the optical bands , the spectral slope in the optical actually rises , with @xmath115 ( * ? ? ?
* and references therein ) .
a similar situation would be seen if the early optical emission is dominated by the low energy tail of the prompt emission , a situation which was seen for grb 041219a and grb 050820a @xcite .
this has also been suggested for grb 050904 @xcite , although the optical flux in the flare is much brighter than the extrapolation of the x - ray spectrum for grb 050904 @xcite , in contrast to the cases of grb 041219a and grb 050820a @xcite .
we thus conclude that the spectrum in the optical bands during the prompt emission was not steeper than at late times , and possibly shallower , i.e. @xmath116 .
on the one hand , this implies that the @xmath95 magnitude of the afterglow at @xmath0 would be brighter than what we derived with the spectral slope of the late afterglow . on the other hand ,
when we apply the cosmological shift to @xmath2 , the @xmath98 shift would become smaller .
we find @xmath117 mag for @xmath114 and @xmath118 mag for @xmath115 ( also correcting for the small host extinction ) , resulting in @xmath119 and @xmath120 , respectively .
thus , for @xmath115 , the extrapolated @xmath95-band magnitude at @xmath2 of the flare would be comparable to that of grb 990123 .
however , we should also note that the same argument may apply to the flare of grb 990123 , so that the flare of grb 050904 truly is brighter with a high probability .
in the literature on grb 050904 , consensus is not reached on whether grb 050904 is , apart from its redshift , a special event .
let us review some aspects of both the prompt emission and the afterglow : in many ways , grb 050904 does not differ from the typical mean values found for grbs . in the source frame , @xmath121 seconds , with @xmath1-ray emission detected up to 69 seconds after the trigger , including the flare seen prominently in the x - rays and the optical @xcite .
the intrinsic luminosities of the x - ray @xcite , the late optical ( as shown in this work ) and the radio afterglow @xcite are typical , as are the light curve decay parameters and the jet break time in the rest frame ( z06 , k06 ) .
but some aspects of grb 050904 do stand out .
the peak energy @xmath122 lies beyond the _ swift _ bat energy range . from joint
fitting of _ swift _ bat and _ suzaku _ wam data , @xcite report a _ preliminary _
@xmath123 kev , translating into @xmath124 kev in the rest frame .
this peak energy is the highest found in the compilation of 56 bursts @xcite , only grb 050717 ( observer frame @xmath125 kev , time - integrated spectrum , * ? ? ?
* ) very probably has a higher peak energy ( the redshift is not known but assumed to be high ) .
the isotropic energy is not exactly known , @xcite derive @xmath126 erg .
the upper end of the range is higher than for any other burst , including grb 990123 @xcite , and even the lower value is higher than for most grbs with well - determined @xmath127 . using a jet break time of @xmath128 days , @xcite derive log @xmath129(erg)@xmath130
this is the second highest beaming - corrected energy ever found @xcite , after grb 050820a ( log @xmath129(erg)@xmath131 , * ? ? ?
* ) .the x - ray afterglow is the most variable ever seen by _ swift _
@xcite , implying that , in the source frame , the central engine was active for several hours after the burst @xcite .
we have shown in this paper that the early optical / nir emission is brighter than for any other grb detected so far .
@xcite derive a high circum - burst density of 680 @xmath132 from broadband afterglow modeling , noting that a density typical for grb environments ( 10 @xmath132 ) is ruled out by the absence of a bright early radio afterglow .
there is no aspect of grb 050904 that exceeds the typical grb parameters so much that one could claim that the progenitor of grb 050904 was markedly different from those of all other long grbs , e.g. a population iii star .
however , it is clear that grb 050904 is one of the most extreme grbs ever observed , comparable only to grb 990123 ( cf .
* ) , and exceeding even this seminal event in some aspects .
we have shown that the early flare of grb 050904 detected by the tarot telescope is the most luminous optical / nir transient ever detected , exceeding even the peak magnitude of the prompt optical flash of grb 990123 . while grb 050904 is not markedly different from other grbs in any single aspect , it is among the most extreme grbs ever detected .
so far , only a single grb at the end of the reionization era has been detected .
it is thus not possible to decide whether grb 050904 is typical for grbs at a very high redshift , or if it is simple observational bias that the first grb to be found at @xmath133 is one of the most extreme events ever detected .
more typical events would not have sufficient follow - up to determine the necessary parameters , especially the redshift .
the _ swift _ satellite , with its high sensitivity in an energy band that is typical for highly redshifted hard bursts and its possibility of long - duration image triggers , is a very powerful instrument for detecting these very distant events .
probably , more very high redshift grbs will be detected during the lifetime of _ swift _ , possibly even breaking the redshift records for quasars @xcite and galaxies @xcite .
indeed , a grb at a higher redshift than grb 050904 may have already been found , grb 060116 @xcite .
while the detection of a grb by _ swift _ is the first and most important step , most parameters characterizing the event can only be obtained by ground - based telescopes .
the determination of the true burst energetics depends on the measurement of the redshift and the jet break time , and grbs at @xmath133 are bright only in the nir .
while optical spectroscopy is still feasible to @xmath134 , any higher redshifts will move the metal lines that are used for an exact determination of @xmath135 beyond the typical limit of optical spectroscopy ( and the region beyond @xmath136 m is already strongly affected by sky lines ) .
nir spectroscopy even at large facilities needs bright sources , and thus far , nir spectroscopy of a grb afterglow has never been successful in detecting any lines .
the dilemma is that observing time at large facilities is precious and observations would only be triggered in case there is already a strong suspicion that the event lies at a very high redshift .
this implies the need for rapid photometric redshift determinations that use the lyman @xmath54 trough , as it has been successfully applied to grb 050904 @xcite .
one telescope that is already in use and has the capability for rapid multi - color photometry is the magnum telescope , which also observed grb 050904 @xcite .
a key element would be the deployment of more moderately large robotic nir telescopes such as pairitel @xcite and rem @xcite , which also has an optical camera .
we thank jochen greiner for useful comments and bruce gendre for a very constructive referee report .
d.a.k . and s.k .
acknowledge support by dfg grant kl 766/13 - 2 and by the german academic exchange service ( daad ) under grant no .
n.m . acknowledges support under crui vigoni programme no .
506 - 2005 .
akerlof , c. et al .
1999 , , 398 , 400 amati , l. 2006 , , 372 , 233 bloom , j. s. , starr , d. l. , blake , c. h. , skrutskie , m. f. , & falco , e. e. 2005 , proc .
astronomical data analysis software & systems xv
( editors c. gabriel , c. arviset , d. ponz and e. solano ) , submitted ( astro - ph/0511842 ) berger , e. et al .
2002 , , 581 , 981 berger , e. et al .
2006 , , submitted ( astro - ph/0603689 ) ber , m. 2006 , , 638 , l71 bolzonella , m. , miralles , j .- m . , & pell , r. 2000 , , 363 , 476 cenko , s. b. et al .
2006 , , in press ( astro - ph/0608183 ) chevalier , r. a. , & li , z .- y .
1999 , , 520 , l29 cusumano , g. et al .
2006 , , 440 , 164 cusumano , g. et al .
2006 , , in press ( astro - ph/0610570 ) ducati , j. r. , bevilacqua , c. m. , rembold , s. b. , & ribeiro , d. 2001 , , 558 , 309 fan , x. et al . 2003 , , 125 , 1649 frail , d. a. et al .
2006 , , 646 , l99 friedman , a. s. , & bloom , j. s. 2005 , , 627 , 1 gendre , b. et al .
2006 , , in press ( astro - ph/0603431 ) grazian , a. et al .
2006 , gcn 4545 , http://gcn.gsfc.nasa.gov/gcn/gcn3/4545.gcn3 haislip , j. b. et al .
2006 , , 440 , 181 ibata , r. a. , lewis , g. f. , irwin , m. j. , lehr , j. & totten , e. j. 1999 , , 118 , 1922 inoue , a. k. , yamazaki , r. , & nakamura , t. 2004 , , 601 , 644 iye , m. et al .
2006 , , 443 , 186 kann , d. a. , klose , s. , & zeh , a. 2006 , , 641 , 993 ( k06 ) kawai , n. et al .
2006 , , 440 , 189 krimm , h. a. et al .
2006 , , 648 , 1117 lamb , d. q. , & reichart , d. e. 2000 , , 536 , 1 liang , e. , & zhang , b. 2005 , , 638 , l67 maiolino , r. et al .
2004 , , 431 , 533 nardini , m. et al .
2006 , , 451 , 821 oke , j. b. , & gunn , j. e. 1983 , , 266 , 713 pei , y. c. 1992 , , 395 , 130 piranomonte , s. et al .
2006 , gcn 4583 , http://gcn.gsfc.nasa.gov/gcn/gcn3/4583.gcn3 price , p. a. et al .
2002 , , 572 , l51 price , p. a. et al .
2006 , , 645 , 851 sandage , a. 1988 , , 26 , 561 sari , r. , piran , t. , & narayan , r. 1998 , , 497 , l17 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 spergel , d. n. et al .
2005 , , 148 , 175 stratta , g. et al .
2005 , , 441 , 83 sugita , s. et al .
2006 , poster shown at the meeting of the royal astronomical society , london , september 2006 tagliaferri , g. et al .
2005 , , 443 , l1 tanvir , n. , levan , a. j. , priddey , r. s. , fruchter , a. s. , & hjorth , j. 2006 , gcn 4602 , http://gcn.gsfc.nasa.gov/gcn/gcn3/4602.gcn3
totani , t. et al .
2006 , , 58 , 485 vestrand , w. t. et al .
2005 , , 435 , 178 vestrand , w. t. et al .
2006 , , 442 , 172 watson , d. et al .
2006 , , 637 , l69 wei , d. m. , yan , t. , & fan , y. z. 2006 , , 636 , l69 wu , x. f. , dai , z. g. , huang , y. f. , lu , t. 2005 , , 619 , 968 zeh , a. , klose , s. , & kann , d. a. 2006 , , 637 , 889 ( z06 ) zerbi , r. m. et al .
2001 , astron .
nachr . , 322 , 275 zhang , b. , & mszros , p. 2004 , int.j.mod.phys .
a19 ( 2004 ) , 2385 zou , y. c. , xu , d. , & dai , z. g. 2006 , , 646 , 1098 | with a redshift of @xmath0 , grb 050904 is the most distant gamma - ray burst ever discovered .
it was an energetic event at all wavelengths and the afterglow was observed in detail in the near - infrared bands .
we gathered all available optical and nir afterglow photometry of this grb to construct a composite nir light curve spanning several decades in time and flux density .
transforming the nir light curve into the optical , we find that the afterglow of grb 050904 was more luminous at early times than any other grb afterglow in the pre-_swift _ era , making it at these wavelengths the most luminous transient ever detected . given the intrinsic properties of grb 050904 and its afterglow , we discuss if this burst is markedly different from other grbs at lower redshifts . |
when we consider the cosmological data from wmap , supernovae ia , galaxy clustering on large scales , and cross - correlations between galaxies and the cmb , we are faced with several possible conclusions .
one is that we live in a spatially flat friedmann - robertson - walker universe currently dominated by either a cosmological constant or repulsive dark energy .
the best fit for the dimensionless energy density parameters are @xmath1 and @xmath2 , within the concordance @xmath3cdm model @xcite .
the advantage of this model is that nearly all available observational data supports it ; the disadvantage is that it requires the vast majority of the energy density in the universe to be in two unknown substances , dark matter and dark energy .
this conclusion rests on the accuracy of our current gravity model , general relativity .
the key equation of general relativity , einstein s equation , relates the curvature and the expansion rate of the universe to its matter and energy content .
the current paradigm is to modify the matter content of the universe , by including dark matter and dark energy , to account for observations .
instead , however , we might modify how the universe curves in response to matter , which would mean modifying our theory of gravity .
there are many possible ways to modify gravity , depending on what one wishes to `` fix '' .
for example , mond @xcite removes the need for dark matter to account for galactic rotation curves and has several other interesting results , but seems to fail at the scale of galaxy clusters , with even optimistic accountings needing roughly as much dark matter as baryonic matter . at the other end
is something like dgp or conformal gravity , which hopes to account for the acceleration of the universe without invoking a cosmological constant @xcite .
in addition , there are other models which seek to unify dark matter and dark energy @xcite .
what we seek to do in this paper is to study the problem more phenomologically .
for example , regardless of how the force law for gravity is modified , it will often be stronger or weaker , relative to the standard model , at larger or shorter scales .
one way to parameterize this is to introduce a modified yukawa - like potential for a point mass : @xmath4 @xcite where @xmath5 indicates the strength and @xmath6 the scale ( constant in physical rather than comoving coordinates ) on which this modification is most relevant . on scales smaller than @xmath6
, @xmath7 reduces to the standard newtonian potential ; on larger scales it transitions to the newtonian potential multiplied by a factor of @xmath8 .
this is similar to the interaction considered in @xcite , in which a long range dark matter interaction is introduced yielding a different yukawa - like potential that instead modifies gravity on short length scales .
note that this is not a cosmological model : there is no prescription for determining things like the expansion factor and the resulting hubble factor . like previous workers
, we will assume that these are the same as in the standard cosmology .
the main goal of studying such a model is to gain intuition for some generic effects of modifications to standard gravity .
for example , in the linear theory of such a model , the growth of fluctuations is @xmath0-dependent @xcite it is not in standard gravity . as a result ,
a smooth spherical region within which the density is the same as the background universe will evolve .
this qualitatively different behavior from standard gravity has not been emphasized so it is worth showing the argument explicitly .
consider the density field smoothed on scale @xmath9 at some early time @xmath10 .
we can write this field in terms of its fourier modes , and the ( fourier transform of the ) smoothing kernel as @xmath11 the linearly evolved field is @xmath12 where @xmath13 is the linear theory growth factor . in standard gravity
, @xmath13 is independent of @xmath0 , so if @xmath14 then @xmath15 also .
but if @xmath13 depends on @xmath0 , then if @xmath16 at some time @xmath17 , it will , in general , be non - zero at other times ( the exception being if the @xmath0-dependence of @xmath18 happens to exactly cancels that of @xmath13 ) .
for the yukawa - like modification considered here , the @xmath0 dependence of the spherical top hat filter does not cancel that of @xmath13 .
thus , we are led to the rather remarkable conclusion that , when the gravitational potential has been modified , then linear theory predicts that a spherical tophat patch within which the density is the same as the background will evolve ! the reason why can be traced to the fact that birkhoff s theorem no longer applies once the newtonian potential has been modified . without this theorem , the spherical top hat filter is no longer special , and our common sense prejudice from standard gravity that initially overdense regions become denser , underdense regions less dense , but regions within which the density is the same as the background do not evolve must be treated with caution .
the evolution of nonlinear clustering in this model has been studied using numerical simulations by @xcite .
our goal in what follows is to provide a framework for understanding this nonlinear evolution more completely . to this end
, we will use the spherical evolution model , which has found extensive use in the standard model
it is used to motivate estimates of the abundance of nonlinear objects @xcite , a crucial ingredient in methods which describe the growth of nonlinear gravitational clustering @xcite .
it also provides a framework for discussing the environmental dependence of clustering @xcite .
section [ model ] summarizes what is known from the linear theory of this yukawa - like model , and then shows our estimate of the key , and sometimes subtle , changes to the spherical evolution model .
section [ sims ] compares this work to the simulations of @xcite .
a final section summarizes .
we begin by considering the evolution of density perturbations .
this can be done by either considering the fluid equations in expanding coordinates , or by considering the conservation of stress - energy @xmath19 .
if we start with a smooth background , add small perturbations , and linearize the resulting equations , we get a second order differential equation for the evolution of the fractional overdensity , @xmath20 that depends on time , scale factor @xmath21 , the hubble parameter @xmath22 , and the potential @xmath23 . in standard gravity
, we would use the poisson equation to set the relationship between @xmath23 and @xmath24 , but here we will assume a modified poisson equation that results in the above potential , equation ( [ initialpotential ] ) .
@xmath25 it is easier to work with the fourier transform of this equation : @xmath26 \left(1-\frac{h_0 ^ 2 \omega_\lambda}{h^2 } \right ) h^2 \delta_k.\ ] ] this can be solved relatively easily to determine the linear growth of structure associated with equation ( [ initialpotential ] ) .
cdm , at @xmath27 , when @xmath28 mpc and @xmath29 ( top ) and @xmath30 ( bottom ) .
dashed lines show this ratio for the model in equation ( [ initialpotential ] ) , and solid lines for equation ( [ potentialeq ] ) . for the solid lines , @xmath31 mpc , and dotted @xmath32 mpc.,title="fig : " ] cdm , at @xmath27 , when @xmath28 mpc and @xmath29 ( top ) and @xmath30 ( bottom ) .
dashed lines show this ratio for the model in equation ( [ initialpotential ] ) , and solid lines for equation ( [ potentialeq ] ) . for the solid lines , @xmath31 mpc , and dotted @xmath32 mpc.,title="fig : " ] the dashed lines in figure [ fig : linear ] show how this growth differs from that in the standard model .
note in particular that @xmath33 , whereas the standard model has no @xmath0 dependence in the growth factor @xmath34 .
the figure shows the effects that one expects to see , namely that for negative @xmath5 the growth factor is smaller at small @xmath0 ( large scales ) , whereas for positive @xmath5 the growth factor is larger at large scales .
both have a region where they deviate from being scale independent , until on very large scales they return to scale independence , though with a different value than in general relativity .
while this is not a significant problem for linear theory , we decided to explicitly force the potential back to normal gravity at large scales for reasons which will become clear in the next section , briefly , because we want to assume that the cosmological model is indistinguishable from @xmath3cdm , on the largest scales we want gravity to be the same as in @xmath3cdm .
the potential we are using from this point onward is : @xmath35 with this potential , the @xmath36 term serves to make explicit the return to normal gravity at large scales .
the linear theory equation becomes @xmath37 \nonumber\\ & & \qquad \times \left(1-\frac{h_0 ^ 2 \omega_\lambda}{h^2 } \right ) h^2 \delta_k,\end{aligned}\ ] ] and the linear growth associated with this solution is shown as the solid lines in figure [ fig : linear ] , when @xmath31 mpc .
the dotted line shows what happens if @xmath32 mpc a scale which is large compared to that probed by baos .
although the analysis which follows uses @xmath31 mpc , the results which follow are not sensitive to this choice .
excursion set methods @xcite are used to estimate the number density of collapsed haloes , the merger rates of haloes , the conditional mass function of progenitors as a function of final halo mass and time , and the nonlinear counts - in - cells distribution , all of which help us to link what we observe about the properties of galaxies and galaxy clusters with cosmology .
essentially , excursion set methods relate the properties of halos today to the initial density fluctuation field .
an advantage of such methods is that we only need a few things in order to use them : one is an assumption that the initial fluctuations are small and gaussian , the other is a model for determining how dense something must have been initially to collapse at a given time .
the first is given to us by wmap , the second is more difficult .
a simple model for how to determine this critical density is given by the spherical evolution model , in which one considers a spherical tophat perturbation in the initial density fluctuation field . in the standard model , the gravitational evolution of such a patch
is determined only by the mass within it , and so one can determine how overdense such a patch needs to be initially in order to collapse by a given time .
this critical overdensity @xmath38 generically depends on the background cosmology @xcite .
the spherical collapse calculation begins with the statement that the force driving the acceleration is related to the gravitational potential by @xmath39 this can be integrated once to get @xmath40 where @xmath41 is the integral of the potential over the mass distribution , and @xmath42 is the total energy of the patch , which is constant .
in standard gravity , the potential of a shell of mass @xmath43 is the same as that of a point mass at the center of the sphere , so @xmath41 reduces to @xmath44 .
the constant @xmath42 can be related to the initial overdensity and/or expansion rate of the patch : the initial expansion rate is given by the hubble expansion rate of the background in which the patch is embedded , namely in comoving coordinates @xmath45 , so @xmath46 , so @xmath47 . including a cosmological constant presents no conceptual difference . in standard gravity
one can directly solve this equation .
the solution is a cycloid for which the critical value of the initial overdensity required for collapse , @xmath38 , does not depend on the initial size of the patch .
this scale independence of @xmath38 is a result of birkhoff s theorem : the evolution of a tophat sphere is the same as that given by friedmann s equations , so the actual size of the patch drops out . when gravity is modified , things are no longer so simple .
for example , when the potential is given by equation ( [ potentialeq ] ) , birkhoff s theorem no longer applies : a particle offcenter in a uniform spherical shell will feel a force from the shell because we no longer have a @xmath48 force law .
this has two consequences .
first , equation ( [ f = gradphi ] ) can still be integrated once to get equation ( [ solveeq ] ) , only now @xmath41 has contributions from both the internal and external mass distributions .
we can get @xmath49 by integrating equation ( [ potentialeq ] ) of the mass distribution , leaving @xmath42 and the initial value for @xmath50 to be determined .
as before , @xmath42 is the total energy ( constant in time ) , and we set @xmath47 .
( this was why we used equation [ potentialeq ] rather than equation [ initialpotential ] . )
second , whereas evenly spaced concentric shells remain evenly spaced in the standard tophat model , this is no longer the case when the potential is modified . as a result , the initial tophat perturbation develops a nontrivial density profile as it collapses . , with @xmath51 ( top ) and @xmath52 ( bottom ) .
the thick solid lines are actually the positions of the shell on the edge of the density perturbation for 3,10 , 20 , 40 , 50 , 60 , 80 , 100 , and 200 shells , note they all lie nearly on top of each other .
the dashed curves show the evolution when @xmath53 .
, title="fig : " ] , with @xmath51 ( top ) and @xmath52 ( bottom ) .
the thick solid lines are actually the positions of the shell on the edge of the density perturbation for 3,10 , 20 , 40 , 50 , 60 , 80 , 100 , and 200 shells , note they all lie nearly on top of each other .
the dashed curves show the evolution when @xmath53 .
, title="fig : " ] of course , neither of these changes prevent us from solving for the evolution of @xmath54 .
our main interest in what follows is not in the details of how the density profile is modified ( this is interesting in its own right ) , but in the modification to the critical density required for collapse . to estimate this in practice ,
two things require care , both of which are related to the breakdown of birkhoff s theorem .
the first is that , because the shape of the perturbation evolves , one must follow the evolution not just of a shell at the boundary of the perturbation , but of a series of concentric shells .
so the question is : how finely spaced must the shells be before one converges to the correct solution ?
the second is that one now cares not only about the mass initially interior to the initial boundary , but the mass exterior as well . in this case
, the question is : how far beyond the initial boundary of the perturbation matters before one reaches convergence ? therefore we start with an initial patch which is substantially larger than that within which there is an initial overdensity , and use a simple 1-dimensional n - body simulation to solve for @xmath54 .
we found that volumes having twice the initial comoving radius or larger were sufficient to account for the lack of birkhoff s theorem , regardless of @xmath5 or @xmath6 .
we also found that for objects of mass up to @xmath55 , using 3 ( linearly spaced ) wide shells provided a @xmath38 that was within 1% of 200 shells ( see figure [ fig : denplot ] ) .
for higher mass objects we found that we needed more than 3 shells , but by 40 shells @xmath38 is within 0.02% of @xmath38 calculated with 200 shells .
cdm , @xmath28 mpc and @xmath31 mpc . from top to bottom , curves show models in which @xmath56 and @xmath57 ( note that @xmath53 is standard gravity ) . ] having determined that our numerical solution had converged , we evaluated @xmath54 at the present time in an @xmath3cdm background model , for a grid of initial sizes and overdensities . by finding which pairs of initial density and size @xmath58 when evolved result in @xmath59 , we obtained @xmath60 .
since the initial overdensity is always small , we can use the fact that @xmath61 to express this critical density as a function of mass rather than initial radius .
this is shown in figure [ fig : barriershapes ] .
notice that @xmath38 depends on mass ; this is not unexpected , because patches which remain smaller than @xmath6 throughout their evolution ( and become small mass halos ) are unlikely to notice any modification , whereas those which are larger than @xmath6 at any time during their evolution will .
this mass - dependence of @xmath38 means that we expect to see a variation in cluster abundances only at masses larger than @xmath62 .
this is a consequence of the fact that @xmath63 mpc , for which the mass scale is about @xmath64 .
( for fixed @xmath5 , the mass scale is proportional to @xmath65 . ) , as a function of mass @xmath66 for models with @xmath29 ( short dashed ) and @xmath30 ( long dashed ) , the solid line is for @xmath53 .
the rapid rise of the negative @xmath5 barrier in figure [ fig : barriershapes ] is why there is a larger effect on the abundance of high mass halos for negative @xmath5 than positive @xmath5 . ]
figure [ fig : barriershapes ] shows that , when the potential is modified , then @xmath38 is no longer scale - independent .
because it depends on mass , the relevant excursion set problem is one with a ` moving ' rather than ` constant ' barrier , so it is of the type first studied by @xcite . figure [ fig : comparison ] shows the result of using this formalism to estimate the abundance of virialized objects .
briefly , making this estimate requires that one generate an ensemble of random walks in the @xmath67 plane , where @xmath68 is the variance in the initial fluctuation field when smoothed on scale @xmath58 . since @xmath69 is a monotonic function of @xmath43 , the variables @xmath70 , @xmath43 and @xmath58 are essentially equivalent to one another . in particular , @xmath71 where @xmath72 .
one then finds the first crossing distribution @xmath73 of the ` barrier ' @xmath74 .
the abundance of objects is @xmath75 .
figure [ fig : comparison ] shows that the abundance of massive objects increases as @xmath5 increases , whereas the abundance of low mass objects is essentially unchanged from when @xmath53 .
we argued above that this makes intuitive sense : the lower mass objects do not feel the change in gravity because they were smaller than @xmath6 throughout their evolution ; the more massive objects are able to become even more massive if @xmath5 is positive , since then gravity is stronger . before moving on ,
it is worth noting that we were careful to describe and implement the excursion set approach in initial rather than linearly evolved variables . in standard gravity ( @xmath53 ) , it is more common to phrase the discussion in terms of linearly evolved variables .
since the linear growth factor is independent of @xmath0 when @xmath53 , this is straightforward .
however , this is no longer the case when @xmath76 , because of the @xmath0-dependence in @xmath77 .
indeed , the barrier shape in figure [ fig : barriershapes ] is qualitatively like that of the linear growth factor in figure [ fig : linear ] , so one might ask if the difference that we see in the mass function is entirely a consequence of the @xmath0-dependence of the linear growth factor . more specifically , the differences shown in figure [ fig : comparison ] are really a consequence of two effects : first , we have assumed that @xmath78 is the same for all @xmath5 . therefore , @xmath79 is not : the linear theory evolution when @xmath80 results in more large scale power than when @xmath81 , so the rms fluctuation in the present day fields are different in the jargon , @xmath82 at @xmath83 is larger for the @xmath80 models . since we know that , in standard gravity , the abundance of massive halos is exponentially sensitive to @xmath82
, one might wonder if this alone accounts for much of the effect .
( later on , we will discuss another consequence of normalizing the models at the initial rather than final time . ) to quantify this , we would like to compare the predicted abundances when the models are normalized so that @xmath79 , rather than @xmath78 , is the same .
this will isolate the effect of a mass - dependent @xmath84 on the halo abundances .
( top ) and @xmath51 ( bottom ) to that when @xmath53 .
solid curve shows this ratio when the excursion set method is correctly implemented , using the initial fluctuation field values @xmath78 and a moving barrier @xmath85 , and @xmath78 is independent of @xmath5 .
dashed curve uses @xmath79 from the linearly evolved field and a constant barrier @xmath86 .
dotted curve uses @xmath78 and @xmath85 , but now @xmath78 is modified so that @xmath79 is the same for both values of @xmath5 . note that all these approaches produce different results .
, title="fig : " ] ( top ) and @xmath51 ( bottom ) to that when @xmath53 .
solid curve shows this ratio when the excursion set method is correctly implemented , using the initial fluctuation field values @xmath78 and a moving barrier @xmath85 , and @xmath78 is independent of @xmath5 .
dashed curve uses @xmath79 from the linearly evolved field and a constant barrier @xmath86 .
dotted curve uses @xmath78 and @xmath85 , but now @xmath78 is modified so that @xmath79 is the same for both values of @xmath5 . note that all these approaches produce different results .
, title="fig : " ] therefore , we used the excursion set approach with the same constant barrier height as one would have in the standard ( @xmath53 ) linearly evolved gravity model , and then evaluated @xmath79 , rather than @xmath78 , using the @xmath0-dependent linear growth factor .
i.e. , @xmath87 the resulting first crossing distribution @xmath88 is the same as in standard gravity ( after all the barrier is constant ) , but when expressed as a function of mass @xmath43 , the abundances are different because the relation between @xmath89 and @xmath43 depends on @xmath90 .
the dashed line in figure [ fig : comparisonlinear ] shows that this method yields a mass function that also differs substantially from the standard one : it drops substantially below unity for large @xmath43 . in this case , however , the drop is entirely due to the @xmath0-dependent growth factor .
in addition , notice that although it is qualitatively like the solid line , for which @xmath38 is mass dependent , it can be substantially different ( i.e. , the ratio of the solid to the dashed line is greater than unity ) at large @xmath43 .
this shows that there is more to the change in the mass function than simply the change in the relation between @xmath91 and mass . whereas the solid line shows results in which the initial power spectra are the same for all @xmath5 ( i.e.
, @xmath78 is the same for all models ) , the dashed line shows what happens if we adjust the shape of the initial power spectrum in the @xmath53 model so that @xmath79 is the same as for the @xmath30 model .
the dotted line shows the result of adjusting instead the initial @xmath92 of the @xmath30 model so that it produces the same @xmath79 as the original @xmath53 calculation . in this case , the @xmath30 initial conditions now have substantially more large scale power , so the predicted abundance of massive halos is larger , until the mass dependence of @xmath38 begins to matter ( this is not evident in the figure , because it happens at @xmath93 ) .
the fact that none of the curves shown in figure [ fig : comparisonlinear ] are unity for all @xmath43 , nor are any two curves the same , means that cluster abundances in modified gravity models can not be mimicked in standard gravity simply by changing the shape of the initial power spectrum so that it agrees with modified linear theory at @xmath83 .
trading ` cmb'-normalization for ` cluster ' normalization does not work , because the cluster mass function depends on the nonlinear physics of gravitational collapse : cluster counts are sensitive to more than the change to linear theory . for cmb - normalized models
, we feel that the appropriate estimate of the effect of modifying gravity is shown by the solid line . in the following section we use numerical simulations to test this prediction .
mpc ratioed to @xmath53 . in all panels ,
data points are from simulation , whereas curves are from our excursion set calculation . in the top left panel , ( blue )
long dashed and ( red ) short dashed curves are for @xmath51 and @xmath94 , respectively , and all models had the same initial fluctuation spectrum ( so that @xmath53 has @xmath95 today ) .
the top right panel shows the ratio of counts in two @xmath53 runs , but with different initial conditions , taylored so that @xmath82 at @xmath83 is the same as in the @xmath51 ( long dashed ) and @xmath94 ( short dashed ) runs .
short dashed curves in bottom panel show the ratio of the @xmath51 counts to that in the @xmath53 run when it has been normalized to have the same ( linear theory ) power spectrum at @xmath83 as the @xmath51 run .
long dashed curve shows a similar analysis of the @xmath30 case .
this panel shows the ratio of the numerators in the previous two panels .
, title="fig : " ] mpc ratioed to @xmath53 . in all panels ,
data points are from simulation , whereas curves are from our excursion set calculation . in the top left panel , ( blue ) long dashed and ( red ) short dashed curves are for @xmath51 and @xmath94 , respectively , and all models had the same initial fluctuation spectrum ( so that @xmath53 has @xmath95 today ) .
the top right panel shows the ratio of counts in two @xmath53 runs , but with different initial conditions , taylored so that @xmath82 at @xmath83 is the same as in the @xmath51 ( long dashed ) and @xmath94 ( short dashed ) runs .
short dashed curves in bottom panel show the ratio of the @xmath51 counts to that in the @xmath53 run when it has been normalized to have the same ( linear theory ) power spectrum at @xmath83 as the @xmath51 run .
long dashed curve shows a similar analysis of the @xmath30 case .
this panel shows the ratio of the numerators in the previous two panels .
, title="fig : " ] mpc ratioed to @xmath53 . in all panels ,
data points are from simulation , whereas curves are from our excursion set calculation . in the top left panel , ( blue )
long dashed and ( red ) short dashed curves are for @xmath51 and @xmath94 , respectively , and all models had the same initial fluctuation spectrum ( so that @xmath53 has @xmath95 today ) .
the top right panel shows the ratio of counts in two @xmath53 runs , but with different initial conditions , taylored so that @xmath82 at @xmath83 is the same as in the @xmath51 ( long dashed ) and @xmath94 ( short dashed ) runs .
short dashed curves in bottom panel show the ratio of the @xmath51 counts to that in the @xmath53 run when it has been normalized to have the same ( linear theory ) power spectrum at @xmath83 as the @xmath51 run .
long dashed curve shows a similar analysis of the @xmath30 case .
this panel shows the ratio of the numerators in the previous two panels .
, title="fig : " ]
we now compare our spherical collapse predictions for halo abundances with measurements in the simulations of @xcite .
these simulations followed the evolution of @xmath96 particles in a periodic box of size @xmath97mpc , for various choices of @xmath5 and @xmath6 . in all cases ,
the background cosmology was flat @xmath3cdm with @xmath98 , and the particle mass was @xmath99 .
in addition , the simulations were always started from the same initial phases , a feature we will exploit shortly .
we identify halos in the standard way using a friends - of - friends algorithm with link length 0.2 times the interparticle separation . in what follows ,
we show results from the @xmath28 mpc runs .
the @xmath53 simulation , with standard initial conditions has @xmath100 at @xmath83 , the corresponding runs for @xmath51 and @xmath30 have @xmath101 and @xmath102 respectively .
following our discussion of how the counts depend on the shape and normalization of the initial power spectrum , we also study results from @xmath53 simulations in which the initial power spectrum was modified so that , at @xmath83 , it has the same shape as the two @xmath76 cases ( @xmath101 and 0.7 ) .
figure [ fig : compareall ] shows how the mass function depends on @xmath5 .
the panels shows the ratio to @xmath53 , and curves show the predictions from our excursion set with moving or standard barrier calculation with standard or modified initial power spectrum .
the calculation is in reasonably good agreement
note in particular that it captures the sense of the trends with @xmath5 . vs. standard
, the bottom is @xmath51 vs. standard .
the points come from selecting big haloes in each realization of the modified gravity simulation and then finding the corresponding halo in the standard gravity simulation , and then the reverse .
n is the number of particles in the identified halos .
the tilt shows that in stronger gravity massive halos are likely to have more particles .
the `` tails '' in the lower left are unimportant.,title="fig : " ] vs. standard , the bottom is @xmath51 vs. standard .
the points come from selecting big haloes in each realization of the modified gravity simulation and then finding the corresponding halo in the standard gravity simulation , and then the reverse .
n is the number of particles in the identified halos .
the tilt shows that in stronger gravity massive halos are likely to have more particles .
the `` tails '' in the lower left are unimportant.,title="fig : " ] because the simulations all had the same initial phases , we were able to perform a slightly more stringent test .
namely , we directly compared the masses of individual halos in the @xmath51 and @xmath30 runs with those when @xmath53 ( i.e. standard gravity ) .
the filled triangles in figure [ fig : nmod ] show the result of selecting the most massive halos in the @xmath76 runs and plotting the number of particles they contain versus the number of particles they contained in the @xmath53 run .
the open squares show the result of making the selection in the @xmath53 run .
the top and bottom panels show results for @xmath103 and @xmath104 respectively .
this illustrates that with stronger gravity ( larger @xmath5 ) , a given halo is more likely to become more massive , but this only matters for halos more massive than about @xmath105 .
we presented a study of nonlinear effects in a model with a modified gravitational potential ( equation [ potentialeq ] ) .
in particular , we showed how the spherical evolution model is modified , and the effect this has on the abundance of virialized objects .
halos are more massive in models where gravity is stronger on large scales ( figure [ fig : nmod ] ) , although this effect is only important for sufficiently massive objects whose evolution brings them close to the scale @xmath6 on which gravity was modified .
the effect this has on the abundance of massive objects depends on how the models are normalized .
if normalized so that the fluctuation field is the same at early - times ( cmb - normalized ) , then the models with @xmath80 have more massive halos ( solid curves in fig .
[ fig : comparisonlinear ] differ from unity ) .
this remains true , although the dependence on @xmath5 is reduced , if the models are normalized to have the same ( @xmath5-dependent ) linear theory rms fluctuations today ( compare solid and dashed curves in fig . [
fig : comparisonlinear ] ) . if normalized to have the same ( linear theory ) rms fluctuations today , whatever the value of @xmath5 , then the trends can be reversed ( compare dotted curves with unity in fig .
[ fig : comparisonlinear ] ) .
this last normalization convention is sometimes known as ` cluster - normalized ' : our work suggests that , in the context of modified gravity models , this jargon is misleading !
we showed that our analysis captures the essence of the trends seen in the simulations ( figure [ fig : compareall ] ) , so , in principle , the abundance of rich clusters should place interesting constraints on modifications to the gravitational potential . in particular , the modification to cluster abundances can not be reproduced by standard gravity with initial conditions modified to match the change in the linear theory power spectrum ; the differences in abundance can be larger than ten percent for sufficiently massive halos ( figure [ fig : comparisonlinear ] ) .
however , to use cluster counts quantitatively in this way , our analysis should be extended to models in which objects form from an ellipsoidal collapse , as was necessary for standard gravity @xcite .
our analysis also helps to understand an interesting fact about the shape of the nonlinear power spectrum in modified gravity theories .
figure 7 in @xcite shows that whereas the large scale power spectrum in modified theories may be rather different than in standard gravity ( because the linear growth factor is modified ) , the power on small scales ( @xmath106 ) is unchanged .
our analysis shows that , because small mass objects were never larger than the scale @xmath6 on which gravity was modified , they are not affected by the modification , so their abundance is the same as in the standard case .
this is not affected by the initial conditions that we choose so that in both when @xmath5 is non - zero , and when the power spectrum is changed so that we end up with a final power spectrum the same as in the case of modified gravity , the abundances of small halos is unaffected .
in addition , their formation histories will also be unchanged , so their internal structural parameters ( shapes , density profiles ) are also unchanged . in the halo model description of large scale structure @xcite ,
the power at @xmath106 is dominated by small mass halos .
since these are the ones for which gravity is essentially unmodified , the small - scale power spectrum is also unchanged .
figure [ fig : contours ] shows the result of a complementary study . in this case
, we selected a massive halo from one of the simulations ( say @xmath51 ) , and then looked at where its particles were in the other runs ( with different @xmath5 ) .
the figure illustrates clearly that when @xmath51 , then the particle distribution is more compact .
e.g. , in the top panel , the large halo in the @xmath51 run is broken up into three smaller haloes in the @xmath103 run .
the bottom panel shows another effect : that the particles which made up a halo in the @xmath30 run are in a different location in the @xmath51 run , suggesting that the peculiar velocities of the most massive halos are also sensitive to @xmath5 .
the sequence of contours associated with the different @xmath5 runs look rather similar to the time evolution of an object in , say , an @xmath53 model .
thus , the most massive halos in models with @xmath107 may be like high - redshift versions of the most massive halos in models with @xmath80 .
therefore , these figures suggest that the galaxy populations in massive clusters may be rather different in models with large @xmath5 than when @xmath5 is small . in particular , it is likely that the difference between the cluster and field galaxy populations increases as the strength of gravity on large scales increases .
this is largely a consequence of the different @xmath82 values in these runs so cluster @xmath108 ratios , currently used to to constrain @xmath82 ( e.g. * ? ? ?
* ; * ? ? ?
* ) , may one day be used to constrain modifications to gravity .
furthermore , in standard gravity models , there is a strong correlation between evolution and environment @xcite .
there are two reasons to suspect that this will be different if the gravitational potential is modified .
first , in the standard model , the correlation between local environment and evolution is a consequence of birkhoff s theorem .
birkhoff s theorem is lost when the force law is modified ( it is this which modified our spherical evolution model from the standard case ) . and
second , figure [ fig : contours ] shows that the time scale for the assembly of objects is modified .
the environmental dependence of galaxy properties is in rather good agreement with the standard model @xcite , so it may be that this will one day provide interesting constraints on @xmath5 .
this is the subject of work in progress . in standard gravity
, the formation and abundance of voids can be estimated using similar methods to those used for clusters @xcite extending our analysis of the modified potential to voids is also work in progress .
we would like to thank francis bernardeau , antonaldo diaferio , lam hui , kyle parfrey and roman scoccimarro for helpful discussions and encouragement .
99 dunkley , j. , et al .
( 2008 ) , arxiv:0803.0586 milgrom , m. , , 270 , 365 ( 1983 ) bekenstein , j. d. , , 70 , 083509 ( 2004 ) dvali , g. , gabadadze , g. , & porrati , m. , phys .
b 485 , 208 ( 2000 ) lue , a. , scoccimarro , r. , & starkman , g. d. , , 69 , 124015 ( 2004 ) mannheim , p. d. , general relativity and gravitation , 22 , 289 ( 1990 ) diaferio , a. , & ostorero , l. ( 2008 ) , arxiv:0808.3707 kamenshchik , a. , moschella , u. , & pasquier , v. , physics letters b , 511 , 265 ( 2001 ) makler , m. , de oliveira , s. q. , & waga , i. , physics letters b , 555 , 1(2003 ) sen , a. a. , & scherrer , r. j. , , 72 , 063511 ( 2005 ) scherrer , r. j. , & sen , a. a. , , 77 , 083515 ( 2008 ) bertacca , d. , bartolo , n. , diaferio , a. , & matarrese , s. , 2008 , jcap , in press ( arxiv:0807.1020 ) sealfon , c. , verde , l. , & jimenez , r. , , 71 , 083004 ( 2005 ) shirata , a. , shiromizu , t. , yoshida , n. , & suto , y. , , 71 , 064030 ( 2005 ) shirata , a. , suto , y. , hikage , c. , shiromizu , t. , & yoshida , n. , , 76 , 044026 ( 2007 ) stabenau , h. f. , & jain , b. , , 74 , 084007 ( 2006 ) nusser , a. , gubser , s. s. , & peebles , p. j. e. , , 71 , 083505 ( 2005 ) press , w. h. , & schechter , p. , , 187
, 425 ( 1974 ) cooray , a. , & sheth , r. , phys .
rept . , 372 , 1 ( 2002 ) mo , h. j. , & white , s. d. m. , mnras , 282 , 347 ( 1996 ) sheth , r. k. , & tormen , g. , mnras , 329 , 61 ( 2002 ) bond , j. r. , cole , s. , efstathiou , g. , & kaiser , n. , , 379 , 440 ( 1991 ) lacey , c. , & cole , s. , mnras , 262 , 627 ( 1993 ) sheth , r. k. , mnras , 300 , 1057 ( 1998 ) gunn , j. e. , & gott , j. r. i. , , 176 , 1 ( 1972 ) sheth , r. k. , & tormen , g. , mnras , 308 , 119 ( 1999 ) sheth , r. k. , mo , h. j. , & tormen , g. , mnras , 323 , 1 ( 2001 ) martino m. c. , & sheth r. k. , mnras , in press tinker , j. l. , weinberg , d. h. , zheng , z. , & zehavi , i. , apj , 631 , 41 ( 2005 ) skibba , r. a. , sheth , r. k. , & martino , m. c. , mnras , 382 , 1940 ( 2007 ) hoyle , f. , rojas , r. r. , vogeley , m. s. , & brinkmann , j. , apj , 620 , 618 ( 2005 ) abbas , u. , & sheth , r. k. , mnras , 378 , 641 ( 2007 ) blanton , m. r. , & berlind , a. a. , apj , 664 , 791 ( 2007 ) sheth , r. k. , & van de weygaert , r. , mnras , 350 , 517 ( 2004 )
this appendix provides fits to the barriers in figure [ fig : barriershapes ] .
first , let @xmath109 , where @xmath43 is measured in units of @xmath110 and @xmath6 is measured in @xmath111mpc , then for positive @xmath5 , @xmath112 whereas for negative @xmath5 , @xmath113 these fits are accurate to a few percent in the range of @xmath114 for @xmath115 , @xmath6 up to @xmath116 mpc , and @xmath5 between @xmath94 and 1 . for @xmath28 mpc , the high end of the range of @xmath115 corresponds to @xmath117 .
the dependence here makes sense , because we expect the mass scale of the modification to scale as @xmath65 . as for the @xmath5 dependence
, we can see from figure [ fig : barriershapes ] that for negative @xmath5 the deviation from standard gravity is stronger , hence the dependence on @xmath5 for positive @xmath5 , and @xmath118 for negative @xmath5 . in principle
, one can use these expressions to generate analytic approximations to the halo mass function as follows . for a given initial power spectrum ,
@xmath115 can be written as a function of @xmath78 ; this specifies the barrier shape in the units which are useful for the excursion set approach . then insert this barrier shape into the expressions for the first crossing distribution given by @xcite .
the ratio of this first crossing distribution to that associated with a barrier of height @xmath38 quantifies the change in halo abundances which is due to @xmath5 and @xmath6 . multiplying this ratio by the actual halo abundances in the standard model @xcite yields an analytic expression for the abundances in the modified model . | modifications to the gravitational potential affect the nonlinear gravitational evolution of large scale structures in the universe . to illustrate some generic features of such changes ,
we study the evolution of spherically symmetric perturbations when the modification is of yukawa type ; this is non - trivial , because we should not and do not assume that birkhoff s theorem applies .
we then show how to estimate the abundance of virialized objects in such models .
comparison with numerical simulations shows reasonable agreement : when normalized to have the same fluctuations at early times , weaker large scale gravity produces fewer massive halos .
however , the opposite can be true for models that are normalized to have the same linear theory power spectrum today , so the abundance of rich clusters potentially places interesting constraints on such models .
our analysis also indicates that the formation histories and abundances of sufficiently low mass objects are unchanged from standard gravity .
this explains why simulations have found that the nonlinear power - spectrum at large @xmath0 is unaffected by such modifications to the gravitational potential .
in addition , the most massive objects in cmb - normalized models with weaker gravity are expected to be similar to the high - redshift progenitors of the most massive objects in models with stronger gravity .
thus , the difference between the cluster and field galaxy populations is expected to be larger in models with stronger large - scale gravity .
[ firstpage ] |
theories with extra dimensions typically give rise to large landscapes of vacua stabilized by flux .
this means that most minima have a great many potential decay paths : they may decay by a _ small step _
( dropping just one or a few units of flux ) ; they may decay by a _ giant leap _ ( dropping many units of flux ) ; they may decay by the giantest leap of all , a _ bubble of nothing _
( dropping every unit of flux , so that the extra dimensions collapse to zero size and spacetime pinches off ) ; they may _ decompactify _ ( so that the extra dimensions run away to infinite size ) ; or they may simultaneously decompactify _ and _ drop flux . in this paper
we will construct the tunneling solutions that describe these processes and compare their relative rates , in the context of an explicit flux compactification . in a previous paper @xcite
, we described effects that generically enhance the rate of giant leaps and showed that _ in the thin - wall approximation _ indeed giant leaps can be the fastest decay route for ` monoflux ' landscapes one - dimensional landscapes built of many units of a single flux .
then , in @xcite , we described additional enhancements specific to ` multiflux ' landscapes multi - dimensional landscapes built of a single unit each of many different fluxes . in section [ sec:6dem ] , we review these arguments , and introduce our explicit flux compactification : 6d einstein - maxwell theory compactified on a two - sphere .
transitions occur by the nucleation of a charged black two - brane ; specifically we assume there are no fundamental charged branes of lower tension . in section
[ sec : fluxinstantons ] , we construct the pertinent instantons , including all thick - wall effects , though still treating the black brane as thin . in section [ sec : bofn ] , we calculate the decay rate as a function of the distance jumped . in monoflux landscapes
, we find that small steps beat giant leaps because thick - wall effects overwhelm the enhancements described in @xcite .
however , in multiflux landscapes , giant leaps beat small steps from almost all starting points , to such an extent that the fastest decay is to dump all the flux at once , jump over the entire landscape , and nucleate a bubble of nothing . in section [ sec : bubbleofnothing ]
, we examine the bubble of nothing in detail and show how it is approached as the limit of flux tunneling in which all units of flux are discharged . in this limit , not only
does the size of the extra dimensions go to zero inside the bubble , but the interior becomes so negatively curved that a 3d slice through the bubble has surface area but no volume .
( witten s original bubble of nothing was constructed for an _
unstabilized _ extra dimension @xcite ; ours is for _ stabilized _ extra dimensions . ) in the appendix , we describe a new decay that combines flux tunneling and decompactification , and argue that it is always subdominant .
( we study , in this paper , transitions out of the four - dimensional vacua ; others have considered transitions out of the six - dimensional vacuum @xcite .
our enhancements for stacks of branes lie not in the prefactor @xcite but in the exponent . despite the enhancements ,
the fastest decay is still generically exponentially suppressed , so we do nt expect percolation @xcite . )
in this section , we review a simple flux compactification , 6d einstein - maxwell theory @xcite , and flux tunneling between the corresponding vacua @xcite .
the six - dimensional action is @xmath0 where @xmath1 and @xmath2 run from @xmath3 to @xmath4 , @xmath5 is the ricci scalar associated with the metric @xmath6 , @xmath7 is the maxwell field strength , @xmath8 is a positive six - dimensional cosmological constant , and we use units for which @xmath9 , @xmath10 , and the reduced 6d planck mass are all 1 .
two of the dimensions are compactified on a sphere , which is buttressed against collapse by flux ; specifically @xmath11 different two - form fluxes , each with field strength @xmath12 where @xmath13 runs from @xmath14 to @xmath11 , @xmath15 is the charge of the @xmath13th magnetic quantum , @xmath16 is the number of units of the @xmath13th magnetic flux , and @xmath17 and @xmath18 span the extra - dimensional two - sphere .
we can choose coordinates @xmath19 such that after integrating out the extra dimensions we end up in einstein frame : @xmath20 where the reduced 4d planck mass is @xmath21 , and @xmath22 is the radion .
we can shift @xmath22 to set @xmath23 , then @xmath24 and @xmath25 the first term in the potential is from flux ; the second is from the curvature of the two - sphere ; and the third , is from the bulk cosmological constant .
the minimum of this potential is at @xmath26 = \frac{2}{3 } \frac{4 \pi m^2 } { { f}^2 }
\left(1 + \sqrt{1 - \frac{3}{4 } \frac { f^2}{4 \pi m^2 } } \right)$ ] , so @xmath27^{\frac{3}{2 } } \right ) \right ] .
\label{eq : vmin}\ ] ] different values of @xmath28 give rise to anti - de sitter ( @xmath29 ) , minkowski ( @xmath30 ) or de sitter ( @xmath31 ) four - dimensional vacua , as shown in fig . 1 ; as @xmath28 increases , the minimum moves to larger values of @xmath22 , and eventually disappears ( @xmath32 ) .
[ fig : vofpsi ] is plotted for different values of @xmath28 .
large @xmath28 gives rise to a de sitter minimum , with @xmath33 and @xmath34 positive .
these minima can decay either by decompactifying ( tunneling over the barrier and rolling @xmath35 ) or by flux tunneling to lower @xmath28 .
as @xmath28 decreases , the minimum first becomes minkowski , with @xmath36 , and then ads , with @xmath33 and @xmath34 negative ; these minima can decay only by flux tunneling . as @xmath37 , the potential destabilizes and @xmath22 can roll towards minus infinity .
this corresponds to a bubble of nothing , as we will see in sec .
[ sec : bubbleofnothing].,title="fig:",width=432 ] e transitions between the minima of 6d einstein - maxwell theory are effected by the nucleation of a charged two - brane that forms a sphere in the large dimensions and sits at a point in the extra dimensions .
the brane unravels flux , so that @xmath22 relaxes to a lower minimum of the effective potential on the inside of the sphere .
if the brane is charged under @xmath38 units of each flux , then the internal field is reduced @xmath39 .
the fastest decay uses the lightest brane with a given charge .
since the theory contains no fundamental branes , these are extremally charged black two - branes .
the tension of such a brane is @xmath40 in @xcite we saw two factors that enhance the rate of giant leaps , both caused by the radion .
the first relates to the brane tension , which , in the thin - brane limit , contributes to the action a term @xmath41 where @xmath42 is the induced metric .
the brane couples to the radion in such a way that , as the extra dimensions swell , the effective tension of the brane redshifts : this means @xmath22 spikes in the vicinity of branes @xcite , and the larger the tension , the bigger the spike .
this brane - radion coupling induces an attractive force between branes stacks of branes bind together , lowering their total tension and making them easier to nucleate .
the second relates to the change in the effective potential @xmath43 .
when flux is discharged , @xmath44 at fixed @xmath22 changes linearly with @xmath45 , as in eq .
[ effpot ] , but the minimum of the potential also relaxes to a more negative @xmath22 , so that @xmath43 scales faster than linearly with @xmath45 .
bigger @xmath43 means the true vacuum is more preferred and the false vacuum decays faster . to illustrate the range of behaviors , we focus on two extreme examples .
* monoflux landscapes * we define as having only a single type of flux ( @xmath46 ) and many units of it ( @xmath47 ) , so that @xmath48 .
nucleating a stack of @xmath49 branes sends @xmath50 and the branes are charged under the same flux .
* multiflux landscapes * we define as having many types of flux ( @xmath51 ) , but only a single unit of each ( @xmath52 ) , each with the same charge ( @xmath53 ) , so that @xmath54 .
nucleating a stack of @xmath49 branes sends @xmath55 and the branes are now charged under different fluxes .
in @xcite we saw two additional factors that enhance the rate of giant leaps , this time specific to multiflux landscapes .
the first relates to eq .
[ eq : tofgn ] and the way the tension of a stack of @xmath49 branes grows with @xmath49 .
monoflux branes all carry the same type of charge , so @xmath56 .
this is because , ignoring the radion , these extremal monoflux branes do not interact their gravitational attraction is precisely cancelled by their magnetic repulsion , and the tensions add .
multiflux branes are charged under different fluxes , so @xmath57 .
this is because the branes bind together there is now no magnetic repulsion to cancel their gravitational attraction .
the second relates to eq .
[ eq : fsquared ] and the way the flux dropped grows with @xmath49 .
for monoflux landscapes , the flux lines are the same and thus repel , so that @xmath58 , which grows slower than linearly with @xmath49 .
for multiflux landscapes , the flux lines are different and thus do not interact , so that , which grows linearly with @xmath49 . for given true and false vacuum energies ,
therefore , the tension of the multiflux brane is larger than that of the monoflux brane : @xmath59 \\ \label{tmulti } t_\text{multi } & = & \frac 2{\sqrt3 } \left(f_\text{false}^2-f_\text{true}^2\right)^{1/2}.\end{aligned}\ ] ] ( the tensions agree for @xmath60 , the bubble of nothing . ) on account of their larger tension , all multiflux decays are suppressed relative to their monoflux counterparts , but the least relatively suppressed are the ones with big @xmath49 , the giant leaps .
the instantons that describe bubble nucleation can be found by passing to euclidean signature and imposing o(4 ) symmetry .
the 4d metric can then be written as @xmath61 and the action is @xmath62 \nonumber \\[12pt ] & + & 2\pi^2 a^3 t e^{-3\psi/2 m } { \big{|}}_{\bar{\rho } } + 2 \pi^2 \int_{\bar{\rho}}^{\ , } d\rho \left [ a^3 \left ( \frac12 \psi'^{\,2}+v_n(\psi)\right)- 3m^2a\bigl(1+(a')^2\bigl)\right],\end{aligned}\ ] ] where we have integrated by parts , so that the action only contains single derivatives ( and there is no gibbons - hawking term ) .
the action has three pieces , corresponding , respectively , to the bubble s interior , the brane , and the bubble s exterior .
the equations of motion are the euler - lagrange equation for @xmath63 , @xmath64 and the gravitational constraint equation for @xmath65 , @xmath66 where @xmath67 is chosen appropriately for inside the brane ( @xmath68 ) and outside the brane ( @xmath69 ) . at the brane , there are two boundary conditions , one from varying @xmath70 @xmath71 and the other ( the israel junction condition ) from varying @xmath72 , @xmath73 this second condition can be better understood by applying eq . and taking the non - gravitational limit ( @xmath74 ) : @xmath75 this is a force - balance equation , balancing the pressure differential , on the left , against the surface tension , on the right . for larger @xmath76 ,
the pressure term wins and the bubble can lower its action by expanding ; for smaller @xmath77 , the surface tension term wins and the bubble can lower its action by contracting this mode is therefore associated with the negative eigenvalue . at the origin @xmath78 and the instanton must be smooth , @xmath79 which implies that @xmath80 .
( except for the bubble of nothing , which satisfies its own smoothness condition . ) for decay from minkowski or ads space , the instanton is infinite , @xmath65 grows without bound , and @xmath63 must asymptote to its false vacuum value far from the bubble . for decay from de sitter space ,
the instanton is compact , @xmath65 has a second zero at the antipode to @xmath81 , where , again by smoothness , @xmath82 , though @xmath22 need not sit in its false vacuum value . ) , minkowski ( @xmath83 ) and ads ( @xmath84 ) .
the radion @xmath63 is near its true vacuum value at @xmath85 and from there interpolates smoothly towards its false vacuum value ( indicated by the dotted line ) , except for a derivative discontinuity at the brane ( indicated by the red dot ) . in the de sitter case , the instanton is compact , so @xmath63 never quite reaches its false vacuum value .
deep inside the bubble , @xmath65 is concave up , as is appropriate to ads ; at large @xmath86 , @xmath65 behaves like @xmath87 , @xmath86 , and @xmath88 respectively.,width=624 ] .
the corresponding brane tension is larger than in the monoflux case , which has two consequences .
first , the bubble must be bigger to compensate for the increased surface tension , so @xmath77 and @xmath72 are bigger .
second , the radion spikes more in the vicinity of the brane , and to such an extent that @xmath63 is nonmonotonic.,width=624 ] in fig .
[ sampleinstantonmono ] , instanton profiles are plotted for three sample monoflux transitions .
all transitions are to the same ads true vacuum , but from different false vacua : respectively , a de sitter , a minkowski and an ads . in all cases , @xmath22 is monotonic in @xmath86 ; though the effect of the brane tension is to locally spike the radion , this effect is overwhelmed by @xmath22 interpolating between its different vacuum values . in fig .
[ sampleinstantonmulti ] , instanton profiles are now plotted for three sample multiflux transitions .
the starting and finishing @xmath89 s are the same as for the monoflux transitions ; the difference is that the brane is heavier in the multiflux case .
this has two visible consequences : multiflux bubbles must be bigger to compensate for the bigger surface tension ; and the heavier brane induces a spike in the radion that is now so large that @xmath22 is nonmonotonic .
( for transitions to deep enough ads , much deeper than shown here , the change in @xmath22 becomes large enough that it does eventually overwhelm the spike caused by the multiflux brane and the profile becomes monotonic . )
the tunneling rate @xmath90 between two vacua is given by @xmath91 in fig .
[ bofn ] , the tunneling exponent @xmath2 is plotted for decays from four false vacua : very high de sitter , intermediate de sitter , minkowski and ads . it is shown as a function of the fraction of the landscape jumped : @xmath92 for a monoflux landscape and @xmath93 for a multiflux landscape . though we have drawn it as a solid line , true vacua only lie at discrete points along this curve , with a spacing set by @xmath94 .
( other than this , the values of @xmath94 and @xmath8 , have no effect . ) for monoflux landscapes , the plots look similar for each starting point .
@xmath2 increases monotonically and then flattens out near @xmath95 .
though in the thin - wall approximation giant leaps appear the most likely decay @xcite , this full calculation shows that actually small steps dominate . for multiflux landscapes
, it depends where you start . from minkowski and ads
vacua , @xmath2 is monotonically _ decreasing _ so that giant leaps are preferred and the best is to drop all units of flux ( which leads to a bubble of nothing , as we will discuss in more detail in the next section ) .
small enough steps are nt just suppressed , they re forbidden : a whole section of the landscape is inaccessible .
an interpretation of this behavior was given by coleman and de luccia @xcite .
quantum decay conserves energy , so the bubble s surface area to volume ratio must be small enough that the energy benefit of the interior can compensate for the energy cost of the surface tension .
but because the interior is negatively curved , its volume scales asymptotically with its area , and the surface area to volume ratio can not be made arbitrarily low . for small steps , the bubble can never be big enough to have zero energy and the instantons are gravitationally blocked . from de sitter vacua , no decay is gravitationally blocked .
@xmath2 rises fast at small @xmath49 and then turns over and falls slowly down towards @xmath96 .
much the same is true from very high de sitter , but with a complication . from very high de sitter decompactification
so dominates flux tunneling for intermediate @xmath49 that the flux tunneling instanton disappears in the range between the stars .
this is not gravitational blocking the instanton has instead been swallowed by a much faster decay .
similar disappearing instantons have been discussed elsewhere in the literature @xcite ; we give a general account of instanton disappearances and a complete account of this particular disappearance in @xcite . from even higher de sitter ,
the stars move apart , eventually swallowing even the bubble of nothing .
if you start very near the top of the multiflux landscape , the most likely decay is decompactification . from slightly farther down the landscape
, it depends on @xmath94 , which controls the proximity of the next vacuum . for a given de sitter space
, there is a critical value of @xmath94 for which the rate to nucleate a single brane is the same as the rate to nucleate a bubble of nothing . at smaller @xmath94
, the first vacuum appears earlier and has smaller @xmath2 , so that the optimal decay path is a small step . at larger @xmath94 ,
giant leaps dominate .
this critical value of @xmath94 decreases as you move down the landscape , so that for fixed @xmath94 , if you start near the top you expect to first take small steps and then , still well within the de sitter regime , make a giant leap to a bubble of nothing . as a function of the number @xmath49 of flux units dropped .
we plot this both for monoflux and multiflux landscapes , and from four different starting points : very high de sitter ( @xmath97 ) , de sitter ( @xmath98 ) , minkowski ( @xmath99 ) and ads ( @xmath100 ) .
discrete true vacua lie along the curves with spacing set by @xmath94 . for monoflux decays , the curves rise monotonically with @xmath49
. for multiflux decays from ads and minkowski , @xmath101 falls monotonically in fact there is a critical @xmath49 , indicated by the vertical dashed line , at which @xmath101 diverges ; decays with smaller @xmath49 are gravitationally forbidden
. for multiflux decays from de sitter , @xmath101 rises sharply and then falls off again . for multiflux decays from very high de sitter ,
decompactification is competitive and its rate is indicated by the dotted horizontal line ; it so dominates flux decay for intermediate @xmath49 that no instanton exists in the range between the two stars .
monoflux transitions are faster than the corresponding multiflux transitions , with the exception of the maximum value of @xmath49 ; they agree on the rate of the bubble of nothing .
the minkowski bubble of nothing has tunneling exponent @xmath102 .
bubbles of nothing from de sitter have larger @xmath2 and those from ads have smaller @xmath2.,width=624 ] two approximations go into these plots .
the brane breaks the spherical symmetry of the extra dimensions .
our first approximation is to ignore the effects of this breaking we treat the shape moduli as fixed and ignore higher kaluza - klein modes of the flux .
though we have eliminated the thin - wall approximation of @xcite , our second approximation is to continue to treat the brane as thin .
this allows us to use the membrane action eq .
[ eq : brane ] .
this approximation is reliable at small and intermediate @xmath49 , establishing that multiflux giant leaps beat small steps ; but it gets steadily worse as we tunnel farther and by the bubble of nothing solution the thickness of the brane ( @xmath103 ) is approaching the size of the extra dimensions and the proper size of the bubble .
we anticipate that moving away from this approximation will increase @xmath2 for the largest values of @xmath49 .
we could move away from this approximation by committing to a particular model that resolves the core of the brane , as in @xcite ; such a model would make the brane tension , and also its thickness , into tunable parameters .
monoflux transitions are faster than the equivalent multiflux transitions , because for a given @xmath45 , the brane tension is smaller . but
as @xmath45 increases , monoflux transitions get slower and multiflux transitions generally get faster . in the limit that all the flux is dropped ,
the rates are the same : the transitions have the same @xmath45 , the same brane tension , and the same instanton profile .
this limit corresponds to the bubble of nothing .
the fastest decay in a multiflux landscape is often achieved in the limit @xmath104 , so that all the flux is discharged , the radion is no longer stabilized against collapse , both @xmath34 and @xmath105 , the extra dimensions shrink to zero size , spacetime pinches off , and we create a bubble of nothing . for comparison ,
we review the original bubble of nothing , discovered by witten @xcite , in which an unstabilized extra dimension shrinks to zero .
the instanton has metric @xmath106 with @xmath107 and @xmath108 .
the left pane of fig .
[ fig : cross ] shows a cross - section through this bubble at the instant of nucleation .
the size of the extra dimension , @xmath109 , is plotted against the area - radius , @xmath110 .
the extra dimension shrinks smoothly to zero at @xmath111 , and spacetime pinches off : for @xmath112 there is literally nothing .
after nucleation , the bubble accelerates outwards approaching the speed of light , and the hole in spacetime grows .
this bubble of nothing describes the collapse of a _ single _ , _ unstabilized _ extra dimension in minkowski .
bubbles of nothing in 6d einstein - maxwell theory describe the collapse of _ two _ , _ stabilized _ extra dimensions from any of the minkowski , de sitter or ads vacua .
the right pane of fig .
[ fig : cross ] shows a cross - section through the bubble of nothing from the minkowski vacuum .
the metric is @xmath113 so the proper size of the extra dimensions is @xmath114 and the proper area - radius is @xmath115 .
the brane sits at finite @xmath116 ; inside the brane , there is no stabilizing flux and the size of the extra dimensions shrinks to zero .
it does so much more precipitously than in the 5d case , as there is now a curvature term driving the collapse . defining the pinch - off point to be @xmath81
, we can expand the solutions to eqs .
[ eompsi ] and [ eomgrav ] at small @xmath117 to give @xmath118 any value of @xmath119 solves the equations of motion , but only one value ensures that @xmath22 settles in at the desired value at infinity , rather than under - shooting or over - shooting . we can find this value numerically and for decays from minkowski it is @xmath120 .. core - resolution is a desirable ( albeit model - dependent ) way to move beyond our thin - brane approximation .
but the substantive difference is that our thin brane sits at finite @xmath22 , which is inconsistent with the ansatz , suggesting the solution in @xcite has too many negative modes we anticipate an instability whereby the points near @xmath81 roll off the crest of the higgs potential and the brane shifts to larger @xmath22 . combining a core - resolution technique with a more general ansatz would be definitive , though computationally intensive . ]
the proper area of the @xmath85 equatorial two - sphere is @xmath121 , which is finite even though @xmath122 and @xmath123 .
the solution is smooth despite not having @xmath124 , as can be seen by defining @xmath125 and expanding near @xmath126 to give @xmath127 the bubble of nothing in 6d einstein - maxwell theory is not an isolated solution it can be smoothly approached as the limit of flux tunneling in which all the flux is removed .
studying this limit gives new insight into the bubble s structure .
, @xmath128 , @xmath129 , and , in bold , the bubble of nothing , @xmath130 .
the plot is made for a monoflux landscape ; a multiflux landscape approaches the same limit from a slightly different direction .
on the outside of the bubble , @xmath131 so @xmath132 . on the inside of the bubble ,
the extra dimensions relax to their true vacuum size.,title="fig:",width=336 ] , @xmath128 , @xmath129 , and , in bold , the bubble of nothing , @xmath130 .
the plot is made for a monoflux landscape ; a multiflux landscape approaches the same limit from a slightly different direction . on the outside of the bubble ,
@xmath131 so @xmath132 .
on the inside of the bubble , the extra dimensions relax to their true vacuum size.,title="fig:",width=364 ] figure [ fig : bubofnot3 ] shows a sequence of solutions that tunnel ever farther down the landscape and approach the bubble of nothing .
the instantons can be roughly divided into three segments . in the inner segment , when @xmath86 is small , @xmath22 sits close to its true vacuum value and @xmath133 , where @xmath134 is the appropriate ads curvature length . in the middle segment
, @xmath22 interpolates from close to the true vacuum to close to the false vacuum and @xmath135 , as in eq .
[ eq : arho ] . in the outer segment , on the other side of the brane
, @xmath22 slowly relaxes to its false vacuum value as @xmath136 and , because the plots are for a minkowski false vacuum , @xmath137 .
as the instantons approach the bubble of nothing , the inner segment gets deeper but shorter the solution spends less long loitering in the vicinity of the true vacuum ( as measured by @xmath86 ) . in the limit that we tunnel all the way down
, @xmath34 disappears off to minus infinity , the inner segment vanishes entirely , and the middle segment stretches all the way to @xmath81 . because @xmath138 , the interior of the bubble has no volume in the _ extra _ dimensions .
but it s more than that : it also has no volume in the _ extended _ directions ( it s not like a flattened pancake with 3-volume but no thickness )
. the interior of the bubble , deep in ads , is negatively curved , so that @xmath139 ; because this is exponentially growing , most of the volume sits within @xmath134 of the surface . in the limit ,
@xmath140 so @xmath141 and there is surface area ( it s a `` bubble '' ) but no volume ( of nothing " ) .
as discussed here and in @xcite , there are two radion - related effects that help giant leaps in both monoflux and multiflux landscapes
. 1 . * tension effect . * in einstein frame
, the 4d effective brane tension shrinks as the radion swells , but at a cost : to swell , the radion must move away from its minimum .
a second brane at the same point accrues the same benefit , without having to pay again the cost swelling the extra dimensions .
thus the radion mediates an attractive force between branes , reducing the tension of large stacks .
2 . @xmath142
* for a given @xmath22 , @xmath67 changes linearly with @xmath28 , but as @xmath28 changes so too does the @xmath22-location of the minimum .
@xmath33 is thus a concave - down function of @xmath28 , so that @xmath44 grows faster than @xmath45 ( indeed @xmath105 as @xmath37 ) .
there are three effects that make giant leaps more favored in multiflux landscapes than in monoflux landscapes
. 1 . * tension effect . * for monoflux branes
, the magnetic repulsion cancels the gravitational attraction , and the tension of a stack grows like @xmath49 ( ignoring the radion ) . for multiflux branes , charged under different fluxes , there is no magnetic repulsion , and the tension grows like @xmath143 .
2 . @xmath144
* effect . * for monoflux decays , at each step you shed the same @xmath145 , but ever less @xmath28 : @xmath146@xmath147 .
but for multiflux decays , the flux lines do nt interact , and at each step you shed the same @xmath28 : @xmath148 .
* thick - wall effect . *
giant leaps have thick walls , and thick walls slow decay
they encroach upon the interior true vacuum and increase @xmath2 . for monoflux transitions , which proceed by small bubbles , this effect is large and in the end sinks giant leaps . however , as we saw in fig .
[ sampleinstantonmulti ] , multiflux transitions proceed by bigger bubbles , which can accommodate the thick walls with a proportionately smaller increase in @xmath2 . incorporating all these effects ,
we have calculated the tunneling rate as a function of @xmath49 , in the thin - brane approximation .
this allows us to track the most likely decay chain . from the highest de sitter states , the fastest decay is to decompactify . from somewhat lower down
, it depends on the type of landscape . for monoflux landscapes , built of many units of a single flux , the most likely decay chain is to discharge one unit at a time , proceeding by small steps all the way to the bottom . for multiflux landscapes , built of a single unit each of many fluxes ,
high de sitter states take a few small steps before a catastrophic giant leap all the way to a bubble of nothing . what about intermediate cases , with many units of many fluxes
? different types of flux still like to discharge together , because they enjoy all the benefits described above .
that said , if the @xmath149 all share a common factor @xmath150 , then the bubble of nothing ( @xmath151 ) is necessarily subdominant .
because monoflux landscapes take small steps , we know that tunneling by @xmath152 must be faster . if the @xmath149 do nt share a common factor and the @xmath15 vary , then the cases proliferate , but for much of the landscape , giant leaps will still be the fastest decay . if , as seems plausible , we ourselves live in a multiflux landscape , then we draw two conclusions .
we arrived here by an exponentially subdominant decay , and we will leave here by a bubble of nothing .
thanks to paul steinhardt and herman verlinde for useful feedback and to jose juan blanco - pillado , handhika ramadhan and benjamin shlaer for helpful discussion on the bubble of nothing .
hat tip to kendrick and mari , much respect .
in sec . 4 , we considered both flux tunneling instantons and decompactification instantons .
but there is one more possibility . in this appendix
, we construct the instanton that simultaneously discharges flux _ and _ decompactifies . in fig .
[ fig : decompactification]a we plot the standard decompactification instanton .
this instanton does nt change the flux , and instead nucleates a bubble inside of which @xmath22 is over the hump in the effective potential of fig . 1 and will classically roll @xmath35 . in fig .
[ fig : decompactification]b we plot the new instanton . as before
, @xmath22 is over the hump , so this instanton corresponds to decompactification .
but there is now a charged brane , so that there is less flux on the inside , the effective potential is reduced , and hump is shrunk . which method of decompactifying is fastest ? on the one hand
the new method lowers the hump in the effective potential , making decompactification easier . on the other hand
, it requires the nucleation of a brane , making decompactification harder .
figure [ fig : bdecompactification ] shows that this second effect is strongest , so that the new instantons are always subdominant to standard decompactification .
this figure shows that decompactification gets slower the more flux is simultaneously dropped . .
at @xmath153 , no flux is dropped , so this is just standard decompactification . since the rate grows with @xmath49 , simultaneously dropping flux and decompactifying is subdominant to standard decompactification .
similar plots , with identical conclusions , can be drawn from other de sitter false vacua and for multiflux landscapes.,width=384 ] indeed , standard decompactification so dominates this new instanton , that eventually the new instanton disappears at a critical @xmath154 ( just as the regular flux tunneling instanton disappears from a high enough de sitter state , as in the top left pane of fig . 4 ) .
see @xcite for an account of the disappearance
. as the false vacuum is raised , the star moves left and eventually swallows everything , including even the standard decompactification instanton .
( at the moment the standard decompactification instanton is swallowed , it is replaced by a hawking - moss instanton@xcite . )
a. r. brown and a. dahlen , `` small steps and giant leaps in the landscape , '' phys .
d * 82 * , 083519 ( 2010 ) [ arxiv:1004.3994 [ hep - th ] ] .
a. r. brown and a. dahlen , `` giant leaps and minimal branes in multi - dimensional flux landscapes , '' phys .
d * 84 * , 023513 ( 2011 ) [ arxiv:1010.5241 [ hep - th ] ] .
e. witten , `` instability of the kaluza - klein vacuum , '' nucl .
b * 195 * , 481 ( 1982 ) .
a. d. linde and m. i. zelnikov , `` inflationary universe with fluctuating dimension , '' phys .
b * 215 * , 59 ( 1988 ) ; s. m. carroll , m. c. johnson and l. randall , `` dynamical compactification from de sitter space , '' jhep * 0911 * , 094 ( 2009 ) [ arxiv:0904.3115 [ hep - th ] ] ; j. j. blanco - pillado , d. schwartz - perlov and a. vilenkin , `` transdimensional tunneling in the multiverse , '' jcap * 1005 * , 005 ( 2010 ) [ arxiv:0912.4082 [ hep - th ] ] . j. l. feng , j. march - russell , s. sethi and f. wilczek , saltatory relaxation of the cosmological constant, nucl . phys .
b * 602 * , 307 ( 2001 ) [ arxiv : hep - th/0005276 ] .
a. h. guth and e. j. weinberg , `` could the universe have recovered from a slow first order phase transition ?
, '' nucl .
b * 212 * , 321 ( 1983 ) .
s. sarangi , g. shiu and b. shlaer , `` rapid tunneling and percolation in the landscape , '' int .
j. mod .
a * 24 * , 741 ( 2009 ) [ arxiv:0708.4375 [ hep - th ] ] .
g. o. freund and m. a. rubin , `` dynamics of dimensional reduction , '' phys .
b * 97 * , 233 ( 1980 ) .
j. j. blanco - pillado , d. schwartz - perlov and a. vilenkin , `` quantum tunneling in flux compactifications , '' jcap * 0912 * , 006 ( 2009 ) [ arxiv:0904.3106 [ hep - th ] ] .
i. s. yang , `` stretched extra dimensions and bubbles of nothing in a toy model landscape , '' phys .
d * 81 * , 125020 ( 2010 ) [ arxiv:0910.1397 [ hep - th ] ] .
s. r. coleman and f. de luccia , `` gravitational effects on and of vacuum decay , '' phys .
d * 21 * , 3305 ( 1980 ) .
m. cvetic and h. h. soleng , `` naked singularities in dilatonic domain wall space times , '' phys .
d * 51 * , 5768 ( 1995 ) [ arxiv : hep - th/9411170 ] ; m. c. johnson and m. larfors , `` an obstacle to populating the string theory landscape , '' phys .
d * 78 * , 123513 ( 2008 ) [ arxiv:0809.2604 [ hep - th ] ] ; a. aguirre , m. c. johnson and m. larfors , `` runaway dilatonic domain walls , '' phys .
d * 81 * , 043527 ( 2010 ) [ arxiv:0911.4342 [ hep - th ] ] .
a. r. brown and a. dahlen , `` the case of the disappearing instanton , '' arxiv:1106.0527 [ hep - th ]
. j. j. blanco - pillado and b. shlaer , `` bubbles of nothing in flux compactifications , '' phys .
d * 82 * , 086015 ( 2010 ) [ arxiv:1002.4408 [ hep - th ] ] .
j. j. blanco - pillado , h. s. ramadhan and b. shlaer , `` decay of flux vacua to nothing , '' jcap * 1010 * , 029 ( 2010 ) [ arxiv:1009.0753 [ hep - th ] ] .
s. w. hawking and i. g. moss , `` supercooled phase transitions in the very early universe , '' phys
. lett .
b * 110 * , 35 ( 1982 ) .
l. g. jensen and p. j. steinhardt , `` bubble nucleation for flat potential barriers , '' nucl
. phys .
b * 317 * , 693 ( 1989 ) . | the rate and manner of vacuum decay are calculated in an explicit flux compactification , including all thick - wall and gravitational effects . for landscapes built of many units of a single flux ,
the fastest decay is usually to discharge just one unit .
by contrast , for landscapes built of a single unit each of many different fluxes , the fastest decay is usually to discharge all the flux at once , which destabilizes the radion and begets a bubble of nothing . by constructing the bubble of nothing as the limit in which ever more flux is removed ,
we gain new insight into the bubble s appearance . finally , we describe a new instanton that mediates simultaneous flux tunneling and decompactification .
our model is the thin - brane approximation to six - dimensional einstein - maxwell theory . |
we wish to thank the engineering and physical sciences research council ( uk ) for supporting this work .
jtn acknowledges support from the isaac newton trust , and dar acknowledges support from toshiba cambridge research centre .
we thank a.v .
khaetskii for a critical reading of the manuscript , and we acknowledge help with the numerical calculations from i. m. castleton and c. h. w. barnes . although portions of the two 2degs close to the ohmic contacts have carrier densities different to @xmath46 and @xmath120 , the resistance of these leads is always much smaller than the tunneling resistance , and the results presented here depend only on the matched carrier densities in the tunneling region . | the tunneling between two parallel two - dimensional electron gases has been investigated as a function of temperature @xmath0 , carrier density @xmath1 , and the applied perpendicular magnetic field @xmath2 . in zero magnetic field
the equilibrium resonant lineshape is lorentzian , reflecting the lorentzian form of the spectral functions within each layer . from the width of the tunneling resonance
the lifetime of the electrons within a 2deg has been measured as a function of @xmath1 and @xmath0 , giving information about the density dependence of the electron - impurity scattering and the temperature dependence of the electron - electron scattering . in a magnetic field
there is a general suppression of equilibrium tunneling for fields above @xmath3 t. a gap in the tunneling density of states has been measured over a wide range of magnetic fields and filling factors , and various theoretical predictions have been examined . in a strong magnetic field , when there is only one partially filled landau level in each layer , the temperature dependence of the conductance characteristics has been modeled with a double - gaussian spectral density .
introduction [ s : intro ] in addition to conserving energy @xmath4 , resonant tunneling between two parallel two - dimensional electron gases ( 2degs ) requires@xcite the conservation of the in - plane momentum @xmath5 . in zero magnetic field
, the conservation of both quantities allows the investigation of the broadening of the electronic states within an individual 2deg , in a way that is not possible with conventional transport measurements .
for example , recent experimental studies@xcite have shown that the lifetime of the two - dimensional electrons can be measured directly by tunneling spectroscopy .
the lifetime broadening is characterized by the electron spectral function @xmath6 , which describes the probability that an electron in a particular @xmath7-state has energy @xmath4 . in general , @xmath6 is strongly peaked near the free - particle energy @xmath8 , with a width @xmath9 determined by the lifetime @xmath10 of the electron . in a strong perpendicular magnetic field ,
when the landau level filling factor @xmath11 is less than unity , it is assumed that the electron traverses the tunnel barrier between the two 2degs in a much shorter timescale than the charge rearrangement around the injected electron ( and around the hole left behind in the emitter layer ) .
the energies of the injection and extraction processes create a gap in the tunneling density of states ( dos ) .
many - body theories predict@xcite that for a fixed filling factor , @xmath12 , the width of the energy gap is approximately @xmath13 ( where @xmath14 is the magnetic length ) , or if the electron liquid forms a regular lattice@xcite the gap energy is @xmath15 ( where @xmath16 is the inter - electron spacing within a layer ) . at fixed filling factor both predicted forms for the gap have the same @xmath17 dependence with magnetic field . in an earlier paper we reported@xcite tunneling characteristics in the high magnetic field regime ( @xmath18 ) .
the theoretical predictions do not match the measured linear field dependence , @xmath19 ( where @xmath20 is the cyclotron energy ) , that we have observed@xcite in high magnetic fields both at fixed filling factor and constant carrier density . from tunneling measurements of 2d-2d samples with different barrier widths , eisenstein _
et al._@xcite argue that the high field gap energy is largely determined by the in - plane coulomb energy , modified by an interlayer excitonic binding energy .
the form of the gap used to fit their data was @xmath21 , where @xmath22 is some universal function of filling factor , and @xmath23 is the quantum well center - to - center separation .
although the interpretation of our data is different to that of eisenstein _ et al._@xcite ( there is no universal @xmath22 which allows us to plot our high field data in this form ) , we measure gaps of comparable energy in lower mobility , but otherwise similar gaas / algaas double - layer structures . in a 2d-3d tunneling system ashoori _
et al._@xcite measured a filling factor independent suppression of tunneling for @xmath24 .
the temperature dependence of the conductance was fitted to a model with a linear variation of the dos about the fermi energy @xmath25 .
the half - width at half - depth of the gap was measured to be linear in magnetic field , with a dependence @xmath26 .
taking into account the difference in the two tunneling systems and the definitions of @xmath27 and @xmath28 , the two gap widths can be argued@xcite to be comparable . here
we consolidate earlier work@xcite at zero and high magnetic fields , and present data taken at intermediate fields .
the rest of this paper is organized as follows .
the samples and measurement details are described in sec .
[ s : exp ] .
the basic tunneling formalism is reviewed in sec .
[ s : formal ] , showing how the tunneling conductance is determined by the underlying electron spectral functions within each 2deg . in sec .
[ s : zerob ] we present tunneling measurements in zero magnetic field , investigating the various scattering mechanisms which are responsible for the linewidth of the resonance .
the tunneling properties in a perpendicular magnetic field are investigated in secs .
[ s : perpb ] and [ s : temp ] .
experimental details [ s : exp ] the measurements presented in this paper were obtained from two samples , c751 and c770 , each consisting of two modulation - doped 180 wide gaas quantum wells separated by a 100 ( c770 ) or 125 ( c751 ) @xmath29 barrier .
the double - well structures were grown by molecular beam epitaxy on an @xmath30 epilayer , which had previously been patterned by a focussed ga ion beam to form backgates.@xcite optical lithography was used to fabricate a mesa , ohmic contacts , and to define surface schottky gates .
small front ( back ) gates near the ohmic contacts were used to selectively deplete the upper ( lower ) 2deg , forming an independent contact to the lower ( upper ) layer .
the two 2degs overlap in a @xmath31m@xmath32 m tunneling region .
the as - grown carrier densities of the upper and lower layers in the tunneling region are @xmath33 @xmath34 and @xmath35 @xmath34 .
the corresponding , as - grown , low - temperature , mobilities of the two layers are @xmath36 @xmath37/vs and @xmath38 @xmath37/vs .
the carrier densities in the tunneling region are independently controlled by voltages @xmath39 and @xmath40 applied to large area front- and backgates .
two - terminal conductance measurements of the sequential depopulation of landau levels ( lls ) in each layer were used to determine the relationship between @xmath41 and the corresponding carrier density @xmath42 .
for example , in c770 the top layer has a density @xmath43 @xmath34 , where @xmath44 v. the carrier density at @xmath45 can vary by as much as 10% from one sample cooldown to another .
by contrast , the constant of proportionality ( 7.1 ) between @xmath39 and @xmath46 is always the same , and is in good agreement with a model where the top 2deg and the front gate form a parallel plate capacitor of separation 1000 .
the differential tunneling conductance @xmath47 was measured using constant voltage ( 0.1 mv ) ac lockin techniques as a function of carrier density @xmath1 , temperature @xmath0 , and applied dc interlayer source - drain voltage @xmath48 .
for measurements in a perpendicular magnetic field @xmath2 the carrier densities were always matched , and the given values of @xmath49 refer to the filling factors in each layer .
tunneling formalism [ s : formal ] tunneling accesses the single - particle dos ( the dos available for the addition of another electron ) , compared to conventional transport measurements which provide information about the dos available for the excitation of an electron from the fermi sea into a conducting state .
zheng and macdonald@xcite showed theoretically that 2d-2d tunneling characteristics in zero magnetic field can be used to determine the spectral function of an electron . within layer @xmath50 [ @xmath51 ( upper , lower ) ] scattering introduces an energy broadening @xmath52 , where @xmath53 is the scattering time . in the born approximation the spectral function within each 2deg will have the lorentzian form@xcite @xmath54 where @xmath4 is measured from the bottom of the 2d subband . using standard notation@xcite the tunneling current @xmath55 between the two layers
is calculated from the two spectral functions to be @xmath56,\end{aligned}\ ] ] where @xmath57 is the fermi distribution function .
the chemical potentials @xmath58 and the 2d subband energies @xmath59 are defined in fig .
[ f : band ] .
the integral is simplified by the following assumptions : the absolute conservation of momentum ( such that the interlayer tunneling matrix element @xmath60 is proportional to @xmath61 ) , low temperature ( @xmath62 , with fermi functions approximated by step functions ) , and weak disorder ( @xmath63 ) . using these assumptions , substitution of lorentzian spectral functions into eq .
[ e : current ] gives the tunneling current @xmath64 where @xmath65 , and @xmath66 is a constant with the dimensions of conductance .
the tunneling linewidth is the sum , @xmath67 , of the widths of the lorentzian spectral functions in the two layers.@xcite in this paper we report differential conductance @xmath68 measurements .
the zero - bias conductance obtained by differentiating eq .
[ e : current ] is @xmath69 using lorentzian spectral functions , the differential conductance as one subband edge ( controlled by a gate voltage @xmath70 ) is swept past the other is@xcite @xmath71 equation [ e : conductance2 ] predicts lorentzian @xmath72 characteristics with a linewidth @xmath9 and peaked with magnitude @xmath66 at @xmath73 , when the carrier densities in the two wells are equal .
we are interested in the differential conductance as a function of @xmath48 , taken when the carrier densities have been matched at @xmath74 .
however , upon application of an interlayer dc voltage the capacitance between the two layers unavoidably mismatches the carrier densities , and it is _ incorrect _ to set @xmath75 in the derivative of eq .
[ e : current2 ] to obtain the conductance @xmath76 there are in fact two contributions to @xmath77 : one component @xmath78 is due to the applied gate voltages ( this is zero when the carrier densities are matched at @xmath74 ) , and the other , @xmath79 , is a capacitive effect .
@xmath79 is determined by the change of carrier density @xmath80 induced by @xmath48 , where @xmath81 is the capacitance per unit area between the two 2degs .
the capacitive change in @xmath77 is @xmath82 , where @xmath83 is the 2d density of states , and @xmath84 ( where the bohr radius is @xmath85 in gaas ) .
substitution of @xmath86 into eq .
[ e : current2 ] , followed by differentiation , gives the modified lineshape @xmath87 setting @xmath88 ( for matched carrier densities at @xmath74 ) gives the lineshape @xmath89 which has a linewidth of @xmath90 , in contrast to the linewidth @xmath9 given in eq .
[ e : noneqbm1 ] .
a typical value of @xmath91 is 0.15 .
therefore , linewidths obtained from @xmath92 measurements are approximately 18% larger than those obtained from equilibrium conductance measurements.@xcite similar capacitive corrections should also be applied to linewidths obtained@xcite from plots of @xmath93 versus @xmath48 .
we can only quantify capacitive effects at @xmath94 , though the magnitude of such effects in a perpendicular magnetic field can be estimated by replacing @xmath95 with @xmath96 , the density of states at the fermi level in the outer landau level .
if the outer ll is half - filled , @xmath97 , then the capacitive change of the fermi energies @xmath98 will be negligible .
the opposite result is obtained near integral filling factors and so we have not made any conclusions from tunneling measurements in this regime . in sec .
[ s : temp ] we model high field tunneling measurements obtained at filling factor @xmath99 .
to model the data we have guessed the form of the spectral density @xmath100 , and have calculated the tunneling current using the expression @xmath101 .
\label{e : spect}\end{aligned}\ ] ] tunneling in zero magnetic field [ s : zerob ] figure [ f : lorentz ] shows the zero - field differential tunneling conductance @xmath102 as a function of @xmath39 , the front gate voltage ; by fixing @xmath40 , the carrier density in the bottom layer was held constant at @xmath103 @xmath34 .
the equilibrium measurements were taken at @xmath104 k ( squares ) and 19.0 k ( circles ) . due to a misalignment between the frontgate and backgate in the tunneling region , there is an unwanted second resonance ( which is independent of @xmath40 ) , as well as a small constant background conductance . in fig .
[ f : lorentz ] both of these weak contributions have been removed , and the corrected data at the two temperatures have been fitted to the lorentzian lineshape @xmath105 which is eq .
[ e : conductance2 ] rewritten explicitly in terms of the front gate voltage @xmath106 .
the fits , shown as solid lines , are with @xmath107 v , @xmath108 v , and @xmath109s at 3 k , and with @xmath110 v , @xmath111 v , and @xmath112s at 19 k. we have also tried fitting the resonances to a gaussian and to the derivative of a fermi function ( @xmath113 $ ] ) . of the three functional forms ,
a gaussian fit is the poorest , @xmath114 is considerably better , but the overall shape and the tails of the resonance are best fit by a lorentzian .
the half width at half maximum of the low - temperature lorentzian resonance , @xmath115 mv , corresponds to an energy broadening @xmath116 mev .
this linewidth @xmath9 is the sum of the widths of the spectral functions in the double - layer system , @xmath117 .
we relate the linewidth to the scattering rate , defining the average scattering rate of the double - layer system to be @xmath118 .
non - equilibrium measurements provide an alternative method for determining the linewidth @xmath9 .
figure [ f : betafig ] shows a comparison of the non - equilibrium @xmath92 ( squares ) and equilibrium @xmath119 characteristics ( circles , and offset vertically ) for the same @xmath120 for sample c751 . to demonstrate the validity of eq .
[ e : noneqbmgen ] , the non - equilibrium data was obtained with a small mismatch of carrier densities , causing a slight shift of the central @xmath92 peak to the left of @xmath74 .
the solid line through the @xmath92 data is a fit to eq .
[ e : noneqbmgen ] with @xmath121 mev , corresponding to a carrier density mismatch of 0.7% .
though small , this mismatch is enough to cause a noticeable asymmetry in the @xmath92 characteristics .
the fits to both the equilibrium and non - equilibrium traces in fig .
[ f : betafig ] give @xmath122 mev and @xmath123s .
the fitted value @xmath124 is obtained from the non - equilibrium trace and corresponds to an interlayer separation of @xmath125 , somewhat larger than the quantum well center - to - center separation of 305 deduced from the growth parameters . however , both parallel magnetic field measurements ( using the method of ref . ) and self - consistent calculations suggest a center - to - center wavefunction separation of @xmath126 - 350 .
the successful fits to both the equilibrium and non - equilibrium resonances in fig .
[ f : betafig ] provide strong evidence for a lorentzian spectral function in zero magnetic field .
we have been able to fit lorentzian lineshapes to the equilibrium tunneling resonances obtained over a wide range of carrier densities , @xmath127 @xmath34 , at temperatures from 3 k to 19 k. from the @xmath1 and @xmath0 dependence of the linewidth @xmath9 , we have investigated the electron - impurity , electron - phonon , and electron - electron contributions to the overall scattering rate @xmath128 figure [ f : widtht ] shows the temperature dependence of the equilibrium linewidth @xmath9 at matched carrier densities of @xmath129 @xmath34 .
the solid lines are fits of the form @xmath130 ; at high @xmath1 the fits are good , but at lower densities the measured linewidth departs from quadratic behavior .
the fitting parameters are both functions of @xmath1 , with typical values @xmath131 - 0.8 mev and @xmath132 mev / k@xmath133 . only the electron - phonon and electron - electron scattering terms in eq .
[ e : rate ] are expected to show significant temperature dependence . in the temperature range considered here ,
the sheet resistivity @xmath134 of each 2deg shows only a weak temperature dependence , indicating that the mobility is dominated by impurity scattering , and that , by comparison , the temperature dependent electron - phonon contribution is small .
we therefore associate the temperature dependent component of the tunneling linewidth with that due to electron - electron scattering @xmath135 .
the electron - electron scattering rate in a clean 2d metal is given by the well - known theoretical expression@xcite @xmath136,\ ] ] where @xmath137 is the 2d thomas - fermi screening wavevector . at fixed carrier density
@xmath138 exhibits an approximate @xmath139 temperature dependence , in agreement with the solid line fits used in fig .
[ f : widtht ] . at low temperatures ,
the electron - electron scattering is negligible , and @xmath140 is dominated by electron - impurity scattering .
figure [ f : logglogn ] shows a log - log plot of the equilibrium linewidth versus carrier density at @xmath141 k. the fitted gradient is @xmath142 , which is approximately the power - law behavior expected@xcite for small - angle scattering from remote ionized impurities , @xmath143 where @xmath144 is the sheet density of impurities , and @xmath145 is the spacer layer thickness .
equation [ e : gold ] is valid for @xmath146 , where @xmath147 is the fermi wavevector . in our samples
the ionized donors are distributed over distances of 400 - 700 from the center of each quantum well . taking the smallest distance , @xmath148 , the condition @xmath149 is easily satisfied for @xmath150 @xmath34 , though the applicability of eq .
[ e : gold ] becomes questionable at the lowest carrier densities . from experiments over a range of carrier densities and temperatures
we have fitted over 300 measured equilibrium linewidths @xmath151 to the expression @xmath152.\ ] ] if @xmath9 is measured in units of mev , @xmath1 in units of @xmath153 @xmath34 , and @xmath0 in k , the best fit was obtained with the numerical values @xmath154 and @xmath155 .
the dashed lines in fig .
[ f : widtht ] are obtained from this best fit .
the magnitude of the electron - impurity scattering rate , @xmath156 in eq .
[ e : ratent ] , is in agreement with other experiments,@xcite but is fifteen times smaller than the theoretical prediction of eq .
[ e : gold ] .
discrepancies between theory and experiment have previously been resolved@xcite by including correlations between scatterers . in the high @xmath1 , low @xmath0 regime , the ratio @xmath157 ( where @xmath158 is the mobility lifetime ) agrees with measurements@xcite of similar mobility single - layer gaas / algaas heterostructures .
furthermore , the scattering time @xmath159 ps is comparable to the dingle time 1.1 ps ( obtained from shubnikov - de haas oscillations measured on a hall bar made from the same wafer ) , confirming the importance of small - angle scattering in determining the tunneling linewidth .
to look at just the electron - electron scattering term , fig .
[ f : univ ] shows @xmath160 plotted versus @xmath161 .
this representation of the data is expected to be nearly universal,@xcite though there is some dependence on carrier density inside the square bracket of eq .
[ e : gnq ] .
this weak dependence on @xmath1 is illustrated by the two solid line traces of 3.06@xmath162[eq .
[ e : gnq ] ] ( our best fit for @xmath135 ) for @xmath163 and @xmath164 @xmath34 ( upper and lower solid curves , respectively ) .
equation [ e : gnq ] is plotted for @xmath165 @xmath34 and without any prefactor ( @xmath166 ) as the dashed line labelled gq in fig .
[ f : univ ] .
since the original derivation by giuliani and quinn@xcite there has been some confusion about the value of @xmath167 that multiplies this expression .
fukuyama and abrahams@xcite calculated the expression @xmath168 and this is shown in fig .
[ f : univ ] as the curve labelled fa.@xcite more recent theories@xcite highlight the mathematical errors in ref . and suggest that due to the nature of the tunneling experiment , there are both electron and hole contributions to the linewidth .
these two considerations result in @xmath169 , giving the third dashed line in fig .
[ f : univ ] labelled jm . without any energy dependent corrections the various predictions ( gq , fa , and jm ) can be compared to our equilibrium measurement @xmath170 .
the difference between the predictions of gq and jm and our fit is purely in the value of the prefactor @xmath167 ; whereas the curve fa has a functional form which does not describe the data very well .
the greatest discrepancies between theory and experiment in fig .
[ f : univ ] occur for large @xmath161 , for which there are various possible reasons .
first , it is assumed in the derivation@xcite of eq .
[ e : gnq ] that @xmath171 , a condition which breaks down at low densities and high temperatures .
second , when the matched carrier densities are reduced there is an increase in the disorder of the two layers , and the 2degs may no longer be considered as clean 2d metals .
the increased level of disorder may also render the born approximation invalid , in which case the spectral functions and the tunneling resonance are no longer lorentzian .
third , the derivation of lorentzian @xmath119 characteristics ( eq . [ e : lorentz2 ] ) from lorentzian spectral functions is only valid when @xmath172 ; the two energies become comparable when @xmath173 @xmath34 .
the fourth reason is experimental : the pinch - off characteristics of the gate ( @xmath174 distort the symmetric equilibrium lineshape , increasing the uncertainty of measured linewidths at low @xmath1 . in our analysis of the temperature dependence of the equilibrium linewidth we have followed an approach similar to that of murphy _
et al._,@xcite where the non - equilibrium linewidths were measured from @xmath93 versus @xmath48 characteristics of double 2deg structures of higher mobility ( @xmath175 @xmath37/vs ) . in our samples the measured electron - electron scattering rate is enhanced over eq .
[ e : gnq ] by a factor @xmath176 , which is two times smaller than that measured in ref . .
the two measurements are different , and non - equilibrium linewidths are expected to be larger than equilibrium linewidths for two reasons : an 18% enhancement can be attributed to charge transfer effects ( see sec .
[ s : formal ] ) , and a further 15% ( ref . ) to 40% ( ref . ) enhancement may be introduced by the energy dependence of @xmath135 . in order to achieve an enhancement factor @xmath177 close to that of murphy _
et al._@xcite recent theories@xcite have concentrated on the energy dependence of @xmath135 , but have neglected the interlayer charge transfer effects . in fig .
[ f : betafig ] we show that , in our samples , we can obtain good agreement between equilibrium and non - equilibrium linewidths by including only charge transfer effects . moreover , in further investigations over a range of matched and mismatched carrier densities , we find that the energy dependent corrections are consistently small .
there are two major differences between our samples and those of ref . that might account for the different measured values of @xmath167 .
the first is the level of disorder ; in our samples the electron - impurity contribution to the linewidth is @xmath178 mev compared to 0.25 mev in ref . .
this greater disorder probably does not affect the applicability of eq .
[ e : gnq ] , but will make the energy dependent corrections to @xmath135 less important.@xcite the second difference is the thickness of the tunnel barrier , which will affect the strength of the interlayer electron - electron interactions . for samples with quantum well widths of 200 and barrier thicknesses in the range 175 - 340
( giving @xmath179 - 540 ) , murphy _
et al._@xcite measure an enhancement @xmath180 which appears to be independent of barrier thickness . in our samples
the interlayer electron - electron distance ( @xmath126 ) is comparable to the lowest used in ref . .
this could suggest that it is the greater disorder in our samples that gives rise to the smaller enhancement for @xmath135 . by way of comparison , we point out that berk _
et al._@xcite have measured electron - electron scattering rates similar to our own ( in the units used here , @xmath181 ) , in 2degs with a similarly low mobility @xmath182-@xmath183
@xmath184vs but more strongly coupled ( @xmath185 ) .
the effect of stronger tunneling ( as quantified by the symmetric - antisymmetric gap @xmath186 ) on @xmath135 has been calculated by slutzky _
et al._@xcite tunneling in a perpendicular magnetic field [ s : perpb ] the application of a perpendicular magnetic field @xmath2 has a marked effect on the resonant tunneling properties between two 2degs .
figure [ f : tunnsdh ] shows the equilibrium conductance @xmath187 , at a matched carrier density of @xmath188 @xmath34 , as a function of magnetic field at 0.1 k ( solid line ) and 1.3 k ( dashed line ) . as @xmath2 is increased , the oscillations in the tunneling conductance mirror the formation of lls in both 2degs .
however , in contrast to sdh oscillations , which are an in - plane transport effect , the oscillations shown in fig . [ f : tunnsdh ] originate from the current flowing perpendicular to the 2degs.@xcite the oscillations in @xmath187 can be interpreted in terms of variations of @xmath96 , the dos at the fermi level in a field @xmath2 .
we will describe the temperature dependence of the equilibrium tunneling in two magnetic field regimes , @xmath189 t and @xmath190 t. below 0.6 t , the formation of lls modulates the dos , creating minima ( maxima ) at integral ( half - integral ) filling factors .
the equilibrium tunneling amplitude depends on the dos in both layers and shows quantum oscillations commensurate with @xmath96 .
when the temperature is increased from 0.1 to 1.3 k , the low field tunneling conductance follows the behavior of @xmath96 ; the amplitude of the conductance oscillations decrease as the modulation in @xmath96 becomes weaker , and @xmath187 tends towards its zero - field value .
for @xmath189 t the magnitude of the conductance maxima increase with magnetic field , following the expected increase in the degeneracy of a half - filled ll .
however , for @xmath190 t the amplitude of the tunneling maxima _ decrease _ with increasing magnetic field ; this can be interpreted as a magnetic - field - induced suppression of the equilibrium tunneling between the two 2degs , with the formation of a gap in the tunneling dos . when the temperature is raised to 1.3
k the suppression of the conductance is reduced , indicating thermal activation across the gap . the fig .
[ f : tunnsdh ] inset shows low - field equilibrium tunneling for a matched carrier density of @xmath191 @xmath34 . at this higher carrier density
the crossover from sdh - like to activated behavior again occurs at @xmath3 t. these measurements at different temperatures show that the signature of the magnetic - field - induced gap is evident at filling factors as high as @xmath192 . clearly , above 0.6 t a mechanism other than the simple formation of lls is needed to explain the unusual temperature dependence observed at half - integral filling factors .
the formation of a magnetic - field - induced gap can also be investigated by non - equilibrium measurements , where at low magnetic fields a dip around zero bias is observed in the @xmath92 characteristics .
figure [ f : gap ] shows @xmath92 characteristics at various magnetic fields for a matched carrier density of @xmath193 @xmath34 . in the trace at @xmath194 t there is a slight dip at @xmath74 ; with increasing magnetic field both the depth of the dip and the separation of the surrounding peaks become larger . to characterize the tunneling gap ,
we define @xmath28 to be the voltage separation of the conductance peaks either side of @xmath74 .
previously , we have measured@xcite @xmath55 characteristics for @xmath195 and have successfully fitted our low voltage data to the expression @xmath196 . at high fields the values of @xmath28 obtained from such fits are precisely the same as those measured from the peak separations.@xcite although at low magnetic fields we can not fit our data to this theoretical expression for @xmath197 , we nevertheless use the peak separation as a straightforward measure of the gap . to test various theoretical predictions
we have investigated @xmath28 at fixed @xmath1 and fixed @xmath49 , both as a function magnetic field @xmath2 .
figure [ f : gapb ] shows the gap parameter @xmath28 , measured from traces similar to those in fig .
[ f : gap ] , over a wider range of @xmath2 . due to the presence of edge states we disregard measurements close to the integer quantum hall ( qh ) regime , as well as those near the fractional qh state at @xmath198 .
the linear fit through the remaining points ( solid circles ) is @xmath199 mev .
figure [ f : gapb ] shows that the gap is most cleanly observed at half - integral filling factors , and in measurements at fixed @xmath1 there are more data points at low @xmath49 than at high @xmath49 . with this proviso
the linear fit in fig .
[ f : gapb ] spans the whole field regime , though it is difficult to identify deviations at low @xmath2 .
we have also performed conductance measurements at a higher matched carrier density of @xmath200 @xmath34 , and fig .
[ f : greyscale ] shows the evolution of the the low field tunneling characteristics at this carrier density presented as a grey - scale plot .
the figure is made up of @xmath92 characteristics , similar to those in fig .
[ f : gap ] , taken at 101 magnetic fields from @xmath94 to 2 t. to accentuate the data each trace has been scaled between 0 ( black ) and 1 ( white ) .
the light structures around the center , marked with solid lines , are the two tunneling peaks surrounding the gap , with a splitting that is linear in magnetic field , @xmath201 mev . within experimental error the gaps measured at the two carrier densities , @xmath202 @xmath34 and @xmath203 @xmath34 , have the same magnetic field dependence . to obtain more information about the gap at higher filling factors , fig .
[ f : gapnu ] shows measurements of @xmath204 taken as a function of @xmath2 while maintaining a half - integral filling factor in the tunneling region .
gap measurements were obtained at @xmath205 for seventeen matched carrier densities in the range @xmath206 @xmath34 . at a given magnetic field , the higher the half - integral filling factor , the smaller the gap .
the dos is not clearly spin - split below @xmath207 t , and so data points for the highest filling factor and lowest fields are subject to some error .
the straight line @xmath208 mev is included in fig .
11 to demonstrate the linear @xmath2 dependence of the tunneling gap at @xmath99 , and the sublinear nature of the gap at higher filling factors . at low carrier densities @xmath209 @xmath34
, measurements of @xmath210 converge to a value similar to that measured at @xmath99 . for the higher carrier concentrations the gap for @xmath211
is reduced in comparison with @xmath212 .
this observation , plus the scarcity of data points at higher filling factors in measurements at fixed @xmath1 accounts for the same measured gap at @xmath213 @xmath34 and @xmath214 @xmath34 .
we stress that , as previously reported,@xcite the high field gap shows the same linear behavior both at fixed @xmath1 and fixed @xmath18 .
figures [ f : gap ] and [ f : greyscale ] also show tunneling between lls of different index . at high magnetic fields ,
the first inter - ll tunneling transition has been observed@xcite as a peak in the @xmath55 characteristics at @xmath215 , and the equivalent conductance measurement@xcite exhibits a peak at @xmath216 .
both measurements show inter - ll transitions occuring at higher than expected bias voltages , and it is proposed@xcite that this enhanced ll spacing is due to an effective mass reduced by many - body effects.@xcite the arrows in fig . [
f : gap ] and the dashed lines superimposed on fig .
[ f : greyscale ] show the position of the first inter - ll tunneling peaks either side of zero bias . over the range @xmath217 t
the conductance peaks occur at @xmath218 , an enhancement that is only slightly less than that observed at higher magnetic fields .
if the enhanced ll spacing is due to a reduced effective mass , our measurements suggest that the mass is reduced at higher filling factors by an amount much greater than that predicted by smith _
et al._@xcite there has been much interest in tunneling in a weak magnetic field where , even for filling factors @xmath219 , it is believed that there is an energy cost associated with the injection and extraction of electrons into and out of a 2deg .
it has been theoretically shown@xcite that at low magnetic fields the bare coulomb interaction ( which typically exceeds the cyclotron energy @xmath220 ) is screened by the filled lls , and the renormalized interaction is sufficiently weak that it does not mix lls .
therefore only interactions between electrons in the outer partially filled ll need to be considered ; this is the starting point for many of the theories in the weak field regime . following the treatment at high magnetic fields , the gap at lower fields and higher filling factors can be considered to be the coulomb energy , where the cyclotron length @xmath221 replaces the magnetic length @xmath222 as the relevant length scale .
the semiclassical energy gap is @xmath223 where @xmath224 is the interaction parameter , and is given by @xmath225 .
levitov and shytov@xcite consider the problem of an electron orbiting in a ring of thickness @xmath222 and radius @xmath226 in one layer , tunneling into a similar ring in the second layer , leaving behind a hole in the first ring .
the electrostatic energy of this two - ring capacitor creates an energy gap that is proportional to eq .
[ e : semigap ] , but multiplied by a logarithmic screening term .
alternatively , using both a hydrodynamic approach@xcite and a hartree - fock calculation,@xcite aleiner and coworkers have obtained a gap of the form @xmath227 it has been shown@xcite that the total gap has contributions from two terms similar in form to @xmath228 and @xmath229 , where @xmath228 is the dominant term . in a recent theory , fogler _ et al._@xcite propose that under certain conditions
the ground state of the 2deg in a weak magnetic field is a charge density wave ( cdw ) superimposed on a uniform background of filled lls .
tunneling into a partially filled ll disturbs some of the correlations of the cdw , and the energy cost of this disruption gives rise to a tunneling gap @xmath230 as @xmath2 is decreased there is a predicted@xcite crossover from the spin - resolved qh regime ( spin - split lls ) where the ground state is a cdw , to the spin - unresolved qh regime ( well separated , but spin - degenerate lls ) .
the cdw ground state can only exist in the former regime , while at lower fields the ground state is a uniform electron liquid which is expected@xcite to have a gap of a different origin . in our samples
the sdh oscillations of the individual 2degs show that the spin - unresolved qh regime occurs over a narrow range of magnetic field .
the fig .
[ f : tunnsdh ] inset shows that at @xmath150 @xmath34 it is possible to observe a suppression of tunneling ( above 0.6 t ) in the absence of spin - splitting ( which becomes distinct for @xmath231 t ) .
moreover , in all our measurements over a wide range of @xmath1 , we see no clear change in the character of the gap once spin - splitting has set in . the @xmath2 dependence of the low field gaps predicted by aleiner _ et al._@xcite and fogler _
et al._@xcite are determined by @xmath229 , which varies as @xmath232 at fixed carrier density , and as @xmath233 at fixed filling factor .
figure [ f : gapb ] shows that at fixed @xmath1 the gap is linear ( not quadratic ) in @xmath2 , and fig .
[ f : gapnu ] shows that at fixed @xmath49 the weak field gap @xmath234 is sublinear in @xmath2 . the semiclassical gap ( eq . [ e : semigap ] ) behaves as @xmath235 at fixed carrier concentration , and as @xmath236 at fixed filling factor . for fixed @xmath1
the measured gaps are indeed linear in @xmath2 , though we find that the best fits are obtained with a negative intercept .
similarly the best fits of @xmath204 to a @xmath17 field dependence are also obtained with a negative intercept , using the expression @xmath237 . in the inset to fig .
[ f : gapnu ] we have plotted @xmath238 versus @xmath239 for @xmath240 . taking @xmath241 mev the data collapse onto a single straight line with slope @xmath242 .
we conclude from measurements at constant @xmath1 and constant @xmath49 that the high filling factor gap is best described by @xmath243 the negative intercept could be interpreted as evidence for an excitonic effect ( similar to that observed@xcite at @xmath99 ) , and with a magnitude which is approximately independent of filling factor for @xmath240 .
we were unable to obtain good agreement between eq .
[ e : modgap ] and the measured gaps at @xmath99 and 3/2 , instead much larger negative intercepts ( @xmath244 and 1.7 mev , respectively ) were required . with
our samples we have obtained@xcite better agreement at @xmath99 using a linear rather square root magnetic field dependence for the gap @xmath212 up to 25 t. temperature studies in high fields [ s : temp ] in this section we use eq .
[ e : spect ] to model high field @xmath92 measurements taken at @xmath99 , results which have been presented elsewhere.@xcite figure [ f : tempdep](a ) shows the conductance characteristics at @xmath245 t at a matched carrier density of @xmath246 @xmath34 . as the temperature is increased from @xmath247 k to 6 k , the equilibrium conductance @xmath248 increases , while the height of the surrounding peaks is reduced .
figure [ f : tempdep](b ) shows @xmath187 in an arrhenius plot , revealing activated behavior at low temperatures with an activation energy of @xmath249 mev . above the activation temperature ( @xmath250 k ) the equilibrium conductance departs from exponential behavior , and for @xmath251 k the conductance decreases .
we have measured @xmath252 at different matched carrier densities and magnetic fields , while maintaining a filling factor of @xmath99 in the tunneling region .
figure [ f : eab ] shows @xmath252 as a function of magnetic field , with a fit which shows that the activation energy is proportional to @xmath2 . at @xmath253
the ratio @xmath254 is @xmath255 , and is approximately independent of @xmath2 .
activation plots of the high field gap were first obtained by eisenstein _
et al . _ , and for comparison we measure from fig . 3 of ref . the ratio @xmath256 at @xmath257 t. temperature can be used to probe the shape of the tunneling dos , and previously ashoori _ et al._@xcite have used a dos with a linear dip at the fermi level to model 2d-3d tunneling measurements .
a linear dos does not give activated behavior , and to obtain such characteristics we have modeled the data using the double - gaussian spectral density @xmath258 where the fermi level @xmath25 lies in the minimum between the two peaks .
this tunneling dos is a modified version of eq . 14 in ref . , which was originally proposed for @xmath259 and @xmath260 . by comparison with the original formula
, we have introduced an adjustable parameter @xmath261 in the denominator of the exponential . as @xmath262
the double - gaussian spectral density consists of two @xmath263-functions at @xmath264 , and so the gap @xmath265 at @xmath266 . at low temperatures the activation energy @xmath267 , and so @xmath28 and @xmath252 share the same magnetic field dependence ( that of @xmath268 ) , in agreement with experiment . to simulate the conductance measurements in fig .
[ f : tempdep ] we have chosen @xmath269 so that the theoretical ratio @xmath270 agrees with the experimental value of 0.047 .
@xmath55 characteristics have been calculated using eq .
[ e : alsdos ] in eq .
[ e : spect ] , and by numerical differentiation of these characteristics we obtain the @xmath92 traces shown in fig .
[ f : gausst](a ) . the value of @xmath271 mev best reproduces the conductance and activation characteristics ( see fig . [
f : gausst](b ) ) at 8 t. from similar simulations of @xmath99 data at different magnetic fields we have measured the @xmath2 dependence of @xmath268 ; the results are displayed in fig .
[ f : eab ] , with a fit showing @xmath268 with a linear dependence on the magnetic field , @xmath272 .
the calculated @xmath92 curves reproduce many features of the experimental data .
first , there are bias voltages which delineate regions of positive and negative @xmath273 ; second , the conductance maxima move closer together with increasing temperature ; third , it shows that at higher temperatures @xmath187 is no longer activated , but reduces with increasing temperature .
there are , however , some differences between the model and experiment . the very low temperature limit of the experimental characteristics differ from those of the model ; in the experimental system disorder introduces a finite width to the dos at @xmath266 .
we have tried to simulate this disorder by introducing a finite offset temperature , that is by replacing @xmath0 with @xmath274 in eq .
[ e : alsdos ] , where @xmath275 is a constant .
however , there was little improvement in the shape of the simulated characteristics .
conclusions [ s : conclude ] we have used both the equilibrium and non - equilibrium tunneling characteristics to show that the underlying spectral function , @xmath6 , within each of the 2degs of a double - layer system is a lorentzian .
equilibrium measurements show that the lorentzian tunneling linewidth has a carrier density and temperature dependence that can be attributed to electron - impurity and electron - electron scattering .
the measured electron - impurity scattering rate is similar in magnitude to the dingle time , and comparison with theory shows that correlations between scatterers need to be taken into account to obtain better agreement with experiment . the mathematical form , but not the magnitude , of the electron - electron scattering rate measured from the equilibrium lineshape is well described by theory . in a perpendicular magnetic field
the equilibrium tunneling conductance oscillates as a function of @xmath2 , reflecting the formation of lls . for @xmath276
t the magnetic field creates a gap in the tunneling dos which suppresses the equilibrium tunneling . for fixed filling factor @xmath277
the tunneling gap has a linear magnetic field dependence , whereas at higher half - integral filling factors the gap is sublinear and is best described by a semiclassical gap .
we have modeled the temperature dependence of the conductance characteristics in a strong magnetic field using a double - gaussian spectral density . |
the cobe / dmr experiment ( @xcite ) has revolutionized the field of structure formation , providing the first robust evidence for primary , large scale anisotropy of the cosmic microwave background ( cmb ) . on these angular scales cmb anisotropies
are expected because of potential fluctuations at last scattering ( @xcite ) .
these linear fluctuations are predicted to be constant in time , and then provide a unique tool for determining the initial conditions out of which large scale structure formed . because of the low cobe / dmr angular resolution and of the correspondingly large cosmic variance ( @xcite ) , these initial conditions can not be yet determined with high precision ( @xcite , @xcite ) . in spite of this , scale - invariant initial conditions , a robust prediction of inflation , are indeed consistent with the cobe / dmr four years data ( @xcite ) .
it is worth remembering that the analysis of these data , based on the assumption of a flat universe , is completely insensitive to the chemistry of the universe .
it seems difficult to reconcile the bulk of the observational data ( cmb anisotropy , bulk flows , galaxy correlation function , etc . ) with a model which is not dynamically dominated by cold dark matter ( cdm ) .
however , it does not seems probable that cdm can provide the closure density : if this were the case , the universe would probably be more inhomogeneous than observed on small scales .
the excess power on these scales can be reduced , for example , by considering a low density ( @xmath7 ) cdm dominated model , which is flat because of a suitable cosmological constant , or a mixture of cold and hot dark matter , i.e. mixed dark matter ( mdm ) models . in this paper
we will focus on the latter scenario .
however , just because of the power reduction on small scales , in a mdm model structure formation tends to be a too recent process .
this difficulty can be alleviated by considering initial conditions that are not exactly scale invariant .
for example , it has been shown that the abundance of rich cluster in mdm models can be better reconciled with the x - ray luminosity function by assuming initial `` blue '' power spectra , i.e. @xmath8 with @xmath9 ( @xcite ) .
blue power spectra could overproduce small scale cmb anisotropy .
however , a possible late reheating of the intergalactic medium could have controlled the level of anisotropy at or below the degree angular scale . if and when the universe underwent through a phase of early reionization of the intergalactic medium is an interesting and still open question , although from the gunn and peterson ( @xcite ) test we know that the universe must have been highly reionized at redshift @xmath10 , and the presence of heavy elements in the intracluster gas suggest that a considerable energy release occurred during the earliest stages of galaxy formation and evolution
the recent detections of cosmic microwave background ( cmb ) anisotropy at degree angular scales provide a new and powerful way for investigating this issue .
in fact , the level of detected anisotropies at these scales can in principle discriminate among different assumptions for the thermal history of the universe .
also , using large and intermediate angular scales measurements increases the lever arm for determining the primordial spectral index . a definitive answer to these and
other open questions will come when space missions ( e.g. cobras / samba and map ) will provide a robust and definitive picture of the intermediate angular scale anisotropy .
meanwhile , the number of experiments reporting detections of anisotropy at degree angular scales has increased up to a couple of tens ( see table i below ) .
because of their sub - degree angular resolution , these experiments are quite effective in testing different reionization histories .
so , the goal of this paper is to discuss if these anisotropy data , together with cobe / dmr , discriminate among different assumptions for the primordial spectral index @xmath11 and for the thermal history of the universe .
we consider a flat universe ( @xmath12 ) composed by baryons ( @xmath13 ) , cold dark matter ( @xmath14 ) , one family of massive neutrinos ( @xmath15 ) , photons and two families of massless neutrinos . for age considerations we fix @xmath16 ,
i.e. @xmath17 . the basic equations for describing the time evolution of density fluctuations in these different cosmic components have been already derived ( @xcite , @xcite , @xcite ) . in fourier space
, they are : @xmath18 @xmath19 @xmath20 @xmath21 @xmath22 @xmath23 @xmath24 @xmath25 @xmath26 where @xmath27 .
eq.(1 ) describes the time evolution of the ( fourier transformed ) fluctuation in the phase space distribution of massive neutrinos : @xmath28 , where @xmath29 , @xmath30 is the ratio between the neutrino momentum @xmath31 and the neutrino temperature ( we use natural units ) , proportional to the cmb temperature , @xmath32 .
the massive neutrino background density is then @xmath33 where @xmath34 , @xmath35 is the number of neutrino families with a non vanishing rest mass ( @xmath36 in our case ) , @xmath37 , and each flavour state has two helicity states .
we expand @xmath38 in legendre polynomials , @xmath39 , and transform eq.(1 ) in a set of coupled differential equations for the multipoles @xmath40 ( see e.g. @xcite ) .
the massive neutrino density contrast reads @xmath41 while @xmath42 \epsilon^{-1}y g_0 .
$ ] for @xmath43 and @xmath44 , we have @xmath45 and @xmath46 , respectively . finally , @xmath47 .
the integrals in @xmath48-space giving @xmath49 , @xmath50 , @xmath51 and @xmath52 are performed numerically with a 16 points gauss - laguerre integration method .
eq.(2 ) describes massless neutrinos , and it is obtained by integrating in @xmath48- space eq .
( 1 ) in the ultra - relativistic limit ( @xmath53 ) : @xmath54 .
again we expand @xmath55 in legendre polynomials , @xmath56 and transform eq .
( 2 ) in a set of 40 coupled differential equations .
the density fluctuations of massless neutrinos is @xmath57 , while @xmath58 .
we numerically follow the evolution of fluctuations in this component only when the perturbation proper wavelength is larger than one tenth of the horizon . afterwards , free streaming rapidly damps fluctuations in this hot component .
eq.(3 ) describes fluctuations in the @xmath59 and @xmath60 stokes parameters of the cmb ( see e.g. @xcite ) and differs from eq .
( 2 ) because of the collisional term , proportional to the thomson cross section , @xmath61 , and to the electronic number density , @xmath62 .
the latter is evaluated assuming a standard recombination history ( @xcite ) , with a helium mass fraction @xmath63 .
we expand @xmath64 and @xmath65 in legendre polinomials : @xmath66 and @xmath67 .
the number of harmonics is automatically increased up to @xmath68 for @xmath69 .
the correlation function ( acf ) of the temperature fluctuations can be written as @xmath70 \eqno(9)\ ] ] where @xmath71 is the dispersion of a gaussian approximating the angular response of the beam , and @xmath72 we define the parameter @xmath73 as the amplitude @xmath74 of the power spectrum ( considered as a free parameter ) in units of @xmath75 , the amplitude needed to reproduce @xmath76 , as observed by cobe - dmr ( @xcite ) . performing the integral in eq.(10 ) with high accuracy at high @xmath77 s
requires a very good sampling of the @xmath78 s in @xmath79-space , and this is a heavy computational task . to avoid this problem we sample the interval @xmath80 with a step @xmath81
, we evaluate the integral in eq.(10 ) , and we use a smoothing algorithm to suppress the high frequency , sample noise .
the @xmath82 s obtained in this way differ from those obtained with a much denser @xmath79-space sampling ( @xmath83 ) by only a fraction of percent up to @xmath84 .
the time evolution of the baryon and cdm density contrasts and of the baryon peculiar velocity are described by eq.(6 ) , ( 8) and ( 7 ) respectively .
the system is closed by eq.(4 ) and ( 5 ) describing the field equations for the trace and the @xmath85 component of the metric perturbation tensor .
we numerically integrate the previous equations from redshift @xmath86 up to the present . in the following
we will also assume that the universe reionized instantaneously at redshift @xmath87 , and remained completely reionized up to the present . because of the low baryonic abundance , we need reionization at high redshifts in order to substantially suppress the anisotropy at degree scales . for @xmath88 and @xmath89 ,
there is a probability lower than 50 @xmath5 for a photon to scatter against a free electron at @xmath90 .
we have selected a set of 20 different anisotropy detections obtained by different experiments , or by the same experiment with different window functions and/or at different frequencies . for each detection , labeled by the index @xmath91 , we report in table 1 the detected mean square anisotropy , @xmath92 , and the corresponding 1@xmath93 error , @xmath94 . in our error estimates , the calibration error was added in quadrature to the statistical error . when not explicitly given , we estimated the mean square anisotropy as follows : @xmath95 - \sigma_j^2 $ ] , where @xmath96 are the published anisotropy data of the @xmath91-th experiment , @xmath97 $ ] is the variance of the data points , and @xmath98 is the mean square value of the instrumental noise . theoretically , the mean ( over the ensemble ) squared anisotropy is given by a weighted sum of the @xmath82 s : @xmath99 = { 1 \over 4\pi } \sum_\ell ( 2 \ell + 1 ) c_\ell w_{\ell , j}= \overline { { c_{\ell_{eff}}}\over{4\pi } } \sum_\ell ( 2 \ell + 1 ) w_{\ell , j } \eqno ( 13)\ ] ] where the windows function @xmath100 contains all the experimental details ( chop , modulation , beam , etc . ) , and @xmath101 is the mean value of the @xmath82 s over the window function .
the effective multipole number @xmath102 is defined as follows : @xmath103 and is listed in table i for a scale invariant model without reionization .
although in principle model dependent , the values of @xmath104 are quite stable because of the narrowness in @xmath77 space of the window functions . using numerical simulations , which take into account scan strategy and experimental noise , we verify that the expected distribution for @xmath105 is well approximated by a gaussian , with mean @xmath106 and a cosmic / sampling variance @xmath107 here @xmath108 represents the fraction of the sky sampled by each experiment and it is also listed in table i. given the gaussian distribution of @xmath109 , we compute the likelihood of the 20 ( assumed independent ) cmb anisotropy detections as follows : @xmath110 } } \exp \bigl\{- { 1 \over 2 } { \bigl[\delta _ j^{(exp ) } - { \cal a}^2 \delta^{(th)}_j\bigl]^2 \over { \cal a}^4 \bigl ( \delta^{(th)}_j \bigr)^2 + \sigma_j ^2 } \bigr\ } \eqno ( 15)\ ] ] as already stated , this is a function of three parameters : the amplitude @xmath111 , the spectral index @xmath11 and the reionization redshift @xmath1 . for each pair
@xmath112 we select the value @xmath113 which maximizes the likelihood ( isolevels of @xmath113 are shown in fig.1 for @xmath114 , respectively ) .
the results of our theoretical calculation are shown in fig.2a , where we plot the @xmath82 s for few models , each of them normalized with its own @xmath113 .
our results show that the difference between a pure cdm and a mdm ( @xmath115 ) models , normalized with the same @xmath113 is very tiny , @xmath116 up to @xmath117 and @xmath118 up to @xmath119 ( @xcite , @xcite , @xcite ) .
so we will make hereafter the assumption that the anisotropy pattern does not depend on @xmath120 , at least for @xmath121 .
this generalizes our assumption that the cmb anisotropy depends only upon three independent variables : @xmath111 , @xmath11 and @xmath1 . for the models shown in fig.1a we also plot the band - power estimate for the @xmath91-th experiment @xmath122 and in fig.1b the corresponding window functions , @xmath123 .
the 2-d , conditional distribution @xmath124 has a quite distinctive peak at @xmath125 and @xmath126 .
a @xmath127-confidence level contour in the @xmath112 is obtained cutting the @xmath128 distribution with the isolevel @xmath129 , and by requiring that the volume below the surface inside @xmath129 is a fraction @xmath127 of the total volume . in fig.3
we plot the @xmath130 and the @xmath131 contours , again for @xmath132 and @xmath133 , respectively ( we actually sampled a region of the @xmath112 plane much larger than shown in figure ) .
it is clear that the considered data set identifies a preferred region of the @xmath134 space .
the shape of the confidence contours can be understood by noting that increasing @xmath11 would overproduce anisotropies : a corresponding increase in @xmath1 is thus required to damp the fluctuations to a level compatible with observations . on the other hand , if @xmath135 is increased , the optical depth increases , thus requiring a lower reionization redshift to produce the detected level of degree - scale anisotropy .
this effect dominates over the smaller increase of primary cmb anisotropies .
the simple analysis carried out here does not take into account details on the scan pattern and/or the beam profile .
moreover , we know that the published error bars could not account properly for correlated errors or other more subtle effects . in order to test the significance of our analysis and its robustness against the published estimates for the experimental errors , we take a drastic point of view : we consider the best fit model ( @xmath136 , @xmath137 ) normalized to cobe / dmr ( i.e. we fix @xmath138 ) . for each degree
scale data point we compute the difference between the measured value and the value expected from the theory .
we then associate to each data point a 1@xmath93 error bar equal to the root mean square of these 19 differences ( instead of the published error bars ) .
the corresponding results of the likelihood analysis are in very good agreement with what is found using the appropriate errors : @xmath139 is slightly better , especially if the reionization redshift is increased .
so we do believe that our results are little affected by possible correlations in the errors and/or systematics .
our main conclusions , derived from fig.3 , are as follows .
the conditional likelihood shows a maximum for @xmath125 and @xmath140 , for @xmath141 , i.e. for baryonic abundances consistent with primordial nucleosynthesis .
we have checked the stability of this result by applying a jack - knife analysis to the considered data set , and we have seen that the results are reasonably stable : the best model has always @xmath142 and @xmath143 , unless cobe data are excluded ( in that case @xmath144 and @xmath126 ) .
the 95@xmath5 confidence contours in the @xmath112 plane include a wide range of parameters combinations .
this means that the presently available data set is not sensitive enough to produce `` precise '' determinations for @xmath11 and @xmath1 ; systematic and statistical errors in the different experiments are still significant .
\3 . if we exclude an early reionization of the intergalactic medium ( @xmath145 ) we get the following @xmath146 confidence level estimates for the spectral index : @xmath147 ( @xmath148 ) ; @xmath149 ( @xmath150 ) ; @xmath151 ( @xmath152 ) .
this has to be compared to the results from cobe - dmr alone : @xmath153 , at the @xmath154 confidence level ( @xcite ) .
so , in spite of their still low signal to noise ratio , the degree scale experiments already allow to better constrain the spectral index , although still at the @xmath155 level .
note that the `` standard '' , flat model with no reionization is close but not always inside the @xmath146 confidence contour .
however , this result is determined mainly by the saskatoon experiment .
in fact , we tested the stability of our results by repeating the analysis several times , excluding one experiment at the time . if we do not consider reionization , the @xmath156 c.l .
interval @xmath149 does not change more than a few @xmath5 excluding either argo , cat , max , msam , south pole or tenerife .
the lower limit drops to @xmath157 if the saskatoon data are excluded , while the upper limit raises to @xmath158 if the cobe data are excluded .
this test also shows that neglecting the correlation due to overlapping sky coverage ( e.g. tenerife and cobe , and/or msam and saskatoon ) does not change significantly the results of our analysis . \4 . for scale invariant models with no reionization ,
the height of the first `` doppler '' peak occurs at @xmath159 and is a factor of @xmath160 higher than the sachs - wolfe plateau at low @xmath77 s . for @xmath161
the peak amplitude is roughly a factor @xmath162 higher than in the scale - invariant case . a complete reionization from @xmath163 up to the present suppress the peak by roughly @xmath164 .
altogether , a model with @xmath161 and @xmath137 has a doppler peak a factor of @xmath165 higher than in the standard scale invariant case without reheating .
so our analysis confirms that a doppler peak in the @xmath82 spectrum centered at @xmath166 is perfectly consistent with the data .
all the models we have worked with have , as stated above , @xmath167 .
in flat models with vanishing cosmological constant we can expect @xmath168 from the estimated age of globular clusters ( @xcite ) .
varying @xmath169 modifies the amplitude and position of the acoustic peaks in the radiation power spectrum .
however , we have verified that , in the region of the spectrum probed by current anisotropy experiments , @xmath164 variations of the hubble constant yeld modifications in the spectrum very similar to those obtained varying @xmath135 between 0.03 and 0.07 . also , we did not consider tensor modes in our analysis , as they are expected to be of negligible amplitude for @xmath139 ( @xcite )
we will address this issue in a forthcoming paper .
abbot l.f . and wise m.b . , 1984 , , 282 , l47 .
bennet c.l . ,
1994 , , 436 , 423 .
bennet c.l .
, et al . 1996 , , 464 , 1 .
bonometto et al . , 1983 , , 123 , 118b .
cheng e.s . , et al.1994 , , 422 , l40 . de bernardis p. , et al .
1994 , , 422 , l33 .
de gasperis , g. , muciaccia , p.f . and vittorio , n. 1995 , , 439 , 1 .
dragovan m. , et al .
1993 , , 427 , l67 .
gates , dodelson and stebbins , astro - ph/9509147 , 1995 gundersen j.o . , et al .
1993 , , 413 , l1 .
, peterson b.a . , 1965 , , 142 , 1633 .
hancock s. , et al .
, 1994 , nature , 367 , 333 .
jones & wyse 1985 , 149 , 144j .
kolb & turner , 1990 , the early universe , addison - wesley , ny . lucchin f. et al .
1996 , preprint astro - ph 9504028 .
bertschinger e. , 1995 , ap.j . , 455 , 7 .
masi s. , et al .
1996 , , 463 , l47 .
melchiorri a. and vittorio n. , 1996 in preparation melchiorri a. , balbi a. , de bernardis p. , de gasperis g. and vittorio n. , 1996 in preparation netterfield et al .
, 1996 , in press , astro - ph 9601197 peebles p.j.e .
1982a , , 263l , 1p .
peebles p.j.e . 1982b , , 259 , 442p .
peebles p.j.e .
1981 , , 248 , 885 .
scaramella , r. & vittorio , n. , 1993 , , 263 , l17 .
sachs r.k . , wolfe a.m. , 1967 , , 147 , 73 .
scott p.f .
et al . , 1996 , , 461 , l1 .
smoot g. , et al .
1992 , , 396 , l1 ( dmr92 ) .
steinhardt p.j .
astro - ph 9502024 .
tanaka s.t .
1996 , , in press , astro - ph 9512067 | we study the anisotropy of the cosmic microwave background ( cmb ) in cold and mixed dark matter ( cdm and mdm ) models , with non scale - invariant primordial power spectra ( i.e. @xmath0 ) and a late , sudden reionization of the intergalactic medium at redshift @xmath1 .
we test these models against recent detections of cmb anisotropy at large and intermediate angular scales .
we find that current cmb anisotropy measurements can not discriminate between cdm and mdm models .
our likelihood analysis indicates that models with blue power spectra ( @xmath2 ) and a reionization at @xmath3 are most consistent with the anisotropy data considered here . without reionization our analysis gives @xmath4 ( 95@xmath5 c.l . ) for @xmath6 . _
p.le a. moro 2 , 00185 roma , italy .
_ _ e - mail : [email protected]_ _ via della ricerca scientifica , 00133 roma , italy _
_ e - mail : [email protected] , [email protected],_ _ p.le a. moro 2 , 00185 roma , italy .
_ _ e - mail : [email protected]_ _ via della ricerca scientifica , 00133 roma , italy _ _ e - mail : [email protected]_ |
while the existence of the galactic bar was pointed out decades ago @xcite , the bar existence received a broader support only following the work of @xcite , based on 2.4 micron light distribution in the inner galaxy as measured by @xcite , and the work of @xcite , based on the asymmetry in the distribution of galactic bulge stars as measured by iras .
the presence of the galactic bar has been observed in many different ways ; recent reviews are provided by @xcite , @xcite , @xcite , and @xcite .
the optical gravitational lensing experiment ( ogle ) is a long term observing project carried out at the las campanas observatory in chile , operated by the carnegie institution of washington .
the instrumentation of the three stages of the project was described by @xcite , @xcite , and @xcite .
a catalogue of over 200,000 variable stars discovered within approximately 11 square degrees in the direction of the galactic bulge was published by @xcite .
the public domain data set contains @xmath8 i - band photometric data points obtained in the observing seasons 1997 , 1998 , 1999 with the ogle - ii instrumentation : a dedicated 1.3-meter telescope with a @xmath9 pixel ccd camera , built and operated by the warsaw university observatory .
the presence of the galactic bar was already detected in the apparent distribution of red clump giants in the ogle - i data ( @xcite , @xcite ) . in this paper
, we set out to determine whether the evidence for the galactic bar may also be found among ogle variables .
we calculated periodograms ( @xcite , @xcite ) for each of the 200,000 objects in the ogle - ii variable star catalogue @xcite that had at least 100 i - band measurements , and accepted as real those for which the probability of a false signal @xcite was smaller than 0.001 . because fields 45 and 46 have only two seasons of observations ( fewer than 100 measurements ) , periodograms were not calculated for any stars in these fields .
we were struck by the two clear sequences apparent in the period amplitude diagram , presented in fig .
[ fig1 ] . for many stars ,
several significant periods were found .
the plotted period has the largest amplitude .
our convention is that amplitudes are defined as half of the peak - to - peak amplitude for a given mode .
originally , we had intended to use mira - type long period variables as possible tracers of the galactic bar .
however , it is clear that the variables defining the two sequences in fig .
[ fig1 ] are more numerous and , having small amplitudes , they appear to be more useful than miras as tracers of the galactic structure . therefore , this paper is devoted to these low amplitude variables only .
in addition to the two sequences present in fig .
[ fig1 ] there is a clump of variables around @xmath10 and @xmath11 ; this is mostly caused by an artefact , and should be ignored .
we divided the variables into types a and b , with the division line shown in fig .
type a is above and to the left , and type b below and to the right , of the division line .
variables within the parallelograms : @xmath12 @xmath13 were selected in 33 ogle - ii fields spanning a large range of galactic longitudes , but only a small range of galactic latitudes .
the colour magnitude diagram , corrected for interstellar reddening and extinction with the maps provided by @xcite , is shown in fig .
[ fig2 ] for the field bul_sc1 .
type a variables are marked with large blue dots ; b variables are marked with green open circles .
a relatively small range of apparent i - band magnitudes and a large number of variables indicates that these stars might be useful ` standard candles , ' and that it might be possible to use them as tracers of the galactic bar .
most variables shown in fig .
[ fig2 ] are bright red giants .
it is very likely that these stars are similar to the variables found by @xcite and @xcite in the ogle catalogue for the lmc and smc , to those found by @xcite , @xcite , and @xcite in the macho database for the lmc , to the m giant variables in ngc 6522 @xcite , and to those found by @xcite in the macho database for the galactic bulge .
minimum variability levels have been detected for red giants from the eso project
long term photometry on variables " ( @xcite ) , and also from hipparcos photometry ( @xcite ) , the cooler stars having larger minimum variability thresholds and larger mean amplitudes .
furthermore , a period - amplitude relation has been noticed for a few red giants stars by @xcite , and also in a large sample of stars from asas data ( @xcite ) .
a possible theoretical mechanism for the oscillations was studied by @xcite and @xcite .
the aim of our paper is to select from the ogle catalogue red giant variables populating the two period amplitude sequences in fig .
[ fig1 ] , to provide many parameters for these variables in electronically accessible tables , and to check if they trace the galactic bar .
we used reddening maps published by @xcite for 48 ogle - ii galactic bulge fields .
all variables of type a in a subset of these fields located at the galactic latitude @xmath14 are shown in fig .
it is clear that they form three distinct groups : luminous red giants , less luminous red giants located near the bulge red clump region , and relatively blue galactic disc stars .
these three groups are well separated .
we selected stars located above and to the right of the line in fig .
[ fig3 ] for our study , and we refer to them as ogle small amplitude red giant variables ( osargs ) .
some type b variables are close to the saturation limit of ogle - ii , and therefore we restrict our study of the galactic structure to osarg type a variables only .
the ogle saturation limit causes a problem for the study of the brightest osargs .
this technical difficulty will be overturned when the asas i - band data @xcite become available , which have a limiting i - mag of about 14 , and a considerable overlap with ogle - ii photometry .
three years of galactic bulge data have been acquired by asas already . for our analysis
we omit fields located far from @xmath14 , and those fields for which a reliable extinction map is not available .
this leaves us with a total of 33 ogle - ii fields .
the osarg type a variables form a nearly horizontal sequence in the colour magnitude diagram shown in fig . [ fig3 ] .
we fitted their distribution with a straight line using the least square method , and looking for the best global solution of the form : @xmath15 , \eqno(2)\ ] ] where @xmath16 and @xmath17 are the reddening corrected average magnitude and colour for variable number @xmath18 in the field @xmath19 , @xmath20 is a constant , different for every field , representing the ` average ' magnitude of osarg type a variables in field @xmath19 , @xmath21 is the same constant for all fields , and @xmath22 is the mean @xmath23 colour for all osarg type a variables in all 33 fields represented in fig .
we computed the value of the parameter @xmath21 as 0.169 .
the values of the parameter @xmath24 and the standard deviations from the relation ( 2 ) are given in table [ tab1 ] for all 33 fields .
this table also lists the total number of type a and type b variables in each field and the galactic coordinates of the field centers .
[ fig4 ] presents the variation of the average magnitude of osarg type a variables with the galactic longitude .
a clear trend is present , practically identical to that shown by the red clump giants ( @xcite , @xcite , @xcite ) , demonstrating that osarg variables are located in the galactic bar .
we repeated the analysis for osarg type b variables and it was qualitatively similar to that shown in fig .
[ fig4 ] , but disturbed by the ogle - ii saturation limit .
multiple periods were detected for most osargs .
the ratios of those periods to the dominant period ( i.e. , the one with the largest amplitude ) are shown in fig .
[ fig5 ] as a function of the dominant period .
several bands corresponding to the period ratios close to 2 , 1.4 , 0.9 , 0.75 , and 0.5 are clearly seen .
the interpretation of this rich spectrum of oscillations is beyond the scope of our paper .
our intent is to present this richness with a hope that it may attract specialists to investigate the nature of the oscillations .
virtually all osargs were identified with objects in the 2mass catalogue , and the values of j , h , and k magnitude are included in the tables available on internet / simbad .
a sample of these tables is shown in our table [ tab2 ] .
these magnitudes were corrected for interstellar extinction using maps developed by @xcite and using extinction ratios between the infrared bands and v - band provided by @xcite .
note that the extinction in k - band is small , and any errors in the estimate are of relatively little consequence .
the average k - band magnitude , corrected for extinction , varies with galactic longitude in a way similar to that seen in fig .
[ fig4 ] . next , extinction corrected @xmath25 magnitudes were also corrected for the galactic bar effect ( as shown in fig .
[ fig4 ] of this paper ) to the distance corresponding to galactic longitude @xmath26 ; we label the corresponding magnitudes @xmath27 .
a relation between the dominant period of oscillations and @xmath27 is shown in fig .
the two groups of osargs that are visible in fig .
[ fig1 ] are also clearly visible in fig .
[ fig6 ] .
the two groups of osargs are also plotted in the infrared color magnitude diagram in fig .
only 20% of all variables are plotted to avoid crowding .
there is almost full overlap between the two groups , distinctly different from the clear separation in fig .
[ fig1 ] and fig .
[ fig6 ] .
[ fig8 ] shows the dependence of the dominant period on the reddening corrected @xmath23 color , and clear period color relations are apparent .
note : k - band magnitude is close to the bolometric magnitude , and hence the period luminosity relations are clearly visible in fig . [ fig6 ] .
however , the bright giants form a nearly horizontal line in the @xmath28 diagram shown in fig .
hence , the @xmath23 color is for them a better absolute luminosity indicator than the @xmath29 magnitude , and this is clearly demonstrated in fig .
[ fig8 ] . finally , it is time for us to present a random sample of type a and type b osargs in fig .
these are 12 variables of each type in the field @xmath30 , sorted with increasing dominant period .
these appear as semi - regular variables because of their rich period structure .
a much longer time baseline of photometric measurements will become available soon when the current ogle - iii data are processed and published ( cf .
udalski et al .
2002a ) . with a longer time baseline
, it will be much easier to decide to what extent the multiple periods we found are stable , and to what extent osargs are multiply - periodic rather than semi - regular variables .
we provide a catalogue of 8970 osarg type a variables and 6399 osarg type b variables in electronic form at : http://www.astro.princeton.edu/@xmath31jwray/osarg/ in the files a.txt and b.txt , respectively .
separate files for each field are available at the same url , and are called a1.txt , a2.txt , , a49.txt and b1.txt , b2.txt , , b49.txt .
the catalogue is a subset of the general ogle - ii catalogue of variable stars in the galactic bulge @xcite .
a sample of the tables , listing data for the first 25 type a stars from the field @xmath32 , is shown in table [ tab2 ] , which is split into two parts because of its considerable width .
the consecutive columns are as follows : the variable star number according to @xcite ; the star number according to ogle maps ( @xcite ) ; 2000 ra and dec ; x and y pixel coordinates ; average v - band magnitude ; v - band extinction ; average ( v - i ) color ; ( v - i ) reddening ; average i - band magnitude ; i - band extinction ; j , h , k - band magnitudes ; the dominant period and its amplitude ; the next three significant periods and amplitudes . in cases where fewer than three significant periods were found ,
the extra columns contain the value -99.99999 .
similarly , for a few stars there was no match found in the 2mass catalogue , so the j , h , k columns for those stars contain the value -9.999 .
the coordinates and i and v - band magnitudes are given following ogle maps ( @xcite ) , the reddening and extinction is given following @xcite , and the j , h , k magnitudes are given following 2mass ( all - sky point source catalogue , release 2003 march 25 ) .
we provide all these data in electronically accessible files to facilitate analysis of osarg variables to all interested users .
there can be no doubt that osargs have been found in the macho database for the galactic bulge @xcite , and that they are related to red giant variables found in the lmc and smc by @xcite , by @xcite and by @xcite . however , there is also a striking difference between the galactic bulge and the lmc and smc : the multiple narrow bands seen in the @xmath33 diagrams for the lmc and smc are not present in the galactic bulge , where we see only two groups of stars .
it appears that our type a and b variables correspond to type @xmath34 and @xmath35 variables , respectively , of ita et al .
( 2003 , fig .
note : the distance modulus to the galactic bulge is 4 mag smaller than that to the lmc , so the red giant tip , which is at @xmath36 in the lmc , is expected to be at @xmath37 in the galactic bulge .
unfortunately , this is close to the i - band saturation limit of ogle , and this complicates the comparison and biases our diagrams .
we have an impression that our two sequences of variables are broader in the apparent magnitude than the corresponding sequences seen in the lmc and smc .
this may be caused by at least two effects : the galactic bar has a relatively large radial depth , which remains an effect even after correction is made for the inclination ( cf .
[ fig4 ] ) , and there is likely also a broad range of metallicities in our galaxy as compared to lmc and smc .
another striking difference between osargs and similar variables in the lmc is the presence of a large number of lmc stars with color @xmath38 , as shown in fig . 3 of @xcite ; these are carbon stars .
we found very few such variables , as is apparent in our fig .
[ fig7 ] , where there are less than a handful of stars off the right hand limit of the figure .
this difference may be due to the higher metallicity of the galactic bulge as compared with the lmc .
the period amplitude relations as presented in our fig . [ fig1 ] have not been reported in print to the best of our knowledge .
however , similar relations were found for the lmc and smc variables by @xcite .
we restrict this paper to the announcement of the osarg variables in the galactic bulge / bar , to the catalogue of 15,369 objects , and to the demonstration that they are located in a bar inclined to the line of sight .
we make no attempt at modelling the bar , as considerably more complete data should be available as soon as asas photometry ( cf .
@xcite ) becomes available , and the problems with the ogle saturation limit are alleviated .
this work was supported with the nsf grant ast-0204908 , and nasa grant nag5 - 12212 .
this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation blitz l. , spergel d. n. , 1991 , apj , 379 , 631 christensen01 christensen - dalsgaard j. , kjeldsen h. , mattei j. a. , 2001 , apj , 562 , l141 cook k. h. et al .
( macho ) , 1997 , in ferlet r. , maillard j .-
p . , raban b. , eds , 12th iap astrophysics meeting , variable stars and the astrophysical returns of microlensing surveys . editions frontires , gif - sur - yvette p. 17
dehnen w. , 2002 , in athanassoula e. , bosma a. , mujica r. , eds , asp conf .
ser . 275 , disks of galaxies : kinematics , dynamics and perturbations .
pac . , san francisco , p. 105
de vaucouleurs g. , 1964 , in kerr f. j. , rodgers a. w. , eds , proc .
20 , the galaxy and the magellanic clouds .
( canberra : australian acad . sci . ) , p. 195
dziembowski w. a. et al . , 2001 ,
mnras , 328 , 601 eyer l. , grenon m. , 1997 , esa sp-402 , 467 eyer l. , blake c. , 2002 , in aerts c. , bedding t. r. , christensen - dalsgaard j. , eds , asp conf .
259 , radial and nonradial pulsations as probes of stellar physics .
, san francisco , p. 160 garzn f. , 1999 , in beckman j. e. , mahoney t. j. , eds , asp conf . ser .
187 , the evolution of galaxies on cosmological timescales . sheridan books , san francisco , p. 31
gerhard o. , 2002 , in da costa g. s. , jerjen h. , eds , asp conf .
273 , the dynamics , structure and history of galaxies : a workshop in honour of prof .
k. freeman .
, san francisco , p. 73
glass i. s. , schultheis m. , 2003 , preprint ( astro - ph/0307366 ) ita y. et al .
, 2003 , preprint ( astro - ph/0310083 ) jorissen a. et al . , 1997 ,
a&a , 324 , 578 kiss l. l. , bedding t. , 2003 , mnras , 343 , l79 lomb n. r. , 1976 , ap&ss , 39 , 447 matsumoto t. et al . , 1982 , in riegler g. , blandford r. , eds , the galactic center .
american institute of physics , new york , p. 48
merrifield m. r. , 2003 , preprint ( astro - ph/0308302 ) nakada y. et al .
, 1991 , nature , 353 , 140 pojmaski g. , 2002 , aca , 52 , 397 press w. h. et al . , 1992 , numerical recipes in fortran .
cambridge university press schlegel d. , finkbeiner d. p. , davis m. , 1998 , apj , 500 , 525 soszyski i. ( ogle ) , 2003 , phd thesis , warsaw university stanek k. z. et al .
( ogle ) , 1994 , apj , 429 , l73 stanek k. z. et al .
( ogle ) , 1997 , apj , 477 , 163 sumi t. , 2003 , preprint ( astro - ph/0309206 ) udalski a. et al .
( ogle ) , 1992 , aca , 42 , 253 udalski a. , kubiak m. , szymaski m. ( ogle ) , 1997 , aca , 47 , 319 udalski a. et al .
( ogle ) , 2002a , aca , 52 , 1 udalski a. et al .
( ogle ) , 2002b , aca , 52 , 217 wood p. r. et al .
( macho ) , 1999 , in le bertre t. , lebre a. , waelkens c. , eds , proc .
191 , asymptotic giant branch stars .
, san francisco , p. 151
wood p. r. , 2000 , publ .
astr . soc .
australia , 18 , 18 woniak p. r. et al .
( ogle ) , 2001 , aca , 51 , 175 | among over 200,000 galactic bulge variable stars in the public domain ogle catalogue , we found over 15,000 red giant variables following two well defined period amplitude relations .
the periods are in the range @xmath0 days , and amplitudes in the range @xmath1 mag in i - band .
the variables cover a broad range of reddening corrected colours , @xmath2 , and a fairly narrow range of extinction corrected apparent magnitudes , @xmath3 .
a subset of variables ( type a ) has a rms scatter of only 0.44 mag .
the average magnitudes for these stars are well correlated with the galactic longitude , and vary from @xmath4 for @xmath5 to @xmath6 for @xmath7 , clearly indicating that they are located in the galactic bar .
most variables have several oscillation periods . [ firstpage ]
galaxy : bulge stars : variables |
metallicity is one of the few progenitor attributes that can affect the luminosity of type ia supernovae ( snia ) , with important consequences for their use as cosmological standard candles @xcite .
recently , @xcite analysed the metallicity as a source of dispersion in the light - curve width - luminosity relationship and found that deriving supernova luminosities from light curve shapes without accounting for the supernova metallicity might lead to systematic errors of up to 0.5 mag . from an observational point of view
, the dependence of snia luminosities on metallicity might originate a systematic error of @xmath5% in the measurement of the dark energy equation of state @xmath6 @xcite , comparable to the current statistical uncertainties . @xcite proposed to include the host properties ( galaxy mass and metallicity ) into the snia light - curve fitters in order to correct for these systematic errors . up to now , attempts to measure the metallicity directly from supernova observations have been scarce and their results uncertain @xcite .
measuring the metallicity , @xmath7 , from the x - ray emission of supernova remnants is a promising alternative but as yet has been only applied to a single supernova @xcite , although there are prospects to extend this analysis to further remnants @xcite .
an alternative venue is to estimate the supernova metallicity as the mean @xmath7 of its environment @xcite - this can be done for statistically significant samples of sne in nearby galaxies by taking advantage of the long evolutionary timescales of their supernova remnants @xcite .
excluding these studies based on individual objects or supernova remnants , the vast majority of our knowledge about the metallicity of type ia sn progenitors has been assembled by measuring the bulk properties of their _ entire _ host galaxies .
@xcite looked for galactic age or metal content correlations with snia luminosity in star - forming hosts , but their results were ambiguous .
@xcite studied the correlations between snia properties and galaxy metallicity or , more precisely , its oxygen abundance as determined from emission lines in the integrated spectra of star - forming hosts .
@xcite looked for systematic trends of snia uv spectra with metallicity of the host galaxy , and found that the spectral variations were much larger than predicted by theoretical models .
@xcite studied the metallicity of all the star - forming galaxies in the sloan digital sky survey ( sdss ) catalogue with redshifts in the range @xmath8 that hosted sne of some kind and noted that snia occur in a wide range of metallicities , as low as 0.25 solar .
@xcite , using data from the sdss and supernova survey concluded that prompt snia are more luminous in metal - poor systems . @xcite and
@xcite , using different methodologies to estimate the metallicity of snia hosts , arrived to opposite conclusions with respect to the dependence of supernova luminosity on @xmath7 , although these discrepancies might have been solved recently @xcite .
we note that most of the previous work has focused on comparisons between the empirical observationally - determined relationship between the _ metallicity of hosts _ and their snia luminosities on the one hand ( e.g. * ? ? ?
* ) and theoretical predictions of correlations between the _ supernova metallicity _ and luminosity on the other hand ( e.g. * ? ? ?
* ) , thus assuming implicitly that both metallicities are somehow tied . estimating the metallicity of an entire galaxy is a complex task , involving many different observational and theoretical challenges . on the observational side ,
different tracers exist for both gas - phase and stellar metallicities , each with their own advantages and disadvantages @xcite . on the theoretical side , complex modelling of the observations
is often required to arrive at a numerical estimate of the metallicity , which introduces biases and uncertainties that depend on the specific technique used ( e.g. * ? ? ?
@xcite measured the host metallicity from absorption lines of fe at two wavelengths , thus testing the gas - phase in early - type galaxies , using single - age single - metallicity stellar population models .
then , to derive metallicity , they assumed solar elemental ratios in the galactic interstellar medium ( ism ) .
@xcite determined the host metallicity indirectly . in a first step
, they determined the host mass from spectral energy distribution fits to galaxy photometry .
next , they applied the well - known mass - metallicity relationship @xcite , also valid for the gas - phase .
the @xcite mass - metallicity relationship was determined in turn through measures of the nebular emission lines of oxygen in star - forming galaxies .
@xcite compared the stellar metallicities to those of the gas - phase ( measured through the oxygen abundance ) for several thousand galaxies in the sdss .
they found that the relationship between both metallicities is non - linear , with a shallower than linear dependence of the gas - phase oxygen abundance on the stellar metallicity , and a large scatter of data points .
the question we want to address in this work is : what is the relationship between the metallicity of a snia and that of its host galaxy ? given the complexity of the problem we want to address it from a statistical point of view , aiming to determine a distribution function for the difference of metallicities between the supernova and its host .
our approach is based on calculations of galactic chemical evolution together with the evolution of the supernova rate . in order
to understand which the best estimators of snia metallicity are we study the correlations between snia metallicity and different observational tracers of host metallicity : fe vs. o in the gas - phase and stellar vs. gas - phase metallicity . in order to keep simplicity
, we use homogeneous one - zone galactic chemical evolution models , disregarding spatial variations within each galaxy . taking into account the spatial variations of metallicity would increase the scatter and corrections modelled in this paper .
the plan of the paper is as follows . in section [ models ] , we describe the theoretical framework we adopt in order to derive the statistical properties of the metallicities of snia compared to that of their host galaxies . in the next section
, we analyse the results , with particular emphasis on the reliability of different host metallicity measures as estimators of snia metallicity . among the tracers that we consider , we rank highest the measure of snia metallicity through the gas - phase metallicity of its host .
we pay special attention to the influence of the assumed delay time distribution ( dtd ) of snia .
it turns out that the metallicity of an average snia is closer to that of its host if the dtd peaks at small time delays , as the dtd of @xcite , than if the snia rate follows the rate of formation of white dwarfs , as proposed by @xcite . in section [ catalogs ]
we explore the possibility of using current galactic catalogues , such as sdss and vespa , to derive the metallicity distribution function of snia .
finally , in section [ toconclude ] we draw our conclusions .
we define the metallicity probability density of snia , @xmath9 as the probability that a snia comes from a progenitor whose metallicity logarithm differs from the logarithm of its host metallicity , @xmath0 , in : @xmath10 similarly , we define the metallicity distribution function ( mdf ) of snia as the cumulative metallicity probability , i.e. the probability that the difference between the logarithm of the metallicity of the snia progenitor and that of the host is smaller than @xmath3 ( note that this means @xmath3 larger in absolute value in the usual case that the snia metallicity is smaller than the host s one ) , @xmath11 we give additional details about the procedure we use to compute @xmath12 at the end of section [ sectinf ] .
we further define the `` metallicity correction '' , @xmath13 , as the mean value of @xmath3 , i.e. : @xmath14 the present rate of snia explosions in a given host , @xmath15 , is the convolution over time of the star formation rate ( sfr ) , @xmath16 , and the dtd , @xmath17 : @xmath18 the probability that a snia that explodes now ( at time @xmath19 ) comes from a progenitor born at time @xmath20 is : @xmath21 where @xmath22 is the time spanned since progenitor birth to the moment the host is being observed , which we will call look - back time , and @xmath19 is the cosmic time .
finally , the knowledge of @xmath23 vs. time and the metallicity evolution , @xmath24 , allows to construct the mdf of snia .
we have used two different prescriptions for the dtd : those of @xcite and ( * ? ? ? * in the following m10b ) . as we will show later in section [ sectdtd ] ,
the dtd is the most influential ingredient for the determination of the mdf of snia .
* in the following p08 ) found that the observational rate of snia is a constant fraction , @xmath25 of the stellar death rate and independent of the sfr .
later , @xcite used a different method to confirm the results of p08 , i.e. the rate of snia traces the rate of formation of white dwarfs ( wd ) , with a constant efficiency factor : @xmath26 where @xmath27 is the initial mass function ( imf ) , @xmath28 is the inverse of the derivative of the lifetime of a star of mass @xmath29 whose lifetime is precisely @xmath22 , and the integral extends down to the evolution time of the maximum mass of a star that ends as a wd , @xmath30 . in this formulation ,
the probability that a snia has an age ( i.e. , its progenitor s age ) @xmath22 is : @xmath31 on the other hand , when the dtd of m10b is applied the probability that a snia has an age @xmath22 is obtained right from eq .
[ eqpia ] : @xmath32 we adopt a salpeter imf @xcite with a small - mass cut - off of @xmath33 m@xmath34 and a high - mass cut - off of @xmath35 m@xmath34 , and the stellar lifetime law proposed by @xcite : @xmath36 with mass in @xmath37 and time in gyr . in section [ sectsens ]
we explore the sensitivity of our results to other choices for the imf @xcite and the stellar lifetime law @xcite .
one of the main ingredients necessary to determine the mdf is the prescription for the stellar yields , including supernova processing . here
, we adopt the yields provided by @xcite for stars from 13 to 40 m@xmath34 and metallicities from 0 to 0.02 .
stars in the above mass range are the main production site of the cno elements .
the @xcite nucleosynthesis calculations take into account the entire lifetime of the stars , from zero - age main sequence to the supernova or hypernova explosion , and provide the mass ejected in the explosion .
we assume that the mass lost by the star during its hydrostatic evolution returns to the ism the same metal fraction with which the star was born .
the yield tables of @xcite provide the ejected mass of each element , @xmath38 , as a function of the star mass , @xmath39 , and metallicity , @xmath40 : @xmath41 . in most of our calculations we interpolate the yields as a function of metallicity
, according to the metallicity of the ism at the time of stellar birth . as in @xcite , we adopt a fraction of hypernovas @xmath42 . in order to simplify ,
we apply the instantaneous recycling approximation to massive stars .
we integrate the yields between a minimum mass of 10 m@xmath34 and a maximum mass of @xmath35 m@xmath34 , thus obtaining a mean yield , @xmath43 . in this work ,
we use two measures of the metallicity , @xmath7 , either the cno elements , whose mean yield is @xmath44 , or all the elements with atomic number larger than five ( we will refer to this last group generically as fe in the following , and they will be our default choice for the measure of metallicity , @xmath7 , unless otherwise stated ) , whose mean yield we generically denote as @xmath45ni and empowers the light curve . on the other side ,
the cno elements present in the progenitor turn into @xmath46ne in the white dwarf , which has a relevant role as a catalyst of the thermonuclear reactions in snia and may also influence the brightness of the supernova ( for a discussion see * ? ? ? * and references therein ) . ] .
the integration is performed numerically by dividing the full mass interval in bins centred in the mass of each of the models present in the tables given by @xcite , then weighting the respective yield by the imf of mass @xmath39 and multiplying by the width of the mass interval . with respect to stars of mass below 10 m@xmath34
, we assume that the difference of mass between the zero - age main sequence star and its end - of - life compact remnant is returned to the ism with the same fractional composition of metals it had at birth .
both , the fraction of the mass of a star that is returned to the ism during its hydrostatic evolution , @xmath47 , and the total fraction of the mass of a star that is returned to the ism ( excluding snia ) , @xmath48 , have been taken from @xcite for stars of @xmath49 m@xmath34 , and from @xcite for stars of larger masses .
stars ending as snia return an additional mass to the ism equal to the chandrasekhar mass and composed 100% of metals , with negligible amounts of cno elements . to study the statistical properties of the mdf of snia we include into our chemical evolution models both a time - dependent sfr and an infall law that account for the interaction with the galactic halo .
for the sfr we adopt the schmidt law as implemented by @xcite : @xmath50 where @xmath51 is the timescale of star formation , and @xmath52 is the mass of gas in the ism at time @xmath19 . in the literature , there are other prescriptions for the sfr law .
for instance , ( * ? ? ?
* see also @xcite ) proposed to use a universal exponential sfr as a function of time : @xmath53\,,\ ] ] with @xmath54 gyr for early - type galaxies , @xmath55 gyr for spirals , and @xmath56 gyr for dwarf irregulars .
however , the use of the sandage sfr law can lead to non - positive gas mass if the constant of proportionality is not accurately chosen and/or other ingredients such as inflow / outflow are present in the model .
it is for this reason that we prefer to adhere to the schmidt law of sfr . to include the infall of material from the galactic halo we follow the conservative model of @xcite and assume that the total mass of the system formed by gas , stars and halo is constant and equal to @xmath57
initially , the whole mass is in the halo , and the mass infall rate as a function of time is : @xmath58\,,\ ] ] where @xmath59 is the timescale of infall . @xcite
propose values of @xmath60 gyr as representative of elliptical galaxies , 10.9 gyr to account for galaxies of type sbc / sc , and 60.5 gyr to represent types scd / sd . in this work ,
we use @xmath51 and @xmath59 as free parameters in order to obtain models for different types of galaxies the set of differential equations to integrate in these models is the following one ( in addition to eqs .
[ eqpia2 ] or [ eqpia3 ] ) : @xmath61 @xmath62 + r_\mathrm{ia}m_\mathrm{ch } \nonumber \\ & & + \int_{m(t)}^{m_\mathrm{up}}{\phi(m)s(t-\tau_m)e_\mathrm{h}(m)x_j(t-\tau_m)m \mathrm{d}m}\ , , \label{eqmj}\end{aligned}\ ] ] where @xmath63 is the inverse of the stellar lifetime function ( eq . [ eqtau ] ) , @xmath64 is the mass fraction of elements @xmath65 in the gas at time @xmath19 , @xmath66 is given by eq .
[ eqtau ] , @xmath16 a function of time through @xmath52 , the subscript @xmath65 in eq .
[ eqmj ] stands for either cno or fe , and @xmath67 is computed using eq . [ eqrwd ] ( for the p08 dtd ) or eq
. [ eqria ] ( for the m10b dtd ) . in this equation , we assume that the infall gas is of primordial composition and does not contribute to the mass of metals . seemingly , the term @xmath68 is only added to the equation of evolution of @xmath69 , but not to that of @xmath70 . the last term in eq .
[ eqmj ] accounts for the fraction of metals in the protostellar nebula that returns to the ism during the hydrostatic stellar evolution .
if the mass fraction @xmath71 were a monotonic function of time , the metallicity probability density might be obtained directly by combining eqs .
[ eqpia ] , [ eqmg ] , and [ eqmj ] , .
we have omitted it to improve readability . ]
@xmath72 where we have used the definition of @xmath3 ( eq . [ eqdz ] ) and the relationship between @xmath23 and @xmath67 , @xmath73 , @xmath71 is the measure of @xmath7 , and @xmath74 however , it is not guaranteed that the metallicity is a monotonic function of time , so we have to resort to a different procedure to obtain the metallicity probability density .
in this work we have computed @xmath75 through a monte carlo calculation .
first , we solve eqs .
[ eqpia ] , [ eqmg ] , and [ eqmj ] to obtain @xmath23 and @xmath71 as functions of time .
second , we generate randomly the distribution of birth times of snia according to @xmath23 , by drawing @xmath76 random numbers , @xmath77 , from a uniform distribution and assigning them birth times @xmath78 such that @xmath79 for each @xmath78 thus obtained , the corresponding @xmath3 is identified and the probability density is easily calculated .
for this last calculation we have used 100 bins in @xmath3 .
the results shown in this paper were generated with @xmath80 random numbers per each @xmath65 each time the chemical evolution of a galaxy was computed .
elliptical galaxies span large luminosity and mass ranges , their light emission is dominated by red giant stars , and contain no gas ( for a recent review see * ? ? ?
* and references therein ) .
ellipticals are metal rich , although their mean stellar metallicity is in the range @xmath81\simeq$ ] from -0.8 to + 0.3 @xcite , in general the metallicity grows with the mass of the galaxy .
the history of star formation in elliptical galaxies is still controversial , the two main scenarios being the single starburst model ( monolithic scenario ) and the hierarchical model . in the former ,
elliptical galaxies are assumed to have formed at high redshift through a short and intense starburst as a result of dissipative collapse of protogalactic gas clouds followed by a supernova driven galactic wind ( e.g. * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ) . in the hierarchical model , elliptical galaxies form at relatively recent epochs through mergers of gaseous galaxies , with continuous star formation through a wide redshift range ( e.g. * ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in every case , elliptical galaxies should be characterized by short and intense starburst(s ) during which a large number of massive stars form and explode in a short time interval , releasing a large energy into the gas leading to massive outflows that disperse the supernova yields into the intergalactic medium , especially for small mass galaxies .
the question of which model best reproduces the observational constraints is not settled yet .
since we do not pretend to address all the complexities of the formation of elliptical galaxies ( see , e.g. , * ? ? ?
* ) , we have just chosen to adopt a simple prescription intended to reproduce the chemical evolution in the monolithic scenario : the so - called `` leaky - box model '' of @xcite . in this prescription
, the effect of the supernova driven outflow is encapsulated in the use of an _ effective yield _
, @xmath82 @xcite , where @xmath83 is the usual yield of element @xmath84 , @xmath85 is a reduced yield that goes to the ism and accounts for all the material lost to outflow motions , and @xmath86 is a constant .
the leaky - box model allows us to simulate elliptical galaxies of different masses , where large values of @xmath87 represent low - mass low - metallicity early - type galaxy models .
we have computed models of early - type galaxies with infall timescale and sfr timescale in the range 0.1 - 0.5 gyr ( see * ? ? ? * for an study of the correlation between both timescales and the size and luminosity of the galaxy ) , and have explored a range of @xmath87 from 0 ( no outflow , a model for giant elliptical galaxies ) to 10 ( meant to represent dwarf ellipticals ) .
our objective is to understand the relationship between the metallicities of snia and those of their hosts .
the sensitivity of this relationship to uncertain ingredients of the theoretical model depends on the way these ingredients affect the host chemistry in comparison to the snia rate .
here we use a simplified treatment of the galactic chemical evolution that allows obtaining mdf independent of the sfr , and the dtd of p08 .
we start by considering the simultaneous evolution of metallicity and snia rates in a closed - box model .
we further simplify the model by taking a constant sfr , @xmath88 , through the complete galactic life , and assuming the instantaneous recycling approximation for stars more massive than 10 m@xmath34 . within these assumptions , the evolution of the mass of gas and of the mass of cno elements in the gas phase , @xmath70 , are given by the following differential equations : @xmath89\ , , \label{eqmgsimple}\ ] ] @xmath90\ , , \label{eqmcnosimple}\ ] ] where @xmath44 incorporates now the last term in eq .
[ eqmj ] . within the same assumptions ,
[ eqpia2 ] and [ eqpia3 ] simplify to : @xmath91 and the mdf can be obtained formally from : @xmath92 using the results of the integration of eqs .
[ eqmgsimple ] and [ eqmcnosimple ] .
the last expression shows that the mdf thus obtained is independent of @xmath88 .
the imf we use is that of @xcite .
the results we obtain are insensitive to the precise values of the cut - off within the range we have explored : @xmath93 m@xmath34 and @xmath94 m@xmath34 .
, width=9 ] besides the imf of @xcite , we have considered as well those of @xcite and @xcite . figure [ figimf ] shows the behaviour of these imf within the range of masses of snia progenitors . in this mass range ,
the normalized imf of @xcite and @xcite are nearly indistinguishable , while the imf of @xcite gives slightly less weight to the more massive snia progenitors , consequently favouring slightly longer snia delay times .
these small differences translate in small variations in the mdf of snia .
the difference in the metallicity correction , @xmath13 , between all these imf is smaller than @xmath95 dex for the whole sample of galactic models we have generated . to address the dependence of our results on the stellar lifetime law ,
we have repeated the calculations using the prescriptions of @xcite instead of @xcite . in general , the lifetimes obtained following @xcite are slightly shorter ( @xmath96% ) than those given by @xcite .
although this leads snia to follow tighter the sfr ( as we are here working with the dtd of p08 ) , the metallicity correction does not change by more than @xmath97 dex between both lifetime laws . in order to test the sensitivity of our results to the prescription of the stellar yields we have repeated the calculations with the yields @xmath38 given by @xcite fixed at two extreme metallicities : either @xmath98 or @xmath99 .
the result is that the mdf of snia is insensitive to the choice of stellar yields .
this insensitivity is a consequence of our definition of the mdf as a function of the ratio of @xmath1 to @xmath0 .
hence , even if the absolute values of both metallicites change with the stellar yield prescription , their ratio is extremely insensitive to them .
in this section , we will present the results obtained with the models including inflow and the models of passive galaxies .
unless otherwise stated , we solve eqs .
[ eqmg ] and [ eqmj ] from @xmath100 to @xmath101 gyr . .
the different curves are meant to represent typical galaxies : model d in table [ tab1 ] ( long - dashed magenta ) belongs to an early - type galaxy , while models b ( dotted black ) and c ( short - dashed blue ) are representative of late - type galaxies , and model a ( solid red ) is a transition galaxy characterized by a short infall timescale and a moderately long star formation rate timescale .
the vertical axis gives the sfr _
normalized _ by the total mass of the system , @xmath102 ( @xmath102 is also the mass of the halo at @xmath100 and should be much smaller for irregulars and dwarf ellipticals than for normal - size galaxies ) .
, width=9 ] [ cols="<,^,^,^,^,^,^,^,^ " , ] we start by analysing the behaviour of typical galaxies meant to represent from ellipticals to late - type galaxies .
table [ tab1 ] gives their main parameters together with the final mass fractions of gas and stars with respect to the total mass of the system , the final metallicity ( both as given by all metals and by cno elements ) of the host , and the statistical properties of the corresponding mdf of snia : metallicity correction and standard deviation of the distribution , @xmath103 .
we computed all the model galaxies discussed in this subsection with no outflow .
models a to c are in good agreement with the models in table 2 of @xcite for galaxies of types e , sc , and sd ( the remaining mass of gas and the mass of stars at @xmath104 of the models , as well as the final metallicities agree reasonably well given that the model of evolution of the rate of snia we use are not the same as their ) .
the sfr of model d is more bursty ( fig .
[ figsfr ] ) , and we include it here as a representative of the early - type galaxy models used in the rest of the paper ( even though it has @xmath105 , which would correspond to a massive early - type galaxy ) . in this and the next subsection , we present the results obtained with the dtd of p08 , while we delay to section [ sectdtd ] the discussion of the differences with respect to the dtd of m10b . in fig .
[ figpiap ] , we show the distribution of birth - times for a snia exploding at @xmath104 .
the curves peak at the present time because the dtd is strongly biased towards small delay times .
this effect is more pronounced in late - type galaxies , for which the star - formation activity is either slightly decreasing or increasing .
early - type galaxies display a similar peak at quite early times , belonging to the epoch of intense star - formation activity , but the importance of this initial peak varies according to the duration of the star - formation epoch .
thus , snia from early - type galaxy models come from two different populations whose properties can be quite different : an old , predominantly low - mass , population belonging to the star - formation peak , and a young population characterized by larger masses
. + figure [ figmdf ] shows the mdf of snia , for the same typical galaxies as those in the left panel of fig .
[ figpiap ] , as a function of @xmath3 . as can be deduced from fig .
[ figpiap ] , the younger the stellar population of a galaxy is ( as in models b and c ) the smaller the scatter in metallicity of potential snia progenitors . the late - type galaxy distribution is strongly biased towards a zero correction and display a short tail , down to @xmath106 dex for a galaxy that is still in its initial phases of star formation ( @xmath107 gyr , meant to represent an irregular galaxy ) , and @xmath108 dex for a galaxy that has already gone through most of the infall process ( @xmath109 gyr ) .
the probability of a snia metallicity differing in less than 0.1 dex ( in absolute value ) from @xmath0 is 75% for the irregular galaxy model and 57% for the model with @xmath109 gyr .
, for the same models as in the left panel of fig .
[ figpiap ] .
, width=9 ] early - type galaxies show a broader mdf implying that snia from these hosts will have a larger intrinsic metallicity dispersion , even if they come from the same host ! again , this behaviour can be understood from the probability density of snia progenitors shown in fig .
[ figpiap ] : models a and d have a substantial contribution of old potential snia progenitors , thus they show a large scatter in their mdf . for the model with @xmath60 gyr and @xmath110
gyr half of the snia have a metallicity differing more than 0.38 dex ( in absolute value ) from @xmath0 , while for the model with @xmath60 gyr and @xmath111 gyr the median of the correction is -0.49 dex . .
, width=9 ] figure [ figcnomdf ] shows the mdf of snia when the metallicity is measured as the abundance of cno elements .
the main difference with the mdf derived for fe is found in the behaviour of the older galaxy model , i.e. the one with @xmath60 gyr and @xmath111 gyr .
the reason of this behaviour can be understood by looking at fig .
[ figpiap ] ( bottom right panel ) , where we display the evolution with time of the mass fraction of cno elements in the gas .
the long - dashed ( magenta ) curve shows the evolution following the short burst of star formation coincident with the infall epoch .
initially the abundance of cno elements grows fast due to the self - enrichment of the gas due to efficient reprocessing driven by massive stars .
once the star formation burst is over , there are no more massive stars present , and intermediate and low mass stars enrich the interstellar medium at the end of their lifetime with yields representative of their birth time , on average less rich on cno elements than at the peak of the curve .
this causes the cno abundance to decrease rapidly until it reaches a stable value at about 60% of the peak cno mass fraction
. therefore , the snia that explode at @xmath112 gyr can have birth metallicites _ larger _ than that of the host , as seen in fig .
[ figcnomdf ] . because snia are observed at different redshifts ,
it is also interesting to know if their mdf varies with the elapsed time of galactic evolution , @xmath113 .
we have integrated again the evolutionary equations with @xmath114 and 13.7 gyr . within this range of @xmath113
the metallicity correction
changes by less than @xmath115 . in this section ,
we use a sample of galactic models with varying @xmath51 , @xmath59 , and @xmath87 to study the statistical properties of the relationship between @xmath1 and @xmath0 .
we have built models for early - type galaxies by using small values of the timescales associated to infall and sfr : either @xmath60 gyr or @xmath116 gyr , and either @xmath111 gyr or @xmath116 gyr , and with different degrees of outflow , controlled by the parameter @xmath87 . thus ,
for each value of @xmath87 we show four models , obtained with the aforementioned combinations of @xmath59 and @xmath51 .
in general , the location of these models in the plane @xmath4 vs @xmath0 is driven by @xmath87 , while the variation with @xmath59 and @xmath51 is small . in order to generate models for late - type galaxies gyr and @xmath105 .
note that this includes models a to c and a to c in table [ tab1 ] .
] we left the infall and sfr timescales vary in the ranges : @xmath117 gyr and @xmath118 gyr . in order to avoid too extreme or unrealistic galaxy models
we have further constrained our sample to have @xmath119 and @xmath120 .
figure [ figziap ] ( left panel : dtd of p08 ) shows the metallicity correction , to be applied to @xmath0 in order to estimate @xmath1 , as a function of the host metallicity .
the two kind of galaxies show different behaviour : while there is an approximately linear dependence of the correction on @xmath0 for late - type galaxies , the correction is about constant for early - type galaxies of every metallicity , i.e. the dispersion of the corrections is much smaller than the range of host metallicities .
the linear function that fits the late - type galaxies passes as well through the early - type ones with no outflow ( @xmath105 ) . as @xmath87 increases , the final metallicities of the early - type hosts decrease while the metallicity correction remains the same .
figure [ figcnoiap ] shows the metallicity correction belonging to cno abundances .
it displays the same gross properties as those seen in fig .
[ figziap ] with one important difference concerning early - type galaxies : their metallicity corrections are now among the lowest of all models displayed .
the reason for this behaviour is again the peculiar mdf of cno elements in snia from this kind of galaxies as seen in fig .
[ figcnomdf ] .
the dispersion of their mdf is , however , similar to that of galaxies with a much larger metallicity correction ( compare models a and d in table [ tab1 ] ) . in many observational studies ,
it is necessary to generate a random set of snia with stochastic metallicities in order to compare their statistical properties with those of the observational sample .
the linear fit of @xmath13 vs. @xmath121 $ ] shown in fig .
[ figziap ] together with the mdf may be used to generate such a random set of snia , with stochastic metallicities related ( but not equal ) to their host metallicities . the simpler way to do
this is to approximate the mdf of snia as a straight line in the semi - logarithmic plot shown in fig .
[ figmdf ] , @xmath122 that determines the probability density , @xmath123 the mean metallicity correction , @xmath124 and the standard deviation , @xmath103 , of the distribution of @xmath3 : @xmath125 in table [ tab1 ] we give the standard deviation of the mdf of our typical galaxy models . as can be seen , the relationship between @xmath13 and @xmath103 derived in eq .
[ eqsigma ] approximately holds for all the models , with the exception of the distribution of cno in models d and d , for which a simple proportionality law as that of eq .
[ eqfitmdf ] is not a good fit ( see fig .
[ figcnomdf ] ) . to generate a stochastic distribution of snia metallicites in a late - type galaxy of known @xmath0
, one would have first to apply the linear function of fig .
[ figziap ] ( or that in fig .
[ figcnoiap ] if what is desired is the distribution of mass fractions of the cno elements ) to obtain the metallicity correction @xmath126 and @xmath103 ( @xmath127 ) . for an early - type host
the procedure is even simpler , as the metallicity correction takes on a nearly constant value for this kind of galaxies .
the delay time distribution is a property specific of snia that has rather small imprint on the galactic chemical evolution of cno elements but affects strongly the temporal distribution of snia , hence it is expected to influence appreciably the mdf .
the dtd of snia is a matter of current debate @xcite .
p08 proposed a model in which the rate of snia was proportional to the wd formation rate .
this model reproduced satisfactorily the trend of snia rates as a function of the specific sfr in galaxies of ages between 1 and 13 gyr .
in a recent work , @xcite presented a method to recover the supernova dtd that simultaneously takes into account supernova data and the reconstructed star formation history of the individual galaxies in a survey , and applied this method to the events present in the loss sn survey and in the vespa database .
independently , @xcite analyzed light curves and host galaxy spectra of 101 snia from the sdss using vespa algorithms , from which they derived the dtd of snia .
@xcite found that their data required a stronger short delay - time component than allowed by the model of p08 , i.e. they called for a higher contribution from large mass progenitors .
the vespa database is organized in temporal bins , whose resolution is better at the present epoch and degrades with look - back time ( for further details see section [ catalogs ] ) , hence the results of @xcite should be more sensitive to the short delay time component of the snia dtd .
thus , the discrepancy between the models proposed by p08 and @xcite might be due to the different sensitivities to the time ranges of the snia dtd ( but see m10b ) .
anyway , it is necessary to check the imprint that the different dtd have on the mdf of snia .
having used a model based on p08 in the previous sections , we now discuss the results obtained with the dtd of m10b , eq . 13 in their paper , @xmath128 where @xmath129 , @xmath19 is measured in gyr , and we have applied a factor 0.7 to the dtd of m10b to convert back from their diet - salpeter imf to ours .
models a to d in table [ tab1 ] were computed using their dtd in eqs . [ eqria ] and [ eqpia ] . the right panel of fig .
[ figpiap ] shows the distribution of birth - times for a snia exploding at @xmath104 for these model galaxies . with respect to the dtd of p08 ( left panel of fig .
[ figpiap ] ) the new dtd increases the contribution from young stars ( small delay - time ) with respect to old stars .
late - type galaxies are scarcely affected at all , but early - type galaxies display a quite different trend in which the contribution of the population belonging to the initial star - formation peak is dramatically reduced . with this new dtd ,
the differences between galaxy - types strongly smooth out . the metallicity correction obtained with the dtd of m10b
is shown in the right panel of figs .
[ figziap ] and [ figcnoiap ] , for both late - type and early - type galaxies .
overall , the correction is significantly smaller than that obtained using the dtd of p08 , as expected given the larger contribution from prompt snia implied by the dtd of m10b . from these results
it is clear that the dtd of snia is the most important factor determining the magnitude of the correction needed to obtain the metallicity of snia from those of their host galaxies .
we stress that any of the above dtd models ( p08 vs. m10b ) is subject to a large degree of observational uncertainty . for instance , m10b discuss two possible approaches to determine the power law exponent in eq .
[ dtdmaoz ] , based on what they call the optimal - iron constraint and the minimal - iron constraint , which lead to different exponents , @xmath130 , in the range from -1.5 to -0.9 ( in this study , we have adopted in eq .
[ dtdmaoz ] the mean , i.e. -1.2 ) .
it is interesting to consider how much would change the plots in figs .
[ figziap ] and [ figcnoiap ] if @xmath130 covered the full range given by m10b .
we have repeated the calculations with @xmath131 and @xmath132 .
for the first case , belonging to an extreme case of the optimal - iron constraint , the straight - line fit to our late - type galaxies changes to @xmath133 + 0.06 $ ] while the correction for early - type galaxies changes by @xmath134 dex . for the last case ,
belonging to an extreme case of the minimal - iron constraint , the fit of the late - type host corrections changes to @xmath135 - 0.16 $ ] , and the corrections for early - type hosts changes by @xmath136 dex .
we can see that the uncertainty in the dtd leads to uncertainties in @xmath13 of the same order as the own correction .
we conclude that an accurate determination of the dtd of snia is a prerequisite to achieve a good understanding of the differences between the metallicities of these supernovae and those of their host galaxies . here
we discuss the reliability of different host metallicity estimators as for establishing the statistical properties of snia : gas - phase vs. stellar metallicity , @xmath137 , and gaseous cno abundance vs. fe abundance .
we have evaluated the mean host stellar metallicity as the mean of the metallicity of the ism at star birth , weighted by the sfr and by the imf , for all the stars whose mean - sequence lifetime is larger than the elapsed time since their formation .
the mean stellar metallicity derived by observational studies is weighted by the contribution of stars in different phases of their life , with weights depending on the particular spectral feature used for measuring the metallicity .
even though both ways to define a mean stellar metallicity may give different values , we expect that both will follow the same qualitative trends .
[ figziazstar ] shows the metallicity correction to apply if the host galaxy metallicity is measured from the stellar population ( and the dtd of m10b is used ) . in comparison with the trend shown in the right panel of fig .
[ figziap ] we notice that the corrections to apply to the metallicities of early - type galaxies continues being independent of host metallicity , but now they cluster around @xmath138 .
late - type galaxies display also a similar behaviour as in fig .
[ figziap ] , with changed sign , but they could not be fitted by a linear function and show larger dispersion for given @xmath0 .
thus , if possible , it is advisable to measure the metallicity of snia in late - type hosts as that of their gas .
these results can be understood by computing the distribution of birth - times of present - day ( at @xmath101 gyr ) stars .
we have calculated the distribution of birth times as the fraction of the mass in present - day stars represented by the stars born at time @xmath19 that are still alive , @xmath139 the distribution of birth - times follows closely the sfr of each model ( see fig . [ figsfr ] ) because 99% of the stellar mass at birth belongs to stars smaller than @xmath140 m@xmath141 ( for our salpeter imf ) , hence the integral in eq .
[ birth ] is always close to @xmath142 . comparing fig .
[ figsfr ] with the probability density of birth times of snia shown in fig .
[ figpiap ] ( top right ) it is easy to see that the average age of snia progenitors is much less than the average age of the stellar population , especially because the peak of @xmath23 at times close to @xmath113 . in galaxies characterized by a monotonic increase of the metallicity ( models a , b , and c ) the immediate consequence is that the metallicity of snia is larger than the average stellar metallicity .
early - type galaxies , like model d , display a non - monotonic evolution of the metallicity , that peaks later than the sfr .
for instance , for model d the peak of the sfr occurs at @xmath143 gyr , while the peak of @xmath144 is found at @xmath145 gyr and that of @xmath146 takes place at @xmath147 gyr .
thus , the stellar population is chiefly drawn from low - metallicity gas , while snia have an appreciable contribution from times close to present , at which the gas metallicity is much larger . , instead of the gas metallicity , @xmath0 .
here we have used the dtd of m10b .
the filled symbols belong to models a , b , c , and d in table [ tab1 ] ( from top to bottom ) .
the error bar at the top left of the figure is the typical uncertainty in observationally determined @xmath148 $ ] ( see discussion at the end of section [ themostrelevant ] ) .
, width=8 ] in fig .
[ figcnoiap ] we show the metallicity correction when the gas - phase metallicity is estimated from the mass fraction of cno elements in the gas . in comparison with fig .
[ figziap ] the cno metallicity correction is nearly identical to the fe correction for late - type galaxies but significantly smaller for early - type hosts , for a given dtd . as can be seen in fig .
[ figpiap ] ( bottom row ) the chemical evolutions of fe and cno elements in the galaxy models a , b , and c are in close agreement , thus the mdf of both element groups are similar ( compare fig .
[ figmdf ] with fig .
[ figcnomdf ] ) .
thus , the use of fe instead of cno abundances does not seem to represent an improvement in the estimates of snia metallicities obtained from those of their hosts .
the host cno abundance of early - type galaxies is in much better agreement with the _ mean _ snia one than are the corresponding total metal abundances , although their dispersions are similar . using again the chemical evolution computed for the prototypical galaxy model d , displayed in the bottom row of fig .
[ figpiap ] , and the probability density of snia shown in fig .
[ figpiap ] , the reason for this closer agreement between cno abundances in snia and the gas , as compared with fe abundances , is that snia from early - type galaxies have non - negligible contributions from quite early times and near present time .
the cno abundances are low at both times , because presently there are no high - mass stars contributing to their synthesis . on the other hand ,
the fe abundance increases continuously during the last 11 gyr due to the contribution of snia , with the result that young snia progenitors are drawn from gas with high fe abundance while old snia progenitors were born from gas with quite small fe abundance .
hence , the average discrepancy between present - day gas fe abundance and that of snia is larger than in the case of the cno abundances . a meaningful evaluation of the corrections we propose to obtain snia metallicities from the different metallicity measures of their hosts can only be done by considering the observational errors in the galactic metallicity and in the dtd of snia .
as discussed in the introduction , galactic metallicities are measured using different methods , all of them subject to a suite of experimental and theoretical uncertainties .
@xcite measured the fe abundance of the gas of early - type hosts of snia with quite different errors , depending on the host .
an average error in @xmath149 $ ] of 0.1 - 0.2 dex can be deduced from their table 4 and from the comparison they perform with metallicities measured by @xcite and @xcite for a subset of galaxies they have in common . on the other hand , the @xcite mass - metallicity relationship , used by @xcite to measure host metallicity , is sensitive to @xmath150 $ ] in the gas phase . the scatter in this relationship is of order @xmath151 dex . finally , the observational determination of @xmath148 $ ] is subject to systematic biases of order @xmath152 dex @xcite .
summarizing , the uncertainty in the measured metallicity of snia hosts is similar for the different metallicity tracers , @xmath153 dex , although it may be substantially larger for a given galaxy .
we have drawn in figs .
[ figziap ] , [ figcnoiap ] , and [ figziazstar ] an error bar indicative of the typical uncertainty in the measured host metallicity .
these uncertainties are not large compared to the range of host metallicities covered in the figures ( horizontal axes ) .
one can also compare the host metallicity uncertainty with the correction necessary to obtain @xmath1 , plotted in the same figures ( vertical axes ) .
looking at the corrections deduced for the dtd of p08 ( left panels of the same figures ) it can be seen that they are in general much larger than @xmath154 . on the contrary , if the dtd of m10b is used ( right panels of figs .
[ figziap ] and [ figcnoiap ] ) , @xmath155 is of the same order as @xmath154 , especially for early - type galaxies .
thus , if this last dtd were the correct choice , the error in the determination of snia metallicity would be dominated by the uncertainty in the measurement of the host metallicity rather than the difference between galactic and supernova metallicities .
in this section , we explore the facilities provided by current galactic catalogues in order to obtain observational constraints to our theoretically derived mdf of snia .
ideally , one might use an observationally determined star formation history ( sfh ) of a galaxy together with a gas metallicity history , determined through metallicity - dependent synthetic stellar population ( ssp ) models , to recover a probability distribution function for the metallicity of snia progenitors . in practice , however , there are limitations due to both the discretisation of the catalogue in time bins of finite size , and the paucity of metallicities used in the ssp models . in section [ probcat ]
we shortly address how to reformulate eqs .
[ eqpia2 ] and [ eqpia3 ] in order to calculate the probability that a snia progenitor was born in a given time bin .
we apply then the new formalism to the vespa catalogue , and discuss the results and the consequences of the paucity of metallicities in the ssp models in section [ rescat ] .
we explain the main characteristics of the vespa catalogue in the next paragraph .
vespa ( versatile spectral analysis , * ? ? ?
* ) is a method to reconstruct star formation and metallicity histories from galactic spectra using ssp models .
@xcite applied their vespa code to all galaxy spectra in the seventh data release of the sdss @xcite , and compiled their results in a publicly accessible catalogue of stellar masses , star formation rates and metallicity histories of nearly 800,000 galaxies .
vespa uses all of the available absorption features and the shape of the continuum ( emission lines are not included in the analysis ) in order to estimate the sfh of each galaxy , with variable time resolution , and the metallicity of the stars born in each time bin , which traces the gas metallicity .
the number of time bins depends on the quality of the data on each galaxy . at the highest resolution
, vespa uses 16 age bins , logarithmically spaced between look - back times 0.002 gyr and 13.7 gyr , however if the data are not of sufficient quality some bins may lack information .
the vespa catalogue contains results from several runs using different ssp models , imf prescriptions , dust models , and galactic samples ( for details see the above - mentioned papers from tojeiro et al ) .
the ssp models are supplied at metallicities of @xmath156 , 0.01 , 0.02 , and 0.04 .
the present rate of snia explosions in a given host is the convolution over time of the sfr and the dtd .
the probability that a snia that explodes now ( at time @xmath19 ) comes from a progenitor born at time @xmath20 is given by eq .
[ eqpia ] .
assuming we know the dtd and that the sfr can be obtained from a galactic catalogue , both in discrete time bins , the probability that the snia was born at time bin @xmath84 is ( for a given galaxy ) : @xmath157 where the addition in the denominator extends to all bins defined for that galaxy , and @xmath158 is the mass of stars formed in time bin @xmath84 .
if the observational dtd is not known with enough resolution ( as compared to the time bin resolution in a given catalogue ) , one can turn to the model of p08 ( eq . [ eqpia2 ] ) or its equivalent based on the dtd of m10b ( eq . [ eqpia3 ] ) .
what we actually need in order to use the galactic catalogue is the probability that a snia exploding now comes from a progenitor born within a time interval given by the temporal bins defined in the catalogue : say between look - back times @xmath159 and @xmath160 : @xmath161 where we have used the dtd of p08 . to use
the catalogued data we transform the above integral using a mean sfr defined from the stellar mass formed in the time interval @xmath84 and the duration of that time interval , @xmath162 : @xmath163 , and combine @xmath164 : @xmath165 the integral in the last equation can be easily computed given the imf , the stellar lifetime law , and the temporal limits of each bin .
note that there is no need to know the proportionality constant for @xmath166 nor even to re - normalize the imf , because we know that @xmath167 , if we add @xmath166 for all the bins in a galaxy .
thus , the probability density can be normalized unambiguously .
if the dtd of m10b is adopted instead , @xmath166 can be obtained from : @xmath168 where @xmath169 is the average dtd between @xmath159 and @xmath160 . from the previously determined probability
that a snia comes from time bin @xmath84 , @xmath166 , and the catalogued metallicity for that same time bin , @xmath40 , the metallicity probability density is given by : @xmath170 this procedure provides us with a discrete set of @xmath3 for which the probability density is known .
here we show an example of the application of the formalism introduced in the last section to real data .
we have accessed the vespa catalogue ( runid=2 ) and classified the galaxies as early or late - type according to the criterion proposed by @xcite on base of their @xmath171 and @xmath172 sdss model magnitudes .
then , we have selected the galaxies with a minimum temporal resolution of three bins , in order to work with reasonable sfhs .
finally , we recovered 90906 early - type galaxies and 40419 late - types . we stress that we have not filtered the catalogue in order to select which galaxies are potential snia hosts . from a theoretical point of view , every single galaxy can house binaries with the appropriate parameters ( total mass , secondary mass , initial separation , metallicity ) to produce snia .
observationally it is another story , because surveys aiming to detect snia have to pre - select target galaxies for which the observations and measurements are more efficient .
this difference has to been kept in mind when comparing between our theoretically derived snia statistics and the correlations found in observational campaigns .
we have applied the formalism developed in the previous section ( eqs .
[ eqpi ] and [ eqppi ] ) to all the galaxies selected from vespa in order to obtain the mdf of each galaxy .
unfortunately , at the present level of resolution of the catalogue the resulting mdf is of little practical use because of the limited spectra of metallicities of the ssp models .
as the mdf is an statistical description of the difference of metallicity between the host and the snia , the use of just four metallicities in the ssp models produces an unreliable extremely clumpy mdf . keeping this in mind , we comment here briefly on the relationship between the metallicity correction , @xmath13 , and the @xmath0 obtained from the catalogue .
linear fits to these quantities are shown as dashed lines in fig .
[ figziap ] . given the uncertainties , it is striking the close coincidence between the points belonging to our theoretical models and the linear fits obtained with the catalogue , especially for late - type galaxies .
the differences between theoretical models of early - type and late - type galaxies are more or less reproduced by the mean metallicities derived from vespa . in view of the strong uncertainty affecting the metallicity statistics ,
we do not follow further with the analysis of the vespa derived mdf , but we emphasize that the formalism presented in the previous subsection can ( and should ) be applied to future galactic catalogues based on improved data and metallicity resolution . in spite of the failure to derive useful mdf from vespa , the data from the catalogue can still be used to show trends of the mean snia metallicity , @xmath4 calculated using the probability defined by eq .
[ eqpi ] , vs. the galactic sfr and age .
figure [ figssfr ] shows @xmath4 as a function of the ssfr ( specific star formation rate ) .
we first notice the sharp cut on top of the points distribution at @xmath173 , which derives from the maximum metallicity fed to the ssp models of vespa .
if we disregard this limitation of the models , we can still fit a tendency line to the data points , with the result that can be seen in the plot : snia exploding in high ssfr galaxies do have smaller @xmath4 than those that explode in passive galaxies . on the other hand , the tendency lines for early and
late - type galaxies do not differ significantly . while the correlation is not strong
, we note that a relationship between ssfr and snia luminosity has been found in some observational studies of snia ( e.g. , to cite just one , * ? ? ?
* who found that snia in low ssfr galaxies appear brighter on average than those in high ssfr galaxies , after applying stretch corrections ) . as a note of caution
, we remark that our galaxy sample was drawn from the whole unfiltered vespa catalogue whereas each observational study of snia properties are based on particular selection function of hosts .
thus , a direct comparison with our results might not be meaningful . as a function of the specific sfr , as derived from the vespa catalogue ( runid=2 ) and the dtd of m10b .
the solid lines are the results of linear fits to the points given by @xmath174 = -0.177\log\left(\mathrm{ssfr}\right ) - 2.10 $ ] ( blue , early - type galaxies ) and @xmath174 = -0.196\log\left(\mathrm{ssfr}\right ) - 2.26 $ ] ( red , late - type galaxies ) .
, width=9 ] figure [ figtback ] shows the distribution of @xmath4 as a function of the look - back time of the host galaxy ( that is : @xmath175 , where @xmath176 is the distance to the galaxy and @xmath177 is the speed of light ) , that is itself a measure of the age of the universe at the local host time , and of the redshift . besides the accumulation of points around @xmath178 ,
the data show a tendency for smaller @xmath4 at lower redshift . the trend of @xmath0 with look - back time ( not shown in the fig . )
is opposite to that of @xmath4 , i.e. older hosts are less metal - rich as expected from popular galaxy evolution models .
the sign of the slope of @xmath4 vs. @xmath179 is difficult to explain , as it implies that snia are on average _ less _ luminous at high redshifts , contrary to what is derived from observational samples ( e.g. * ? ? ?
* their fig .
anyway , we can not derive any strong conclusion given the limitations of the present analysis , as the trend we detect might be a result of observational biases of the catalogued data . as a function of the look - back time of the host galaxy , @xmath179 , derived from the vespa catalogue ( runid=2 ) and the dtd of m10b .
the solid lines are the results of linear fits to the points given by @xmath174 = 0.147 t_\mathrm{back } - 0.524 $ ] ( blue , early - type galaxies ) and @xmath174 = 0.064 t_\mathrm{back } - 0.251 $ ] ( red , late - type galaxies ) .
, width=9 ] we have checked that the above trends are qualitatively independent of the assumptions concerning the dtd and the specific runs of vespa that can be selected ( galactic samples , ssp models , imf prescriptions , and dust models ) .
we have also tested the scatter of the parameters of the tendency lines in figs .
[ figssfr ] and [ figtback ] as a result of the uncertainties in the galactic properties stored in the catalogue ( errors in the metallicity and star formation rates in each galactic bin ) . to this end , we have computed again the tendency lines for 200 random realizations of the galactic sets ( early - types and late - types ) by adding random noise according to the catalogued metallicity and sfr errors in every time bin of each galaxy of our vespa sample . for each random realization , we have computed the tendency line and thereafter we have analyzed the distribution of slopes , their mean and standard deviation .
the standard deviation of the slope of the tendency lines of ssfr and @xmath179 with respect to @xmath4 is typically less than 0.002 .
we have calculated the chemical evolution of homogeneous galaxy models together with the evolution of the supernova rates in order to evaluate the metallicity distribution function of snia , @xmath2 , i.e. the probability that the logarithm of the metallicity of a snia exploding now differs in less than @xmath3 from that of its host .
we have analysed several model galaxies aimed to represent from active to passive galaxies , including dwarf galaxies prone to experience supernova driven outflows .
our results show a remarkable degree of coincidence ( in an statistical sense ) between the mean @xmath1 and @xmath0 .
the dispersion of @xmath1 in active galaxy models is quite small , meaning that @xmath0 is a quite good estimator of the supernova metallicity , while passive galaxies present a larger dispersion .
we have devised a procedure to correct the difference in metallicity between snia and their hosts ( sect .
[ sectstat ] ) , based on a linear fit given in fig . [ figziap ] .
these results are insensitive to the choice of imf , stellar lifetime and stellar yields , within the ranges we have explored .
in contrast , the delay time distribution of snia remains one of the main ingredients influencing the difference of metallicity between snia and their hosts .
we have discussed the use of different metallicity indicators ( fe vs. o , gas - phase metallicity vs. stellar metallicity ) . using o ( assuming its evolution is representative of that of the cno group ) as a metallicity measure for late - type galaxies does not change appreciable the metallicity correction with respect to using fe ( assuming its evolution is representative of that of all metals ) .
it is remarkable that the metallicity correction for early - type galaxies ( @xmath180 gyr ) is quite different for o than for fe . on the other hand ,
the metallicities of snia are much better represented by those of the gas - phase than by the mean stellar metallicities .
thus , when possible , it is advisable to measure the host metallicity through that of its gas - phase . finally , the results of the application of our formalism to a galactic catalogue ( vespa ) suggest that snia come , in average , from smaller metallicity progenitors both at low redshifts and in galaxies with large star - formation activity
. the paucity of metallicities used in the original stellar - population synthesis models used in the construction of the galactic catalogue does not allow us to build a significant mdf of snia based on the galactic histories stored in the database .
however , in spite of the large uncertainty in the metallicity derived from the catalogue , the gross trends of @xmath4 vs. @xmath0 obtained from vespa for different galaxy types are roughly consistent with our theoretical estimates .
the derivation of improved , observationally based , mdf will be possible in the future if the ssp models include further refinements in their grid of metallicities .
one of the main drawbacks of our approach is the use of homogeneous one - zone galactic chemical evolution models .
actual galaxies are heterogeneous , and radial gradients of chemical composition are routinely measured within them .
a possible improvement of our model would be to divide the galaxy in several non - interacting shells as in , e.g. , @xcite , then computing the galactic mdf accounting for the probability that a snia originates in each shell . then
, however , a complication would arise when comparing to observations of far away galaxies for which there is no possibility of measuring the metallicity gradient : what is the meaning of _ the _ galaxy metallicity in a context where each independent shell has its own chemical history ?
what is clear is that the derived mdf of snia would be characterized by a larger dispersion than the mdf we have obtained . in some sense ,
the true mdf of snia reflects a double dependence : in space , i.e. the position of the snia progenitor within its host galaxy , and in time , i.e. the moment in the galactic history when the progenitor was born .
it is just this last dependence what we have studied in the present work .
we thank the referee for his / her comments that have helped to improve substantially the present study . this work has been partially supported by a mec grant , by the european union feder funds , and by the generalitat de catalunya .
cb thanks benoziyo center for astrophysics for support .
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 . , l. & cappellaro , e. 2009 , in american institute of physics conference series , vol . 1111 , american institute of physics conference series , ed . g. giobbi , a. tornambe , g. raimondo , m. limongi , l. a. antonelli , n. menci , & e. brocato , 477484 | the efficient use of type ia supernovae ( snia ) for cosmological studies requires knowledge of any parameter that can affect their luminosity in either systematic or statistical ways .
observational samples of snia commonly use the metallicity of the host galaxy , @xmath0 , as an estimator of the supernova progenitor metallicity , @xmath1 , that is one of the primary factors affecting snia magnitude . here , we present a theoretical study of the relationship between @xmath1 and @xmath0 .
we follow the chemical evolution of homogeneous galaxy models together with the evolution of the supernova rates in order to evaluate the metallicity distribution function , @xmath2 , i.e. the probability that the logarithm of the metallicity of a snia exploding now differs in less than @xmath3 from that of its host .
we analyse several model galaxies aimed to represent from active to passive galaxies , including dwarf galaxies prone to experience supernova driven outflows .
we analyse as well the sensitivity of the mdf to the most uncertain ingredients of our approach : imf , star - formation law , stellar lifetime , stellar yields , and snia delay - time distribution ( dtd ) .
our results show a remarkable degree of agreement between the mean @xmath4 in a galaxy and its @xmath0 when they both are measured as the cno abundance , especially if the dtd peaks at small time delays , while the average fe abundance of host and snia may differ up to 0.4 - 0.6 dex in passive galaxies .
the dispersion of @xmath1 in active galaxy models is quite small , meaning that @xmath0 is a quite good estimator of the supernova metallicity .
passive galaxies present a larger dispersion , which is more pronounced in low mass galaxies .
we present a procedure to generate random snia metallicities , given the host metallicity .
we also discuss the use of different metallicity indicators : fe vs. o , and gas - phase metallicity vs. stellar metallicity .
finally , the results of the application of our formalism to a galactic catalogue ( vespa ) suggest that snia come , in average , from small metallicity progenitors both at low redshifts ( contrary to expectations ) and in galaxies with high star - formation activity . in spite of large uncertainties in the metallicities derived from the catalogue , the gross trends of @xmath4 vs. @xmath0 obtained from vespa for different galaxy types are roughly consistent with our theoretical estimates .
catalogues distance scale galaxies : abundances supernovae : general . |
over the last three decades , the low - dimensional quantum spin models with competing ( frustrated ) interactions have attracted considerable research interest especially due to their extraordinary diverse ground - state behaviour .
_ geometrically frustrated spin systems _ constitute a special sub - class of the frustrated models that can be distinguished by incapability of spins , inherent in their lattice positions , to simultaneously minimize the ground - state energy of each individual spin - spin interaction @xcite . as a rule , the quantum spin systems affected by a rather strong geometric frustration often exhibit an exotic non - magnetic ground state ( which does not have its classical analogue ) in addition to a rich variety of the semi - classically ordered ones @xcite .
it is worthy to notice , moreover , that a subtle interplay between the geometric frustration and quantum fluctuations gives rise to a number of intriguing phenomena including the macroscopic degeneracy of ground state @xcite , order by disorder effect @xcite , chirality @xcite , quantum phase transitions @xcite-@xcite , quantized plateaux in the magnetization curves @xcite-@xcite , double - peak specific heat curves @xcite-@xcite , enhanced magnetocaloric effect @xcite-@xcite , etc . despite a significant amount of effort
, there are only few frustrated spin-@xmath0 quantum heisenberg models , such as majumdar - ghosh model on double chain @xcite-@xcite , sawtooth ( @xmath2 ) chain @xcite-@xcite , or shastry - sutherland model @xcite , for which precise analytic solution is available leastwise for the ground state .
nevertheless , it should be pointed out that frustrated quantum systems are in general rather difficult to deal with , since extensive numerical methods must be used in order to obtain a reliable estimate of their magnetic properties . from this point of view , the one - dimensional ( 1d ) frustrated spin systems are the simplest systems with respect to accurate treatment . of these systems , the spin-@xmath0 quantum heisenberg model with
_ diamond chain _
topology is currently actively engaged in the investigation of geometric frustration .
interestingly , this simple quantum system turned out to have rather complex ground state ; apart from the usual ferrimagnetic phase there are in fact several quantum dimerized and plaquette states involved in the zero - field ground - state phase diagram @xcite-@xcite .
further studies devoted to the spin-@xmath0 quantum heisenberg model on the diamond chain have provided fairly accurate results for the ground - state phase diagram in a presence of the external field @xcite-@xcite , the spin gap @xcite , the magnetization and susceptibility @xcite .
another remarkable finding relates to the observation of an inversion phenomenon , which can be induced in the frustrated diamond chain through the exchange anisotropy @xcite .
note that the ground state and thermodynamics of the mixed - spin diamond chains containing also higher - spin sites have already been particularly examined as well @xcite-@xcite .
it is worthwhile to remark that 1d frustrated spin systems have initially been introduced purely as toy models suitable for investigating the effect of spin frustration .
however , a recent progress achieved in the design and controlled synthesis of molecular - based magnetic materials afforded another stimulus for testing 1d frustrated spin systems by overcoming the lack of desirable model compounds .
as a matter of fact , the rapidly expanding field of molecular engineering have led to fabrication of several coordination polymers , which can be regarded as genuive examples of the frustrated spin models @xcite . with the help of structural data known long ago @xcite , kikuchi and
co - workers @xcite-@xcite have recognized cu@xmath3(co@xmath3)@xmath4(oh)@xmath4 ( azurite ) as an appropriate candidate for the diamond chain compound .
experimental data measured for the high - field magnetization , susceptibility , specific heat , nmr @xcite-@xcite and esr data @xcite-@xcite have indeed confirmed kikuchi s conjecture and it is now quite well established that the azurite represents the actual material for the frustrated diamond chain .
it should be also mentioned that sufficiently strong frustration found in the azurite clearly manifests itself in its quantum features ; the high - field magnetization shows quantum plateau at one third of the saturation magnetization , the susceptibility turned out to have two round peaks at relatively low temperatures and esr data have proven the gap - less excitation spectrum .
theoretical interest focused on the diamond chain structure also enhances its another experimental realizations provided by the polymeric compounds such as cu@xmath4oso@xmath5 @xcite and m@xmath3(oh)@xmath4 ( m = ni , co , mn ) @xcite-@xcite . with this background
, we shall investigate in the present article a simplified version of the frustrated ising - heisenberg diamond chain , which can exactly be solved by applying an accurate map based on the generalized decoration - iteration transformation @xcite .
it is noteworthy that this relatively simple and straightforward analytical approach has recently been adapted to study an appearance of quantized plateaux in the magnetization process of the trimerized @xcite and tetramerized @xcite ising - heisenberg linear chains .
as we shall show hereafter , the same strategy can also be used to explore the frustrated ising - heisenberg diamond chain , which represents another particular realization of models tractable within the generalized algebraic maps .
the main goal of present work is to exploit a grand advance of the method used for obtaining exact results for all possible ground states as well as all basic thermodynamic quantities .
the rest of this paper is organized as follows .
the model hamiltonian and major features of the mapping method are presented in section 2 .
section 3 is divided into two sub - sections . in the former one , we provide rigorous results for the spin-@xmath0 ising - heisenberg diamond chain , while the latter one comprises magnetic data of the mixed spin-@xmath0 and spin-1 ising - heisenberg diamond chain .
altogether , exact results for ground - state phase diagrams , magnetization , entropy , susceptibility and specific heat are derived and particularly discussed for both investigated diamond chains .
a cooling rate of adiabatic demagnetization is also explored in connection with the enhanced magnetocaloric effect .
finally , some concluding remarks are drawn in section 4 .
let us begin by considering two kinds of spins regularly distributed on the 1d lattice composed of the diamond - shaped units as diagrammatically depicted in fig . 1 .
to ensure an exact tractability of this spin system , we shall further suppose that each diamond - shaped plaquette consists of two dumbbell _ heisenberg _ spins ( @xmath1 ) , which are placed in between two _ ising _ spins ( @xmath6 ) residing corner - sharing positions on the diamond motifs .
for further convenience , we shall write the total hamiltonian as a sum over plaquette hamiltonians , i.e. @xmath7 , where each plaquette hamiltonian @xmath8 involves all interaction terms associated with one diamond - shaped unit ( see fig
. 1 ) @xmath9 + j_{\rm i } ( \hat{s}_{3k-1}^{z } + \hat{s}_{3k}^{z})(\hat{\mu}_{3k-2}^{z } + \hat{\mu}_{3k+1}^{z } ) \nonumber \\ \fl & - h_{\rm h } ( \hat{s}_{3k-1}^{z } + \hat{s}_{3k}^{z } ) - h_{\rm i } ( \hat{\mu}_{3k-2}^{z } + \hat{\mu}_{3k+1}^{z})/2.\end{aligned}\ ] ] above , @xmath10 and @xmath11 ( @xmath12 ) denote spatial components of the spin-@xmath0 and spin-@xmath1 operators , the parameter @xmath13 stands for the anisotropic xxz interaction between the nearest - neighbouring heisenberg ( dumbbell ) spins and the parameter @xmath14 accounts for the ising - type coupling between the heisenberg spins and their nearest ising neighbours .
the parameter @xmath2 allows to control the heisenberg exchange interaction @xmath13 between the easy - axis ( @xmath15 ) and easy - plane ( @xmath16 ) regime and finally , the last two terms incorporate coupling of the ising and heisenberg spins to an external longitudinal magnetic field @xmath17 and @xmath18 , respectively .
the crucial point of our calculation represents calculation of the partition function . by making use of the commutation rule between different plaquette hamiltonians , i.e. by exploiting @xmath19 = 0 $ ] valid for each @xmath20 ,
the partition function of ising - heisenberg diamond chain can be partially factorized into the following product @xmath21 where @xmath22 , @xmath23 is being boltzmann s constant , @xmath24 the absolute temperature , the symbol @xmath25 means a summation over all available configurations of ising spins @xmath26 and @xmath27 stands for a trace over spin degrees of freedom of two dumbbell heisenberg spins belonging to @xmath28th diamond unit . after performing this partial trace
, the structure of relation ( [ eq : z ] ) immediately implies a possibility of applying the generalized decoration - iteration mapping transformation @xcite @xmath29 g \bigl [ \beta j_{\rm i } ( \mu_{3k-2}^{z } + \mu_{3k+1}^{z})/2 \bigr ] \nonumber \\ & = & a \exp \bigl[\beta r \mu_{3k-2}^{z } \mu_{3k+1}^{z } + \beta h_{0}(\mu_{3k-2}^{z } + \mu_{3k+1}^{z})/2 \bigr],\end{aligned}\ ] ] which converts the ising - heisenberg diamond chain into the uniform spin-@xmath0 ising linear chain with an effective nearest - neighbour coupling @xmath30 and an effective external field @xmath31 .
the expression @xmath32 in the first line of eq .
( [ eq : dit ] ) depends on the spin of heisenberg atoms and its explicit form is given in appendix for two particular spin cases @xmath33 and @xmath34 .
it is noteworthy that a general validity of the mapping relation ( [ eq : dit ] ) necessitates a self - consistency condition to be satisfied , what means , that it must hold independently of spin states of both ising spins @xmath35 and @xmath36 .
it can readily be proved that a substitution of four possible configurations of the ising spins @xmath35 and @xmath36 into the formula ( [ eq : dit ] ) indeed gives just three independent equations , which unambiguously determine the unknown mapping parameters @xmath37 , @xmath30 and @xmath31 @xmath38 here , we have defined the functions @xmath39 , @xmath40 and @xmath41 in order to write the transformation parameters @xmath37 , @xmath30 and @xmath31 in more abbreviated and general form . now , a direct substitution of the transformation ( [ eq : dit ] ) into the expression ( [ eq : z ] ) yields the equality @xmath42 which establishes an exact mapping relationship between the partition function @xmath43 of the ising - heisenberg diamond chain and , respectively , the partition function @xmath44 of the uniform spin-@xmath0 ising linear chain defined by means of the nearest - neighbour coupling @xmath30 and the effective field @xmath31 .
notice that the exact solution for the partition function of the latter system is well - known ( see for instance ref .
@xcite ) and hence , the relation ( [ eq : zdz0 ] ) can readily be utilized for calculation of some important quantities ( magnetization , entropy , specific heat , susceptibility ) by making use of the standard thermodynamical - statistical relations .
it should be mentioned , however , that there also exists another alternative approach that is of particular importance if some relevant physical quantity can not be obtained within this procedure .
actually , the problem connected with the calculation of correlation functions and/or quadrupolar moment can be simply avoided by employing the following exact spin identities @xmath45 in above , the function @xmath46 depends exclusively on the ising spin variables @xmath26 , while the function @xmath47 may depend on any spin variable belonging to @xmath28th diamond plaquette .
the superscripts @xmath48 label spatial components of the appropriate spin operators and finally , the symbols @xmath49 and @xmath50 stand for the standard canonical average performed over the ensemble defined on the ising - heisenberg diamond chain and its corresponding ising chain , respectively . by applying the exact spin identities ( [ eq : f1 ] ) and
( [ eq : f2 ] ) , one easily attains rigorous results for the sub - lattice magnetization @xmath51 and @xmath52 reduced per one ising and heisenberg spin , respectively , the pair correlation functions @xmath53 , @xmath54 , @xmath55 , as well as , the quadrupolar moment @xmath56 @xmath57 as one can see , all aforelisted quantities can be expressed in terms of the single - site magnetization @xmath58 and the nearest - neighbour correlation @xmath59 of the spin-@xmath0 ising linear chain given by @xmath30 and @xmath31 . since exact analytical formulae for those quantities
have been retrieved in the literature many times before @xcite , we shall restrict ourselves for brevity to listing the coefficients @xmath60 , @xmath61 and @xmath62 ( @xmath63 = 13 ) emerging in the set of eqs .
( [ eq : miz])-([eq : qhzz ] ) @xmath64 an explicit representation of the functions @xmath65 is too cumbersome to write it down here and it is therefore left for appendix . finally , we just simply quote the well - known thermodynamical - statistical relations , which have been utilized for calculating gibbs free energy @xmath66 , entropy @xmath1 , specific heat @xmath67 and susceptibility @xmath68 .
accurate results for these quantities have been obtained with the help of the precise mapping relation ( [ eq : zdz0 ] ) and using @xmath69 where @xmath70 is referred to as the gibbs free energy of corresponding spin-@xmath0 ising chain @xcite .
before proceeding to a discussion of the most interesting results it is worthy to mention that the results derived in the preceding section hold regardless whether ferromagnetic or antiferromagnetic interactions are assumed . as we are mainly interested in examination of the spin frustration effect , in what follows we restrict both exchange parameters @xmath13 and @xmath14 to positive values in order to match the situation in the frustrated antiferromagnetic diamond chain .
furthermore , it is convenient to reduce the number of free parameters by assuming equal @xmath71-factors for the ising and heisenberg spins , i.e. by imposing the same parameter representing the effect of external field @xmath72 . to simplify further discussion
, we shall also introduce a set of reduced parameters @xmath73 , @xmath74 and @xmath75 as describing dimensionless temperature , external field and the strength of frustration , respectively .
first , let us take a closer look at the ground state of the spin-@xmath0 ising - heisenberg diamond chain . for illustration ,
two ground - state phase diagrams are displayed in fig .
2 , the one in the @xmath76 space for the system without external magnetic field ( fig . 2a ) and the other one in the @xmath77 space for the system placed in the non - zero external field under the assumption @xmath78 ( fig .
both the figures clearly demonstrate that a competition between the interaction parameters @xmath79 , @xmath2 and @xmath80 gives rise to three possible ground states .
apart from the usual ferrimagnetic phase ( fri ) and the frustrated phase ( fru ) found both in a presence as well as absence of the external field , the system ends up in the saturated paramagnetic phase ( spp ) once the external field is being above its saturation value .
spin order within fri , fru and spp can be distinguished from one another according to their attributes ( physical quantities included in the set of eqs .
( [ eq : miz])-([eq : qhzz ] ) ) , as well as , through their wave functions @xmath81
@xmath82 @xmath83 the first product in the aforelisted eigenfunctions is carried out over all ising spins , the second one runs over all pairs of the heisenberg dumbbell spins and @xmath84 denotes standard ket vector assigned to @xmath85th projection of the ising ( @xmath86 ) and heisenberg ( @xmath87 ) spins .
it is quite evident from the set of eqs .
( [ fria ] ) that fri displays the classical ferrimagnetic spin arrangement usually observed in the pure ising systems , actually , all the results clearly indicate antiparallel alignment between the nearest - neighbouring ising and heisenberg spins . it should be emphasized , however , that the classical ferrimagnetic order originating from the antiferromagnetic ising interaction @xmath14 can be destroyed through the competing heisenberg interaction @xmath88 that brings a frustration into play . as a matter of fact ,
the spin order dramatically changes when the frustration parameter @xmath79 exceeds the boundary value @xmath89 in such case , all heisenberg spin pairs create singlet dimers and on account of this singlet pairing , all ising spins become completely free to flip .
in other words , the ising spins are thoroughly uncorrelated in fru on account of the frustration arising from the singlet pairing between the heisenberg spins as also suggested by ( [ frua ] ) .
owing to this fact , it might be concluded that the diamond spin chain splits into a set of independent monomers ( ising spins ) and dimers ( heisenberg spin pairs ) whenever the frustration parameter is stronger than its boundary value @xmath90 .
thus , fru is virtually being macroscopically degenerate monomer - dimer state with the residual entropy @xmath91 proportional to the total number of frustrated ising spins . for completeness
, it should be also noticed that sufficiently strong external field stabilizes the standard spp regardless whether fri or fru constitutes the zero - field ground state ( see fig .
as could be expected , all ising as well as heisenberg spins tend to align into the external - field direction above the saturation field , which represents lower bound for an occurrence of spp .
next , we turn our attention to the magnetization process at zero as well as non - zero temperatures . for this purpose , two typical magnetization vs. field dependences
are plotted in fig . 3 for several dimensionless temperatures .
it can be readily understood by comparing the displayed magnetization curves with the phase diagram shown in fig . 2b that two different zero - temperature limits obviously reflect both possible types of field - induced transitions fri - spp ( fig .
3a ) and fru - spp ( fig .
since the ground state is being formed in the former ( latter ) case by fri ( fru ) , the zero - temperature magnetization curve depicted in fig .
3b ) starts from non - zero ( zero ) magnetization in the limit of vanishing external field .
contrary to this , both magnetization curves always start from zero magnetization ( disordered state ) at any finite temperature according to the one - dimensional character of the investigated spin system
. it should be emphasized , moreover , that the magnetization jumps to be observed in the magnetization curves strictly at @xmath92 are gradually smeared out when temperature is raised from zero .
in addition , the higher the temperature , the smaller the width of magnetization plateaux ( horizontal regions in the magnetization vs. field dependence ) , which entirely disappear from the magnetization curves above a certain temperature .
finally , it is quite interesting to mention that the identified magnetization plateaux at one third of the saturation magnetization satisfy the oshikawa - yamanaka - affleck rule @xcite proposed for the formation of quantized plateaux .
now , let us step forward to a discussion concerned with the thermal dependence of the zero - field susceptibility times temperature ( @xmath93 ) data as displayed in fig .
4 . if the frustration parameter @xmath79 is selected so that fri constitutes the ground state ( fig .
4a ) , then , @xmath93 data exhibit a round minimum prior to exponential divergence appearing on temperature decrease . as it can be clearly seen from fig .
4a , the stronger the frustration parameter @xmath79 , the deeper the notable minimum whose position is simultaneously shifted towards lower temperatures .
it should be mentioned , moreover , that an appearance of the round minimum in the @xmath94 dependence is a typical feature of quantum ferrimagnets , since @xmath93 product monotonously decreases with temperature for ferromagnets , while it monotonously increases for antiferromagnets @xcite .
accordingly , a location of the round minimum can be regarded as a point that determines ferromagnetic - to - antiferromagnetic crossover in view of thermal exitations . on the other hand , if the frustration parameter @xmath79 drives the system into the disordered fru ground state ( fig .
4b ) , the @xmath93 product then exhibits a round minimum before it tends to the constant value @xmath95 by approaching zero temperature .
notice that this zero - temperature value can be explained in compliance with the curie law of the frustrated ising spins , which effectively form isolated spins ( monomers ) in fru .
another quantity , which is important for overall understanding of thermodynamics , is being the specific heat .
temperature variations of the zero - field specific heat are depicted in figs .
5a - b for @xmath78 and several values of the frustration parameter @xmath79 .
according to these plots , there still emerges at least one round maximum , which can be thought of as the usual schottky - type maximum irrespective whether fri or fru constitutes the ground state .
if the frustration parameter is selected sufficiently close to fri - fru phase boundary given by @xmath90 , however , there also appears an additional second maximum located in the low - temperature part of the specific heat .
apparently , the low - temperature peak becomes the more pronounced , the closer is @xmath79 selected to @xmath90 .
when the frustration parameter @xmath79 drives the system into the disordered fru ground state , then , the striking second maximum gradually disappears upon further strenghtening of @xmath79 , since the low - temperature peak shifts towards higher temperatures until it entirely merges with the high - temperature schottky - type maximum ( see the curves labeled as @xmath96 and @xmath97 ) .
these observations would suggest that the double - peak structure in the specific heat curves originates from thermal excitations between the ground - state spin configuration and the ones close enough in energy to the ground state .
the situation becomes even more intriguing on applying the external magnetic field .
as one can see from fig .
5c , the rising external field generally causes a gradual increase in the height of the low - temperature peak and moves it towards the higher temperatures .
similar trend is also seen in a change of the size and position of the high - temperature maximum , albeit this schottky - like peak moves towards higher temperature more slowly than the low - temperature one . as a result ,
both maxima coalesce at a certain value of the external field and above this value , the specific heat exhibits just single rounded maximum ( see for instance the curve @xmath98 ) . apart from these rather trivial findings
, a remarkable triple - peak specific heat curves can also be detected when small but non - zero external field is applied to the system driven by the frustration into the disordered fru state ( the case @xmath99 in fig .
5c ) . besides two aforedescribed peaks ,
whose origin has been resolved earlier , there also appears an additional third peak to be located at lowest temperature .
there are strong indications that an appearance of this additional sharp maximum can be explained through the field - induced splitting of the energy levels related to the frustrated ising spins . in accordance with this statement , the insert of fig .
5d clearly demonstrates how this peculiar third maximum gradually shifts towards higher temperatures with increasing the field strength until it coalesces with the second low - temperature peak ( see the curve labeled @xmath100 ) . at last
, we shall briefly discuss the adiabatic demagnetization curves studied in connection with the enhanced magnetocaloric effect .
some interesting results for adiabatic processes keeping entropy constant are presented in fig . 6 in the form of the temperature vs. external field dependence .
two depicted sets of demagnetization curves reflect two available adiabatic scenarios related to spp@xmath101fri ( fig .
6a ) and spp@xmath101fru ( fig .
6b ) transitions .
apparently , the maximal cooling rate emerges in the vicinity of critical fields and zero field , where zero temperature is in principle reached whenever the entropy is set equal to or less than its residual value @xmath102 .
it should be pointed out that a relatively fast heating , which occurs when the external field is lowered from its critical value , prevents a practical use of the whole demagnetization curves for a cooling purpose .
in addition , the cooling effect becomes of technological relevance only if a cooling rate exceeds the one of paramagnetic salts . from this point of view
, the enhanced magnetocaloric effect is observed only if the frustration drives the system into disordered fru ground state ( fig .
6b ) . even under this condition , the cooling rate of paramagnetic materials
is exceeded only if the entropy is chosen close enough to its residual value @xmath102 and the external field is below @xmath103 .
this limitation would imply that temperatures in the sub - kelvin range are in principle accessible provided that the frustrated diamond chain compound with exchange constants of the order of few tents kelvin , such as azurite , is used as refrigerant . in this part , we shall turn our attention to the mixed spin-@xmath0 and spin-@xmath34 ising - heisenberg diamond chain with the aim to clarify the impact of integer - valued heisenberg spins on the magnetic behaviour of the frustrated diamond chain .
we start our discussion repeatedly with the ground - state analysis . the phase diagram constructed in an absence of the external field ( fig .
7a ) implies existence of three possible ground states .
besides the semi - classically ordered ferrimagnetic phase ( fri ) , there also appear the quantum ferrimagnetic phase ( qfi ) and the frustrated phase ( fru ) .
fri , qfi and fru can be characterized by means of @xmath104 @xmath105 @xmath106_{3k-1 , \ , 3k } , \nonumber \\
m_{\rm i}^z = 0 , m_{\rm h}^z = 0 , q_{\rm hh}^{zz } = -c_{\rm hh}^{zz } = ( 1 + \delta^{-1})/2 , c_{\rm hh}^{xx } = -2 \delta \delta^{-1 } , c_{\rm ih}^{zz } = 0 ; \label{fru } \end{aligned}\ ] ] where @xmath107 , the first product in the aforelisted eigenfunctions is taken over all ising spins ( @xmath84 stands for @xmath86 ) and the second product runs over all pairs of the heisenberg dumbbell spins ( @xmath108 is assigned to @xmath109 ) .
analytic expressions for the phase boundaries depicted in fig .
6a read @xmath110 the most significant difference between the two investigated diamond chains apparently rests in a presence of qfi located in between fri and fru .
this observation would suggest that the geometric frustration initially favours uprise of qfi before it finally energetically stabilizes fru .
accordingly , it might be concluded that there may not occur in the mixed spin-@xmath0 and spin-1 diamond chain a direct frustration - induced transition between fri and fru except the one observable in the ising limit ( @xmath111 ) .
next , the ground - state phase diagram reflecting the effect of external field is shown in fig .
this phase diagram suggests that in response to the applied external field there also may arise the quantum ferromagnetic phase ( qfo ) and the saturated paramagnetic phase ( spp ) besides the aforementioned fri , qfi and fru phases @xmath112 @xmath113 before proceeding further , let us make few comments on all these possible ground states . since fri and spp
are commonly observed also in the semi - classical ising spin systems , we should therefore concentrate on qfi , qfo and fru , in which quantum entanglement of the heisenberg spin pairs indicates the quantum nature of these phases .
it can easily be understood from eqs .
( [ qfi ] ) and ( [ qfo ] ) that qfi and qfo are quite similar one to each other . as a matter of fact ,
all pairs of heisenberg spins reside in both phases the same eigenstate and thus , the only difference between them consists in the spin arrangement of their ising counterparts . in qfo , which is stable at stronger fields ,
the ising spins tend to align towards the external field , whilst they are oriented in opposite to the external field in qfi , which is stable at relatively weaker fields .
notice that the quantum entanglement between the pairs of heisenberg spins that occurs in qfi and qfo can also be understood within the valence - bond - solid picture @xcite .
if spin-@xmath34 sites are decomposed into two spin-@xmath0 variables , then , one of the decomposed spins at each spin-@xmath34 site forms the singlet - dimer with its nearest - neighbouring spin-@xmath34 site , while the other one is polarized by the external field . as a result of this incomplete pairing
, each spin-@xmath34 site effectively acts in qfi and qfo as it would be the spin-@xmath0 variable .
further , it should be also remarked that all ising spins are completely free to flip ( frustrated ) in fru on behalf of a preferred antiferromagnetic alignment between each pair of the heisenberg spins .
owing to this fact , fru can be viewed as a state characterized by a complete randomization of the ising spins , what consequently leads to a macroscopic degeneracy of fru resembled in its residual entropy @xmath91 . to clarify the magnetization scenario available for the mixed spin-@xmath0 and spin-@xmath34 diamond chain , we depict in fig .
8 all possible types of magnetization curves . in agreement with the phase diagram shown in fig .
7b , there are in total four different types of magnetization curves reflecting the transitions fri - spp ( a ) , qfi - fri - spp ( b ) , fru - fri - spp ( c ) , and fru - qfo - spp ( d ) .
it should be stressed , nevertheless , that the system undergoes true transitions between those phases merely at zero temperature , where indeed emerge stepwise magnetization curves with abrupt change(s ) of the magnetization at critical field(s ) .
however , it is worthy to notice that there are no real magnetization jumps at any finite temperature and the sharp stepwise magnetization curves to be observed at zero temperature are gradually smeared out as temperature increases . it can be clearly seen from fig . 8
that an increase in temperature actually shrinks also the width of plateaux until the plateau states completely disappear from the magnetization curves above a certain temperature .
the most notable magnetization curves are those with the zero - field ground state corresponding to fru as shown in figs .
8c - d . according to these plots
, it can easily be realized that all frustrated ising spins tend to align to the external - field direction for any finite but non - zero external field provided that temperature is set to zero . at non - zero temperatures , on the contrary ,
the magnetization rises much more steadily in the vicinity of zero field in comparison with the magnetization curves having the long - range - ordered fri ( fig . 8a ) and qfi ( fig .
8b ) phases in the ground state .
finally , it is also worthwhile to remark that the observed magnetization plateaux at @xmath114 ( qfi and fru ) and @xmath115 ( qfo ) of the saturation magnetization satisfy the oshikawa - yamanaka - affleck rule @xcite .
now , let us investigate in particular an influence of the spin frustration on thermal variations of the zero - field susceptibility times temperature ( @xmath93 ) data .
the temperature dependence of @xmath93 product is displayed in fig .
9a for several values of the frustration parameter @xmath79 that determine the ground state to be either of fri or qfi type .
interestingly , @xmath93 data then exhibit a round minimum upon cooling , which is followed by an exponentially steep increase that appears by approaching zero temperature .
it is quite obvious from fig . 9a
that the stronger the parameter of frustration @xmath79 , the deeper the notable minimum whose position is simultaneously shifted towards lower temperatures .
it is worthy to remember , moreover , that the minimum in the @xmath116 plot is being a typical feature of quantum ferrimagnets @xcite and also in our case , the minimum becomes especially marked by selecting @xmath117 when qfi constitutes the ground state . on the other hand ,
the temperature dependence of @xmath93 data are depicted in fig .
9b for the case when the frustration parameter @xmath118 drives the system into the disordered fru ground state . under these circumstances ,
the susceptibility diverges as @xmath119 at low temperatures and @xmath93 product tends to constant value 1/12 nearby zero temperature .
this value can be interpreted in terms of the curie law of the frustrated ising spins , since the spin arrangement that appears in fru can be viewed as being composed of an independent set of the antiferromagnetic heisenberg dimers and isolated ising monomers , whereas the former ones do not contribute to @xmath93 product in the zero - temperature limit . to gain an insight into overall thermodynamics ,
let us turn our attention to a thermal behaviour of the specific heat . for this purpose ,
the zero - field specific heat is plotted against temperature in figs .
10a - b for several values of the frustration parameter @xmath79 .
it can be clearly seen from fig .
10a that a rounded schottky - type maximum observable at smaller values of @xmath79 ( e.g. @xmath120 ) gradually changes , as the frustration strengthens , to a striking dependence with the shoulder superimposed on this round maximum ( @xmath121 ) .
it is noteworthy that the shoulder becomes the more pronounced , the closer is @xmath79 selected to @xmath122 determining the phase boundary between fri and qfi . under further increase of the frustration
the shoulder merges with the schottky - type maximum what eventually gives rise to a peculiar non - rounded maximum with almost constant value of the specific heat over the wide temperature range ( @xmath123 ) .
next , fig .
10b shows how the specific heat recovers its double - peak structure when @xmath79 approaches another phase boundary between qfi and fru to appear at @xmath124 ( see for instance the curves for @xmath125 and @xmath126 ) .
in addition , it is quite evident from fig .
10b that repeated strengthening of the frustration results in a suppression of the low - temperature peak until it finally merges with the high - temperature one .
note furthermore that the high - temperature peak has in general tendency to enhance in magnitude ( both in height as well as in width ) as @xmath79 increases and its position is shifted towards higher temperatures .
altogether , it might be concluded that the remarkable double - peak structure of the specific heat arises just when the frustration parameter is close enough to a phase boundary .
this result is taken to mean that the observed double - peak specific heat curves always originate from thermal excitations to a spin configuration rather close in energy to the ground state .
even more striking situation emerges by turning on the external field .
the overall trends of the external field is to increase height and width of the low - temperature peak and to shift it towards higher temperatures until it coalesces with the higher - temperature maximum .
it should be remarked , nevertheless , that the most notable dependences of the specific heat arise from when the frustration leads to the disordered fru ground state . besides the aforedescribed general trends illustrated in fig .
10c , there also appears a remarkable kind of the specific heat curve with the triple - peak structure as particularly drawn in fig . 10d and its insert .
apparently , the additional third peak observable at very low temperatures occurs on applying the small but non - zero external field .
it is therefore quite reasonable to conjecture that an origin of this low - temperature peak lies in the field - induced splitting of energy levels related to the frustration of the ising spins . actually , the stronger the external field , the greater the splitting caused by the external field and consequently , the position of this peak steadily shifts towards higher temperatures . above certain external field ,
the additional third peak vanishes because of merging with the low - temperature peak observable also in the zero - field case .
finally , we shall close our discussion with an exploration of the adiabatic demagnetization examined in connection with the enhanced magnetocaloric effect .
adiabatic processes keeping the entropy constant are plotted in fig .
11 in the form of the temperature vs. external field dependence .
two displayed sets of demagnetization curves evidently reflect adiabatic change of temperature , which accompanies the transitions spp@xmath127fri@xmath127qfi ( fig .
11a ) and spp@xmath127qfo@xmath127fru ( fig .
11b ) . when comparing these results with the ones formerly discussed for the spin-@xmath0 diamond chain ( fig .
6 ) , one easily finds some similarities between the two investigated diamond chains . indeed ,
the most obvious drop in temperature is retrieved once again in the vicinity of critical fields and zero field , where zero temperature is in principle reached whenever the entropy is set equal to or less than its residual value @xmath102 . moreover , the enhanced magnetocaloric effect with the cooling rate exceeding the one of paramagnetic salts occurs just as the disordered fru phase constitutes the ground state and the external field is below @xmath103 .
by contrast , the most obvious difference between the two investigated spin systems consists in the greater diversity of the adiabatic process of the latter ( mixed - spin ) system , which exhibits dependences with two critical fields in addition to the ones with one critical field only .
exactly solvable frustrated spin systems are currently much sought after in the field of condensed matter physics , since they can serve as useful model systems for in - depth understanding of the effect of geometric frustration still not fully elucidated , yet . in the present work ,
we have provided rigorous results for one notable example of such system , the mixed spin-@xmath0 and spin-@xmath1 ising - heisenberg diamond chain tractable within the generalized decoration - iteration map .
it is worthy to mention that this rather simple model system has primarily been developed to predict and to understand the behaviour of insulating magnetic materials , in which the heisenberg dimers interact with the ising monomers in such a way that they form the diamond chain .
notice that the coordination polymers consisting of the pairs of interacting transition - metal elements ( heisenberg dimers ) coupled to the rare - earth elements ( ising monomers ) represent perspective experimental realizations of the proposed system .
although we are not aware of any real coordination compound , which would meet this requirement , the recent progress in the design and controlled synthesis of the molecular - based magnetic materials supports our hope that it would be possible to prepare such polymeric chains in the near future .
it is worthwhile to remark that a special emphasis is in our study laid on the investigation of geometric frustration generated by the competition between heisenberg- and ising - type interactions .
our results clearly demonstrate that an interplay between the frustration and quantum fluctuations , both arising from the heisenberg exchange interaction , is being at the origin of interesting behaviour not commonly observed in the semi - classical ising spin systems .
indeed , this interplay gives rise to several peculiar ground states with entangled states of heisenberg spins and quantum effects turned out to play a substantial role in determining their finite - temperature properties , as well . among other matters ,
we have found rigorous evidence for appearance of the quantized plateux in the magnetization curves , the round minimum in temperature dependence of the susceptibility times temperature data , the double - peak zero - field specific heat curves , the enhanced magnetocaloric effect and so on . in our opinion ,
the most interesting finding to emerge from our study is a direct evidence of the triple - peak specific heat curve that appears when applying small external field to the system driven by the frustration into the disordered state . to the best of our knowledge ,
the discovery and possible explanation of the triple - peak structure in the specific heat curve has not been reported for any frustrated system hitherto .
last but not least , it should be remarked that the relative ease of generalized mapping method used here implies a possibility of further extensions .
actually , this approach can straightforwardly be extended to account also for the single - ion anisotropy effect , the biquadratic exchange interaction , the next - nearest - neighbour interaction , the multispin interactions , etc .
it is also noteworthy that the applied procedure is not constrained neither by the lattice topology and thus , it can be utilized for investigating the effect of geometric frustration on the planar ising - heisenberg lattices composed of the diamond - shaped units @xcite . in this direction
continuous our next work .
before we list explicit expressions for the function @xmath32 and @xmath65 ( @xmath63 = 14 ) , which enter into the transformation formulae ( [ eq : dit])-([eq : ar ] ) and the set of eqs .
( [ eq : klm1])-([eq : klm4 ] ) , respectively , let us establish the notation @xmath128 , @xmath129 and @xmath107 .
+ ( a ) expressions for the spin case @xmath33 : @xmath130 @xmath131/\bigl[2g(x)\bigr ] , \nonumber \\
f_3(x ) & = & - \exp(j_{\rm h}/4 ) \sinh ( j_{\rm h}\delta/2)/\bigl[2g(x)\bigr ] .
\nonumber \end{aligned}\ ] ] \(b ) expressions for the spin case @xmath132 : @xmath133 \nonumber \\ & + & \exp(j_{\rm h } ) + 4 \cosh(x - h_{\rm h } ) \cosh(j_{\rm h } \delta ) , \nonumber \end{aligned}\ ] ] @xmath134 + 2\sinh(x - h_{\rm h } ) \cosh(j_{\rm h } \delta)\}/g(x ) , \nonumber \\
f_2(x ) & = & \{2 \exp(-j_{\rm h } ) \cosh[2(x - h_{\rm h } ) ] - \exp(j_{\rm h } ) - \delta \mathcal{c } - \mathcal{s}\}/g(x ) , \nonumber \\
f_3(x ) & = & - \{2 \cosh(x - h_{\rm h } ) \sinh(j_{\rm h } \delta ) + 4 \delta \mathcal{s}\}/g(x ) , \nonumber \\
f_4(x ) & = & \{2 \exp(-j_{\rm h } ) \cosh[2(x - h_{\rm h } ) ] + \exp(j_{\rm h } ) \nonumber \\ & + & 2 \cosh(x - h_{\rm h } ) \cosh(j_{\rm h } \delta ) + \delta \mathcal{c } + \mathcal{s}\}/g(x ) , \nonumber \end{aligned}\ ] ] where @xmath135 and @xmath136 .
the authors express gratitude to scientific grant agencies vega and apvv supporting this work under the grants vega 1/2009/05 and apvt 20 - 005204 .
100 richter j , krger s , farnell d and bishop r 2001 quantum phase transitions in spin systems _ series on advances in quantum many - body theory _ vol 3 , ed bishop r f , gernoth k a and walet n r ( singapore : world scientific ) pp 239246 * diagrammatic representation of the mixed - spin ising - heisenberg diamond chain .
the empty ( filled ) circles denote the lattice positions of the ising ( heisenberg ) spins , the ellipse demarcates @xmath28th diamond - shaped plaquette . *
\(a ) ground - state phase diagram in the @xmath137 plane for the system in an absence of the external magnetic field ; ( b ) ground - state phase diagram in the @xmath138 plane for @xmath78 .
* the total magnetization reduced with respect to its saturation value versus the external magnetic field at various temperatures @xmath139 = 0.0 , 0.1 , 0.3 , 0.5 in ascending order along the direction of arrows . * thermal dependence of the zero - field susceptibility times temperature data for @xmath78 , @xmath79 = 0.25 , 0.5 , 0.75 , 0.9 , 1.0 ( fig .
4a ) and @xmath78 , @xmath79 = 1.1 , 1.25 , 1.5 , 2.0 ( fig .
4b ) in ascending order along the direction of arrows . for clarity , the case @xmath140 = 1.0 corresponding to the fri - fru phase boundary
is depicted as broken line . *
temperature variations of the specific heat when @xmath78 is fixed . figs .
5a - b illustrate the effect of frustration parameter @xmath79 on the shape of zero - field specific heat and figs .
5c - d display the effect of the applied external magnetic field when the frustration parameter @xmath141 drives the system into the disordered fru state .
* adiabatic demagnetization in the form of temperature versus external field dependence for @xmath78 , @xmath142 ( fig .
6a ) and @xmath78 , @xmath143 ( fig .
6b ) . for better orientation , broken curve
depicts the dependence when entropy is fixed at the residual value @xmath91 of fru . *
\(a ) ground - state phase diagram in an absence of the external field ; ( b ) ground - state phase diagram in the @xmath138 plane for @xmath78 .
* the total magnetization reduced with respect to its saturation value as a function of the external field at various temperatures @xmath139 = 0.0 , 0.1 , 0.3 , 0.5 in ascending order along the direction of arrows . * thermal dependence of the zero - field susceptibility times temperature data for @xmath78 , @xmath79 = 0.25 , 0.4 , 0.5 , 0.6 , 0.75 , 0.9 , 1.0 ( fig .
9a ) and @xmath78 , @xmath79 = 1.1 , 1.25 , 1.5 , 1.75 ( fig .
9b ) in ascending order along the direction of arrows . for clarity ,
the cases @xmath144 = 0.5 and @xmath145 = 1.0 corresponding to the phase boundaries between fri - qfi and qfi - fru , respectively , are depicted as broken lines . *
temperature variations of the specific heat when @xmath78 is fixed .
10a - b illustrate the effect of frustration parameter @xmath79 on the shape of zero - field specific heat , whereas figs .
10c - d display the effect of applied external field when the frustration parameter @xmath141 drives the system into the disordered fru state .
* adiabatic demagnetization in the form of temperature versus external field dependence when @xmath78 , @xmath125 ( fig .
11a ) and @xmath78 , @xmath146 ( fig .
11b ) . for clarity , broken curve
depicts the dependence when entropy is fixed at the residual value @xmath91 of fru . | ground - state and finite - temperature properties of the mixed spin-@xmath0 and spin-@xmath1 ising - heisenberg diamond chains are examined within an exact analytical approach based on the generalized decoration - iteration map .
a particular emphasis is laid on the investigation of the effect of geometric frustration , which is generated by the competition between heisenberg- and ising - type exchange interactions .
it is found that an interplay between the geometric frustration and quantum effects gives rise to several quantum ground states with entangled spin states in addition to some semi - classically ordered ones . among the most interesting results to emerge from our study
one could mention a rigorous evidence for quantized plateux in magnetization curves , an appearance of the round minimum in the thermal dependence of susceptibility times temperature data , double - peak zero - field specific heat curves , or an enhanced magnetocaloric effect when the frustration comes into play .
the triple - peak specific heat curve is also detected when applying small external field to the system driven by the frustration into the disordered state . |
the unitary extensions of chiral perturbation theory @xmath6 have brought new light in the study of the meson - baryon interaction and have shown that some well known resonances qualify as dynamically generated , or in simpler words , they are quasibound states of a meson and a baryon , the properties of which are described in terms of chiral lagrangians .
after early studies in this direction explaining the @xmath7 and the @xmath0 as dynamically generated resonances @xcite , more systematic studies have shown that there are two octets and one singlet of resonances from the interaction of the octet of pseudoscalar mesons with the octet of stable baryons @xcite .
the @xmath0 belongs to one of these two octets and plays an important role in the @xmath8 interaction with its coupled channels @xmath9 , @xmath10 and @xmath11 @xcite . here , we adopt and extend the ideas of ref .
@xcite for the reaction @xmath12 and study the analogous reaction @xmath2 where the @xmath13 final state can form the @xmath0 resonance . besides processes relevant in @xmath14 @xcite
, we also include @xcite the contribution from the @xmath1 resonance which qualifies as dynamically generated through the interaction of the @xmath15 meson octet and the @xmath16 baryon decuplet as recent studies show @xcite . at the same time we also study the @xmath3 reaction and
make predictions for its cross section , taking advantage of the fact that it appears naturally within the coupled channels formalism of the @xmath2 reaction and leads to a further test of consistency of the ideas explored here .
both reactions are currently being analyzed at elsa@xcite .
before turning to the two - meson photoproduction , we give a short review of the properties of the @xmath0 in the meson - baryon sector , where this resonance shows up clearly in the spin isospin @xmath17 channel . the model of ref .
@xcite provides an accurate data description of elastic and quasielastic @xmath8 scattering in the @xmath18 channel . within the coupled channel approach in the @xmath19 representation of ref .
@xcite , not only the @xmath8 final state is accessible , but also @xmath20 , @xmath21 , and @xmath22 in a natural way .
the corresponding @xmath23 coefficients for the @xmath24-wave interaction kernel in the charge @xmath25 sector are found in ref .
besides the interaction from the lowest order chiral lagrangian @xcite , also the @xmath26 two loop contribution is included . in ref .
@xcite it is shown that the same model for the resonance provides a good data description close to threshold for the one-@xmath27 photoproduction , @xmath28 , when including some basic photoproduction mechanisms , which makes us confident to use the model of the @xmath0 for the more complex reaction with two mesons in the final state in the next section . in fig .
[ fig : overview_production ] ( 300,130 ) ( -45,-5 ) final state.,title="fig:",width=453 ] the relevant photoproduction mechanisms are shown .
diagrams ( c ) shows the pole terms that accompanies the contact term ( b ) in order to ensure gauge invariance .
similar terms are included for diagrams ( a ) , ( d ) , and ( f ) but not separately drawn . in all diagrams , the gray blob signifies the dynamically generated @xmath0 that provides the final state interaction @xmath30 with the intermediate meson - baryon state in the loop given by the list of channels from above . here , we can only give a qualitative description of the processes and the amplitudes which are denoted in detail in ref .
@xcite .
diagram ( a ) uses a contact term in a similar way as for the @xmath31 photoproduction in ref .
@xcite , additionally taking into account the anomalous magnetic moment of the nucleon with the effective lagrangian from ref .
diagrams ( b ) and ( c ) are a straightforward extension of the basic photoproduction mechanisms in @xmath32 @xcite for the final state with an additional @xmath33 .
it is known that the inclusion of explicit resonances plays an important role in the two - pion photoproduction @xcite , and therefore we include the relevant mechanisms as sub - processes in the calculation : in diagram ( e ) we show the contribution from the @xmath1 decay into @xmath34 and @xmath35 . in the same way , the @xmath36 decay from ref .
@xcite is taken into account .
( f ) shows the corresponding @xmath37-kroll - ruderman term ( @xmath38 ) and also the @xmath39-kroll - ruderman term ( @xmath40 ) .
additionally , we include the contribution from the dynamically generated @xmath5 resonance from ref .
@xcite which has been identified with the @xmath1 . in particular , the decay channels @xmath41 in diagram ( e ) are possible , whose strengths are not measured yet in experiment but a genuine prediction of the model @xcite . from these couplings ,
also a tree level diagram originates for the photoproduction : @xmath42 $ ] .
the invariant mass spectra for @xmath43 and @xmath44 and cross section for the @xmath45 reaction are predicted which can be directly compared to the forthcoming experiments at the elsa facility . in fig .
[ fig : bigger_window ] the spectrum for @xmath46 is shown for several lab photon energies @xmath47 , taking into account all processes from fig .
[ fig : overview_production ] plus the tree level process described at the end of sec .
[ sec : production ] .
the solid and dashed curves correspond to the full and reduced model ( no @xmath48 channel , no vector exchange in @xmath49-channel , see ref .
@xcite , @xcite ) for the @xmath0 resonance , and we take the difference as a hint for the theoretical error . (
300,150 ) ( -45,-5 ) $ ] as a function of @xmath43 [ mev ] for various photon lab energies @xmath47 .
solid and dashed lines : full and reduced model for the @xmath0 , respectively.,title="fig:",width=491 ] the @xmath0 is clearly visible in the spectrum , although it interferes with the tree level process at low @xmath47 which shifts its position slightly . at higher @xmath47 , we observe a second peak which moves with the energy .
this is a reflection of the @xmath50 in the tree level process which is on shell around these invariant masses . for a more detailed discussion ,
the @xmath51 invariant mass spectrum is not shown here , but in ref .
@xcite . there
, the @xmath50 shows up as a clean signal as expected .
the total cross section is plotted in fig .
[ fig : cross_section ] .
( 300,110 ) ( -45,-5 ) for the @xmath52 reaction .
right side : @xmath53 for the @xmath54 reaction .
solid line : full model for the @xmath0 . dashed line : reduced model . dashed dotted line ( left side ) : phenomenological potential for the @xmath55 transition ( only available up to @xmath56 gev).,title="fig:",width=215 ] ( 140,-5 ) for the @xmath52 reaction
right side : @xmath53 for the @xmath54 reaction .
solid line : full model for the @xmath0 . dashed line : reduced model . dashed dotted line ( left side ) : phenomenological potential for the @xmath55 transition ( only available up to @xmath56 gev).,title="fig:",width=226 ] additionally to the full and reduced model
, we show the outcome when using the phenomenological meson - baryon @xmath57 transition from the pwa from ref . @xcite . from the processes with @xmath0 , diagram ( e ) from fig .
[ fig : overview_production ] with @xmath58 , @xmath34 gives the largest contribution which is due to the strong @xmath59 transition in the model for the @xmath0 . also , the tree level diagram contributes significantly as we have already seen in fig .
[ fig : bigger_window ] .
the cross section is relatively independent of the chosen model for the @xmath0 ( solid vs. dashed line ) , and even the the phenomenological potential does not change much the result .
the @xmath60 reaction is calculated in a similar way as for the @xmath61 final state by replacing the @xmath62 final state of the @xmath0 by the @xmath63 one and including a new tree level diagram in analogy to the process described at the end of sec .
[ sec : production ] : @xmath64 $ ] .
the cross section for @xmath65 is much smaller than for the @xmath61 final state as fig .
[ fig : cross_section ] shows . this is among other reasons @xcite due to the fact that the reaction takes places at much higher photon energies where the dynamically generated @xmath0 is off shell .
in this paper we have studied the reactions @xmath66 and @xmath3 within a chiral unitary framework which considers the interaction of mesons and baryons in coupled channels and dynamically generates the @xmath0 . from the various processes , we find dominant the decay of the dynamically generated @xmath1 resonance into @xmath67 , followed by the unitarization , or in other words , the @xmath68 decay .
a similar term provides also a tree level process which leads , together with the @xmath0 , to a characteristic double hump structure in the @xmath62 and @xmath51 invariant masses at higher photon energies .
a virtue of this approach and a test of the nature of the resonance as a dynamically generated object is that one can make predictions about cross sections for the production of the resonance without introducing the resonance explicitly in the formalism since only its components in the @xmath69 and @xmath70 meson - baryon base is what matters , together with the coupling of the photons to these components and their interaction in a coupled channel formalism .
in particular , the reactions studied here probe decay channels of these resonances such as @xmath71 or transitions like @xmath72 which are predicted by the model and not measured yet .
the measurement of both cross sections is being performed at the elsa / bonn laboratory and hence the predictions are both interesting and opportune and can help us gain a better insight in the nature of some resonances , particularly the @xmath0 and the @xmath1 in the present case .
this work is partly supported by dgicyt contract number bfm2003 - 00856 , and the e.u .
euridice network contract no .
hprn - ct-2002 - 00311 .
this research is part of the eu integrated infrastructure initiative hadron physics project under contract number rii3-ct-2004 - 506078 .
99 n. kaiser , p. b. siegel and w. weise , phys .
b * 362 * ( 1995 ) 23 .
n. kaiser , t. waas and w. weise , nucl .
a * 612 * ( 1997 ) 297 e. oset and a. ramos , nucl .
a * 635 * ( 1998 ) 99 j. c. nacher , a. parreno , e. oset , a. ramos , a. hosaka and m. oka , nucl .
a * 678 * ( 2000 ) 187 j. a. oller and u. g. meissner , phys .
b * 500 * ( 2001 ) 263 d. jido , j. a. oller , e. oset , a. ramos and u. g. meissner , nucl .
a * 725 * ( 2003 ) 181 c. garcia - recio , m. f. m. lutz and j. nieves , phys .
b * 582 * ( 2004 ) 49 t. inoue , e. oset and m. j. vicente vacas , phys .
c * 65 * ( 2002 ) 035204 j. c. nacher , e. oset , h. toki and a. ramos , phys .
b * 455 * ( 1999 ) 55 j. c. nacher , e. oset , m. j. vicente and l. roca , nucl .
a * 695 * , 295 ( 2001 ) m. doring , e. oset and d. strottman , arxiv : nucl - th/0510015 .
e. e. kolomeitsev and m. f. m. lutz , phys . lett .
b * 585 * , 243 ( 2004 ) s. sarkar , e. oset and m. j. vicente vacas , nucl .
a * 750 * , 294 ( 2005 ) v. metag and m. nanova , private communication . j. gasser and h. leutwyler , nucl .
* b250 * ( 1985 ) 465 , 517 , 539 .
u. g. meissner , rep .
56 ( 1993 ) 903 ; v. bernard , n. kaiser and u. g. meissner , int
e4 ( 1995 ) 193 .
g. ecker , prog .
35 ( 1995 ) 1 .
d. jido , a. hosaka , j. c. nacher , e. oset and a. ramos , phys .
c * 66 * ( 2002 ) 025203 r. a. arndt , w. j. briscoe , i. i. strakovsky , r. l. workman and m. m. pavan , phys .
c * 69 * , 035213 ( 2004 ) | using a chiral unitary approach for meson - baryon scattering in the strangeness zero sector , where the @xmath0 and @xmath1 resonances are dynamically generated , we study the reactions @xmath2 and @xmath3 at photon energies at which the final states are produced close to threshold . among several reaction mechanisms , we find the most important is the excitation of the @xmath1 state which subsequently decays into a pseudoscalar meson and the @xmath0 .
hence , the reaction provides useful information with which to test current theories of the dynamical generation of the low - lying @xmath4 and @xmath5 states .
predictions are made for cross sections and invariant mass distributions which can be compared with forthcoming experiments at elsa . |
the study of glueballs currently deserves much interest from a theoretical point of view , either within the framework of lattice qcd or effective models .
their experimental detection is also an active field of research for recent reviews , see refs .
@xcite . among the various effective approaches which have been proposed ( bag model @xcite , qcd in coulomb gauge @xcite , )
, many studies have been devoted to potential models of glueballs . in this framework , glueballs are seen as bound states of two or more constituent gluons interacting via a phenomenological potential .
early works on this subject are quoted in refs .
@xcite , and more recent studies can be found in refs .
@xcite . actually , the relevance of using a potential model to describe a glueball is still controversial . assuming that a potential model can be used
, two basic questions then appear : what is the mass of a constituent gluon , and what is the potential ?
the answer to these questions differ from one approach to another .
let us begin by the problem of the mass .
on the one hand , it is argued in some works that a gluon is a massless particle , which acquires a dynamical mass given by @xmath2 because of the confining interaction .
relativistic , spin - dependent , corrections , are then developed in powers of @xmath3 @xcite . in this picture
, the constituent gluon is _ a posteriori _ massive , because it is confined into a glueball , and this mass is state - dependent .
it is worth mentioning that , more generally , both quarks and gluons can acquire a constituent mass from renormalization theory .
this constituent mass runs with the momentum : one can look in particular at the coulomb gauge approach of ref .
@xcite , where it is shown that massless gluons acquire a running mass which is about @xmath1 gev at zero momentum .
the same approach can also be applied to compute the constituent mass of light quarks @xcite . on the other hand
, it is often assumed in different studies that a constituent gluon has to be _ a priori _ considered as massive @xcite .
the idea underlying these last approaches is roughly that the nonperturbative effects of qcd causes a mass term to appear in the gluon propagator .
consequently , the gluons should be seen as massive particles , with a fixed mass which is typically assumed to be around @xmath4 gev @xcite .
the relativistic corrections are then expanded as usual in powers of @xmath5 .
interestingly , the typical value of @xmath6 in the second approach is compatible with the dynamical mass @xmath7 for the ground state in the first approach , and with the constituent gluon mass at zero momentum of the coulomb gauge model of ref .
@xcite .
we turn now our attention to the potentials appearing in the various existing models of qcd . in a two - body system ,
the best - known phenomenological potential is the cornell one , which is roughly of the form @xmath8 , @xmath9 being the separation between the confined particles .
it is worth mentioning that @xmath10 is the energy of a straight string of energy density @xmath11 , also called the flux tube , linking the quark to the antiquark and encoding the confining interaction .
the coulomb part is the lowest order contribution of the one gluon exchange processes .
the cornell potential arises from qcd in the case of a quark - antiquark bound state , as it can be shown by the wilson loop technique @xcite .
lattice qcd computations of the energy between a static quark - antiquark pair also support this potential @xcite .
furthermore , background perturbation theory tells that the potential between two massless constituent gluons should also be of the cornell form @xcite . bound states of gluons with the cornell potential
have been investigated for example in refs .
but , as a linearly rising potential neglects string breaking effects , which have been observed in lattice qcd @xcite between static quarks , another confining potential is also often used , that is a saturating one of the form @xmath12 .
as for the cornell one , models built on such a potential have been applied to usual hadrons @xcite , but also to glueballs @xcite .
let us note that the short - range part corresponding to massive constituent gluons is not a coulomb term , but is proportional to the yukawa potential @xmath13 .
> from this discussion , we can conclude that the best way of dealing with constituent gluons is still controversial .
consequently , it is of particular interest to try to obtain relevant informations from more fundamental approaches such as lattice qcd .
if the mass spectrum of pure gauge qcd the glueball spectrum is now accurately computed in lattice qcd @xcite , the potential energy between two constituent gluons has been much less studied than the quark - antiquark one . up to know ,
the only method to obtain this energy with lattice qcd is to compute the energy between two static sources , these sources being in the adjoint representation of su(3 ) @xcite .
the cornell shape is then favored , as in the quark - antiquark case .
however , nowadays , both the masses and wave functions of glueballs can be computed by lattice calculations @xcite .
we propose in this paper a new method for extracting the potential between two constituent gluons from these lattice qcd data .
such a method has the conceptual advantage of dealing with
physical " glueballs rather than with somewhat artificial static sources .
it is a direct application of the lagrange mesh procedure that we presented in ref .
@xcite .
our paper is organized as follows . in sec .
[ latti ] , we recall the main lattice qcd results concerning the lightest scalar glueball .
then , we describe the method to compute the effective gluon - gluon potential in sec .
[ compu ] , and we comment our results in sec . [
sec : res ] .
we finally draw some conclusions in sec .
[ conclu ] .
an su(3 ) lattice calculation in glueball spectroscopy shows that the lightest glueball is a scalar particle , whose quantum numbers are @xmath14 , and whose mass is given by @xmath15 gev @xcite .
theoretical arguments also support this point @xcite .
the su(2 ) wave function of this scalar glueball has been first computed in ref .
@xcite , and its mass was found to be around @xmath16 gev , which is lower than the currently accepted su(3 ) value .
more recently , the su(3 ) scalar glueball wave function has been computed @xcite . in this last work
, it is found that @xmath17 in agreement with the result of ref .
@xcite .
the @xmath0 radial wave function which is computed in ref .
@xcite seems thus to be a reliable result . as only a few points of this wave function
are available , it is more convenient for latter calculations to fit them by the trial function @xmath18.\ ] ] the size parameter @xmath19 fm @xmath20 gev@xmath21 is interpreted as the glueball radius in ref . @xcite . its introduction allows to deal with dimensionless fit parameters @xmath22 and @xmath23 .
a fit with the levenberg - marquardt algorithm gives @xmath24 with a satisfactory agreement since the coefficient of determination is equal to @xmath25 , close to the optimal value of @xmath26 .
the result is plotted in fig .
let us note that @xmath27 is normalized in such a way that @xmath28 .
consequently , the lattice qcd calculations of ref .
@xcite provide us not only with the @xmath0 glueball mass @xmath29 ( [ p1 ] ) , but also with a radial wave function , denoted as @xmath27 and given by eqs .
( [ fit ] ) and ( [ p2 ] ) .
in this work , we want to find a hamiltonian @xmath30 such that @xmath31 where @xmath27 and @xmath29 come from lattice qcd . this eigenequation is the main equation to be solved in the following . to our knowledge , this problem has never been addressed before .
it is of particular interest in order to check the relevance of the approaches with constituent gluons . in terms of potential models ,
a glueball can be a bound state of two or more constituent gluons coupled in a color singlet .
we point out that all the potential models agrees with the fact that the lowest @xmath0 should be a mainly two - gluon bound state @xcite .
moreover , the quasiparticle picture of ref .
@xcite , based on the qcd hamiltonian in coulomb gauge , also favors a bound state of two constituent gluons for the lowest @xmath0 .
we will thus assume for @xmath30 a two - body standard form , i.e. [ hamdef ] @xmath32 with a local potential @xmath33 and a semirelativistic kinetic term given by @xmath34 once the mass of the gluon , @xmath6 , as well as the radial and orbital quantum numbers @xmath35 and @xmath36 are specified ( @xmath37 for the lightest scalar glueball ) , the only unknown quantity in eq .
( [ eigen ] ) is the local central potential @xmath33 , that we will compute in the following .
it is worth underlining an important assumption that we make in this work : motivated by the standard representation of the @xmath0 glueball as a two - gluon bound state , we identify the lattice wave function @xmath27 , computed from plaquette operators , with the gluon - gluon component of the scalar glueball wave function .
this hypothesis can be intuitively understood as follows . a completely relativistic wave equation for a two - body system is provided by the well - known bethe - salpeter equation @xcite .
in this formalism , the relativistic wave function @xmath38 ( also known as the bethe - salpeter wave function ) for a bound state @xmath39 of two particles @xmath40 and @xmath41 is defined as @xcite @xmath42 in this last equation , @xmath43 is the vacuum state and @xmath44 is the creation operator for a particle @xmath45 .
moreover , @xmath46 is the spacetime coordinate of particle @xmath45 , @xmath47 , and @xmath48 and @xmath49 are the @xmath50-momentum and the spacetime coordinate of the center of mass of the system respectively . the relativistic wave function ( [ bswf1 ] ) can easily be linked to a schrdinger - like wave function .
let us indeed work in the rest frame of the bound state @xmath39 .
then , @xmath51 . in the usual nonretarded approximation @xmath52 ,
one has @xmath53 since the time coordinate @xmath54 only appears as a global phase , only the spatial separation @xmath55 , computed in the rest frame of the configuration , is now relevant .
formula ( [ bswf2 ] ) can then straightforwardly be applied to the gluon - gluon component of a glueball state @xmath56 . indeed , if @xmath57 denotes the creation operator of a gluon , the corresponding bethe - salpeter wave function at @xmath58 is given by @xcite @xmath59 where the summation on @xmath60 and the integration on the angular part @xmath61 enforce a particular spin and angular symmetry of the wave function , following the state which is considered .
equation ( [ bswf3 ] ) can thus be identified as the gluon - gluon wave function for a particular stationary glueball state , in the rest frame of the system .
interestingly , the wave function ( [ bswf3 ] ) is precisely what is computed in the lattice qcd studies of refs .
@xcite for the lightest glueball states .
the key point to perform such a calculation is the evaluation of the two - gluon operator @xmath62 , which is achieved on the lattice by computing the correlation matrix between two plaquettes at the different points @xmath63 and @xmath64 . from this discussion , we can justify the identification of the lattice wave function of ref .
@xcite with the schrdinger - like wave function of a bound state made of two gluons .
several methods exist to compute the equivalent local potential from a given wave function and its corresponding energy ( see ref .
but , these methods are only applicable to the case of nonrelativistic kinematics .
we gave in ref .
@xcite a procedure , relying on the lagrange mesh method , to make such computations with a semirelativistic kinematics of the form ( [ tdef ] ) .
we recall here the main points of this method , but refer the reader to ref .
@xcite for a detailed study .
the lagrange mesh method is a very accurate and simple numerical procedure to compute eigenvalues and eigenfunctions of a two - body schrdinger equation @xcite .
it is also applicable to a semirelativistic kinetic operator , i.e. the spinless salpeter equation @xcite . in the case of radial equations ,
a lagrange mesh is formed of @xmath65 mesh points @xmath66 which are the zeros of the laguerre polynomial @xmath67 of degree @xmath65 @xcite . the lagrange basis is then given by a set of @xmath65 regularized lagrange functions , @xmath68 satisfying the condition @xmath69 @xcite and @xmath70 .
the weights @xmath71 are linked to the mesh points @xmath66 through a gauss quadrature formula which is used to compute all the integrals over the interval @xmath72 , that is @xmath73 the regularized wavefunction , given by @xmath74 , is then developed in the lagrange basis . the semirelativistic kinetic matrix elements @xmath75 for the operator ( [ tdef ] )
can be accurately computed in this basis @xcite .
we showed in ref .
@xcite that , starting from a given regularized wave function @xmath74 , obtained in our case from eq .
( [ fit ] ) , and its corresponding energy @xmath29 , the equivalent local potential is accurately given at the mesh points by @xmath76 in the above equation , @xmath77 is a scale parameter chosen to adjust the size of the mesh to the domain of physical interest . the angular orbital momentum @xmath36 has to be a priori specified in eq .
( [ effpot1 ] ) since the matrix elements @xmath75 depend on @xmath36 @xcite . as we deal with the lightest glueball
, we assume @xmath78 and @xmath79 in order to obtain a @xmath0 state .
equation ( [ effpot1 ] ) requires the knowledge of the wave function at the mesh points @xmath80 .
so , it is more efficient to work with the trial wave function ( [ fit ] ) than with interpolated points between the few available lattice data . the first numerical parameter is the number of mesh points , @xmath65 , that we set equal to @xmath81 ( although @xmath82 already gives a good picture of the potential @xcite ) .
the second parameter is the scale parameter @xmath77 . in eq .
( [ effpot1 ] ) , we use a dimensionless variable @xmath83 , with @xmath84 .
a relevant value of @xmath77 is obtained thanks to the relation @xmath85 , where @xmath86 is the last mesh point and @xmath87 is a physical radius located in the asymptotic tail of the wave function .
this radius has to be a priori estimated , but not with a great accuracy , since the method is not variational in @xmath77 @xcite . to determine @xmath77
, we impose that @xmath88=\epsilon$ ] , with @xmath27 given by eq .
( [ fit ] ) , and @xmath89 a small number that we arbitrarily fix at @xmath90 .
this way of estimating @xmath77 has already given good results @xcite . as it can be seen from eq .
( [ fit ] ) , @xmath91=1 $ ] , and we obtain @xmath92^{1/b}.\ ] ] let us note that , strictly speaking , the effective potential @xmath33 is not unique . indeed , if @xmath93 is a state such that @xmath94 with @xmath95 , then the whole class of potentials @xmath96 satisfies the eigenequation ( [ eigen ] ) , @xmath97 being an arbitrary number .
but , these potentials are clearly nonlocal .
up to our knowledge , there is no physical indication that the interaction between two gluons could be non local .
we thus compute the unique local potential @xmath33 for the @xmath0 state that we are studying .
consequently , the only possibility is @xmath98 , for which @xmath99 .
we begin by setting @xmath100 in the kinetic operator ( [ tdef ] ) .
this corresponds to a model in which the gluons are assumed to be massless . with the scale parameter defined by eq .
( [ hdef ] ) , we can apply formula ( [ effpot1 ] ) .
results are plotted in fig .
[ fig2 ] .
the potential obtained by using the optimal values of @xmath29 , @xmath22 , and @xmath23 clearly exhibits a confining long - range part , and a rapidly decreasing short - range part .
the errors on these three parameters ( see eqs .
( [ p1 ] ) and ( [ p2 ] ) ) allow the true " potential to be located between two extremal curves .
we have checked that the curves obtained remain stable for different values of @xmath89 , @xmath65 , and @xmath77 .
these numerical results are compatible with the following cornell potential @xmath101 in this expression , @xmath102 is the strong coupling constant , that we set equal to @xmath103 as in the case of the static quark - antiquark potential @xcite .
the @xmath104 factor is the color factor coming from the one gluon exchange between two gluons when the pair is in a color singlet . in the confining part , @xmath105 is the fundamental quark - antiquark flux tube energy density .
it is usually assumed to be around @xmath106 gev@xmath107 @xcite .
the constant @xmath108 indicates the scaling of the energy density which is different for a gluon - gluon system or a quark - antiquark pair .
it is generally assumed to be given by @xmath109 ( casimir scaling ) @xcite , or by @xmath110 ( square root of casimir scaling ) @xcite .
the last constant , @xmath111 , is used to fit the height of potential ( [ corn ] ) on the numerically computed optimal potential .
the cornell potential with @xmath112 is closer to the optimal curve than the one with the casimir scaling .
however , the interaction with @xmath112 demands that @xmath113 gev .
this positive constant is somewhat surprising . indeed
, one expects that the parameter @xmath111 encodes some relativistic corrections to the cornell potential .
talking of such relativistic corrections , like spin - orbit or spin - spin terms , is meaningful even in the case of massless gluons , since these corrections are expressed in terms of powers of @xmath3 , with a rather large value for @xmath114 , the dynamical gluon mass generated by the confinement @xcite .
indeed , a computation of @xmath7 with the fitted wave function ( [ fit ] ) leads to @xmath115 gev .
since we consider here a @xmath116 state , the only remaining relativistic corrections are retardation terms and contact ( spin - spin ) interactions , which are both negative in this case @xcite .
the casimir scaling seems thus more satisfactory since it is compatible with a vanishing value of @xmath111 and it is still located in the allowed region .
one can see in fig .
[ fig2 ] that the very short - range part of the numerically computed potential is less singular than the coulomb potential .
this part is actually very sensitive to the short - range behavior of the wave function .
it has been shown that the wave function of a true semirelativistic coulomb problem diverges in @xmath117 @xcite , but such a divergence can not be computed in lattice calculations . anyway
, more points should be necessary to elucidate this short - range behavior .
let us note that the ground state masses of the hamiltonian ( [ hamdef ] ) with the two fitted potentials ( [ corn ] ) are respectively @xmath118 gev and @xmath119 gev with @xmath120 and @xmath112 .
these values are contained in the error bars of @xmath29 .
the corresponding wave function with @xmath112 is nearly indistinguishable of the optimal wave function ( [ fit ] ) , while the one for @xmath120 decreases more quickly .
this is coherent with fig .
[ fig2 ] , where we can see that the cornell potential with the casimir scaling increases faster than the optimal curve , consequently leading to a faster decreasing wave function .
as mentioned in the introduction , it is argued in many works that the constituent gluons should have a fixed nonzero mass , typically around @xmath4 gev @xcite .
the effective gluon - gluon potential can also be computed by using our method . in order to clearly see the changes between @xmath100 and @xmath121 ,
we have computed the effective potential for a rather large but still relevant value of @xmath6 , namely @xmath1 gev .
the result is plotted in fig .
[ fig3 ] . a comparison with fig .
[ fig2 ] shows that the asymptotic behavior of @xmath33 drastically changes when @xmath121 . instead of a monotonically increasing potential
, it now seems to saturate at some energy scale .
such a saturation suggests an interpretation in terms of string breaking effects , which have been measured in lattice qcd in the case of a static quark - antiquark pair .
thus , the optimal potential of fig . [ fig3 ] should no longer be compared to a cornell form .
however , in a pioneering work about glueballs as bound states of massive gluons by cornwall and soni @xcite , it is suggested that the gluon - gluon potential at lowest order reads @xmath122 where @xmath123 is the spin of the glueball .
this potential clearly saturates at a value @xmath124 , then forbidding bound states with a mass greater than @xmath125 . for @xmath126 ,
the confining part reduces to @xmath127 , and a linearly rising confining potential is recovered . by comparison with the cornell potential ( [ corn ] ) , one could expect that @xmath128 which provides us with a definition of @xmath129 in terms the more intuitive parameters @xmath6 and @xmath105 .
some useful remarks can also be made about the short - range part of the potential ( [ cornw ] ) , i.e the part which is proportional to @xmath130 .
firstly , it involves a yukawa potential instead of a coulomb one , because both exchanged gluons and constituent gluons are massive .
moreover , it is spin - dependent even at the lowest order .
this is a characteristic feature of effective potential with massive gluons .
other spin - dependent corrections in @xmath5 also exist , but we do not mention them since they are absorbed in the constant @xmath111 by definition .
again , this constant is fitted so that the absolute height of the optimal potential is recovered .
it is readily observed in fig .
[ fig3 ] that potential ( [ cornw ] ) fits rather well the numerically computed potential , in particular the long - range behavior , for @xmath131 gev@xmath21 . for comparison
, we notice that @xmath132 gev@xmath21 when @xmath120 and @xmath133 gev@xmath21 when @xmath112 .
but , as we argued in the previous section , the short - range part is only poorly known with the current lattice data .
the difference between @xmath120 and @xmath110 is less important than for the cornell potential even if the casimir scaling seems closer to the optimal curve when the usual value @xmath134 gev@xmath107 is used .
it is worth mentioning that we fitted @xmath135 , which is higher than in the case of massless gluons .
but , such a value is commonly used in models with massive gluons @xcite .
we finally note that the ground state masses of the hamiltonian ( [ hamdef ] ) with the two fitted potentials ( [ cornw ] ) are respectively @xmath136 gev and @xmath137 gev with @xmath120 and @xmath112 .
again , these values are contained in the error bars of @xmath29 . in this case , the corresponding wave functions with both values of @xmath108 are very similar to each other , but with a slightly larger spatial extension than the optimal wave function ( [ fit ] ) . a plot of the effective potential for different values of @xmath6 is given in fig .
[ fig4 ] in order to see more clearly the evolution of the potential shape with an increasing gluon mass .
it is readily observed that there is a transition between a linearly rising regime at zero mass and a saturating one at high gluon mass .
as we already pointed out , the short - range behavior of the computed potentials is rather poorly known because of the current precision of the lattice data .
however , interesting informations can be deduced concerning the long - range part of the potential , typically for @xmath138 gev@xmath21 .
one can find indeed that the different potentials computed in fig .
[ fig4 ] can be correctly described at large @xmath9 by the following form @xmath139 with @xmath140 a fitted parameter .
such a form has the advantage of having a non trivial limit for @xmath100 , which is simply the linearly rising potential @xmath141 . when @xmath121 however , it saturates at the value @xmath142 .
potential ( [ vfit ] ) has been plotted for the different values of @xmath6 that we considered , with the previously found values of @xmath105 , that is @xmath106 gev@xmath107 .
we fixed @xmath112 because we already saw in sec .
[ massless ] that the optimal numerically computed potential was better reproduced with that value of @xmath108 .
but , again , we stress that the precision on the potential does not allow to decide whose scaling law is the best one .
the constant @xmath111 have been fitted for each @xmath6 in order for the potential ( [ vfit ] ) to have the right absolute height . in the previous section
, we showed that the potential obtained with @xmath143 gev can be well reproduced by the form ( [ cornw ] ) , whose long - range part is a priori inequivalent to the potential ( [ vfit ] ) .
actually , both expressions coincide if @xmath144 with @xmath134 gev@xmath107 , this relation states that potentials ( [ cornw ] ) and ( [ vfit ] ) are identical for @xmath145 gev when @xmath112 , and @xmath146 gev for @xmath120 .
remarkably , this corresponds to a gluon mass around @xmath1 gev , that is the case that we treated previously .
in this work , we have shown how to compute the effective potential between two gluons from the mass and wave function of the @xmath0 state obtained in lattice qcd , reasonably assuming that the lattice wave function mainly gives the gluon - gluon part of the glueball schrdinger - like wave function .
this method , which is here used for the first time , has the advantage of allowing to deal with semirelativistic kinematics , which is necessary to include massless gluons systems in the discussion .
the relevance of potential models to describe gluons is still a matter of controversy nowadays . in particular
, the mass of a constituent gluon is an open problem : should it be zero or not ?
arguments favoring both hypothesis can be found in the literature , so we computed the effective gluon - gluon potential in both cases .
when the constituent gluons are massless , the potential we find is compatible with a cornell one for standard values of the parameters .
when the constituent gluons are massive however , the effective potential is merely compatible with a saturating one with a yukawa - type short - range part .
these results go beyond the usual computation of the static potential between two static sources in the adjoint representation . in this work indeed , we started from pure gauge results in lattice qcd and we showed that the constituent gluon picture naturally emerges from these lattice data .
this could yield an a posteriori justification of the success of potential models in the description of the glueball spectrum .
moreover , we showed that the allowed shape of the potential is constrained by the mass of the constituent gluon ( see eq .
( [ vfit ] ) ) : the linear confinement or the saturating one are only valid for massless or massive gluons respectively .
although the casimir scaling seems favored , the current precision of the lattice data does not allow to decide which scaling law is correct .
this problem could be solved by making high precision computations of the wave function .
we mention however the recent work of ref .
@xcite , where the behavior of a quark - antiquark - constituent gluon system is investigated on the lattice , with results in agreement with the casimir scaling hypothesis . an important application of our method is that , provided accurate wave functions for glueballs with a higher total spin are computed ( too few points are currently available for the @xmath147 wave function @xcite ) , spin - dependent terms vanishing for the scalar glueball could be studied .
this would be a check of the relevance of approaches with massless or massive constituent gluons .
indeed , potentials coming from both models already differ at the lowest order for what concerns spin - dependent terms .
we consequently think that a very accurate determination of the wave functions of the lightest @xmath0 and @xmath147 glueballs in lattice qcd , especially at short range , could serve as a check to determine the most relevant potential approach of glueballs .
we expect that , provided these wave functions are known , our method should lead to numerically computed potentials with a common long - range behavior confinement is indeed not supposed to depend on @xmath148 . on the contrary , the short - range behavior should be different , and give important informations about the relativistic corrections .
we hope to present such a study in future works .
h. b. meyer , hep - lat/0508002 ; w. ochs , hep - ph/0609207 . c. e. carlson , t. h. hansson , and c. peterson , phys .
d * 27 * , 1556 ( 1983 ) .
a. szczepaniak , e. s. swanson , c .- r .
ji , and s. r. cotanch , phys .
lett . * 76 * , 2011 ( 1996 ) [ hep - ph/9511422 ] ; a. szczepaniak and e. s. swanson , phys .
b * 577 * , 61 ( 2003 ) [ hep - ph/0308268 ] .
h. fritzsch and p. minkowski , nuovo cimento * 30 * a , 393 ( 1975 ) ; t. barnes , z. phys .
c * 10 * , 275 ( 1981 ) .
j. m. cornwall and a. soni , phys .
b * 120 * , 431 ( 1983 ) .
a. b. kaidalov and yu .
a. simonov , phys .
b * 636 * , 101 ( 2006 ) [ hep - ph/0512151 ] .
f. buisseret , arxiv:0705.0916 .
f. brau and c. semay , phys .
d * 70 * , 014017 ( 2004 ) [ hep - ph/0412173 ] , and references therein ; v. mathieu , c. semay , and b. silvestre - brac , phys .
d * 74 * , 054002 ( 2006 ) [ hep - ph/0605205 ] .
f. buisseret and c. semay , phys .
d * 73 * , 114011 ( 2006 ) [ hep - ph/0511270 ] .
f. j. llanes - estrada , p. bicudo , and s. r. cotanch , phys .
lett . * 96 * , 081601 ( 2006 ) [ hep - ph/0507205 ] .
f. j. llanes - estrada and s. r. cotanch , phys .
lett . * 84 * , 1102 ( 2000 ) [ hep - ph/9906359 ] .
hou and a. soni , phys .
d * 29 * , 101 ( 1984 ) ; w .- s .
hou , c .- s .
luo , and g .-
wong , phys .
d * 64 * , 014028 ( 2001 ) [ hep - ph/0101146 ] ; f. brau and c. semay , phys .
d * 72 * , 078501 ( 2005 ) [ hep - ph/0411108 ] .
a. c. aguilar , a. mihara , and a. a. natale , phys .
d * 65 * , 054011 ( 2002 ) [ hep - ph/0109223 ] .
n. brambilla , p. consoli , and g. m. prosperi , phys .
d * 50 * , 5878 ( 1994 ) .
g. s. bali , phys .
343 * , 1 ( 2001 ) [ hep - ph/0001312 ] .
a. simonov , hep - ph/9911237 ; a. b. kaidalov and yu .
a. simonov , phys .
b * 477 * , 163 ( 2000 ) [ hep - ph/9912434 ] .
g. s. bali _
et al . _ ,
d * 71 * , 114513 ( 2005 ) [ hep - lat/0505012 ] .
j. vijande , p. gonzalez , h. garcilazo , and a. valcarce , phys .
d * 69 * , 074019 ( 2004 ) [ hep - ph/0312165 ] ; p. gonzalez , a. valcarce , j. vijande , and h. garcilazo , int .
a * 20 * , 1842 ( 2005 ) [ hep - ph/0409202 ] . c. j. morningstar and m. peardon , phys .
d * 60 * , 034509 ( 1999 ) [ hep - lat/9901004 ] ; y. chen _ et al .
d * 73 * , 014516 ( 2006 ) [ hep - lat/0510074 ] .
p. de forcrand and k
liu , phys .
* 69 * , 245 ( 1992 ) . m. loan and y. ying , prog .
phys . * 116 * , 169 ( 2006 ) [ hep - lat/0603030 ] .
f. buisseret and c. semay , phys .
e * 75 * , 026705 ( 2007 ) [ physics/0611146 ] .
e. e. salpeter and h. a. bethe , phys . rev .
* 84 * , 1232 ( 1951 ) ; j. schwinger , proc .
* 37 * , 455 ( 1951 ) .
g. c. wick , phys . rev . *
96 * , 1124 ( 1954 ) ; r. cutkosky , phys . rev . * 96 * , 1135 ( 1954 ) . g. b. west , phys . rev
. lett . * 77 * , 2622 ( 1996 ) [ hep - ph/9603316 ] .
a. lovell and k. amos , phys .
c * 62 * , 064614 ( 2000 ) [ nucl - th/0007031 ] , and references therein .
d. baye and p .- h .
heenen , j. phys . a * 19 * , 2041 ( 1986 ) .
m. vincke , l. malegat , and d. baye , j. phys .
b * 26 * , 811 ( 1993 ) ; d. baye , j. phys .
b * 28 * , 4399 ( 1995 ) . c. semay , d. baye , m. hesse , and b. silvestre - brac , phys
e * 64 * , 016703 ( 2001 ) .
g. s. bali , phys .
d * 62 * , 114503 ( 2000 ) [ hep - lat/0006022 ] ; s. kratochvila and p. de forcrand , nucl .
b * 671 * , 103 ( 2003 ) [ hep - lat/0306011 ] ; c. semay , eur .
j. a * 22 * , 353 ( 2004 ) [ hep - ph/0409105 ] .
k. johnson and c. b. thorn , phys .
d * 13 * , 1934 ( 1976 ) ; g. martens , c. greiner , s. leupold , and u. mosel , phys .
d * 70 * , 116010 ( 2004 ) .
a. b. kaidalov , yu .
a. simonov , phys .
b * 477 * , 163 ( 2000 ) [ hep - ph/9912434 ] . v. mathieu and f. buisseret , hep - ph/0702226 .
b. durand and l. durand , phys .
d * 28 * , 396 ( 1983 ) ; * 50 * , 6662(e ) ( 1994 ) ; p. castorina , p. cea , g. nardulli , and g. paiano , phys .
d * 29 * , 2660 ( 1984 ) .
p. bicudo , m. cardoso , and o. oliveira , hep - ph/0704.2156 . | starting from the @xmath0 glueball mass and wave function computed from lattice qcd , we compute the local potential between two constituent gluons . since the properties of constituent gluons are still a matter of research , we allow for them to be either massless , or massive with a mass around @xmath1 gev .
both pictures are actually used in the literature .
when the gluons are massless , the corresponding local potential is shown to be compatible with a cornell form , that is a linear confinement plus a short - range coulomb part , with standard values for the flux tube energy density and for the strong coupling constant . when the gluons are massive , the confining potential is a saturating one , commonly used to simulate string - breaking effects .
these results fill a gap between lattice qcd and phenomenological models : the picture of the scalar glueball as a bound state of two constituent gluons interacting via a phenomenological potential is shown to emerge from pure gauge lattice qcd computations .
moreover , we show that the allowed potential shape is constrained by the mass of the constituent gluons . |
the stellar content and evolutionary history of blue compact dwarf ( bcd ) galaxies have been a puzzle for some time .
the low metal abundances derived from the h - ii regions of bcds ( between about z@xmath2/3 and z@xmath2/50 , thuan et al .
1994 ) , combined with their high rates of star formation and large h - i masses ( thuan & martin 1981 ) , have raised the question of whether bcds are truly young galaxies which formed their first stars recently , or old galaxies which occasionally light up with bright , compact starburst regions ( searle & sargent 1972 , searle et al .
1973 , izotov & thuan 1999 ) . in their original definition of a young " galaxy , searle & sargent specified
this is a galaxy which formed _ most _ of its stars in recent times . in subsequent years
the argument over bcd ages has evolved into a more polarized one , because it is important to understand whether or not there might exist any local galaxies which are only now forming their very first generation of stars out of pristine gas .
such bcds that are making _ all _ of their stars in the on - going starburst could be considered to be local examples of primeval galaxies ( thuan & izotov 1998 ) and they would be fundamentally different from old galaxies which began forming their first stars over 10 gyrs ago at high redshifts .
several lines of evidence support the old - galaxy hypothesis .
thuan ( 1983 ) found indications for the presence of evolved stars from the integrated , near - ir colors of bcds .
fanelli et al . ( 1988 ) used uv spectra to show that the star - formation rates of bcds are discontinuous .
a deep ccd imaging survey by loose & thuan ( 1986 ) first revealed that the bright , compact star - forming regions of bcds are embedded in much fainter , redder halos with elliptical outer isophotes .
nearly all galaxies in their sample ( @xmath095% ) show such an underlying , low - surface - brightness component .
subsequent ccd surveys have confirmed this result ( e.g. , kunth et al .
1988 , papaderos et al .
1996 , telles et al .
1997 , meurer 1999 ) .
evolutionary synthesis models ( e.g. , krger et al .
1991 ) and chemodynamical models ( e.g. , rieschick & hensler 1998 ) favor the old - galaxy view as well . hubble space telescope ( hst ) observations of several extremely metal - poor bcds recently led thuan at al . ( 1997 ) and thuan & izotov ( 1998 ) to propose that primeval bcds may yet be found locally .
their argument is based in part on the blue colors observed for the unresolved underlying disks .
this is seen in i zw 18 ( z@xmath2/50 ) by hunter & thronson ( 1995 ) , in sbs 0335 - 052 ( z@xmath2/41 ) and sbs 0335 - 052w ( around z@xmath2/50 ) by thuan at al .
( 1997 ) , papaderos et al .
( 1998 ) and lipovetsky et al .
( 1999 ) , and in sbs 1415 + 437 ( z@xmath2/21 ) by thuan et al .
based on abundance measurements , izotov & thuan ( 1999 ) suggest that in fact _ all galaxies with z @xmath3 z@xmath2/20 are young _ , with ages not exceeding 40 myr , while those with z @xmath3 z@xmath2/5 are no older than @xmath4 1 - 2 gyr . in the definition of izotov & thuan then , a young " galaxy is a galaxy which has not experienced any star - forming events prior to the current one ( which might have enriched its gas beyond that which is observed ) . by most accounts ,
even galaxies with ages @xmath1 2 gyr would be considered to be relatively young galaxies , as their formation redshift would not have been co - eval with that of the large or other dwarf galaxies .
the controversy over the age of bcds reflects our current knowledge of galaxy formation and evolution .
did dwarf galaxies form from primordial density fluctuations at z @xmath5 5 ( ikeuchi & norman 1987 ) ? are the dwarfs observed today the leftover building blocks of large galaxies ( white & rees 1978 , dekel & silk 1986 ) ? is the formation of dwarfs delayed until z @xmath4 1 when the uv background cooled sufficiently for gas to collapse within small dark matter halos and form stars ( blanchard et al .
1992 , babul & rees 1992 ) ?
are dwarfs rapidly evolving for z @xmath1 1 , in which case they could account for the rapid evolution in the galaxy luminosity function necessary to explain the faint blue excess ( broadhurst et al . 1988 , babul & ferguson 1996 , spaans & norman 1997 , ferguson & babul 1998 , guzmn et al .
where are the local , isolated h - i clouds still awaiting their first generation of stars ( briggs 1997 ) ? resolved stellar populations are the fossil record of a galaxy s star - formation history ( sfh ) .
they can also be used to provide a definition of a galaxy s age .
we shall consider young " a galaxy which contains only young " stars , stars with ages @xmath1 100 myr .
intermediate - age stars signal a galaxy is at least of intermediate age ; we consider this to be the case when stars that are older than a few hundred myr are detected .
since the time - resolution of color - magnitude diagrams ( cmd ) of distant galaxies rapidly decreases if the stellar ages exceed 1 gyr , a detection of stars with ages of about 1 gyr is often considered sufficient to show the presence of old stars .
we , however , prefer to consider all stars with ages of up to 10 gyr as intermediate - age stars , and reserve the term old " for stars of the kind that inhabit galactic globular clusters and have ages @xmath0 10 gyr .
an old " galaxy is therefore one that has formed at least some stars more than 10 gyrs ago .
the dwarf galaxies in the local group exhibit an astounding variety of sfhs , yet all contain old stars ( mateo 1998 ) in the above sense .
however , there are no bcds known in the local group .
vii zw 403 is hence a key object ; it is so close that hst observations resolve it into individual stars ( schulte - ladbeck et al .
1998 , hereafter sch98 , lynds et al . 1998 ) .
thus it has become possible for the first time to derive the distance of a bcd using a stellar distance indicator ( rather than just a recession velocity ) .
this is crucial for the interpretation of the stellar content .
recent determinations of the present - day metallicity of the ionized gas in vii zw 403 ( or rather , its o / h ratio ) , were published by martin ( 1997 ) and izotov , thuan & lipovetsky ( 1997 ) .
martin finds log ( o / h ) = -4.42(0.06 ) , and izotov , thuan & lipovetsky give -4.31(0.01 ) .
both groups employ the metallicity scale for which the sun has log ( o / h ) = -3.07 .
we note that izotov & thuan ( 1999 ) also use this solar value . on this scale ,
the metallicity of vii zw 403 is between 1/22 and 1/17 of solar . despite its low metallicity
( @xmath4 z@xmath2/20 ) sch98 argued that vii zw 403 is not a young galaxy , based on the detection of a red giant branch ( rgb ) with a well defined tip and a red asymptotic giant branch ( agb ) .
vii zw 403 also exhibits extended outer isophotes ( loose & thuan 1986 , hopp & schulte - ladbeck 1995 ) with a red color that is consistent with an old , metal poor population ( schulte - ladbeck & hopp 1998 ) .
the spectrum of the background sheet displays absorption lines that indicate an evolved population underlying the recent starburst ( hopp et al .
1998 ) . in this paper ,
we provide further arguments that the existence of an early epoch of star - formation may be gleaned from the morphology of vii zw 403 .
we demonstrate that in radial bins outward from its starburst center , the contribution to the cmd by young stars decreases , while the red tangle that contains the old and metal - poor rgb ( aparicio & gallart 1994 ) becomes a narrow feature . for radii larger than about 1 kpc ,
the young stars are absent and the stellar content is well described by a globular - cluster - like stellar population .
we argue that this resolved old stellar population supports earlier suggestions that the faint halos of bcds harbor old stars .
hst wfpc2 observations of vii zw 403 were obtained in the continuum ( f336w , f555w and f814w filters , approximately the johnson - cousins u , v , and i bands ) and in the h@xmath6 emission line ( f656n filter ) .
the relevant exposures are listed in table 1 of sch98 .
as described in sch98 , we re - ran the pipeline with improved calibration files , removed cosmic rays , corrected for cte , and corrected for geometric distortion .
we conducted photometry on h@xmath6-subtracted images .
we used daophot to perform psf photometry on each of the four wfpc2 chips .
we calibrated the photometry using the most up - to - date synphot tables . after determining the positions of each point source , we merged the object catalogs of the four chips into a single file of positions and instrumental magnitudes in the vega magnitude system .
the accuracy of these coordinates relative to the guide star system is about 0.5 " . in figures 1a , b
, we provide the errors for the point - source measurements in the continuum filters , as well as the results of our completeness tests .
daophot / allstar residuals in f555w and f814w filters on all chips can be summarized by stating that they reach 0.1 mag at magnitudes of about 26 and 25.5 , respectively .
we checked the completeness of our photometry for each chip by adding a distribution of false stars consistent with the magnitude distribution of the real stars .
the percentage of recovered stars indicates that completeness is about 50% at m@xmath7 = 26 , m@xmath8 = 25 .
4 of sch98 indicates there are holes " in the distribution of red stars on the pc due to incompleteness effects ; this presumably also produces a less well populated red tangle in the first two panels of fig . 3 below .
) the foreground galactic reddening for vii zw 403 is e(b - v ) = 0.025 ( burstein & heiles 1984 ) ; we correct for the corresponding extinction using the tables provided in holtzman et al .
( 1995 , tables 12a , b ) . a small and patchy internal reddening ( e(b - v)@xmath10.16 ) can not be ruled out ( lynds et al .
1998 ) , but is not central to the arguments presented below ( the location of the blue plume in the cmd is consistent with no or low internal reddening for the majority of the young stars detected , the old stars in baade s red sheet " are measured throughout the halo where we find no evidence for internal extinction ) .
final colors are transformed u , v , and i magnitudes using the color terms in holtzman et al .
( 1995 , table 7 ) .
figure 2 is a position plot of all stars detected in both v and i ( 5459 objects ) .
we used the distribution of sources detected in both u and v to locate the center of star fomation in vii zw 403 ( r.a.[@xmath9 = 172.002 , dec.[@xmath9 = 78.994 , j2000 ) .
we then cast concentric circles about this location . in figure 3 , we display the [ ( v - i)@xmath10 , i@xmath10 ] cmds observed within each of six radial bins marked in fig . 2 . the cmds of the first and sixth bins are also superimposed with stellar evolutionary tracks .
these tracks use the stellar atmospheres of f. castelli ( see , e.g. , bessell et al .
1998 ) for a metallicity of z@xmath2/30 , which were folded with the hst filter / system response and kindly made available to us by l. origlia .
stellar evolution is based on the fagotto et al .
( 1994 ) isochrones for z=0.0004 ( z@xmath2/50 ) .
these models were chosen because they most closely represent a compromise " metallicity for the stellar populations which we observe ( i.e. , without trying to model the enrichment history , on which we comment below , in detail as well ) .
as is always the case , the comparison of theoretical tracks ( or synthetic cmds ) with data depends on the accuracy with which we can ascertain either one and thus map one onto the other .
we will consider as safe , conclusions which do not depend on the choice of a particular model .
for instance , a well populated rgb is usually only observed if a stellar population with an age upwards of about 1 gyr is present , irrespective of the stellar metallicities adopted ( cf . also , sweigart , greggio & renzini 1990 ) .
we will point out where appropriate , when inferences are based on what we consider more uncertain model results .
the location of the tracks on the observed cmds employs the distance modulus derived below .
the changes in stellar content with position are quite striking .
the center displays a cmd that has a dominant blue plume of main - sequence ( ms ) stars and blue supergiants ( bsg ) or blue - loop ( bl ) stars , a prominent red supergiant ( rsg ) plume , a few , very red agb stars , and a weak red tangle . in the second bin , this red tangle and the region of intermediate - mass bl stars are much more populated . we note that h@xmath6 emission is detected only in the first and second bins . by the third bin ,
both the blue plume and the rsg plume have weakened whereas the red tangle is strong . by the fourth bin ,
most stellar indicators of young ages have disappeared .
our detection limit of v@xmath10 @xmath4 28 is low enough to show that ms stars younger than about 200 myr are absent .
the few remaining bl stars , which suggest ages of a few hundred myr , are still present in this bin . in the fifth and sixth bins , we see the outer regions of the galaxy . here
, the red tangle has become a narrow band .
there are seven objects that are red and brighter than i@xmath10=23 .
a likely explanation is that these are galactic foreground stars .
the number of galactic foreground stars within the wfpc2 images is expected to be very small ( mndez & guzmn 1998 ) ; since the fifth and sixth bins cover the largest area they may contain a few .
the cmds of the fifth and sixth bins suggest that only intermediate - age and old agb and rgb stars are found in the outer regions of vii zw 403 .
in sch98 we used the tip - of - the - red - giant - branch ( trgb ) method to obtain the distance to vii zw 403 . using the red tangle in the star - forming region
, we estimated a metallicity [ fe / h ] of -1.2 , resulting in m@xmath12=-4.10 for the trgb and a distance modulus of 28.4 mag ( with a total random error of 0.09 , and a dominant systematic error of 0.18 ) .
this corresponds to a distance of 4.8 mpc .
we now re - derive the distance using the cmd of the outer bins , only .
the new i@xmath13 , [email protected] , is not significantly different from our old value . following lee et al .
( 1993 ) , we can estimate [ fe / h ] from the v - i color either just below the trgb , at m@xmath12=-3.5 , or better , at m@xmath12=-3.0 .
we have done both and find a consistent ( v - i)@xmath10 color of 1.28 , with an rms error of 0.03 ( for m@xmath12=-3.5 ) and a dispersion of 0.21 .
this translates into a mean metallicity of @xmath1[fe / h]@[email protected] ( or z=0.00024 or z@xmath2/83 ) with a spread or metallicity range ( corrected for measurement error ) of @xmath140.7 .
we now find the trgb at m@xmath12=-3.98 , or m - m=28.23 , or a distance of about 4.4 mpc , in excellent agreement with the value of 4.5 mpc derived by lynds et al .
( 1998 ) . at this distance ,
1 `` corresponds to about 21.5 pc .
the holmberg diameter of vii zw 403 was measured by schulte - ladbeck & hopp ( 1998 ) as 2a@xmath15=146 '' , and translates into a physical diameter of 3.1 kpc .
3 shows that the young stars are contained within the inner few hundred pc , and are absent beyond a radius of 1 kpc .
not surprisingly , this agrees with the typical " size of a bcd ( thuan & martin 1981 ) . in fig . 2 we colored in green the two radii which clearly separate the young and old stellar components of vii zw 403 .
in fig . 4 , we display the cmds inside of the inner ( 386 pc ) and outside of the outer ( 1352 pc ) radius in blue and red , respectively .
this illustrates the change in width of the red tangle due to a diminishing contribution by intermediate - mass stars .
we mark the position of the trgb , and provide an absolute - magnitude scale .
we also overlay the empirical globular - cluster ridge - lines of da costa & armandroff ( 1990 ) , which indicate that the halo population of vii zw 403 is similar to the population of galactic globular clusters , the prototypes of population ii .
4 shows that agb stars , which populate a strip from [ ( v - i)@xmath10 , m@xmath12@xmath10]@xmath4[1 , -4 ] to [ 3.5 , -6.5 ] are found both at small and large radii .
the strip may contain stars as old as 10 gyr and z=0.0004 ( z@xmath2/50 ) at its bright , blue end .
the extent of the agb to the red requires 1/5@xmath0z@xmath2@xmath01/20 ; here the average location of stars is well described by the 4 gyr isochrones ( bertelli et al .
note that theoretical models of the agb phase are very uncertain .
this is because a ) their atmospheres contain difficult - to - model molecules , and b ) mass - loss is an important but ill - known parameter that determines their evolution .
if the agb models can be believed , then vii zw 403 has had an interesting chemical history .
schulte - ladbeck & hopp ( 1998 ) showed that their surface - brightness profiles of vii zw 403 could be fit with an exponential law , with scale lenghts in b of 25.7 `` , r of 25.2 '' . in fig . 5
, we display the i - band surface density of resolved stars versus radius , derived from starcounts in the complete quadrant marked in fig . 2 by blue lines .
an exponential law describes well the distribution of the resolved stars outside of the innermost region where there is incompleteness .
the scale length for all stars in i is 25.6 `` ( or about 550 pc ) , in excellent agreement with our ground - based results . for the red stars ( 0.6@xmath1v - [email protected] )
only , the scale length is 31.8 '' ( or about 680 pc ) .
owing to our observation that the cmd of the halo of vii zw 403 is very reminiscent of that of a dwarf spheroidal galaxy ( dsph ) , and in connection with the interesting question of what is the nature of the faint blue galaxies , we compare the scale length of vii zw 403 to that of dsphs .
we find that the scale length is about 1.5 times larger than those of the largest local group dsph like ngc 147 , ngc 185 , ngc 205 , or fornax ( mateo 1998 ) .
we may ask how typical is this scale length for bcds in general ?
one note of caution other bcds do not have direct distance determinations , and the recession velocities in the literature are scaled with ( different values of ) ho , resulting in fairly unreliable scale lengths .
nevertheless , the distance errors might average out over a large sample of scale lengths .
we compared the scale length of vii zw 403 to values derived for the bcd - dominated samples of bothun et al .
( 1989 ) and telles et al .
( 1997 ) . in both papers ,
deep ccd exposures to relatively large angular diameters were used for detailed surface photometry and exponential scale lengths were derived .
the central starburst contribution was ignored in the determination .
bothun et al .
used r - band data , telles et al .
, v - band observations .
we transformed both data sets to the same distance scale with ho = 75 km / s / mpc . in the bothun et al .
sample , 24% of the objects have scale lengths of 0.7 kpc and smaller .
telles et al . found 38% of their objects in this regime .
surface photometry of 93 galaxies from the emission line galaxy sample of popescu et al .
( 1997 ) was performed by vennik , hopp & popescu ( 1999 ) . again , this sample is dominated by bcds , but contains a few galaxies of other types .
the scale lengths are derived from r images , and 44% of the galaxies show scale lenghts of 0.7 kpc and smaller .
papaderos et al .
( 1996 ) studied a sample of 14 galaxies in much more detail than the other authors . here
, the derived r - band scale lengths range from 0.17 to 2.3 kpc . despite the various and different selection effects of the four samples from the literature , we conclude from this comparison that the scale length found for the underlying galaxy in vii zw 403
is fairly typical for that of bcds .
several pieces of evidence now suggest that vii zw 403 has an old stellar halo , and this has implications for the formation of bcds . while it has previously been assumed that red colors at large radii point to the presence of evolved stellar background sheets in bcds , the rsg , agb and rgb stars actually overlap over a wide range in effective temperatures and hence colors ( cf . fig .
therefore , a red background sheet can not be considered as evidence of an old , underlying host galaxy unless it can be proven that the stellar population which dominates the integrated color of a bcd is an old one .
this has now been accomplished with the detection of red giants in the halo of vii zw 403 .
however , owing to the well - known age - metallicity degeneracy , comparatively young ( @xmath4 1 gyr ) and metal rich rgb stars can populate the same region of a cmd as old ( 10 - 15 gyr ) and metal poor rgb stars , based on optical , broad - band colors .
we interpret the color dispersion of the rgb with a variation in chemical abundance ; however , age dispersion ( and/or differential reddening ) can also broaden the rgb ( e.g. , bertelli et al . 1994 ) . in the case of vii
zw 403 we argue that the well - defined rgb tip , and the narrowing of the red tangle from the center toward the outer regions of the galaxy where young stars are absent , indicate that our interpretation is appropriate
. conclusive evidence could be obtained from the detection of old horizontal - branch stars or the turn - off of an old ms , but these are too faint to be reached with the hst .
vii zw 403 does not appear to have a globular - cluster system .
recently , the stellar halos of several dwarf irregular ( dirr ) and transition dirr / dsph galaxies in the local group have been resolved into stars ( wlm , minniti & zijlstra 1996 , ngc 3109 , minniti et al .
1999 , antlia , aparicio et al .
these observations have been interpreted to indicate the presence of old and metal - poor , population ii halos .
vii zw 403 joins the ranks of an increasing number of local , star - forming dwarf galaxies with resolved old stellar halos .
as is pointed out by mateo ( 1998 ) , all suitably studied star - forming dwarfs of the local group show evidence for extended , smooth , and symmetric distributions for their older stars .
most bcds show extended , smooth and red outer isophotes as well . our results for vii zw 403
imply the identification of the background sheets with old stellar halos is justified for these bcds . the controversy over the age of bcds
whether they are young or old galaxies may be resolved in the following way : there is a continuum of star - formation and chemical - enrichment histories among the bcds .
it is likely that the vast majority of bcds have an ancient ( @xmath0 10 gyr ) stellar population substratum , and must thus be recognized as old galaxies .
this is based on the observation that over 95% of bcds in the loose & thuan sample show extended background sheets of red color . comparing observations of the background - sheet colors for a sample of bcds and dirrs with the population synthesis models of schmidt et al .
( 1985 ) , schulte - ladbeck & hopp ( 1998 ) suggest that complex star - formation histories prevail in these galaxies .
thus , depending on just how much mass is involved in the on - going starburst , and how the morphology of the star - forming regions compares to that of the older stellar substratum , the colors of some bcds might be entirely consistent with those of young " galaxies in the sense that they are currently experiencing a strong starburst .
if the outer isophotes are sufficiently red in color , they can be interpreted to indicate that such bcds formed some stars at epochs similar to that of galactic globular cluster formation ( kunth et al .
1988 , papaderos et al .
1996 , telles et al .
1997 , schulte - ladbeck & hopp 1998 , meurer 1999 ) .
the range of background - sheet colors suggests that the sfhs of most bcds since their formation at high redshift have probably been diverse , depending on the frequency , duration , and intensity of the star - formation events .
this is similar to what has been derived for dwarf galaxies within our local group . in a few of the extremely metal - poor bcds , the color gradients are small and the outer isophotes remain fairly blue ( hunter & thronson 1995 , thuan at al .
1997 , papaderos et al .
1998 , lipovetsky et al . 1999 ,
thuan et al .
the argument that such _
blue _ background - sheet colors indicate a galaxy is only now making its first generation of stars seems untenable to us in the light of population synthesis models . where the young and old stellar populations are spatially co - existent , schmidt et al .
( 1985 ) show that a starburst completely dominates integrated optical colors for up to 50 myr , even if as little as 0.1% of the dwarf galaxy s mass is involved ; hence a young burst may render the underlying old population undetectable .
it is therefore possible ( although in the absence of deep cmds not yet demonstrated ) that the old populations in these few extremely metal - poor bcds elude us due to a contrast problem . on the other hand ,
the cmds of the few bcds which have been resolved with hst into single stars indicate they do contain stellar generations which predate the present starburst . in figure 6 , we show the evolution of vii zw 403 in a series of synthetic snapshot " cmds .
these synthetic cmds were computed with the above z@xmath2/50 evolutionary tracks .
we used the bologna code , which was recently adapted by greggio et al .
( 1998 ) for the simulation of hst data .
the synthetic cmds help to illustrate the general features of stellar evolution at low metallicity , and may also serve as templates for future cmds of extremely metal - poor bcds .
the panels of fig . 6 show very well that the observed changes in stellar population with radius ( fig .
3 ) can be interpreted as a change of the stellar ages with distance from the core .
we use synthetic cmds to place a lower limit on the age of the red giants that we see in the halo of vii zw 403 .
we find a well defined trgb first appears for ages @xmath0 3 gyr ( and of course continues to be present up to 15 gyrs ) . beyond an age of about 3 gyr
, we lose age resolution in the cmd ; and we can not constrain , from the location of the rgb alone , the presence of stars with ages in excess of 10 gyr .
clearly , vii zw 403 has had a rich history of star formation .
aloisi et al .
( 1999 ) recently used synthetic cmds to investigate deep hst cmds of i zw 18 , the most metal - poor bcd known ( z@xmath2/50 , from the ionized gas ) .
they find that the present burst is not the first one to occur in this galaxy either ; the data require a prior burst 500 myr to 1 gyr ago .
while by most accounts a galaxy with an age below 1 gyr would be considered a young galaxy , these results for i zw 18 are in contradiction with the primeval galaxy hypothesis of izotov & thuan ( 1999 ) based on abundance analyses . unfortunately , there are no extremely metal - poor bcds known that are closeby enough to allow for look - back times of a large fraction of the hubble time . whether or not we are observing a very small percentage of extremely metal - poor bcds while they are undergoing their very first starburst at the present epoch is an interesting suggestion which remains to be investigated further ; it will depend critically on the cycling of their gas ( see below ) .
vii zw 403 shows evidence for at least three eras " of star formation .
the rgb stars suggest the first event occured at a look - back time of at least 3 gyr and probably @xmath010 gyr , the agb stars indicate star - formation also happened at around 4 gyr ago ( with considerable uncertainty ) , and the young stars testify to activity which took place less than about 1 gyr ago .
the data also allow us to infer the chemical enrichment history of vii zw 403 . the rgb stars are consistent with a metallicity of the order of z@xmath2/100 , while the ionized gas and ( by assumption ) the young stars are at @xmath4z@xmath2/20 ( martin 1997 , izotov et al .
curiously , an extended agb is not expected unless z / z@xmath2@xmath01/20 ( see sch98 , lynds et al .
1998 ) . the enrichment history of vii
zw 403 , at face value , is therefore inconsistent with closed - box models of galaxy evolution and seems to require the loss of enriched gas or accretion of metal - poor gas .
vii zw 403 is isolated from massive neighbors , so an accrection scenario la silk et al .
( 1987 ) seems unlikely .
papaderos et al .
( 1994 ) claim the x - ray detection of an outflow of hot gas from this galaxy .
employing our new distance , we can estimate the total gas ( h - i mass ) from the measurements of thuan & martin ( 1981 ) and tully et al .
( 1981 ) to be about 7x10@xmath16 m@xmath2 .
this places vii zw 403 in the mass range for which models suggest that gas may be blown out , but the entire gas reservoir may not be blown away ( mac low & ferrara 1999 ) .
the bcd vii zw 403 exhibits a radial population gradient .
young , blue stars and h@xmath6 emission are confined to the core . the core region has a diameter which equals the defining size of a bcd . intermediate - age and old ,
red stars are distributed throughout an extended background sheet or halo .
the halo stars show an rgb with a well defined tip and are interpreted to be an old and metal - poor population , similar to that in galactic globular clusters .
bcds were once recognized as `` the first metal - poor systems of population i to be discovered '' ( searle & sargent 1972 ) .
vii zw 403 is the first bcd with compelling evidence for the existence of a population ii halo .
vii zw 403 is also the first bcd for which a direct comparison has been possible between the results from population synthesis of the integrated halo color and resolved stellar content .
the detection of red giants at large distances from the starburst center verifies previous identifications of red halos with old stellar populations .
if all bcd halos harbor old stars , then they must have formed at high redshift and survived re - heating ; bcds would not require the delayed - formation scenario .
work on this project was supported through an archival research grant and guest observer hst grants to rsl and mmc ( go-7859.01 - 96a ) .
uh acknowledges financial support from sfb375 .
we thank livia origlia for supplying us with data prior to publication .
aloisi , a. , tosi , m. , greggio , l. 1999 , aj , in press aparicio , a. , gallart , c. 1994 , in third ctio / eso workshop on the local group : comparative and global properties , ed .
a. layden , r. c. smith & j. storm ( garching : eso ) , 115 aparicio , a. , dalcanton , j.j . ,
gallart , c. , martinez - delgado , d. 1997 , aj , 114 , 1447 babul , a. , ferguson , h.c .
1996 apj , 458 , 100 babul , a. , rees , m.j . ,
1992 mnras , 255 , 346 bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. , nasi , e. 1994 , a&as , 106 , 275 bessell , m.s . , castelli , f. , plez , b. 1998 , a&a , 333 , 231 blanchard , a. , valls - gabaud , d. , mamon , g.a .
1992 a&a , 264 , 365 bothun , g.d . , schmitz , m. , halpern , j.p . ,
lonsdale , c.j .
, impey , c. 1989 , apjs , 70 , 271 briggs , f.h .
1997 , apj , 484 , 618 broadhurst , t.j . , ellis , r.s . , shanks , t.s .
1988 , mnras , 235 , 827 burstein , d. , heiles , c. , 1984 , apjs , 54 , 33 da costa , g.s . , armandroff , t.e .
1990 , aj , 100 , 162 dekel , a. , silk , j. , 1986 , apj , 303 , 39 fagotto , f. , bressan , a. , bertelli , g. , chiosi , c. 1994 , a&as , 104 , 365 fanelli , m.n .
, oconnell , r.w . , thuan t.x .
1988 , apj , 334 , 665 ferguson , h.c . ,
babul , a. 1998 , mnras , 296 , 585 greggio , l. , tosi , m. , clampin , m. , de marchi , g. , leitherer , c. , nota , a. , sirianni , m. 1998 , apj , 504 , 725 guzmn , r. , jangren , a. , koo , d.c . , bershady , m.a . ,
simard , l. 1998 , apj , 495 , l13 holtzman , j.a .
, burrows , c.j . , casertano , s. , hester , j.j . ,
trauger , j.t . ,
watson , a.m. , worthey , g. 1995 , pasp , 107 , 1065 hopp , u. , schulte - ladbeck , r.e .
1995 , a&as , 111 , 527 hopp , u. , schulte - ladbeck , r.e . , crone , m.m .
1998 , in bonn / bochum - graduiertenkolleg workshop , eds .
t. richtler & j.m .
braun ( shaker verlag ) , 161 hunter , d.a . ,
thronson , jr . ,
h.a . , 1995 , apj , 452 , 238 ikeuchi , s , norman , c. 1987 , apj , 312 , 485 izotov , y.i .
, thuan , t.x . ,
lipovetsky , v. 1997 , apjs , 108 , 1 izotov , y.i . ,
thuan , t.x . , 1999 ,
apj , 511 , 639 kunth , d. , maurogordato , s. , vigroux , l. 1988 , a&a , 204 , 10 krger , h. , fritze - von alvensleben , u. , loose , h .- h . , fricke , k.j .
1991 , a&a , 242 , 343 lee , m.g . ,
freedman , w.l .
, madore , b.f . , 1993 apj , 417 , 553 lipovetsky , v.a . , chaffee , f.h . , izotov , y.i . , foltz , c.b .
, kniazev , a.y . , hopp , u. 1999 , preprint ( astro - ph/9902368 ) loose , h.h .
, thuan , t.x .
1986 , in star - forming dwarf galaxies and related objects , eds .
d. kunth , t.x .
thuan , and j.t.t .
van ( gif - sur - yvette : ditions frontires ) , 73 lynds , r. , tolstoy , e. , oneil jr .
, e.j . , hunter , d.a .
1998 , aj , 116 , 146 mac low , m .-
m . , ferrara , a. 1999 , apj , 513 , 142 martin , c. 1997 , apj , 491 , 561 mateo , m. 1998 , ara&a , 36 , 435 mndez , r.a .
, guzmn , r. 1998 , a&a , 333 , 106 meurer , g.r . 1999 , in dwarf galaxies and cosmology " , eds .
thuan , c. balkowski , v. cayatte , j. tran than van ( gif - sur - yvette : ditions frontires ) , in press minniti , d. , zijlstra , a.a .
1996 , apj , 467 , l13 minniti , d. , zijlstra , a.a . , alonso , m.v . 1999 ,
aj , 117 , 881 papaderos , p. , fricke , k.j . , thuan , t.x . ,
loose , h .- h .
1994 , a&a , 291 , l13 papaderos , p. , loose , h .- h . , thuan , t.x . , fricke , k.j .
1996 , a&as , 120 , 207 papaderos p. , izotov y.i . ,
fricke k.j . ,
thuan t.x .
, guseva n.g .
1998 , a&a 338 , 43 popescu , c.c . , hop , u. , elssser , h. 1997 , a&a , 325 , 881 rieschick , a. , hensler , g. 1998 , in bonn / bochum - graduiertenkolleg workshop , eds . t. richtler & j.m .
braun ( shaker verlag ) , 193 schmidt , a. a. , alloin , d. , bica , e. 1995 , mnras , 273 , 945 schulte - ladbeck , r.e . , crone , m.m . , hopp , u. 1998 , apj , 493 , l23 ( sch98 ) schulte - ladbeck , r.e . ,
hopp , u. 1998 , aj , 116 , 2886 searle , l. , sargent , w.l.w .
1972 , apj , 173 , 25 searle , l. , sargent , w.l.w . ,
bagnuolo , w.g .
1973 , apj , 179 , 426 silk , j. , wyse , r.f.g .
, shields , g.a .
1987 , apj , 322 , l59 spaans , m. , norman , c.a .
1997 , apj , 483 , 87 sweigart a.v . , greggio , l. , renzini a. 1990 , apj , 364 , 527 telles , e. , melnick , j. , terlevich , r. 1997 , mnras , 288 , 78 thuan , t.x .
1983 , apj , 268 , 667 thuan , t.x . ,
izotov , y.i .
, lipovetsky , v. , pustilnik , s.a . , 1994 , in eso / ohp workshop on dwarf galaxies , eds .
g. meylan & p. prugniel ( garching : eso ) , 421 thuan , t.x .
, izotov , y. , lipovetsky , v. 1997 , apj , 477 , 661 thuan , t.x .
, izotov , y.i . , 1998 , in bonn / bochum - graduiertenkolleg workshop , eds
. t. richtler & j.m .
braun ( shaker verlag ) , 91 thuan t.x .
, izotov y.i . , foltz c.b . 1999 ,
apj , submitted thuan , t.x . , martin , g.e .
1981 , apj , 247 , 823 tully , r.b .
, boesgaard , a.m. , dyck , h.m . , schempp , w.v .
1981 , apj , 246 , 38 vennik , j. , hopp , u. , popescu , c.c .
1999 , in iau colloq .
171 , `` the low surface brightness universe '' , ed .
j. davies , in preparation white , s.d.m . , rees , m.j .
1978 , mnras , 183 , 341 | we present evidence for the existence of an old stellar halo in the blue compact dwarf galaxy vii zw 403 .
vii zw 403 is the first blue compact dwarf galaxy for which a clear spatial segregation of the _ resolved _ stellar content into a core - halo " structure is detected .
multicolor hst / wfpc2 observations indicate that active star formation occurs in the central region , but is strikingly absent at large radii .
instead , a globular - cluster - like red giant branch suggests the presence of an old ( @xmath0 10 gyr ) and metal poor ( @xmath1[fe / h]@xmath0=-1.92 ) stellar population in the halo .
while the vast majority of blue compact dwarf galaxies has been recognized to possess halos of red color in ground - based surface photometry , our observations of vii zw 403 establish for the first time a direct correspondence between a red halo color and the presence of old , red giant stars . if the halos of blue compact dwarf galaxies are all home to such ancient stellar populations , then the fossil record conflicts with delayed - formation scenarios for dwarfs . |
preparing a massive object in a spatial quantum superposition over distances comparable to its size is a tantalizing possibility .
this has been achieved in seminal experiments on matter - wave interferometry of molecules @xcite , which hold a mass record of @xmath1 amu @xcite .
it would be particularly intriguing to prepare large quantum superpositions of even larger masses such that one could enter into the _ gravitational quantum regime _
( gqr ) , which we define as follows . for a solid sphere of radius @xmath2 and mass @xmath3 ,
we define the timescale @xmath4 , where @xmath5 is the planck s constant and @xmath6 the newton s gravitational constant .
@xmath7 has two interpretations : ( i ) it is the conjectured lifetime of a quantum superposition state of a single sphere delocalized over a distance @xmath2 , according to the parameter - free in the mass density is typically introduced , thereby making the model parameter - dependent . in that case the quantum superposition lifetime for a sphere of radius @xmath2 separated by a distance of the order @xmath2 would then be reduced by a factor @xmath8 , namely it would be given by @xmath9 .
thus the lifetime @xmath7 discussed in the text is a parameter - free upper - bound to the lifetime predicted by gravitationally - induced collapse models .
see @xcite for further details . ]
gravitationally - induced decoherence collapse model proposed by disi and penrose @xcite , ( ii ) @xmath10 is the kinetic energy of two spheres equivalent to their gravitational interaction energy when separated by a distance @xmath11 , a situation that could be used to measure @xmath6 @xcite .
moreover , @xmath7 is related to the much longer timescale @xmath12 required to entangle the center - of - mass motion of two spheres spatially localized within a length scale @xmath13 and separated by a distance @xmath14 using the gravitational interaction @xcite . for a given mass density , one can plot @xmath7 as a function of @xmath3 , see .
we define the gqr as the regime in which @xmath15 , where @xmath16 is a characteristic time scale of the quantum system for each of the two interpretations of @xmath7 . in particular : in ( i ) @xmath16
is the coherence time of the superposition taking into account standard sources of decoherence , in ( ii ) @xmath17 could be the trap frequency of a harmonic potential where two spheres are cooled near to their respective ground states .
choosing @xmath16 to be of the order of @xmath18 second , which is an ambitious timescale for quantum experiments , one sees in that @xmath19 pertains to masses @xmath20 .
the goal of this article is to propose an earth - based experimental scenario that allows one to access the gqr . preparing a large mass ( @xmath20 ) in a large quantum superposition ( comparable to its size )
is an ambitious challenge that we think can be addressed via two different approaches : either the _ bottom - up approach _ , consisting in gradually increasing the mass of the object that is prepared in a large superposition , or the _ top - down approach _ , that gradually increases the delocalization distance of a large mass that is brought into the quantum regime .
matter - wave interferometry has traditionally followed the bottom - up approach .
starting in the 30 s with the quantum interference of single electrons , the mass of the objects that can be interfered has remarkably increased up to molecules of @xmath1 amu @xcite .
the top - down approach is more recent and has been triggered by the field of quantum nano- and micromechanics @xcite .
recent achievements in this field include bringing objects with masses ranging from @xmath21 amu to @xmath22 amu to the quantum regime by cooling their associated mechanical degree of freedom to its quantum ground state @xcite .
however , while the mass is definitely large , the delocalization distance of the ground state in such systems is typically smaller than the size of a single atom .
such tiny delocalization distances prevent current quantum micromechanical oscillators to enter into the gqr , specially in the context of ( i ) .
levitated mechanical oscillators , in particular optically levitated dielectric nanospheres in high finesse optical cavities @xcite , have been suggested as an ideal nanomechanical system to implement the top - down approach by releasing the mechanically cooled nanosphere from the trap to increase the delocalization distance @xcite .
however , as shown in fig .
[ fig1 ] , these all - optical hybrid schemes @xcite using the best of matter - wave interferometry and quantum optomechanics seem , even in the most ambitious proposals requiring a space environment @xcite , to be limited to masses between @xmath23 and @xmath21 amu , putting them still away from the gqr . here
we propose to attain spatially large quantum superpositions of masses @xmath24 amu , well inside the gqr , by abandoning the use of lasers and using instead an _ all - magnetic _ on - chip architecture .
this approach combines the following salient features : ( i ) cryogenic temperatures both for environment and the massive particle to minimize decoherence due to emission , scattering , and absorption of black - body radiation , ( ii ) the use of static magnetic potentials created by persistent currents to magnetically levitate the sphere without creating decoherence ( space environment is not required ) as well as to exponentially speed - up quantum dynamics by using inverted potentials , and ( iii ) coupling the position of the sphere to quantum circuits on an integrated on - chip configuration to cool the center - of - mass motion to the ground state , to make a double - slit smaller than the size of the sphere by measuring the squared center - of - mass position , and to measure the interference pattern downstream .
these tools permit , in principle , to make an all - magnetic young s double slit experiment , with a single run - time of less than a second , on - earth , with a superconducting microsphere of radius @xmath25 m , mass @xmath22 amu , and with a slit separation of half a micrometer .
as discussed before , this ambitious goal could be experimentally addressed from a top - down approach by first implementing on - chip quantum magnetomechanics with superconducting microspheres @xcite , and then progressively implementing the key ingredients of the proposal .
the article is organized as follows . in
we discuss the main challenges that a proposal aiming at preparing large quantum superpositions of a massive object has to overcome .
the discussion of these challenges aims at motivating both the protocol and the implementation of our proposal . in
we describe the quantum micromechanical interferometer protocol as well as how to analytically calculate all the steps of the protocol taking into account decoherence .
this enables us to analytically obtain the interference pattern as a function of all the physical parameters of the protocol and environmental decoherence . in
we propose and analyze an all - magnetic on - chip implementation of the quantum micromechanical interferometer protocol for a superconducting microsphere . in we analyze a case study with a set of experimental parameters , whose complete list is given in the appendix .
while the experimental parameters have not been fine - tuned , this is an example which shows the feasibility to enter in the gravitational quantum regime such that parameter - free gravitational - induced decoherence could be unambiguously falsified .
we draw our conclusions and provide some final remarks in .
let us discuss three fundamental challenges that will need to be overcome by proposals aiming to prepare large superpositions of a massive object in the gqr : unavoidable decoherence , slow quantum dynamics , and the double - slit .
these challenges are discussed in a general fashion without considering any particular implementation .
they will serve as a motivation to introduce the quantum micromechanical interferometer protocol in and the implementation in that we propose to overcome them .
consider a sphere of radius @xmath2 whose center - of - mass position along the @xmath26-axis is described by the quantum state @xmath27 .
most sources of decoherence of the center - of - mass position are of the position localization ( pld ) type ( see , for example @xcite ) .
pld makes the off - diagonal terms of the density matrix in the position basis decay exponentially in time . in the so called long wavelength ( lw )
limit , where decoherence events provide only partial which - path information , it can be shown that @xmath28 \bra{x } \hat \rho(0 ) \ket{x'}$ ] , where @xmath29 is the localization parameter . in this regime decoherence
is modeled by the master equation ( t ) = - [ eq : pldecoh ] , where @xmath30 is the hamiltonian of the system . in the short wavelength
( sw ) limit , where decoherence events provide a full which - path information , one has that @xmath31 \bra{x } \hat \rho(0 ) \ket{x'}$ ] . as discussed in @xcite
most of the paradigmatic collapse models predict , at the level of density matrices , a pld .
in particular , the effect of the parameter - free gravitationally induced collapse model , conjectured by penrose and disi @xcite , can be described with the lw limit of the pld with a localization parameter given by @xcite _ g = .[eq : dplam ] here @xmath6 is the newton s gravitational constant .
note that @xmath32 is parameter - free , namely , it does not depend on any phenomenological parameter , since a homogeneous mass density is assumed .
as already remarked in a footnote of the introduction , such choice provides a lower bound on the gravitationally - induced decoherence rate , hence by falsifying this case all other parameter - dependent forms of gravitationally - induced decoherence would also be falsified , see @xcite for further details
. combining eqs .
[ eq : pldecoh ] and [ eq : dplam ] , we can immediately see that the correlations in a spatial quantum superposition of extent @xmath2 decay as @xmath33 \bra{r/2 } \hat \rho(0 ) \ket{-r/2}$ ] , where @xmath34 is the timescale used in to define the gqr . in order to falsify such an exotic source of decoherence , which predicts the breakdown of standard quantum mechanics
, one has to account for and control the standard sources of decoherence that are expected in an implementation .
we discuss the following inevitable sources of decoherence : scattering of air molecules , emission , absorption , and scattering of black - body radiation , and stochastic forces due to vibrations .
decoherence due to scattering of air molecules , with a thermal wavelength typically smaller than the superposition size of interest , is described by pld in the sw limit @xcite .
the decoherence rate is given by _
a = , where @xmath35 is the environmental pressure , @xmath36 is the environmental temperature , and @xmath37 is the mass of an air molecule . in order to preserve coherence , the quantum micromechanical interferometer should be implemented in a time @xmath38 , namely , it is required that the sphere does not scatter any air molecules during each run of the protocol . for a sphere of @xmath39 in an optimal cryogenic environment with @xmath40 and @xmath41 , the coherence time
is given by @xmath42 .
hence , a single run of the quantum micromechanical interferometer has to be performed in a timescale of the order of a second or less . combined with the slow quantum dynamics of large masses , as we discuss below , this poses a challenge .
we remark that pressures as low as @xmath43 can be achieved in cryogenic environments .
this was shown , for example , in the experiments with antiproton lifetimes by gabrielse _
et al _ @xcite .
note that @xmath44 is an upper - bound for the time scale @xmath16 used to define the gqr , especially regarding interpretation ( i ) .
decoherence due to scattering , absorption , and emission of black - body radiation is described by pld in the lw limit , since the wavelength of thermal photons is typically much larger than the superposition size .
the localization parameter is given by @xmath45 , where @xcite ^s_bb c r^6 ^9 ^2 is the contribution due to scattering and [ eq : lbbe ] ^e(a)_bb c r^3 ^6 are the contributions due to emission ( e ) and absorption ( a ) of thermal photons .
the last two contributions are the same when the internal bulk temperature @xmath46 of the sphere is the same as the environmental @xmath36 .
these localization parameters depend on the dielectric permittivity @xmath47 of the sphere at the thermal frequency @xmath48 . for a sphere of @xmath39 and density @xmath49
, one can show that & & 10 ^ -15 ^9 ^ 2 + & & 10 ^ -7 ^6 hence , in order to minimize the effect of decoherence due to black - body radiation , such that gravitationally - induced decoeherence would dominate , the quantum micromechanical interferometer should be implemented in a cryogenic environment .
any experiment will suffer from unavoidable vibrational noise .
let us consider vibrations along the @xmath26-axis described by the fluctuating variable @xmath50 , which has length units and fulfills @xmath51 and @xmath52 .
the vibrational noise is characterized by the power spectral density s_xx ( ) = e^ , which has units of @xmath53 . for a given external potential @xmath54 that depends on the @xmath26-coordinate of the center - of - mass of the sphere
, the vibrational noise leads to a time - dependent stochastic force given by @xmath55 $ ] .
the hamiltonian of the system can then be written as @xmath56 . for a harmonic potential of the form @xmath57 with the stochastic force due to vibrations given by @xmath58
, one can show @xcite that by averaging the stochastic force , one obtains a master equation for the dynamics .
for the particular case of @xmath59 , this master equation is of the pld form in the lw limit with a localization parameter given by _
v = s_xx(_p ) .
note that the decoherence due to vibrations scales quadratically with the mass of the sphere and quartically with the trapping frequency since stochastic force is proportional to @xmath60 for a harmonic potential . as expected , the stochastic force is zero when the particle is not in an external potential . in order to falsify gravitationally - induced decoherence in the presence of an external harmonic potential
, the decoherence induced by such fluctuations has to satisfy @xmath61 , which becomes increasingly challenging with the size of the sphere .
in particular , for a sphere of @xmath39 one requires @xmath62 ^ 2 \times 10^{-15}~\text{m}/\sqrt{\text{hz}}$ ] .
thus , in order to suppress the effects of vibration - induced decoherence , the implementation of the quantum micromechanical interferometer necessitates a minimal use of external potentials in addition to substantial vibration isolation .
in any quantum micromechanical interferometer scheme there are two important dynamical processes that require some time : the generation of a large enough coherence length , to be defined below , and the generation of an interference pattern with sufficient visibility , namely that the separation between the interference fringes is larger than the resolution with which the position can be measured . in order to define the coherence length ,
let us first recall some properties of an important class of states
gaussian states that can be used to describe the quantum state of the sphere along the @xmath26-axis . a quantum state @xmath27 with @xmath63 = \tr [ \pop \hat \rho ] = 0 $ ] is called gaussian if it is completely determined by the following three real parameters @xmath64 $ ] , @xmath65 $ ] , and @xmath66/2 $ ] @xcite .
indeed , the density matrix of a gaussian state in the position basis is given by [ gausstate ] = e^ -a_1 x^2 - a_1^ * x^2 - a_2 x x , where v_x a_1 & - , + v_x a_2 & .
here we have used that the purity @xmath67 $ ] of a gaussian state is given by @xmath68^{-1/2}$ ] . note that the heisenberg uncertainty relation for a gaussian state reads as @xmath69 .
the coherence length @xmath70 for a gaussian state is defined as @xmath71/\sqrt{2 \pi v_x}$ ] , and is given by the compact expression .
the coherence length scales with the position variance as @xmath72 , as expected , with a scaling factor given by the purity of the state . for a gaussian state with a coherence length @xmath70 incident on a double slit of separation @xmath73
, one will observe a coherent superposition resulting in interference fringes for @xmath74 and an incoherent mixture for @xmath75 .
it is straightforward to show that during free expansion of the wavefunction without decoherence , or in other words , under unitary evolution with @xmath76 , the coherence length grows at a speed given by _
( t ) = . starting out in the ground state of a harmonic potential with frequency @xmath77 , such that @xmath78 and @xmath79 , then @xmath80 .
note that the larger the mass the slower the coherence length grows .
for a sphere of @xmath39 and mass density @xmath81 initially in a harmonic trap of frequency @xmath82 , the speed at which the coherence length grows is @xmath83 .
since we would like to have @xmath70 of the order of @xmath2 , this speed in free expansion poses a challenge for propagation times of the order of a second , as required to prevent decoherence due to scattering of air molecules .
even more challenging is the time required to generate sufficiently separated fringes .
it is known in matter - wave interferometery that after the gaussian state passes through a double slit of separation @xmath84 , one obtains a coherent superposition which after free expansion downstream generates , in the absence of decoherence , an interference pattern with a fringe spacing x_f =
. the fringe separation speed is thus given by @xmath85 .
for a sphere of @xmath39 , mass density @xmath49 , and assuming a double slit of separation @xmath86 , the speed is given by @xmath87 which is very small .
for example , assuming a position measurement resolution of @xmath88 , one would require @xmath89 to obtain resolvable fringes .
this definitely poses a very important challenge .
the final challenge is to prepare a quantum superposition of the center - of - mass position of the sphere , which should ideally be done without any decoherence .
this is a non - gaussian operation that transforms a gaussian state with a positive wigner function into a non - gaussian state with a negative wigner function .
the preparation of non - gaussian states has been a major goal and a rather formidable challenge in quantum nanomechanics @xcite .
a double slit in a young s experiment is a beautiful example of how this can be done .
however , in a double slit one is typically forced to prepare spatial superpositions which are larger than the size of the object , namely @xmath90 . for large masses
this condition might be too demanding as it could compromise the conditions required to meet the challenges discussed before .
for instance the larger the superposition the longer it takes for fringes to be generated for the final measurement .
hence it might be advantageous to prepare superpositions with an extent not necessarily larger than the diameter of the sphere . implementing a double slit with a slit separation smaller than the size of the sphere is a seemingly naive challenge , but an important one to meet when taking into account the constraints posed by the slow quantum dynamics and the unavoidable sources of decoherence .
in this section we propose and analyze a quantum micromechanical interferometer protocol for a massive object .
the protocol will be designed in a manner such that two of the challenges discussed in , namely that of slow quantum dynamics and of the double slit , are addressed .
a detailed implementation of the protocol for the specific case of a superconducting sphere is proposed and analysed in .
this implementation is designed to overcome the unavoidable decoherence challenge .
thus in this section we keep the discussion general without reference to any particular implementation .
the quantum micromechanical interferometer protocol is divided into 7 steps ( see ) ; 1 . *
cooling*. cooling the center - of - mass motion in a harmonic trap of frequency @xmath91 for a duration @xmath92 to a final phonon occupation number @xmath93 .
* boost*. evolution of the center - of - mass motion in an inverted harmonic potential of frequency @xmath94 for a time @xmath95 which exponentially boosts the sphere s kinetic energy .
3 . * free*. free evolution for a time @xmath96 to delocalize the center - of - mass over large distances .
4 . * split*. a squared position measurement for a time @xmath97 to prepare a quantum superposition state . 5 .
* rotation*. a short evolution for a duration @xmath98 in a harmonic trap of frequency @xmath99 to give opposite momentum to the two wave packets in the superposition .
6 . * inflation*. evolution in an inverted quadratic potential of frequency @xmath100 for a duration @xmath101 to exponentially generate the interference fringes
* measure*. a squared position measurement of time @xmath102 to unveil the interference pattern .
the total time of the protocol after cooling is thus given by @xmath103 .
the position of the center - of - mass of the sphere is given by @xmath104 , where @xmath105 is the position of the center of along the direction in which the quantum superposition is prepared .
the other degrees of freedom are assumed to be in a thermal classical state with a temperature as low as possible .
steps 1 to 3 of the protocol are used to create a large and coherent matter - wave of the center of mass of the sphere along the @xmath26-axis .
these are the steps used to generate an observable difference between the state of the sphere when one includes the existence of gravitationally - induced decoherence and when one does not .
in particular , the coherence length @xmath70 after step 3 , takes a smaller value @xmath106 when gravitationally - induced decoherence is taken into account . in order to falsify gravitationally - induced decoherence we aim at operating the micromechanical interferometer in a regime where @xmath107 , for a given slit separation @xmath73 .
this implies that the double slit prepares a coherent superposition in the absence of gravitationally - induced decoherence , and an incoherent mixture otherwise .
steps 4 to 7 consist of performing the double slit operation and unveiling the interference fringes for the case in which the coherent superposition has been prepared .
these steps have to be done unitarily such that if one prepares a coherent superposition the visibility of the interference fringes is not washed out .
thus , we emphasize that steps 1 to 3 are used to create the difference between the state with and without the existence of gravitationally induced decoherence , and steps 4 to 7 are used to demonstrate this difference . in the following
we show how the position probability distribution along the @xmath26-axis after step 7 , namely @xmath108 , can be analytically calculated taking into account pld in the lw limit and all the parameters describing the protocol .
we divide the discussion into the separate steps of the protocol . in this step , which takes place from @xmath109 to @xmath110 , the center - of - mass motion of the microsphere along the @xmath26-axis
is cooled using a linear quantum nanomechanical coupling to a cavity mode @xcite .
the particle is confined in a harmonic trap of frequency @xmath111 , such that its zero - point motion length is given by @xmath112 .
the total hamiltonian describing the coupling of the particle to a driven cavity mode of frequency @xmath113 and creation ( annihilation ) mode operators @xmath114 ( @xmath115 ) is given by [ coolham ] = & - + _ 1 + g_l(+ ) + & - e_1 ( - ) .
the above hamiltonian is given in a rotating frame with the drive frequency @xmath116 . here
@xmath117 with @xmath118=1 $ ] , @xmath119 , @xmath120 is the strength of the coherent drive that couples to the cavity mode , and @xmath121 is the linear quantum nanomechanical coupling .
we linearize the system by making the change of variables @xmath122 and @xmath123 , and choosing @xmath124 and @xmath125 such that assuming that @xmath126 and neglecting the decoherence in the center - of - mass motion , one can eliminate the cavity mode degree of freedom @xmath115 and write an effective master equation for the reduced density matrix of the sphere .
this master equation leads to the following rate equation for the phonon number of the sphere center - of - mass degree of freedom : with the additional assumption that the bath attached to the cavity mode is at zero temperature . in this case the damping and driving terms of the above rate equation are given by with the shifted detuning @xmath127 .
if at @xmath110 the initial number of phonons is @xmath128 , the number of phonons at a later time @xmath92 is given by |n = + e^-(a_--a_+ ) t_1 .
an additional heating source in the center - of - mass motion at a rate @xmath129 can be easily included by replacing @xmath130 with @xmath131 .
note that when the coupling to the cavity mode is turned on for the cooling process , the equilibrium position of the sphere along @xmath26 shifts to the mean field value @xmath132 .
one can expect that as the coupling is turned off slowly the cooled state of the particle will be adiabatically shifted to @xmath133 . at @xmath110 the center - of - mass motion along the @xmath26-axis
is thus prepared in a gaussian state given by [ eq : solstep1 ] v_x(t=0 ) & = _ 1 ^ 2 ( 2 |n + 1 ) , + v_p(t=0 ) & = ( 2 |n + 1 ) , + c(t=0 ) & = 0 .
+ this state has a purity @xmath134 and an initial coherence length given by @xmath135 .
owing to the large mass , the coherence length at this step will be typically much smaller than the size of a single atom .
our goal is to expand it to a value of the order of the radius of the sphere , that is , by many orders of magnitude .
this is what we do in the next two steps of the protocol .
as discussed in , in free space the coherence length of the state grows with a speed given by @xmath136 . in order to increase this speed , in step 2 ( boost ) the center - of - mass evolves in an inverted harmonic potential of frequency @xmath137 , given by the hamiltonian @xmath138 , from @xmath110 to @xmath139 such that @xmath140 is boosted by a factor of @xmath141 $ ] .
assuming pld in the lw limit , with a localization parameter @xmath142 , the dynamics preserves the gaussian character of the state at @xmath110 since the hamiltonian is quadratic and the jump operators in the master equation are linear in @xmath105 and @xmath143 .
thus the parameters characterizing the gaussian state at @xmath139 can be obtained by analytically solving [ eq : dynamics ] v_x ( t ) & = c(t ) , + v_p(t ) & = 2 m _ 2 ^ 2 c(t ) + 2 ^2 _ 2 , + c ( t)&= m _ 2 ^ 2 v_x(t)+ v_p(t ) . with the initial conditions given by .
this leads to v_x(t_2 ) = & + & + + & + , v_p(t_2 ) = & + & + + & + , c ( t_2)= & - + & + + & + . note the exponential dependence on @xmath144 due to dynamics in an inverted potential . after boosting the momentum fluctuations of the center of mass along the @xmath26-axis ,
the center - of - mass motion is let evolve freely from @xmath139 to @xmath145 in the presence of pld with a localization parameter @xmath146 . denoting @xmath147 , at @xmath148 the time - evolved gaussian state
is given by solving with @xmath149 , one obtains v_x(t_3 ) = & v_x(t_2 ) + t_3 + t_3 ^ 2 + & + t_3 ^ 3 , + v_p(t_3 ) = & v_p(t_2 ) + 2 _ 3 ^2 t_3 , + c(t_3 ) = & c(t_2 ) + t_3 + . note that @xmath150 grows linearly in time until decoherence due to the @xmath146 term dominates , then @xmath151 . in this regime
the purity decays as @xmath152 causing the coherence length decay as @xmath153 .
note that so far no approximations have been made with regards to the treatment of the dynamics and thus the state of the center - of - mass can be fully determined .
the difference in the size of the coherence length with or without the presence of gravitationally - induced decoherence is expected to be apparent at this point . in this step
the spatially delocalized wavefunction is split by means of a double slit into a spatial superposition of two localized wavepackets .
as discussed in , it would be convenient to obtain a slit separation @xmath73 smaller than the diameter of the sphere .
we propose to achieve this by performing a continuous time quantum measurement of @xmath154 .
the measurement outcome will determine the slit separation , and the strength of the measurement the width of the slits .
the idea of using an @xmath154 measurement of a nanosphere to prepare a quantum superposition state was introduced in @xcite . in those cases the measurement was approximated to be instantaneous , here
we address the more general situation considering that the measurement takes a finite time @xmath97 .
the measurement is modeled as follows .
we consider the sphere to be quadratically coupled with a cavity mode that is resonantly driven with a strength @xmath155 .
the quadratic coupling strength is given by @xmath156 , and in this step it is convenient to use the length unit @xmath157 , namely we define @xmath158 .
here @xmath159 is the purity of the state after free expansion @xmath160 .
the hamiltonian in the frame rotating at the cavity resonance frequency is given by the decay rate of the cavity is given by @xmath161 .
we assume that a homodyne measurement of the phase quadrature of the output cavity field is performed , such that the evolution of the density matrix @xmath162 describing the center - of - mass motion and the cavity mode is given by the following stochastic master equation @xcite [ eq : homdynesme ] _ t = & - t - _ 4 t + & + 2 _ 4 t + & - , where @xmath163 is the wiener increment .
again , we assume pld with a localization parameter @xmath164 . after a displacement operation on the cavity field by an amount given by the classical field amplitude in the empty cavity , namely @xmath165 , we can perform an adiabatic elimination ( see @xcite ) of the cavity mode when @xmath166 .
the resulting equation for the reduced density matrix of the @xmath26-coordinate of the center - of - mass position of the sphere is given by [ eq : smepos2 ] d = & - t + & - _ 4 t + & - t + & + .
the effective measurement strength representing the continuous measurement of @xmath167 @xcite is given by @xmath168 .
the quadratic coupling leads to an effective harmonic potential of frequency given by @xmath169 . to solve
it is useful to write the density matrix in the position basis as [ eq : aphi ] ( t ) = a(x , x,t ) , where @xmath170 , @xmath171 , such that @xmath172 , and @xmath173 .
note that since @xmath174 \leq 1 $ ] , then @xmath175 . by introducing in ,
one obtains [ eq : diffeqs1 ] a = & -_a t + & -2 _ t + & -_4 ^2_4 ( x - x)^2a t + & -(x^2-x^2)^2 a t + & + a , and [ eq : diffeqs2 ] = & - _ t + & + a t + & - t. where we have used @xmath176 . at this step
, one can perform the phase - amplitude separation approximation which corresponds to neglecting the second term in , thereby making the evolution of @xmath177 independent of @xmath178 .
this is justified since @xmath177 evolves much faster than @xmath178 as initially @xmath179 with @xmath180 .
recall that to keep the purity constant , one requires @xmath181 to be comparable to @xmath182 which at @xmath183 , after expanding the coherence length many orders of magnitude , is much larger than @xmath184 . using @xmath185 as the initial condition , @xmath177 is given by @xmath186 with [ eq : soldiffeqs3 ] f(t)= - . using
this one can thus solve to obtain the time - evolved density matrix . by neglecting the pld term and the contribution of phase gradients to the amplitude dynamics in , which lead to negligible contributions for the experimental parameters given in , and enforcing that @xmath187 , the density matrix at @xmath97 is given by ( x , x,t_4 ) m_(x ) m _
( x ) a(x , x,t_3 ) , where m_(x ) = ^1/4 .
here @xmath188 is a normally distributed random number with a variance @xmath97 and a zero - mean that parametrizes the outcome @xmath189 of the homodyne measurement .
the measurement strength is given by @xmath190 , where the inequality is for the adiabatic elimination condition to hold .
note that the position probability distribution after the action of the measurement operator is given by two gaussian peaks separated by and a width given by [ eq : peakwidth ] _ d & = = .
we have assumed here that the measurement outcome is such that @xmath191 , which is the condition required to make the two peaks distinguishable .
thus the continuous time measurement of @xmath154 effectively acts as a double slit .
the probability distribution for the outcomes @xmath192 , when the measurement operator is applied on the input state @xmath193 , is given by : one aspect of realising such a double slit is that the slit separation depends on the measurement outcome @xmath194 and will vary from run to run . as a result in our proposal we have to post - select a suitable slit - width that satisfies @xmath195 which will enable us to falsify gravitationally - induced decoherence . to this end
it is advantageous also to have at hand the probability distribution for the slit separation @xmath73 for a given measurement strength given by using the above distribution we can now estimate the probability for the slit separation to fall in a given range of interest @xmath196 . choosing the minimum slit separation as @xmath197 ensures that the two peaks in the post - measurement state are well resolved , while the condition @xmath198 naturally leads to the choice @xmath199 . in the limit of large enough measurement strengths such
that all the corresponding outcomes satisfy @xmath200 , we get for this probability @xmath201 - \ \erf\,\left[\frac{2\mathcal{p}_3}{\sqrt{\chi_4 } } \right],\end{aligned}\ ] ] which for large measurement strength @xmath202 , tends to approach @xmath203\approx 0.8 $ ] . at this step ,
the state after a double slit of separation @xmath73 has been prepared . as discussed in
, the generation of fringes in free expansion would require an unfeasibly long time .
in the next two steps , we propose a method to overcome this challenge . in the next two steps ,
step 5 ( rotation ) , and step 6 ( inflation ) , we will ensure that the fringes separate exponentially in time instead of linearly as implied by dynamics in free space . to achieve this , first , in step 5 ( rotation ) , we let the system evolve with the hamiltonian @xmath204 during a time @xmath98 , such that @xmath105 and @xmath143 evolve in the heisenberg picture as [ eq : rotation ] ( t_5)&=(t_4 ) ( _ 5 t_5)+(_5 t_5 ) , + ( t_5)&=(t_4 ) ( _ 5 t_5 ) -m_5 ( t_4 ) ( _ 5 t_5 ) . here , for simplicity
, we have ignored decoherence for the moment . in step 6 ( inflation )
, the system evolves with the hamiltonian @xmath205 for a time @xmath101 , such that @xmath105 and @xmath143 evolve in the heisenberg picture as ( t_6)&=(t_5 ) ( _ 6 t_6)+(_6 t_6 ) , + ( t_6)&=(t_5 ) ( _ 6 t_6 ) + m_6 ( t_5 ) ( _ 6 t_6 ) .
note that for @xmath206 , then ( t_6 ) .
using , and choosing @xmath98 such that [ eq : tof_condition ] ( _ 5 t_5 ) = ( _ 5 t_5 ) , then ( t_6 ) e^_6 t_6 ( t_4 ) .
thus @xmath207 is proportional to @xmath208 with a coefficient that grows exponentially with @xmath209 .
therefore one can perform an exponentially fast time - of - flight measurement of the momentum distribution at @xmath210 by measuring the position distribution at @xmath211 .
note that for @xmath212 the condition is satisfied with @xmath213 .
we remark that should one be interested in measuring the probability distribution of another quadrature at @xmath210 , one could adjust @xmath98 accordingly . motivated by the discussion above ,
let us calculate the state obtained after rotation for a time @xmath98 .
since after step 4 ( split ) , the state is non - gaussian , its evolution has to be calculated more carefully . in this context , it is convenient to use the wigner function w(x , p , t)= _ - ^y ( t ) . in the presence of pld in the lw limit , with localization parameter @xmath214 , and the hamiltonian @xmath215 , the evolution of the wigner function
is given by = -+m _ 5 ^ 2 x+^2_5 .
to solve this eqution it is useful to define the fourier transform |w ( k_x , k_p , t ) = _ -^w(x , p , t)e^-i k_x xe^-i k_p px p , such that the dynamical equation reads = -m _ 5 ^ 2 k_p-^2_5 k_p^2|w . [ recpw ]
let us now make the ansatz where @xmath216 ( @xmath217 ) is the solution of with @xmath218 ( @xmath219 ) . plugging into leads to the following equation for the convolution function @xmath220 = -m _ 5 ^ 2 k_p-^2_5 k_p^2|g .
[ recpg ] this equation together with the initial condition @xmath221 has the following solution where g(k_x , k_p , t ) & + & \ { -2 t _ 5 - . + & . + } .
this gives the exact solution of the evolved wigner function including pld .
note that for a quadratic hamiltonian the evolution of the wigner function in the absence of decoherence is given by @xmath222 , where @xmath223 and @xmath224 are the classical solutions given by the hamilton equations . in order to quantify the effect of decoherence during the rotation ,
it is convenient to define dimensionless variables : @xmath225 , @xmath226 and @xmath227 , for some given length and momentum units @xmath228 and @xmath229 . then , with a_1 & , + a_2 & , + b & 2|t + + & - .
+ with the experimental parameters used in the case study , and with @xmath230 and @xmath231 ( recall and ) , one can show that typically @xmath232 and thus decoherence in this step can be neglected . in step 6
, the center of mass evolves in an inverted potential of the form @xmath233 and pld with localization parameter @xmath234 from @xmath235 to @xmath211 .
the state can be calculated using the wigner function , as discussed in step 5 replacing @xmath236 and @xmath237 .
after the evolution , we expect to observe fringes in the position probability distribution , namely in @xmath238 . using the results of the previous section
, one can show [ eq : fringes ] p_0=_-^p_= 0dy , where under the action of @xmath239 [ eq : blurring ] _
= is the blurring distance due to decoherence .
one can show that the fringe separation grows exponentially in time approximately as @xmath2402 \pi \hbar/(md \w_6)$ ] .
it will thus be important that decoherence during the inflation step is low enough such that @xmath241 at @xmath242 so that the interference pattern is not washed out .
note that the position probability distribution of the center - of - mass after the step 6 ( inflation ) can be calculated analytically taking into account all the experimental parameters and pld in the lw limit .
this is used in to obtain the interference pattern for a specific experimental implementation of the protocol .
the final step ( measure ) consists in performing a continuous time measurement of @xmath154 in order to unveil the fringes .
note that since the interference pattern has to be symmetric , it is more convenient to directly measure @xmath154 , as is done in step 4 ( split ) .
hence , in this step the hamiltonian is given by where the convenient length scale unit is now given by @xmath243^{1/2}$ ] .
note that the cavity parameters in this step are given by : the resonant driving strength @xmath244 , the decay rate of the cavity given by @xmath245 , and the classical field amplitude in the empty cavity @xmath246 .
a homodyne measurement of the phase quadrature of the cavity field is performed assuming , again , the adiabatic approximation , namely @xmath247 .
further , invoking the phase - amplitude separation approximation , which is valid again due to the large phase accumulated during step 5 ( inflation ) , the measurement strength is given by @xmath248 . the position resolution that can be achieved in the measurement
is bounded from below by x .
this position resolution ignores measurement back - action which is justified in the parameter regime we are interested in . in order to resolve the fringes
we require @xmath249 .
in this section we propose and analyze a physical set - up for the implementation of the quantum micromechanical interferometer protocol discussed in .
the proposal is illustrated in .
we consider a superconducting sphere that is magnetically levitated on top of a superconducting chip , at some distance @xmath250 , and skates along the @xmath251-axis .
the main ingredients of the implementation are the following : * the different steps of the protocol are implemented by letting the sphere skate along the @xmath251-axis in a static potential landscape as sketched in b. the potentials along the @xmath26-axis , @xmath54 , act during the time required for the sphere to skate through the particular region in the potential landscape .
this is advantageous since the potential is static and does not need to be switched - on and off which could easily cause additional decoherence .
* the potential landscape is generated by persistent superconducting currents that give rise to a static magnetic field potential that interacts with a superconducting microsphere in the meissner state @xcite .
being persistent , the intensity of the currents does not fluctuate and hence the magnetic potentials are stable , minimising decoherence arising from trap fluctuations . *
the cavities required for cooling , splitting , and measuring the position of the state are implemented with flux - dependent microwave quantum cavities containing a squid .
this leads to a quantum magnetomechanical coupling @xcite between the sphere and the cavities .
these quantum circuits are positioned on a chip at @xmath252 , where the persistent currents are located as well .
the sphere skates along the @xmath251-axis , levitated at some distance @xmath250 .
the squids are positioned in the appropriate region of the magnetic landscape such that when the particle skates over them , the desired flux - dependent couplings to the particle position ( or squared position ) are implemented .
note that the driving fields of the cavities can be switched on throughout since the interaction of the sphere with the cavity will only take place while the sphere is skating near them . in order to control the speed
at which the sphere skates along the @xmath251-axis , the chip could be conveniently inclined .
a superconducting sphere with optimal properties could be loaded on a superconducting chip using a magnetic conveyor as has been experimentally realized with ultracold atoms @xcite .
being an on - chip implementation , one could imagine having a system in which the same sphere can be reused for each run of the experiment . at the end of each run the sphere could be brought back to step 1 with an on - chip magnetic conveyor .
this would prevent the run - to - run variations that would necessitate an average over mass and magnetic properties of the different spheres used to obtain the interference pattern . in the following
we discuss how the magnetic landscape in b , which we refer to as a _ magnetic skatepark _ , can be implemented with persistent currents .
subsequently we discuss how steps 1 , 4 , and 7 can be performed using a quantum magnetomechanical coupling to a microwave resonator . then we consider decoherence due to magnetic field fluctuations arising from the surface , a source of decoherence which is introduced by choosing this particular physical implementation . finally , we argue that this implementation can be addressed by a set of progressively challenging experiments in the field of quantum micromechanical oscillators . for a perfect superconducting sphere of volume @xmath253 in an applied magnetic field @xmath254 ,
the hamiltonian describing the interaction between the magnetic field and the induced dipole @xcite of the sphere , whose center of mass is at position @xmath255 in a gravitational field along the @xmath256-axis , is given by @xcite [ eq : magpot ] v()= | ( ) |^2 + m g z , where @xmath257 , @xmath258 is the vacuum magnetic permeability , and @xmath259 is the effective gravitational acceleration along the @xmath256-axis taking into account the small inclination of the chip .
this expression assumes that the superconducting sphere is in the meissner state , namely that @xmath260 in the interior of the sphere
. such assumption requires the radius of the sphere to be much larger than the fundamental length scales in superconductivity , namely the london s penetration depth and the superconducting coherence length @xcite .
the magnetic field used to engineer the potential required in each of the steps of the protocol will be created by 4 parallel wires @xcite .
let us start by considering 4 ideal infinite wires ( neglecting their thicknesses ) placed along the @xmath251-direction at @xmath252 .
the two interior wires , separated by a distance @xmath261 , carry a current of strength @xmath262 in opposite directions .
the two outer wires , separated by a distance @xmath263 , carry a current @xmath264 , as sketched in a. these wires create a magnetic field , @xmath265 , that does not depend on @xmath251 . in absence of gravity , analytical expressions of the trapping position can be derived @xcite .
considering gravity , we numerically study the properties of the traps .
in particular , we study the normalized potential seen by the sphere as a function of its coordinates @xmath266 , where @xmath267 is the field created by the two inner wires at the origin of coordinates , @xmath268 , and @xmath269 . here
@xmath270 is the mass density of the sphere .
the properties of the magnetic potential can be modified by tuning the distances and intensities of the wires .
for the protocol discussed in , the following potential distributions are needed : * strong attractive potential ( step 1 ) : strong confinement in the @xmath26-direction , while the particle is trapped at @xmath271 and @xmath272 . *
inverted potential ( steps 2 , 6 ) : inverted potential in the @xmath26 direction while the particle is levitating at @xmath273 . *
free propagation potential ( steps 3 , 4 , 7 ) : flat potential in the @xmath26-direction while the particle is levitating at @xmath273 .
* weak attractive potential ( step 5 ) : weak confinement in the @xmath26-direction , with a frequency value similar to that for the inverted potential while the particle is trapped at @xmath271 and @xmath273 .
all of the different potentials required above can be obtained by fixing the current intensities through the wires and the distance between the interior wires , and only tuning the distance between the outer ones . in b
we plot the absolute value of the trapping frequency in @xmath26 as a function of the distance between the outer wires ( for the particular distances and currents used in the case study ) . in c
we plot the corresponding trapping heights @xmath274 .
as can be seen from these plots , the magnetic potential can be modified by only tuning the outer distance between wires , allowing to create the different distributions required for the protocol .
the strong attractive potential is generated placing the outer wires at a distance @xmath275 .
this generates a quadrupole - like field distribution , and the trapping takes place slightly below the zero field point , @xmath276 , given by @xcite @xmath277 a tight trap both in @xmath26 and @xmath256-directions is then created .
on the contrary , the inverted potential is obtained by placing the outer wires closer to the inner ones . in this case , as @xmath278 , the potential is the same as that created by a single pair of wires with current ( @xmath279 ) , giving rise to an inverted potential in @xmath26 and a stable trapping point in @xmath256 due to the gravitational force . between these two cases
, one can find a @xmath280 value for which the potential in the @xmath26 direction at @xmath281 changes from being inverted to attractive , see b. this allows us to design a flat potential permitting free expansion of the state along the @xmath26-axis .
a possible set of experimental numbers for implementing the protocol are discussed in .
these results have been obtained assuming that the wires are infinitely long .
however , we have performed calculations considering the complete protocol with the different finite regions as sketched in a. both trapping heights and potential profiles agree well with the results for infinite guides and transitions from one region to the other can be made adiabatic by controlling the lengths of the connections between the different regions . however , taking into account these smooth transitions in the protocol , and choosing the experimental parameters accordingly in the case study discussed in lies beyond the scope of this manuscript . in steps 1 , 4 , and 7 a coupling to a cavity mode is required .
this is achieved by a magnetomechanical coupling between the center - of - mass of the superconducting microsphere and a microwave resonator with a squid @xcite .
the frequency of the cavity mode is flux - dependent due to the presence of the squid @xcite , and it is given by @xmath282 , where @xmath114 ( @xmath115 ) is the creation ( annihilation ) cavity mode operator , @xmath283^{1/2}- e_c$ ] , and @xmath284 here @xmath285 is the flux quantum , @xmath286 is the position of the superconducting sphere , @xmath287 is the charging energy of a single electron stored in the capacitance , and @xmath288 the energy associated with an electron tunneling across one of the two identical junctions .
the regime @xmath289 is assumed so that the resonator is linear .
note that the magnetomechanical coupling arises because the flux threading the pick - up coil @xmath290 depends on @xmath286 . as before we denote the position of the sphere as @xmath255 with @xmath105 corresponding to the quantum mechanical position operator of the center of mass motion along the @xmath26-axis .
expanding the hamiltonian up to second order in @xmath105 , one has @xmath291 where @xmath292 , and @xmath293 is the position scaled by a length scale which we choose in each step of the protocol and is related to @xmath294)^{1/2}$ ] .
the linear and quadratic magnetomechanical couplings are then given by g_l = & - _ l , + g_q = & - , where we have defined the frequency @xmath295 , @xmath296 $ ] , @xmath297 $ ] , @xmath298 , and the magnetomechanical dimensionless parameters _ l & .
|_x=0 , + _ q & .
|_x=0 . note that to have a quadratic coupling the pick - up coil is placed such that @xmath299 and @xmath300 . whenever @xmath301 and @xmath300 , the linear coupling is much larger than the quadratic since @xmath302 and @xmath303 for the taylor expansion to be valid .
finally , the microwave photons in the cavity decay with a rate @xmath304 , where @xmath305 is the cavity quality factor . the dissipative term in the master equation describing this decay is given by [ dissterm ] [ ] = 2 .
thus the hamiltonian and the dissipative term are of the same form as those used in to discuss steps 1 , 4 , and 7 .
as was remarked in @xcite , the non - linear cavity cooperativity due to the quadratic coupling is greatly enhanced due to the large scale given by @xmath294)^{1/2}$ ] , many orders of magnitude larger than the typical zero point motion in quantum nanomechanical systems where non - linear couplings are typically very weak @xcite .
let us now discuss the appropriate location for the pick - up coils on the chip such that the steps 1,4 , and 7 of the protocol are passively implemented in the magnetic skatepark .
we choose to place the pick - up coils on the chip _
i.e. _ in the @xmath252 plane .
we consider rectangularly shaped coils with lengths along the @xmath251 and @xmath26-axis given by @xmath306 and @xmath307 respectively .
the magnetic flux threading the pick up coil @xmath308 has a contribution given by the field created by the induced currents of the superconducting sphere and a contribution given by the field created by the wires . with the classical position of the sphere given by @xmath309 and the center of the coil at the origin @xmath310 , the first derivative of the flux respect to the @xmath26-position of the sphere is zero due to the symmetry of the coil and the magnetic field .
thus , to obtain a non - zero linear coupling , @xmath311 , the coil has to be shifted from the center in the @xmath26 direction . restricting to the available space between the inner wires forming the magnetic guide ( @xmath312 as used in the case study )
, the linear coupling @xmath311 can be easily optimized over @xmath313 and @xmath307 resulting in a maximum for @xmath314 and @xmath315 .
on increasing the length @xmath306 of the coil , the coupling increases until @xmath316 and saturates thereafter .
note that the total flux that crosses the coil , which determines the constant flux coefficient @xmath317 entering into the quantum magnetomechanical couplings , depends on its geometrical properties . regarding the quadratic coupling , @xmath318 , we consider a centered coil in @xmath26 , which ensures that the first derivative of the flux with respect to the @xmath26-position of the sphere is zero .
also in this case , both the magnetomechanical coefficient and the constant flux coefficient depend on the dimensions of the coil .
the particular experimental numbers used in the case study are given in the appendix .
we emphasize that the couplings are switched - on and off dynamically by letting the sphere skate over the area where the pick - up coils are located without changing the intensity of the driving fields , something that can be advantageous to minimize decoherence and noise . in the on - chip implementation of the quantum micromechanical interferometer protocol
, potential surface - induced sources of decoherence could play a role .
in particular , as the microsphere couples to the fluctuations of the electromagnetic ( em ) field , both quantum and thermal , we expect its motion to decohere as a consequence .
such decoherence can be studied using the quantum brownian motion ( qbm ) model @xcite with the center of mass motion as the system , coupled linearly to the bath of em field modes . in this section ,
we consider the decoherence induced by the interaction of a diamagnetic particle with the fluctuations of the electromagnetic field near a surface , which could potentially change the density of bath modes to a large extent .
looking at the interaction between a diamagnetic particle and the field as in , in the presence of an externally applied classical magnetic field @xmath265 with fluctuations @xmath319 , one can write the interaction hamiltonian up to bilinear terms as [ hintb ] _
i , where @xmath320 is the classical trajectory of the particle , and @xmath321 stands for the quantum fluctuations of the center of mass position along the @xmath26-axis .
we have ignored here the contribution from the interaction terms of third and higher order in fluctuations of the magnetic field and of the center of mass position .
one can then express the magnetic field fluctuations in the presence of the surface as @xcite where @xmath322_{il } = \epsilon_{ijk}\partial_j [ \bar{\bar{g}}_\lambda(\rr,\rr',\omega)]_{kl}$ ] .
the operators @xmath323 and @xmath324 refer to the bosonic creation and annihilation operators for the medium - assisted field that follow canonical commutation relations , and the coefficients @xmath325 are defined as with @xmath326 and @xmath327 as the space - dependent permittivity and permeability , and @xmath328 as the green s tensor for a particle near a planar semi - infinite surface @xcite . rewriting the interaction hamiltonian as @xmath329 , where one can identify the operator @xmath330 as the bath operator pertaining to the qbm model
, one can readily see that the interaction will yield a pld as the bath monitors the system s position coordinate via the linear coupling .
assuming that the particle is in a harmonic trap of frequency @xmath331 , the decoherence resulting from the quantum brownian motion can be written as @xcite where @xmath332}$ ] stands for the noise correlation function of the thermal bath at temperature @xmath36 . expressing the magnetic field fluctuations in terms of the medium - assisted field operators and simplifying further
, we obtain the decoherence to be where the effective spectral density of the vacuum modes @xmath333 in the presence of a surface is given as @xcite here we have defined the tensor considering the total green s tensor as the sum of free space and scattering components @xmath334 , the free space green s tensor @xmath335 yields the motional decoherence rate of the diamagnet in the absence of the surface and the scattering green s tensor @xmath336 gives the modifications to the decoherence due to scattering off the surface .
we find that for the case of a particle trapped in a harmonic potential of frequency @xmath337 near a perfect conductor , the localization parameter is given by where @xmath338 is the externally applied magnetic field evaluated at the center of mass position of the particle . here
we have assumed that the distance of the particle from the surface is much smaller compared to the trap wavelength @xmath339 , which corresponds to the quasi - static limit as the response of the surface is evaluated at a frequency @xmath340 that is much smaller than @xmath341 .
we find that the decoherence rate for the motion of a particle near a perfectly conducting surface is vanishingly small relative to the gravitationally induced decoherence @xmath342 , using typical experimental numbers , see appendix .
further , observing that a superconductor behaves almost like a perfect conductor in the quasi - static limit @xcite , we find that the localization parameter for a microsphere moving near a superconducting surface is very well approximated by that for a perfect conductor @xmath343 , and is negligibly small as well . for a metal surface with a skin depth @xmath344 , assuming a small trap frequency , such that @xmath345 , one obtains a localization parameter that goes as @xmath346 . for a superconducting sphere of radius @xmath347 , environmental temperature of @xmath348 , and @xmath349 ,
one has that @xmath350 , where @xmath351 is the skin depth of gold .
thus , for typical non - superconducting metals , we find that decoherence due to magnetic field fluctuations near the surface is much larger than gravitationally - induced decoherence .
therefore , to be able to observe gravitationally - induced decoherence , one needs to place the microsphere near a superconducting surface .
that is , one should use a superconducting atom chip @xcite .
the magnetic skatepark implementation of the quantum micromechanical interferometer protocol discussed above is ambitious and therefore experimentally challenging , specially when operating in the regime presented in the case study of .
as discussed in the introduction of the article , we suggest to address this goal from a top - down approach , namely to gradually increase the delocalization distance of a superconducting microsphere brought to the quantum regime .
to do so , one could perform a set of progressively challenging experiments that could be themselves relevant in the the field of quantum nano- and micromechanics @xcite . a possible experimental roadmap could be to show : * magnetic levitation of a superconducting microsphere in vacuum . * magnetomechanical coupling between a superconducting object and a quantum circuit .
the superconducting object does not need to be levitating , see for instance @xcite .
* ground state cooling of a magnetically levitated superconducting sphere using a quantum magnetomechanical coupling by combining the two previous points . *
the preparation of non - trivial quantum states of the center of mass of a magnetically levitated superconducting microsphere using a quantum magnetomechanical coupling to a quantum circuit .
these states still have a small delocalization distance comparable to the zero point motion . * quantum coherent expansion of the center - of - mass motion by making the trap shallower after ground state cooling . * exponential speed - up of the coherent expansion by inverting the quadratic potential after ground state cooling . * the implementation of ground state cooling and coherent expansion as proposed in step 1 , 2 , and 3 of the skatepark . *
perform a squared continuous - time quantum measurement using the coupling to a quantum circuit to prepare a double slit . *
the implementation of the quantum micromechanical interferometer with the 7 steps with small slit separations .
one can then gradually increase the slit separation distance .
note that this roadmap would also allow to carefuly characterize the standard sources of decoherence arising from scattering of air molecules , vibrations , and magnetic field fluctuations , as well as other sources of decoherence that may have been overlooked .
in this section we consider a particular set of experimental parameters and analytically calculate the interference pattern that would be obtained in the proposed quantum micromechanical interferometer .
this is done considering standard sources of decoherence and with ( red curve in ) or without ( grey curve in ) gravitationally - induced decoherence . with the parameters explicitly listed in the appendix , one obtains the interference pattern plot in , indicating that the faint parameter - free gravitationally - induced decoherence could be unambiguously falsified .
we emphasize that the set of parameters that we chose , see appendix , has not been fine - tuned .
it is only a representative example and we anticipate that there would be other sets of parameters that would lead to the same conclusion , namely that gravitationally - induced decoherence could be falsified .
of course , some of the important parameters have been chosen to ensure that the interferometer protocol is faithfully implemented . in the rest of this section
we discuss and justify the same .
we have considered a superconducting sphere in the meissner state with a radius of @xmath39 and a mass of @xmath352 , well inside the gravitational quantum regime , as shown in .
the superposition size is determined by the slit separation @xmath353 and slit width @xmath354 .
the total time of the quantum micromechanical interferometer after cooling ( which takes @xmath355 ) is @xmath356 .
this time is distributed between the different steps as @xmath357 , @xmath358 , @xmath359 , @xmath360 , @xmath361 , and @xmath362 .
the harmonic potentials in steps 2 , 5 , and 7 have a frequency of @xmath363 . note that the longest duration in the operation of the interferometer is in the free expansion of step 3 , where standard sources of decoherence are minimal .
the coherence length after step 3 is given by @xmath364 without gravitationally - induced decoherence , and @xmath365 including it .
hence note that the condition @xmath195 is fulfilled .
the latter is the key condition in order to falsify gravitationally - induced decoherence , as shown in .
steps 4 , 5 , 6 , and 7 are intended to build the interference pattern , in case of a coherent superposition , and thus have to be performed as fast as possible in order to prevent decoherence from washing - out the interference pattern . with the parameters used ,
the blurring distance is given by @xmath366 with the presence of gravitationally - induced decoherence and @xmath367 with only standard sources of decoherence .
note that the difference is minimal and the blurring is small enough to not wash - out the interference fringes generated when gravitationally - induced decoherence is absent , since the fringe separation takes the value @xmath368 for the chosen parameters .
a larger mass of the sphere has not been considered mainly because of two reasons : ( i ) the magnetic field gradients are limited by the fact that the field at the surface of the sphere has to be smaller than the first critical field of the superconducting material .
the larger the mass , the smaller the field gradients and thus the smaller the trap frequency in step 1 , which then prevents one from achieving the resolved - sideband regime and thus preparing a sufficiently pure initial state .
namely , we require the initial thermal mean phonon number occupation to be smaller than 1 , and @xmath369 for our parameter choice .
( ii ) the larger the mass , the more vibration isolation is required to ensure that gravitationally - induced decoherence is the main source of decoherence .
we have already considered a very good isolation vibration with values of @xmath370 at frequencies below @xmath371 , as required in steps 3 , 4 , and 7 , and @xmath372 at frequencies of @xmath363 , as required in steps 2,5 and 6 . in a cryogenic environment
these values are challenging , and comparable to vibration isolation in present - day gravitational wave detectors .
nevertheless , we remark that in future cryogenic gravitational wave detectors , schemes for achieving vibration isolation at the incredible numbers of @xmath373 at low frequencies are being developed , see for instance the kagra project @xcite .
vibrations are the main source of decoherence in our model provided that the decoherence from magnetic field fluctuations can be ignored .
the latter is the case if the surface is superconducting , for example , if one uses a superconducting atom chip @xcite .
note that the sphere is skating at a distance from the surface of the order of @xmath374 .
the distance slightly varies in each step of the protocol , something that we have not optimized since the smooth transitions between steps have not been included in our calculations .
should the decoherence from the surface be a hindering factor , one could also consider implementing the interferometer protocol in free fall @xcite with external magnetic potentials speeding - up the dynamics , as in step 2 and 6 of the protocol . as mentioned before
, we refer to the appendix for the complete list of experimental parameters used to analytically calculate , as an example , the interference pattern shown in .
we have proposed and analyzed a quantum micromechanical interferometer scheme for preparing and probing quantum superpositions of a sphere in the micrometer scale with a mass @xmath375 .
the quantum superposition state is prepared by using a young s double slit with slit separation comparable to the radius of the sphere . in particular , we have proposed an all - magnetic on - chip scheme where a superconducting microsphere skates in a magnetic landscape generated by persistent currents .
the chip also houses some squids that are used to cool , implement a double slit , and measure final interference fringes . as the sphere skates through such a magnetic skatepark an all - magnetic young s double slit experiment
is implemented passively , namely without the need to switch on and off magnetic potentials and to actively control the quantum circuits . using a feasible set of experimental parameters as a case study
, we have shown that one can enter into the gravitational quantum regime where the faint parameter free gravitationally - induced collapse model , proposed by disi and penrose @xcite , could be unambiguously falsified .
while we have focused on this paradigmatic parameter - free collapse model as a figure of merit , we remark that our proposal would falsify the continuous - spontaneous - localization collapse model @xcite with a collapse frequency phenomenological parameter as small as @xmath376 , several orders of magnitude below the original value proposed by ghirardi , rimini , and weber in the 80 s @xcite . in our analysis , especially regarding the experimental feasibility , we have modelled the experiment with some necessary degree of simplification .
for example , we have not considered the many potential sources of noise and decoherence that could be encountered in a real experimental implementation .
instead , we have analyzed the aspects that we consider unavoidable in any experimental implementation and that would pose more than a technological challenge . while the experimental implementation of our proposal is ambitious and therefore challenging , we have nevertheless argued that it could be addressed by a set of progressively challenging experiments that would be themeselves relevant in the field of quantum nano- and micromechanics .
we believe that in general , experiments aiming to explore the quantum superposition principle of very massive objects would benefit from the conditions that we have analysed and determined to be crucial for this proposal : ( i ) a cryogenic environment that minimizes decoherence due to black - body radiation and due to scattering of air molecules because of the ultra - high vacuum present in such environments .
we note , for example , that pressures less than @xmath377 mbar have been achieved in cryogenic environments @xcite , ( ii ) the possibility to use static fields ( ideally using persistent currents ) to generate inverted potentials that speed - up the otherwise very slow quantum dynamics , ( iii ) a very good vibration isolation to avoid decoherence when using external potentials , ( iv ) superconducting surfaces in case of near - surface proposals since low frequency electromagnetic field fluctuations are absent , and ( v ) the use of quantum circuits to cool , to engineer non - linearities , and to measure the position of the massive object .
we remark that one of the main ingredients to perform quantum interference of such massive objects in less than a second is the use of inverted potentials that speed - up the quantum dynamics .
this idea could also be implemented in a free fall experiment that would require a cryogenic high - vacuum tower of few meters @xcite , employing the same protocol as described here . in our protocol
we propose to perform the young s double slit by performing a quantum measurement of the squared position . alternatively
, one could also consider preparing the quantum superposition using a non - gaussian ( more than quadratic dependence in position ) magnetic potential .
also , one could consider a magnetic grating formed , for instance , using the field created by an array of pinned superconducting vortices or nanomagnets .
if the magnetic lattice has field maxima larger than the critical field of the sphere , superconductivity is prevented for spheres passing through these areas .
such a magnetic grating could be used for near - field matter - wave interferometry , as an all - magnetic version of the all - optical talbot - lau matter - wave interferometers currently being implemented @xcite . in this all - magnetic scenario
, one could again use static fields to speed - up the dynamics without creating decoherence and thus increase the mass of the particle used in near - field matter - wave interferometers without requiring a space environment @xcite .
our proposal could also be extended to exploit the other interpretations of the gravitational quantum regime .
in particular , one could consider two cold superconducting spheres moving along parallel magnetic waveguides and measure the mutual gravitational interaction which can then be used to determine @xmath6 .
that is a levitation version of a recently proposed experiment for measuring @xmath6 with milligram masses @xcite .
one could also consider entangling the levitated spheres using the gravitational interaction .
we hope that the proposal presented in this manuscript , namely an earth - based on - chip all - magnetic quantum micromechanical interferometer exploring quantum mechanics in a hitherto unexplored parameter regime , will trigger further table - top experimental and theoretical research on the interplay between quantum mechanics and gravity .
this work is supported by the european research council ( erc-2013-stg 335489 qsupermag ) and the austrian federal ministry of science , research , and economy ( bmwfw ) .
ori acknowledges inspiring discussions with m. aspelmeyer , j. i. cirac , and s. dimopoulos .
in this section we provide a list of experimental parameters . we emphasize that the choice of parameters is not optimized , it is simply a representative example .
this set of parameters has been used to plot .
we label the free parameters that we have chosen with the item symbol @xmath378 .
the parameters that are not free , either because they are calculated from the chosen free parameters or because they are fixed by the choice ( mass density once we have chosen the material ) , are labeled with the item symbol @xmath379 .
* scattering of air molecules @xmath389 .
* black - body decoherence due to scattering is 9.4 10 ^ -27 ^2 * black - body decoherence due to absorption and emission is 2.1 10 ^ -15 * localization parameter due to vibrations in a harmonic potential of frequency @xmath390 is 1.4 10 ^ 10 * position pick - up coil in step 1 : @xmath394 and @xmath395 . *
size pick - up coil in step 1 : @xmath396 and @xmath397 .
* position pick - up coil in step 4 and 7 : @xmath398 . *
size pick - up coil in step 4 and 7 : @xmath399 and @xmath400 .
* distance outer wires @xmath401 * trap distance @xmath402 and frequency @xmath403 along the @xmath256-axis .
* attractive potential frequency @xmath404
. * cavity frequency @xmath405 . * cavity decay rate @xmath406
. * constant flux coefficient @xmath407
. * magnetomechanical coefficient @xmath408 .
* linear magnetomechanical single - photon coupling @xmath409 . * intra - cavity mean - field photon number @xmath410 . *
detuning of the driving field @xmath411 .
* adiabatic elimination condition @xmath412 .
* cooling time @xmath355 . *
initial mean number of phonons @xmath413 .
* final mean number of phonons @xmath414 .
* coherence length @xmath415 .
* distance outer wires @xmath416 * trap distance @xmath417 and frequency @xmath418 along the @xmath256-axis .
* inverted potential frequency @xmath419 . *
boosting time @xmath357 .
* momentum gain @xmath420
. * standard vs gravitational - induced decoherence @xmath421 .
* position variance without @xmath32 : @xmath422 . * coherence length without @xmath32 : @xmath423 . * coherence length with @xmath32 : @xmath423 .
* distance outer wires @xmath424 * trap distance @xmath425 and frequency @xmath426 along the @xmath256-axis .
* attractive potential frequency @xmath427 .
* free time @xmath358 .
* position variance without @xmath32 : @xmath428 .
* position variance with @xmath32 : @xmath428 . * coherence length without @xmath32 : @xmath429 . * coherence length with @xmath32 : @xmath430 .
* distance outer wires @xmath424 * trap distance @xmath425 and frequency @xmath426 along the @xmath256-axis . * length unit @xmath431
. * cavity frequency @xmath432 . * cavity decay rate @xmath433
. * constant flux coefficient @xmath434
. * magnetomechanical coefficient @xmath435 .
* quadratic magnetomechanical single - photon coupling @xmath436 .
* intra - cavity mean - field photon number @xmath437 .
* adiabatic elimination condition @xmath438 .
* measurement time @xmath359 .
* initial phase @xmath439 .
* final phase @xmath440 . * measurement strength @xmath441 . * slit separation @xmath442 . *
slit width @xmath443 .
* distance outer wires @xmath444 * trap distance @xmath445 and frequency @xmath446 along the @xmath256-axis .
* attractive potential frequency @xmath447 . * rotation time @xmath448 . * standard vs gravitational - induced decoherence @xmath449 . * convolution factors that justify neglecting decoherence @xmath450 and @xmath451 .
* distance outer wires @xmath416 * trap distance @xmath417 and frequency @xmath418 along the @xmath256-axis .
* inverted potential frequency @xmath452 .
* inflation time @xmath361 .
* standard vs gravitational - induced decoherence @xmath453 .
* blurring coefficient without @xmath32 : @xmath454 .
* blurring coefficient with @xmath32 : @xmath455 . *
fringe separation at @xmath242 : @xmath456 * distance outer wires @xmath424 * trap distance @xmath425 and frequency @xmath426 along the @xmath256-axis . *
length unit without @xmath32 : @xmath457 . * length unit with @xmath32 : @xmath458 .
* cavity frequency @xmath459 . * cavity decay rate @xmath460
. * constant flux coefficient @xmath434
. * magnetomechanical coefficient @xmath435 .
* quadratic magnetomechanical single - photon coupling @xmath461 .
* intra - cavity mean - field photon number @xmath462 . *
adiabatic elimination condition @xmath463 .
* measurement time @xmath362 . * measurement strength @xmath464
. * position resolution @xmath465 .
* total protocol time after cooling @xmath466 .
m. arndt , o. nairz , j. vos - andreae , c. keller , g. van der zouw , and a. zeilinger , _ wave particle duality of c60 molecules _ , http://www.nature.com/nature/journal/v401/n6754/abs/401680a0.html[nature ( london ) * 401 * , 680 ( 1999 ) . ]
k. hornberger , s. gerlich , p. haslinger , s. nimmrichter , and m. arndt , _ colloquium : quantum interference of clusters and molecules _ , http://journals.aps.org/rmp/abstract/10.1103/revmodphys.84.157[rev .
* 84 * , 157 ( 2012 ) . ]
a. d. cronin , j. schmiedmayer , and d. e. pritchard , _ optics and interferometry with atoms and molecules _ , http://journals.aps.org/rmp/abstract/10.1103/revmodphys.81.1051[rev .
* 81 * , 1051 ( 2009 ) . ]
s. eibenberger , s. gerlich , m. arndt , m. mayor , and j. txen , _ matter wave interference of particles selected from a molecular library with masses exceeding 10 000 amu _ , http://pubs.rsc.org/en/content/articlelanding/2013/cp/c3cp51500a#!divabstract[phys .
chem . phys . *
15 * , 14696 ( 2013 ) . ]
j. schmle , m. dragosits , h. hepach , m. aspelmeyer , _ a micromechanical proof - of - principle experiment for measuring the gravitational force of milligram masses _ , http://arxiv.org/abs/1602.07539[arxiv:1602.07539 . ] c. anastopoulos and b. l. hu , _ probing a gravitational cat state _ , http://iopscience.iop.org/article/10.1088/0264-9381/32/16/165022/meta;jsessionid=a483988d12959a94fbd046ef68e1f306.c2.iopscience.cld.iop.org[classical quant
. grav . *
32 * , 165022 ( 2015 ) . ]
j. d. teufel , t. donner , d. li , j. w. harlow , m. s. allman , k. cicak , a. j. sirois , j. d. whittaker , k. w. lehnert , and r. w. simmonds , _ sideband cooling of micromechanical motion to the quantum ground state _ , http://www.nature.com/nature/journal/v475/n7356/full/nature10261.html[nature ( london ) * 475 * , 359 ( 2011 ) .
] j. chan , t. p. mayer alegre , a. h. safavi - naeini , j. t. hill , a. krause , s. grblacher , m. aspelmeyer , and o. painter , _ laser cooling of a nanomechanical oscillator into its quantum ground state _ , http://www.nature.com/nature/journal/v478/n7367/abs/nature10461.html[nature ( london ) * 478 * , 89 ( 2011 ) . ]
o. romero - isart , m. l. juan , r. quidant , and j. i. cirac , _ toward quantum superposition of living organisms _ , http://iopscience.iop.org/article/10.1088/1367-2630/12/3/033015/meta[new j. phys .
* 12 * , 033015 ( 2010 ) . ]
d. e. chang , c. a. regal , s. b. papp , d. j. wilson , j. ye , o. painter , h. j. kimble , and p. zoller , _ cavity opto - mechanics using an optically levitated nanosphere _ , http://www.pnas.org/content/107/3/1005[proc .
u. s. a. * 107 * , 1005 ( 2010 ) . ]
n. kiesel , f. blaser , u. deli , d. grass , r. kaltenbaek , and m. aspelmeyer , _ cavity cooling of an optically levitated submicron particle _ , http://www.pnas.org/content/110/35/14180.abstract[proc .
u. s. a. * 110 * , 14180 ( 2013 ) . ]
p. asenbaum , s. kuhn , s. nimmrichter , u. sezer , and m. arndt , _ cavity cooling of free silicon nanoparticles in high vacuum _ , http://www.nature.com/ncomms/2013/131106/ncomms3743/full/ncomms3743.html[nat
. commun . * 4 * , 2743 ( 2013 ) . ] j. millen , p. z. g. fonseca , t. mavrogordatos , t. s. monteiro , and p. f. barker , _ cavity cooling a single charged levitated nanosphere _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.114.123602[phys .
lett . * 114 * , 123602 ( 2015 ) . ] j. gieseler , b. deutsch , r. quidant , and l. novotny , _ subkelvin parametric feedback cooling of a laser - trapped nanoparticle _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.109.103603[phys .
* 109 * , 103603 ( 2012 ) . ]
o. romero - isart , a. c. pflanzer , f. blaser , r. kaltenbaek , n. kiesel , m. aspelmeyer , and j. i. cirac , _ large quantum superpositions and interference of massive nanometer - sized objects _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.107.020405[phys .
lett . * 107 * , 020405 ( 2011 ) . ] j. bateman , s. nimmrichter , k. hornberger , and h. ulbricht , _ near - field interferometry of a free - falling nanoparticle from a point - like source _ , http://www.nature.com/ncomms/2014/140902/ncomms5788/full/ncomms5788.html[nat . commun . * 5 * , 4788 ( 2014 ) . ] r. kaltenbaek , g. hechenblaikner , n. kiesel , o. romero - isart , k. c. schwab , u. johann , and m. aspelmeyer , _ macroscopic quantum resonators ( maqro ) _ , http://link.springer.com/article/10.1007%2fs10686-012-9292-3[exp
. astron . * 34 * , 123 ( 2012 ) . ] o. romero - isart , l. clemente , c. navau , a. sanchez , and j. i. cirac , _ quantum magnetomechanics with levitating superconducting microspheres _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.109.147205[phys .
* 109 * , 147205 ( 2012 ) . ]
g. gabrielse , x. fei , l. a. orozco , r. l. tjoelker , j. haas , h. kalinowsky , t. a. trainor , and w. kells , _ thousandfold improvement in the measured antiproton mass _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.65.1317[phys .
lett . * 65 * , 1317 ( 1990 ) . ] g. gabrielse , _ comparing the antiproton and proton , and opening the way to cold hydrogen _ , http://www.sciencedirect.com/science/article/pii/s1049250x01800379[advances in atomic , molecular , and optical physics * 45 * , 1 ( 2001 ) . ] i. wilson - rae , n. nooshi , w. zwerger , and t. j. kippenberg , _ theory of ground state cooling of a mechanical oscillator using dynamical backaction _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.99.093901[phys .
* 99 * , 093901 ( 2007 ) . ]
f. marquardt , j. p. chen , a. a. clerk , and s. m. girvin , _ quantum theory of cavity - assisted sideband cooling of mechanical motion _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.99.093902[phys .
* 99 * , 093902 ( 2007 ) . ]
c. genes , d. vitali , p. tombesi , s. gigan , and m. aspelmeyer , _ ground - state cooling of a micromechanical oscillator : comparing cold damping and cavity - assisted cooling schemes _ , http://journals.aps.org/pra/abstract/10.1103/physreva.77.033804[phys .
a * 77 * , 033804 ( 2008 ) . ]
i. wilson - rae , n. nooshi , j. dobrindt , t. j. kippenberg , and w. zwerger , _ cavity - assisted backaction cooling of mechanical resonators _ , http://iopscience.iop.org/article/10.1088/1367-2630/10/9/095007/meta[new
. j. phys .
* 10 * , 095007 ( 2008 ) . ]
m. cirio , g. k. brennen , and j. twamley , _ quantum magnetomechanics : ultrahigh - q - levitated mechanical oscillators _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.109.147206[phys .
lett . * 109 * , 147206 ( 2012 ) . ] g. via , g. kirchmair , and o. romero - isart , _ strong single - photon coupling in superconducting quantum magnetomechanics _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.114.143602[phys .
* 114 * , 143602 ( 2015 ) . ]
s. minniberger , f. diorico , s. haslinger , c. hufnagel , c. novotny , n. lippok , j. majer , c. koller , s. schneider , and j. schmiedmayer , _
magnetic conveyor belt transport of ultracold atoms to a superconducting atomchip _ , http://link.springer.com/article/10.1007%2fs00340-014-5790-5[app .
b * 116 * , 1017 ( 2014 ) . ] j. h. thywissen , m. olshanii , g. zabow , m. drndic , k. s. johnson , r. m. westervelt , and m. prentiss , _ microfabricated magnetic waveguides for neutral atoms _ , http://epjd.epj.org/articles/epjd/abs/1999/10/d8303/d8303.html[eur .
j. d * 7 * , 361 ( 1999 ) . ] j. koch , t. m. yu , j. gambetta , a. a. houck , d. i. schuster , j. majer , a. blais , m. h. devoret , s. m. girvin , and r. j. schoelkopf , _ charge - insensitive qubit design derived from the cooper pair box _ , http://journals.aps.org/pra/abstract/10.1103/physreva.76.042319[phys . rev .
a * 76 * , 042319 ( 2007 ) . ] c. tokoku , _ cryogenic system for the interferometric cryogenic gravitationalwave telescope , kagra - design , fabrication , and performance test - _ , http://scitation.aip.org/content/aip/proceeding/aipcp/10.1063/1.4860850[aip conf .
proc . * 1573 * , 1254 ( 2014 ) . ] b. k. skagerstam , u. hohenester , a. eiguren , and p. k. rekdal , _ spin decoherence in superconducting atom chips _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.97.070401[phys .
* 97 * , 070401 ( 2006 ) . ]
t. nirrengarten , a. qarry , c. roux , a. emmert , g. nogues , m. brune , j .-
raimond , and s. haroche , _ realization of a superconducting atom chip _ , http://journals.aps.org/prl/abstract/10.1103/physrevlett.97.200405[phys .
lett . * 97 * , 200405 ( 2006 ) . ]
a. bassi , k. lochan , s. satin , t. p. singh , and h. ulbricht , _ models of wave - function collapse , underlying theories , and experimental tests _ , http://journals.aps.org/rmp/abstract/10.1103/revmodphys.85.471[rev .
phys . * 85 * , 471 ( 2013 ) . ]
p. haslinger , n. drre , p. geyer , j. rodewald , s. nimmrichter , and m. arndt , _ a universal matter - wave interferometer with optical ionization gratings in the time domain _ , http://www.nature.com/nphys/journal/v9/n3/abs/nphys2542.html[nat
* 9 * , 144 ( 2013 ) . ]
l. disi , _ notes on certain newton gravity mechanisms of wavefunction localization and decoherence _ , http://iopscience.iop.org/article/10.1088/1751-8113/40/12/s07/meta;jsessionid=6a3176bc7bf382793f5870bcaaef0da7.c4.iopscience.cld.iop.org[j .
a : math . theor . * 40 * , 2989 ( 2007 ) ] . | we propose and analyze an all - magnetic scheme to perform a young s double slit experiment with a micron - sized superconducting sphere of mass @xmath0 amu .
we show that its center of mass could be prepared in a spatial quantum superposition state with an extent of the order of half a micrometer .
the scheme is based on magnetically levitating the sphere above a superconducting chip and letting it skate through a static magnetic potential landscape where it interacts for short intervals with quantum circuits . in this way a protocol for fast quantum interferometry using quantum magnetomechanics is passively implemented .
such a table - top earth - based quantum experiment would operate in a parameter regime where gravitational energy scales become relevant .
in particular we show that the faint parameter - free gravitationally - induced decoherence collapse model , proposed by disi and penrose , could be unambiguously falsified .
the ambitious goal of the proposal could be experimentally addressed by implementing a set of intermediate , progressively challenging , experiments that might be themselves relevant in the field of quantum miromechanics . |
finding the first stars and galaxies remains a major observational challenge in astrophysics .
searches for metal - free ( pop iii ) or extremely metal - poor individual stars or for ensembles of stars ( clusters , proto - galaxies , populations of galaxies , etc . )
are ongoing both nearby ( pop iii stars in the halo of our galaxy ) , and in galaxies out to the highest redshifts currently known .
astronomers have been quite inventive in searching for pop iii or metal - poor stars , exploiting many possible direct and indirect signatures ; some of these methods will be discussed below . a more detailed account on primeval galaxies , some of the physics related to these objects , an overview on searches etc .
is presented in the lectures of schaerer ( 2007 ) .
although some studies may have found signatures of pop iii stars , none of those claims is very strong ( see discussion below ) , and my personal opinion taking a somewhat conservative attitude is that pop iii still remain to be found .
however , there is good hope that this goal should be reached in the quite near future with current or forthcoming instrumentation , as i will sketch below .
many authors have modeled the evolution of pop iii stars in the 1980s and before . since
then the interest in these stars has been revived , and new generations of models computed .
subsequently i will use predictions from the stellar evolution and non - lte atmosphere models as well as the evolutionary synthesis models computed by schaerer ( 2002 , 2003 ) . for
a detailed description of the input physics and references to earlier work the reader is referred to these papers .
one of the main distinctive features of pop iii stars is their extreme compactness close to and on the zero age main sequence ( zams ) , implying much higher effective temperatures ( up to @xmath3 100 kk ) than usual . for a given mass
this leads to a considerable increase of the ionising flux and to a much harder ionising spectrum ( cf .
tumlinson et al .
2001 , schaerer 2002 ) .
in addition , if the conditions in metal - poor environments favour the formation of massive stars ( cf .
bromm & larson 2004 , and contributions by abel , bromm , tan , and others in these proceedings ) , the integrated spectrum of young , zero and very low metallicity stellar populations can be quite different than that of populations at `` normal '' metallicities , as shown e.g. by schaerer ( 2002 , 2003 ) .
in particular , the strong ly@xmath0 line emission and the presence of nebular he ii emission lines are probably the best / easiest features to search for very metal - poor and pop iii objects .
for example in figure 1 ( left ) we show the increase of the ionising ( lyman continuum ) photon flux @xmath4 normalised to the uv continuum luminosity with decreasing metallicity for different imfs ( salpeter , scalo ; b , c as in schaerer 2003 , and other imfs ) . for a fixed imf
the increase from solar to zero metallicity is typically a factor of 2 ; assuming that more extreme imfs may be valid for pop iii , this increase can be up to a factor @xmath3 10 ( compared to solar and salpeter ) .
note that this prediction also depends on the exact star - formation ( sf ) history , which affects in particular the uv continuum output .
the right panel of fig . 1 shows the increase of the hardness of the ionising flux , expressed by the ratio of he@xmath5 to h ionising flux , @xmath6 , with decreasing metallicity .
as already mentioned , the strong lyman continuum flux at low metallicities implies a strong intrinsic ly@xmath0 emission ( cf .
schaerer 2002 , 2003 ) .
for example , the _ maximum ly@xmath0 equivalent width
_ predicted for an integrated stellar population increases from @xmath3 200 - 300 to 500 - 1000 or even larger ( depending on the imf ) from solar to zero metallicity , as shown in fig .
of course it must be reminded that after its emission inside the ionised region surrounding the starburst ly@xmath0 photons are scattered and may be absorbed , thereby altering the intrinsic value of @xmath7 ( see e.g. schaerer 2007 ) .
the increase of the hardness of the ionising radiation shown in fig .
1 ( right ) translates into the increase of he ii recombination lines , such as he ii @xmath11640 , shown in fig . 2 ( right ) .
for obvious reasons their strength depends on the exact sf history , age , and in particular on the imf , as shown by these figures ( see also schaerer 2003 ) . while clearly a promising signature of very metal - poor populations as pointed out earlier1640 emission can also of different origin ( stellar cf .
3.1 , agn or shocks ) and nebular he ii is observed in some normal h ii regions ( see e.g. thuan & izotov 2005 ) or in other peculiar objects ( cf .
fosbury et al .
2003 and discussion in schaerer 2003b ) .
] , this strong sensitivity to the presence of the most massive stars
i.e. to the imf and age may complicate the interpretation of he ii lines , especially when dealing with non - detections ( cf . below ) .
once the properties of individual pop iii clusters or galaxies are known , important questions are of course how many such objects are expected and what is their distribution , e.g. in mass and luminosity ? answering these questions
requires in particular a description of the transport and mixing of metals to determine the duration ( locally ) of pop iii sf events .
several authors have addressed this using different approaches ( e.g. scannapieco et al .
2003 , yoshida et al . 2004 , tornatore et al . 2007 ) . the predicted pop iii sf rate , defined by sf in regions with @xmath8 , as a function of redshift from the recent computations of tornatore et al . ( 2007 )
is shown in fig .
3 . predicted number counts from choudhury & ferrara ( 2007 ) are shown in fig . 5 for illustration .
it is possible that galaxies containing pop iii or very metal - poor stars have already been found in different samples . here
i will critically examine some of these results .
for example , jimenez & haiman ( 2006 ) have recently proposed an explanation to several apparent puzzles concerning lyman break galaxies ( lbgs ) and related objects at @xmath9 34 .
they propose that their stellar populations contain @xmath3 1030 % of primordial stars , which would explain the observed he ii @xmath11640 emission in lbgs , the existence of some galaxies with very high w(ly@xmath0 ) , the excess of lyman continuum flux seen in some lbgs , and the nature of ly@xmath0 blobs .
although the idea to tackle simultaneously four problems is in principle attractive , the proposed solution is probably not unique and not the most likely one , and some of the 4 problems may not be robust observationally . first , the strength of the he ii @xmath11640 line in the composite spectrum of lbgs is quite modest ( @xmath10 ) , and small enough to be explained by wolf - rayet stars in normal stellar populations ( cf .
models of schaerer & vacca 1998 , and brinchmann et al .
furthermore the observed line is broad , as observed in wolf - rayet stars .
second , the existence of a population of galaxies at @xmath9 45 with high ly@xmath0 equivalent widths found by the lala survey , does not seem to be confirmed by other groups or is at least very uncertain ( see sect .
third , other explanations are proposed for extended ly@xmath0 blobs ( e.g. matsuda et al .
finally , the proposed explanation may also have difficulties with mixing time scales ( pan & scalo 2007 ) .
the excess of lyman continuum flux in some lbgs remains still puzzling ( see also shapley et al .
2006 , iwata et al .
2008 ) a few years ago malhotra & rhoads ( 2002 ) found ly@xmath0 emitters ( lae ) at @xmath11 with an unusual ly@xmath0 equivalent width distribution from their lala survey .
they suggested that their large fraction of objects with a high w(ly@xmath0 ) ( @xmath12 200 - 300 ) could be due to very metal - poor or pop iii objects , a very unusual imf , or to agn .
follow - up observations of these objects have been undertaken , also at x - rays , and several papers have already adressed these results ( see discussion in schaerer 2007 ) . the agn hypothesis has been rejected ( wang et al .
deep spectroscopy aimed at detecting other emission lines , including the he ii @xmath11640 line indicative of a pop iii contribution ( cf .
sect . 4 ) , have been unsuccessful ( dawson 2004 ) , although the achieved depth may not be sufficient .
if taken at face value , the origin of these high w(ly@xmath0 ) remains thus unclear .
however , there is some doubt on the reality of these equivalent widths measured from nb and broad - band imaging , or at least on them being so numerous even at @xmath13 .
first of all the objects with the highest @xmath14 have very large uncertainties since the continuum is faint or non - detected .
second , the determination of @xmath14 from a nb and a centered broad - band filter ( @xmath15-band in the case of malhotra & rhoads 2002 ) may be quite uncertain , e.g. due to unknowns in the continuum shape , the presence of a strong spectral break within the broad - band filter etc .
( see hayes & oestlin 2006 for a quantification , and shimasaku 2006 ) .
furthermore other groups have not found such high @xmath7 objects ( e.g. hu 2004 , ajiki 2003 and compilation in verhamme et al .
2008 ) suggesting also that this may be related to insufficient depth of the lala photometry . recently shimasaku ( 2006 ) and dijkstra & wyithe ( 2007 ) have again , used @xmath14 distributions to argue for a non - negligible contribution of pop iii stars in laes , this time in objects at @xmath16 .
in fact the _ observed _ restframe values of @xmath14 lae samples at @xmath17 and 6.5 obtained with subaru are considerably lower than those of malhotra & rhoads at @xmath13 , and only few objects show @xmath18 , the approximate limit expected for `` normal '' stellar populations ( cf .
fig . 2 ) .
however , if the igm transmission @xmath19 affects half of the ly@xmath0 line , as often assumed ( see also hu et al .
2004 ) , this would imply true intrinsic equivalent widths @xmath20 times higher than the observed value , i.e. typically 4 times higher at @xmath17 ! for this reason
shimasaku suggest that these may be young galaxies or objects with pop iii contribution .
based on this reasoning and on modeling the ly@xmath0 luminosity function and igm transmission , dijkstra & wyithe ( 2007 ) argue for the presence of @xmath3 410 % of pop iii sf in @xmath3 half of ly@xmath0 selected galaxies and in few percent of i - drop galaxies at @xmath21 5.7 .
although possible , this conclusion rests strongly on the assumption that the igm transmission ( due to individual or overlapping ly@xmath0 forest clouds ) really cuts out a significant fraction of the ly@xmath0 line emerging from the galaxy .
this is by no means clear , e.g. since it implies the need for cold gas close ( and/or infalling ) to the galaxies , which has not been observed yet .
the lbg cb58 studied in depth , savaglio et al .
( 2002 ) find no ly@xmath0 absorption close to the galaxy ( at @xmath22 2000 km s@xmath23 ) .
also verhamme et al .
( 2008 ) find no need for igm absorption in their ly@xmath0 profile modeling of lbg / lae at @xmath9 35 .
furthermore , absorbing a fraction of ly@xmath0 by the igm is even more difficult if the intrinsic ly@xmath0 profile emerging from the galaxy is already redshifted ( and asymmetric ) , as expected if outflows are as ubiquitous as in lbgs at lower redshift .
assessing these important issues remains to be done . in the meantime conclusions depending strongly on corrections to the observed ly@xmath0 equivalent widths ( or lf )
should probably be taken with caution .
several groups have tried to use the expected he ii @xmath11640 emission ( cf .
2 ) to search for pop iii stars , so far with no positive detection .
however , the limits obtained from most recent survey begin to provide interesting constraints .
follow - up spectroscopy of lala sources by dawson et al .
( 2004 ) ( cf .
3.2 ) have yielded an upper limit of @xmath24 25 ( 3 @xmath25 ) and he ii @xmath11640/ly@xmath0@xmath26 0.130.20 at 23 @xmath25 from their composite spectrum ( 11 objects ) .
nagao ( 2005 ) searched for he ii @xmath11640 with deep spectroscopy of one strong ly@xmath0 emitter at @xmath27 .
currently the lowest limit on he ii emission is that measured by ouchi et al.(2008 ) in composite spectra of 36 and 25 @xmath28 and 3.7 laes .
they reach @xmath29he ii @xmath11640@xmath30ly@xmath0@xmath31 0.02 and 0.06 ( @xmath32 ) at these two redshifts , which translates to @xmath33 and @xmath34 respectivelyhe ii @xmath11640@xmath30ly@xmath0@xmath35 ( cf .
s03 ) . ] .
although no definite conclusions can be drawn from that , comparison with fig .
1 ( right ) show that these limits are already close to the maximum or slightly below the predicted values for metal - free populations with imfs including very massive stars . using an original approach combining narrow / intermediate - band filters to select dual ly@xmath0 + he ii @xmath11640 emitters at @xmath36 and 4.6 ,
nagao et al . (
2008 ) have carried out a survey of 875 arcmin@xmath37 using subaru in the subaru deep field ( see also nagao , these proceedings , for more details ) .
no such dual emitters were found , down to a flux limit of @xmath38 erg s@xmath23 @xmath39for he ii @xmath11640 . with some assumptions on the imf and using the models of s03
, the observed flux limits translate to an upper limit for the pop iii sfr of @xmath40 m@xmath41 yr@xmath23 , and a sfr density of @xmath42 m@xmath41 yr@xmath23 mpc@xmath43 at @xmath44 ( nagao et al . 2008 ) .
interestingly this first pop iii sfr density determination turns out to be very close to the theoretical prediction of tornatore et al .
( 2007 ) shown in fig .
clearly further work and deeper observations are required to track the elusive population iii at high - redshift and future theoretical work may provide a more detailed / accurate picture of the expectations .
the impressive depth down to few times
@xmath46 erg s@xmath23 @xmath39 reached already in emission line searches with subaru , vlt , and keck is illustrated in fig . 4 , where ly@xmath0 measurements and upper limits obtained in the visible and near - ir domain are compiled .
in the case of lensed galaxies the intrinsic depths can be considerably larger , as e.g. shown for the candidate high - z galaxies of stark et al .
( 2007 ) . in several cases of @xmath45 galaxies or candidates , spectroscopic follow - up observations
have already been undertaken to search for various emission lines , including the potential pop iii indicator he ii @xmath11640 discussed above ( e.g. pell et al . 2005 , stark et al .
2007 , richard et al .
2008 , in preparation ) .
since the average metallicity of galaxies must decrease with increasing redshift , searches for primeval galaxies and/or pop iii galaxies are also naturally focussed towards the highest redshift . at the same time galaxy searches at @xmath45
are also of great interest to identify the sources of cosmic reionisation . reviewing the status of galaxy searches / observations at @xmath45 is beyond the scope of the present contribution .
for an overview see e.g. schaerer ( 2007 and references therein ) .
the most recent results from searches for @xmath47 galaxies are presented by bouwens et al .
( 2008 ) and richard et al .
( 2008 ) , who use surveys in blank fields or fields with massive galaxy clusters benefiting from strong gravitational lensing . at the bright end of the lf the number density of galaxies at @xmath48 , and hence also the total sfr density , remains controversial ( cf .
richard et al .
2006 , bouwens et al .
2008 ) . at fainter magnitudes ( @xmath49 27 ) the results from these two approaches ( blank fields and lensing clusters ) give consistent number densities ( see richard et al .
the lensing cluster technique allows one to extend the observations to the faintest levels , currently down to an effective magnitude of @xmath50 2930 .
clearly the two techniques , one with an additional gain in depth by lensing over a small area and the other over wide fields , are complementary and both approaches will likely play an important role in finding distant , primeval , maybe even pop iii dominated objects .
the lensing studies show a source density of 3100 arcmin@xmath51@xmath52 for @xmath53 from @xmath54 to 30 ( richard et al .
this source density is also in agreement with the theoretical predictions from stiavelli et al .
( 2004 ) and choudhury & ferrara ( 2007 ) , shown in fig .
5 . if spectroscopy of objects down to magnitudes of 30 or even 32 become feasible with extremely large telescopes ( cf .
4 ) this means that we should expect very high source densities , e.g. several hundred objects ( ! ) in a field of view of several arcmin@xmath55as foreseen for elt instruments .
high multiplex multi - object spectrographs reaching with an optical and near - ir coverage up to @xmath56 m , such as eve proposed for the e - elt , should therefore be very efficient in detecting ly@xmath0 and he ii @xmath11640 lines up to redshift 13 and 9.4 respectively .
great progress can be expected from searches for the sources of cosmic reionisation and primeval / pop iii galaxies in the fairly near future !
choudhury , t. r. , & ferrara , a. 2007 , _ mnras _ , 380 , l6 dawson , s. , et al . 2004 , _
apj _ , 617 , 707 dijkstra , m. , & wyithe , j. s. b. 2007 , _ mnras _ , 379 , 1589 fosbury , r. a. e. , et al . 2003 , _
apj _ , 596 , 797 hayes , m. , stlin , g. 2006 , _ a&a _ , 460 , 681 hu , e. m. , et al .
2004 , _ aj _ , 127 , 563 iwata , i. , et al . , 2008 ,
apjl , submitted , arxiv:0805.4012 jimenez , r. , & haiman , z. 2006 , nature , 440 , 501 kashikawa , n. , et al .
2006 , apj , 648 , 7 kennicutt , r.c . , 1998 , _
araa _ , 36 , 189 malhotra , s. , rhoads , j.e . , 2002 ,
_ apj _ , 565 , l71 matsuda , y. , et al . , & petitpas , g. r. 2007 , _ apj _ , 667 , 667 nagao , t. , et al .
2008 , _ apj _ , 680 , 100 ouchi , m. , et al .
apj__s , 176 , 301 pan , l. , & scalo , j. 2007 , _ apj _ , 654 , l29 pell , r. , schaerer , d. , richard , j. , le borgne , j .- f . , & kneib , j .-
2005 , in `` gravitational lensing impact on cosmology '' , iau symp . 225 , 373 richard , j. , stark , d. p. , ellis , r. s. , et al . & smith , g. p. 2008 , _ apj _ , in press , arxiv:0803.4391 scannapieco , e. , schneider , r. , & ferrara , a. 2003 , _ apj _ , 589 , 35 schaerer , d. 2002 , _
, 382 , 28 schaerer , d. 2003 , _ a&a
_ , 397 , 527 schaerer , d. 2003b , in `` multi - wavelength cosmology '' , astro - ph/0309528 schaerer , d. , 2007 , in `` the emission line universe '' , xviii canary islands winter school of astrophysics , ed .
j. cepa , cambridge univ . press , arxiv.0706.0139 schaerer , d. , & vacca , w. d. 1998 , _ apj _ , 497 , 618 schneider , r. , ferrara , a. , & salvaterra , r. 2004 , _ mnras _ , 351 , 1379 shapley , a. e. , steidel , c. c. , pettini , m. , adelberger , k. l. , & erb , d. k. 2006 , _ apj _ , 651 , 688 shimasaku , k. , et al . 2006 , _
pasj _ , 58 , 313 stark , d.p . , ellis , r.s .
, richard , j. , kneib , j .-
p . , smith , g.p . ,
santos , m.r .
, 2007 , _ apj _ , 663 , 10 stiavelli , m. , fall , s. m. , & panagia , n. 2004 , _ apj _ , 600 , 508 thuan , t. x. , & izotov , y. i. 2005 , _ apjs _ , 161 , 240 tornatore , l. , ferrara , a. , & schneider , r. 2007 , _ mnras _ , 382 , 945 tumlinson , j. , giroux , m.l . , shull , j.m .
, 2001 , _ apj _ , 550 , l1 | we review the expected properties of pop iii and very metal - poor starburst and the behaviour the ly@xmath0 and he ii @xmath11640 emission lines , which are most likely the best / easiest signatures to single out such objects .
existing claims of pop iii signatures in distant galaxies are critically examined , and the searches for he ii @xmath11640 emission at high redshift are summarised .
finally , we briefly summarise ongoing and future deep observations at @xmath2 aiming in particular at detecting the sources of cosmic reionisation as well as primeval / pop iii galaxies . |
why do we observe magnetic fields in the sky ?
why do we live in a magnetized universe ?
a variety of observations imply that stars , planets , galaxies , clusters of galaxies are all magnetized .
the typical magnetic field strengths . for practical reasons , in cosmic ray physics and in cosmology
it is also useful to express the magnetic field in @xmath0 ( in units @xmath1 ) .
recalling that the bohr magneton is about @xmath2 the conversion factor will then be @xmath3 .
] range from few @xmath4 g ( in the case of galaxies and galaxy clusters ) to few g ( in the case of planets , like the earth ) up to @xmath5 g ( in the case of neutron stars ) .
physical phenomena are characterized depending upon the typical time and length scales where they take place .
magnetic fields of stars and planets are related to length - scales which are much smaller than the diameter of the milky way ( of the order of @xmath6 kpc ) or of the local supercluster . in this sense , a magnetic field of
the order of @xmath4 g is minute on the terrestrial scale but it is sizable over the scale of the supercluster . in the present review
only large - scale magnetic fields will be considered , i.e. magnetic fields whose typical scale exceeds the au ( @xmath7 ) .
large - scale magnetic fields must be understood and treated as an essential part of the largest structures observed in the sky .
legitimate questions arising in the study of magnetized structures concern their formation and evolution .
the answers to these questions are still not completely settled .
specific observations will eventually allow to discriminate between the different competing explanations .
magnetic fields of distant spiral galaxies are in the @xmath4 g range .
there is also compelling evidence of large - scale magnetic fields which are not associated with individual galaxies .
this empirical coincidence reminds a bit of one of the motivations of the standard hot big - bang model , namely the observation that the light elements are equally abundant in rather different parts of our universe .
the approximate equality of the abundances implies that , unlike the heavier elements , the light elements have primordial origin .
the four light isotopes @xmath8 , @xmath9 , @xmath10 and @xmath11 are mainly produced at a specific stage of the hot big bang model named nucleosynthesis occurring below the a typical temperature of @xmath12 mev when neutrinos decouple from the plasma and the neutron abundance evolves via free neutron decay .
the abundances calculated in the simplest big - bang nucleosythesis model agree fairly well with the astronomical observations .
in similar terms it is plausible to argue that large - scale magnetic fields have similar strengths at large scales because the initial conditions for their evolutions were the same , for instance at the time of the gravitational collapse of the protogalaxy . the way the initial conditions for the evolution of large - scale magnetic fields are set
is generically named magnetogenesis .
there is another comparison which might be useful . back in the seventies
the so - called harrison - zeldovich spectrum was postulated .
later , with the developments of inflationary cosmology the origin of a flat spectrum of curvature and density profiles has been justified on the basis of a period of quasi - de sitter expansion named inflation .
it is plausible that in some inflationary models not only the fluctuations of the geometry are amplified but also the fluctuations of the gauge fields .
this happens if , for instance , gauge couplings are effectively dynamical .
as the harrison - zeldovich spectrum can be used as initial condition for the subsequent newtonian evolution , the primordial spectrum of the gauge fields can be used as initial condition for the subsequent magnetohydrodynamical ( mhd ) evolution which may lead , eventually , to the observed large - scale magnetic fields .
cosmologists and theoretical physicists would like to understand large - scale magnetization in terms of symmetries which are broken .
there are other two different motivations leading , independently , to the study of large - scale magnetic fields .
for instance it was observed long ago by fermi that if cosmic rays are in equilibrium with the galaxy , their pressure density is comparable with the one of the magnetic field of the galaxy .
large - scale magnetic fields are also extremely relevant for astrophysics
. magnetic fields in galaxies are sufficiently intense to affect the dynamics of interstellar gas both on the galactic scale and also on smaller scales characteristic of star formation processes .
this is the third motivation leading to the study of large scale - magnetic fields : the astrophysical motivation which tries to combine our knowledge of the universe in powerful dynamical principles based on the microscopic laws of nature .
the cross - disciplinary character of the physical phenomena addressed in this review is apparent from the table of content . in section 2 a brief historical account of this fifty years old problem
will be given .
then , to clarify the nature of the various theoretical constructions , the empirical evidence of large scale magnetic fields will be discussed in section 3 .
section 4 contains some background material on the evolution of magnetic fields in globally neutral plasmas in flat space .
section 5 and 6 address the problem of the evolution of large scale magnetic fields in curved backgrounds and of their origin . in section 7
the possible effects of magnetic fields on the thermodynamical history of the universe are scrutinized .
section 8 will be concerned with the effects of large scale magnetic fields on the cosmic microwave background radiation ( cmb ) and on the relic background of gravitational waves . in section 9
the rle of gravitating magnetic fields in cosmological solutions will be swiftly pointed out .
the perspective of present review is theoretical . in this sense various ( very important )
experimental results ( for instance concerning optical polarization ) will receive only a swift attention .
in january 1949 large - scale magnetic fields had no empirical evidence .
the theoretical situation can be summarized as follows : * the seminal contributions of alfvn @xcite convinced the community that magnetic fields can have a very large life - time in a highly conducting plasma ; * using the new discoveries of alfvn , fermi @xcite postulated the existence of a large - scale magnetic field permeating the galaxy with approximate intensity of @xmath4 g and , hence , in equilibrium with the cosmic rays ; * rather surprisingly alfvn @xcite did not appreciate the implications of the fermi idea and was led to conclude ( incorrectly ) that cosmic rays are in equilibrium with stars disregarding completely the possibility of a galactic magnetic field ; * in 1949 hiltner @xcite and , independently , hall @xcite observed polarization of starlight which was later on interpreted by davis and greenstein @xcite as an effect of galactic magnetic field aligning the dust grains . according to the presented chain of events it is legitimate to conclude that * the discoveries of alfvn were essential in the fermi proposal ( who was already thinking of the origin of cosmic rays in 1938 before leaving italy ) ; * the idea that cosmic rays are in equilibrium with the galactic magnetic fields ( and hence that the galaxy possess a magnetic field ) was essential in the correct interpretation of the first , fragile , optical evidence of galactic magnetization . the origin of the galactic magnetization , according to @xcite , had to be somehow primordial .
this idea was further stressed in two subsequent investigations of fermi and chandrasekar @xcite who tried , rather ambitiously , to connect the magnetic field of the galaxy to its angular momentum . in the fifties
various observations on polarization of crab nebula suggested that the milky way is not the only magnetized structure in the sky .
the effective new twist in the observations of large - scale magnetic fields was the development ( through the fifties and sixties ) of radio - astronomical techniques . from these measurements , the first unambiguous evidence of radio - polarization from the milky way ( mw )
was obtained @xcite .
it was also soon realized that the radio - zeeman effect ( counterpart of the optical zeeman splitting employed to determine the magnetic field of the sun ) could offer accurate determination of ( locally very strong ) magnetic fields in the galaxy .
the observation of lyne and smith @xcite that pulsars could be used to determine the column density of electrons along the line of sight opened the possibility of using not only synchrotron emission as a diagnostic of the presence of a large - scale magnetic field , but also faraday rotation . in the seventies
all the basic experimental tools for the analysis of galactic and extra - galactic magnetic fields were present . around this epoch
also extensive reviews on the experimental endeavors started appearing and a very nice account could be found , for instance , in the review of heiles @xcite .
it became gradually clear in the early eighties , that measurements of large - scale magnetic fields in the mw and in the external galaxies are two complementary aspects of the same problem .
while mw studies can provide valuable informations concerning the _ local _ structure of the galactic magnetic field , the observation of external galaxies provides the only viable tool for the reconstruction of the _ global _ features of the galactic magnetic fields . as it will be clarified in the following sections , the complementary nature of global and local morphological features of large - scale magnetization may become , sometimes , a source of confusion . since the early seventies
, some relevant attention has been paid not only to the magnetic fields of the galaxies but also to the magnetic fields of the _ clusters_. a cluster is a gravitationally bound system of galaxies . the _ local group _ ( i.e. _ our _ cluster containing the mw , andromeda together with other fifty galaxies ) is an _ irregular _ cluster in the sense that it contains fewer galaxies than typical clusters in the universe .
other clusters ( like coma , virgo ) are more typical and are then called _ regular _ or abell clusters . as an order of magnitude estimate
, abell clusters can contain @xmath13 galaxies . in the nineties magnetic fields
have been measured in single abell clusters but around the turn of the century these estimates became more reliable thanks to improved experimental techniques . in order to estimate magnetic fields in clusters , an independent knowledge of the electron density along the line of sight is needed ( see sec .
recently faraday rotation measurements obtained by radio telescopes ( like vla in socorro ( new mexico ) ] ) have been combined with independent measurements of the electron density in the intra - cluster medium .
this was made possible by the maps of the x - ray sky obtained with satellites measurements ( in particular rosat@xmath14 kev .
a catalog of x - ray bright abell clusters was compiled . ] ) .
this improvement in the experimental capabilities seems to have partially settled the issue confirming the measurements of the early nineties and implying that also clusters are endowed with a magnetic field of @xmath15 g strength which is _ not associated with individual galaxies_. the years to come are full of interesting experimental questions to be answered .
these questions , as it will be discussed in the following , may have a rather important theoretical impact both on the theory of processes taking place in the local universe ( like ultra - high energy cosmic ray propagation ) and on the models trying to explain the origin of large scale magnetic fields .
last but not least , these measurements may have an impact on the physics of cmb anisotropies and , in particular , on the cmb polarization .
in fact , the same mechanism leading to the faraday rotation in the radio leads to a faraday rotation of the cmb _ provided _ the cmb is linearly polarized .
one of the important questions to be answered is , for instance , the nature and strength of the supercluster magnetic field and now more careful statistical studies are starting also along this important direction .
superclusters are gravitationally bound systems of clusters .
an example is the local supercluster formed by the local group and by the virgo cluster .
together rosat , various observations with the einstein , exosat , and ginga satellites showed the presence of hot diffuse gas in the local supercluster .
the estimated magnetic field in this system is , again , of the order of the @xmath4 g but the observational evidence is still not conclusive .
another puzzling evidence is the fact that lyman-@xmath16 systems ( with red - shifts @xmath17 ) are endowed with a magnetic field @xcite . from the historical development of the subject
a series of questions arises naturally : * what is the origin of large - scale magnetic fields ? * are magnetic fields primordial as assumed by fermi more than 50 years ago ? * even assuming that large - scale magnetic fields are primordial , is there a theory for their generation ?
before describing the theory of something ( not of everything ) it is highly desirable , in natural sciences , to understand in some detail the empirical evidence of the subject under discussion . while technical accounts of the various experimental techniques exist already @xcite it seems useful , in the present context , to give an account of the experimental situation by emphasizing the physical principles of the various measurements
there are valuable books dealing directly with large - scale magnetism @xcite .
furthermore excellent reviews of the morphological features of large scale magnetism can be found in @xcite . even if the present review will not be concerned with planetary magnetic fields , analogies with physical situations arising closer to the earth
are often useful and , along this perspective , magnetic field observations in the solar system have been recently reviewed @xcite .
magnetic fields observed in the universe have a homogeneous ( or uniform ) component and a non - uniform component .
it is then useful , for the purposes of this section , to write , in a schematic notation , that @xmath18 where @xmath19 is the uniform component and @xmath20 the non - uniform component .
both components are phenomenologically very relevant .
different experimental techniques can probe different components of large - scale magnetic fields ( i.e. either the total magnetic field or its homogeneous component ) .
another important distinction is between @xmath21 and @xmath22 : there are measurements ( like synchrotron emission ) which are sensitive to @xmath21 i.e. the transverse component of the magnetic field ; there are other measurements ( like the faraday rotation measure or the zeeman splitting of spectral lines ) which are sensitive to @xmath22 i.e. the magnetic field along the line of sight . in order to measure large scale magnetic fields , one of the first effects coming to mind ,
is the zeeman effect .
the energy levels of an hydrogen atom in the background of a magnetic field are not degenerate .
the presence of a magnetic field produces a well known splitting of the spectral lines : @xmath23 where @xmath24 denotes the _ uniform component of the magnetic field along the line of sight_. from the estimate of the splitting , the magnetic field intensity can be deduced .
indeed this technique is the one commonly employed in order to measure the magnetic field of the sun @xcite .
the most common element in the interstellar medium is neutral hydrogen , emitting the celebrated @xmath25-cm line ( corresponding to a frequency of @xmath26 mhz ) . if a magnetic field of @xmath4 g strength is present in the interstellar medium , according to eq .
( [ zee ] ) , an induced splitting , @xmath27 , can be estimated .
zeeman splitting of the @xmath25-cm line generates two oppositely circular polarized spectral lines whose apparent splitting is however sub - leading if compared to the doppler broadening .
in fact , the atoms and molecules in the interstellar medium are subjected to thermal motion and the amount of induced doppler broadening is roughly given by @xmath28 where @xmath29 is the thermal velocity @xmath30 where @xmath31 is the mass of the atom or molecule .
the amount of doppler broadening is @xmath32 khz which is much larger than the zeeman splitting which should be eventually detected .
zeeman splitting of molecules and recombination lines should however be detectable if the magnetic field strength gets larger with the density .
indeed in the interstellar medium there are molecules with an unpaired electron spin . from eq .
( [ zee ] ) it is clear that a detectable zeeman splitting ( i.e. comparable or possibly larger than the doppler broadening ) can be generically obtained for magnetic fields where @xmath33 g , i.e. magnetic fields of the order of the m g. molecules with an unpaired electron spin include oh , cn , ch and some other . in the past
, for instance , magnetic fields have been estimated in oh clouds ( see @xcite and references therein ) .
magnetic fields of the order of @xmath34 m g have been detected in interstellar @xmath35 maser clumps ( with typical densities @xmath36 ) @xcite .
more recently attempts of measuring magnetic fields in cn have been reported @xcite .
the possible caveat with this type of estimates is that the measurements can only be very _ local _ : the above mentioned molecules are much less common than neutral hydrogen and are localized in specific regions of the interstellar medium . in spite of this caveat , zeeman splitting measurements can provide reliable informations on the local direction of the magnetic field .
this determination is important in order to understand the possible origin of the magnetic field .
this aspect will be discussed , in more detail , when describing the magnetic field of the milky way .
the first experimental evidence of the existence of large - scale magnetic fields in external galaxies came , historically , from the synchrotron emission of crab nebula @xcite . the emissivity formula ( i.e. the energy emitted from a unit volume , per unit time , per unit frequency interval and per unit solid angle ) for the synchrotron is a function of @xmath21 and and of the relativistic electron density , namely @xmath37 where @xmath38 is the frequency
, @xmath39 is a proportionality constant depending only upon the spectral index @xmath16 which also determines the ( isotropic ) relativistic electron number density distribution for electrons with energies in the range @xmath40 : @xmath41 a useful estimate for the maximal emission from electrons of energy @xmath42 is given by @xcite @xmath43 in generic terms the synchrotron emission is then sensitive to the _ total _ transverse magnetic field . in oversimplified terms
the measurement proceeds in three steps : * estimate of the emission intensity ; * from the estimate of the emission intensity , defining as @xmath44 the typical size of the source , the quantity @xmath45 can be deduced ; * the magnetic field can then be obtained by specifying the electron density distribution of the source .
while the second step is rather well defined ( @xmath16 can indeed be determined from the observed emission spectrum are between @xmath46 and @xmath47 .
an important caveat to this statement , relevant for foreground extraction ( for instance in studies of cmb anisotropies ) is that the synchrotron spectrum may have a break . ] ) the third step has to be achieved in a model - dependent way .
the relativistic electron density is sometimes estimated using equipartition , i.e. the idea that magnetic and kinetic energy densities may be , after all , comparable .
equipartition is not always an empirical evidence , but can certainly be used as a working hypothesis which may or may not be realized in the system under consideration . for instance equipartition probably holds for the milky way but it does not seem to be valid in the magellanic clouds @xcite .
the average equipartition field strengths in galaxies ranges from the @xmath48 g of m33 up to the @xmath49 g of ngc2276 @xcite . in order to illustrate the synchrotron emission
take for instance the mw .
away from the region of the galactic plane synchrotron emission is the dominant signal of the galaxy for frequencies below @xmath50 ghz . between @xmath51 and @xmath50 ghz the spectral dependence is @xmath52 .
it should be noticed that the spectral index of synchrotron emission may change in different frequency ranges and also in different regions of the sky due to the variation of the magnetic field @xcite .
this aspect is crucial in the experimental analysis of cmb anisotropy experiments in the ghz region .
in fact , in the context of cmb studies the synchrotron is a _ foreground _ which should be appropriately subtracted .
the way this is done is by extrapolation of lower frequency measurements .
the problem of proper subtraction of synchrotron foreground is particularly acute in the case of the planck mission to @xmath53 ghz . ]
the synchrotron has an intrinsic polarization which can give the orientation of the magnetic field , but not _ the specific sign of the orientation vector_. the strength of synchrotron polarization is proportional to @xmath54 , i.e. the ratio between the magnetic energy densities of the uniform and total magnetic field .
thus , according to eq .
( [ btot ] ) , the synchrotron polarization is sensitive to the _ random component _ of the large - scale magnetic field . to infer the magnitude of the magnetic field strength faraday effect has been widely used .
when a polarized radio wave passes through a region of space of size @xmath55 containing a plasma with a magnetic field the polarization plane of the wave gets rotated by an amount @xmath56 which is directly proportional to the square of the plasma frequency ) in the following section . ]
@xmath57 ( and hence to the electron density ) and to the larmor frequency @xmath58 ( and hence to the magnetic field intensity ) .
a linear regression connecting the shift in the polarization plane and the square of the wavelength @xmath59 , can be obtained : @xmath60 by measuring the relation expressed by eq .
( [ regr ] ) for two ( or more ) separate ( but close ) wavelengths , the angular coefficient of the regression can be obtained and it turns out to be @xmath61 in units of @xmath62 when all the quantities of the integrand are measured in the above units .
the explicit dependence of the red - shift can be also easily included in eq .
( [ fr ] ) .
notice , in general terms that the rm is an integral over distances .
thus the effect of large distances will reflect in high values of the rm .
furthermore , the faraday effect occurs typically in the radio ( i.e. @xmath63 ) , however , some possible applications of faraday effect in the microwave can be also expected ( see section 8 of the present review ) .
the shift in the polarization plane should be determined with an accuracy greater than @xmath64 .
otherwise ambiguities may arise in the determination of the angular coefficient appearing in the linear regression of eq .
( [ regr ] ) .
this aspect is illustrated in fig .
[ farfig ] which is rather standard but it is reproduced here in order to stress the possible problems arising in the physical determination of the rm if the determination of the shift in the polarization plane is not accurate . the rm defined in eq .
( [ fr ] ) not only the magnetic field ( which should be observationally estimated ) , but also the column density of electrons . from the radio - astronomical observations ,
different techniques can be used in order to determine @xmath65 .
one possibility is to notice that in the observed universe there are pulsars .
pulsars are astrophysical objects emitting regular pulses of electromagnetic radiation with periods ranging from few milliseconds to few seconds . by comparing the arrival times of different radio pulses at different radio wavelengths , it is found that signals are slightly delayed as they pass through the interstellar medium exactly because electromagnetic waves travel faster in the vacuum than in an ionized medium .
hence , from pulsars the column density of electrons can be obtained in the form of the dispersion measure , i.e. @xmath66 . dividing the rm by dm , an estimate of the magnetic field can be obtained .
due to their abundance , pulsars lead to the best determination of the magnetic field in the galactic disk @xcite . in fig .
[ f1a ] ( adapted from @xcite ) a map of the antisymmetric rm sky is reported . in the picture
the open circles denote negative rm while filled circles denote positive rm .
the size of the circle is proportional to the magnitude of the rm .
the convention is , in fact , to attribute negative rm to a magnetic field directed away from the observer and positive rm if the magnetic field is directed toward the observer . as in the case of synchrotron emission also faraday rotation measurements
can be used as a diagnostic for foreground contamination .
the idea would be , in this context to look for cross - correlations in the faraday rotation measure of extra - galactic sources and the measured microwave signal at the same angular position .
a recent analysis has been recently reported @xcite . if magnetic field or the column density change considerably over the integration path of eq .
( [ fr ] ) one should probably define and use the two - point function of the rm , i.e. @xmath67 the suggestion to study the mean - squared fluctuation of the rm was proposed @xcite .
more recently , using this statistical approach particularly appropriate in the case of magnetic fields in clusters ( where both the magnetic field intensity and the electron density change over the integration path ) , newmann , newmann and rephaeli @xcite quantified the possible ( statistical ) uncertainty in the determination of cluster magnetic fields ( this point will also be discussed later ) .
the rather ambitious program of measuring the @xmath68 power spectrum is also pursued @xcite . in @xcite
the analysis of correlations in the rm has been discussed . since the early seventies @xcite the magnetic field of the milky way ( mw ) was shown to be parallel to the galactic plane .
rm derived from pulsars allow consistent determinations of the magnetic field direction and intensity @xcite . in the milky way ,
the uniform component of the magnetic field is in the plane of the galactic disk and it is thought to be directed approximately along the spiral arm .
there is , though , a slight difference between the northern and southern hemisphere .
while in the southern hemisphere the magnetic field is roughly @xmath69 g , the magnetic field in the northern hemisphere is three times smaller than the one of the southern hemisphere .
the magnetic field of the mw is predominantly toroidal .. ] however , looking at the center of the galaxy , a poloidal component ( typically @xmath70 of the toroidal one ) has been detected .
in fact , in the central @xmath71 pc of the mw there are in fact pooloidal ( dipole ) fields whose origin is probably primordial @xcite ( see also @xcite ) .
the reason for this statement is that through the usual plasma physics mechanisms ( like the dynamo theory to be discussed later in section 4 ) it is hard to account for a sizable poloidal component , even if localized in the central region of the mw . the common practice is to classify large - scale magnetic fields in spiral galaxies according to the parity of the toroidal field under reflections of the azimuthal angle @xcite .
according to some authors this distinction palys a crucial rle in order to assess the primordial ( or non - primordial ) nature of the observed field ( see , for instance , @xcite ) .
in short , the virtues and limitations of the previous distinction can be summarized as follows .
suppose to plot the toroidal component of the measured magnetic field ( for instance of the mw ) as a function of the azimuthal angle @xmath72 .
this is , in fact , equivalent , to plot the @xmath73 for instance near the equatorial plane of the galaxy .
then two qualitatively different situations can arise . in the first case @xmath73 upon rotation of @xmath74 around the galactic center
is even and this corresponds to axisymmetric spiral galaxies ( ass ) . in the second case
@xmath75 upon rotation of @xmath74 around the galactic center is odd and this corresponds to bisymmetric spiral galaxies ( bss ) .
this sharp distinction is invalidated by two experimental facts .
first of all there may be extra phases so that it is not clear , in practice , if @xmath73 is , predominantly , even or odd . in spite of possible ambiguities ( which are
, however , an experimental problem ) the distinction may be very useful : if the magnetic fields would originate through some dynamo amplification the preferred configuration will be of the ass - type and not of the bss - type .
there are strong indications that the mw is a bss spiral @xcite .
a related issue is the _ reversal _ of the magnetic field from one spiral arm to the other . in @xcite
@xmath76 rotation measures from polarization observations of southern pulsars have been reported .
for the galactic disk , convincing evidence of field reversal near the _ perseus _ arm is presented . in the solar circle
three reversals are observed : near the _ carina - sagittarius _ arm , near the _ crux - scutum _ arm , and possibly a third one in the _ norma _ arm .
these reversals are claimed to be consistent with bss models . since this is a relevant point , the possible controversies arising in the analysis of the magnetic field of the mw
will now be swiftly mentioned .
it is appropriate , for this purpose , to give a geometrical characterization of the arm structure of the mw @xcite .
the spiral structure of the galaxy can be described in terms of a two - dimensional coordinate system @xmath77 whose origin @xmath78 is at the galactic center . in this coordinate system
the sun is located at @xmath79 where the coordinates are expressed in kpc . for each spiral arm
the equation reduces to four curves , each rotated by @xmath80 of the form : @xmath81 in eq .
( [ spiralarms ] ) @xmath82 , @xmath83 , @xmath84 . in fig .
[ f2 ] the map of the spiral arms is illustrated according to the model of eq .
( [ spiralarms ] ) .
[ f2 ] is adapted from the recent paper of valle @xcite .
the region with dashed lines is poorly known from the observational point of view ( `` zona galactica incognita '' in the terminology of valle @xcite ) . in terms of fig .
[ f2 ] the basic question related to the magnetic field structure concerns the _ relative _ orientation of the magnetic field direction between one arm and the nearest one .
as mentioned above@xcite some studies suggest that three reversals are present .
some other studies @xcite based on a statistical re - analysis of the most recent data from 1995 to 2001 suggest that _ only one _ reversal is observed .
the two large arrows in fig .
[ f2 ] represent the field directions in the carina - sagittarius and in the perseus arms . in @xcite
it has been argued that the magnetic field of the mw has the structure of axisymmetric spiral .
however , the presence of one field reversal seems to be not fully consistent with this interpretation .
recently @xcite han , manchester , lyne and qiao , using the parkes multibeam pulsar survey , provided further observational support for the detection of a counterclockwise magnetic field in the norma spiral arm .
the morphological properties of the magnetic field in the carina - sagittarius and in the crux - scutum arms have been confirmed @xcite .
these results were obtained from the analysis of pulsar rotation measures .
if the magnetic field of the mw flips its direction from one spiral arm to the other , then , as pointed out by sofue and fujimoto @xcite the galactic magnetic field should probably be associated with a bss model . in the sofue - fujimoto model the radial and azimuthal components of the magnetic field in a bisymmetric logarithmic spiral configuration is given through @xmath85 where @xmath86 kpc is the galactocentric distance of the maximum of the field in our spiral arm , @xmath87 and @xmath88 is the pitch angle of the spiral .
the smooth profile @xmath89 can be chosen in different ways .
a motivated choice is @xcite ( see also @xcite ) @xmath90 where @xmath91 kpc is the distance of the sun to the galactic center and @xmath92 kpc .
the original model of sf does not have dependence in the @xmath93 direction , however , the @xmath93 dependence can be included and also more complicated models can be built @xcite . typically , along the @xmath93 axis
, magnetic fields are exponentially suppressed as @xmath94}$ ] with @xmath95 kpc . in the bss model
there are then three issues to be determined in order to specify the parameters@xcite : * the number and location of field reversals ; * the value of the pitch angle @xmath88 introduced in eq .
( [ bss ] ) ; * the field strength in the disk .
the answers to these three questions are , in short , the following : * 3 to 5 field reversals have been claimed ; * the pitch angle is determined to be @xmath96 @xcite ; * the strength of the regular field is , as anticipated @xmath97 @xmath98 .
it is interesting to notice that the _ total _ field strength , in the notation of eq .
( [ btot ] ) can reach even 6 @xmath4 g indicating , possibly , a strong stochastic component .
differently from other spirals , the milky way has also a large _ radio halo_. the radio - halo indicates the large - scale height of the magnetic field and its origin is unclear as far as the stochastic component of the galactic magnetic field is concerned , the situation seems to be , according to the reported results , still unclear @xcite .
it is , at present , not fully understood if the stochastic component of the galactic magnetic field is much smaller than ( or of the same order of ) the related homogeneous part as implied by the estimates of the total field strength .
the structure of magnetic fields can be relevant when investigating the propagation of high - energy protons @xcite as noticed already long ago @xcite .
this aspect leads to crucial ( and structural ) ambiguities in the analysis of the propagation of charged particles of ultra - high energy .
the current observations of ultrahigh energy cosmic rays ( uhecr ) do not lead to a firm evidence of the existence of a cutoff of the cosmic ray spectrum between @xmath99 and @xmath100 ev .
this is the celebrated greisen , zatsepin , kuzmin ( gzk ) cutoff @xcite . if the source distribution of uhecr is isotropic and homogeneous because photoproduction interactions on the microwave background , then the gzk cutoff should be present .
the isotropization of uhecr can be explained by scattering in large magnetic fields .
this suggestion can be achieved in the presence of a sizable galactic halo .
magnetic fields may change not only the local intensity of uhecr but also their energy spectrum @xcite .
various papers analyzed the acceleration and propagation of uhecr in magnetized structures ( see , for instance , @xcite and references therein )
. an excellent presentation of the evidence of cosmic magnetism in nearby galaxies has been given by beck et al.@xcite and , in a more concise form , by widrow @xcite .
the past few years have witnessed a long discussion on the primordial nature of magnetic fields in nearby galaxies whose features are partially reported in @xcite . in analogy with the problem discussed in the case of the mw two different opinions have been confronted .
theorists believing the primordial nature of the galactic magnetic field supported the bss model . on the contrary , dynamo theorists supported the conclusion that most of the nearby galaxies are ass .
as previously described , the variation of the @xmath68 as a function of the azimuthal angles has been used in order to distinguish ass from bss @xcite .
the simplistic summary of the recent results is that there are galaxies where bss structure dominates ( like m81,m51 ) @xcite . at the same time in some galaxies
the ass dominates @xcite .
already the short summary of the main experimental techniques used for the detection of large scales magnetic fields shows that there may be problems in the determination of magnetic fields right outside galaxies .
there magnetic fields are _ assumed _ to be often of @xmath101 strength .
however , due to the lack of sources for the determination of the column density of electrons , it is hard to turn the assumption into an experimental evidence .
magnetic fields in clusters have been recently reviewed by giovannini @xcite and , more extensively , by carilli and taylor @xcite . since various theoretical speculations
suggest that also clusters are magnetized , it would be interesting to know if regular abell clusters posses large scale magnetic fields .
different results in this direction have been reported @xcite ( see also @xcite ) .
some studies during the past decade @xcite dealt mainly with the case of a single cluster ( more specifically the coma cluster ) .
the idea was to target ( with faraday rotation measurements ) radio sources inside the cluster .
however , it was also soon realized that the study of many radio sources inside different clusters may lead to experimental problems due to the sensitivity limitations of radio - astronomical facilities .
the strategy is currently to study a sample of clusters each with one or two bright radio - sources inside . in the past
it was shown that regular clusters have cores with a detectable component of rm @xcite .
recent results suggest that @xmath4 gauss magnetic fields are indeed detected inside regular clusters @xcite . inside
the cluster means in the intra - cluster medium .
therefore , these magnetic fields can not be associated with individual galaxies .
regular abell clusters with strong x - ray emission were studied using a twofold technique @xcite . from the rosat full sky survey ,
the electron density has been determined @xcite .
faraday rm ( for the same set of 16 abell clusters ) has been estimated through observations at the vla radio - telescope .
the amusing result ( confirming previous claims based only on one cluster @xcite ) is that x - ray bright abell clusters possess a magnetic field of @xmath4 gauss strength .
the clusters have been selected in order to show similar morphological features .
all the 16 clusters monitored with this technique are at low red - shift ( @xmath102 ) and at high galactic latitude ( @xmath103 ) .
these recent developments are rather promising and establish a clear connection between radio - astronomical techniques and the improvements in the knowledge of x - ray sky .
there are various satellite missions mapping the x - ray sky at low energies ( asca , chandra , newton and @xmath34 kev and it is flying since february 1993 .
chandra ( nasa mission ) and newton ( esa mission ) have an energy range comparable with the one of asca and were launched , almost simultaneously , in 1999 . ] ) .
there is the hope that a more precise knowledge of the surface brightness of regular clusters will help in the experimental determination of large scale magnetic fields between galaxies .
it should be however mentioned that evidence for the presence of relativistic electrons and magnetic fields in clusters was directly available even before , from measurements of extended regions of radio synchrotron emission ( for frequencies @xmath104 ghz ) @xcite .
it is interesting to notice that intra - cluster magnetic fields of @xmath4 g strength can induce faraday rotation on cmb polarization . by combining informations from sunyaev - zeldovich effect and x - ray emission from the same clusters
, it has been recently suggested that a richer information concerning electron column density can be obtained @xcite . in fig .
[ f1 ] the results reported in @xcite are summarized . in fig .
[ f1 ] the rm of the sample of x - ray bright abell clusters is reported after the subtraction of the rm of the galaxy . at high galactic latitude ( where all the observed clusters are )
the galactic contribution is rather small and of the order of @xmath105 .
the results reported in @xcite in fig .
[ f1 ] the open points represent sources viewed through the thermal cluster gas , whereas the full points represent control sources at impact parameters larger than the cluster gas .
the excess in rm attributed to clusters sources is clearly visible . using the described techniques large scale magnetic fields
can be observed and studied in external galaxies , in clusters and also in our own galaxy . while the study of external galaxies and clusters may provide a global picture of magnetic fields , the galactic observations performed within the milky way are more sensitive to the local spatial variations of the magnetic field . for this reasons local and global observations are complementary .
the flipped side of the coin , as we will see in the second part of the present section , is that the global structure of the magnetic field of our galaxy is not known directly and to high precision but it is deduced from ( or corroborated by ) the global knowledge of other spiral galaxies .
magnetic fields in the local supercluster are notoriously difficult to measure . in the absence of direct measurements
it is the opinion of the author that these fields can not be assumed to be small , i.e. @xmath106 . in this situation
it seems useful to consider the possibility that these fields are _ large_. in @xcite this point of view has been taken in connection with the propagation of the most energetic cosmic rays .
this approach is desirable since the ( few ) reliable claims of measurements of large - scale magnetic fields in the local supercluster seem to indicate that these fields are rather strong and more in the range of the @xmath4 g than in the range of the n g. recently @xcite the possible existence , strength and structure of magnetic fields in the intergalactic plane , within the local supercluster , has been scrutinized .
the local supercluster is centered approximately at the virgo cluster , about @xmath107 mpc from the local group .
a statistically significant faraday screen acting on the radio - waves coming from the most distance sources has been found .
this analysis supports the existence of a regular magnetic field of @xmath108 g in the local supercluster . in the past detection of radio emission in the coma supercluster
has been reported @xcite .
the plane of the coma supercluster is defined by the coma cluster and by the abell cluster 1367 .
the observed magnetic field has been extimated to be of the order of @xmath109 g. in @xcite using numerical simulations the spectrum of protons above @xmath99 ev has been determined under the assumption that the injection spectrum is determined by a discrete set of continuously emitting sources .
the sources follow the profile of the local supercluster .
it has been shown that if @xmath110 the source distribution assumed in the supercluster is inconsistent with the observations .
on the other hand , if @xmath111 , 10 sources in the local supercluster would lead to spectra consistent with the observations .
this analysis seems to fit with @xcite and with the general arguments put forward in @xcite . in a complementary perspective , the authors of @xcite claim to be able to simulate the evolution of large - scale magnetic fields from reasonably high red - shifts . the obtained results , according to @xcite , correctly reproduce the present large - scale structure of cluster and supercluster magnetic fields both in their observed and unobserved features .
the simplified discussion of the plasma evolution is often related to magnetohydrodynamics ( mhd ) which is a one - fluid plasma description holding under very specific assumptions .
the goal of the first part of the present section ( together with the related appendix a ) will be to present the reduction of the common two - fluids plasma dynamics to the simplified case of one - fluid description provided by mhd .
consider , as a starting point , a globally neutral plasma of electrons and ions ( in the simplest case protons ) all at the same temperature @xmath112 and with mean particle density @xmath113 .
suppose , for simplicity , local thermodynamical equilibrium .
the charge densities will be given , respectively , by @xmath114 where @xmath115 is the electrostatic potential .
( [ weakly ] ) it has been assumed that the plasma is weakly coupled , i.e. @xmath116 . a test charge @xmath117 located in the origin will then experience the electrostatic potential following from the poisson equation , @xmath118 - 4\pi q_{\rm t } \delta^{(3)}(\vec{x } ) , \label{pois}\ ] ] or , using eqs .
( [ weakly ] ) into eq .
( [ pois ] ) @xmath119 where @xmath120 is the debye length . for a test particle
the coulomb potential will then have the usual coulomb form but it will be suppressed , at large distances by a yukawa term , i.e. @xmath121 . in the interstellar medium there are three kinds of regions which are conventionally defined , namely @xmath122 regions ( where the hydrogen is predominantly in molecular form ) , @xmath123 regions ( where hydrogen is in atomic form ) and @xmath124 regions ( where hydrogen is ionized ) .
sometimes @xmath124 regions are denoted with hi and @xmath122 regions with hii . in the @xmath124 regions the typical temperature @xmath112 is of the order of @xmath34@xmath125 ev
while for @xmath113 let us take , for instance , @xmath126 .
then @xmath127 .
for @xmath128 the coulomb potential is screened by the global effect of the other particles in the plasma .
suppose now that particles exchange momentum through two - body interactions .
their cross section will be of the order of @xmath129 and the mean free path will be @xmath130 , i.e. recalling eq .
( [ debye ] ) @xmath131 .
this means that the plasma is a weakly collisional system which is , in general , _ not _ in local thermodynamical equilibrium .
this observation can be made more explicit by defining another important scale , namely the plasma frequency which , in the system under discussion , is given by @xmath132 where @xmath133 is the electron mass . the plasma frequency is the oscillation frequency of the electrons when they are displaced from their equilibrium configuration in a background of approximately fixed ions .
recalling that @xmath134 is the thermal velocity of the charge carriers , the collision frequency @xmath135 is always much smaller than @xmath136 .
thus , in the idealized system described so far , the following hierarchy of scales holds : @xmath137 which means that before doing one collision the system undergoes many oscillations , or , in other words , that the mean free path is not the shortest scale in the problem .
usually one defines also the _ plasma parameter _
@xmath138 , i.e. the number of particles in the debye sphere . in the approximation of weakly coupled plasma @xmath139 which also imply that the mean kinetic energy of the particles is larger than the mean inter - particle potential , i.e. @xmath140 in the language of eq .
( [ weakly ] ) . from the point of view of the evolution of large - scale magnetic fields ,
the galaxy is globally neutral system of charged particles with typical rotation period of @xmath141 yrs .
two complementary descriptions of the plasma can then be adopted .
the first possibility is to study the full system of kinetic equations ( the vlasov - landau description @xcite for the one particle distribution function ) . in the kinetic description
the `` observables '' of the plasma dynamics are related to the various moments of the distribution function . in a complementary perspective
, it is also possible to study directly the evolution equations of the various moments of the distribution function ( charge and matter densities , momentum transfer etc . ) .
the resulting description is an effective one since not _ all _ the moments of the distribution function can be kept .
the necessity of a consistent truncation in the hierarchy of the various moments is usually called _
closure problem_. mhd is an effective description which holding for @xmath142 where @xmath44 and @xmath143 are the typical length and time scales of the problem .
the mhd approximation holds for typical length scales larger than the debye scale and typical frequencies much smaller than the plasma frequency .
in particular , the spectrum of plasma excitations obtained from the full kinetic description matches with the spectrum of mhd excitations but only at sufficiently small frequencies ( typically of the order of the ionoacoustic waves ) @xcite .
if the plasma is weakly collisional and out of thermodynamical equilibrium , an approximation where the evolution equations are still tractable is given in terms of the one - particle distribution function @xmath144 where the subscript denotes the given charge carrier ( for instance electrons and ions ) .
the one - particle distribution obeys the vlasov - landau equation for each particle species of charge @xmath145 and of mass @xmath146 : @xmath147 the term appearing at the right - hand side of eq .
( [ vlasov1 ] ) is the binary collision term . in a weakly coupled plasma , the long - range force
is given by the coulomb interaction , while the short - range component arises thanks to binary collisions .
in other words the electric and magnetic fields appearing in eq .
( [ vlasov1 ] ) are _ mean fields _ obeying _ mean _ maxwell s equations : @xmath148 where @xmath149 is the mean particle density .
in the context of the kinetic approach the evolution equation should be solved self - consistently with the maxwell s equations .
clearly , in a system where the density of particles is small the mean field given by the distant particles is more important than the the force produced by the closest particles . in practice
the vlasov - landau system can be linearized around some equilibrium value of the one - particle distribution function and specific examples will be discussed later on .
the one - particle distribution function is not directly observable .
the directly observable quantities are the the various moments of the one - particle distribution , for instance , the particle density , @xmath150 and the related matter and charge densities for each particle species ( i.e. @xmath151 and @xmath152 ) or summed over the different species ( i.e. @xmath153 and @xmath154 ) . however , the dynamical informations on the plasma evolution are not only encoded in the moment of order zero but also in the higher moments ( the first , the second and even higher ) . for some class of problems
the kinetic approach is to be preferred .
for instance the kinetic approach is mandatory for all the high frequency phenomena ( like , for instance , equilibration of charge and current density fluctuations ) . as far as the low - frequency phenomena are concerned one can imagine to obtain a reduced ( fluid - like ) description .
the evolution equations for the one - fluid variables are obtained from the two - species kinetic description by some algebra which is summarized in the first part of appendix a. the bottom line of the derivation is that the moments of the one - particle distribution function obey a set of partial differential equations whose solutions can be studied in different approximations .
one of these approximations is the mhd description .
the set of one - fluid equations ( derived in appendix a ) can be written as @xmath155 \vec{v } = \vec{j } \times \vec{b } - \vec{\nabla } p , \label{mt}\\ & & \vec{e } + \vec{v}\times\vec{b } = \frac{1}{\sigma } \vec{j } + \frac{1}{e n_{\rm q } } ( \vec{j } \times \vec{b } - \vec{\nabla } p_{e } ) .
\label{genohm1}\end{aligned}\ ] ] supplemented by the maxwell s equations @xmath156 eqs .
( [ cm ] ) and ( [ cq ] ) are the mass and charge conservation , eq . ( [ mt ] ) is the equation for the momentum transfer and eq .
( [ genohm1 ] ) is the generalized ohm law . in eqs .
( [ cm])([genohm1 ] ) @xmath157 is the one - fluid velocity field , @xmath158 is the total pressure and @xmath159 is the pressure of the electrons . if the plasma is globally neutral the charge density can be consistently neglected .
however , global neutrality is plausible only for typical lengths much larger than @xmath160 where the collective properties of the plasma are the leading dynamical effect . under this assumption ,
( [ cq ] ) and ( [ mx1 ] ) imply @xmath161 which means that mhd currents are , in the first approximation , solenoidal . the second step to get to a generalized form of the mhd equations is to neglect the terms responsible for the high frequency plasma excitations , like the displacement current in eqs .
( [ mx3 ] ) whose form becomes then @xmath162 given the two sets of consistent approximations discussed so far , the form taken by the generalized ohm law may not be unique and , therefore , different mhd descriptions arise depending upon which term at the right hand side of eq .
( [ genohm1 ] ) is consistently neglected .
the simplest ( but often not realistic ) approximation is to neglect all the terms at the right hand side of eq .
( [ genohm1 ] ) .
this is sometimes called _ ideal mhd approximation_. in this case the ohmic electric field is given by @xmath163 . this ideal description
is also called sometimes _ superconducting mhd approximation _ since , in this case , the resistive terms in eq .
( [ genohm1 ] ) , i.e. @xmath164 go to zero ( and the conductivity @xmath165 ) . in the early universe it is practical to adopt the superconducting approximation ( owing to the large value of the conductivity ) but it is also very dangerous since all the @xmath166 effects are dropped . a more controllable approximation is the _ real ( or resistive ) mhd_.
the resistive approximation is more controllable since the small expansion parameter ( i.e. @xmath167 ) is not zero ( like in the ideal case ) but it is small and finite .
this allows to compute the various quantities to a a given order in @xmath167 even if , for practical purposes , only the first correction is kept . in this framework the ohmic current is not neglected and the generalized ohm law takes the form @xmath168 comparing eq .
( [ genohm2 ] ) with eq .
( [ genohm1 ] ) , it is possible to argue that the resistive mhd scheme is inadequate whenever the hall and thermoelectric terms are can not be neglected . defining @xmath169 as the larmor frequency
, the hall term @xmath170 can be neglected provided @xmath171 where @xmath44 us the typical size of the system and @xmath172 the typical bulk velocity of the plasma . with
analogous dimensional arguments it can be argued that the pressure gradient , i.e. the thermoelectric term , can be consistently neglected provided @xmath173 .
specific examples where the hall term may become relevant will be discussed in the context of the biermann battery mechanism .
the set of mhd equations , as it has been presented so far , is not closed if a further relation among the different variables is not specified @xcite .
typically this is specified through an equation relating pressure and matter density ( adiabatic or isothermal closures ) .
also the incompressible closure ( i.e. @xmath174 ) is often justified in the context of the evolution of magnetic fields prior to recombination .
it should be stressed that sometimes the adiabatic approximation may lead to paradoxes .
it would correspond , in the context of ideal mhd , to infinite electrical conductivity and vanishingly small thermal conductivity ( i.e. the electrons should be extremely fast not to feel the resistivity and , at the same time , extremely slow not to exchange heat among them ) .
the ideal and resistive mhd descriptions are rather useful in order to illustrate some global properties of the evolution of the plasma which are relevant both in the evolution of large - scale magnetic fields prior to recombination and in the discussion of the gravitational collapse in the presence of magnetic fields .
these global properties of the plasma evolution go under the name of alfvn theorems @xcite ( see also @xcite ) . in the ideal limit both the magnetic flux and the magnetic helicity
are conserved .
this means , @xmath175 where @xmath176 is an arbitrary closed surface which moves with the plasma . )
has been derived without the @xmath177 term at the right hand side because of different conventions . ] if @xmath165 the expression appearing at the right hand side is sub - leading and the magnetic flux lines evolve glued to the plasma element .
the other quantity which is conserved in the superconducting limit is the magnetic helicity @xmath178 where @xmath179 is the vector potential whose presence may lead to think that the whole expression ( [ h1 ] ) is not gauge - invariant . in fact @xmath180 is not gauge invariant but , none the less , @xmath181 is gauge - invariant since the integration volume is defined in such a way that the magnetic field @xmath182 is parallel to the surface which bounds @xmath183 and which we will call @xmath184 .
if @xmath185 is the unit vector normal to @xmath184 then @xmath186 on @xmath184 and the gauge dependent contribution to the integral appearing in eq .
( [ h1 ] ) vanishes . in physical terms
the integration may always be performed imagining that the space is sliced in flux tubes of the magnetic field .
this procedure is correct provided the magnetic flux is , at least approximately , conserved as implied by eq .
( [ flux ] ) .
the magnetic gyrotropy @xmath187 it is a gauge invariant measure of the diffusion rate of @xmath181 . in fact , in the resistive approximation @xcite @xmath188 as in the case of eq .
( [ h1 ] ) , for @xmath189 the magnetic helicity is approximately constant .
the value of the magnetic gyrotropy allows to distinguish different mechanisms for the magnetic field generation .
some mechanisms are able to produce magnetic fields whose flux lines have a non - trivial gyrotropy
. the properties of turbulent magnetized plasmas may change depending upon the value of the gyrotropy and of the helicity @xcite .
the physical interpretation of the flux and magnetic helicity conservation is the following . in mhd
the magnetic field has to be always solenoidal ( i.e. @xmath190 ) .
thus , the magnetic flux conservation implies that , in the superconducting limit ( i.e. @xmath165 ) the magnetic flux lines , closed because of the transverse nature of the field , evolve always glued together with the plasma element . in this approximation , as far as the magnetic field evolution is concerned , the plasma is a collection of ( closed ) flux tubes .
the theorem of flux conservation states then that the energetical properties of large - scale magnetic fields are conserved throughout the plasma evolution . while the flux conservation concerns the energetical properties of the magnetic flux lines , the magnetic helicity concerns chiefly the _ topological _ properties of the magnetic flux lines .
if the magnetic field is completely stochastic , the magnetic flux lines will be closed loops evolving independently in the plasma and the helicity will vanish
. there could be , however , more complicated topological situations @xcite where a single magnetic loop is twisted ( like some kind of mbius stripe ) or the case where the magnetic loops are connected like the rings of a chain . in both cases the magnetic helicity will not be zero since it measures , essentially , the number of links and twists in the magnetic flux lines .
furthermore , in the superconducting limit , the helicity will not change throughout the time evolution .
the conservation of the magnetic flux and of the magnetic helicity is a consequence of the fact that , in ideal mhd , the ohmic electric field is always orthogonal both to the bulk velocity field and to the magnetic field . in the resistive mhd approximation
this is no longer true .
if the conductivity is very large ( but finite ) , the resistive mhd approximation suggests , on one hand , that the magnetic flux is only approximately conserved . on the other hand
the approximate conservation of the magnetic helicity implies that the closed magnetic loops may modify their topological structure .
the breaking of the flux lines , occurring at finite conductivity , is related to the possibility that bits of magnetic fluxes are ejected from a galaxy into the inter - galactic medium .
this phenomenon is called _ magnetic reconnection _ and it is the basic mechanism explaining , at least qualitatively , why , during the solar flares , not only charged particles are emitted , but also magnetic fields . in the context of large scale magnetic fields
the approximate ( or exact ) magnetic flux conservation has relevant consequences for the rle of magnetic fields during gravitational collapse . from eqs .
( [ mhd1 ] ) and ( [ genohm2 ] ) the ohmic electric field can be expressed as @xmath191 which inserted into eq .
( [ mx2 ] ) leads to the magnetic diffusivity equation @xmath192 the first term of eq .
( [ magndiff1 ] ) is the dynamo term . the second term of eq .
( [ magndiff1 ] ) is the magnetic diffusivity term whose effect is to dissipate the magnetic field . by comparing the left and the right hand side of eq .
( [ magndiff1 ] ) , the typical time scale of resistive phenomena is @xmath193 where @xmath44 is the typical length scale of the magnetic field . in a non - relativistic plasma
the conductivity @xmath194 goes typically as @xmath195 @xcite . in the case of planets , like the earth
, one can wonder why a sizable magnetic field can still be present .
one of the theories is that the dynamo term regenerates continuously the magnetic field which is dissipated by the diffusivity term @xcite . in the case of the galactic disk the value of the conductivity . in the case of the galactic disks @xmath196 .
the variable @xmath197 denotes , in the present review , the conformal time coordinate . ]
is given by @xmath198 .
thus , for @xmath199 @xmath200 . in eq .
( [ difftime ] ) the typical time of resistive phenomena has been introduced .
( [ difftime ] ) can also give the typical resistive length scale once the time - scale of the system is specified .
suppose that the time - scale of the system is given by @xmath201 where @xmath202 is the present value off the hubble parameter .
then @xmath203 leading to @xmath204 .
the scale ( [ diffscale ] ) gives then the upper limit on the diffusion scale for a magnetic field whose lifetime is comparable with the age of the universe at the present epoch .
magnetic fields with typical correlation scale larger than @xmath205 are not affected by resistivity . on the other hand , magnetic fields with typical correlation scale @xmath206
are diffused .
the value @xmath207 is consistent with the phenomenological evidence that there are no magnetic fields coherent over scales smaller than @xmath208 pc .
the dynamo term may be responsible for the origin of the magnetic field of the galaxy .
the galaxy has a typical rotation period of @xmath209 yrs and comparing this number with the typical age of the galaxy , @xmath210 , it can be appreciated that the galaxy performed about @xmath6 rotations since the time of the protogalactic collapse . from eq .
( [ magndiff1 ] ) the usual structure of the dynamo term may be derived by carefully averaging over the velocity filed according to the procedure of @xcite . by assuming that the motion of the fluid is random and with zero mean velocity the average is taken over the ensemble of the possible velocity fields . in more physical terms
this averaging procedure of eq .
( [ magndiff1 ] ) is equivalent to average over scales and times exceeding the characteristic correlation scale and time @xmath211 of the velocity field .
this procedure assumes that the correlation scale of the magnetic field is much bigger than the correlation scale of the velocity field which is required to be divergence - less ( @xmath212 ) . in this approximation
the magnetic diffusivity equation can be written as : @xmath213 where @xmath214 is the so - called dynamo term which vanishes in the absence of vorticity . in eqs .
( [ dynamored])([alphafirst ] ) @xmath182 is the magnetic field averaged over times longer that @xmath211 which is the typical correlation time of the velocity field . it can be argued that the essential requirement for the consistence of the mentioned averaging procedure is that the turbulent velocity field has to be `` globally '' non - mirror symmetric @xcite .
if the system would be , globally , invariant under parity transformations , then , the @xmath16 term would simply vanish .
this observation is related to the turbulent features of cosmic systems . in cosmic turbulence
the systems are usually rotating and , moreover , they possess a gradient in the matter density ( think , for instance , to the case of the galaxy ) .
it is then plausible that parity is broken at the level of the galaxy since terms like @xmath215 are not vanishing @xcite . the dynamo term , as it appears in eq .
( [ dynamored ] ) , has a simple electrodynamical meaning , namely , it can be interpreted as a mean ohmic current directed along the magnetic field : @xmath216 this equation tells us that an ensemble of screw - like vortices with zero mean helicity is able to generate loops in the magnetic flux tubes in a plane orthogonal to the one of the original field .
consider , as a simple application of eq .
( [ dynamored ] ) , the case where the magnetic field profile is given by @xmath217 for this profile the magnetic gyrotropy is non - vanishing , i.e. @xmath218 . from eq .
( [ dynamored ] ) , using eq .
( [ knot1 ] ) @xmath219 obeys the following equation @xmath220 admits exponentially growing solutions for sufficiently large scales , i.e. @xmath221 .
notice that in this naive example the @xmath16 term is assumed to be constant .
however , as the amplification proceeds , @xmath16 may develop a dependence upon @xmath222 , i.e. @xmath223 $ ] . in the case of eq .
( [ f1 ] ) this modification will introduce non - linear terms whose effect will be to stop the growth of the magnetic field .
this regime is often called saturation of the dynamo and the non - linear equations appearing in this context are sometimes called landau equations @xcite in analogy with the landau equations appearing in hydrodynamical turbulence . in spite of the fact that in the previous example the velocity field has been averaged , its evolution obeys the navier - stokes equation @xmath224 = - \vec{\nabla}p + \vec{j}\times \vec{b } , \label{ns1}\ ] ] where @xmath38 is the thermal viscosity coefficient . since in mhd
the matter current is solenoidal ( i.e. @xmath225 ) the incompressible closure @xmath226 , corresponds to a solenoidal velocity field @xmath227 .
recalling eq .
( [ mhd1 ] ) , the lorentz force term can be re - expressed through vector identities and eq .
( [ ns1 ] ) becomes @xmath224 = - \vec{\nabla}\biggl [ p + \frac{|\vec{b}|^2}{8\pi } \biggr ] + ( \vec{b}\cdot\vec{\nabla } ) \vec{b}. \label{ns2}\ ] ] in typical applications to the evolution of magnetic fields prior to recombination the magnetic pressure term is always smaller than the fluid pressure is usually defined . if the plasma is confined , then @xmath228 is of order 1 . on the contrary , if @xmath229 , as implied by the critical density bound in the early universe , then the plasma may be compressed at higher temperatures and densities . ] , i.e. @xmath230 . furthermore , there are cases where the lorentz force term can be ignored .
this is the so - called force free approximation . defining the kinetic helicity as @xmath231 , the magnetic diffusivity and navier - stokes equations
can be written in a rather simple and symmetric form @xmath232 in mhd various dimensionless ratios can be defined .
the most frequently used are the magnetic reynolds number , the kinetic reynolds number and the prandtl number : @xmath233 where @xmath234 and @xmath235 are the typical scales of variation of the magnetic and velocity fields . in the absence of pressure and density perturbations the combined system of eqs .
( [ mhd1 ] ) and ( [ ns2 ] ) can be linearized easily .
using then the incompressible closure the propagating modes are the alfvn waves whose typical dispersion relation is @xmath236 where @xmath237
. often the lundqvist number is called , in plasma literature @xcite magnetic reynolds number .
this confusion arises from the fact that the lunqvist number , i.e. @xmath238 , is the magnetic reynolds number when @xmath239 coincides with the alfvn velocity . to have a very large lundqvist number implies that the the conductivity is very large . in this sense
the lunqvist number characterizes , in fusion theory , the rate of growth of resistive instabilities and it is not necessarily related to the possible occurrence of turbulent dynamics . on the contrary , as large reynolds numbers are related to the occurrence of hydrodynamical turbulence , large _ magnetic _
reynolds numbers are related to the occurence of mhd turbulence @xcite . according to the naive description of the dynamo instability presented above the origin of large - scale magnetic fields in spiral galaxies
can be reduced to the following steps : * during the @xmath6 rotations performed by the galaxy since the protogalactic collapse , the magnetic field should be amplified by about @xmath6 e - folds ; * if the large scale magnetic field of the galaxy is , today , @xmath240 the magnetic field at the onset of galactic rotation might have been even @xmath6 e - folds smaller , i.e. @xmath241 ; * assuming perfect flux freezing during the gravitational collapse of the protogalaxy ( i.e. @xmath165 ) the magnetic field at the onset of gravitational collapse should be @xmath242 g over a typical scale of 1 mpc .
this picture is oversimplified and each of the three steps mentioned above will be contrasted with the most recent findings .
the idea that a celestial body may acquire a magnetic field by differential rotation can be traced back to the paper of larmor of 1919 @xcite .
one of the early ancestors of the dynamo mechanism , can be traced back to the model of fermi and chandrasekar @xcite . in @xcite
the attempt was to connect the existence off the galactic magnetic field with the existence of a galactic angular momentum . later on dynamo theory
has been developed in greater detail ( see @xcite ) and its possible application to large - scale magnetic fields has been envisaged .
the standard dynamo theory has been questioned in different ways .
piddington @xcite pointed out that small - scale magnetic fields can grow large enough ( until equipartition is reached ) to swamp the dynamo action .
the quenching of the dynamo action has been numerically shown by kulsrud and anderson @xcite .
more recently , it has been argued that if the large - scale magnetic field reaches the critical value is the velocity field at the outer scale of turbulence . ]
@xmath243 the dynamo action could also be quenched @xcite .
( [ magndiff1 ] ) is exact , in the sense that both @xmath157 and @xmath182 contain long and short wavelength modes .
the aim of the various attempts of the dynamo theory is to get an equation describing only the `` mean value '' of the magnetic field . to this end
the first step is to separate the exact magnetic and velocity fields as @xmath244 where @xmath245 and @xmath246 are the averages over an ensemble of many realizations of the velocity field @xmath157 . in order to derive the standard form of the dynamo equations few important assumptions
should be made .
these assumptions can be summarized as follows : * the scale of variation of the turbulent motion @xmath247 should be smaller than the typical scale of variation of @xmath248 . in the galactic problem
@xmath249 is the differential rotation of the galaxy , while @xmath247 is the turbulent motion generated by stars and supernovae .
typically the scale of variation of @xmath157 is less than @xmath71 pc while the interesting scales for @xmath245 are larger than the kpc ; * the field @xmath250 is such that @xmath251 .
* it should happen that @xmath252 .
* magnetic flux is frozen into the plasma ( i.e. magnetic flux is conserved ) . from the magnetic diffusivity equation ( [ magndiff1 ] ) , and using the listed assumptions , it is possible to derive the typical structure of the dynamo term by carefully averaging over the velocity field according to the procedure outlined in @xcite . inserting eq .
( [ sep ] ) into ( [ magndiff1 ] ) and breaking the equation into a mean part and a random part , two separate induction equations can be obtained for the mean and random parts of the magnetic field @xmath253 where the ( magnetic ) diffusivity terms have been neglected . in eq .
( [ mean ] ) , @xmath254 is called mean field ( or turbulent ) electromotive force and it is the average of the cross product of the small - scale velocity field @xmath247 and of the small scale magnetic field @xmath250 over a scale much smaller than the scale of @xmath245 but much larger than the scale of turbulence . sometimes , the calculation of the effect of @xmath255 is done in the case of incompressible and isotropic turbulence . in this case @xmath256 .
this estimate is , however , not realistic since @xmath248 is not isotropic .
more correctly @xcite , @xmath255 should be evaluated by using eq .
( [ random ] ) which is usually written in a simplified form @xmath257 where all but the first term of eq .
( [ random ] ) have been neglected . to neglect the term @xmath258 does not pose any problem since it corresponds to choose a reference frame where @xmath246 is constant . however , the other terms , neglected in eq .
( [ inran ] ) , are dropped because it is assumed that @xmath259 .
this assumption may not be valid all the time and for all the scales .
the validity of eq .
( [ inran ] ) seems to require that @xmath194 is very large so that magnetic diffusivity can keep always @xmath250 small @xcite . on the other hand
@xcite one can argue that @xmath250 is only present over very small scales ( smaller than @xmath71 pc ) and in this case the approximate form of eq .
( [ inran ] ) seems to be more justified . from eqs .
( [ mean])([inran ] ) it is possible to get to the final result for the evolution equation of @xmath245 @xcite as it is usally quoted @xmath260 where @xmath261 where @xmath16 is the dynamo term , @xmath262 is the diffusion term and @xmath263 is the typical correlation time of the velocity field .
the term @xmath16 is , in general , space - dependent .
the standard lore is that the dynamo action stops when the value of the magnetic field reaches the equipartition value ( i.e. when the magnetic and kinetic energy of the plasma are comparable ) . at this point
the dynamo `` saturates '' .
the mean velocity field can be expressed as @xmath264 where @xmath265 is the angular velocity of differential rotation at the galactocentric radius @xmath266 . in the case of flat rotation curve @xmath267 which also implies that @xmath268 .
( [ fin ] ) can then be written in terms of the radial and azimuthal components of the mean magnetic field , neglecting , for simplicity , the diffusivity term : @xmath269 the second equation shows that the @xmath16 effect amplifies the radial component of the large - scale field .
then , through eq .
( [ azim ] ) the amplification of the radial component is converted into the amplification of the azimuthal field , this is the @xmath270 effect .
usually the picture for the formation of galactic magnetic fields is related to the possibility of implementing the dynamo mechanism . by comparing the rotation period with the age of the galaxy ( for a universe with @xmath271 , @xmath272 and @xmath273 ) the number of rotations performed by the galaxy since its origin is approximately @xmath6 . during these @xmath6 rotations the dynamo term of eq .
( [ fin ] ) dominates against the magnetic diffusivity . as a consequence an instability develops
this instability can be used in order to drive the magnetic field from some small initial condition up to its observed value .
( [ fin ] ) is linear in the mean magnetic field .
hence , initial conditions for the mean magnetic field should be postulated at a given time and over a given scale .
this initial mean field , postulated as initial condition of ( [ fin ] ) is usually called seed .
most of the work in the context of the dynamo theory focuses on reproducing the correct features of the magnetic field of our galaxy .
the achievable amplification produced by the dynamo instability can be at most of @xmath274 , i.e. @xmath275 .
thus , if the present value of the galactic magnetic field is @xmath276 gauss , its value right after the gravitational collapse of the protogalaxy might have been as small as @xmath277 gauss over a typical scale of @xmath6@xmath71 kpc . there is a simple way to relate the value of the magnetic fields right after gravitational collapse to the value of the magnetic field right before gravitational collapse . since the gravitational collapse occurs at high conductivity the magnetic flux and the magnetic helicity
are both conserved .
right before the formation of the galaxy a patch of matter of roughly @xmath278 mpc collapses by gravitational instability .
right before the collapse the mean energy density of the patch , stored in matter , is of the order of the critical density of the universe .
right after collapse the mean matter density of the protogalaxy is , approximately , six orders of magnitude larger than the critical density .
since the physical size of the patch decreases from @xmath278 mpc to @xmath6 kpc the magnetic field increases , because of flux conservation , of a factor @xmath279 where @xmath280 and @xmath281 are , respectively the energy densities right after and right before gravitational collapse .
the correct initial condition in order to turn on the dynamo instability would be @xmath282 gauss over a scale of @xmath278 mpc , right before gravitational collapse .
this last estimate is rather generous and has been presented just in order to make contact with several papers ( concerned with the origin of large scale magnetic fields ) using such an estimate .
the estimates presented in the last paragraph are based on the ( rather questionable ) assumption that the amplification occurs over thirty e - folds while the magnetic flux is completely frozen in . in the real situation ,
the achievable amplification is much smaller .
typically a good seed would not be @xmath277 g after collapse ( as we assumed for the simplicity of the discussion ) but rather @xcite @xmath283 the possible applications of dynamo mechanism to clusters is still under debate and it seems more problematic @xcite .
the typical scale of the gravitational collapse of a cluster is larger ( roughly by one order of magnitude ) than the scale of gravitational collapse of the protogalaxy .
furthermore , the mean mass density within the abell radius ( @xmath284 mpc ) is roughly @xmath13 larger than the critical density .
consequently , clusters rotate much less than galaxies . recall that clusters are formed from peaks in the density field .
the present overdensity of clusters is of the order of @xmath13 .
thus , in order to get the intra - cluster magnetic field , one could think that magnetic flux is exactly conserved and , then , from an intergalactic magnetic field @xmath285 g an intra cluster magnetic field @xmath286 g can be generated .
this simple estimate shows why it is rather important to improve the accuracy of magnetic field measurements in the intra - cluster medium discussed in section 3 : the change of a single order of magnitude in the estimated magnetic field may imply rather different conclusions for its origin .
recent numerical simulations seem to support the view that cluster magnetic fields are entirely primordial @xcite .
as the temperature increases above @xmath278 mev , a relativistic plasma becomes almost a perfect conductor .
there are two complementary effects associated with large - scale magnetic fields in curved backgrounds : * large - scale magnetic fields may evolve over a rigid background space - time determined , for instance , by the dynamics of barotropic fluids and , in this case the energy density of the magnetic field must always be smaller than the energy density of the fluid sources ; * magnetic fields may be so intense to modify the structure of the space - time and , in this case , their energy density is comparable with the energy density of the other sources of the geometry . the first effect will be examined in the present section , while the discussion of the second effect , which is more speculative , will be confined to section 9 . the standard model of cosmological evolution rests on three important assumptions @xcite .
the first assumption is that over very large length scales ( greater than @xmath287 mpc ) the universe is described by a homogeneous and isotropic friedmann - robertson - walker ( frw ) line element : @xmath288 , \label{metric}\ ] ] where @xmath289 is the scale factor and @xmath197 the conformal - time coordinate ( notice that eq .
( [ metric ] ) has been written , for simplicity , in the conformally flat case ) .
the second hypothesis is that the sources of the evolution of the background geometry are perfect fluid sources . as a consequence
the entropy of the sources is constant .
the third and final hypothesis is that the dynamics of the sources and of the geometry is dictated by the general relativistic frw equations : will be used . ]
@xmath290 where @xmath291 and the prime related to conformal time as @xmath292 . ] denotes the derivation with respect to @xmath197 .
recall also , for notational convenience , that @xmath293 where @xmath294 is the conventional hubble parameter .
the various tests of the standard cosmological model are well known @xcite .
probably one of the most stringent one comes from the possible distortions , in the rayleigh - jeans region , of the cmb spectrum .
the absence of these distortions clearly rules out steady - state cosmological models . in the standard cosmological model one
usually defines the proper distance of the event horizon at the time @xmath295 @xmath296 this distance represents the maximal extension of the region over which causal connection is possible .
furthermore , the proper distance of the particle horizon can also be defined : @xmath297 if the scale factor is parametrized as @xmath298 , for @xmath299 the universe experiences a decelerated expansion , i.e. @xmath300 and @xmath301 while the curvature scale decreases , i.e. @xmath302 .
this is the peculiar behaviour when the fluid sources are dominated either by dust ( @xmath303 ) or by radiation ( @xmath304 ) . in the case of the standard model ,
the particle horizon increases linearly in cosmic time ( therefore faster than the scale factor ) .
this implies that the cmb radiation , today observed with a temperature of @xmath305 over the whole present horizon has been emitted from space - time regions which were not in causal contact .
this problem is known as the horizon problem of the standard cosmological model .
the other problem of the standard model is related to the fact that today the intrinsic ( spatial ) curvature @xmath306 is smaller than the extrinsic curvature , i.e. @xmath307 . recalling that @xmath308 it is clear that if @xmath309 , @xmath310 increases so that the intrinsic curvature could be , today , arbitrarily large .
the third problem of the standard cosmological model is related to the generation of the large entropy of the present universe .
the solution of the kinematical problems of the standard model is usually discussed in the framework of a phase of accelerated expansion @xcite , i.e. @xmath311 and @xmath312 . in the case of inflationary dynamics
the extension of the causally connected regions grows as the scale factor and hence faster than in the decelerated phase .
this solves the horizon problem .
furthermore , during inflation the contribution of the spatial curvature becomes very small .
the way inflation solves the curvature problem is by producing a very tiny spatial curvature at the onset of the radiation epoch taking place right after inflation .
the spatial curvature can well grow during the decelerated phase of expansion but is will be always subleading provided inflation lasted for sufficiently long time .
in fact , the minimal requirement in order to solve these problems is that inflation lasts , at least , @xmath313-efolds .
the final quantity which has to be introduced is the hubble radius @xmath314 .
this quantity is local in time , however , a sloppy nomenclature often exchanges the hubble radius with the horizon .
since this terminology is rather common , it will also be used here . in the following applications it will be relevant to recall some of the useful thermodynamical relations . in particular , in radiation dominated universe , the relation between the hubble parameter and the temperature is given by @xmath315 where @xmath316 is the effective number of relativistic degrees of freedom at the corresponding temperature .
( [ frw1 ] ) , implying @xmath317 , has been used in eq .
( [ therm1 ] ) together with the known relations valid in a radiation dominated background @xmath318 where @xmath319 . in eq .
( [ therm2 ] ) the thermodynamical relation for the number density @xmath320 has been also introduced for future convenience .
the inflationary dynamics can be realized in different ways .
conventional inflationary models are based either on one single inflaton field @xcite or on various fields @xcite ( see @xcite for a review ) .
furthermore , in the context of single - field inflationary models one oughts to distinguish between small - field models @xcite ( like in the so - called new - inflationary models ) and large - field models @xcite ( like in the case of chaotic models ) . in spite of their various quantitative differences , conventional inflationary models
are based on the idea that during the phase of accelerated expansion the curvature scale is approximately constant ( or slightly decreasing ) .
after inflation , the radiation dominated phase starts .
it is sometimes useful for numerical estimates to assume that radiation suddenly dominates at the end of inflation . in this case
the scale factor can be written as @xmath321 where @xmath16 is some effective exponent parameterizing the dynamics of the primordial phase of the universe .
notice that if @xmath322 we have that the primordial phase coincides with a de sitter inflationary epoch .
the case @xmath322 is not completely realistic since it corresponds to the case where the energy - momentum tensor is simply given by a ( constant ) cosmological term . in this case
the scalar fluctuations of the geometry are not amplified and the large - scale angular anisotropy in the cmb would not be reproduced .
the idea is then to discuss more realistic energy - momentum tensors leading to a dynamical behaviour close to the one of pure de sitter space , hence the name quasi - de sitter space - times .
quasi - de sitter dynamics can be realized in different ways .
one possibility is to demand that the inflaton slowly rolls down from its potential obeying the approximate equations @xmath323 valid provided @xmath324 and @xmath325 .
there also exist exact inflationary solutions like the power - law solutions obtainable in the case of exponential potentials : @xmath326 since , from the exact equations , @xmath327 the two slow - roll parameters can also be written as @xmath328 in the case of the exponential potential ( [ pot ] ) the slow - roll parameters are all equal , @xmath329 .
typical potentials leading to the usual inflationary dynamics are power - law potentials of the type @xmath330 , exponential potentials , trigonometric potentials and nearly any potential satisfying , in some region of the parameter space , the slow - roll conditions .
inflation can also be realized in the case when the curvature scale is increasing , i.e. @xmath331 and @xmath332 .
this is the case of superinflationary dynamics .
for instance the propagation of fundamental strings in curved backgrounds @xcite may lead to superinflationary solutions @xcite .
a particularly simple case of superinflationary solutions arises in the case when internal dimensions are present . a model of background evolution
can be generically written as @xmath342 in the parameterization of eq .
( [ back ] ) the internal dimensions grow ( in conformal time ) for @xmath343 and they shrink for @xmath344
. must be effectively constant for practical purposes . ]
superinflationary solutions are also common in the context of the low - energy string effective action @xcite . in critical superstring theory
the electromagnetic field @xmath345 is coupled not only to the metric ( @xmath346 ) , but also to the dilaton background ( @xmath347 ) . in the low energy limit such interaction
is represented by the string effective action @xcite , which reads , after reduction from ten to four expanding dimensions , @xmath348 \label{action4}\end{aligned}\ ] ] were @xmath349 controls the tree - level four - dimensional gauge coupling ( @xmath115 being the ten - dimensional dilaton field , and @xmath350 the volume of the six - dimensional compact internal space ) .
the field @xmath194 is the kalb - ramond axion whose pseudoscalar coupling to the gauge fields may also be interesting . in the inflationary models based on the above effective action
@xcite the dilaton background is not at all constant , but undergoes an accelerated evolution from the string perturbative vacuum ( @xmath351 ) towards the strong coupling regime , where it is expected to remain frozen at its present value . the peculiar feature of this string cosmological scenario ( sometimes called pre - big bang @xcite ) is that not only the curvature evolves but also the gauge coupling .
suppose , for the moment that the gauge fields are set to zero .
the phase of growing curvature and dilaton coupling ( @xmath352 , @xmath353 ) , driven by the kinetic energy of the dilaton field , is correctly described in terms of the lowest order string effective action only up to the conformal time @xmath354 at which the curvature reaches the string scale @xmath355 ( @xmath356 is the fundamental length of string theory ) .
a first important parameter of this cosmological model is thus the value @xmath357 attained by the dilaton at @xmath354 .
provided such value is sufficiently negative ( i.e. provided the coupling @xmath358 is sufficiently small to be still in the perturbative region at @xmath354 ) , it is also arbitrary , since there is no perturbative potential to break invariance under shifts of @xmath347 .
for @xmath359 high - derivatives terms ( higher orders in @xmath360 ) become important in the string effective action , and the background enters a genuinely stringy " phase of unknown duration .
an assumption of string cosmology is that the stringy phase eventually ends at some conformal time @xmath361 in the strong coupling regime . at this time
the dilaton , feeling a non - trivial potential , freezes to its present constant value @xmath362 and the standard radiation - dominated era starts
. the total duration @xmath363 , or the total red - shift @xmath364 encountered during the stringy epoch ( i.e. between @xmath365 and @xmath361 ) , will be the second crucial parameter besides @xmath366 entering our discussion .
for the purpose of this paper , two parameters are enough to specify completely our model of background , if we accept that during the string phase the curvature stays controlled by the string scale , that is @xmath367 ( @xmath368 is the planck mass ) for @xmath369 . during the string era @xmath370 and @xmath371 are approximately constant , while , during the dilaton - driven epoch @xmath372 here @xmath373 represents the possible effect of internal dimensions whose radii @xmath374 shrink like @xmath375 for @xmath376 ( for the sake of definiteness we show in the figure the case @xmath377 ) .
the shape of the coupling curve corresponds to the fact that the dilaton is constant during the radiation era , that @xmath370 is approximately constant during the string era , and that it evolves like @xmath378 during the dilaton - driven era .
an interesting possibility , in the pre - big bang context is that the exit to the phase of decelerated expansion and decreasing curvature takes place without any string tension correction .
recently a model of this kind has been proposed @xcite .
the idea consists in adding a non - local dilaton potential which is invariant under scale factor duality .
the evolution equations of the metric and of the dilaton will then become , in 4 space - time dimensions , @xmath379 where @xmath380 is the shifted dilaton . a particular solution to these equations
will be given by @xmath381^{1/\sqrt{3 } } , \label{scalefactor1}\\ & & \overline{\varphi } = - \frac{1}{2 } \log { ( 1 + \tau^2 ) } + \varphi_{0 } , \label{dilaton1}\end{aligned}\ ] ] where @xmath382 this solution interpolates smoothly between two self - dual solutions .
for @xmath383 the background superinflates while for @xmath384 the decelerated frw limit is recovered . in a frw metric of the type ( [ metric ] )
the evolution equations of the gauge fields are invariant under weyl rescaling of the four - dimensional metric .
this property is often named conformal invariance .
consider , for simplicity , the evolution equation of sourceless maxwell fields : @xmath385 = 0 . \label{conformal}\
] ] since the metric ( [ metric ] ) is conformally flat , the time dependence arising in eq .
( [ conformal ] ) from @xmath386 , is always compensated by the controvariant indices of the fields strength .
maxwell fields in conformally flat metrics obey the same equations obtained in the case of minkowski space - time provided the conformal time coordinate is used and provided the fields are appropriately rescaled .
this property holds not only for maxwell fields but also for chiral fermions in frw backgrounds . using weyl invariance , the generalization of mhd equations to the case of conformally flat frw space - times
can then be easily obtained : @xmath387 where @xmath388 are the curved space fields expressed as a function of the corresponding flat space quantities . in the case of a radiation dominated background ,
the navier - stokes equation takes the form @xmath389 ' + \vec{v}\cdot\vec{\nabla}[(\overline{p } + \overline{\rho})\vec{v } ] = - \vec{\nabla}\biggl[\overline{p } + \frac { |\vcb|^2]}{8\pi } \biggr ] + \frac{[\vcb\cdot\vec{\nabla}]\vcb}{4\pi } + ( \overline{p } + \overline{\rho})\nu \nabla^2 \vec{v } , \label{cns}\ ] ] where the incompressible closure , i.e. @xmath390 , has been already adopted and where the vector identity @xmath391 \vcb$ ] has been used . in eq .
( [ cns ] ) @xmath392 and @xmath393 . in terms of the rescaled
pressure and energy densities the continuity equation becomes @xmath394 = 0 , \label{conteq}\ ] ] where a barotropic equation of state has been assumed for the background fluid , i.e. @xmath395 . the energy density is homogeneous and hence the last term in eq .
( [ conteq ] ) can be consistently dropped within the incompressible closure .
however , one may ought to linearize eq .
( [ conteq ] ) and this is why the last term may be kept for future considerations . if @xmath396 , then eq . ( [ conteq ] ) implies @xmath397 . for generic barotropic index eq .
( [ cns ] ) becomes @xmath398\vcb}{4\pi ( \gamma + 1 ) \overline{\rho } } + \nu \nabla^2 \vec{v } , \label{cnsgen}\ ] ] where the magnetic pressure term has been dropped since it is negligible with respect to @xmath399 .
recalling that for @xmath400 the magnetic diffusivity equation can be written , using vector identities , as @xmath401 \vcb= [ \vcb\cdot\vec{\nabla}]\vec{v } + \frac{1}{4\pi\sigma } \nabla^2 \vcb . \label{dif}\ ] ] eqs .
( [ cmh1])([cns ] ) form a closed system sharing analogies with their flat space counterpart .
thanks to conformal invariance , various solutions can be lifted from flat to curved space like , for instance @xcite , the following fully nonlinear solution of eqs .
( [ dif ] ) and ( [ cns ] ) @xmath402 valid if thermal and magnetic diffusivities are neglected and if @xmath403 .
( [ cmh1])([cns ] ) present also some differences with respect to the flat space case : since the universe is expanding , the relative balance between diffusion scales may change as time goes by .
all the propagation of mhd excitations can be then generalized to the curved - space case .
in various papers @xcite this program has been achieved ( see also @xcite for earlier attempts ) . in @xcite the evolution of magnetic fields with non - vanishing magnetic helicity has been studied using the resistive mhd approximation in curved space . in @xcite the evolution of non - linear alfvn waves with the purpose of deriving possible observable effects @xcite . in order to analyze the high - frequency spectrum of plasma excitations
, the vlasov - landau approach can be generalized to curved spaces @xcite .
there are two options in order to achieve this goal .
the first option is to assume that the magnetic field breaks the isotropy of space - time .
this approach leads to the study of plasma effects in a magnetic field in full analogy with what is done in terrestrial tokamaks .
the second choice is to discuss magnetic fields which are stochastic and which do not break the isotropy . later in the present review
the effects of magnetic fields breaking the isotropy of space - time will be considered but , for the moment , the attention will be concentrated on the second approach .
a typical problem which should be discussed within the kinetic approach is the relaxation of charge and current density fluctuations in a relativistic plasma @xcite .
the problem is then to compute what magnetic field is induced from charge and current density fluctuations .
consider an equilibrium homogeneous and isotropic conducting plasma , characterized by a distribution function @xmath404 common for both positively and negatively charged ultrarelativistic particles ( for example , electrons and positrons ) .
the normalization is chosen in such a way that @xmath405 . within the kinetic approach , the initial charge and current density fluctuations
are related to the perturbations of the initial value of the distribution function around the equilibrium distribution ( [ dist0 ] ) , i.e @xmath406 where @xmath407 refers to positrons and @xmath408 to electrons , and @xmath409 is the conformal momentum .
the vlasov equation defining the curved - space evolution of the perturbed distributions can be written as @xmath410 where the two terms appearing at the right hand side of each equation are the collision terms .
this system of equation represents the curved space extension of the vlasov - landau approach to plasma fluctuations @xcite .
all particle number densities here are related to the comoving volume .
notice that , in general @xmath411 . in the ultra - relativistic limit @xmath412 and the vlasov equations are conformally invariant .
this implies that , provided we use conformal coordinates and rescaled gauge fields , the system of equations which we would have in flat space @xcite looks exactly the same as the one we are discussing in a curved frw ( spatially flat ) background @xcite .
the evolution equations of the gauge fields coupled to the two vlasov equations can be written as @xmath413 , \nonumber\\ & & \vec{\nabla}\times \vec{\cal e } + \vec{b } ' = 0 , \nonumber\\ & & \vec{\nabla}\cdot \vec{\cal b } = 0 , \nonumber\\ & & \vec{\nabla}\times \vec{\cal b } -\vec{\cal e}'= 4\pi e \int d^3 p \vec{v } [ f_{+}(\vec{x } , \vec{p},\eta ) - f_{-}(\vec{x } , \vec{p},\eta ) ] .
\label{max}\end{aligned}\ ] ] in order to illustrate the physical relevance of this description , eqs .
( [ max ] ) will be used later in this section to study the fate of charge and current density fluctuations . when mhd is discussed in curved spce - times ,
the evolution of the geometry reflects in the time evolution of the physical scales of the problem . from eq .
( [ dif ] ) , the comoving momentum corresponding to the magnetic diffusivity scale is given by the relation @xmath414 in the case of a relativistic plasma at typical temperatures @xmath415 , the flat - space conductivity ( see eq .
( [ curvflat ] ) ) is given by @xmath416 where @xmath417 is the cross section . while the conductivity is @xmath194 .
] then , it can be shown @xcite that @xmath418 where @xmath419 is a slowly increasing function of the temperature .
in @xcite it has been convincingly argued that @xmath420 at @xmath421 mev while it becomes @xmath422 at @xmath423 ( for a recent analysis of conductivity in high temperature qed see @xcite ) .
recalling eq .
( [ curvflat ] ) the physical diffusivity scale @xmath424 will be , in units of the hubble parameter , @xmath425 discounting for the effect due to the possible variations of the effective relativistic degrees of freedom , the ( physical ) magnetic diffusivity length scale is much smaller than the hubble radius at the corresponding epoch .
for instance , if @xmath426 , then @xmath427 .
this means that there are ten decades length - scales where the magnetic field spectrum will not experience diffusion due to the finite value of the conductivity .
the estimate of the magnetic diffusivity scale will now be repeated in a different range of temperatures , namely for @xmath428 when weak interactions have fallen out of thermal equilibrium . in this case
the relevant degrees of freedom are the three neutrino species and the photon , i.e. @xmath429 . in this case
the conductivity will be given by @xmath430 which leads to @xmath431 if @xmath432 and @xmath417 .
( [ nr ] ) leads to the known @xmath195 dependence of the conductivity from the temperature , as expected from a non - relativistic plasma .
the physical magnetic diffusivity momentum is , in this case @xmath433 eqs .
( [ omsig ] ) and ( [ nrdiffscale ] ) refer to two different ranges of temperatures .
the magnetic diffusivity scale in units of the hubble parameter obtained in eq .
( [ nrdiffscale ] ) is smaller than the one given by in eq .
( [ omsig ] ) .
hence , by lowering the temperature across the weak equilibration temperature makes the conductivity length scale larger in units of the hubble radius : _ inertial range _ of length scales ( i.e. the region where conductivity effects are negligible ) is larger above the weak equilibration temperature . summarizing @xmath434 where @xmath435 . in the evolution of the plasma
it is sometimes relevant to estimate the thermal diffusivity scale setting a bound on the coherence of the velocity field and not of the magnetic field .
there are however phenomena , like the propagation of mhd waves , where the excitations of a background magnetic field are coupled to the excitations of the velocity field . generally speaking the typical correlation scale of the velocity field
is always smaller , in various physical situations , than the typical correlation scale of the magnetic fields .
the range of physical momenta affected by thermal diffusivity is bounded from above by the thermal diffusivity scale which can be obtained from the following relations @xmath436 were @xmath437 is the photon mean free path , i.e. @xmath438 the mean free path measures the efficiency of a given process to transfer momentum . from eqs .
( [ cns ] ) and ( [ thdiffmom ] ) the typical thermal diffusivity momentum is : @xmath439 where @xmath440 is the physical momentum as opposed to the comoving momentum @xmath441 .
the scale defined in eq .
( [ thdif2 ] ) is sometimes called silk scale . notice that the mean free path given in eq . ( [ mfp ] )
increases as @xmath442 .
since the physical scale of an inhomogeneous region increases as @xmath289 , the photon mean free path can become larger than the physical size of an inhomogeneous region .
this is the so - called free streaming regime .
if @xmath443 , the photon mean free path is given by eq .
( [ mfp ] ) where @xmath444 is the thompson cross section and @xmath445 is the electron number density depending both on the critical fractions of baryons , @xmath446 and on the ionization fraction @xmath447 .
the ionization fraction is the ratio between the number density of protons and the total number density of baryons .
the value of @xmath447 drops from @xmath278 to @xmath208 in a short time around recombination .
( [ ne1 ] ) is simply derived by noticing that , after electron positron annihilation , @xmath448 where @xmath449 is the baryon - to - photon ratio , the free parameter of big - bang nucleosynthesis calculations @xcite .
thus @xmath450 for temperatures larger than the mev the photon mean free path can be estimated by recalling that , above the mev , the thompson cross section is replaced by the klein - nishina cross section ( i.e. @xmath451 ) .
thus , from eq .
( [ mfp ] ) , the photon mean free path is @xmath452 as before , it is useful to refer @xmath453 to the hubble rate and to compute the typical diffusion distances for @xmath454 @xmath455 and for @xmath456 @xmath457 at temperatures @xmath458 , corresponding roughly to @xmath459 @xmath460 annihilation , eqs .
( [ difm ] ) and ( [ difm ] ) differ by a factor @xmath461 .
this drop of five orders of magnitude of the thermal diffusivity scale reflects the drop
( by ten orders of magnitude ) of the photon mean free path as it can be appreciated by comparing eqs .
( [ mfp1 ] ) and ( [ mfp2 ] ) at @xmath462 . for temperatures smaller than the electroweak temperature ( i.e. @xmath463 gev ) down to the temperature of neutrino decoupling
, neutrinos are the species with the largest mean free path and , therefore , the most efficient momentum transporters .
neglecting possible variations of the number of relativistic species the neutrino mean free path will be simply given by @xmath464 where @xmath465 and @xmath466 are the leptons and quark number densities . using this estimate
the diffusion scale will be @xmath467 which should be compared with eq .
( [ difm ] ) . clearly @xmath468 .
it seems then difficult to get observable effects from inhomogeneities induced by mhd excitations . in @xcite
the damping rates of alfvn and magnetosonic waves have been computed and it has been shown that mhd modes suffer significant damping ( see , however , @xcite ) . to avoid confusion , the results reported in ref .
@xcite apply to the damping of mhd waves and not to the damping of magnetic fields themselves as also correctly pointed out in @xcite .
in standard big bang cosmology there are three rather close ( but rather different ) epochs , * the matter - radiation equality taking place at @xmath469 ev , i.e. @xmath470 sec ; * the decoupling of radiation from matter occurring at @xmath471 i.e. @xmath472 * the recombination occurring , for a typical value of the critical baryon fraction , at @xmath473 , i.e. @xmath474 . after plasma recombines ,
the baryons are no longer prevented from moving by viscosity and radiation pressure .
thus , the only term distinguishing the evolution of the baryons from the evolution of a cold dark matter fluid is the presence of the lorentz force term in the navier - stokes equation . in this situation ,
it is interesting to investigate the interplay between the forced mhd regime and the generation of gravitational inhomogeneities .
it was actually argued long ago by wasserman @xcite and recently revisited by coles @xcite that large - scale magnetic fields may have an important impact on structure formation . assuming that , for illustrative purposes , the matter fluid consists only of baryons and cold dark matter , for @xmath475 the evolution of the system is described by the following set of non relativistic equations
@xmath476 , \label{phi}\end{aligned}\ ] ] where @xmath477 is the newtonian potential . eqs .
( [ vbaryon ] ) come from eq .
( [ cnsgen ] ) for @xmath478 in the case of homogeneous pressure density .
( [ vcold ] ) can be obtained , in the same limit , from eq .
( [ cnsgen ] ) by setting to zero the lorentz force . from eq .
( [ conteq ] ) , by linearizing the continuity equations for the two species , the relation between the divergence of the velocity field and the density contrast can be obtained @xmath479 where @xmath480 and @xmath481 thus , taking the divergence of eqs .
( [ vbaryon ] ) and ( [ vcold ] ) and using , in the obtained equations , eqs .
( [ phi ] ) and ( [ deltav ] ) the following system can be obtained @xmath482}{4\pi \rho_{\rm b } a^4(t ) } , \label{deltab}\\ & & \delta_{\rm c } '' + { \cal h}\delta_{\rm c } - \frac{a^2}{2 m_{\rm p}^2 } \rho_{{\rm m } } \delta = 0 , \label{deltac}\end{aligned}\ ] ] where @xmath483 and @xmath484 is the total matter density .
multiplying eq .
( [ deltab ] ) by @xmath446 and eq .
( [ deltac ] ) by @xmath485 , the sum of the two resulting equations can be written as @xmath486}{4\pi \rho_{\rm b } a^4(t ) } , \label{deltadec}\ ] ] where it has been used that the background geometry is spatially flat .
going to cosmic time , eq .
( [ deltadec ] ) reads @xmath487}{4\pi \rho_{\rm b } a^6(t)}. \label{finalcontrast}\ ] ] the homogeneous term of eq .
( [ finalcontrast ] ) can be easily solved .
the source term strongly depends upon the magnetic field configuration @xcite .
however , as far as time evolution is concerned , @xmath488 is constant in the inertial range of scales , i.e. @xmath489 .
hence , @xmath490 and @xmath491 . assuming a stochastic magnetic field
the solution of equation ( [ finalcontrast ] ) can be obtained integrating from @xmath492 and recalling that @xmath493 .
the result is @xmath494.\ ] ] where @xmath495 is the typical length scale of variation of the magnetic field .
the total density contrast grows as @xmath496 with an initial condition set by the lorentz force , i. e. @xmath497 to seed galaxy directly the effective @xmath498 should be of the order off @xmath499 for a flat universe , implying that @xmath500 also the converse is true
. namely one can read eq .
( [ sform ] ) from left to right . in this case
( [ sform ] ) will define the typical scale affected by the presence of the magnetic field .
this scale is sometimes called magnetic jeans scale @xcite : it can be constructed from the jeans length @xmath501 where @xmath502 is replaced by @xmath503 , i.e. the alfvn velocity . as a simple application of kinetic techniques ,
it is useful to understand which magnetic field can be obtained from an initial inhomogeneous distribution of charge and current density fluctuations .
consider then the system of eqs .
( [ max ] ) under the assumption that , initially , electric and magnetic fields vanish : @xmath504 using eqs .
( [ vl+])([max ] ) the induced magnetic fields can be obtained .. this problem is discussed , in greater detail , in @xcite .
subtracting eqs .
( [ vl+ ] ) and ( [ vl- ] ) , and defining @xmath505 the vlasov - landau system can be written as @xmath506 where @xmath507 is the initial profile of the distribution function and where @xmath508 is the typical collision frequency .
the subscripts @xmath509 denote the fourier transform with respect to space and the laplace transform with respect to time of the corresponding quantity . after enforcing the gauss constraint at the initial time
, the electric field can be separated in its polarizations parallel and transverse to the direction of propagation of the fluctuation .
the transverse current provides a source for the evolution of transverse electric field fluctuations @xmath510 whereas the charge fluctuations provide a source for the evolution of longitudinal electric field fluctuations @xmath511 in eqs .
( [ perp ] ) and ( [ parallel ] ) the longitudinal and transverse part of the electric field fluctuations have been defined : @xmath512 the solution of eq .
( [ vl1 ] ) is given by @xmath513 , \label{f}\ ] ] where we used that @xmath514 .
the longitudinal and transverse components of the electric fluctuations can be obtained by inserting eq .
( [ f ] ) into eqs .
( [ perp])-([parallel ] ) @xmath515 where @xmath516 and @xmath517 are , respectively , the longitudinal and transverse part of the polarization tensor @xmath518 now , the general expression for the generated magnetic field is @xmath519 \frac{g_{\vec{k}}(\vec{p } ) } { ( \vec{k } \cdot \vec{v } - \omega -i\tilde{\nu})}. \label{bg}\ ] ] the space - time evolution of the magnetic fluctuations can be determined by performing the inverse laplace and fourier transforms : @xmath520\int d^3 p \frac { g_{\vec{k}}(\vec{p } ) } { ( \vec{v}\cdot \vec{k } - \omega -i\tilde{\nu})}. \label{bf}\ ] ] in order to perform this integral , the explicit relations for the polarization tensors should be given .
they depend on the equilibrium distribution function @xmath521 , which we take to be the expression given in eq .
( [ dist0 ] ) then we have for transverse polarization @xmath522 \ln{\frac{k - \omega - i\tilde{\nu}}{k + \omega + i\tilde{\nu } } } - 2 \frac{\omega+i\tilde{\nu}}{k}\biggr\ } , \label{perpapp}\ ] ] and for the longitudinal polarization : @xmath523 consider now the case of very small momenta @xmath524 and @xmath525 , relevant for long - ranged magnetic fields .
then the computation of the integral ( [ bf ] ) in the large time limit and with the use of explicit form of the transverse polarization tensor in ( [ perpapp ] ) gives : , the equation @xmath526 defining the poles of the inverse laplace transform implies @xmath527 . ]
@xmath528 where @xmath194 is the plasma conductivity in the relaxation time approximation , @xmath529 and initial electric current is given by @xmath530 the obtained results assumed that the linearization of the vlasov equation is consistent with the physical assumptions of our problem .
this is indeed the case . in order to safely linearize the vlasov equation we have to make sure that the perturbed distribution function of the charge fluctuations is always smaller than the first order of the perturbative expansion ( given by the distribution of eq .
[ dist0 ] ) . in other words
we have to make sure that @xmath531 if the relativistic plasma frequency is defined , @xmath532 the condition expressed by eq .
( [ cond ] ) can be restated , for modes @xmath533 , as @xmath534 ( where we essentially took the square modulus of eq .
( [ cond ] ) ) .
this last inequality expresses the fact that the energy density associated with the gauge field fluctuations should always be smaller than the critical energy density stored in radiation .
the linear treatment of the vlasov equation is certainly accurate provided the typical modes of the the field are smaller than the plasma frequency and provided the energy density in electric and magnetic fields is smaller than @xmath535 , i.e. the energy density stored in the radiation background .
a relevant topic in the evolution of large scale magnetic fields concerns the concepts of _ direct _ and _ inverse _ cascades .
a direct cascade is a process , occurring in a plasma , where energy is transferred from large to small length scales .
an inverse cascade is a process where the energy transfer goes from small scales to large length scales .
one can also generalize the the concept of energy cascade to the cascade of any conserved quantity in the plasma , like , for instance , the helicity .
thus , in general terms , the transfer process of a conserved quantity is a cascade .
the concept of cascade ( either direct or inverse ) is related with the concept of turbulence , i.e. the class of phenomena taking place in fluids and plasmas at high reynolds numbers , as anticipated in section 4 .
non - magnetized fluids become turbulent at high _ kinetic _ reynolds numbers . similarly
, magnetized fluids should become turbulent for sufficiently large _ magnetic _
reynolds numbers .
the experimental evidence of this occurrence is still poor .
one of the reasons is that it is very difficult to reach , with terrestrial plasmas , the physical situation where the magnetic and the kinetic reynolds numbers are both large but , in such a way that their ratio is also large i.e. @xmath536 the physical regime expressed through eqs .
( [ ratio ] ) rather common in the early universe .
thus , mhd turbulence is probably one of the key aspects of magnetized plasma dynamics at very high temperatures and densities .
consider , for instance , the plasma at the electroweak epoch when the temperature was of the order of @xmath71 gev .
one can compute the reynolds numbers and the prandtl number from their definitions given in eqs .
( [ magnrey])([pr ] ) .
in particular , @xmath537 which can be obtained from eqs .
( [ magnrey])([pr ] ) using as fiducial parameters @xmath538 , @xmath539 , @xmath540 and @xmath541 for @xmath542 . at high reynolds and prandtl numbers
the evolution of the large - scale magnetic fields is not obvious since the inertial regime is large .
there are then two scales separated by a huge gap : the scale of the horizon ( @xmath543 cm at the electroweak epoch ) and the diffusivity scale ( about @xmath544 cm ) .
one would like to know , given an initial spectrum of the magnetic fields , what is the evolution of the spectrum , i.e. in what modes the energy has been transferred .
the problem is that it is difficult to simulate systems with huge hierarchies of scales .
suppose , for instance , that large scale magnetic fields are generated at some time @xmath545 and with a given spectral dependence .
the energy spectrum of magnetic fields may receive , for instance , the dominant contribution for large comoving momentum @xmath546 ( close to the cut - off provided by the magnetic diffusivity scale ) .
the `` initial '' spectrum at the time @xmath545 is sometimes called injection spectrum .
the problem will then be to know the how the spectrum is modified at a later time @xmath143 .
if inverse cascade occurs , the magnetic energy density present in the ultra - violet modes may be transferred to the infra - red modes . in spite of the fact that inverse cascades are a powerful dynamical principle ,
it is debatable under which conditions they may arise . the possible occurrence of inverse cascades in a magnetized universe has been put forward , in a series of papers , by brandenburg , enqvist and olesen @xcite ( see also @xcite for an excellent review covering also in a concise way different mechanisms for the generation of large - scale magnetic fields ) .
mhd simulations in @xmath547 dimensions @xcite seem to support the idea of an inverse cascade in a radiation dominated universe .
two - dimensional mhd simulations are very interesting but also peculiar . in @xmath547 dimensions ,
the magnetic helicity does not exist while the magnetic flux is well defined and conserved in the ideal mhd limit .
it is therefore impossible to implement @xmath547-dimensional mhd simulations where the magnetic flux and the magnetic helicity are constrained to be constant in the ideal limit .
this technical problem is related to a deeper physical difference between magnetized plasmas with and without helicity .
consider then , for the moment , the situation where the helicity is strictly zero .
it is indeed possible to argue that , in this case , @xmath547-dimensional mhd simulations can capture some of the relevant features of @xmath548-dimensional mhd evolution . in @xcite the scaling properties of mhd equations
have been investigated in the force - free regime ( i.e. in the approximation of negligible lorentz force ) . in the absence of helical components at the onset of the mhd evolution
a rather simple argument has been proposed by olesen @xcite in order to decide under which conditions inverse cascade may arise . consider the magnetic diffusivity and navier - stokes equations in the limit where @xmath549 .
then , the evolution equations are invariant under the following similarity transformations : @xmath550 for simplicity it is also useful to concentrate the attention on the inertial range of momenta where the dynamics is independent on the scale of dissipation . in order to understand correctly the argument under discussion the volume average of the magnetic energy density has to be introduced ( see @xcite for a lucid discussion on the various possible averages of large - scale magnetic fields and see @xcite for a complementary point of view ) .
a generic rotationally and parity invariant two - point function for the magnetic inhomogeneities can be written as @xmath551 defining now @xmath552 from eqs .
( [ g11])([g22 ] ) the following relation can be obtained @xmath553 where @xmath554 is related to @xmath555 and @xmath183 is the averaging volume .
the idea is now to study the scaling properties of @xmath554 under the similarity transformations given in eq .
( [ ol0 ] ) .
following @xcite , it is possible to show that @xmath556 with @xmath557 . in eq .
( [ ol1 ] ) @xmath554 is related to the magnetic energy density and @xmath558 is a function of the _ single argument _
if , initially , the spectrum of the magnetic energy density is @xmath560 , then at later time the evolution will be dictated by @xmath558 . since @xmath561 is approximately constant , because of flux conservation , then the comoving momentum will evolve , approximately , as @xmath562 while the physical momentum will scale like @xmath563 the conclusion of this argument is the following : * if @xmath564 the cascade will be direct ( or forward ) ; * if @xmath565 the comoving momentum moves toward the infra - red and the cascade will be inverse ; * if @xmath566 , then , @xmath567 and , from eqs .
( [ ol0])-([ol1 ] ) , the momentum and time dependence will be uncorrelated .
the olesen scaling argument has interesting implications : * if the initial spectrum of the magnetic energy density is concentrated at high momenta then , an inverse cascade is likely ; * from eqs .
( [ ol2 ] ) and ( [ ol3 ] ) the typical correlation scale of the magnetic field may , in principle , evolve faster than in the case where the inverse cascade is not present . in a frw universe
the correlation scale evolves as @xmath568 from a given initial time @xmath569 where magnetic energy density is injected with spectral dependence @xmath560 . if @xmath570 , eqs .
( [ ol2])([ol3 ] ) imply that @xmath571 where the second equality follows in a radiation dominated phase of expansion . in flat space ,
the analog of eqs .
( [ ol2])([ol3 ] ) would imply that @xmath572 .
finally , in the parametrization of olesen @xcite the case of a gaussian random field corresponds to @xmath573 , i.e. @xmath574 . in this case
the correlation scale will evolve as @xmath575 in the context of hydrodynamics the scaling law expressed through eq.([corr2 ] ) is also well known @xcite .
it is appropriate , at this point , to recall that hogan @xcite was the first one to discuss the evolution of the correlation scale produced by a gaussian random injection spectrum ( i.e. @xmath576 ) .
hogan was also the first one to think of the first order phase transitions as a possible source of large scale magnetic fields @xcite .
an argument similar to the one discussed in the case of mhd can be discussed in the case of ordinary hydrodynamics . in this case
, the analog of @xmath554 will be related to the kinetic energy density whose spectrum , in the case of fully developed turbulence goes as @xmath577 , i.e. the usual kolmogorov spectrum .
the evolution of the correlation scale derived in eq .
( [ corr2 ] ) can be obtained through another class of arguments usually employed in order to discuss inverse cascades in mhd : the mechanism of selective decay @xcite which can be triggered in the context of magnetic and hydrodynamical turbulence .
the basic idea is that modes with large wave - numbers decay faster than modes whose wavenumber is comparatively small .
son @xcite applied these considerations to the evolution of large - scale magnetic fields at high temperatures and his results ( when the magnetic field configuration has , initially , vanishing helicity ) are in agreement with the ones obtained in @xcite . in @xcite the topics discussed in @xcite
have been revisited and in @xcite the possible occurrence of inverse cascades has been criticized using renormalization group techniques .
the main assumption of the arguments presented so far is that the evolution of the plasma occurs in an inertial regime .
as discussed above in this section , at high temperatures the inertial regime is rather wide both from the magnetic end thermal points of view .
however , in the case where the magnetic and thermal diffusivity scales are _ finite _ the arguments relying on the inertial regime can not be , strictly speaking , applied .
in this situation numerical simulations should be used in order to understand if the diffusive regime either precedes or follows the inverse cascade .
if the magnetic and thermal diffusivity scales are finite , two conceptually different possibilities can be envisaged .
one possibility is that the inverse cascade terminates with a diffusive regime .
the other possibility is that the onset of diffusion prevents the occurrence of the inverse cascade .
this dilemma can be solved by numerical simulations along the lines of the ones already mentioned @xcite .
the numerical simulations presented in @xcite are in @xmath578 dimensions .
the results support the occurrence of inverse cascade in the presence of large kinetic viscosity .
the moment at which the cascade stops depends on the specific value of the viscosity coefficient . in @xcite @xmath579-dimensional mhd simulations have shown evidence of inverse cascade .
another possible limitation is that in the early universe the magnetic and kinetic reynolds numbers are both large and also their ratio , the prandtl number , is very large .
this feature seems to be difficult to include in simulations .
the considerations presented so far do not take into account the possibility that the initial magnetic field configuration has non vanishing helicity .
if magnetic helicity is present the typical features of mhd turbulence can qualitatively change @xcite and it was argued that the occurrence of inverse cascades may be even more likely @xcite ( see also @xcite ) .
the conservation of the magnetic flux is now supplemented , in the magnetic sector of the evolution , by the conservation of the helicity obtained in eqs .
( [ h1 ] ) and ( [ h2 ] ) .
the conservation of the helicity is extremely important in order to derive the mhd analog of kolmogorov spectrum .
in fact , the heuristic derivation of the kolmogorov spectrum is based on the conservation of the kinetic energy density .
if the conservation of the magnetic helicity is postulated , the same qualitative argument leading to the kolmogorov spectrum will lead in the case of fully developed mhd turbulence to the so - called iroshnikov - kraichnan spectrum @xcite whose specific form , is , in terms of the quantity introduced in eq .
( [ ol1 ] ) @xmath580 .
suppose , for simplicity , that the injection spectrum has a non - vanishing helicity and suppose that the helicity has the same sign everywhere . in this case @xmath581 .
since the helicity is conserved the magnetic field will scale according to eqs .
( [ ol0 ] ) but with @xmath582 corresponding to @xmath583 .
hence , the correlation scale will evolve , in this case , as @xmath584 i.e. faster than in the case of eq .
( [ corr2 ] ) .
it is difficult to draw general conclusions without specifying the nature and the features of the mechanism producing the injection configuration of the magnetic field .
the results critically summarized in this section will be important when discussing specific ideas on the origin of large - scale magnetic fields . in fig .
[ f3 ] the cases of different injection spectra are compared .
it is clear that an injection spectrum given by a gaussian random field leads to a rather inefficient inverse cascade .
since the magnetic helicity vanishes at the onset of the evolution , the rate of energy transfer from small to large scales is rather slow . on the contrary
if the injection spectrum is helical , the rate increases significantly . for comparison the situation where the injection spectrum corresponds to the iroshnikov - kraichnan case , i.e. the case of fully developed mhd turbulence ,
is also reported .
the dashed line of fig .
[ f3 ] illustrates the evolution of the correlation scale dictated by the the universe expansion . at small temperatures and small densities of the different fermionic charges the @xmath585
is broken down to the @xmath586 and the long range fields which can survive in the plasma are the ordinary magnetic fields .
for sufficiently high temperatures ( and for sufficiently high values of the various fermionic charges ) the @xmath585 is restored and non - screened vector modes correspond to hypermagnetic fields .
in fact , abelian electric fields decay within a typical time scale @xmath167 where @xmath194 is the conductivity .
the long - ranged non - abelian magnetic fields ( corresponding , for instance , to the color @xmath587 or to the weak @xmath588 ) can not exist because at high temperatures the non - abelian interactions induce a `` magnetic '' mass gap @xmath589 where @xmath590 is the gauge coupling constant .
also the non - abelian electric fields decay because of the finite value of the conductivity as it occurs for abelian electric fields .
therefore , the only long scale field that can exist in the plasma for enough time must be associated with some abelian u(1 ) group .
this statement , valid to all orders in perturbation theory , has been confirmed non - perturbatively for the electroweak theory by recent lattice studies in @xcite . under normal conditions ( i.e. small temperatures and small densities of the different fermionic charges )
the su(2)@xmath591u(1)@xmath592 symmetry is broken " down to u(1)@xmath593 , the massless field corresponding to u(1)@xmath593 is the ordinary photon and the only long - lived field in the plasma is the ordinary magnetic one . at sufficiently high temperatures , @xmath594 , the su(2)@xmath591u(1)@xmath592 symmetry is restored " , and non - screened vector modes @xmath595 correspond to the u(1)@xmath592 hypercharge group . hence ,
if primordial fields existed at @xmath596 , they did correspond to hypercharge rather than to u(1)@xmath593 . at the electroweak epoch
the typical size of the hubble radius is of the order of @xmath543 cm .
the typical diffusion scale is of the order of @xmath597 cm .
therefore , over roughly eight orders of magnitude hypermagnetic fields can be present in the plasma without being dissipated @xcite .
the evolution of hypermagnetic fields can be obtained from the anomalous magnetohydrodynamical ( amhd ) equations .
the amhd equations generalize the treatment of plasma effects involving hypermagnetic fields to the case of finite fermionic density@xcite .
there are essential differences between the interactions of magnetic fields and the ones of hypermagnetic fields with matter .
the ordinary electromagnetic field has a vector - like coupling to the fermions , while the coupling of the hypercharge fields is chiral .
thus , if hyperelectric ( @xmath598 ) and hypermagnetic ( @xmath599 ) fields are present simultaneously , they cause a variation of the fermionic number according to the anomaly equation , @xmath600 ( here @xmath601 the hypercharge gauge coupling constant ) .
now , the presence of _ non - homogeneous _
hypermagnetic fields in the ew plasma with ( hyper)conductivity @xmath602 always implies the existence of a related electric field , @xmath603 . since for a general stochastic magnetic background @xmath604 , the non - uniform hypermagnetic field may absorb or release fermions and produce , ultimately , baryon and lepton density perturbations because of the anomaly equation .
the behaviour of _ cold _ fermionic matter with non - zero anomalous abelian charges was considered in @xcite where it was pointed out that the anomalous fermionic matter is unstable against the creation of abelian gauge field with non - zero chern - simons number , which eats up fermions because of the anomaly .
the right electron number density may be converted to the hypercharge field because of a similar effect .
also the opposite effect is possible : hypercharge fields may be converted into fermions in a _ hot _ environment .
the electroweak plasma in _ complete _ thermal equilibrium at a temperature @xmath605 can be characterized by @xmath606 chemical potentials @xmath607 corresponding to the exactly conserved global charges @xmath608 ( @xmath609 is the lepton number of the @xmath610-th generation , @xmath611 is the baryon number , and @xmath612 is the number of fermionic generations ) .
one should also introduce a chemical potential @xmath613 corresponding to weak hypercharge , but its value is fixed from the requirement of the hypercharge neutrality of the plasma , @xmath614 .
it is interesting to study this plasma slightly out of thermal equilibrium , for instance in the situation where a non - uniform distribution of the hypermagnetic field is present . because of the anomaly , this field
is coupled to the fermionic number densities . in principle , different chemical potentials can be assigned to all the fermionic degrees of freedom of the electroweak theory ( @xmath615 if @xmath616 ) and the coupled system of boltzmann - type equations for these chemical potentials and the hypercharge fields may be written . since we are interested in the slow processes in the plasma , this is not necessary .
if the coupling , corresponding to some slow process , is switched off , then the electroweak theory acquires an extra conserved charge and a further chemical potential should be added to the system given in eq .
( [ 2.11 ] ) .
an interesting observation ( which turns out to be quite important in our context ) has been made in @xcite , where it was noticed that perturbative reactions with right - handed electron chirality flip are out of thermal equilibrium at temperatures higher than some temperature @xmath617 .
tev @xcite . ]
thus , the number of right electrons is perturbatively conserved at temperatures @xmath618 and the chemical potential @xmath619 can be introduced for it . on the other hand , this charge is not conserved because of the abelian anomaly , @xmath620 and it is therefore coupled to the hypermagnetic field .
here @xmath621 and @xmath622 are , respectively , the @xmath623 hypercharge field strengths and their duals , @xmath601 is the associated gauge coupling and @xmath624 is the hypercharge of the right electron .
now we are ready to derive the anomalous mhd equations in flat space @xcite . the effective lagrangian density describing the dynamics of the gauge fields at finite fermionic density is @xcite : @xmath625 ( @xmath590 is the determinant of the metric defined in ( [ metric ] ) ; @xmath626}$ ] ; @xmath627 is the covariant derivative with respect to the metric ( [ metric])[notice that in the metric ( [ metric ] ) @xmath628 } = \partial_{[\alpha}y_{\beta]}$ ] ] ; @xmath601 is the abelian coupling constant ) .
the first and the last terms in eq .
( [ 2.15 ] ) are nothing but the curved space generalization of the flat - space effective lagrangian for the hypercharge fields at finite fermion density @xcite , @xmath629 is the ohmic current .
the equations of motion for the hyperelectric and hypermagnetic fields are then @xmath630 with the same notations introduced in the case of the conventional mhd equations
. to eqs .
( [ 2.16 ] ) , the evolution equation of the right electron chemical potential , accounting for the anomalous and perturbative non - conservation of the right electron number density ( @xmath631 ) , must be added : @xmath632 where @xmath633 is the perturbative chirality - changing rate , @xmath634 , @xmath635 is the equilibrium value of the right electron number density , and the term proportional to @xmath636 is the right electron anomaly contribution .
finally , the relationship between the right electron number density and the chemical potential must be specified .
this relation depends upon the particle content of the theory . in the case of the minimal standard models ( msm )
the evolution equation of the chemical potential becomes @xcite @xmath637 at finite hyperconductivity ( in what we would call , in a mhd context , resistive approximation ) we have that from eq .
( [ 2.16 ] ) the induced hyperelectric field is not exactly orthogonal to the hypermagnetic one and , moreover , an extra `` fermionic '' current comes in the game thanks to the fact that we are working at finite chemical potential .
therefore in our context the resistive ohm law can be written as @xmath638 in the bracket appearing in eq .
( [ ohm ] ) we can identify two different contributions .
one is associated with the curl of the magnetic field .
we will call this the mhd contribution , since it appears in the same way in ordinary plasmas .
the other contribution contains the chemical potential and it is directly proportional to the magnetic field and to the chemical potential .
this is a typical finite density effect .
in fact the extra ohmic current simply describes the possibility that the energy sitting in real fermionic degrees of freedom can be transferred to the hypermagnetic field .
it may be of some interest to notice the analogy between the first term of eq .
( [ ohm ] ) and the typical form of the ohmic current discussed in eq .
( [ genohm1 ] ) appearing in the context of the dynamo mechanism . in the latter case
the presence of a current ( proportional to the vorticity through the @xmath16 dynamo term ) was indicating that large length scales magnetic fields could grow by eating up fluid vortices . by inserting @xmath639 obtained from the generalized ohm law ( [ ohm ] ) in the evolution equations ( [ 2.16 ] ) of the hypercharge fields , we obtain the generalized form of the magnetic diffusivity equation : @xmath640 in order to be consistent with
our resistive approach we neglected terms containing time derivatives of the electric field , which are sub - leading provided the conductivity is finite . in our considerations
we will also make a further simplification , namely we will assume that the ew plasma is ( globally ) parity - invariant and that , therefore , no global vorticity is present .
therefore , since the length scale of variation of the bulk velocity field is much shorter than the correlation distance of the hypermagnetic field , the infrared modes of the hypercharge will be practically unaffected by the velocity of the plasma , which will be neglected when the large - scale part of the hypercharge is concerned .
this corresponds to the usual mhd treatment of a mirror symmetric plasma ( see , e.g. eq . ( [ dynamored])([vort ] ) , when @xmath583 ) . eqs . ( [ hyperdiffusivity ] ) and ( [ mur ] )
form a set of amhd equations for the hypercharge magnetic field and right electron chemical potential in the expanding universe .
the abelian nature of the hypercharge field does not forbid that the hypermagnetic flux lines should have a trivial topological structure .
this situation is similar to what already encountered in the case of conventional electromagnetic fields with the important difference that the evolution equations of hypermagnetic fields are different from the ones of ordinary magnetic fields . after a swift summary of the properties of hypermagnetic knots ( based on @xcite ) ,
some interesting applications of these hypercharge profiles will be reviewed . in the gauge @xmath641 , @xmath642 ,
an example of topologically non - trivial configuration of the hypercharge field is the chern - simons wave @xcite @xmath643 this particular configuration is not homogeneous but it describes a hypermagnetic knot with _ homogeneous _ helicity and chern - simons number density @xmath644 where @xmath645 , @xmath646 ; @xmath601 is the @xmath647 coupling .
it is possible to construct hypermagnetic knot configurations with finite energy and helicity which are localized in space and within typical distance scale @xmath648 .
let us consider in fact the following configuration in spherical coordinates @xcite @xmath649 ^ 2 } , \nonumber\\ y_{\theta}({\cal r},\theta ) & = & \frac{2 b_0 } { \pi l_{s } } \frac { \sin{\theta}}{\bigl [ { \cal r}^2 + 1\bigr]^2 } , \nonumber\\ y_{\phi}({\cal
r},\theta ) & = & - \frac { 2 b_0 } { \pi l_{s } } \frac {
n { \cal r}\sin{\theta}}{\bigl[{\cal r}^2 + 1\bigr]^2 } , \label{conf2}\end{aligned}\ ] ] where @xmath650 is the rescaled radius and @xmath651 is some dimensionless amplitude and @xmath340 is just an integer number whose physical interpretation will become clear in a moment .
the hypermagnetic field can be easily computed from the previous expression and it is @xmath652 ^ 2 } , \nonumber\\ & & h_{\theta}({\cal r } , \theta ) = - \frac{4 b_{0}}{\pi~ l_{s}^2}\frac{{\cal r}^2 -1}{\bigl [ { \cal r}^2 + 1\bigr]^3}n \sin{\theta } , \nonumber\\ & & h_{\phi}({\cal r } , \theta ) = - \frac{8 b_0 } { \pi~ l_{s}^2}\frac { { \cal r } \sin{\theta}}{\bigl [ { \cal r}^2 + 1\bigr]^3}. \label{knot}\end{aligned}\ ] ] the poloidal and toroidal components of @xmath653 can be usefully expressed as @xmath654 and @xmath655 .
the chern - simons number is finite and it is given by @xmath656 ^ 4 } = \frac{g'^2 n b^2_0}{32 \pi^2}. \label{cs}\ ] ] the total helicity of the configuration can also be computed @xmath657 we can compute also the total energy of the field @xmath658 and we discover that it is proportional to @xmath659 .
this means that one way of increasing the total energy of the field is to increase the number of knots and twists in the flux lines .
this type of configurations can be also obtained by projecting a non - abelian su(2 ) ( vacuum ) gauge field on a fixed electromagnetic direction @xcite .
the resulting profile of the knot depends upon an arbitrary function of the radial distance .
these configurations have been also studied in @xcite . in particular , in @xcite
, the relaxation of hk has been investigated with a technique different from the one employed in @xcite but with similar results .
the problem of scattering of fermions in the background of hypermagnetic fields has been also studied in @xcite .
hypermagnetic knots with large correlation scale can be also generated dynamically provided an unknotted hypermagnetic background is already present .
let us assume that dynamical pseudoscalar particles are evolving in the background geometry given by eq .
( [ metric ] ) .
the pseudoscalars are _ not _ a source of the background ( i.e. they do not affect the time evolution of the scale factor ) but , nonetheless , they evolve according to their specific dynamics and can excite other degrees of freedom . the action describing the interaction of a dynamical pseudoscalar with hypercharge fields can be written as @xmath660 .
\label{action}\ ] ] this action is quite generic . in the case
@xmath661 eq .
( [ action ] ) is nothing but the curved space generalization of the model usually employed in direct searches of axionic particles @xcite .
the constant in front of the anomaly is a model - dependent factor .
for example , in the case of axionic particles , for large temperatures @xmath662 , the abelian gauge fields present in eq .
( [ action ] ) will be hypercharge fields and @xmath663 where @xmath664 and @xmath665 is a numerical factor of order @xmath278 which can be computed ( in a specific axion scenario ) by knowing the peccei - quinn charges of all the fermions present in the model @xcite . for small temperatures
@xmath666 we have that the abelian fields present in the action ( [ action ] ) will coincide with ordinary electromagnetic fields and @xmath667 where @xmath668 is the fine structure constant and @xmath669 is again a numerical factor .
we want now to study the amplification of gauge field fluctuations induced by the time evolution of @xmath670 .
then , the evolution equation for the hypermagnetic fluctuations @xmath675 can be obtained by linearizing eqs .
( [ system ] ) .
we will assume that any background gauge field is absent .
in the linearisation procedure we will also assume that the pseudoscalar field can be treated as completely homogeneous ( i.e.@xmath676 ) .
this seems to be natural if , prior to the radiation dominated epoch , an inflationary phase diluted the gradients of the pseudoscalar . in this approximation
, the result of the linearization can be simply written in terms of the vector potentials in the gauge @xmath677 and @xmath678 : @xmath679 by combining the evolution equations for the gauge fields we can find a decoupled evolution equation for @xmath675 , @xmath680 from this equation is already apparent that the pseudo - scalar vertex induces an interaction in the two physical polarizations of the hypermagnetic field . giving initial conditions which are such that @xmath681 with @xmath682 a profile with @xmath683 can be generated provided @xmath684 .
since the flux is conserved the ratio between the magnetic energy density , @xmath685 and the energy density sitting in radiation , @xmath686 is almost constant and therefore , in terms of this quantity ( which is only scale dependent but not time dependent ) , the most simplistic dynamo requirement can be rephrased ( see section 3 ) as @xmath687 to be compared with the value @xmath688 which would lead to the galactic magnetic field only thanks to the collapse and without the need of dynamo action ( this would be the case when the magnetic field is fully primordial ) .
notice that eq .
( [ dyn ] ) assumes that the magnetic flux is exactly frozen into the plasma element . as previously pointed out during the discussion of the dynamo mechanism , the requirement expressed by eq .
( [ dyn ] ) is unrealistic : it would correspond to thirty e - folds of amplification during galactic rotation and perfect flux freezing during the collapse of the protogalaxy . in a more realistic situation situation , taking into account the effectively achievable amplification of the dynamo action the requirement would rather be , prior to collapse , @xmath689 in spite of the richness of the theoretical models , the mechanisms for magnetic field generation can be divided , broadly speaking , into two categories : astrophysical and cosmological .
the cosmological mechanisms can be divided , in turn , into _ causal _ mechanisms ( where the magnetic seeds are produced at a given time inside the horizon ) and _ inflationary _ mechanisms where correlations in the magnetic field are produced outside the horizon .
the first attempts in this direction have been made by biermann @xcite and harrison @xcite .
the biermann battery mechanism is easy to understand from the form of the generalized ohm law discussed in eq .
( [ genohm1 ] ) .
mhd equations are linear and homogeneous in the magnetic field intensities .
the idea is then to look for a natural source term which seemed to be provided by the thermoelectric current already introduced in the context of the generalized ohm law .
consider , indeed , eq .
( [ genohm1 ] ) in the approximation where the hall term is neglected ( the argument can be also generalized to the case of non vanishing hall term ) .
then , the ohmic electric field will not be the one simply obtained in resistive mhd but it will get a contribution from the thermoelectric term : @xmath690 using the other mhd equations in the incompressible approximation the new form of the magnetic diffusivity equation can be derived : @xmath691 the magnetic diffusivity equation has now a source terms which does not vanish provided the gradients of the charge density of the electrons is not parallel to the pressure gradients . as a consequence of the presence of the source term
, the magnetic field will grow linearly in time until the thermoelectric source term is comparable with the dynamo term .
the estimate of the magnetic field intensity prior to the onset of galactic rotation will then be @xmath692 leading to magnetic fields @xmath693 g over typical scales of the order of @xmath34 kpc ( see also @xcite for a recent analysis in a similar framework ) . in a complementary perspective
, harrison@xcite discussed the possibility that the pressure gradients are strictly vanishing , i.e. @xmath694 .
this is a reasonable assumption , for instance , in a radiation - dominated stage of expansion where pressure gradients are expected to be negligible by the global homogeneity of the background . in this situation , an evolution equation for the vorticity
can be derived @xmath695 where @xmath696 and @xmath697 is the ion mass .
if we now postulate that some vorticity was present prior to decoupling , then eq . ( [ harrison2 ] ) can be solved and the magnetic field can be related to the initial vorticity as @xmath698 it is clear that if the estimate of the vorticity is made prior to equality ( as originally suggested by harrison@xcite ) of after decoupling as also suggested , a bit later@xcite can change even by two orders of magnitude .
prior to equality @xmath699 and , therefore , @xmath700 g . if a similar estimate is made after decoupling the typical value of the generated magnetic field is of the order of @xmath701 g. so , in this context , the problem of the origin of magnetic fields is circumvented by postulating an appropriate form of vorticity whose origin must be explained .
the idea was employed , later on , in the context of the physics of topological defects .
vachaspati and vilenkin @xcite have suggested that cosmic strings with small scale structure may be a source of the wanted vorticity .
the argument is , in short , that since matter flow in baryonic wakes is turbulent , velocity gradients will be induced in the flow by the small - scale wiggles which produced the required vorticity .
furthermore , dynamical friction between cosmic strings and matter may provide a further source of vorticity @xcite .
there have been also studies trying to generate large - scale magnetic fields in the context of superconducting cosmic strings ( see , for instance , @xcite and references therein ) .
recently the possible generation of large - scale magnetic fields prior to hydrogen recombination has been discussed in @xcite .
the vorticity required in order to produce the magnetic fields is generated , according to @xcite , by the photon diffusion at second order in the temperature fluctuations . in a similar perspective hogan @xcite got much less optimistic estimates which , according to @xcite , should be attributed to different approximation schemes employed in the analysis . different astrophysical objects , of different physical sizes , have comparable magnetic fields .
this coincidence is hard to explain in the context of causal mechanisms of generation .
the analogy with structure formation , already presented in the introduction , is here useful . in the late seventies , prior to the formulation of the inflationary paradigm , the initial conditions for the density contrast were taken as primordial input .
later on , in the context of inflationary models , the primordial spectrum of curvature and density fluctuations could be calculated . during inflation , in fact
, fields of various spin are present and they can be excited by the dynamical evolution of the geometry . in the context of inflationary models of particular relevance
are the scalar and tensor fluctuations of the geometry , corresponding respectively , to fluctuations of the scalar curvature and to gravitational waves .
gravitational waves and curvature perturbations obey evolution equations which are not invariant under weyl rescaling of the four - dimensional metric .
then the quantum mechanical fluctuations in the corresponding fields will simply be amplified by the evolution of the background geometry .
it is interesting to speculate that large scale magnetic fields could be produced thanks to a similar mechanism .
the major obstruction to this type of models is that gauge fields are not amplified thanks to the evolution of the background geometry the reason being that the evolution equations of gauge fields are invariant under weyl rescaling of the metric . when conformal invariance is broken , by some means , one is often led to estimate the amplitude of gauge field fluctuations arising as a result of the breaking .
clearly the detailed amount of amplified gauge fields will be specific of the particular model . in the following various ways of breaking conformal invariance
will be listed without getting through the details of the calculations . before doing
so it is anyway instructive to introduce some general considerations stressing the analogy between the production of magnetic inhomogeneities and the production of gravitational inhomogeneities whose late time evolution leads the anisotropies in the cmb .
a similar equation holds for the canonical normal mode for the scalar fluctuations of the geometry , i.e. @xmath711 v_{k } = 0 .
\label{vvar}\ ] ] the variable @xmath712 is defined as @xmath713 where @xmath714 is the fluctuation of the inflaton , @xmath715 is the curvature perturbation and @xmath670 is the scalar fluctuation of the geometry in the conformally newtonian gauge @xcite . if the inflaton has an exponential potential @xmath716 . in the case of gauge fields , each of the two polarization of the appropriately rescaled vector potential obeys , the following equation @xmath717 { \cal a}_{k } = 0 .
\label{scheq}\ ] ] in specific models , @xmath718 may be associated with the gauge coupling .
however @xmath719 can also be viewed as a generic pump field arising as a result of the breaking of conformal invariance .
the time dependence of the potential term is rather common to different models : it goes to zero for @xmath720 as @xmath721 .
the numerical coefficient appearing in the potential determines the strength and spectrum of the amplified gauge fields whose subsequent evolution has to be however computed at finite conductivity . in fig .
[ f4 ] the typical form of the potential term appearing in eq .
( [ scheq ] ) is illustrated .
different comoving frequencies go under the barrier at different times and the amount of amplification is roughly proportional to the time spent under the barrier .
clearly , given the generic form of the barrier , smaller frequencies are more amplified than the frequencies comparable with the height of the barrier at @xmath722 . in fig .
[ f4 ] the explicit numerical value of the height of the barrier corresponds to a ( present ) frequency of @xmath723 hz which can be realized if the pump field goes to a constant right after the end of a conventional inflationary phase followed by a radiation - dominated stage of expansion . in the case of gravitational waves , and , with some caveats , in the case of scalar metric fluctuations , the `` potential barriers '' appearing , respectively , in eqs .
( [ tens ] ) and ( [ vvar ] ) may be related with the inverse hubble radius .
hence , in the case of metric fluctuations , a mode which is under the barrier is also , with a swift terminology , outside the horizon ( see fig .
this is the reason why , following the conventional nomenclature , in fig .
[ f4 ] the moment when a given scale gets under the barrier has been denoted by @xmath724 ( i.e. horizon exit ) . according to the same convention , the moment when a given scale gets out the potential barrier
is labeled by @xmath725 ( i.e. horizon re - entry ) .
the mixing coefficients @xmath732 , determining the parametric amplification of a mode @xmath733 , computed by matching these various branches of the solution reported in eq .
( [ solution ] ) .
one finds : @xmath734 similar calculations can be performed in order to obtain the spectrum of scalar and tensor fluctuations of the geometry @xcite .
suppose now to make a simple estimate .
assume , for instance , that @xmath590 is evolving prior to @xmath722 according to the dynamics dictated by a given model .
suppose also that after @xmath722 the universe is suddenly dominated by radiation , and @xmath735 for @xmath736 .
in this situation all the modes reenter during radiation and the amplification will be roughly given , to leading order by @xmath737 if the function @xmath738 is identified with the evolving gauge coupling this result suggest that in order to have large amplification , @xmath738 has to grow from smaller to larger values .
this is what happens , for instance , in the case of pre - big bang models where @xmath739 and @xmath347 is the for - dimensional dilaton field @xcite . since we ought to estimate the amplification of an initial quantum mechanical fluctuations , a fully quantum mechanical treatment is certainly appropriate also in view of the discussion of the correlation properties of the obtained fluctuations .
this analysis has been performed in @xcite where the squeezing properties of the obtained photons have also been discussed .
the perturbed effective lagrangian density @xmath740 , ~~l(\eta ) = \int d^3 x { \cal l}(\vec{x},\eta ) , \label{action2}\ ] ] describes the evolution of the two ( @xmath741 ) transverse degrees of freedom defined by the coulomb gauge condition @xmath742 and @xmath743 ( the prime denotes differentiation with respect to conformal time ) .
the fields @xmath744 have kinetic terms with canonical normalization and the time evolution given in eq .
( [ scheq ] ) stems from the euler - lagrange equations derived from eq .
( [ action2 ] ) . by functionally deriving the the action
the canonically conjugated momenta can be obtained leading to the hamiltonian density and to the associated hamiltonian @xmath745 , \nonumber\\ & & h(\eta ) = \int d^3 x { \cal h}(\vec{x}).\end{aligned}\ ] ] the operators corresponding to the classical polarizations appearing in the hamiltonian density @xmath746 , \nonumber\\ & & \hat{\pi}_{\alpha } ( \vec{x } , \eta ) = \int { d^3 k } \frac{1}{(2 \pi)^{3/2 } } \hat{\pi}_{\alpha}(k,\eta ) e^{i \vec{k}\cdot{\vec{x } } } , \nonumber\\ & & \hat{\cal \pi}_{\alpha}(k,\eta ) = i \sqrt{\frac{k}{2}}\bigl [ \hat{a}_{k,\alpha}(\eta)- \hat{a}^{\dagger}_{-k,\alpha}(\eta)\bigr],\end{aligned}\ ] ] obey canonical commutation relations and the associated creation and annihilation operators satisfy @xmath747 = \delta_{\alpha\beta}\delta^{(3 ) } ( \vec{k } - \vec{p})$ ] .
the ( two - modes ) hamiltonian contains a free part and the effect of the variation of the coupling constant is encoded in the ( hermitian ) interaction term which is quadratic in the creation and annihilation operators whose evolution equations , read , in the heisenberg picture @xmath748 the general solution of the previous system of equations can be written in terms of a bogoliubov - valatin transformation , @xmath749 where @xmath750 and @xmath751 . notice that the bogoliubov coefficients are the quantum analog of the mixing coefficients discussed in the semiclassical approach to the problem .
unitarity requires that the two complex functions @xmath752 and @xmath753 are subjected to the condition @xmath754 which also implies that @xmath752 and @xmath753 can be parameterized in terms of one real amplitude and two real phases @xmath755 ( @xmath266 is sometimes called squeezing parameter and @xmath756 is the squeezing phase ; from now on we will drop the subscript labeling each polarization if not strictly necessary ) .
the total number of produced photons @xmath757 is expressed in terms of @xmath758 , i.e. the mean number of produced photon pairs in the mode @xmath546 .
inserting eqs .
( [ heis])([num ] ) into eqs .
( [ heiseq ] ) we can derive a closed system involving only the @xmath759 and the related phases : @xmath760 where @xmath761 .
in quantum optics @xcite the coherence properties of light fields have been a subject of intensive investigations for nearly half a century . in the present context the multiparticle states described
so fare are nothing but squeezed states of the electromagnetic field @xcite .
in fact , up to now the heisenberg description has been adopted . in the schrdinger picture the quantum mechanical states obtained as a result of the time evolution
are exactly squeezed vacuum states @xcite .
magnetic fields over galactic scales have typical frequency of the order @xmath762@xmath763 hz which clearly fall well outside the optical range .
thus , the analogy with quantum optics is only technical .
the same quantum optical analogy has been successfully exploited in particle @xcite and heavy - ions physics @xcite of pion correlations ( in order to measure the size of the strongly interacting region ) and in the phenomenological analysis of hadronic multiplicity distributions .
the interference between the amplitudes of the magnetic fields ( young interferometry @xcite , in a quantum optical language ) estimates the first order coherence of the magnetic background at different spatial locations making use of the two - point correlation function whose trace over the physical polarizations and for coincidental spatial locations is related to the magnetic energy density .
( [ i])([iii ] ) can be solved once the pump field is specified but general expressions can be also obtained @xcite . in conventional inflationary models
it is very difficult to produce large scale magnetic fields with phenomenologically relevant strength .
this potential difficulty has been scrutinized in various investigations @xcite .
turner and widrow @xcite listed a series of field theory models in de sitter space with the purpose of finding a natural way of breaking conformal invariance .
the first suggestion was that conformal invariance may be broken , at an effective level , through the coupling of photons to the geometry @xcite .
typically , the breaking of conformal invariance occurs through products of gauge - field strengths and curvature tensors , i.e. @xmath764 where @xmath31 is the appropriate mass scale ; @xmath765 and @xmath766 are the riemann and ricci tensors and @xmath767 is the ricci scalar . if the evolution of gauge fields is studied during a phase of de sitter ( or quasi - de sittter ) expansion , then the amplification of the vacuum fluctuations induced by the couplings listed in eq .
( [ curv ] ) is minute .
the price in order to get large amplification should be , according to @xcite , an explicit breaking of gauge - invariance by direct coupling of the vector potential to the ricci tensor or to the ricci scalar , i.e. @xmath768 in @xcite two other different models were proposed ( but not scrutinized in detail ) namely scalar electrodynamics and the axionic coupling to the abelian field strength .
dolgov @xcite considered the possible breaking of conformal invariance due to the trace anomaly .
the idea is that the conformal invariance of gauge fields is broken by the triangle diagram where two photons in the external lines couple to the graviton through a loop of fermions .
the local contribution to the effective action leads to the vertex @xmath769 where @xmath770 is a numerical coefficient depending upon the number of scalars and fermions present in the theory .
the evolution equation for the gauge fields , can be written , in fourier space , as @xmath771 and it can be shown that only if @xmath772 the gauge fields are amplified . furthermore , only is @xmath773 substantial amplification of gauge fields is possible . in a series of papers @xcite the possible effect of the axionic coupling to the amplification of gauge fields has been investigated .
the idea is here that conformal invariance is broken through the explicit coupling of a pseudo - scalar field to the gauge field ( see section 5 ) , i.e. @xmath774 where @xmath775 is the dual field strength and where @xmath669 is a numerical factor of order one .
consider now the case of a standard pseudoscalar potential , for instance @xmath776 , evolving in a de sitter ( or quasi - de sitter space - time ) .
it can be shown , rather generically , that the vertex given in eq .
( [ pseudo ] ) leads to negligible amplification at large length - scales .
the coupled system of evolution equations to be solved in order to get the amplified field is similar to eqs .
( [ system ] ) already introduced in the duscussion of hypermagnetic fields @xmath777 from eq .
( [ bpsi ] ) , there is a maximally amplified physical frequency @xmath778 where the second equality follows from @xmath779 ( i.e. @xmath780 ) . the amplification for @xmath781 is of the order of @xmath782}$ ] where @xmath371 is the hubble parameter during the de sitter phase of expansion . from the above expressions
one can argue that the modes which are substantially amplifed are the ones for which @xmath783 .
the modes interesting for the large - scale magnetic fields are the ones which are in the opposite range , i.e. @xmath784 .
clearly , by lowering the curvature scale of the problem the produced seeds may be larger and the conclusions much less pessimistic @xcite .
another interesting idea pointed out by ratra @xcite is that the electromagnetic field may be directly coupled to the inflaton field . in this case
the coupling is specified through a parameter @xmath16 , i.e. @xmath785 where @xmath347 is the inflaton field in planck units . in order to get sizable large - scale magnetic fields
the effective gauge coupling must be larger than one during inflation ( recall that @xmath347 is large , in planck units , at the onset of inflation ) .
in @xcite it has been suggested that the evolution of the abelian gauge coupling during inflation induce the growth of the two - point function of magnetic inhomogeneities .
this model is different from the one previously discussed @xcite . here
the dynamics of the gauge coupling is not related to the dynamics of the inflaton which is not coupled to the abelian field strength . in particular , @xmath786 can be as large as @xmath787 . in @xcite
the mhd equations have been generalized to the case of evolving gauge coupling . recently
a scenario similar to @xcite has been discussed in @xcite . in the perspective of generating large scale magnetic fields
gasperini @xcite suggested to consider the possible mixing between the photon and the graviphoton field appearing in supergravity theories ( see also , in a related context @xcite ) .
the graviphoton is the massive vector component of the gravitational supermultiplet and its interaction with the photon is specified by an interaction term of the type @xmath788 where @xmath789 is the filed strength of the massive vector .
large - scale magnetic fields with @xmath790 can be obtained if @xmath791 and for a mass of the vector @xmath792 .
bertolami and mota @xcite argue that if lorentz invariance is spontaneously broken , then photons acquire naturally a coupling to the geometry which is not gauge - invariant and which is similar to the coupling considered in @xcite .
finally davis and dimopoulos @xcite considered the possibility of phase transitions taking place during inflation .
they found that sizable large - scale magnetic fields can be generated provided the phase transition occurs in the last 5 e - foldings of the inflationary stage of expansion . while the coupling of electromagnetic field to the metric and to the charged fields is conformally invariant , the coupling of the charged scalar field to gravity is not .
thus , vacuum fluctuations of the charged scalar field can be amplified during inflation at super - horizon scales , leading potentially to non - trivial correlations of the electric currents and charges over cosmological distances .
the fluctuations of electric currents , in turn , may induce magnetic fields through maxwell equations at the corresponding scales .
the role of the charged scalar field may be played by the higgs boson which couples to the hypercharge field above the electroweak phase transition ; the generated hypercharged field is converted into ordinary magnetic field at the temperatures of the order of electroweak scale .
no detailed computations were carried out in @xcite in order to substantiate this idea .
the suggestion of @xcite was further developed quite recently in @xcite for the standard electroweak theory with an optimistic conclusion that large scale magnetic fields can be indeed generated . in @xcite
a supersymmetric model was studied . in @xcite ,
previous treatments have been further scrutinized by computing , with higher accuracy , the amplification of the charged scalar field and the damping induced by the conductivity .
it turns out that the resulting magnetic fields are insufficient in order to provide reasonable seeds for the dynamo amplification .
introducing appropriately rescaled fields the action of the abelian - higgs model in a conformally flat frw space - time can be written as @xmath793 .
\label{actionab}\ ] ] now , since the evolution equation of the charged scalar is not conformally invariant , current density and charge density fluctuations will be induced .
then , by employing a vlasov - landau description similar to the one introduced in section 5 , the resulting magnetic field will be of the order of @xmath794 which is , by far , too small .
later it has been proposed that much larger magnetic fields may be obtained in the context of the abelian - higgs model @xcite ( see however @xcite for a detailed criticism of this proposal ) .
if internal dimensions are dynamical , then weyl invariance may be naturally broken @xcite .
consider a pure electromagnetic fluctuation decoupled from the sources , representing an electromagnetic wave propagating in the @xmath339-dimensional external space such that @xmath795 , @xmath796 . in the metric given in eq .
( [ metric ] ) the evolution equation of the gauge field fluctuations can be written as @xmath797 where @xmath798}$ ] is the gauge field strength and @xmath799 is the determinant of the @xmath800 dimensional metric .
notice that if @xmath801 the space - time is isotropic and , therefore , the maxwell s equations can be reduced ( by trivial rescaling ) to the flat space equations .
if @xmath802 we have that the evolution equation of the electromagnetic fluctuations propagating in the external @xmath339-dimensional manifold will receive a contribution from the internal dimensions which can not be rescaled away @xcite .- dimensional metric . in string theories , quite generically , the one - form fields are also coupled to the dilaton field .
this case has been already analyzed in the context of string inspired cosmological scenarios and will be discussed later .
] in the radiation gauge ( @xmath803 and @xmath804 ) the evolution the vector potentials can be written as @xmath805 the vector potentials @xmath806 are already rescaled with respect to the ( conformally flat ) @xmath807 dimensional metric . in terms of the canonical normal modes of oscillations
@xmath808 the previous equation can be written in a simpler form , namely @xmath809 in order to estimate the amplification of the gauge fields induced by the evolution of the internal geometry we shall consider the background metric of eq .
( [ metric2 ] ) in the case of maximally symmetric subspaces @xmath810 , @xmath811 .
suppose now that the background geometry evolves along three different epochs .
during the first phase ( taking place for @xmath812 -\infty , -\eta_1]$ ] ) the evolution is truly multidimensional . at @xmath813
the multidimensional dynamics is continuously matched to a radiation dominated phase turning , after decoupling , into a matter dominated regime of expansion . during
the radiation and matter dominated stages the internal dimensions are fixed to their ( present ) constant size in order not to conflict with possible bounds arising both at the bbn time and during the matter - dominated epoch .
the evolution of the external dimensions does not affect the amplification of the gauge fields as it can be argued from eq .
( [ eq1 ] ) : in the limit @xmath814 ( i.e. conformally invariant background ) eq .
( [ eq1 ] ) reduces to the flat space equation .
a background with the features described above has been introduced in eq .
( [ back ] ) . defining , respectively , @xmath815 and @xmath816 as the size of the internal dimensions at the bbn time and at the present epoch
, the maximal variation allowed to the internal scale factor from the bbn time can be expressed as @xmath817 where @xmath818 @xcite .
the bounds on the variation of the internal dimensions during the matter dominated epoch are even stronger . denoting with an over - dot the derivation with respect to the cosmic time coordinate
, we have that @xmath819 where @xmath202 is the present value of the hubble parameter @xcite . the fact that the time evolution of internal dimensions is so tightly constrained for temperatures lower of @xmath278 mev does not forbid that they could have been dynamical prior to that epoch @xcite . in the parameterization of eq .
( [ back ] ) the internal dimensions grow ( in conformal time ) for @xmath343 and they shrink must be effectively constant for practical purposes . ] for @xmath344 . by inserting this background into eq .
( [ eq1 ] ) we obtain that for @xmath820 @xmath821 whereas @xmath822 for @xmath823 .
since @xmath824 goes to zero for @xmath825 we can define , in both limits , a fourier expansion of @xmath826 in terms of two distinct orthonormal sets of modes .
the amplification of the quantum mechanical fluctuations of the gauge fields can then be computed using the standard techniques ( see , for instance , @xcite ) . in @xcite
it has been shown that in simple models of dimensional decoupling magnetic fields can be generated with strength compatible with the bound of eq .
( [ dyn2 ] ) .
in particular there is the interesting possibility that large - scale magnetic fields are produced in the case when internal dimensions also expand while the external ones also expand but at a different rate .
it should be mentioned that the interplay between gauge fields and large ( or infinite ) extra - dimensions is still under investigation .
there is , at the moment , no definite model leading to the generation of large - scale magnetic field in a framework of infinite internal dimensions .
the main reason is that it is difficult to build reasonable cosmological models with large extra - dimensions .
more specifically , one would like to have a model where gauge fields arise naturally in the bulk . in six- dimensions
interesting models can be constructed @xcite where localized gauge zero modes may be naturally present ( in analogy with what happens in the case of six - dimensional warped models @xcite ) .
large scale magnetic fields may also be produced in the context of string cosmological models @xcite ( see also @xcite for the analysis of the amplification of magnetic fields during a phase of coherent dilaton oscillations ) . in many respects
the case of string cosmological models is related to the one where the gauge coupling is dynamical .
however , in the string cosmological context , the dynamics of the gauge coupling and of the geometry are connected by the solutions of the low - energy @xmath228-functions .
the basic evolution of the background and of the gauge coupling has been discussed in section 5 . in order to achieve a large amplification of the quantum fluctuations of the gauge fields
, the gauge coupling should be sufficiently small when the typical scale of the gravitational collapse hits the `` potential barrier '' . in particular
, @xmath827 is the dynamo requirement has to be satisfied .
it is difficult to produce large scale magnetic fields with reasonable amplitudes if the pre - big bang phase matches immediately to the post - big bang evolution @xcite .
however , phenomenologically consistent pre - big bang models lead to a sufficiently long stringy phase when the dilaton and the curvature scale are roughly constant . if this phase is included the conditions expressed by eqs .
( [ dyn ] ) and ( [ dyn2 ] ) are easily satisfied . furthermore , there are regions in the parameter space of the model where @xmath828 which implies that the galactic magnetic field may be fully primordial @xcite , i.e. no dynamo action will be required .
the experimental evidence concerning large - scale magnetic fields suggests that magnetic fields should have similar strength over different length - scales . in this
sense inflationary mechanisms seem to provide a rather natural explanation for the largeness of the correlation scale . at the same time
, the typical amplitude of the obtained seeds is , in various models , rather minute .
primordial magnetic fields can also be generated through physical mechanisms operating inside the hubble radius at a given physical time .
particularly interesting moments in the life of the universe are the epoch of the electroweak phase transition ( ewpt ) or the qcd phase transition where magnetic fields may be generated according to different physical ideas . in the following the different proposals emerged so far will be reviewed . at the time of the ewpt
the typical size of the hubble radius is of the order of @xmath543 cm and the temperature is roughly @xmath71 gev . before getting into the details of the possible electroweak origin of large - scale magnetic fields it is useful to present a kinematical argument based on the evolution of the correlation scale of the magnetic fields @xcite .
suppose that , thanks to some mechanism , sufficiently large magnetic fields compatible with the critical density bound are generated inside the hubble radius at the electroweak epoch .
assuming that the typical coherence length of the generated magnetic fields is maximal , the present correlation scale will certainly be much larger but , unfortunately , it does not seem to be as large as the mpc scale at the epoch of the gravitational collapse .
as already mentioned , the growth of the correlation scale may be enhanced , by various processes such as inverse cascade and helical inverse cascade .
for instance , if the injection spectrum generated at the electroweak epoch is gaussian and random a simple estimate shows that the present correlation scale is of the order of @xmath71 au which is already larger than what one would get only from the trivial expansion of length - scales ( i.e. @xmath278 au ) @xcite . if , in a complementary perspective , the injection spectrum is strongly helical , then the typical correlation scale can even be of the order of @xmath71 pc but still too small than the typical scale of the gravitational collapse
. large - scale magnetic fields can be generated at the electroweak epoch in various ways .
consider first the case when the phase transition is strongly first order .
hogan @xcite originally suggested the idea that magnetic fields can be generated during first - order phase transitions .
since during the phase transition there are gradients in the radiation temperature , similar thermoelectric source terms of mhd equations ( which were discussed in the context of the biermann mechanism ) may arise .
the magnetic fields , initially concentrated on the surface of the bubbles , are expelled when bubbles collide thanks to the finite value of the conductivity . the idea that charge separation can be generated during first - order phase transitions has been exploited in @xcite .
the suggestion is again that there are baryon number gradients at the phase boundaries leading to thermoelectric terms . in the process of bubble nucleation and collisions turbulence
is then produced . in spite of the fact that the produced fields are sizable
, the correlation scale , as previously pointed out , is constrained to be smaller than @xmath71 pc . in a first - order phase transition
the phases of the complex order parameter of the nucleated bubbles are not correlated .
when the bubbles undergo collisions a phase gradient arises leading to a source terms for the evolution equation of the gauge fields .
kibble and vilenkin @xcite proposed a gauge - invariant difference between the phases of the higgs field in the two bubbles .
this idea has been investigated in the context of the abelian - higgs model @xcite .
the collision of two spherical bubbles in the abelian - higgs model leads to a magnetic field which is localized in the region at the intersection of the two bubbles .
the estimate of the strength of the field depends crucially upon the velocity of the bubble wall .
the extension of this idea to the case of the standard model @xmath829 has been discussed in @xcite .
a relevant aspect to be mentioned is that the photon field in the broken phase of the electroweak theory should be properly defined . in @xcite
it has been shown that the definition employed in @xcite is equivalent to the one previously discussed in @xcite .
it has been argued by vachaspati @xcite that magnetic fields can be generated at the electroweak time even if the phase transition is of second order .
the observation is that , provided the higgs field fluctuates , electromagnetic fields may be produced since the gradients of the higgs field appear in the definition of the photon field in terms of the hypercharge and @xmath588 fields .
two of the arguments proposed in @xcite have been scrutinized in subsequent discussions .
the first argument is related to the averaging which should be performed in order to get to the magnetic field relevant for the mhd seeds .
enqvist and olesen @xcite noticed that if line averaging is relevant the obtained magnetic field is rather strong .
however @xcite ( see also @xcite ) volume averaging is the one relevant for mhd seeds .
the second point is related to the fact that the discussion of @xcite was performed in terms of gauge - dependent quantities .
the problem is then to give a gauge - invariant definition of the photon field in terms of the standard model fields .
as already mentioned this problem has been addressed in @xcite and the proposed gauge - invariant is equivalent @xcite to the one proposed in @xcite . in
@xcite a mechanism for the generation of magnetic fields at the electroweak epoch has been proposed in connection with the amhd equations .
the idea is to study the conversion of the right - electron chemical potential into hypercharge fields . in this context
the baryon asymmetry is produced at some epoch prior to the electroweak phase .
the obtained magnetic fields are rather strong ( i.e. @xmath830 at the ew epoch ) but over a small scale , i.e. @xmath831 dangerously close to the diffusivity scale .
the final point to be mentioned is that , probably , the electroweak phase transition is neither first order nor second order but it is of even higher order at least in the context of the minimal standard model @xcite . this conclusion has been reached using non - perturbative techniques and the relevant point , in the present context , is that for higgs boson masses larger than @xmath832 the phase transition seems to disappear and it is possible to pass from the symmetric to the broken phase without hitting any first or second order phase transition .
there have been also ideas concerniing the a possible generation of magnetic fields at the time of the qcd phase transition occurring roughly at @xmath833 , i.e. at the moment when free quarks combine to form colorless hadrons .
the mechanism here is always related to the idea of biermann with thermoelectric currents developed at the qcd time.since the strange quark is heavier than the up and down quarks there may be the possibility that the quarks develop a net positive charge which is compensated by the electric charge in the leptonic sector . again , invoking the dynamics of a first - order phase transition , it is argued that the shocks affect leptons and quarks in a different way so that electric currents are developed as the bubble wall moves in the quark - gluon plasma . in @xcite
the magnetic field has been estimated to be @xmath834 at the time of the qcd phase transition and with typical scale of the order of the meter at the same epoch . in @xcite
it has been pointed out that , probably , the magnetic fields generated at the time of qcd phase transition may be much stronger than the ones estimated in @xcite .
the authors of @xcite argue that strong magnetic fields may be generated when the broken and symmetric phase of the theory coexist .
the magnetic fields generated at the boundaries between quark and hadron phases can be , according to the authors , as large as @xmath835 g over scales of the order of the meter at the time of the qcd phase transition .
recently , in a series of papers , boyanovsky , de vega and simionato @xcite studied the generation of large scale magnetic fields during a phase transition taking place in the radiation dominated epoch .
the setting is a theory of @xmath836 charged scalar fields coupled to an abelian gauge field that undergoes a phase transition at a critical temperature much larger than the electroweak scale . using non - equilibrium field theory techniques
the authors argue that during the scaling regime ( when the back - reaction effects are dominant ) large scale magnetogenesis is possible .
the claim is that the minimal dynamo requirement of eq .
( [ dyn ] ) is achievable at the electroweak scale .
furthermore , much larger magnetic fields can be expected if the scaling regime can be extended below the qcd scale . from the above discussion
it is apparent that the mechanisms producing magnetic fields at some moment in the life of the universe ( and inside the hubble radius ) have to justify the large correlation scale of the magnetic seeds required by subsequent mhd considerations .
the possibility of turbulent behaviour in the early universe , even though physically justified , seems to be difficult to quantify even if rather reasonable quantitative estimates appeared so far . on the other hand
, inflationary scenarios can lead to magnetic seeds with large correlation scale . a useful point to be stressed
is that there may exist `` mixed '' mechanisms . consider the following example which has been proposed in @xcite and @xcite .
suppose that during the inflationary phase a stochastic background of gauge fields is produced thanks to the breaking of conformal invariance .
the spectral amplitude of the generated field will be constrained by the critical density bound and by all the other dynamical considerations related to the finite values of the diffusion scales .
the generated magnetic field spectrum will be distributed over a large interval of frequencies . in particular
, the energy spectrum of the magnetic field will be non negligible at some intermediate scale , like , for instance , the electroweak scale , i.e. the physical frequency @xmath837 .
in other words , if magnetic fields are generated thanks to some inflationary mechanism , the amplified magnetic inhomogeneities will be reentering at different epochs during the radiation dominated phase and the scales which left the horizon later during inflation will reenter earlier . given the background of abelian gauge fields
, the dynamics of the electroweak epoch may add interesting physical features .
for instance the magnetic field of inflationary origin will be , as discussed , topologicaly trivial : its magnetic gyrotropy will vanish .
however , during the ew epoch , thanks to a pseudoscalar vertex due either to a pseudo goldstone boson or to the chemical pootential , the topologically trivial hypermagnetic background may be turned into a topologically non - trivial background and large magnetic helicity may be generated .
in the context of various mechanisms , the magnetic fields generated at large at the scale of the protogalactic collapse may be rather small . in this section
the possible effects of magnetic present over much smaller length - scales will be analyzed .
two important scales are the hubble radius at the electroweak epoch and at the big - bang nucleosynthesis epoch .
hypermagnetic fields present for temperatures @xmath838 have a twofold effect : * they may affect the phase diagram of the electroweak theory ; * they may offer a mechanism to seed the baryon asymmetry of the universe ( bau ) .
the physical picture of the possible effects of magnetic fields on the electroweak phase diagram is exactly the same as the macroscopic description of superconductors in the presence of an external magnetic field .
the normal - superconducting phase transition , being of second order in the absence of magnetic fields , becomes of first order if a magnetic field is present .
the reason for this is the meissner effect , i.e. the fact that the magnetic field can not propagate inside a superconducting cavity , and , therefore , creates an extra pressure acting on the normal - superconducting boundary .
our consideration below explores this simple picture .
consider the plain domain wall that separates broken and symmetric phase at some temperature @xmath605 , in the presence of a uniform hypercharge magnetic field @xmath839 .
far from the domain wall , in the symmetric phase , the non - abelian su(2 ) field strength ( @xmath840 ) is equal to zero , because of a non - perturbative mass gap generation . inside the broken phase , the massive @xmath841 combination of @xmath839 and @xmath842
, @xmath843 must be equal to zero , while the massless combination , corresponding to photon @xmath844 , survives .
the matching of the fields on the boundary gives @xmath845 .
thus , an extra pressure @xmath846 acts on the domain wall from the symmetric side . at the critical temperature it must be compensated by the vacuum pressure of the scalar field .
if we neglect loop corrections associated with the presence of magnetic fields , then the condition that determines the critical temperature is : @xmath847 where @xmath848 is the effective potential in the absence of magnetic field , @xmath849 is the location of the minimum of the potential at temperature @xmath605 .
the above consideration was dealing with the uniform magnetic fields .
clearly , it remains valid when the typical distance scale of inhomogeneities of the magnetic field are larger than the typical bubble size .
this is the case for bubbles smaller than the magnetic diffusion scale , and , in particular , at the onset of the bubble nucleation .
thus , the estimate of the critical temperature coming from ( [ pressure ] ) is applicable . for bubbles larger than the diffusivity scale
, the presence of a stochastic magnetic field will considerably modify their evolution .
in particular , the spherical form of the bubbles is very likely to be spoiled .
these considerations were presented in @xcite . later @xcite perturbative estimates were performed in order to corroborate the proposed picture . in @xcite
a full non - perturbative analysis of the phhase diagram of the electroweak theory in the presence of an hyermagnetic background has been performed . as previously discussed @xcite for values of the higggs boson mass larger than the w boson mass the electroweak phase diagram seems to exhibit a cross - over regime .
the inclusion of a constant hypermagnetic background with typical strength @xmath850 does modify the electroweak phase diagram but does not seem to make the phase transition strongly first order for @xmath851 as expected from perturbative considerations .
furthermore for @xmath852 ( but still compatible with the critical density bound ) a new ( inhomogeneous ) phase has been observed .
the analysis performed in @xcite included a net hypermagnetic flux .
it would be interesting to repeat the same calculation in the presence of a non - vanishing hypermagnetic helicity ( or gyrotropy ) .
depending upon the topology of the flux lines , hypermagnetic fields can have two distinct effects : * topologically non - trivial configurations of the hypermagnetic flux lines lead to the formation of hypermagnetic knots whose decay might seed the baryon asymmetry of the universe ( bau ) ; * even if the electrowak plasma has vanishing hypermagnetic gyrotropy , i.e. @xmath853 , matter antimatter fluctuations may be generated .
consider first the situation where the electroweak plasma contains , for @xmath854 a network of hypermagnetic knots of the type of the ones described in eqs .
( [ conf1])([knot ] ) . in section 5
these configurations have been named hypermagnetic knots ( hk ) and they are chern - simons condensates carrying a non - vanishing ( averaged ) hypermagnetic helicity . as discussed , hk can be dynamically generated from a background of hypermagnetic fields with trivial topology provided a ( time - dependent ) pseudo - scalar is present in the plasma ( see , for instance , eq .
( [ action ] ) ) . in order to seed the bau a network of hk should be present at high temperatures .
in fact for temperatures larger than @xmath855 the fermionic number is stored both in hk and in real fermions . for @xmath856 , the hk should release real fermions since the ordinary magnetic fields ( present _ after _ ew symmetry breaking ) do not carry fermionic number . if the ewpt is strongly first order the decay of the hk can offer some seeds for the bau generation .
this last condition can be met in the minimal supersymmetric standard model ( mssm ) .
the integration of the @xmath647 anomaly equation gives the cs number density carried by the hk which is in turn related to the density of baryonic number @xmath857 for the case of @xmath612 fermionic generations .
in fact , using eqs .
( [ nr])([ohm ] ) and after some algebra @xcite it can be shown that @xmath858 ( @xmath601 is the @xmath647 coupling and @xmath859 is the entropy density ; @xmath316 , at @xmath855 , is @xmath860 in the msm ; @xmath861 ) . in eq .
( [ bau ] ) @xmath633 is the perturbative rate of the right electron chirality flip processes ( i.e. scattering of right electrons with the higgs and gauge bosons and with the top quarks because of their large yukawa coupling ) which are the slowest reactions in the plasma and @xmath862 is the rate of right electron dilution induced by the presence of a hypermagnetic field . in the msm we have
that @xmath863 whereas in the mssm @xmath633 can naturally be larger than @xmath864 . unfortunately , in the msm a hypermagnetic field can modify the phase diagram of the phase transition but can not make the phase transition strongly first order for large masses of the higgs boson .
therefore , we will concentrate on the case @xmath865 and we will show that in the opposite limit the bau will be anyway small even if some ( presently unknown ) mechanism would make the ewpt strongly first order in the msm .
it is interesting to notice that , in this scenario , the value of the bau is determined by various particle physics parameters but also by the ratio of the hypermagnetic energy density over the energy density sitting in radiation during the electroweak epoch .
consider now the complementary situation where the electroweak plasma , for @xmath866 , is filled with topologically trivial hypermagnetic fields . in this case , fluctuations in the baryon to entropy ratio will be induced since , in any case @xmath867 .
these fluctuations are of _ isocurvature _ type and can be related to the spectrum of hypermagnetic fields at the ewpt . defining as @xmath868 the fluctuations in the ratio of baryon number density @xmath857 to the entropy density @xmath869 at @xmath870 @xcite , the value of @xmath871 can be related to the hypermagnetic spectrum which is determined in terms of its amplitude @xmath872 and its slope @xmath770 .
a physically realistic situation corresponds to the case in which the green s functions of the magnetic hypercharge fields decay at large distance ( i. e. @xmath873 ) and this would imply either `` blue '' ( @xmath874 ) or `` violet '' ( @xmath875 ) energy spectra .
the fluctuations of the baryon to entropy ratio generated at the electroweak epoch may survive until the bbn epoch @xcite . the possibility of survival of these fluctuations is related to their typical scale which must exceed the neutron diffusion scale appropriately blue - shifted at the electroweak epoch .
the implications of these fluctuations will be discussed in a moment .
large scale magnetic fields possibly present at the bbn epoch can have an impact on the light nuclei formation . by reversing the argument ,
the success of bbn can be used in order to bound the magnetic energy density possibly present at the time of formation of light nuclei .
magnetic fields , at the bbn epoch may have four distinct effects : * they can enhance the rate of the universe expansion at the corresponding epoch in a way proportional to @xmath876 ; * they can affect the reaction rate in a way proportional to @xmath877 ; * they can affect the phase space of the electrons ; * they can induce , prior to the formation of the light nuclei , matter antimatter fluctuations . while the first three effects are direct , the fourth effect is mainly caused by hypermagnetic fields present even before bbn .
the idea that magnetic fields may increase the universe expansion and , consequently , affect directly the abundance of @xmath878 ( which is directly sensitive to the expansion rate ) has been pointed out long ago by greenstein @xcite and by matese and oconnel @xcite . from a qualitative point of view the effects over the expansion should be leading in comparison with the effects over the rate of interactions ( which are suppressed by @xmath879 ) .
thanks to detailed numerical simulations performed indipendently by different groups ( i.e. cheng et al .
@xcite ; kernan , starkman and vachaspati @xcite ; grasso and rubinstein @xcite ) , the common opinion is that the universe expansion is the leading effect even if this conclusion has been reached only recently and not without some disagreements on the details of the calculations ( see , for instance , @xcite ) . in order to prevent the universe from expanding too fast at the bbn epoch @xmath880 where @xmath881 is the energy density contributed by the standard three light neutrinos with masses much smaller than the mev .
the bbn bound is physically relevant in many situations since it is a bit more constraining than the critical density bound .
the bbn bounds discussed so far are derived assuming a stochastic magnetic field at the bbn epoch .
this means that the isotropy of the geometry is not affected .
however , it is not unreasonable to think of the possibility that a magnetic field with a preferred direction may break the isotropy of the background
. thus bounds on the isotropy of the expansion at the bbn epoch may be turned into bounds on the shear induced by the presence of a magnetic field ( see , for instance , @xcite and @xcite ) .
let us now come to the last point .
the matter
antimatter fluctuations created at the electroweak epoch thanks to the presence of hypermagnetic fields may survive down to the epoch of bbn . since the fluctuations in the baryon
to entropy ratio are , in general , not positive definite , they will induce fluctuations in the baryon to photon ratio , at the bbn epoch .
the possible effect of matter
antimatter fluctuations on bbn depends on the typical scale of of the baryon to entropy ratio at the electroweak epoch . recalling that for @xmath882 gev the size of the electrowek horizon is approximately @xmath883 cm , fluctuations
whose scale is well inside the ew horizon at @xmath855 have dissipated by the bbn time through ( anti)neutron diffusion .
the neutron diffusion scale at @xmath855 is @xmath884 the neutron diffusion scale at @xmath885 is @xmath461 m , while , today it is @xmath208 pc , i.e. of the order of the astronomical unit .
antimatter fluctuations smaller than @xmath461 m annihilate before neutrino decoupling and have no effect on bbn . [ cols="^,^ " , ] two possibilities can then be envisaged .
we could require that the matter
antimatter fluctuations ( for scales @xmath886 ) are small .
this will then imply a bound , in the ( @xmath872,@xmath770 ) plane on the strength of the hypermagnetic background . in fig .
[ ad ] some typical baryon number fluctuations have been reported for different choices of the parameters @xmath872 and @xmath770 . in fig .
[ f5 ] such an exclusion plot is reported with the full line . with the dashed line
the bound implied by the increase in the expansion rate ( i.e. @xmath887 ) is reported .
finally with the dot - dashed line the critical density bound is illustrated for the same hypermagnetic background . for @xmath888 ( full line).,height=240 ]
the second possibility is to study the effects of large matter
matter domains .
these studies led to a slightly different scenario of bbn @xcite , namely bbn with matter
antimatter regions .
this analysis has been performed in a series of papers by rehm and jedamzik @xcite and by kurki - suonio and sihvola @xcite ( see also @xcite ) .
the idea is to discuss bbn in the presence of spherically symmetric regions of anti
matter characterized by their radius @xmath889 and by the parameter @xmath767 , i.e. the matter / antimatter ratio .
furthermore , in this scenario the net baryon - to - photon ratio , @xmath890 , is positive definite and non zero .
antimatter domains larger than @xmath461 m at @xmath278 kev may survive until bbn and their dissipation has been analyzed in detail @xcite .
antimatter domains in the range @xmath891 at 1 kev annihilate before bbn for temperatures between 70 kev and @xmath278 mev .
since the antineutrons annihilate on neutrons , the neutron to proton ratio gets smaller . as a consequence ,
the @xmath10 abundance gets reduced if compared to the standard bbn scenario .
the maximal scale of matter
antimatter fluctuations is determined by the constraints following from possible distortions of the cmb spectrum .
the largest scale is of the order of @xmath71 pc ( today ) , corresponding to @xmath5 m at @xmath278 kev .
yield is illustrated in the @xmath892 plane for @xmath893 .
as the matter / antimatter ratio decreases , we recover the standard @xmath10 yield.,height=240 ] suppose that matter
antimatter regions are present in the range of eq .
( [ range ] ) .
then the abundance of @xmath10 get reduced .
the yield of @xmath878 are reported as a function of @xmath767 , the matter
antimatter ratio and @xmath889 .
now , we do know that by adding extra - relativistic species the @xmath878 can be increased since the universe expansion gets larger .
then the conclusion is that bbn with matter
antimatter domains allows for a larger number of extra - relativistic species if compared to the standard bbn scenario .
this observation may have implications for the upper bounds on the stochastic gw backgrounds of cosmological origin @xcite since the extra - relativistic species present at the bbn epoch can indeed be interpreted as relic gravitons .
the vacuum fluctuations of the gauge fields present during the inflationary stage of expansion may be amplified if conformal invariance is broken .
it is then plausible that magnetic inhomogeneities are amplified not only at the scale of protogalactic collapse but also at larger length - scales .
agnetic fields can be generated over all physical scales compatible large - scale magnetic fields may then have various interesting implications on the physics of the cmb and of its anisotropies .
the following possible effects have been discussed through the years : * distortion of the planckian spectrum ; * shift of the polarization plane ( provided the cmb is linearly polarized ) ; * shift in the position of the first doppler peak ; * generic increase of the amount of the ( primary ) anisotropy . on top of these effects , magnetic fields can also modify the evolution of the tensor fluctuations of the geometry for typical length scales much smaller than the ones probed by cmb anisotropy experiments . the possible distortions of the planckian spectrum of cmb are usually discussed in terms of a chemical potential which is bounded , by experimental data , to me @xmath894 . magnetic field dissipation at high red - shift implies the presence of a chemical potential .
hence bounds on the distorsion of the planckian spectrum of the cmb can be turned into bounds on the magnetic field strength at various scales .
in particular @xcite the obtained bound are such that @xmath895 g for comoving coherence lengths between @xmath896 kpc and @xmath897 kpc .
large scale magnetic fields can also afffect the poosition of the doppler peak . in @xcite
this analysis has been performed in a non - relativistic approximation where the scalar perturbations of the geometry obey linearized newtonian equations of motion .
it has been found that , in this approximation , the effect of the presence of the magnetic fields is an effective renormalization of the speed of sound of the baryons .
large scale magnetic fields present prior to the recombination epoch may act as a source term in the evolution equation of the cosmological perturbations .
if magnetic fields are created over large length scales ( possibly even larger than the mpc ) it is rather plausible that they may be present after matter - radiation equality ( but before decoupling ) when primordial fluctuations are imprinted on the cmb anisotropies .
in fact , the relevant modes determining the large scale temperature anisotropies are the ones that are outside the horizon for @xmath898 where @xmath899 and @xmath900 are , respectively , the equality and the decoupling times .
after equality there are two complementary approaches which can be used in order to address phenomenological implications of large scale magnetic fields : * large scale magnetic fields are completely _ homogeneous _ outside the horizon after equality ; * large scale magnetic fields are _ inhomogeneous _ after equality .
the idea that _ homogeneous _ magnetic fields may affect the cmb anisotropies was originally pointed out by zeldovich @xcite and further scrutinized by grishchuk , doroshkevich and novikov @xcite .
since this idea implies that large - scale magnetic fields are weakly gravitating the details of the discussion ( and of the possible generalization of this idea ) will be postponed to the following section .
the suggestion that _ inhomogeneous _ magnetic fields created during inflation may affect the cmb anisotropies was originally proposed in @xcite . in @xcite
it was noticed that large scale magnetic fields produced during inflation may also lead to large fluctuations for modes which are outside the horizon after equality . in @xcite
it was noticed that either the produced magnetic fields may be used to seed directly the cmb anisotropies , or , in a complementary persepective , the cmb data can be used to put constraints on the production mechanism .
if large - scale magnetic fields are inhomogeneous , their energy momentum tensor will have , in general terms , scalar , vector and tensor modes @xmath901 which are decoupled and which act as source terms for the scalar , vector and tensor modes of the geometry @xmath902 whose gauge - invariant description is well explored @xcite(see also @xcite ) .
the effect of the scalar component of large scale magnetic fields should be responsible , according to the suggestion of @xcite , of the large scale temperature anisotropies .
the direct seeding of large - scale temperature anisotropies seems unlikely .
in fact , the spectrum of large scale gauge fields leading to plausible values of the magnetic energy density over the scale of protogalactic collapse is typically smaller than the value required by experimental data .
however , even if the primordial spectrum of magnetic fields would be tailored in an appropriate way , the initial conditions for the plasma evolution compatible with the presence of a sizable ( but undercritical ) magnetic field are of _ isocurvature _ type @xcite . in order to illustrate this point consider the evolution equations for the gauge - invariant system of scalar perturbations of the geometry expressed in terms of the two bardeen potential @xmath477 and @xmath903 corresponding , in the newtonian gauge , to the fluctuations of the temporal and spatial component of the metric @xcite .
the linearization of the @xmath904 , @xmath905 and @xmath906 components of the einstein equations leads , respectively , to : @xmath907 = \frac{a^2}{2 m_{\rm p}^2 } ( 1 + w ) \rho \partial_{i } u+ \frac{a^2}{2 m_{\rm p}^2 } \delta { \cal t}_{i}^{0 } , \label{0i}\\ & & \biggl\ { \psi '' + { \cal h } ( \phi ' + 2 \psi ' ) + ( { \cal h}^2 + 2 { \cal h } ' ) \phi + \frac{1}{2 } \nabla^2(\phi - \psi ) \biggr\}\delta_{i}^{j } \nonumber\\ & & - \frac{1}{2 } \partial_{i}\partial^{j } ( \phi - \psi ) = -\frac{a^2}{2 m_{\rm p}^2 } \delta { \cal t}_{i}^{j } + w \frac{a^2}{2 m_{\rm p}}^2 \rho \delta , \label{ij}\end{aligned}\ ] ] where @xmath908 and @xmath909 is the usual barotropic index , and @xmath910 , \nonumber\\ & & \delta { \cal t}_{i}^{0 } = \frac{\vcb \times \vec{\nabla } \times \vcb}{16 \pi^2 a^4\sigma}. \label{pertb}\end{aligned}\ ] ] are the fluctuations of the energy - momentum tensor of the magnetic field .
( [ 00an])([ij ] ) have to be supplemented by the perturbation of the covariant conservation equations for the fluid sources : @xmath911 if the magnetic field is force - free , i.e. @xmath912 the above system simplifies .
notice that eq .
( [ u ] ) has been already written in the case where the lorentz force is absent .
however , the forcing term appears , in the resistive mhd approximation , in eq .
( [ pertb ] ) . the term @xmath912 is suppressed by the conductivity , however , it can be also set exactly to zero . in this case ,
vector identities imply @xmath913 i.e. @xmath914 . with this identity in mind the system of eqs .
( [ 00])([ij ] ) and ( [ delta])([u ] ) can be written , in fourier space , as @xmath915 where @xmath916 combining eqs .
( [ 00k])([ijk ] ) with eq .
( [ shk ] ) the following decoupled equation can be obtained @xmath917 .
\label{genpsi}\ ] ] consider , for simplicity , the dark - matter radiation fluid after equality and in the presence of a force - free magnetic field . in this case
( [ genpsi ] ) can be immediately integrated : @xmath918 the solutions given in eq .
( [ solpsieq ] ) determine the source of the density and velocity fluctuations in the plasma , i.e. @xmath919 the solution for the velocity fields and density contrasts can be easily obtained by integrating eqs .
( [ dr1])([udm ] ) with the source terms determined by eq .
( [ solpsieq ] ) .
the purpose is not to integrate here this system ( see for instance @xcite for the standard case without magnetic fields ) . rather it is important to notice that there are two qualitatively different situation .
the first situation is the one where , @xmath920 namely the case where the constant mode of the bardeen potential is negligible if compared to the contribution of the magnetic field .
the derivative of the bardeen potential will be , however , non vanishing . in this case
the cmb anisotropies are said to be seeded by isocurvature initial conditions for the fluid of radiation and dark - matter present after equality . in the opposite case , nemely @xmath921 the constant mode of the bardeen potential
is provided , for instance , by inflation and the magnetic field represent a further parameter to be taken into account in the analysis of cmb anisotropies .
assuming , roughly , that @xmath922 , eq .
( [ secondq ] ) implies that for the typical scale of the horizon at decoupling the critical fraction of magnetic energy density should be smaller than @xmath499 . in the case of eq .
( [ secondq ] ) the leading order relations among the different hydrodynamical fluctuations are well known and can be obtained from eqs.([dr1])([udm ] ) if the magnetic field is approximately negligible , i.e. @xmath923 this result has a simple physical interpretation , and implies the adiabaticity of the fluid perturbations .
the entropy per matter particle is indeed proportional to @xmath924 , where @xmath925 is the number density of matter particles and @xmath605 is the radiation temperature . the associated entropy fluctuation , @xmath926 , satisfies @xmath927 where we used the fact that @xmath928 and that @xmath929 , where @xmath31 is the typical mass of the particles in the matter fluid . equation ( [ dmrad ] ) thus implies @xmath930 , in agreement with the adiabaticity of the fluctuations . after having computed the corrections induced by the magnetic field on the ( leading ) adiabatic initial condition
the temperature anisotropies can be computed using the sachs - wolfe effect @xcite . in terms of the gauge - invariant variables introduced in the present analysis , the various contributions to the sachs
wolfe effect , along the @xmath185 direction , can be written as @xmath931(\eta_{\rm dec } , \vec{x}(\eta_{\rm dec } ) ) - \int_{\eta_0}^{\eta_{\rm dec } } ( \phi ' + \psi')(\eta , \vec{x}(\eta ) ) d\eta , \label{sw}\ ] ] where @xmath569 is the present time , and @xmath932 is the unperturbed photon position at the time @xmath197 for an observer in @xmath933 .
the term @xmath934 is the peculiar velocity of the baryonic matter component .
= 8 cm in fig .
[ fcl1 ] the @xmath935 are plotted in the case of models with adiabatic initial conditions .
= 8 cm the data points reported in figs .
[ fcl1 ] and [ fcl2 ] are those from cobe @xcite , boomerang @xcite , dasi @xcite , maxima @xcite and archeops @xcite .
notice that the data reported in @xcite fill the `` gap '' between the last cobe points and the points of @xcite . in fig .
[ fcl2 ] the recent wmap data are reported @xcite .
recall , that the temperature fluctuations are customarily discussed in terms of their legendre transform @xmath936 where the coefficients @xmath937 define the angular power spectrum @xmath938 by @xmath939 and determine the two - point ( scalar ) correlation function of the temperature fluctuations , namely @xmath940 in fig .
[ fcl1 ] and [ fcl2 ] the experimental points are presented in terms of the @xmath935 spectra . in fig .
[ fcl1 ] and [ fcl2 ] the curves fitting the data are obtained by imposing adiabatic initial conditions .
if magnetic fields would seed directly the cmb anisotropies a characteristic `` hump '' would appear for @xmath941 which is not compatible with the experimental data . if , on the contrary , the condition ( [ secondq ] ) holds , then the modifications due to a magnetic field can be numerically computed for a given magnetic spectral index and for a given amplitude .
this analysis has been performed by koh and lee @xcite .
the claim is that by modifying the amplitude of the magnetic field in a way compatible with the cosmological constraint the effect on the scalar @xmath935 can be as large as the @xmath942 for a given magnetic spectral index .
the main problem is order to detect large - scale magnetic fields from the spectrum of temperature fluctuations is the parameter space . on top of the usual parameters common in the cmb analysis
at least two new parameters should be introduced , namely the magnetic spectral index and the amplitude of the magnetic field .
the proof of this statement is that there is no available analysis of the wmap data including also the presence of large scale magnetic fields in the initial conditions according to the lines presented in this section . to set bounds on primordial magnetic fields from cmb anisotropies it is often assumed that the magnetic field is fully homogeneous .
in this case the magnetic field amplitude has been bounded to be @xmath943 g at the decoupling time @xcite ( see also @xcite for a complementary analysis always in the case of a homogeneous magnetic field ) .
there is the hope that since magnetic fields contribute not only to the scalar fluctuations , but also to the vector and tensor modes of the geometry useful informations cann be also obtained from the analysis of vector and tensor power spectra . in @xcite temperature and polarization power spectra induced by vector and tensor perturbations
have been computed by assuning a power - law spectrum of magnetic inhomogeneities . in @xcite
the tensor modes of the geometry have been specifically investigated . in @xcite
it is argued that magnetic fields can induce anisotropies in the polarization for rather small length scales , i.e. @xmath944 . as already pointed out in the present section , magnetic fields
may be assumed to be force free in various cases .
this is an approximation which may also be relaxed .
however , even if magnetic fields are assumed to be force - free there is no reason to assume that their associated magnetic helicity ( or gyrotropy ) vanishes . in @xcite
the posssible effects of helical magnetic fields on cmb physics have been investigated .
if helical flows are present at recombination , they would produce parity violating temperature - polarization correlations .
the magnitude of helical flows induced by helical magnetic fields turns out to be unobservably small , but better prospects of constraining helical magnetic fields come from maps of faraday rotation measurements of the cmb @xcite
. large scale magnetic fields present at the decoupling epoch can also depolarize cmb @xcite .
the polarization of the cmb represents a very interesting observable which has been extensively investigated in the past both from the theoretical @xcite and experimental points of view @xcite .
forthcoming satellite missions like planck @xcite seem to be able to achieve a level of sensitivity which will enrich decisively our experimental knowledge of the cmb polarization with new direct measurements . if the background geometry of the universe is homogeneous but not isotropic the cmb is naturally polarized @xcite .
this phenomenon occurs , for example , in bianchi - type i models @xcite . on the other hand
if the background geometry is homogeneous and isotropic ( like in the friedmann - robertson - walker case ) it seems very reasonable that the cmb acquires a small degree of linear polarization provided the radiation field has a non - vanishing quadrupole component at the moment of last scattering @xcite . before decoupling photons ,
baryons and electrons form a unique fluid which possesses only monopole and dipole moments , but not quadrupole .
needless to say , in a homogeneous and isotropic model of frw type a possible source of linear polarization for the cmb becomes efficient only at the decoupling and therefore a small degree of linear polarization seems a firmly established theoretical option which will be ( hopefully ) subjected to direct tests in the near future .
the linear polarization of the cmb is a very promising laboratory in order to directly probe the speculated existence of a large scale magnetic field ( coherent over the horizon size at the decoupling ) which might actually rotate ( through the faraday effect ) the polarization plane of the cmb .
consider , for instance , a linearly polarized electromagnetic wave of physical frequency @xmath945 travelling along the @xmath946 direction in a plasma of ions and electrons together with a magnetic field ( @xmath947 ) oriented along an arbitrary direction ( which might coincide with @xmath946 in the simplest case ) .
if we let the polarization vector at the origin ( @xmath948 , @xmath949 ) be directed along the @xmath950 axis , after the wave has traveled a length @xmath951 , the corresponding angular shift ( @xmath952 ) in the polarization plane will be : @xmath953 from eq .
( [ faraday1 ] ) by stochastically averaging over all the possible orientations of @xmath182 and by assuming that the last scattering surface is infinitely thin ( i.e. that @xmath954 where @xmath955 is the thompson cross section ) we get an expression connecting the rms of the rotation angle to the magnitude of @xmath19 at @xmath956 @xmath957 ( in the previous equation we implicitly assumed that the frequency of the incident electro - magnetic radiation is centered around the maximum of the cmb ) .
we can easily argue from eq .
( [ faraday2 ] ) that if @xmath958 the expected rotation in the polarization plane of the cmb is non negligible .
even if we are not interested , at this level , in a precise estimate of @xmath952 , we point out that more refined determinations of the expected faraday rotation signal ( for an incident frequency @xmath959 ) were recently carried out @xcite leading to a result fairly consistent with ( [ faraday1 ] ) .
then , the statement is the following . _
if _ the cmb is linearly polarized and _ if _ a large scale magnetic field is present at the decoupling epoch , _ then _ the polarization plane of the cmb can be rotated @xcite .
the predictions of different models can then be confronted with the requirements coming from a possible detection of depolarization of the cmb @xcite .
magnetic fields can source the evolution equations of the fluctuations of the geometry also over length - scales much smaller than the ones where cmb anisotropy experiments are conduced .
this suggests that magnetic inhomogeneities may leave an imprint on the relc background of gravitational waves .
if a hypermagnetic background is present for @xmath960 , then , as discussed in section 5 and 6 , the energy momentum tensor will acquire a small anisotropic component which will source the evolution equation of the tensor fluctuations of the metric .
suppose now , that @xmath961 has constant amplitude and that it is also homogeneous .
then as argued in @xcite we can easily deduce the critical fraction of energy density present today in relic gravitons of ew origin @xmath962 ( @xmath963 is the redshift from the time of matter - radiation , equality ) . because of the structure of the amhd equations , stable hypermagnetic fields will be present not only for @xmath964 but for all the range @xmath965 where @xmath966 is the diffusivity frequency .
let us assume , for instance , that @xmath967 gev and @xmath968 .
then , the ( present ) values of @xmath969 is @xmath970 thus , @xmath971 .
suppose now that @xmath972 gev ; than we will have that @xmath973 hz
. suppose now , that @xmath974 as , for instance , implied by the analysis of the electroweak phase diagram in the presence of a magnetized background .
this requirement imposes @xmath975@xmath976 and , consequently , @xmath977 notice that this signal would occurr in a ( present ) frequency range between @xmath208 and @xmath13 hz .
this signal satisfies the presently available phenomenological bounds on the graviton backgrounds of primordial origin .
the pulsar timing bound ( which applies for present frequencies @xmath978 hz and implies @xmath979 ) is automatically satisfied since our hypermagnetic background is defined for @xmath980 .
the large scale bounds would imply @xmath981 but a at much lower frequency ( i.e. @xmath982 hz ) .
the signal discussed here is completely absent for frequencies @xmath983 .
notice that this signal is clearly distinguishable from other stochastic backgrounds occurring at much higher frequencies ( ghz region ) like the ones predicted by quintessential inflation @xcite .
it is equally distinguishable from signals due to pre - big - bang cosmology ( mainly in the window of ground based interferometers @xcite ) .
the frequency of operation of the interferometric devices ( virgo / ligo ) is located between few hz and 10 khz @xcite .
the frequency of operation of lisa is well below the hz ( i.e. @xmath984hz , approximately ) . in this model
the signal can be located both in the lisa window and in the virgo / ligo window due to the hierarchy between the hypermagnetic diffusivity scale and the horizon scale at the phase transition . in fig .
[ f7 ] the full thick line illustrates the spectrum of relic gravitational waves produced in a conventional model for the evolution of the universe .
the flat plateau corresponds to modes which left the horizon during the inflationary stage of expansion and re - entered duriing the radiation dominated phase .
the decreasing slope between @xmath985 and @xmath701 hz is due to modes leaving the horizon during inflation and re - entering during the maatter dominated stage of expansion .
clearly , the signal provided by a background of hypermagnetic fields can be even 7 order of magnitude larger than the inflationary prediction .
the interplay between gravitational waves and large - scale magnetic fields has been also the subject of recent interesting investigations @xcite .
up to now large scale magnetic fields have been treated in different frameworks but always within a perturbative approach . however
, magnetic fields can also break explicitely the isotropy ( but not the homogeneity ) of the background geometry suppose that at some time @xmath295 the universe becomes transparent to radiation and suppose that , at the same time , the four - dimensional background geometry ( which we assume , for simplicity , spatially flat ) has a very tiny amount of anisotropy in the scale factors associated with different spatial directions , namely @xmath986 , \label{1}\ ] ] where @xmath987 , as it will be clear in a moment , has to be only slightly different from @xmath988 . the electromagnetic radiation propagating along the @xmath989 and @xmath990 axes will have a different temperature , namely @xmath991 where @xmath992 and @xmath993 are the two hubble factors associated , respectively with @xmath988 and @xmath987 .
the temperature anisotropy associated with the electromagnetic radiation propagating along two directions with different expansion rates can be roughly estimated , in the limit @xmath994 , as @xmath995 dt= \frac{1}{2 } \int \varepsilon(t ) d\log{t}\ ] ] where , in the second equality , we assumed that the deviations from the radiation dominated expansion can be written as @xmath996 with @xmath997 .
the function @xmath998 can be connected with the shear parameter , i.e. @xmath999/[h(t ) + 2 f(t)]$ ] which measures the anisotropy in the expansion . in the standard context
@xcite einstein equations are solved in the metric ( [ 1 ] ) with fluid sources and in the presence of a magnetic field .
the shear parameter is then connected with the magnetic field intensity .
the dynamical origin of the primordial anisotropy in the expansion can be connected with the existence of a primordial magnetic field ( not dynamically generated but postulated as an initial condition ) or with some other sources of anisotropic pressure and , therefore , the possible bounds on the temperature anisotropy can be translated into bounds on the time evolution of the anisotropic scale factors @xcite . today
the amount of anisotropy in the expansion must be very small because of the effect we just described . in the present section
some highly speculative considerations on the possible rle of magnetic fields in the early universe will be intrtoduced . in spite of the fact that magnetic fields must be sufficiently small today ,
in the past history of the universe they might have been very large even modifying the dynamics of the geometry . during the magnetic phase
the anisotropy in the expansion will be constant ( or even grow ) .
however , after the magnetic phase a sufficienly long radiation dominated phase may isotropize the background leading to a tiny amount of anisotropy in the expansion .
this is the basic scenario invoked long ago by zeldovich @xcite . in modern approaches to cosmology
, the dynamics of the universe at very high densities is often discussed in terms of the low - energy string effective action .
it is then useful to analyze the zeldovich proposal in terms of the modern perspective . in the low energy limit , the dilaton field
is directly coupled to the kinetic term of the maxwell field @xcite @xmath1000 \label{actionan}\ ] ] where @xmath1001}$ ] is the maxwell field strength and @xmath1002 is the covariant derivative with respect to the string frame metric @xmath346 .
notice that @xmath1003 is the antisymmetric field strength . in eq .
( [ actionan ] ) the ellipses stand for a possible ( non - perturbative ) dilaton potential and for the string tension corrections parameterized by @xmath1004 . in eq .
( [ actionan ] ) @xmath345 can be thought as the maxwell field associated with a @xmath1005 subgroup of @xmath1006 .
consider a spatially flat background configuration with vanishing antysimmetric field strangth ( @xmath1007 ) and vanishing dilaton potential .
the dilaton depends only on time and the metric will be taken fully anisotropic since we want to study possible solutions with a homogeneous magnetic field which is expected to break the isotropy of the background : @xmath1008 by varying the effective action with respect to @xmath347 , @xmath346 and the vector potential @xmath1009 we get , respectively @xmath1010 \nonumber\\ & & - \frac{1}{2 } \delta_{\mu}^{\nu } g^{\alpha\beta}\partial_{\alpha}\varphi \partial_{\beta } \varphi - \nabla_{\mu } \nabla^{\nu } \varphi + \delta_{\mu}^{\nu } \frac{1}{\sqrt{-g}}\partial_{\alpha } [ \sqrt{-g } g^{\alpha\beta}\partial_{\beta } \varphi ] , \label{g}\\ & & \nabla_{\mu}\biggl [ e^{-\varphi } f^{\mu\nu}\biggr ] = 0 .
\label{b}\end{aligned}\ ] ] where @xmath1002 denotes covariant differentiation with respect to the metric of eq .
( [ metrican ] ) . inserting eq .
( [ phi ] ) into eq .
( [ g ] ) we get that eq .
( [ g ] ) can be expressed as @xmath1011 consider now a homogeneous magnetic field directed along the @xmath989 axis .
the generalized maxwell equations ( [ b ] ) and the associated bianchi identities can then be trivially solved by the field strength @xmath1012 .
the resulting system of equations ( [ phi ] ) , ( [ g2 ] ) can than be written , in the metric of eq .
( [ metrican ] ) , as @xmath1013 where eqs .
( [ 00an])([zz ] ) correspond , respectively to the @xmath904 and @xmath1014 components of eq .
( [ g2 ] ) .
these solutions represent the generalization of pre - big bang solutions @xcite to the case where a constant magnetic field is included .
generalizations of the solution ( [ c1 ] ) can be obtained in various models anisotropic ( but homogeneous ) models of the bianchi class @xcite like bianchi i , ii , iii , vi@xmath1018 and vii@xmath1019 according to the usual classification which can be found , for instance , in @xcite .
further generalizations of the solution ( [ c1 ] ) are possible in the kantowski - sachs [ ks ] metric with a magnetic field aligned alond the radial coordinate @xcite .
magnetized cosmological solutions can be used in order to describe some early phase in the life of the universe . this highly speculative idea
may be implemented in semi - realistic models provided the large anisotropy in the expansion decays at later epochs . in the standard context
@xcite the only source of isotropization is the evolution in a radiation dominated background since the shear parameter decays if the background is dominated by radiation .
if quadratic corrections are present , an initially anisotropic solution becomes isotropic as a consequence of the dynamical properties of the evolution equations as discussed in detail in @xcite .
gauge fields have also been studied as a source of inflationary evolution by ford @xcite and later by lidsey @xcite .
the gauge kinetic term is supplemented by a potential which is a function of @xmath1020 .
both exponential inflation @xcite and power - law inflation @xcite can be realized in these models .
the topics reported in the present review suggest various open problems on the nature of large - scale magnetic fields in the present and in the early universe : * in principle the morphology of magnetic fields in spiral galaxies could be used in order to understand the origin of galactic magnetism but in practice the observations offer answers which are often contradictory ; * recent observations of magnetic fields inside abell clusters confirm the presence of strong magnetic fields in the @xmath4 g range ; * there is evidence that also superclusters are magnetized ; * in spite of the fact that magnetic fields may modify the patterns of the cmb anisotropies and , eventually , induce anisotropies in the polarization , the theoretical analysis has been rarely corroborated by the comparison with the available experimental data ; * faraday rotation of the cmb polarization would be a powerful tool to study the possible exsitence of large scale magnetic fields prior to recombination ; * the imprints of magnetic fields on the relic background of gravitational waves ( in the window of wide band interferometers ) allows to test the existence of a background of abelian gauge fields at the electroweak epoch .
it would be important , for the theorist , a reasonable accuracy ( within one order of magnitude ) in the experimental determination of large - scale magnetic fields .
for instance even one order of magnitude in the intensity of the intra - cluster magnetic field ( i.e. @xmath276 g rather than @xmath1021 ) makes a difference as far as the problem of the origin is concerned . in the same perspective , more accurate determinations of the typical correlation scales of the intra - cluster fields would be desirable
. there is , at the moment , no compelling reason why large - scale magnetic fields should not be primordial , at least to some degree : * in the early universe magnetic fields are easily produced during phase transitions but their typical correlation scales are small ; * inflation greatly helps in producing correlations over very large scales and the ( early ) variation of gauge couplings during the inflationary phase allows the generation of intense large - scale fields ; * primordial magnetic fields may have an impact on the thermodynamical history of the universe , but , in practice , the obtained constraints improve only marginally , in various interesting cases , the critical density bound imposed on the magnetic energy density ; * if magnetic fields are generated during phase transitions , a growth in their correlation scale can not be excluded because of turbulent phenomena whose existence can be reasonably expected ( but not firmly predicted ) on the basis of our terrestrial knowledge of magnetized plasmas . in the next decade a progress of the empirical knowledge is expected in apparently unrelated areas like @xmath989 and @xmath1022-ray astronomy , radio - astronomy , cmb physics , detection of relic gravitational waves .
all these areas are connected with the existence and with the properties of large - scale magnetic fields in the early and in the present universe .
this connection is sometimes rather fragile since it is mediated by various theoretical assumptions .
however , there is the hope that , in a not too distant future , some of the puzzles related to the origin and existence of large - scale magnetic fields may be resolved and some of the current theoretical guessworks firmly ruled out .
some of the ideas presented in this review have been elaborated and assembled on the occasion of different sets of lectures delivered through the years . the author wishes to express his gratitude to d. babusci , h. de vega , m. gasperini , h. kurki - suonio , e. keihnen , n. sanchez , m. shaposhnikov , g. veneziano , a. vilenkin .
a special thank is due to g. cocconi for important remarks which improved the first draft .
in a two - fluid plasma description the charge carriers are the ions ( for simplicity we can think of them as protons ) and the electrons .
the two fluid equations treat the ions and the electrons as two conducting fluids which are coupled as in eqs .
( [ vlasov1])([transv ] ) . given the two- fluid description , one - fluid variables can be defined directly in terms of the two - fluid variables @xmath1023 , \nonumber\\ & & \vec{j } = e [ n_{p } \vec{v}_{p } - n_{e } \vec{v}_{e}].\end{aligned}\ ] ] in the case of a globally neutral plasma @xmath1024 and @xmath1025 .
the ion and electron equations then become @xmath1026\vec{v}_{\rm p } = e n_{q } ( \vec{e } + \vec{v}_{p } \times \vec{b } ) - \vec{\nabla } p_{p } + { \cal c}_{p e } , \label{prot}\\ & & m_{e } n_{q}\biggl [ \frac{\partial}{\partial t } + \vec{v}_{\rm e}\cdot \vec{\nabla } \biggr]\vec{v}_{\rm e } = -e n_{q } ( \vec{e } + \vec{v}_{e } \times \vec{b } ) - \vec{\nabla } p_{e } + { \cal c}_{e p } , \label{elec}\end{aligned}\ ] ] where @xmath1027 and @xmath1028 are the collision terms . in the globally neutral case the center of mass velocity becomes @xmath1029 and the one - fluid mass and charge density conservations become @xmath1030 summing up eqs .
( [ prot ] ) and ( [ elec ] ) leads , with some algebra involving the continuity equation , to the momentum transport equation in the one - fluid theory : @xmath1031 \vec{v } = \vec{j } \times \vec{b } - \vec{\nabla } p , \label{momt}\ ] ] where @xmath1032 . in eq .
( [ momt ] ) the collision term vanishes if there are no neutral particles , i.e. if the plasma is fully ionized .
the final equation of the one - fluid description is obtained by taking the difference of eqs .
( [ prot ] ) and ( [ elec ] ) after having multiplied eq .
( [ prot ] ) by @xmath1033 and eq .
( [ elec ] ) by @xmath1034 .
this procedure is more tricky and it is discussed in standard textbooks of plasma physics@xcite .
the key points in the derivation are that the limit for @xmath1035 must be taken .
the problem with this procedure is that the subtraction of the two mentioned equations does not guarantee that viscous and collisional effects are negligible .
the result of this procedure is the so - called one - fluid generalized ohm law : @xmath1036 the term @xmath1037 is nothing but the _ hall current _ and @xmath1038 is often called thermoelectric term . finally the term @xmath164 is the resistivity term and @xmath194 is the conductivity of the one - fluid description . in eq .
( [ genohm ] ) the pressure has been taken to be isotropic .
this is , however , not a direct consequence of the calculation presented in this appendix but it is an assumption which may ( and should ) be relaxed in some cases . in the plasma physics literature @xcite the anisotropic pressure contribution is neglected for the simple reason that experiments terrestrial plasmas show that this terms is often negligible .
consider and arbitrary closed surface @xmath176 which moves with the plasma .
then , by definition of the bulk velocity of the plasma ( @xmath157 we can also write @xmath1039
. the ( total ) time derivative of the flux can therefore be expressed as @xmath1040 where @xmath1041 is the boundary of @xmath176 .
using now the properties of the vector products ( i.e. @xmath1042 ) we can express @xmath1043 though the ohm law given in eq .
( [ genohm2 ] ) and we obtain that @xmath1044 using now eq .
( [ first ] ) together with the stokes theorem , the following expression can be obtained @xmath1045\cdot d\vec{\sigma } - \frac{1}{4\pi \sigma } \int_{\sigma } \vec{\nabla } \times\vec{\nabla } \times\vec{b}\cdot d\vec{\sigma}. \label{second}\ ] ] from the maxwell s equations the first part at the right hand side of eq .
( [ second ] ) is zero and eq .
( [ flux ] ) , expressing the alfvn theorem , is recovered . with similar algebraic manipulations ( involving the use of various vector identities ) , the conservation of the magnetic helicity can be displayed .
consider a closed volume in the plasma , then we can write that @xmath1046 where @xmath185 is the unit vector normal to @xmath176 ( the boundary of v , i.e. @xmath1047 ) and @xmath1048 is the component of the bulk velocity orthogonal to @xmath184
. the ( total ) time derivative of the magnetic helicity can now be written as @xmath1049 + \int_{\partial v=\sigma}\vec{a}\cdot\vec{b } \vec{v}_{\perp}\cdot\vec{n } d\sigma . \label{ddthel}\ ] ] the partial derivative at the right hand side of eq .
( [ ddthel ] ) can be made explicit .
then the one - fluid mhd equations should be used recalling that the relation between the electromagnetic fields and the vector potential , for instance in the coulomb gauge . finally using again the ohm law and transforming the obtained surface integrals into volume integrals ( through the divergence theorem ) eq .
( [ h2 ] ) , expressing the conservation of the helicity , can be obtained . in spite of the fact that the conservation of the magnetic helicity can be derived in a specific gauge ,
the magnetic helicity is indeed a gauge invariant quantity .
consider a gauge transformation @xmath1050 then the magnetic helicity changes as @xmath1051\ ] ] ( in the second term at the right hand side we used the fact that the magnetic field is divergence free ) . by now using the divergence theorem
we can express the volume integral as @xmath1052 = \int_{\partial v=\sigma } \xi\vec{b } \cdot\vec{n } d\sigma.\ ] ] now if , as we required , @xmath186 in @xmath184 , the integral is exactly zero and @xmath1053 is gauge invariant .
the condition @xmath186 is not specific of a particular profile of the magnetic field .
it can be always achieved by slicing the volume of integration in small flux tubes where , by definition , the magnetic field is orthogonal to the walls of the flux tube . | cosmology , high - energy physics and astrophysics are converging on the study of large - scale magnetic fields . while the experimental evidence for the existence of large - scale magnetization in galaxies , clusters and superclusters is rather compelling , the origin of the phenomenon remains puzzling especially in light of the most recent observations .
the purpose of the present review is to describe the physical motivations and some of the open theoretical problems related to the existence of large - scale magnetic fields .
cern - th/2003 - 307 2.cm massimo giovannini 1.cm _ cern , theoretical physics division , ch-1211 geneva 23 , switzerland _ |
tunnel current transport in superconducting junctions is a classical topic of interest and research.@xcite theoretical description of the superconducting tunneling given by the tunnel model@xcite is based on the assumption about equilibrium in the junction electrodes .
the resulting equations for the tunnel current are expressed through the nonperturbed density of states ( dos ) and equilibrium distribution function .
such an approach includes only single - particle tunnelling processes , and it is sufficient for describing current - voltage characteristics at large applied voltage @xmath2 , and within the subgap voltage region @xmath3 at nonzero temperature . at very small temperature , multiparticle tunneling processes must be taken into account.@xcite however , these processes are exponentially weak at small voltage , and if the temperature is not particularly small , @xmath4 , only tunneling of thermally excited quasiparticles plays significant role .
it is clear , however , that quasiparticle tunnelling generates nonequilibrium distribution in the electrodes , and that the effect is enhanced in diffusive junctions because of the scattering by impurities .
traditionally , this effect is considered to be negligibly small since it is of a higher order in the junction transparency .
how small is the nonequilibrium effect actually ?
in voltage biased superconducting junctions , the phase of the order parameter has different time dependencies in different electrodes , and the interference of the order parameters induced by the tunnelling leads to the time oscillations of the dos .
the character of these oscillations can be qualitatively understood from comparison with the well studied case of ballistic tunnel junctions.@xcite in the latter , the andreev levels are formed in the vicinity of the tunnel barrier ; their energies lie within the energy gap in the bulk electrodes and oscillate in time following evolution of the superconducting phase difference .
similarly , one may expect that in diffusive junctions the time oscillation of the dos will have the form of a `` breathing '' potential well localized in the vicinity of the tunnel barrier ( `` andreev band '' ) , which periodically , with the josephson period , traps and releases quasiparticles .
thus the quasiparticles spend larger time within the junction area compared to the travelling time , which should generate strongly nonequilibrium quasiparticle distribution and , as a result , an appreciable change of the tunnelling current
. the effect should be most pronounced for a planar junction geometry sketched in fig .
[ model](a , b ) , when the tunnel junction is connected to bulk electrodes via diffusive superconducting wires whose length exceeds the size of the andreev band .
otherwise , as in diffusive point contacts , the proximity of large equilibrium reservoirs will suppress the dos oscillations and hence formation of the andreev band because of rapid spreading out of the current .
= 8.0 cm in this paper we demonstrate that the nonequilibrium quasiparticle distribution generated by nonstationary process of formation of the andreev band during the tunneling considerably modifies the dc tunnel current at small applied voltage .
we show that the nonequilibrium effect results in a time dependent amplitudes of all the three current components given by the tunnel model,@xcite @xmath5 ( josephson current , quasiparticle current , and the interference current , respectively ) .
all the amplitudes have nonharmonic time dependence , and generally all of them contribute to the dc current .
however , at small voltage , the major contribution , nonanalytic in junction transparency and voltage , comes from the sine term ( josephson current ) .
more precisely , this additional dc current results from a non - adiabatic component of the distribution function associated either with inelastic relaxation of quasiparticles , or with the quasiparticle diffusion away from the tunnel barrier .
the additional dc current may considerably exceed the magnitude given by the tunnel model , and moreover , leads to a _ non - monotonous _ net dc current .
such effect has close qualitative similarity to the nonequilibrium effects in transparent weak links [ superconductor - normal metal - superconductor ( sns ) junctions and superconducting bridges ] studied earlier.@xcite there are however the differences : in superconducting bridges , the potential well appears due to suppression of the order parameter by the current concentration ( depairing effect),@xcite while in the tunnel junctions this effect does not play any essential role . in long sns junctions ,
the potential well is pre - prepared by the proximity effect , and it exists in the absence of the transport current ; as a result , the oscillations develop at small energy,@xcite while in the tunnel junctions the oscillations occur at large energies close to the gap edge in the electrodes .
organization of the paper is the following . in sec .
ii we formulate the theoretical model and microscopic equations describing current transport through the tunnel junction with diffusive leads . in sec .
iii we introduce an adiabatic approach for solving the time - dependent problem in the limit of low applied voltage .
this approach is applied to calculation of the spectral characteristics of the junction ( sec .
iv ) and the quasiparticle distribution function ( sec . v ) . in sec .
vi we calculate the ac and dc components of the net josephson current , and then discuss and summarize the results in sec .
vii and viii .
the model of tunnel junction we are going to study is depicted in fig . [ model](c ) and consists of a tunnel barrier with the transparency @xmath6 attached to bulk superconducting electrodes via two superconducting leads ( @xmath7 and @xmath8 ) .
we will consider diffusive limit , in which the elastic scattering length @xmath9 is much smaller than the coherence length @xmath10 , where @xmath11 is the diffusion coefficient ( we assume @xmath12 ) .
the length @xmath13 of the leads is assumed to be much larger than @xmath14 ( long junction ) , and their width will be supposed to be much smaller than the josephson penetration depth which implies homogeneity of the current along the junction .
similar model was considered in refs . and in study of the dc josephson effect in tunnel structures . under these conditions ,
the microscopic calculation of the electric current @xmath15 requires solution of the one - dimensional diffusive equations of nonequilibrium superconductivity@xcite ( see also a review@xcite ) for the @xmath16 matrix two - time keldysh - green s function @xmath17 in the leads , @xmath18 & = i \mathcal{d } \partial_x\check{j } , \;\ ; \check{j } = \check{g}\circ\partial_x \check{g } , \;\ ; \check{g}^2 = \delta(t_1 - t_2),\label{gk } \\
\nonumber \check{h } & = \bigl[i\sigma_z \partial_{t_1 } - e\varphi + \hat{\delta}(t_1)\bigr]\delta(t_1-t_2),\end{aligned}\ ] ] where @xmath19 , @xmath20 is the electric potential , @xmath21 and @xmath22 are the modulus and the phase of the order parameter , respectively , @xmath23 are the pauli matrices , and @xmath24 in eq .
( [ defg ] ) , @xmath25 are the @xmath26 nambu matrix retarded and advanced green s functions , and @xmath27 is the matrix distribution function ( we use ` check ' for @xmath16 and ` hat ' for @xmath26 matrices ) .
the multiplication procedure in eqs .
( [ gk])- is defined as the time convolution , @xmath28 in eqs .
( [ gk ] ) , we neglect the inelastic collision term , which will be taken into account later , in sec . vi .
the boundary conditions for the function @xmath29 and the matrix current @xmath30 at the left ( @xmath31 ) and the right ( @xmath32 ) sides of the tunnel junction are given by relation@xcite @xmath33,\label{kl}\ ] ] where @xmath34 is the junction resistance and @xmath35 is the normal conductivity of the leads per unit length .
the electric current is related to the keldysh component of the matrix current @xmath30 as @xmath36 and thus it can be expressed through the boundary value @xmath37 , @xmath38^k(t , t ) .
\label{curr2}\ ] ] equations ( [ gk ] ) can be decomposed into the diffusion equations for the green s functions , @xmath39 = i \mathcal{d}\partial_x \hat{j } , \;\ ; \hat{j } = \hat{g}\circ\partial_x \hat{g } , \;\ ; \hat{g}^2 = \delta(t_1 - t_2),\label{usadel}\ ] ] and the equation for the keldysh component @xmath40 , @xmath41 = i \mathcal{d}\partial_x \hat{j}^k , \ ; \hat{j}^k = \hat{g}^r \circ\partial_x\hat{g}^k + \hat{g}^k \circ\partial_x \hat{g}^a . \label{kineticg}\ ] ] the boundary conditions for the functions @xmath42 and @xmath40 at the tunnel barrier follow from eq .
( [ kl ] ) , [ boundary2 ] @xmath43 , \label{boundaryg } \\
\hat{j}_{0}^k & = ( w / \xi_0)\bigl[\check{g}_{-0 } , { \check{g}}_{+0}\bigr]^k .
\label{boundarygk}\end{aligned}\ ] ] in eqs .
( [ boundary2 ] ) , the transparency parameter @xmath44 is defined as @xmath45 where @xmath46 is the normal resistance of the lead per length @xmath14 . it has been shown in refs . and that this quantity rather than the barrier transparency @xmath6 plays the role of a true transparency parameter in the theory of diffusive tunnel junctions ( see also discussion in sec .
vii ) . in this paper
, we will consider the limit @xmath47 , which corresponds to the conventional tunneling concept . in this case , according to eqs .
( [ boundary2 ] ) , the gradients of all functions are small . within the tunnel model , which assumes @xmath44 to be the smallest parameter in the theory , these gradients are neglected , and the functions @xmath42 and @xmath48 are assumed to be local - equilibrium within the leads . in our consideration , we will lift this assumption and suppose the local - equilibrium form of these functions only within the bulk electrodes ( reservoirs ) . attributing the reference point for the phase , @xmath49 , to the left electrode ,
@xmath50 , these functions in the right electrode , @xmath51 , are given by relations , [ equil ] @xmath52 written in terms of the mixed wigner representation @xmath53 of the two - time functions , @xmath54 in eq .
( [ wigner ] ) , the variable @xmath55 has the meaning of the quasiparticle energy , and @xmath56 is a real time .
similar equations , with @xmath49 and @xmath57 , apply to the left electrode , @xmath50 . because of the small value of the tunneling parameter @xmath44 one can neglect the variation of the electric potential and the superconducting phase along the leads , and assume the voltage @xmath58 and the phase difference @xmath59 between the reservoirs to be directly applied to the tunnel barrier . following this argument
, one can also neglect the variation of the charge imbalance function @xmath60 proportional to a small electric field ( @xmath61 ) penetrating the superconducting leads .
furthermore , the small value of the superfluid momentum in the superconducting leads,@xcite @xmath62 , enables us to neglect a small effect of the suppression of the energy gap,@xcite [ @xmath63 , by the superfluid momentum . in such approximation , the coefficients in the equation within the left lead , @xmath64 , are time - independent functions , similar to the value of @xmath29 at the left electrode . at @xmath65 , using the gauge transformation , @xmath66 , of the function @xmath29 with a unitary operator @xmath67$],@xcite we exclude the time - dependent phase and the electric potential from the equations for the function @xmath68 and the boundary conditions at @xmath51 , which then become similar to the equations for @xmath69 at @xmath64 and the boundary conditions at @xmath50 .
this results in the symmetry relation @xmath70 , which allows us to replace the function @xmath71 in the boundary condition and in the expression for the electric current by the inverse gauge transformation of the function @xmath72 , @xmath73 as the result , the problem is reduced to the solution of a static equation for the function @xmath69 within the left lead with the time - dependent boundary condition at the tunnel barrier .
similar approach is used in the theory of ballistic point contacts , where the josephson coupling is described by an effective time - dependent matching condition for the gauge - transformed bogolyubov - de gennes equations in the leads . in a general nonstatic case , the function @xmath74 consists of a set of harmonics @xmath75 , @xmath76 , which are coupled to each other through a complicated set of recursive equations following from eqs .
( [ gk ] ) ( see discussion in ref . ) .
the problem essentially simplifies if the distance @xmath77 between the harmonics is much smaller than the smallest scale @xmath78 of variations in the quasiparticle spectrum , @xmath79 the magnitude of @xmath78 will be indicated below , see eq .
( [ depth ] ) .
in the limit of low applied voltages , the evolution of the quasiparticle spectrum and distribution function can be described within the adiabatic approximation using expansion over small josephson frequency . in the static case ,
@xmath80 , the function @xmath29 depends only on the time difference @xmath81 , and the wigner transformation reduces eqs .
( [ usadel ] ) to the standard usadel equations,@xcite [ usadelst ] @xmath82 for the scalar components of the green s function @xmath83 the functions @xmath84 and @xmath85 determine the spectral characteristics of the system ; in particular , the quantity @xmath86 is the dos normalized over its value @xmath87 in the normal state . in what follows
, we will express the advanced green s functions through the retarded ones , @xmath88 , using the general relation @xmath89 , and omit the superscript @xmath34 , assuming all green s functions to be retarded . at small applied voltages ,
we proceed to the wigner representation of the two - time functions and expand the time convolutions in eqs .
( [ usadel])- to first order in @xmath77 , @xmath90 where @xmath91 denotes the poisson brackets in the energy - time space . within such approximation ,
the green s function @xmath92 holds the matrix structure , and the gauge - transformed functions @xmath93 and @xmath94 [ see eq .
( [ defbar ] ) ] , which enter the boundary conditions , read [ overline ] @xmath95 the expression for the electric current obtained from eqs .
( [ curr2 ] ) and consists of the three terms , [ currents ] @xmath96 where @xmath97 is the boundary value of the distribution function @xmath98 , and @xmath99 , @xmath100 , and @xmath101 are the spectral densities of the different current components . such structure of the current is similar to the result of the tunnel model:@xcite indeed , when the current spectral densities and distribution function approach nonperturbed equilibrium form , the first term in eq .
( [ i ] ) describes the ac josephson current , the term @xmath102 is the dissipative quasiparticle current which approaches the value @xmath103 in the normal state , and @xmath104 is the interference current .
however , eqs .
( [ currents ] ) are more general , they include the josephson oscillations of the spectral characteristics together with a nonequilibrium form of the distribution function .
to first order in @xmath77 , the spectral functions @xmath105 and @xmath106 obey static usadel equations with time - dependent boundary condition following from eqs .
( [ boundaryg ] ) and , @xmath107 this boundary condition was found in refs . and
for a similar structure with the time - independent phase difference .
thus , at low enough voltages , the spectral characteristics of the junction adiabatically follow time variations of @xmath108 .
equations and can be supplemented by helpful identities following from the unity components of the matrix equations and , @xmath109 in order to satisfy the condition , we introduce usual parametrization @xmath110 , @xmath111 . furthermore , we will neglect a small effect of suppression of the order parameter near the barrier , assuming @xmath21 to be homogeneous [ see comment to eq .
( [ depth ] ) ] . then the equation and the boundary condition for the spectral angle @xmath112 , following from eqs .
( [ usadel1 ] ) and , take the form in a dimensionless variable @xmath113 , @xmath114^{-1/2 } , \label{eqtheta } \\
& \left(\partial_z \theta\right)_0 = -2\gamma(t)\sinh 2\theta_0 , \quad \gamma(t)=w \sin^2[\phi(t)/2 ] .
\label{boundarytheta1}\end{aligned}\ ] ] the quantity @xmath115 in eq .
( [ boundarytheta1 ] ) has the meaning of a time - dependent depairing factor related to the discontinuity of the superconducting phase @xmath108 at the tunnel barrier .
when the phase approaches multiple of @xmath116 , this factor turns to zero , and the spectral angle becomes homogeneous and equal to its bulk value @xmath117 . at arbitrary @xmath22 , equations and describe deviation of the spectral angle from @xmath118 at the distance @xmath119 from the barrier ; thus , in a long junction , @xmath120 , we can apply the solution for a semi - infinite lead,@xcite @xmath121 = \tanh\bigl[(\theta_0 - \theta_s)/{4}\bigr]\exp(k z ) .
\label{soltheta}\ ] ] = 8.0 cm the equation for the boundary value @xmath122 of the spectral angle follows from eqs .
( [ eqtheta])- , @xmath123 the behavior of the dos calculated by numerical solution of eq .
( [ eqtheta0 ] ) is shown in figs .
[ dosex ] and [ dos2d ] . at @xmath49
( we consider the phase within the period @xmath124 ) , the dos approaches the bcs energy dependence depicted by the dashed line in fig .
[ dosex](a ) .
the current through the junction affects the singularity of dos at the bulk energy gap , which becomes a beak - shaped peak with root singularities of the derivatives resulting from the divergence of the decay length @xmath125 at @xmath126 [ see eq .
( [ eqtheta ] ) ] .
this divergence@xcite is an analog of the long - range proximity effect at @xmath127 in ns structures ( see , e.g. , a review@xcite ) .
the dos identically turns to zero inside an oscillating minigap , at @xmath128 .
similar conclusions can be drawn regarding behavior of the current spectral densities shown in fig .
[ spectral](a c ) . within the subgap region , @xmath129 ,
the dos decreases at the distance @xmath130 from the barrier ( more precisely , at @xmath131 , due to the long - range proximity effect mentioned above ) , as shown in fig .
[ dosex](b ) .
this implies that the subgap states form an oscillating cluster of bound states , `` andreev band '' , which has triangular shape in the @xmath132-space as shown in fig .
[ dos2d ] ( similar cluster appears near the lower gap edge , @xmath133 )
. the energy depth of the cluster is @xmath134 [ see eq . ( [ e*ipt ] ) below ] , and it spreads over the distance of several @xmath14 from the tunnel barrier .
we note that because of comparatively large value of @xmath78 , we may neglect in the spectral characteristics a smaller ( @xmath135 ) effect of local suppression of the order parameter @xmath21 in the vicinity of the barrier.@xcite obviously , the quantity @xmath78 plays the role of the smallest characteristic energy of the spectrum in eq .
( [ cond ] ) . the small value of the parameter @xmath44 enables us to apply a perturbative approach for solving eq .
( [ eqtheta0 ] ) . at the energies far enough from the gap edge , where @xmath112 is of the order of unity , the quantity @xmath136 is close to @xmath118 , and eq .
( [ eqtheta0 ] ) leads to the following asymptotic relation for @xmath136 , @xmath137 however , when the energy approaches @xmath21 , this expansion fails due to divergence of @xmath118 at the bulk gap edge .
this requires modification of the perturbative theory within the region @xmath138,@xcite where the quantities @xmath136 and @xmath118 are large , while their difference , @xmath139 , may have arbitrary value . using these arguments
, we hold only large exponents , @xmath140 , in the hyperbolic functions in the right - hand side ( rhs ) of eq .
( [ eqtheta0 ] ) and use the asymptotic expression , @xmath141 at small @xmath142 . then , introducing a dimensionless energy variable @xmath143 and the normalized spectral function @xmath144 according to relations [ eqipt ] @xmath145^{1/5 } , \label{eqy } \\ & \exp\theta_0 = p(t)y(\epsilon ) , \qquad & \exp\theta_s= p(t)\epsilon^{-1/2 } , \label{expipt}\end{aligned}\ ] ] we reduce eq .
( [ eqtheta0 ] ) to a numerical algebraic equation for @xmath144 , @xmath146 according to eqs .
( [ eqipt ] ) , the parameter @xmath147 determines a characteristic scale of the spectral functions @xmath148 in the vicinity of the bulk gap edge .
= 8.5 cm the choice of the relevant complex root @xmath144 of eq .
( [ eqy1 ] ) is determined by the requirement for the asymptotic behavior at @xmath149 , which must coincide with the energy dependence given by direct perturbative expansion at @xmath150 . within the main approximation
, the function @xmath144 turns to zero at large @xmath143 as @xmath151 . at the energies @xmath143 smaller than @xmath152 , the function @xmath144 becomes imaginary , and
therefore the spectral functions @xmath153 turn to zero at @xmath154 , where the minigap edge @xmath155 is determined by relation following from eq .
( [ eqy ] ) , @xmath156 , \;\ ; c = { 25}/({3\cdot 6^{1/5 } } ) \approx 5.82 .
\label{e*ipt}\ ] ] we note that due to moderately small value @xmath157 of the transparency parameter , the minigap values , obtained by numerical solution of eq .
( [ eqtheta0 ] ) and shown in figs .
[ dosex ] and [ spectral ] , slightly differ from their approximate values found from eq .
( [ e*ipt ] ) ( by some 10% ) . using the ipt approximation
, however , significantly simplifies both the analytical and numerical calculations , allowing us to factorize the time and energy dependencies of the spectral functions . applying this approximation to eq .
( [ soltheta ] ) , we obtain the spatial dependence of the spectral angle , @xmath158 , \label{thetaipt } \\
\tanh^2 s(\epsilon ) & = y(\epsilon)\sqrt{\epsilon } , \quad k(\epsilon , t ) = \bigl[2\sqrt{\epsilon}/ip(t)\bigr]^{1/2}. \label{sipt}\end{aligned}\ ] ] it is interesting to note that the spectral functions are inhomogeneous within the leads at @xmath159 and finite at any finite distance from the barrier , as it is seen in fig .
[ dos2d ] , which seems to contradict the divergence of the characteristic decay length @xmath125 of the spectral angle at the gap edge .
the explanation to this effect is the following : at @xmath160 ( @xmath161 ) , the functions @xmath162 and @xmath163 in eq .
( [ sipt ] ) turn to zero as @xmath164 , which cancels the divergence of the prefactor in eq .
( [ thetaipt ] ) and results in the following expression for the spectral angle at the gap edge , @xmath165 ^ 2.\end{aligned}\ ] ] according to eq .
( [ e = d ] ) , the spectral functions at @xmath166 and @xmath167 diverge , which restores their bcs - divergence at the gap edge within the bulk superconductor .
to first order in @xmath77 , equation in the wigner representation has the form of a diffusive kinetic equation,@xcite @xmath168 where @xmath169 is the dimensionless diffusion coefficient . at the reservoir ,
the distribution function approaches equilibrium value , @xmath170 .
the boundary condition at the tunnel barrier following from eq .
( [ boundarygk ] ) , @xmath171 determines the boundary value of the flow @xmath172 of nonequilibrium quasiparticles escaping from the barrier . in this equation ,
the function @xmath173 describes the source of the nonequilibrium - the energy exchange between the quasiparticles and the superfluid condensate due to the oscillations the supercurrent spectral density @xmath174 .
this results in the dos oscillations which can be expressed through @xmath174 by integrating the identity over the left lead , taking into account eq .
( [ equal2 ] ) and extending the integration over the whole negative semi - axis due to rapid decay of the time derivatives of all quantities at a small distance @xmath175 , @xmath176 due to the existence of oscillating andreev band in the vicinity of the barrier , the behavior of the quasiparticles strongly depends on their energy .
indeed , most of excitations with the energy @xmath177 spend a short time @xmath178 near the barrier , they rapidly diffuse away along the leads and escape into the reservoirs ( travelling quasiparticles ) . the quasiparticles with low energy approaching the contact at the distance @xmath179 while the depth of the andreev band increases ( @xmath180 ) are trapped by the andreev band and spend much larger time @xmath181 inside the band , following the spectrum oscillations . during the next half - period ( @xmath182 ) , the depth of the andreev band decreases , and the trapped quasiparticles are pushed out to the extended states .
such a physical picture is similar to the quasiparticle dynamics within the surface skin layer of a superconductor irradiated by rf electromagnetic field.@xcite below we will perform separate analysis of the kinetics of the travelling and trapped quasiparticles . at @xmath177 , the coefficients @xmath183 and @xmath184 in eq .
( [ kin1 ] ) rapidly vary in the near vicinity @xmath119 of the barrier and then , at @xmath185 , approach their values in a bulk superconductor , @xmath186 , @xmath187 . correspondingly , the solution of eq .
( [ kin1 ] ) has the form of a slowly varying function @xmath188 with a small but rapidly varying addition @xmath189 , which vanishes at the distances @xmath185 from the barrier .
thus , within the main approximation , the population of the extended states , @xmath190 , satisfies eq .
( [ kin1 ] ) with the asymptotic values of the coefficients at @xmath185 , @xmath191 to derive effective boundary condition to this equation , we subtract eq .
( [ eqf > ] ) from the initial kinetic equation , keeping large spatial derivative of the function @xmath189 , @xmath192 , \label{bound3}\ ] ] and then integrate eq .
( [ bound3 ] ) over @xmath193 along the left lead . since all terms in this equation vanish at @xmath185 , we extend the integration over the whole semi - axis , similar to eq .
( [ equaln ] ) , and take the smooth function @xmath194 in its left - hand side ( lhs ) at @xmath195 .
then , using eq .
( [ bound1 ] ) , we obtain the boundary condition , @xmath196 the distribution function @xmath194 consists of a static part @xmath197 and a dynamic ( time - dependent ) part , @xmath198 .
the static part linearly varies along the lead and approaches equilibrium distribution function @xmath199 at @xmath50 , @xmath200 .
\label{solg}\ ] ] the dynamic component @xmath201 has the form , @xmath202.\nonumber\end{aligned}\ ] ] here @xmath203 denotes time averaging over the period @xmath204 , and @xmath205 are the fourier harmonics of the function @xmath206 , @xmath207 equation was obtained assuming the decay length @xmath208 of the oscillations of the distribution function to be much smaller than the junction length @xmath13 ; this enables us to apply the solution for a semi - infinite lead . in the opposite case , @xmath209
, the reservoirs effectively suppress these oscillations , and the function @xmath210 rapidly decays while the voltage decreases .
the equation for the harmonics @xmath205 follows from eq .
( [ bound > ] ) , @xmath211 the first two terms in the lhs of eq .
( [ eqhn1 ] ) , proportional to small applied voltage @xmath77 , can be neglected as compared to the fourth term .
then , averaging eq .
( [ eqhn1 ] ) over @xmath212 and using the identity , we obtain the expression for the third term , @xmath213 which therefore is also proportional to small @xmath77 and can be neglected as well . as the result ,
the approximate solution of eq .
( [ eqhn1 ] ) reads @xmath214 , i.e. , the dynamic part @xmath201 of the distribution function is expressed through the boundary value of the static part , @xmath215 .
it follows from this consideration that the dynamic part of the distribution is generated by the oscillations of the quasiparticle spectrum at @xmath177 .
these oscillations provide the energy transfer from the electric field to quasiparticles , which generally results in the ` heating ' of quasiparticles , i.e. , their redistribution towards higher energies ( pumping effect ) .
the heating is described by a diffusion equation for @xmath216 in the energy space following from the eqs .
( [ third ] ) and , @xmath217 similar equation has been obtained in refs . for mar - induced heating in long sns junctions , where the nonequilibrium is constrained by inelastic collisions .
in our strongly inhomogeneous case , the role of relaxation factor in eq .
( [ solg0 ] ) is played by the diffusive flow @xmath218 of nonequilibrium quasiparticles from the junction .
the intensity of the heating effect is determined by the estimate of the diffusion term in eq .
( [ solg0 ] ) with the equilibrium function @xmath219 , @xmath220 . due to the presence of the two small factors ,
this term is small for reasonable lengths of the leads , which enables us to approximate the function @xmath216 with the equilibrium distribution function , @xmath221 .
then the boundary value of the distribution function of travelling quasiparticles reads , @xmath222 at small voltages , @xmath223 , the spatial size @xmath14 of the andreev band is much smaller than the smallest kinetic length @xmath224 , therefore the main part @xmath188 of the distribution function of the trapped quasiparticles is spatially uniform , @xmath225 .
then the kinetic equation takes the form , @xmath226 by averaging eq .
( [ eqf<1 ] ) over @xmath193 and using the boundary condition , we obtain partial differential equation ( pde ) for the function @xmath227 , @xmath228 as long as the coefficients of this pde satisfy the identity , the equation @xmath229 for its characteristic curves @xmath230 in the @xmath231 space is easily integrated , @xmath232 this allows us to reduce the pde to the ordinary differential equation for the function @xmath227 along the characteristics , @xmath233 the quantity @xmath234 has the meaning of averaged number of states with the energy smaller than given energy @xmath55 , normalized over @xmath87 ; in the bulk superconductor , it approaches @xmath235 .
the bottom @xmath236 of the andreev band represents the reference point , @xmath237=0 $ ] . within the ipt approximation , eq .
( [ xi ] ) reads [ jn ] @xmath238^{1/5}j(\epsilon ) , \quad j(\epsilon ) = \int_{\epsilon_\ast}^\epsilon d\epsilon ' n(\epsilon ' ) , \label{j } \\
n(\epsilon ) & = |\epsilon|^{-1/2}\operatorname{\mathop{\mathrm{im}}}\sqrt { iy(\epsilon ) } , \label{n}\end{aligned}\ ] ] where the function @xmath239 is related to the averaged dos as @xmath240^{-1/5}n(\epsilon)$ ] . as follows from eq .
( [ n ] ) , the function @xmath241 plotted in fig .
[ spectral](d ) exhibits a singularity at the bulk gap edge , due to the long - range proximity effect in the superconductor [ see comments to eq .
( [ eqtheta0 ] ) ] .
= 8.5 cm according to eq .
( [ eqfxi ] ) , the distribution of trapped excitations is a function of the `` integral of motion '' @xmath242 , @xmath243 , \label{fxi}\ ] ] and therefore holds a constant value along the trajectories @xmath230 in the @xmath231 space shown in fig .
[ levels ] .
these trajectories can be interpreted as diffusive analogue of andreev levels adiabatically moving in a slowly varying external field ( cf .
ref . ) . from this
we conclude that the trapped quasiparticles are completely dragged by the spectrum oscillations , and their energy periodically changes , in accordance with the relation following from eqs .
( [ xi ] ) and , @xmath244 since the bound states at the tunnel barrier appear and vanish periodically , their population is imposed by the distribution of travelling quasiparticles at the bulk gap edge .
this is expressed by the matching condition @xmath245 , \;\ ; 0<\phi < \pi . \label{bound<>}\ ] ] here @xmath246 is the moment of the crossing the gap edge by quasiparticle with the given value of @xmath247 ( see fig . [ levels ] ) , i.e.
, it is the smallest solution of the equation @xmath248 within the period of the dos oscillations .
since the trapped quasiparticles spend long time at the barrier , the heating effect for them becomes well pronounced , in contrast to the travelling quasiparticles .
indeed , as follows from eq .
( [ f > ] ) , the time - dependent part @xmath210 of the distribution function turns to zero at @xmath249 due to singularity of @xmath250 , and therefore the distribution function of trapped quasiparticles turns into a plateau ( i.e. , does not depend on energy ) , @xmath251 non - adiabatic correction to the distribution function is produced by inelastic relaxation . to include the effect of inelastic collisions on the quasiparticle kinetics
, we will add a collision term in the @xmath252-approximation , @xmath253 $ ] , to the rhs of the kinetic equation .
within such a model , the kinetic equation for the trapped quasiparticles reads @xmath254- f_<(\xi , t)\bigr\}.\ ] ] the initial condition to eq .
( [ eqtau1 ] ) follows from eq .
( [ bound < > ] ) , @xmath255
proceeding to the calculation of the electric current , we start from the analysis of the ac josephson current @xmath256 . first , we note that the effect of the dos oscillations on the amplitude @xmath257 , calculated with the equilibrium distribution function @xmath219 , gives rise only to small corrections to the tunnel model result , @xmath258.@xcite indeed , in this case , the integral in eq .
( [ i1 ] ) is determined by the imaginary matsubara energies @xmath259 far from the gap edges , which allows us to apply ordinary perturbation approach [ see eq .
( [ opt ] ) ] to the calculation of the spectral density @xmath174 .
this results in the correction of the order of @xmath44 to @xmath260,@xcite which will be neglected below . in
what follows , we will focus on the nontrivial contributions of nonequilibrium quasiparticles to the josephson current @xmath257 .
such contributions , @xmath261 and @xmath262 , come from the dynamic part @xmath210 of the distribution of travelling quasiparticles and the distribution @xmath227 of trapped quasiparticles , [ di1 ] @xmath263\ , -\
, n_{\textit{eq}}(e)\}. \label{i1<}\end{aligned}\ ] ] due to rapid convergence of the integral in eq .
( [ i1h ] ) at @xmath264 , the function @xmath265 can be taken at @xmath159 .
for similar reason , we apply the ipt expression for the spectral density @xmath174 , which leads to the following result , @xmath266 where @xmath267 ^ 2 \approx 0.15,\;\;\ ; \nu=0.2,\end{aligned}\ ] ] and @xmath268 is the euler s beta function .
the time average of the current ( [ it ] ) does not vanish , @xmath269 and , moreover , it exhibits a strongly nonlinear ( @xmath270 ) voltage dependence .
we recall that the validity of these results is restricted by the condition of small diffusion length as compared to the junction length , @xmath271 , or , equivalently , @xmath272 , where @xmath273 is the thouless energy [ see comments to eq .
( [ solh ] ) ] . in the opposite case , @xmath274 ,
the oscillations of the distribution function are damped by equilibrium reservoirs , and the nonequilibrium dc current , produced by travelling quasiparticles , rapidly vanishes .
similar damping effect is caused by inelastic collisions , when the inelastic scattering length becomes smaller than the diffusion length @xmath224 .
the dissipative dc current of travelling quasiparticles results from non - adiabatic ( diffusive ) evolution of the distribution function @xmath275 . similar conclusion is also true for the trapped quasiparticles : only the non - adiabatic part of the distribution function contributes to the dissipative dc current . indeed , inserting the adiabatic distribution function @xmath276 , eq .
( [ plateau ] ) , into eq .
( [ i1 < ] ) , and taking advantage of a small integration interval , which allows us to expand the difference of the equilibrium functions over @xmath277 , we get , @xmath278 this equation contains only odd harmonics of the ac josephson current .
the dc current results from the non - adiabatic correction @xmath279 to the distribution function , found from the solution of the kinetic equation , @xmath280 \bigr\rangle_t \!\!= \frac{\delta f_1}{er } w^{4/5 } k(ev).\end{aligned}\ ] ] the factor @xmath281 in eq .
( [ ieps ] ) is a complicated numerical function of the relaxation parameter @xmath282 . in the limits of weak ( @xmath283 ) and strong ( @xmath284 ) relaxation
, this function decreases as @xmath285 and @xmath282 , respectively . in the spirit of our modelling approach to the problem of inelastic scattering ,
it is reasonable to approximate @xmath281 by usual relaxation factor @xmath286 where the parameters @xmath287 are to be evaluated from the asymptotics of @xmath281 in both limiting cases . in the weak relaxation limit , @xmath283 , the adiabatic distribution function
is given by eq .
( [ plateau ] ) , @xmath288 ; then the non - adiabatic correction @xmath279 is determined by the local energy @xmath289 averaged along the trajectories @xmath290 , @xmath291.\ ] ] substituting @xmath279 into the expression and using the ipt relations for the function @xmath234 , we obtain the coefficient @xmath292 in the relaxation factor @xmath281 , @xmath293 \approx 1.38,\end{aligned}\ ] ] where @xmath294 is the solution of equation @xmath295 , and the function @xmath296 is defined in eq .
( [ j ] ) . in the opposite limit of strong relaxation , @xmath297
, the initial condition is quickly ` forgotten ' , and the adiabatic part of the distribution is the equilibrium function of the local energy , @xmath298 $ ] . in this case , the non - adiabatic addition @xmath279 is proportional to the time derivative of @xmath289 along the trajectory , @xmath299 using the expression for @xmath300 and calculating @xmath301 in eq .
( [ ieps ] ) in the ipt approximation , we arrive at the following relation between the coefficients @xmath292 and @xmath302 , @xmath303 ^ 2}{\operatorname{\mathop{\mathrm{im}}}\sqrt{iy(\epsilon)}}\approx 1.74 .
\nonumber\ ] ] now we turn to evaluation of the quasiparticle current @xmath102 and the interference current @xmath304 .
these currents are proportional to small applied voltage , and therefore they can be calculated within the equilibrium approximation for the quasiparticle distribution neglecting small nonequilibrium correction .
the contribution of the trapped quasiparticles is small , because of the small occupied phase volume , and can be neglected .
moreover , in the limit of weak relaxation , this contribution identically turns to zero because of energy - independent distribution @xmath227 . within the bcs approximation for the spectral functions @xmath305 and @xmath306 , the integrals in eqs .
( [ i2 ] ) and logarithmically diverge at the gap edge .
the effect of the phase difference on the quasiparticle spectrum eliminates this divergence and effectively cuts the spectral functions at the value @xmath307 [ see comments to eq .
( [ eqipt ] ) ] , which is equivalent to the effective cut - off in energy , @xmath308 . using the ipt expressions for the spectral functions at @xmath309 ,
we obtain to the main order in @xmath44 , @xmath310 , \label{resi2 } \\
i_3(t ) & = \frac{v}{r } \bigl [ f_1 \ln \sqrt{a\bigl/{\gamma^{4/5}(t ) } } -f_2 \bigr ] , \label{resi3}\end{aligned}\ ] ] where @xmath311 , \nonumber \\ \nonumber \ln a & = \int_0^\infty d\epsilon \bigl[\operatorname{\mathop{\mathrm{re}}}^2 y(\epsilon ) - \frac{\theta(\epsilon-1)}{\epsilon}\bigr]+ \frac{2}{5 }
\ln 2 \approx 0.13.\end{aligned}\ ] ] we note that within the tunnel model,@xcite the role of the cut - off factor is played by the quantity @xmath77 , which enters large logarithmic factor in the expression for the current . in our case , according to the adiabaticity criterion , the effective cut - off factor , @xmath312 , is much larger . from this
we conclude that within the region of applicability of the adiabatic approach , @xmath313 , the quasiparticle and the interference currents are logarithmically smaller as compared to the results of the tunnel model . while the voltage increases and exceeds the depth of the andreev band , @xmath314 , the tunnel model approximation for the currents @xmath102 and @xmath104 becomes valid . due to the presence of a time - dependent factor @xmath115 in the logarithmic terms , the currents @xmath102 and @xmath104 oscillate in time and exhibit logarithmic singularities when @xmath115 turns to zero , i.e. , at @xmath315 . in the vicinity of these points ,
the andreev band vanishes , which violates the condition of adiabaticity .
following remarks to the equations and , we can semi - quantitatively describe the whole shape of the oscillations of the quasiparticle and interference currents by adding a small regularization term @xmath316 to the denominator of the logarithmic argument in eq .
( [ resi3 ] ) . by averaging eq .
( [ resi2 ] ) over time , we obtain the dissipative part of the current @xmath317 , @xmath318 , \\
\nonumber \ln b & = \ln a + \int_0^\pi \frac{d\tau}{\pi } \ln \sin^{-8/5}\tau = 1.24 , \\
\nonumber f_3(t ) & = 1- n_{\textit{eq}}(\delta)-f_2(t).\end{aligned}\ ] ] the interference current also has a dc component due to the oscillations in @xmath104 , though its magnitude , @xmath319 is small with respect to the logarithmic term in @xmath320 .
the sum of the quasiparticle and interference dc currents is given by expression , @xmath321.\end{aligned}\ ] ] the temperature dependencies of the functions @xmath322 , @xmath323 and @xmath324 are shown in fig .
analyzing different calculated current components , we conclude that the most significant deviation from the results of the tunnel model comes from the nonequilibrium josephson current , while the quasiparticle and interference currents undergo minor ( logarithmic ) changes .
there are the two features to be mentioned .
the first is a non - monotonous voltage dependence of the current of trapped quasiparticles described by the relaxation factor @xmath281 [ see eqs .
( [ ieps ] ) and ] , which has a maximum at @xmath325 . due to rather large value of the parameter @xmath326 in most superconductors
, this current may exceed linear contributions of the quasiparticle and interference currents at moderately small values of the transparency parameter @xmath44 . in this case
, the @xmath327-@xmath58 characteristic exhibits an @xmath328-like feature , as shown in fig .
[ discurr](a ) , and the linear conductance at small bias considerably ( by the factor @xmath329 ) exceeds the tunnel model conductance , which is relevant for larger bias [ zero bias conductance peak , fig .
[ discurr](b ) ] .
the second feature is a nonlinear , proportional to @xmath1 , contribution to the current - voltage characteristics produced by the travelling quasiparticles .
we note , however , that because of small numerical value of the constant @xmath330 in eq .
( [ < ij > ] ) , the crossover of @xmath331 to nonlinear behavior for reasonable values of @xmath44 actually occurs at very small voltage .
the reason for this is a numerically small magnitude of the supercurrent spectral density @xmath332 above the bulk gap edge , as is obvious from fig .
[ spectral](a ) .
physically , this is due to a rapid decrease of the probability of the `` over - the - barrier '' andreev reflection ( reflection at the energy outside the energy gap ) . as the result
, the density of nonequilibrium quasiparticles produced by the oscillations of @xmath174 at @xmath333 appears to be small as compared with the `` natural '' width @xmath334 of the perturbed spectral region , which results in smaller contribution of the travelling quasiparticles compared to the trapped ones . in order to estimate the characteristic parameters of the junction for the transparency parameter @xmath157 accepted in our numerical calculations
, we will assume the junction area to be @xmath335 nm and the thickness of the leads as well as the elastic scattering length to be @xmath336 nm .
for al leads , this results in the sheet resistance @xmath337 @xmath338 and @xmath339 @xmath338 at @xmath340 nm .
then , according to eq .
( [ w ] ) , the tunneling probability , the junction resistance and the critical current approach the values @xmath341 , @xmath342 @xmath338 , and @xmath343 @xmath344a , respectively .
thus , the characteristic voltage region , at which the nonequilibrium dc current dominates , is of the order of several microvolts .
= 8.5 cm it follows from the presented estimates and also from fig .
[ discurr ] that the nonequilibrium effect appears at rather small applied voltage , which makes it difficult to observe in practice because of the jumps to the josephson branch . to facilitate the observation ,
the net josephson current must be suppressed by applying magnetic field or using the dc squid setup .
remarkably , the effect survives even in the absence of the net josephson current : local josephson currents flowing through different junctions ( or parts of the junction ) generate dissipative dc current flowing in the same direction , which is determined by the applied voltage ( i.e. , time derivative of the phase ) rather than the local values of the phase difference .
indeed , let us consider two junctions connected in parallel ( dc squid ) and apply half - integer magnetic flux .
then the phase differences at the junctions will be shifted by @xmath345 .
this will lead to different signs of the current spectral density in eq .
( [ ieps ] ) for different junctions , however , the nonequilibrium distribution function will also change the sign , which can be proven by direct calculation .
actually , the equality of dc currents in both junctions follows from a simple fact that the constant phase shift is equivalent to the change of the time reference point , which obviously can not affect the value of the time - averaged ( dc ) current .
thus , the net dc current will double while the ac current disappears .
such situation resembles the case of a long sns junction,@xcite where the josephson effect is suppressed due to decay of the superconducting correlations inside the normal part of the junction , and the nonequilibrium effect can be observed at very small applied voltage .
it is instructive to compare the mechanism of nonequilibrium tunnel current in diffusive junctions with mesoscopic picture of superconductive tunneling in point contacts given by the mar theory.@xcite it has been already mentioned in the introduction , that the andreev band is the qualitative analog of the andreev bound levels in point contacts , the both are formed by the same physical mechanism
andreev reflection by the jump of the superconducting phase at the junction .
the time oscillations of the andreev levels in voltage biased point contacts , and the quasiparticle exchange between the levels and the continuum is the adiabatic description in the time domain of the coherent mar.@xcite the latter also include the quasiparticle transitions across the energy gap , which in the real - time picture correspond to the landau - zener tunneling between the andreev levels ( bands).@xcite such tunneling provides the flow of quasiparticles along the energy axis ( spectral flow ) , proportional to the gradient of the distribution function in the energy space , @xmath346 , within the subgap region , @xmath347 .
these processes are weak at small voltage compared to the continuum - bound level exchange , and they are neglected within the adiabatic approximation adopted in the present paper . for this reason , the spectral flow of the trapped quasiparticles is blocked , which results in a flat ( energy - independent ) distribution of these quasiparticles in the absence of inelastic relaxation . however , there is considerable quantitative difference between the point contacts and diffusive junctions : in the point contacts , most of the quasiparticles reflected by the tunnel barrier escape to the reservoir , while in the long diffusive lead the quasiparticles multiply collide with the barrier due to the impurity backscattering .
this essentially increases the probability of the coherent tunneling and andreev reflection@xcite and , correspondingly , enhances the effect of the phase difference on the junction spectrum . as the result , the depth of the andreev band in our long - arm geometry , @xmath348 in eq .
( [ depth ] ) , considerably exceeds the depth @xmath349 of the andreev level in a point contact with comparable transparency .
it is interesting to mention that generally the role of the junction transparency in tunnel junctions with diffusive electrodes is played by the parameter @xmath44 rather than the barrier transparency @xmath6.@xcite indeed , comparison of the right- and left - hand sides of eqs .
( [ boundary2 ] ) shows that as long as the ` natural ' scale of the currents @xmath350 and @xmath351 is proportional to the gradients of the green s and distribution functions in the vicinity of the barrier , @xmath352 , it is the magnitude of @xmath44 ( not @xmath6 itself ) which determines the effective barrier strength . for example , at large @xmath353 , the commutators in eqs .
( [ boundary2 ] ) are to be small , which implies continuity of all functions at the barrier . from this
we conclude that at @xmath353 , the phase and voltage are continuously distributed along the whole structure , and the barrier does not affect the current transport even if its transparency is small , @xmath354 , provided @xmath355 . in other words , at large @xmath44
, the critical current of the tunnel junction , formally estimated by the ambegaokar - baratoff formula for a low - transparent junction,@xcite exceeds the critical current of diffusive superconducting leads .
physically , this effective ` blooming ' of an opaque barrier results from the multiple coherent backscattering of quasiparticles by impurities mentioned above.@xcite
we have studied the current transport through the voltage biased tunnel junction with long diffusive superconducting leads at low applied voltage , @xmath0 , and finite temperatures .
in contrast to the tunnel model,@xcite we consider the nonequilibrium effects in the spectrum and distribution of quasiparticles in the vicinity of the barrier . using the small value of the josephson frequency with respect to other characteristic energies of the problem , we apply a quasi - adiabatic approach to the analysis of the time evolution of quasiparticles . within such approach
, we obtain a physically transparent picture of the quasiparticle spectrum adiabatically following the time - dependent difference of the superconducting phases .
this results in local oscillations of the spectrum of travelling quasiparticles ( @xmath177 ) and formation of oscillating andreev bound states ( andreev band ) within the subgap region , @xmath356 , at the distance of the order of the coherence length @xmath14 near the barrier .
the quasiparticles trapped by the andreev band are completely dragged by oscillations of the junction spectrum which reflects complete multiple andreev reflection ( mar ) of the subgap quasiparticles and results in generation of higher odd harmonics of the ac josephson current .
the inelastic relaxation of the trapped quasiparticles produces a non - adiabatic component of their distribution which manifests itself in the nonequilibrium contribution to the dc current . at low enough voltages
, this contribution may considerably exceed the quasiparticle dc current given by the tunnel model ; by this reason , the resulting @xmath327-@xmath58 characteristic shows @xmath328-like feature , with the maximum at @xmath357 .
the travelling quasiparticles also deviate from equilibrium due to partial drag ( which corresponds to the over - the - barrier mar ) confined by their fast diffusion from the barrier .
this results in the additional contribution to the dc current , proportional to @xmath1 .
we note that our approach can be easily extended to the current bias regime ; in this case , arbitrary time - dependent phase must be assumed in the calculation .
effect of travelling quasiparticles is interesting , in principle it dominates at small voltage because of the non - analytical voltage dependence .
unfortunately , this contribution is numerically small in the case of the tunnel barrier considered here , because of the small spectral density of the josephson current above the gap .
however , there are no fundamental reasons to expect this contribution to be small in other junctions
. it would be interesting to find favorable junction geometry .
galaiko and v.s .
shumeiko , fiz .
nizk . temp .
* 1 * , 1253 ( 1975 ) [ sov . j. low temp .
* 1 * , 599 ( 1975 ) ] ; a.yu .
azovskii , e.v .
bezuglyi , and v.p .
galaiko , _ ibid . _ * 10 * , 818 ( 1984 ) [ sov .
j. low temp .
* 10 * , 428 ( 1984 ) ] . | we study nonequilibrium effects in current transport through voltage biased tunnel junction with long diffusive superconducting leads at low applied voltage , @xmath0 , and finite temperatures . due to a small value of the josephson frequency ,
the quasiparticle spectrum adiabatically follows the time evolution of the superconducting phase difference , which results in the formation of oscillating bound states in the vicinity of the tunnel junction ( andreev band ) .
the quasiparticles trapped by the andreev band generate higher even harmonics of the josephson ac current , and also , in the presence of inelastic scattering , a nonequilibrium dc current , which may considerably exceed the dc quasiparticle current given by the tunnel model .
the distribution of travelling quasiparticles also deviates from the equilibrium due to the spectrum oscillations , which results in an additional contribution to the dc current , proportional to @xmath1 . |
for sn@xmath1p@xmath1s@xmath2 ferroelectrics , as follows from the first principles calculations @xcite , the spontaneous polarization is determined by three - well local potential for the order parameter fluctuations .
continuous phase transition can be described as second order jahn - teller ( sojt ) effect which is related to the stereoactivity of sn@xmath6 cations placed inside of sulfur ions polyhedron @xcite . for the sn@xmath1p@xmath1s@xmath2 crystal lattice ( fig .
[ fig1 ] ) in mechanism of the ferroelectricity , the hybridization of sn 5s orbitals and s 3p orbitals is a main driving force .
thermodynamics of such transition can be described in well - known blume - emery - griffith ( beg ) model @xcite with two order parameters ( dipolar and quadrupolar ) and three possible values of pseudospin : -1 , 0 , + 1 .
the most important future of beg model is a presence of tricritical point ( tcp ) on states diagram @xcite while the decrease of temperature of phase transition by pressure or by alloying in mixed crystals , the transition becomes of first order . indeed , below tricritical `` waterline temperature '' about 220 - 240 k and at compression above 0.6 gpa , the ferroelectric transition in sn@xmath1p@xmath1s@xmath2 evolves to first order @xcite .
also , at tin by lead substitution in the ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals , for y@xmath7 and below 220 k , the hysteresis which is related to the paraelectric and ferroelectric phase coexistence appears that give evidence about a discontinuous character of the phase transition @xcite .
one more future of discussed @xmath8 and @xmath9 phase diagrams is finite range of the ferroelectric phase existence . under compression ,
the temperature of ferroelectric transition decreases to @xmath10 k at 1.2 gpa in sn@xmath1p@xmath1s@xmath2 crystal @xcite .
the observed @xmath11 dependence at linear extrapolation reaches 0 k near the pressure of 1.5 gpa . for the ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals the paraelectric phase becomes stable till 0 k above [email protected] @xcite .
comparison of @xmath11 and @xmath13 dependencies ( fig .
[ fig2 ] ) shows , that stability of paraelectric state in pb@xmath1p@xmath1s@xmath2 compound can be similar to behavior of sn@xmath1p@xmath1s@xmath2 crystal under a pressure of 2.2 gpa . with increasing pressure the elastic energy increases what make unfavorable the sojt effect and ferroelectric state disappears in sn@xmath1p@xmath1s@xmath2 compound . at transition from sn@xmath1p@xmath1s@xmath2 to more ionic pb@xmath1p@xmath1s@xmath2 compound at normal pressure ,
the sojt effect weakens because pb 6s orbitals have energy about 1 ev lower than in the case of sn 5s orbitals .
this rise of energy distance between s 3p orbitals and pb 6s orbitals determines suppression of pb@xmath6 cations stereoactivity @xcite .
p@xmath1s@xmath2 type crystals in paraelectric ( p2@xmath14/c ) phase with shown shifts of cation atoms in ferroelectric ( pc ) phase.,width=340 ] so , according to the experimental data @xcite , the paraelectric ground state can be stable at @xmath15 gpa for sn@xmath1p@xmath1s@xmath2 , and at @xmath16 and normal pressure for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals . at 0 k ,
the quantum critical point ( qcp ) can be reached by variation of pressure or composition @xcite .
early the qcp have been obviously found in the quantum paraelectric srtio@xmath3 at increase of @xmath17o isotope concentration only @xcite .
here the ferroelectric phase appears almost continuously at near 45% content of @xmath17o , and for srti@xmath17o@xmath3 the ferroelectric phase appears below t@xmath18 k. at cooling to qcp the reciprocal dielectric susceptibility is proportional to @xmath19 , what is the most important criterion of the quantum critical behavior of susceptibility .
but for the most experimental observations , the ferroelectric or magnetic ordering appears by first order transition and @xmath19 behavior of the related susceptibilities is observed only in a some finite temperature interval .
such transformation of qcp into the first order phase transition at @xmath20 k can be explained by an interaction between different order parameters that break different symmetry elements of paraelectric phase @xcite . for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals with compositions @xmath21 and @xmath22 , which are closed to the transition at zero temperature from polar phase ( [email protected] ) to paraelectric one ( [email protected] ) , the dielectric susceptibility demonstrates the quantum critical behavior @xmath25 in vicinity of the first order transitions with @xmath26 k and 20 k , respectively @xcite .
it is interesting to understand the nature of quantum paraelectric state at high pressures in sn@xmath1p@xmath1s@xmath2 compound or at high lead concentration in ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals .
appearance of spontaneous polarization at low temperatures can be mostly related to electronic correlations .
early in the extended falikov - kimbal ( efk ) model it was shown @xcite , that hybridization between orbitals of localized carriers , at the top of valence band , and itinerant electrons , at the bottom of conduction band , can destroy the crystal lattice symmetry center . in this case
the spontaneous polarization in the ground state appears together with excitonic condensate @xcite .
the efk model can be related to hubbard model @xcite which also take into account the spins of fermions .
it is interesting that the last one is usually converted into anisotropic xyz heisenberg model , and for special set of parameters into isotropic heisenberg xxz model @xcite .
it should be noted that isotropic heisenberg model can be equivalent to beg model @xcite , and in fermionic presentation the beg model is also related to efk model @xcite . up to now nobody found experimental evidences about electronic ferroelectricity . a long history of so - called vibronic theory of ferroelectric phase transitions @xcite stimulates searching of the electronic origin of crystal lattices spontaneous polarization .
the vibronic models are mostly connected to jahn - teller effect @xcite and they are based on electron - phonon interaction
. obviously the jahn - teller effect can be leading at description of compounds with ferroelectric phase transitions at high temperatures . both , sojt effect and models of electronic correlations ,
should be obviously involved for description of ferroelectric systems with three - well local potential . for this case of two order parameters ( dipolar and quadrupolar ) , at lowering of second order transition temperature ( by compression or by variation in chemical composition ) the tcp is reached .
further the line of first order transitions goes down to 0 k. here the quantum paraelectric state is evidently realized and possibility of electronic correlation weak effects can be checked .
of course , a some phonon contribution to the electronic correlations is expected , what is accounted in the holshtein - hubbard type models @xcite .
the relaxations of paired in a real space anderson s electrons @xcite or phononic kondo screening @xcite can also be involved in some approaching for description of phonon - like excitations in the low temperature quantum paraelectric state .
p@xmath1s@xmath2 crystal and temperature - composition phase diagram for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals .
the tricritical points are located near 220 k for pressure @xmath27 gpa @xcite or for concentration @xmath28 @xcite . for pb@xmath1p@xmath1s@xmath2 crystal
the calculated temperature - pressure phase diagram is shown .
the metastable region in ground state ( @xmath20 k ) with possible coexistence of paraelectric and ferroelectric phases is shaded .
, width=453 ] in this paper we present the results of dielectric susceptibility and raman scattering studies of pb@xmath1p@xmath1s@xmath2 crystals .
some growth of dielectric susceptibility and softening of the lowest energy phonon optic mode are observed at cooling from room temperature till 20 k in the paraelectric phase of pb@xmath1p@xmath1s@xmath2 .
experimental data on @xmath8 phase diagram of sn@xmath1p@xmath1s@xmath2 and on @xmath9 phase diagram of ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 are compared and involved for explanation of unusual behavior of paraelectric phase of pb@xmath1p@xmath1s@xmath2 crystal .
the electron and phonon spectra of pb@xmath1p@xmath1s@xmath2 crystal are calculated in gga approximation of density functional theory and compared with experimental data . for this crystal
the transition from paraelectric state into ferroelectric one under negative pressure is modeled by first - principles calculations .
taking into account the ideas about valence skipping @xcite and charge disproportionation @xcite , the electronic origin of possible spontaneous polarization of pb@xmath1p@xmath1s@xmath2 crystal lattice is analyzed in the frame of efk model @xcite .
also the approximations of hubbard model and related beg model @xcite are used for consideration of simplified picture of half - filled lattice with sites that incorporate the snps@xmath3 atomic groups with 27 valence electrons . the minimal energetic model
is proposed that considers the 26 valence electrons on hybridized bonding orbitals of snps@xmath3 structure units as core and only one electron is considered at every site .
the @xmath0 charge disproportionation process is energetically favorable for p@xmath29 phosphorus ions valence skipping and @xmath30 recharging .
the on - site coulomb repulsion @xmath31 acts in opposite way to this disproportionation process whith energy gain @xmath32 .
the orbitals hybridization complicates this picture , but also induce acentric state .
we estimate the energy of disproportionation which is main origin of acentric ground state at condition of suppressed sojt effect . at @xmath33
ev , where @xmath34 is a energy gap , the efk model was used for estimation of electronic contribution to dielectric susceptibility .
the calculated electron spectra are compared with parameters of minimal beg like energetic model that is derived from hubbard hamiltonian @xcite and agrees with observed @xmath8 and @xmath9 diagrams of ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 ferroelectrics .
finally , the quantum anharmonic oscillators models , that is based on statistics of phonon - like excitations in the lattice with three - well local potential , is proposed for description of @xmath8 and @xmath9 state diagrams of ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 ferroelectrics and explanation of their dielectric susceptibility growth at cooling in the quantum paraelectric state .
the dielectric susceptibility and raman scattering were investigated for pb@xmath1p@xmath1s@xmath2 crystals that was grown by vapor transport method .
temperature dependence of dielectric susceptibility was investigated by bridge scheme at frequencies range @xmath35 hz @xcite .
the raman scattering was excited by he - ne laser and analyzed by dfs-24 spectrometer with 1 @xmath36 frequency resolution .
the spectra were fitted by voight shape spectral counters . the temperature dependence of real part of dielectric susceptibility is shown at fig .
[ fig3 ] for different frequencies .
the susceptibility increases about ten percepts at cooling in the region of three hundred kelvins without frequency dispersion in the studied diapason .
p@xmath1s@xmath2 dielectric susceptibility at different frequencies.,width=377 ] according to temperature evolution of raman spectra ( fig .
[ fig4 ] ) , the lowest energy optic mode softens about 2 @xmath36 at cooling from room temperature to 79 k. at temperature decrease , the low frequency spectral tail and damping of optical phonons growth also .
such behavior coincides with earlier found @xcite increase of pb@xmath1p@xmath1s@xmath2 crystal lattice anharmonicity according to the rise of gruneisen parameter at cooling on the data of brillouin scattering by acoustic phonons .
the observed rise of dielectric susceptibility about 10% at cooling correlates with a lowering of optic mode frequency by 1 @xmath36 .
p@xmath1s@xmath2 crystals at 293 k ( a ) and 79 k ( b ) , and temperature dependence of spectral lines frequencies ( c ) and damping ( d).,title="fig:",width=302 ] p@xmath1s@xmath2 crystals at 293 k ( a ) and 79 k ( b ) , and temperature dependence of spectral lines frequencies ( c ) and damping ( d).,title="fig:",width=302 ] p@xmath1s@xmath2 crystals at 293 k ( a ) and 79 k ( b ) , and temperature dependence of spectral lines frequencies ( c ) and damping ( d).,title="fig:",width=302 ] p@xmath1s@xmath2 crystals at 293 k ( a ) and 79 k ( b ) , and temperature dependence of spectral lines frequencies ( c ) and damping ( d).,title="fig:",width=302 ]
for characterization of pb@xmath1p@xmath1s@xmath2 paraelectric phase and to search a possible origin of its instability , the electron and phonon spectra were calculated in gga approach of dft .
the monoclinic elementary cell ( fig . [ fig1 ] ) contains two formula units and belong to p2@xmath14/c space group @xcite .
structure of valence zone ( fig .
[ fig5 ] ) correlates with earlier @xcite described for sn@xmath1p@xmath1s@xmath2 analog .
but here the contribution of pb@xmath6 cations 6s orbitals into the density of states at the top of valence band is smaller as in the case of sn@xmath6 5s orbitals presence .
this fact reflects smaller stereoactivity of pb@xmath6 cations in more ionic pb@xmath1p@xmath1s@xmath2 lattice .
such peculiarity is evidently determined by bigger energy distance between sulfur 3p orbitals and lead 6s orbitals .
indeed , the pb 6s orbitals are placed mostly below -8 ev by 1 ev lower in compare with energetic position of sn 5s orbitals in sn@xmath1p@xmath1s@xmath2 crystal .
it is important to mention that for both edges of valence band and conductivity band , the orbitals of sulfur , phosphorous and lead atoms are presented .
obviously the electronic structure of pb@xmath1p@xmath1s@xmath2 compound can be considered as bounding and antibounding counterparts of pbps@xmath3 atomic groups orbitals .
the electron spectra calculations in gga approach are in good enough agreement for the energy gap value - calculated one ( about 2.2 ev ) is closed to 2.45 ev gap according to optic absorption measurements at 4.2 k @xcite . p@xmath1s@xmath2 crystal .
, title="fig:",width=377 ] + p@xmath1s@xmath2 crystal .
, title="fig:",width=377 ] in gga approach the phonon spectra for pb@xmath1p@xmath1s@xmath2 crystal lattice were calculated also .
the phonon branches and partial phonon densities of states are presented at fig .
long wave phonon frequencies are in good agreement with experimental data what give evidence about adequate gga approach for the energetic spectra analysis . below 70 @xmath36 , the translations of pb ions relatively to p@xmath1s@xmath2 structural groups mostly contribute into phonon eigenvectors as follows from the partial phonon densities of states . above 150 @xmath36 the internal vibrations of ( p@xmath1s@xmath2)@xmath37 anions dominate and in higher frequency interval 530 - 590 @xmath36 mostly phosphorous atoms participate in the phonon eigenvectors . the lo - to splitting has the biggest values for the translational lattice polar modes with frequencies 43 , 58 and 70 @xmath36 .
it is remarkable that big lo - to splitting also was found for the ( p@xmath1s@xmath2)@xmath37 internal valence vibrations with frequencies 563 , 566 and 578 @xmath36 .
adequate calculations of electronic and phonon spectra of pb@xmath1p@xmath1s@xmath2 crystal at normal pressure permit to suppose a possibility for obtaining of reasonable calculation results at negative pressure .
it was found that the transition into ferroelectric phase occurs near -1.95 gpa ( fig .
[ fig7 ] ) .
it is important that the acentric deviation appears already at pressure about -1.7 gpa .
the ground state of pb@xmath1p@xmath1s@xmath2 crystal at normal pressure is similar to the sn@xmath1p@xmath1s@xmath2 ground state at 2.2 gpa ( fig .
[ fig2 ] ) .
some similarity of sn@xmath1p@xmath1s@xmath2 compound @xmath8 diagram at pressures @xmath38 gpa with calculated states diagram for pb@xmath1p@xmath1s@xmath2 crystal in pressure range from -2 to 0 gpa permits us to suppose that temperature behavior of pb@xmath1p@xmath1s@xmath2 crystal ( at normal pressure ) could be similar to sn@xmath1p@xmath1s@xmath2 temperature behavior at 2.2 gpa .
this behavior can be determined by some polar fluctuations , like in quadrupole phase according to mc simulations @xcite , and some growth of dielectric susceptibility at cooling can be expected .
let s try to found some explanation for quantum paraelectric behavior in pb@xmath1p@xmath1s@xmath2 crystal .
p@xmath1s@xmath2 crystal lattice.,width=377 ]
earlier @xcite , for sn@xmath1p@xmath1s@xmath2 ferroelectrics the stereoactivity of sn@xmath6 cations , which are placed inside sulfur octahedron , was only considered as origin of local dipoles that exist already in paraelectric phase , because tin cations occupy general position in the monoclinic elementary cell .
this stereoactivity is determined by covalency between tin atomic orbitals and molecular orbitals of p@xmath1s@xmath2 groups , and can be considered as second order jahn - teller effect , which involve phosphorous orbitals also . at weaking of sojt effect by pressure , or at tin by lead substitution , another origins of crystal acentricity can be observed .
the first of them is disproportionation of p@xmath29 cations : @xmath0 .
this process is also related to recharging of surrounding sulfur anions that is accounted by polarizability of nearest space @xcite .
the energy of disproportionation can be determined as @xmath39 .
here @xmath40 is the difference in ionization potentials for p@xmath41 and p@xmath29 ions in vacuum .
the polarization energy is determined by born formula @xcite @xmath42 , where @xmath43 . here
the phenomenological parameters are included
anions and cations radii @xmath44 and @xmath45 , respectively , and dielectric constants @xmath46 and @xmath47 .
the value @xmath48 ev was estimated by using the ionization energies @xmath49 ev for p@xmath41 cations and @xmath50 ev for p@xmath29 cations @xcite .
the value @xmath51 was found for next set of parameters : s@xmath52 anion radius @xmath53 , according s - p interatomic distance @xcite , and p@xmath29 cation radius @xmath54 ; dielectric constants @xmath55 and @xmath56 @xcite . at such parameters , the polarization energy @xmath57 ev
have been found and , sequently , @xmath58 ev . negative value of @xmath59 reflects the valence skipping effect that favors 3s@xmath60 and 3s@xmath61 electronic configuration of p@xmath41 and p@xmath62 cations instead of 3s@xmath63 configuration of p@xmath29 ions .
the phosphorous ions charge disproportionation , or valence ordering process , can be viewed as a lattice of anderson electron pairs @xcite which is stabilized by polarizing of surrounding sulfur polyhedrons .
the energy to displace each s atom by @xmath64 can be written as @xcite @xmath65 where @xmath66 is the charge difference between the two nearest neighbor phosphorous atoms , @xmath67 and @xmath68 are coupling constants .
the hubbard type hamiltonian @xcite can be applied for description of electrons hopping in band with contribution of phosphorous orbitals .
such hamiltonian contains the intra - site and inter - site coulomb interactions .
the first one is determined by @xmath31 constant . the inter - site interaction for simplicity
can be presented by the short - range interaction @xmath69 only , where @xmath70 is madelung constant , @xmath71 is lattice parameter , @xmath72 is charge of electron and @xmath73 is dielectric permittivity . at condition of small band width limit in the spin presentation the hubbard model
can be reflected onto the beg model @xcite : @xmath74 here parameters are derived from hubbard model of anderson s electron pairs
@xmath75 and @xmath76 .
pseudospin variable @xmath77 has values + 1 , 0 , -1 and can be are related to p@xmath62 , p@xmath29 and p@xmath41 states of phosphorous cations .
general view of calculated for beg model phase diagram ( fig .
[ fig8 ] ) correlates with experimental observations .
indeed , under compression the lattice stiffness constant @xmath68 increases , lattice period @xmath71 lowers and following the on - site energy @xmath78 rises and the intersite interaction energy @xmath79 remains almost unchanged .
calculated in the mean - field approximation on the beg model ( 1 ) .
solid lines denote second order and dashed lines first order transitions and the dot , a tricritical point @xcite.,width=377 ] at tin by lead substitution , the stiffness decrease a little @xcite , but the lattice period demonstrates some rise @xcite .
so again , the intercell interaction @xmath79 remains almost unchanged .
but in lead compound the ionicity is bigger than in the case of sn@xmath1p@xmath1s@xmath2 , and the onsite coulomb repulsion @xmath31 is obviously higher . at this
the on - site energy @xmath78 also growth at sn by pb substitution . such influence of chemical composition on the system ground state qualitatively agrees with derived in hubbard approximation @xcite phase diagram where phase transition from paraelectric phase into state with dipole ordering is governed by difference between energies of filled orbitals of two type cations in the crystal lattice . in our case , according to gga calculations ( fig .
[ fig5 ] ) , the difference between p 3p orbitals energy and sn 5s or pb 6s orbitals energies increases nearly 1 ev at transition from sn@xmath1p@xmath1s@xmath2 to pb@xmath1p@xmath1s@xmath2 .
evidently the growth of ionicity stabilizes the paraelectric phase what agrees with experimental observation .
finally , we can estimate some parameters by using the following characteristics : the second order phase transition temperature ( @xmath80 k ) for sn@xmath1p@xmath1s@xmath2 crystal , the coordinates of tricritical points ( @xmath81 k ) on @xmath9 and @xmath8 diagrams , the composition @xmath82 or pressure @xmath83 gpa at which the phase transition temperature goes down to zero . by comparing the experimental diagram ( fig .
[ fig1 ] ) with calculated one ( fig .
[ fig8 ] ) , it can be found that @xmath84 k for considered ferroelectrics . at almost constant value of intersite interaction energy ( parameter @xmath79 )
, the on - site energy parameter @xmath78 mostly determines the global temperature - pressure - composition phase diagram .
the @xmath78 parameter has small value and is governed by balance of two significant characteristics the on - site coulomb repulsion @xmath31 and disproportionation energy @xmath32 .
the first one have the value above 2 ev according to _ ab - initio _ gga+u calculations @xcite that explain observed optic gap width @xmath85 ev for sn@xmath1p@xmath1s@xmath2 crystal .
the disproportionation energy @xmath32 can be compared with ratio @xmath86 and equals about -2 ev .
obviously , for sn@xmath1p@xmath1s@xmath2 crystal a some small value of @xmath78 can be found ( fig .
[ fig8 ] ) and @xmath87 for some pressure @xmath88 or lead content @xmath89 . at @xmath87 or for condition @xmath90 ,
the valence fluctuations are strongly developed , and possibly , they are still enough strong at rise of @xmath78 value to @xmath91 and phase transition temperature @xmath92 going down to 0 k. so , the valence or charge fluctuations can contribute to growth of dielectric susceptibility for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 crystals with @xmath16 and for pb@xmath1p@xmath1s@xmath2 compound .
recent calculations in hubbard model show @xcite that in paraelectric ground state , the charge density at different cations ( phosphorous and lead in our case ) strongly depends on temperature .
this peculiarity can be evidently related to the temperature dependence of dielectric susceptibility .
fluctuations of anderson s 2e electron pairs as origin of dielectric susceptibility growth at cooling in wide temperature interval can be compared with kondo screening phenomena @xcite . at incoherent 2e pairs
fluctuations the susceptibility increases with cooling as @xmath93 , but below some temperature the coherence of electron pairs fluctuations can appear and susceptibility reaches some constant value .
the model of kondo effect with involving of phonon channel definitely predicts @xcite a rise of dielectric susceptibility at the system cooling . at low enough temperatures , near 0 k
, electronic correlations themselves can induce the system acentricity .
the electronic ferroelectricity has been predicted @xcite as result of hybridization of itinerant electron wave function with hole at the top of valence band and appearance of excitonic condensate .
the excitonic order parameter correlates with acentric space distribution of the charge density , as it is follows from the efk model . for ground state in paraelectric phase at approaching to transition into polar phase , by variation of itinerant electrons energy level @xmath94
, the dielectric susceptibility is expected to growth according relation @xcite : @xmath95 where @xmath96 is the interband dipole element , @xmath97 is the volume of unit cell , @xmath98 is the number of sites , @xmath99 is the bandwidth .
it was taken @xmath100 ev and @xmath101 ev according to estimations which were made from gga calculations ( fig .
[ fig5 ] ) .
as was mentioned above , for sn@xmath1p@xmath1s@xmath2 crystals at normal pressure very delicate energetic balance @xmath90 serves for development of charge fluctuations .
moreover , the energy gap between valence and conductivity gap @xmath102 ev is also comparable with estimates values @xmath103 ev for @xmath31 and @xmath104 .
such conditions permit expect possibility of electronic origin for increased at cooling dielectric susceptibility in paraelectric phase .
estimation of dielectric susceptibility for pb@xmath1p@xmath1s@xmath2 at @xmath20 k within efk model ( 2 ) gives value @xmath105 for the interband transition dipole moment @xmath106 . at heating
the susceptibility will decrease because of thermal disordering of the local electric dipoles .
now we propose to use the aqo model @xcite that consider phonon - like bosonic excitations .
these excitations can be regarded as incorporation of electron - phonon interaction in generalized holshtein - hubbard model @xcite or in bose - habbard model @xcite . in the aqo model ,
crystal is representing as one dimensional system of anharmonic oscillators which interact via quadratic interaction term .
so , hamiltonian of such a system is @xmath107 where @xmath108 and @xmath109 are operators of kinetic and potential energy , respectively , @xmath110 is a @xmath111 oscillator s displacement , @xmath112 are coupling constants between @xmath111 and @xmath113 oscillators . in the mean - field approach , the last term of ( [ eq1 ] )
can be replaced by another one - @xmath114 where @xmath115 is average position of all other oscillators in the lattice . in this case , the hamiltonian ( [ eq1 ] ) can be represented as a sum of independent single - particle hamiltonians : @xmath116 such effective particle is an oscillator under the influence of linear symmetry - breaking field which is calculated self - consistently . solving schrdinger equation with hamiltonian ( [ eq2 ] ) one obtain a set of eigen energies @xmath117 of levels and its wave functions @xmath118 which are used for self - consistent calculation of average expectation value @xmath115 . in our calculations
we used matrix representation form for position and momentum operators to solve schrdinger equation @xcite .
p@xmath1s@xmath2crystal at different pressures @xcite . ,
width=340 ] it should be noted that temperature enters in hamiltonian ( [ eq2 ] ) indirectly through boltzmann distribution for the occupation of the levels .
really , from definition of @xmath115 it follows @xmath119 where @xmath120 is the occupation number for @xmath121 level , @xmath122 is a boltzmann constant , @xmath123 is average value of displacement for @xmath121 level .
obviously , in the paraelectirc phase for average desplacement , which is proportional to a order parameter , one obtain @xmath124 .
on the other hand , below phase transition temperature @xmath125 we have @xmath126 .
so , one can model a changes on phase diagram at ionic substitution or pressure by varying coupling constant or shape of potential energy . here
we use a transformation of the local three - well potential @xmath109 ( fig .
[ fig10 ] ) as function of pressure according to the results of @xcite . in present calculations
the mass of oscillator was equal to tin atomic mass .
moreover , it is possible to calculate a dielectric response of such the system of oscillators using following relation @xcite : @xmath127 where @xmath128 , and effective charge of oscillator @xmath129 and its volume @xmath130 can be used as fitting parameters . at pressure @xmath131 gpa ( a ) ; calculated ( continuous line ) and experimental ( points )
temperature dependence of dielectric susceptibility for pb@xmath1p@xmath1s@xmath2 quantum paraelectric ( b).,title="fig : " ] at pressure @xmath131 gpa ( a ) ; calculated ( continuous line ) and experimental ( points ) temperature dependence of dielectric susceptibility for pb@xmath1p@xmath1s@xmath2 quantum paraelectric ( b).,title="fig : " ] examples of the qao model solutions are shown at fig . [ fig11 ] . calculated in this model state diagram
is shown on fig.[fig12 ] .
in addition to the paraelectric and ferroelectric phases , it also contain a coexistence region for stable and metastable solutions .
this diagram reflects the main features of experimental diagram . for sn@xmath1p@xmath1s@xmath2 crystal under compression
, the second order phase transition temperature lowers from 337 k ( at normal pressure ) and the tricritical point is reached near 0.6 gpa and at 220 k. the coexistence region of ferroelectric and metastable solutions is delayed till 1.5 gpa . for higher pressure , the paraelectric phase is stabilized .
the pb@xmath1p@xmath1s@xmath2 compound at normal pressure has quantum paraelectric ground state . by negative pressure below -1.7
gpa the crystal lattice acentricity can be induced .
k ( b ) for the aqo model with paraelectric ( pe ) and ferroelectric ( fe ) phases , and with intermediate region of coexisting stable ferroelectric phase and metastable ( ms ) solutions .
the thermodynamic ways are shown : for sn@xmath1p@xmath1s@xmath2 crystal under compression in the range @xmath132 gpa and for a rise of lead concentration at normal pressure in ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals in the range @xmath133.,title="fig:",width=377 ] k ( b ) for the aqo model with paraelectric ( pe ) and ferroelectric ( fe ) phases , and with intermediate region of coexisting stable ferroelectric phase and metastable ( ms ) solutions .
the thermodynamic ways are shown : for sn@xmath1p@xmath1s@xmath2 crystal under compression in the range @xmath132 gpa and for a rise of lead concentration at normal pressure in ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals in the range @xmath133.,title="fig:",width=377 ] the dielectric susceptibility temperature dependence in quantum paraelectric pb@xmath1p@xmath1s@xmath2 have been calculated in the aqo model and compared with experimental data ( fig .
[ fig11 ] , b ) .
the dielectric susceptibility increase at cooling from room temperature to 20 k obviously reflects strong anharmonicity of the local potential in this compound .
the valence fluctuations play important role in the nature of ferroelectric and quantum paraelectric states in sn(pb)@xmath1p@xmath1s@xmath2 semiconductors .
the charge disproportionation of phosphorous ions @xmath0 can be related to recharging of snps@xmath3 ( or pbps@xmath3 ) structural groups .
this approximation permits to consider a simplified model of the crystal lattice as set of the half - filled sites .
experimental temperature - pressure phase diagrams for sn@xmath1p@xmath1s@xmath2 crystal and temperature - composition one for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals with tricritical point and with decrease of phase transitions lines to 0 k , together with the data about some softening of low energy optic phonons and rise of dielectric susceptibility at cooling in quantum paraelectric state of pb@xmath1p@xmath1s@xmath2 crystal , are analyzed by first principles electron and phonon calculations and compared with electronic correlations models .
the anharmonic quantum oscillators model is developed for description of temperature - pressure - composition phase diagram shape .
temperature dependence of dielectric susceptibility is also described in this model .
the chemical bonds covalence obviously complicates picture of the charge disproportionation which can be presented as @xmath134 , and parameter @xmath64 is interesting for determination at the further development of electronic correlation models .
authors acknowledge prof . ihor v. stasyuk for fruitful discussions . 10 k.z .
rushchanskii , y.m .
vysochanskii , and d. strauch , phys .
99 , 207601 ( 2007 ) .
k. glukhov , k. fedyo , j. banys , and yu . vysochanskii , int . j. mol .
13 , 14356 ( 2012 ) .
m. blume , v.j .
emery , and r.b .
griffiths , phys .
a 4 , 1071 ( 1971 ) .
c. ekiz , m. keskin , and o. yalcin , physica a 293 , 215 ( 2001 ) .
ondrejkovic , m. kempa , y. vysochanskii , p. saint - gregoire , p. bourges , k.z .
rushchanskii , and j. hlinka , phys .
b 86 , 224106 ( 2012 ) .
ondrejkovic , m. guennou , m. kempa , y. vysochanskii , g. garbarino , and j. hlinka , j. phys . : condens .
matter 25 , 115901 ( 2013 ) . k.z .
rushchanskii , r.m .
bilanych , a.a .
molnar , r.m .
yevych , a.a .
kohutych , s.i .
perechinskii , v. samulionis , j. banys , and y.m .
vysochanskii , phys .
status solidi b 253 , 384 ( 2016 ) .
k. moriya , k. iwauchi , m. ushida , a. nakagawa , k. watanabe , s. yano , s. motojima , and y. akagi , j. phys .
64 , 1775 ( 1995 ) .
rowley , l.j .
spalek , r.p .
smith , m.p.m .
dean , m. itoh , j.f .
scott , g.g .
lonzarich , and s.s .
saxena , nature physics 10 , 1 ( 2014 ) .
doi : 10.1038/nphys2924 d. rytz , u.t .
hochli , and h. bilz , phys .
b 22 , 359 ( 1980 ) . l.m .
falicov and j.c .
kimball , phys .
22 , 997 ( 1969 ) .
t. portengen , th .
ostreich , and l.j .
sham , phys . rev .
b 54 , 17452 ( 1996 ) .
t. portengen , th .
ostreich , and l.j .
sham , phys .
76 , 3384 ( 1996 ) .
batista and a.a .
aligia , phys .
92 , 246405 ( 2004 ) .
rice and l. sneddon , phys .
47 , 689 ( 1981 ) .
kristofel and p.i .
konsin , soviet physics uspekhi 19 , 962 ( 1976 ) .
i.b . bersuker and b.g .
vekhter , ferroelectrics 19 , 137 ( 1978 ) .
h. fehske , a.p .
kampf , m. sekania , and g. wellein , eur .
j. b 31 , 11 ( 2003 ) .
michele fabrizio , alexander o. gogolin , and alexander a. nersesyan , phys .
83 , 2014 ( 1999 ) .
anderson , phys .
34 , 953 ( 1975 ) .
a. taraphder and p. coleman , phys .
66 , 2814 ( 1991 ) .
hiroyasu matsuura and kazumasa miyake , journal of the physical society of japan 81 , 113705 ( 2012 ) .
takashi hotta and kazuo ueda , phys .
108 , 247214 ( 2012 ) .
takahiro and takashi hotta , journal of the korean physical society 62 , 1874 ( 2013 ) .
varma , phys .
61 , 2713 ( 1988 ) .
i.a . drabkin , b.ya .
moiges , and s.g .
suprun , fiz .
tela , 27 , 2045 ( 1985 ) ( in russian ) .
rostyslav m. bilanych , ruslan m. yevych , anton a. kohutych , sergij i. perechinskii , ivan m. stoika , and yulian m. vysochanskii , cent .
, 12(9 ) , 611 ( 2014 ) .
r. becker , w. brockner , and h. schafer , z. naturforsch .
38a , 874 ( 1983 ) .
carpentier and r. nitsche , mater .
res . bull . 9 , 401 ( 1974 ) .
agekyan , l.n .
muzyka , fiz .
tela 28 , 3217 ( 1986 ) ( in russian ) .
rushchanskii , a. molnar , r. bilanych , r. yevych , a. kohutych , yu .
vysochanskii , v. samulionis , and j. banys , arxiv:1409.6364v1 [ cond-mat.mtrl-sci ] 22 sep 2014 .
nist online databases , http://srdata.nist.gov/gateway/gateway?dblist=1 y.m .
vysochanskii , t. janssen , r. currat , r. folk , j. banys , j. grigas , and v. samulionis , phase transitions in ferroelectric phosphorous chalcogenide crystals ( vilnius university publishung house , vilnius , 2006 ) , p. 453 .
makoto naka , hitoshi seo , and yukitoshi motome , phys .
116 , 056402 ( 2016 ) .
m. makowska - janusik , t. babuka , k. glukhov , and yu . vysochanskii , journal of alloys and compounds ( 2016 ) ( in press ) . h.j .
bakker , s. hunsche , and h. kurz , phys . rev .
b 48 , 9331 ( 1993 ) .
stasyuk , o.v .
velychko , and o. vorobyov , condensed matter physics 18 , 43004 ( 2015 ) . h.j .
korsch , and m glck , eur .
23 , 413 ( 2002 ) .
vaks , v.m .
galitskii , and a.i .
larkin , sov .
jetp 24 , 1071 ( 1967 ) . | the valence fluctuations which are related to the charge disproportionation of phosphorous ions @xmath0 are the origin of ferroelectric and quantum paraelectric states in sn(pb)@xmath1p@xmath1s@xmath2 semiconductors .
they involve recharging of snps@xmath3 or pbps@xmath3 structural groups which could be represented as half - filled sites in the crystal lattice .
temperature - pressure phase diagram for sn@xmath1p@xmath1s@xmath2 compound and temperature - composition phase diagram for ( pb@xmath4sn@xmath5)@xmath1p@xmath1s@xmath2 mixed crystals , which include tricritical points and where a temperature of phase transitions decrease to 0 k , together with the data about some softening of low energy optic phonons and rise of dielectric susceptibility at cooling in quantum paraelectric state of pb@xmath1p@xmath1s@xmath2 are analyzed by gga electron and phonon calculations and compared with electronic correlations models .
the anharmonic quantum oscillators model is developed for description of phase diagrams and temperature dependence of dielectric susceptibility .
* keywords : * mixed valency , quantum paraelectrics , ferroelectrics |
the characteristics of ultracold bosonic two - body scattering processes provide a key ingredient for modeling and understanding of the properties of degenerate atomic gases .
the dominant s - wave interactions of the two - body collisions can be conveniently modeled by huang - fermi s pseudopotential @xcite . over the last decade
, this pseudopotential has become an essential theoretical tool in particular for analytical studies of ultracold systems , e.g. , see busch et .
@xcite for two atoms in harmonic traps , olshanii for atomic collisions in cylindrical waveguides @xcite , where the so - called confinement - induced resonances ( cirs ) was predicted and experimentally observed in @xcite , or in refs .
@xcite which employed approaches based on the energy dependent scattering lengths , @xmath0 .
the theory of pseudopotentials has been developed further in order to describe scattering processes involving higher - partial waves , i.e. , with relative angular momentum @xmath1 @xcite .
higher - partial wave interactions represent an intriguing perspective in research on ultracold gases and are expected to provide novel many - body phenomena , such as unconventional superfluidity and superconductivity @xcite .
so far , theoretical studies with pseudopotentials for @xmath2 have mainly been used to describe fermionic systems , e.g. , in quasi - one - dimensional @xcite or quasi - two - dimensional @xcite geometries , which have been studied experimentally in @xcite . for the case of bosons
an analytical description of collisions in a planar waveguide has been derived , where the cross - section possesses s- and d - wave resonances @xcite . in harmonic waveguides
higher - partial wave interactions were shown to have important impact on bosonic scattering @xcite through the interference of d - wave resonances ( dwrs ) with the underlying continuum .
a theoretical treatment of d - wave scattering of bosons in the case of quasi-1d geometries is , however , still needed . in this work
we derive an approximative analytical solution for bosonic collisions in a harmonic waveguide , which goes beyond mere s - wave scattering .
our starting - point is to assume that the transversal confinement frequency is weak enough in order to avoid a strong interference between the free space d - wave shape resonance and the s - wave scattering .
the latter permits to model the interatomic potential through a zero - range regularized d - wave pseudopotential suggested in @xcite , which is parametrized by a single parameter , the energy dependent d - wave scattering length ( @xmath3 ) .
the associated regularization operator contains exclusively derivatives .
taking into account the virtual excitations occurring during the scattering process we analytically describe the appearance of the dwr .
an important aspect of our approximative formula for the d - wave scattering amplitude is that one can map the total hamiltonian onto an effective one - dimensional hamiltonian , where the interaction is modeled by a one - dimensional contact potential .
the strength of the contact potential depends on the above - mentioned d - wave scattering amplitude , similar to the case of the s - wave cir @xcite . in order to verify the validity of the approximative analytical results we compare them with corresponding numerical calculations , where the interatomic interaction is modeled by a lennard - jones potential .
a good agreement is obtained only for a sufficiently weak confinement . for strong confinement
the pseudopotential does not capture the s- and d - wave interference and therefore lacks the description of the numerically obtained fano asymmetric lineshape of the transmission coefficient @xcite .
the sectioning of this paper is as follows . in sec .
ii we review the recent developments of pseudopotential theory for higher partial wave interactions . in sec .
iii we describe and discuss the d - wave scattering process .
iv contains a comparison of the analytical results and the numerical calculations and sec .
v provides a brief summary and conclusions .
the pseudopotential approximation to the interatomic interaction is an ubiquitous tool for the theoretical description of ultracold gases : the atomic interactions are modeled in terms of a contact potential .
huang and yang @xcite derived a generalized pseudopotential for all partial waves , which represents a multipolar expansion over delta function contributions .
the huang - yang approach ( hy ) was revised by derevianko @xcite , who corrected an algebraic error for @xmath2 .
however , other derivations have been developed for higher partial wave pseudopotentials , which differ from the hy - approach , by using a more rigorous treatment of the derivatives of the delta - functions in the framework of distribution theory @xcite yielding the following representation : @xmath4 ^ 2}{4\pi(2\ell)!\ell!}g_{\ell}\frac{\delta^{(\ell)}(r)}{r^2}\\ & \times\left[\dfrac{\partial^{2\ell+1}}{\partial { r'}^{2\ell+1 } } { r'}^{\ell+1 } \int d \mathbf{\omega ' } p_{\ell } ( \hat{\mathbf{r } } \cdotp \hat{\mathbf{r } } ' ) \psi(\mathbf r')\right]_{\mathbf{r'}=0 } , \label{1}\end{aligned}\ ] ] where @xmath5 , with @xmath6 defined as the @xmath7-th energy - dependent scattering length @xcite .
here , @xmath8 is the @xmath7-th derivative of the delta function and the integral over the solid angle @xmath9 acts as a projection operator of the total wavefunction on each @xmath10state . for @xmath11 one obtains from eq.([1 ] ) the huang - fermi pseudopotential .
however , the eq.([1 ] ) for @xmath2 differs by a factor @xmath12 from huang and yang s result @xcite , whereas if we substitute in eq.([1 ] ) the relations @xmath13 and @xmath14 we obtain the pseudopotential of derevianko @xcite . additionally , stampfer and wagner @xcite introduce the following pseudopotential for higher partial waves : @xmath15_{\mathbf{r'}=0 } , \label{2}\end{aligned}\ ] ] where the operator @xmath16 , introduced by maxwell @xcite , is obtained from the spherical harmonic polynomial @xmath17 , by replacing the cartesian coordinates @xmath18 with the partial derivatives @xmath19 .
note that eq.([2 ] ) is equal to eq.([1 ] ) as it was proven in @xcite by using the relation @xmath20 the above - mentioned pseudopotentials can be simplified @xcite in order to describe interactions of a single partial wave character .
this simplification is based on the fact that in general the atomic interaction is dominated by the @xmath7-partial wave scattering in the vicinity of the corresponding @xmath7-partial wave resonance .
thus , one can focus on a single partial wave , which yields the following pseudopotential : @xmath21_{\mathbf{r'}=0}. \label{4}\end{aligned}\ ] ] furthermore , the sum w.r.t .
@xmath22 can be simplified with the help of the following summation formula : @xmath23 } c_{k}\left(\bm{\nabla } \cdot \bm{\nabla } ' \right)^{\ell-2k } \nabla^{2k } ( \nabla ' ) ^{2k } , \label{5}\ ] ] where @xmath24 $ ] is the largest , smaller or equal integer of @xmath25 and @xmath26/\left[2^{\ell}k!(\ell - k)!(\ell-2k)!\right]$ ] are the coefficients of the legendre polynomials @xmath27}c_{k}x^{\ell-2k}$ ] .
eqs.([4 ] ) and ( [ 5 ] ) constitute pseudopotentials for any @xmath7-partial wave .
these pseudopotentials are expressed in terms of differential operators which are more convenient in analytical derivations than the pseudopotential with the projection operator ( see eqs.([1 ] ) and ( [ 2 ] ) ) .
in the following we consider ultracold collisions of identical bosons in the presence of an external harmonic confinement . the scattering process is taking place in the low energy regime , @xmath28 , where @xmath29 is the relative momentum in the unconfined degree of freedom and @xmath30 is the oscillator length defined via the trap frequency @xmath31 and the reduced mass @xmath32 of the colliding bosonic pair .
furthermore , in order to go beyond s - wave physics we assume that the momentum @xmath29 is very small but not equal to zero . consequently
, as it was shown in @xcite except for the well known s - wave cir @xcite , an additional resonance emerges in this system , the so - called dwr , which is based on the interference of a free space d - wave shape resonance with s - wave scattering .
the coexistence of s- and d - wave resonances implies interactions between the bosons of s- and d - wave symmetries ( see fig.1(c ) in ref.@xcite ) .
the main goal of this paper is to present an analytical model for the dwr effect in a harmonic waveguide .
however , the modeling of s- and d - wave interactions with pseudopotentials in systems of non - spherical symmetry , as the above - mentioned one , is not a trivial task . as the symmetry of the system is cylindrical
the angular momentum is not conserved and consequently , the application of pseudopotentials , eqs .
( [ 1 ] ) and ( [ 2 ] ) , leads to an explicit integration over all @xmath7-partial waves .
though , as it was proposed in @xcite one can avoid such difficulties by considering a single @xmath7-partial wave in the vicinity of the @xmath7-wave resonance , where contributions from other partial waves are sufficiently suppressed .
the latter means that in the vicinity of the s - wave resonance , we assume that the interaction can be modeled by the following pseudopotential with the help of eqs .
( [ 4 ] ) and ( [ 5 ] ) : @xmath33 , \label{6}\ ] ] which is the huang - fermi pseudopotential , with the s - wave scattering length @xmath34 being energy dependent .
note that eq.([6 ] ) coincides with the pseudopotential proposed in @xcite .
equivalently , in the vicinity of the d - wave resonance , we assume that the interaction can be modeled by the following pseudopotential @xcite : @xmath35 where @xmath36 with the indices @xmath37 and @xmath38 is the energy dependent d - wave scattering length . the bidirectional operator , @xmath39 @xmath40 denotes the differential operator in cartesian coordinates , that acts to the left ( right ) of the pseudopotential @xmath41 .
we emphasize that application of the expressions ( [ 6 ] ) and ( [ 7 ] ) simultaneously is inconsistent and incompatible since the pseudopotentials @xmath42 and @xmath41 are valid for the cases where the wavefunction exhibits only @xmath43- or @xmath44- singularities for small @xmath45 , respectively .
we assume two - body collisions of identical bosons in a waveguide with their transversal degrees of freedom being confined by a harmonic oscillator potential .
the harmonic confinement permits the separation of the center of mass and relative motion yielding a single - particle hamiltonian with a scatterer fixed at the origin for the relative degree of freedom . @xmath46 where @xmath47 is the relative coordinate of the two bosons , and @xmath48 is the pseudopotential which models the arbitrary @xmath7-wave inter - atomic interaction .
the symmetry of the transversal potential , @xmath49 , implies that the scattering solutions should be expanded in an axially symmetric basis .
such a basis set are the eigenstates of the two - dimensional ( 2d ) harmonic oscillator which satisfy the schrdinger equation @xmath50\tilde{\varphi}_{n , m_z}(\rho,\phi)\\ = e_{n , m_{z}}\tilde{\varphi}_{n , m_z}(\rho,\phi ) , \normalsize{\label{9}}\end{aligned}\ ] ] where @xmath51 are the eigenfunctions of the 2d harmonic oscillator , @xmath52 is the azimuthal quantum number and @xmath53 is the transversal energy spectrum of the 2d harmonic oscillator , with @xmath54 being the principal quantum number .
the spherical symmetry imprinted by the pseudopotential , @xmath48 , leads to a separation of the azimuthal @xmath55 angle from the @xmath56 , @xmath57 coordinates in eq.([9 ] ) .
the latter means that during the collision no virtual excitations will occur with respect to the @xmath52 quantum number .
however , virtual excitations will emerge with respect to the principal quantum number @xmath58 , which will be taken into account in the calculation given below .
we focus on the single mode regime , where the collision takes place in the ground state of the transversal confinement with angular momentum @xmath59 and represents the only `` open channel '' .
the latter means that the total energy of the pair collision is limited between the ground and the first axially symmetric transversal state @xmath60 furthermore , the above - mentioned independence on the angle @xmath55 results in the following simplified schrdinger equation : @xmath61\psi(z,\rho)\\ = e \psi(z,\rho ) , \label{11}\end{aligned}\ ] ] where @xmath62 is the total colliding energy , which is defined as the sum of the longitudinal kinetic energy @xmath63 and the energy of the transverse ground state @xmath64 . due to the constraint of eq.([10 ] ) the momentum @xmath29 is limited according to the relation @xmath65 . here , @xmath66 denotes the full 2d axially symmetric solution , which can be expanded in the @xmath67 basis @xmath68 where we sum over all even @xmath58 due to the bosonic symmetry . substituting eq.([12 ] ) into eq.([11 ] ) , multiplying by @xmath69 , integrating and using eq.([9 ] ) yields the following equation for the functions @xmath70 : by solving eq.([13 ] ) for @xmath72 we obtain the scattering wavefunction @xmath73 in the open channel which has the following asymptotic form : @xmath74 where @xmath75 is the one - dimensional scattering amplitude of the @xmath7-wave interaction which describe the even scattered waves . for @xmath76 , eq.([13 ] ) provides us with the solution for the virtual excitations associated with the closed channels , which decay exponentially according to the relation @xmath77 where the coefficients @xmath78 refer to the @xmath7-wave interaction and denote the transition amplitudes from the transverse ground state to the @xmath58-th excited state . in the case of the d - wave pseudopotential we substitute eq.([14 ] ) into eq.([13 ] ) for @xmath72 and eq.([15 ] ) into eq.([13 ] ) for @xmath76 and integrate over the @xmath56 variable , respectively , in the interval @xmath79 $ ] with @xmath80 .
finally , we obtain the following relations for the d - wave scattering and transition amplitudes : @xmath81 where @xmath82 is the regularized part of the wavefunction @xmath83 in the limit @xmath84 , @xmath85\bigg\ } \bigg| _ { z \rightarrow0^+ } , \label{18}\end{aligned}\ ] ] with @xmath57 given by the relation @xmath86 .
then the expression for the wavefunction reads @xmath87\varphi_{0}(\rho)\\ + \frac{a_{d}^5(k ) \sqrt{\pi}}{48a_{\perp}^2}\eta_d \lambda_1(z,\rho ) , \label{19}\end{aligned}\ ] ] where @xmath88 is defined as follows : @xmath89 let us now to proceed with the explicit definition of the regular part @xmath82 of the wavefunction , being a solution of eq.([18 ] ) .
the action of the operator @xmath90 on @xmath91 for @xmath92 results in the following relation : @xmath93 -\frac{a_{d}^5(k)}{24a_{\perp}^5}\eta_d \lambda_2(z,\rho=0 ) , \label{21}\end{aligned}\ ] ] where @xmath94 reads : @xmath95 here , we should note that the @xmath96-series at @xmath97 are not converging uniformly as @xmath98 .
the latter means that the partial derivative w.r.t .
@xmath56-coordinate can not be interchanged with the sum over all even @xmath58 .
thus , firstly one has to perform the summation and then differentiate with respect to the @xmath56 coordinate .
however , with the help of euler s summation formula @xcite one can express the @xmath96-series as an expansion with respect to the @xmath56 variable , where the singular part of the @xmath96-series will be written in closed form yielding the following expressions : @xmath99 where @xmath100 , @xmath101 , @xmath102 and @xmath103 are constants calculated via euler s summation formula .
consequently , by substituting eqs.([19])-([24 ] ) into eq.([18 ] ) we obtain @xmath82 , which is now free of divergences .
this in turn yields the following expression for the one - dimensional scattering amplitude : @xmath104 where the constant @xmath105 is defined by the relation @xmath106 .
the resonance condition for @xmath107 then reads @xmath108{10b_2}. \label{26}\ ] ] eq.([25 ] ) shows that the d - wave scattering in the presence of the waveguide becomes resonant at an off - resonant value of the free space @xmath3 scattering length , due to the significant contributions of the virtual excitations .
the latter holds equally for the case of s - wave cir .
therefore , one can conclude that in general the appearance of resonances in systems with transversal confinement are generated by the virtual excitations in the transverse modes . by applying the s - wave pseudopotential , given in eq.([6 ] ) , and performing the derivation as mentioned above , we will obtain the same relation for the one - dimensional scattering amplitude as in @xcite with the difference that in this case the s - wave scattering length is energy dependent @xcite : @xmath109 the constant @xmath110 in eq.([27 ] ) is equal to @xmath111 .
in this section we compare the analytical results of the previous subsection for the s- and d - wave pseudopotentials with the corresponding numerical simulations .
first we evaluate the d- and s - wave energy dependent scattering lengths by solving numerically a model of cs atoms in free space interacting via a lennard - jones potential , @xmath112 .
note that cs atoms are ideal canditate for studing scattering phenomena , since they provide a rich spectrum of resonances @xcite .
the dispersion coefficient @xmath113 has been taken from @xcite and @xmath114 is a free parameter , which controls the values for the s- and d - wave scattering lengths .
this yields a parametrization of the one - dimensional s- and d - wave scattering amplitudes , eqs.([27 ] ) and ( [ 25 ] ) , respectively , in terms of @xmath114 . in order to numerically solve the cs - cs collisions in the harmonic waveguide
we employ the units @xmath115 , where @xmath116 is the mass of the cs atom and @xmath117 mhz .
the longitudinal energy is set to @xmath118 and the transversal energy is varied within the interval @xmath119 , corresponding to a range @xmath120 for the waveguide confinement frequency .
we thereby focus on the low energy regime , characterized by @xmath121 . in the following
we calculate the transmission coefficient given by : @xmath122 where @xmath123 is the one - dimensional scattering amplitude for the @xmath7-wave . , which shows the dwr on the l.h.s and the s - wave cir on the r.h.s .. the analytically calculated ( b ) s - wave transmission coefficient , @xmath124 , and ( c ) d - wave transmission coefficient , @xmath125 .
the transversal confinement is @xmath126 .
[ fig1 ] in fig .
[ fig1 ] we show the transmission coefficient as a function of @xmath114 with the transversal frequency @xmath127 .
@xmath128 in fig .
[ fig1](a ) denotes the numerically calculated transmission coefficient , as it was shown in @xcite .
we observe the appearance of two minima ( @xmath129 ) due to resonances .
the minimum on the r.h.s . of fig .
[ fig1](a ) refers to the s - wave cir and the minimum on the l.h.s . to the dwr . due to the centrifugal barrier
the width of the dwr is much narrower than that of the s - wave cir .
additionally , we observe that both resonances yield a lorentzian - like lineshape in the transmission spectrum .
this occurs due to the fact that the weak confinement suppresses the interference effects between the s - wave and d - wave interactions . in fig .
[ fig1](b ) the analytically calculated transmission coefficient @xmath124 , as given by eqs.([27 ] ) and ( [ 28 ] ) , is shown which refers to s - wave interactions only .
@xmath124 describes very accurately the s - wave cir ( see fig .
[ fig1](a)-(b ) ) , whereas there is no trace of the dwr , since contributions from higher partial waves are not included in eq.([27 ] ) . in fig .
[ fig1](c ) the analytic expression of the transmission coefficient @xmath125 is depicted according to eqs.([25 ] ) and ( [ 28 ] ) .
we observe that the dwr is described qualitatively accurate enough ( see also insets of figs.[fig2](a)-(b ) ) , despite the fact that in the analytical calculations we have neglected the contributions of s - wave scattering , since the transversal confinement is very weak .
we note that the results shown in figs .
[ fig1](b ) and [ fig1](c ) being obtained analytically for the pure s- and d - wave scattering cases lead to an accurate description of the scattering processes in the presence of a weak external harmonic potential . versus @xmath114 parameter for several confinement frequencies
@xmath130 , ( right panel ) high - resolution figure of the dwr for @xmath127 where the inwards - pointing arrows denote the width of the resonance .
( b)(left panel ) the analytically calculated transmission coefficient @xmath125 as a function of the @xmath114 parameter calculated with the d - wave pseudopotential for the corresponding values of @xmath130 , ( right panel ) high - resolution graph of the dwr for @xmath127 .
the yellow dashed line in the right panels indicates the minimum of the numerical calculated transmission coefficient ( @xmath129 ) in comparison with the corresponding analytical result . ]
let us now determine the regime of validity of the approximative analytical results for the d - wave pseudopotential scattering by comparing it with the corresponding numerical simulations . in fig .
[ fig2 ] we show the transmission coefficient @xmath131 calculated numerically ( fig . [ fig2](a ) ) and analytically ( fig . [ fig2](b ) ) for several transversal confinement frequencies @xmath130 . in fig .
[ fig2](a ) for a strong confinement , e.g. @xmath132 , we first observe that the numerically calculated position of the resonance deviates substantially from the analytical one .
however , as @xmath130 decreases the numerical results for the position of the dwr converges to the corresponding analytical one .
the origin of this behavior is the fact that we considered only d - wave interactions in the analytical calculations and neglected contributions from the s - wave interactions .
this approximation is eligible in the regime of a weak confinement , e.g. @xmath133 , where the s- and d - wave free space resonances are weakly coupled , and in the vicinity of the dwr the d - wave interactions are dominant over the s - wave . in the regime of strong confinement the transmission spectrum @xmath131 , in the vicinity of the dwr , exhibits a strong asymmetric fano - lineshape , with @xmath128 changing abruptly from zero to one .
the latter is an interference effect of the strong s- and d - wave interactions in the confinement and can not be described by this analytical pure d - wave model .
however , this asymmetric profile gets suppressed with decreasing @xmath130 yielding a lorentzian - like lineshape .
the right panels of fig .
[ fig2 ] show a high - resolution graph of the dwr calculated analytically and numerically for @xmath127 , where the dashed vertical line in both plots indicates the position of the minimum of the numerically calculated transmission coefficient @xmath128 .
we observe that the d - wave pseudopotential approximation qualitatively describes the numerical simulations in the regime of weak confinement . however , in this small scale of the @xmath114 values the analytical results show that the width of the dwr is by two orders wider than the numerical predictions , as well as there is a slight difference in the position of the dwr .
these quantitative deviations again occur due to s - wave interactions , which act as corrections in the transmission coefficient @xmath128 despite of the weak coupling of the s - wave and d - wave interactions .
we have derived and analyzed analytical expressions for resonant d - wave scattering in a harmonic waveguide , in the framework of pseudopotential theory .
the interatomic interaction has been modeled by a pure d - wave pseudopotential introduced in @xcite .
we observe that the analytical pseudopotential approach provides results which are in qualitative agreement with the corresponding numerical results in the regime of weak confinement due to the very small interference effects of s- and d - wave interactions.the appearance of the dwr in a waveguide is strongly affected by the virtual excitations over all the closed excited modes of the transversal confinement . for
a strong transversal confinement the analytically and numerically calculated lineshape and position of the dwr deviate from each other .
the interference effect of the s- and d - wave free space resonances in the confinement yields an asymmetric fano profile which can not be derived by a single partial wave model .
our analytical result eq.([25 ] ) for weak confinement can be used for the analysis of resonant scattering processes in confined geometries beyond s - wave physics .
finally , we remark that our work clearly demonstrates the necessity of an extended pseudopotential theory capable of describing systems which possess a non - spherical symmetry that yield interference phenomena between different partial wave scattering amplitudes .
p.g . thanks c. morfonios and f.k .
diakonos for valuable comments and discussions .
v.s.m . acknowledges financial support by the deutsche forschungsgemeinschaft and the heisenberg - landau program .
p.s . acknowledges the deutsche forschungsgemeinschaft for financial support .
99 e. fermi , nuovo cimento , * 11 * , 157 ( 1934 ) .
k. huang and c. n. yang , phys . rev . * 105 * , 767 ( 1957 ) .
k. huang , _ statistical mechanics _ , ( john wiley & sons , new york , 1963 ) . t. busch , b .-
englert , k. rzazewski , and m. wilkens , found .
* 28 * , 549 ( 1998 ) .
m. olshanii , phys .
. lett . * 81 * , 938 ( 1998 ) .
t. kinoshita , t. wegner , and d.s .
weiss , science * 305 * , 1125 ( 2004 ) .
b. paredes _ et al .
_ , nature ( london ) * 429 * , 277 ( 2004 ) .
e. haller , m. gustavsson , m.j .
mark , j.g .
danzl , r. hart , g. pupillo , and h.c .
ngerl , science * 325 * , 1224 ( 2009 ) .
e. tiesinga , c.j .
williams , f.h .
mies , and p.s .
julienne , phys .
a * 61 * , 063416 ( 2000 ) .
d. blume , and c.h .
greene , phys .
a * 65 * , 043613 ( 2002 ) .
e. l. bolda , e. tiesinga , and p. s. julienne , phys .
a * 66 * , 013403 ( 2002 ) .
a. derevianko , phys .
rev . a * 72 * , 044701 ( 2005 ) .
r. stock , a. silberfarb , e. l. bolda , and i. h. deutsch , phys .
lett . * 94 * , 023202 ( 2005 ) .
z. idziaszek and t. calarco , phys .
96 * , 013201 ( 2006 ) .
z. idziaszek , phys .
a * 79 * , 062701 ( 2009 ) .
f. stampfer and p. wagner , j. math .
appl . * 342 * , 202 ( 2008 ) .
l. pricoupenko , phys .
lett . * 96 * , 050401 ( 2006 ) .
k. huang , int .
phys . a * 4 * , 1037 ( 1989 ) .
y. n. demkov and g. f. drukarev , sov .
jetp * 54 * , 650 ( 1981 ) .
m. v. frolov , n.l .
manakov , e. a. pronin , and a. f. starace , phys .
* 91 * , 053003 ( 2003 ) .
a. m. rey , r. sensarma , s. flling , m. greiner , e. demler , and m. d. lukin , europhys .
87 * , 60001 ( 2009 ) .
b. deb and j. hazra , phys .
lett . * 103 * , 023201 ( 2009 ) .
w. hofstetter , j. i. cirac , p. zoller , e. demler , and m. d. lukin , phys .
rev . lett . * 89 * , 220407 ( 2002 ) .
k. kanjilal , and d. blume , phys .
a * 70 * , 042709 ( 2004 ) .
k. gnter , t. stferle , h. moritz , m. khl , and t. esslinger , phys .
lett . * 95 * , 230401 ( 2005 ) .
p.giannakeas , v. s. melezhik and p. schmelcher , phys .
a * 84 * , 023618 ( 2011 ) .
u. fano , phys . rev . * 124 * , 1866 ( 1961 )
. t. m. macrobert , _ spherical harmonics _ , ( dover publications , new york , 1948 ) .
t. m. apostol , _ the american mathematical monthly _ , vol .
106 , no . 5 , 409 - 418 ( 1999 ) . c. chin _ et al .
a * 70 * , 032701 ( 2004 ) .
_ , phys . rev .
a * 76 * , 042514 ( 2007 ) .
a. derevianko , j.f .
babb , a. dalgarno , phys .
a * 63 * , 052704 ( 2001 ) .
porsev and a. derevianko , phys .
a * 65 * , 020701(r ) ( 2002 ) . v. s. melezhik , j. comput
. phys . * 92 * , 67 ( 1991 ) .
s. saeidian , v. s. melezhik , and p. schmelcher , phys .
a * 77 * , 042721 ( 2008 ) . | we analyze d - wave resonances in atom - atom scattering in the presence of harmonic confinement by employing a higher partial wave pseudopotential .
analytical results for the scattering amplitude and transmission are obtained and compared to corresponding numerical ones , which employ the lennard - jones potential .
qualitative agreement is observed for weak confinement . for strong confinement
the pseudopotential does not capture the s- and d - wave interference phenomena yielding an asymmetric fano profile for the transmission resonance . |
relations between mechanics and thermodynamics have been several times pointed out and pedagogically discussed in the literature . in particular we have studied in detail systems whose physics description requires both mechanical and thermodynamical approaches such as the cartesian diver @xcite and the drinking bird @xcite . also many toys ,
whose physics was revisited in a recent publication @xcite , require a mechanical - thermodynamical analysis in order to fully understand their behaviours
. however , in this work we want to emphasize the use of two energy equations " , namely the pure mechanical one , relating the variation of the centre - of - mass kinetic energy with the pseudo - work performed by the resultant external force , and the other one , which is simply the equation expressing the first law of thermodynamics : the variation of the total energy of a system is accounted for by all types of works upon the system and by the heat transfers .
the level of readership of this article is intended for undergraduate students and our aim is to help students and professors to better address thermodynamical issues in the context of mechanical problems .
the classical dynamics of a system of particles of mass @xmath4 is governed by newton s second law which can be expressed by _
ext t = m _ cm , [ newtonpt ] where @xmath5 is the resultant of the external forces and @xmath6 the center - of - mass velocity .
the equation , valid in a given inertial reference frame , remains invariant upon inertial frame transformations .
the previous equation also allows us to conclude that the infinitesimal variation of the kinetic energy of the centre - of - mass equals the so called pseudo - work " @xcite . actually , by multiplying both sides of eq .
( [ newtonpt ] ) by @xmath6 one concludes that the dot product of the resultant of the external forces by the infinitesimal displacement of the centre - of - mass is _ ext _ cm = 12 m v_cm^2 .
[ eq-3 ] the infinitesimal pseudo - work " is the dot product of the resultant force applied to a system and of the corresponding infinitesimal centre - of - mass displacement .
it should not be confused with the infinitesimal work " of a force which is the dot product of the force and of its own infinitesimal displacement .
the above two equations are equivalent , though they may provide complementary information .
the first equation will be referred to as the impulse - momentum equation and states that the infinitesimal impulse of @xmath5 equals the infinitesimal variation of the linear momentum of the center - of - mass , @xmath7 .
the second equation , which will be referred to as the centre - of - mass equation , states that the infinitesimal pseudo - work equals the infinitesimal variation of the centre - of - mass kinetic energy , @xmath8 .
note that , in general , @xmath9 is not the total kinetic energy of the system , @xmath10 , which may also include internal kinetic energy , @xmath11 , for instance , rotational energy . of course , from eqs .
( [ newtonpt ] ) and ( [ eq-3 ] ) , one readily obtains the corresponding integral equations
. the resultant force acting on the system can be the sum of conservative and non - conservative forces .
on the other hand , if there are , for instance , thermal effects , these are not described by the centre - of - mass equation ( [ eq-3 ] ) @xcite .
therefore , in order to study the motion of extensive deformable bodies one needs additional equations to describe rotations @xcite and processes involving production @xcite and dissipation @xcite of mechanical energy . to keep the discussion in this paper at a simple level we do not consider rotations .
here we are particularly interested in addressing the physical description of the kind of processes that also need the first law of thermodynamics @xcite . for any system , its internal energy variation ,
@xmath12 , receives contributions from the variation of the internal kinetic energy , @xmath13 ( including rotational kinetic energy or kinetic energy with respect to the centre - of - mass ) , from any internal work , @xmath14 ( _ i.e. _ work performed by the internal forces ) @xcite , from the internal energy variations related to temperature variations , @xmath15 ( @xmath16 is the specific heat ) , from the internal energy variations related to chemical reactions @xcite , etc . in thermodynamics
both work and heat contribute to the internal energy variation of a system . for a general process on a macroscopic deformable body
, whose analysis needs to combine mechanics and thermodynamics @xcite , instead of the centre - of mass equation one should consider the equation _ cm + = _ j f_ext , j _ j + [ totale ] which is nothing but an expression of the first law of thermodynamics @xcite : it is more general than the centre - of - mass equation ( [ eq-3 ] ) , but both are valid . in ( [ totale ] ) the infinitesimal heat is denoted by a @xmath17 since it is not an exact differential .
each term in the sum on the right - hand side of eq .
( [ totale ] ) is work associated with each _ external _ force @xmath18 , and @xmath19 is the infinitesimal displacement of the force @xmath20 itself ( and not , anymore , the displacement of the centre - of - mass ) .
therefore , @xmath21 is actual work and not pseudo - work .
the above equations are valid in any inertial frame .
this means that the amount of information provided by the set of equations is _ exactly _ the same , irrespective of the chosen inertial frame there are no privileged inertial frames and the above equations are galilean invariant . in many situations there are forces with zero spatial displacement in a certain inertial reference frame ( e.g. the force exerted by the ground on the foot of a walking person ) . in these cases
the force produces no work but there is an impulse and a corresponding linear momentum variation : @xmath22 .
let us consider an example from thermodynamics : a gas of small mass ( hence , gravitational forces are negligible ) is in equilibrium inside a vertical cylinder enclosed by a piston at the top .
a force @xmath23 is applied to the piston and , hence , by the piston to the gas .
this force is always compensated by an opposite force , @xmath24 , exerted upon the gas by the bottom of the cylinder ( the horizontal forces produced by the cylinder walls have zero resultant and can be ignored ) .
the displacement of this force is always zero .
for a quasi - static process the impulse - momentum equation ( [ newtonpt ] ) and the centre - of - mass equation ( [ eq-3 ] ) applied to this system yields \ { l m v_cm = ( f - f ) t + 12 m v_cm^2 = ( f - f ) y_cm . , [ dfcb ] leading to @xmath25 : if the cylinder is initially at rest , @xmath26 during the process , though the position of the centre - of - mass might change , @xmath27 .
on the other hand , in the energy equation ( [ totale ] ) , which incorporates the first law of thermodynamics , one has @xmath28 and , since the bottom force does not do any work , one obtains u = -f y + q ( @xmath29 is the displacement of the force exerted by the piston ) equivalent to the well known equation @xmath30 ( here , @xmath31 is the pressure and @xmath32 is the volume ) .
for a body that behaves like an elementary particle ( no rotation , no deformation , etc . ) the centre - of - mass equation and the first law of thermodynamics provide the same information .
but when the body does not behave like an elementary particle , such as the gas in the cylinder , the two equations provide complementary information . summarizing , besides the impulse - momentum equation , one has at our disposal the centre - of - mass equation and the energy equation @xcite . both equations are always valid and , therefore , they should be compatible .
they may provide the same information if the problem under consideration is a purely mechanical one .
however , if the situation is within the scope of thermodynamics , the second equation , which embodies the first law , is more general and provides new information with respect to the centre - of - mass equation .
of course , one still needs the second law of thermodynamics , which states that , observed the first law , only processes compatible with a non - decrease of the entropy of the universe , @xmath33 , are allowed . to illustrate the usage of the impulse - momentum equation together with the centre - of - mass or with the more general energy equation , we discuss , in the next sections , a set of examples ranging from mechanics to thermodynamics .
we start with a pure mechanical example , continue with problems involving destruction or production of mechanical energy and conclude with a typical thermodynamical problem .
this paper is organized as follows . in section [ sec : boai ] we consider the motion of a block sliding down an incline without friction . in section 3
we study a totally inelastic collision . in section [ sec : boyscat ] we study an example of mechanical energy production by considering the motion of a person on rollers pushing against a wall . in section 5
we treat an inherently inelastic process
the pulling of a chain .
finally , in section [ sec : jte ] we address a typical thermodynamical problem the joule - thomson experiment . with these examples , ranging from a pure mechanical situation to a typical thermodynamical one
, we show the appropriateness of the first law of thermodynamics with respect to the newton s second law in the mechanical description of the system .
a rigid block of mass @xmath4 slides down an incline without friction , as fig .
[ fig : blockinclinenf ] shows .
we describe the process in a reference frame with horizontal @xmath34 and vertical @xmath35 . [
sec : boai ] rigid block on an incline.,width=340 ] there are two forces , namely the weight , @xmath36 and the normal force , @xmath37 , where @xmath38 is the angle between the incline and the horizontal plane .
the block is a rigid body and there are no rotations , hence the displacement of each force is also the displacement of the centre - of - mass .
the infinitesimal displacements are _ g= _
cm = ( l , -l , 0 ) with @xmath39 the elementary displacement along the incline . in the absence of rotations or any other change in the internal energy of the body ( and no heat transfers ) @xcite , the energy equation ( [ totale ] ) is exactly the centre - of - mass equation ( [ eq-3 ] ) .
for the finite process , the block , initially at rest , slides down a distance @xmath40 during the time interval @xmath41 .
the integration of the impulse momentum and of the centre - of - mass equations are straightforward and lead to n = mg ; t_0 = ( 2 l g ) ^1/2 ; v_0 = ( 2 l g ) ^1/2 .
[ erop ] where @xmath42 is the final centre - of - mass velocity .
these results were obtained in the chosen reference frame , but any other inertial frame could be used , leading to equivalent expressions .
in particular the frame with @xmath34 along the incline and @xmath35 perpendicular to it is commonly used .
the simple mechanical process studied in this section is reversible , _
i.e. _ , if the body were launched with opposite velocity from the basis of the incline , it would reach the same original height .
this means that @xmath43 .
this is not the case if there were kinetic friction and it is also not the case in some of the following examples .
the next example is a perfect inelastic collision of a plasticine ball with the ground @xcite and the situation is represented in fig . [
fig : bolachocasuelo ] .
we are only interested in the part of the motion after the initial contact with the soil , until the ball stops completely .
inelastic impact of a plasticine ball with the floor ( i initial contact ; f ball stops ) .
, width=491 ] there are two forces : the weight , @xmath44 , and the force , @xmath45 , exerted by the ground , which , for simplicity , is assume to be constant ( can be regarded as an average force ) . with the @xmath35 axis
directed upwards , the two forces are @xmath46 and @xmath47 .
we denote by @xmath48 the vertical displacement of the centre - of - mass and by @xmath41 the time interval since the initial contact of the ball with the floor until its gets totally stopped .
the displacements associated with the forces and with the centre - of - mass are ( note that the displacement of the normal force is zero @xcite ) : _ g = _ cm= ( 0 , -h , 0 ) _ n = ( 0,0,0 ) . [ rgn ] after integration , the centre - of - mass equation ( [ eq-3 ] ) leads to @xmath49 and , denoting by @xmath50 the velocity of the centre - of - mass when the ball reaches the floor , one obtains ( @xmath51 ) , for the impulse - momentum and center - of - mass equation , \
{ rl m v_0 & = ( n - mg ) t_0 , + 12 m v_0 ^ 2 & = ( n - mg ) h . .
[ 850o ] note that the second equation , namely @xmath52 , is not a ` work - energy ' equation since @xmath53 is pseudo - work .
let s now exploit the energy equation ( [ totale ] ) , which can also be readily integrated ( forces are constant ) , assuming the form _
cm + = g_g + n_n + q . [ 87uy ] in the absence of other effects ( rotations , work of internal forces ) the variation of the internal energy of the ball is assumed to be = m c ( t_f - t_i ) , where @xmath54 are the initial and final temperatures of the ball and @xmath16 its specific heat .
this and @xmath55 in ( [ 87uy ] ) accounts for thermal effects in this type of inelastic collisions . the impulse - momentum equation ( [ newtonpt ] ) and eq .
( [ 87uy ] ) allow us to write \
{ rl m v_0 & = ( n - mg ) t_0 , + -12 m v_0 ^ 2 + m c ( t_f - t_i ) & = m g h + q . .
[ eq : pbcnsl ] to experimentally confirm these equations , now one would need a thermometer , in addition to a watch and a ruler .
the two sets of equations ( [ 850o ] ) and ( [ eq : pbcnsl ] ) are both valid and must be compatible . from the second equations of each set
one concludes that @xmath56 @xcite . assuming that the temperature change of the ball is negligible , @xmath57 , one concludes that there is an energy transfer , @xmath58 , from the system .
this energy is thermal energy ( heat ) exchanged with the thermal reservoir at temperature @xmath59 and , therefore , it is a positive quantity from the reservoir point of view .
the variation of the entropy of the universe is solely the variation of the entropy of the thermal reservoir and it is given by s_=nh t > 0 . for a deformed ball initially at rest on the ground , one could think about an injection of the amount of energy @xmath60 from the heat reservoir , e.g. through the action of the force @xmath61 .
this process , opposite to the one that we have studied , would be allowed by the first law of thermodynamics .
however , it would lead to a decrease of the entropy of the universe , which is prohibited by the second law of thermodynamics @xcite .
therefore this inverse process can not happen spontaneously @xcite .
this is not the case for the pure mechanical example considered in the previous section , which does not increase the entropy .
let us now consider the description of the same inelastic process in the reference frame s@xmath62 , moving vertically , along the common @xmath63 axes , with velocity @xmath32 with respect to s. the forces are identical in both reference frames but the displacements in s@xmath62 , @xmath64 , differ from ( [ rgn ] ) in the @xmath35 component , where there is an extra term @xmath65 : _g = _cm= ( 0 , -h - v t_0 , 0 ) , _n = ( 0 , -v t_0 , 0 ) the impulse - momentum and the center - of - mass equations in the new frame are \ { rl m v_0 & = ( n - mg ) t_0 , + mv_0 v+12 m v_0 ^ 2 & = ( n - mg ) ( h+vt_0 ) . .
[ 850pl ] which reduces exactly to ( [ 850o ] ) .
formally , in any situation , the equations in reference s@xmath62 are always linear combinations of the equations in reference s , and vice - versa , according to the principle of relativity , and we have explicitly shown this for a particular case . therefore , by changing to a new inertial reference frame , one does not gain or loose information ( though one reference frame might be better suited than another one regarding technical aspects ) next we explore the energy equation also in s@xmath62 : _cm + = w_g + w_n + q , the internal energy variation and the heat are the same in both frames : @xmath66 and @xmath67 . the impulse - momentum and the energy equation
are readily obtained and altogether lead to \
{ rcl - m v + m ( v_0 + v ) & = & ( n - m g ) t_0 + -12 m v_0 ^ 2 - m v v_0 + m c ( t_f - t_i ) & = & m g h + m g v t_0-
n v t_0 + q . .
[ eq : pbfltsb ] again , these equations reduce to ( [ eq : pbcnsl ] ) as expected .
the discussion on thermal effects , heat transfer and increase in the entropy of the universe , presented above , still applies here the temperature and the entropy variation , @xmath68 , are galilean invariants .
it is worth noticing that eq .
( [ totale ] ) can still be written as @xmath69 and both sides of the equation are , _ per se _ , galilean invariant too .
a boy on rollers , initially at rest , pushes against a wall ( with infinite mass ) and slides away over the floor @xcite .
the mass of the system is @xmath4 and no dissipative forces between the rollers and the floor are considered . an external horizontal force , @xmath70 ,
is exerted on him by the wall and , as a consequence , his translational kinetic energy increases .
the vertical forces the weight and the normal force from the ground cancel each other and they play no role in the discussion .
the force @xmath70 does not do any work because its displacement is zero @xcite ( fig .
[ fig : personscaters ] ) .
person on rollers pushing against a wall .
, width=340 ] this is again a one dimensional problem and we describe it in reference frame s with the @xmath34 axis pointing to the right .
the force along the @xmath34 direction is @xmath71 , represented in the figure .
we assume @xmath72 to be constant ( @xmath72 can be considered an average force ) .
the displacements of the centre - of - mass and of the horizontal force are _ cm = ( x_cm , 0 , 0 ) , _ f = ( 0,0,0 ) where @xmath73 is the spatial displacement of the centre - of - mass occuring in the time interval @xmath41 , the duration of the action of the horizontal force upon the person .
denoting by @xmath42 the velocity of the boy s centre - of - mass when the contact with the wall ceases , the impulse - momentum and the centre - of - mass equation lead to \ { rl m v_0 & = f t_0 , + 12 m v_0 ^ 2 & = f x_cm . .
[ 8670 ] we stress that the second equation is not an expression of an ` energy - work ' theorem since the external normal force exerted by the wall does not produce any work @xcite . if the force does not do any work , where is the source of energy responsible for the increase of the centre - of - mass kinetic energy ?
these energetic issues are better discussed using the energy equation . during the process of pushing against the wall ,
the work of @xmath70 vanishes but not its impulse .
there are biochemical reactions in the person s muscles .
when a chemical reaction @xmath74 is produced inside the body , these may cause variations in its internal energy , @xmath75 , volume , @xmath76 , entropy , @xmath77 , etc .
if a chemical reaction takes place at constant external pressure , @xmath31 , in diathermal contact with a heat reservoir at temperature @xmath59 , part of the internal energy is used for the expansion against the external pressure and part must be exchanged with the heat reservoir in order to ensure that the entropy of the universe does not decrease .
the work , @xmath78 , and the heat , @xmath79 , are : _
p = - p v _ ; q_=
t s _ , when @xmath80 , the external pressure performs work on the system and when @xmath81 the heat reservoir increases the internal energy of the system .
we assume that no other energy is exchanged between the person and the environment ( no extra heat is exchanged between the person and the environment ) . the energy equation for this process eq .
( [ totale ] ) is : _ cm + u_= w_f + w_p + q , and , together with the impulse - momentum equation , \ { rl m v_0 & = f t_0 , + 12 m v_0 ^ 2 + u_&= -p v_+ ts _ . .
[ eq : fltboys ] the second equation provides new information with respect to the second equation in ( [ 8670 ] ) . in particular , the second equation in eq .
( [ eq : fltboys ] ) can also be expressed as m v_0 ^ 2 = - g _ , where : g_= u_+p v_- ts _ , is the gibbs free energy variation , which is symmetric of the pseudo - work associated with @xmath70 .
this equation shows that , in effect , the person increases its centre - of - mass kinetic energy thanks to the internal biochemical reactions @xcite .
for given temperature and pressure , the variation of the gibbs free energy is the maximum useful work that can be obtained : @xmath82 . in the reference frame s@xmath62 , in standard configuration with respect to s , the centre - of - mass equation is _cm = f_cm , which , again with the impulse - momentum equation results in ( in s@xmath62 the displacement of the force exerted on the person s hand is @xmath83 ) \
{ rcl - m v - m ( v_0-v ) & = & f t_0 , + 12 m v^2 - 12 m ( v_0-v)^2 & = & f ( x_cm- v t_0 ) , .
equivalent to ( [ 8670 ] ) .
the same happens with the energy equation in reference frame s@xmath62 : from @xmath84 one obtains ( [ eq : fltboys ] ) .
the process studied in this section implies no entropy increase of the universe and , therefore , it is reversible .
in fact , the kinetic energy of the system can be completely used , at least in principle , to increase by @xmath85 the free energy of any chemical reaction .
real bodies , which are articulated or made of elastic materials , may acquire accelerations ( changes of centre - of - mass momentum and kinetic energy ) as a consequence of the action of internal forces when they are in contact with an infinite mass body ( a wall , the floor , etc . ) ; of course , internal forces , by themselves , can not accelerate the center - of - mass of an isolated body @xcite .
the final kinetic energy in the example studied in this section , or in the case of a person that jumps , is not due to the work of the contact forces since they have zero displacement @xcite . any explicit or implicit assumption that forces with zero displacement applied to a moving body ( forces on the foot of a person that walks , on the tires of a car that accelerates , on the foot of a person that jumps , etc . )
do work is incorrect .
in fact , the total kinetic energy of a composite , articulated body in contact with an infinite mass body can change , even when no work is done by the external forces applied to it . in the description of processes with production of mechanical energy ,
the role of the chemical reactions ( food consumption in people that walk or jump , fuel combustion in cars , etc . ) plays a central role . in this section
we analised a case of mechanical energy production . in examples with mechanical energy destruction ( e.g. the plasticine ball that stops and deforms after its impact with the floor , or the inherently elastic process of pulling a chain to be treated in the next section )
we come out to the conclusion that the entropy of the universe increases .
a chain of length @xmath40 and mass density @xmath86 lies in a heap on the floor @xcite .
its end is grabbed and pulled horizontally with a force @xmath70 , as fig .
[ fig : cadenaquesedesenrolla ] shows . as a result , every piece of the moving unwrapped chain acquires a _ constant _ velocity @xmath87 .
a chain unwraps under the action of a force @xmath70 moving with constant velocity @xmath88 .
, width=453 ] we denote by @xmath34 the position of the end of the chain , so that @xmath89 is the length of the chain that remains on the heap .
the centre - of - mass of the chain is located at position ( we assume that the heap is narrow ) x_cm = x ( x/2)l = x^22l .
the centre - of - mass velocity is the time derivative and it is given by v_cm = x v_0l . the centre - of - mass momentum ( product of the total mass by the centre - of - mass velocity ) is p_cm = l v_cm = x v_0=p . meaning that the linear momentum of the moving part of the chain is always equal to the linear momentum of the centre - of - mass .
the centre - of - mass kinetic energy is ( half of the total mass times @xmath90 ) k_= 12 l v_cm^2 = 12 , which is , in general , different from the total kinetic energy given by ( half of the mass of the already moving part times @xmath91 ) : k = 12 x v^2_0 .
when the end of the chain moves by @xmath92 , the above mentioned magnitudes change accordingly , and their infinitesimal changes are given by x_cm & = & x lx , + v_cm & = & 1 l v _ 0x , + p_cm & = & _ 0 x = p , [ eqpa ] + k_&= & v^2_0 x , [ eqpx ] + k & = & 12 v^2_0 x .[eqpb ] this is a one dimensional motion and the force is given by @xmath93 .
its infinitesimal displacement and the infinitesimal centre - of - mass displacement , occurring in time @xmath94 are _ f = ( x , 0 , 0 ) _ cm = ( x lx , 0 , 0 ) .
the infinitesimal variation of the linear momentum and of the centre - of - mass kinetic energy [ see eqs .
( [ eqpa ] ) and ( [ eqpx ] ) ] immediately lead to the following set of equations \
{ rcl v_0 x & = & f t + v^2_0 x & = & f x l x . .
[ yu70 ] finally , one concludes that the pulling force must be constant and given by @xcite f = v^2_0 .
[ forcav2 ] the energy equation is @xmath95 .
if one assumes that the chain does not change its temperature during the process , the internal energy is then just the internal kinetic energy , _
i.e. _ the kinetic energy relative to the centre - of - mass ) and therefore @xmath96 ( kning s theorem @xcite )
. hence = f _ f + where @xmath97 is the kinetic energy variation .
this equation , together with the impulse - momentum equation [ see ( [ eqpa ] ) and ( [ eqpb ] ) ] allows us to arrive at \ { rl v_0 x & = f t + 12 v^2_0 x & = f x + q . .
[ yu71 ] from the first equation we obtain the information already included in ( [ yu70 ] ) : with @xmath98 , one gets ( [ forcav2 ] ) . from the second equation in ( [ yu71 ] ) one obtains v^2_0 x = v^2_0 x + q . therefore ,
thermal effects are accounted for through the equation : q = q = - 12 v^2_0 _ 0^x x = - 12 v^2_0 x = -12 f x. this means that half of the work , @xmath99 , produced by the pulling force is dissipated as thermal energy , with the corresponding increase of the entropy of the universe .
the displacement of chains articulated bodies formed by rigid links and with no internal sources of energy , is an inherently irreversible process .
there are links of the chain that vary their speeds abruptly , and these processes destroy mechanical energy @xcite .
each link in the chain that is , abruptly , placed into motion can be regarded as an inelastic collision , producing a loss of mechanical energy and making the process irreversible @xcite .
the description of the process in reference s@xmath62 in standard configuration is also very instructive for the illustration of the principle of relativity at work .
william thomson ( who became lord kelvin ) proposed a method to study the dependence of the internal energy of a gas with its volume . together with joule , they carried out a series of experiments using a device sketched in fig .
[ fig : joth - tr - fig01 ] @xcite . a cylindrical tube , with cross - section of area @xmath100 , is divided into two parts by a porous - plug . in the joule - kelvin ( or porous plug )
process a gas ( usually a real gas ) , of mass @xmath4 , occupying the initial volume @xmath101 , on the left hand side , and pressure @xmath102 , is forced to pass through a porous plug or a valve against a final constant pressure @xmath103 , reaching a final volume @xmath104 , on the right hand side . during the process the pressures on the left and on the right ( see figure ) are kept constant but the gas temperature may change from @xmath105 to @xmath106 .
the joule - thomson or porous - plug ( pp ) experiment.,width=340 ] there are three forces acting on the gas : the left side force , @xmath107 , the right side force , @xmath108 , and the zero displacement force , @xmath109 , exerted on the gas by the porous plug @xcite .
the displacements are _ = ( v_/ a , 0 , 0 ) , _ = ( v_/ a , 0 , 0 ) , _ n= ( 0,0,0 ) , _
cm= ( x_cm , 0,0 ) .
we denote by @xmath73 the horizontal displacement of the centre - of - mass , @xmath110 during the time interval @xmath41 , the duration of the process .
for the sake of generality let us consider initial and final centre - of - mass velocities denoted by @xmath111 ( in a more restricted analysis they could be simply set to zero ) .
the impulse momentum equation , @xmath112 and the centre - of - mass equation , @xmath113 yield \ { rl m ( v_f - v_i ) & = ( p_a - p_a + n ) t_0 + 12 m ( v_f^2 - v_i^2 ) & = ( p_a - p_a + n ) x_cm . .
[ yu70yh ] if the initial and final centre - of - mass velocities are zero , both equations lead to the same equation for @xmath61 , namely n= ( p_- p _ ) a [ nvbg ] the force exerted upon the gas by the porous - plug is directed to the left because @xmath114 .
now we consider the energy equation , @xmath115 and assume zero centre - of - mass velocity @xcite , hence eq .
( [ nvbg ] ) holds . in such case , \
{ rl 0 & = ( p_a - p_a + n ) t_0 + u & = p_v_- p_v_+ q . .
[ yu70yo ] the joule - thomson experiment is carried on under adiabatic conditions , _
i.e. _ the device ( cylinder and pistons ) is thermally isolated so there is no heat involved in the process . in this case the second eq .
( [ yu70yo ] ) reduces to u_- u_= p_v_- p_v_and , since the enthalpy is related to the internal energy through @xmath116 , one concludes that the process is isoenthalpic : @xmath117 . we can finally show that the entropy of the universe increases in this irreversible process . from @xmath118 , for an isoenthalpic process s_h
= -vt p_h > 0 since @xmath119 is always negative ( the pressure decreases in the process ) .
the joule - thomson experiment is a typical thermodynamical process in which all external forces have zero resultant ( and information resulting from the centre - of - mass equation ) . if the process is studied in reference frame s@xmath62 , moving with velocity @xmath32 along the horizontal direction , the porous - plug moves with velocity @xmath120 and the result set of equations is \
{ rl 0 & = ( p_a - p_a + n ) t_0 + u & = p_v_- p_v_- v(f_l - f_r + n ) t_0 + q , .
[ yu70yosp ] totally compatible with the equations in reference s.
the centre - of - mass equation , resulting from the integration of the newton s second law , and the first law of thermodynamics , establishing the energy balance in a process , correspond to two absolutely different physical hypothesis .
both equations are valid but their applicability deserves special attention .
we addressed this issue , in a pedagogical way , by firstly presenting the equations and then applying them to selected examples .
situations for which there is mechanical energy production are particularly interesting .
these include , among others , problems related with human activities ( a person jumping , climbing , cycling , etc . ) or with living beings ( a horse pushing a chariot , etc . ) and with thermal engines ( a car that accelerates due to an internal combustion , etc . ) , which are sometimes incompletely or even incorrectly addressed .
this was actually a motivation for the present work where we have discussed interesting physical situations ranging from mechanics to thermodynamics . together with the impulse - momentum vector equation we used a scalar energy equation , which is either the centre - of - mass equation or the first law of thermodynamics , emphasizing the additional information that can be obtained from the second one with respect to the other .
we also paid some attention to the frame transformation : the set of equations appropriate to discuss a problem provide the same information irrespective of the chosen inertial reference frame ( principle of relativity ) .
this was illustrated using various examples .
of course , the choice of the reference frame is usually determined by the simplicity of the resulting formalism .
intrinsic dissipative processes , such as processes with chains , should be related with thermodynamics when they are solved . when a chain is unwrapped or raised , a significant part of the work provided by the external force that displaces the chain is dissipated as heat , an effect that deserves formalization into an equation , and which is not included when attention is only paid to the mechanical part of the problem .
we also considered a couple of examples involving forces acting without displacing their application points . in a certain reference frame , such a zero displacement force produces no work but it does have an impulse . in another inertial frame
there might be both impulse and work . whenever pertinent we discussed other thermodynamical aspects , namely the irreversibility of some processes in connection with the second law .
on the contrary , we have discussed the forces that act in thermodynamical processes , something that is really uncommon in thermodynamical textbooks .
for instance in the joule - kelvin process the porous plug exerts a force on the gas and , even though it does not do any work , it plays a crucial role in the description of the process . through the discussion of the five examples presented in this article we aimed at clarifying the additional information provided by the more general first law of thermodynamics with respect to newton s second law .
99 j gmez , c fiolhais , m fiolhais , _ the cartesian diver and the fold catastrophe _ , am .
j. phys * 70 * 710 - 714 ( 2002 ) j gmez , r valiente , c fiolhais , m fiolhais , _
experiments with the drinking bird _ , am .
j. phys * 71 * 1257 - 1263 ( 2003 ) j gmez , c fiolhais , m fiolhais , _ toys in physics lectures and demonstrations a brief review _ , phys .
edu . * 44 * 53 - 64 ( 2009 ) c m penchina , _ pseudowork - energy principle _ , am .
* 46 * , 295 - 296 ( 1978 ) j w jewett , jr _ energy and the confused student i : work _ , phys .
46 * , 38 - 43 ( 2008 ) j w jewett , jr _ energy and the confused student v : the energy / momentum approach to problems involving rotating and deformable systems _ , phys .
46 * , 269 - 274 ( 2008 ) h s leff , a. j. mallinckrodt , _ stopping objects with zero external work : mechanics meets thermodynamics , _ am .
j. phys . * 61 * , 121 - 127 ( 1993 ) j a g mcclelland , _ a very persistent mistake _ , phys .
educ . * 46 * 469 - 471 ( 2011 ) j w jewett , jr , _ energy and the confused student iv : a global approach to energy _ , phys .
46 * , 210 - 217 ( 2008 ) c e mungan , _ thermodynamics of a block sliding across a frictional surface _ ,
45 * 288 - 291 ( 2007 ) a j mallinckrodt , h. s. leff , _ all about work , _ am . j. phys . * 60 * 356 - 365 ( 1992 ) p atkins , _ the laws of thermodynamics . a very short introduction _ , oxford u. press , oxford uk 2010 , pp . 65 - 72 h erlichson , _ internal energy in the first law of thermodynamics _ , am . j. phys .
* 52 * 623 - 625 ( 1984 ) u besson , _ work and energy in the presence of friction : the need for a mesoscopic analysis _ , * 22 * 613 - 622 ( 2001 ) b a sherwood , _ pseudowork and real work _ , am . j. phys . *
51 * , 597 - 602 ( 1983 ) r p bauman , _ physics that textbook writers usually get wrong
. i. work _ ,
* 30 * , 264 - 269 ( 1992 ) a b arons , _ developing the energy concepts in introductory physics _ , phys .
* 27 * 506 - 517 ( 1989 ) j koser , _ laboratory activity : specific heat by change in internal energy of silly putty _ , phys .
* 49 * 574 - 575 ( 2011 ) p a smith , _ the first lecture on entropy , _ am . j. phys .
* 51 * , 1049 ( 1983 ) h s leff , _ removing the mystery of entropy and thermodynamics - part ii _ , phys .
50 * 87 - 90 ( 2012 ) r j dufresne , w j gerace and w j leonard , _
springbok : the physics of jumping _ , phys .
teach . * 39 * 109 - 115 ( 2001 ) w h bernard , _ internal work : a misinterpretation _ , am . j. phys .
* 52 * , 253 - 254 ( 1984 ) j a g mcclelland , _ misconceptions in mechanics and how to avoid them _ phys .
* 20 * 159 - 162 ( 1985 ) c a de sousa , _ nonrigid systems : mechanical and thermodynamic aspects _ , eur
. j. phys .
* 23 * 433 - 440 ( 2002 ) d morin , _ introduction to classical mechanics . with problems and solution .
_ cambridge university press .
cambridge 2008 , pp . 167 - 172 j m knudsen , p g hjorth , _ elements of newtonian mechanics . including nonlinear dynamics _ , springer heidelberg 2000 m w zemansky , r h dittman , _ heat and thermodynamics : an intermediate textbook , _
, mcgraw - hill international editions ( 1997 ) s penn , _ comment on : thermodynamics and classical relativity _ , am .
* 35 * , 780 ( 1967 ) r j tykodi , _ thermodynamics and classical relativity _ , am . j. phys .
* 35 * , 250 - 253 ( 1967 ) | we present and discuss a selected set of problems of classical mechanics and thermodynamics .
the discussion is based on the use of the impulse - momentum equation simultaneously with the centre - of - mass ( pseudo - work ) equation or with the first law of thermodynamics , depending on the nature of the problem .
thermodynamical aspects of classical mechanics , namely problems involving non - conservative forces or variation of mechanical energy are discussed , in different reference frames , in connection with the use of one or the other energy equation , and with the compliance of the principle of relativity .
0.55 cm plus 1pt minus 1pt = 1500 * from mechanics to thermodynamics
analysis of selected examples * 0.4 cm j. gmez@xmath0 , m. fiolhais@xmath1 0.1 cm _
@xmath2 departamento de fsica aplicada _
+ _ universidad de cantabria _ + _ e-39005 santander , spain _ + 0.1 cm _
@xmath3 departamento de fsica and centro de fsica computacional
_ + _ universidade de coimbra _ + _ p-3004 - 516 coimbra , portugal _ |
during their initial outburst , gamma - ray bursts ( grbs ) are the most luminous source of electromagnetic radiation in the universe , releasing isotropic - equivalent energies of the order of @xmath8@xmath9 erg in @xmath4-rays alone on timescales of @xmath10@xmath11 s. the favoured mechanism to produce such a large amount of energy on such a short timescale involves accretion onto a newly formed black hole produced either in a merger of two compact objects or from the gravitational core collapse of a massive star
. there appear to be two distinct populations of grbs , distinguished by the duration of their prompt emission and their spectral properties @xcite , whereby long grbs ( prompt emissions lasting @xmath12 s ) are typically softer than short grbs ( @xmath13 s ) @xcite . the most robust model for long
, soft grbs is the collapsar model @xcite , in which a grb results from the energy released during the death of a massive star .
evidence supporting this model includes the large number of long grbs located in regions of active star - formation ( e.g. * ? ? ?
* ; * ? ? ?
* ) , and the association of several nearby grbs with supernovae , e.g. grb 980425/sn1998bw @xcite , grb 030329/sn2003dh @xcite and grb 060218/sn2006aj @xcite there is evidence to suggest that collimated outflows reduce the amount of energy emitted by a grb from isotropic emission by @xmath14 orders of magnitude , where the beam - corrected energy is clustered at @xmath15@xmath16 erg ( frail et al . 2001 ; bloom , frail & kulkarni 2003 ) .
bipolar jets are a natural consequence of accretion onto a black hole and are observed in many high - energy astrophysical phenomena ( e.g. agn , pulsars , micro - quasars ) , where the ejecta take the easiest exit path , along the rotational axis .
a signature of highly relativistic jets is an achromatic break in the broadband afterglow light curves , caused by the expansion of the radiation beaming angle as the bulk lorentz factor , @xmath17 , decreases ( rhoads 1999 ; sari , piran & halpern 1999 ) . once the lorentz factor has decreased such that @xmath18 , where @xmath19 is the half jet opening angle , the rate of decay increases , resulting in a break in the light curve .
one of the scientific aims of the grb dedicated mission , _ swift _
@xcite , was to pin down the range of collimation present in grbs .
_ swift _ is a rapid - response mission with multi - wavelength capabilities designed specifically to study the early time evolution of grbs and their afterglows .
the spacecraft is equipped with three telescopes ; the burst alert telescope ( bat ; * ? ? ?
* ) has a 15150 kev energy range with a partially coded response up to 350 kev , the x - ray telescope ( xrt ; * ? ? ?
* ) which covers the 0.310 kev energy range , and the ultra - violet and optical telescope ( uvot ; * ? ? ?
* ) , which can observe in six colours and a broadband white band filter , covering the wavelength range from 1600 to 6500 . _ swift _
s slewing capabilities provide grb afterglow light curves from @xmath20 s after the trigger , and well - sampled x - ray and optical light curves have revealed far fewer achromatic breaks than had been expected prior to launch ( ghirlanda , ghisellini & lazzati 2004 ; obrian et al .
this may be an indication that jet breaks occur later than previously believed , at a time when the afterglow is no longer detected above the background level . in the x - ray band
this is at around @xmath21 s after the prompt emission .
the time of the jet break , @xmath22 , increases with the jet opening angle , @xmath19 , as @xmath23 erg , for a constant density circumburst medium with density @xmath24 and grb kinetic energy @xmath25 .
large average jet opening angles , small circumburst densities or large grb kinetic energies may , therefore , all contribute to delaying the time of the jet - break to when the afterglow is no longer detectable by _ swift _ , although @xmath22 is only weakly dependent on the latter two parameters listed .
grb 061007 is an example of a burst that showed no evidence of a break in the x - ray or uv / optical afterglow from 80 s after the prompt emission , until it was no longer detected above background , at @xmath26 s. prompt observations by ground based observatories measured the redshift of this burst to be z=1.26 ( osip , chen & prochaska 2006 ; jakobsson et al .
2006 ) . at this redshift ,
grb 061007 had one of the highest isotropic - equivalent energies ever seen , releasing @xmath27 erg in @xmath4-radiation alone , with an optical afterglow comparable to that of grb 990123 @xcite , with @xmath28 mag at 71.7 s after the bat trigger . in this paper
we describe in detail the @xmath4-ray , x - ray and optical observations taken by _
swift_. from the high resolution early time data we constrain the type of environment in which the burst occurred and determine the conditions that distinguish this burst from others . in section [ sec : obs_anlysis ]
we describe the data reduction and analysis techniques used , followed by an analysis on the possible physical parameters and emission mechanisms that could explain the observations in section [ sec : fireballmod]-[sec : ssc ] . in these subsections
we investigate both a wind and ism circumburst environment , as well as spherical - equivalent emission and collimated emission .
we also use the afterglow spectral energy distribution ( sed ) to analysis the properties of the circumburst medium of grb 061007 in section [ sec : host ] .
our results are discussed in section [ sec : model_disc ] and in section section [ sec : conc ] we summarise our conclusions . throughout the paper all errors quoted
are @xmath29 unless otherwise specified and the temporal decay rates , @xmath30 , and spectral indices , @xmath31 , are defined such that @xmath32 and @xmath33 , respectively .
luminosities and the energy release are calculated assuming a standard cosmology model with @xmath34 km s@xmath35 mpc@xmath35 , @xmath36 and @xmath37 .
the bat on board _ swift _ triggered on grb 061007 on @xmath38 october , 2006 at 10:08:08 ut , resulting in an immediate slew to point the narrow field instruments in the direction of the burst . from here on we shall refer to the time of the trigger as t. the refined bat position for grb 061007 is ra = @xmath39 , dec = @xmath40 ( j2000 ) with an uncertainty ( systematic and statistical ) of @xmath41 at 90% containment @xcite .
the bat light curve consists of three large and distinct peaks and is shown in fig .
[ fig : batlc ] .
the time - interval over which 90% of the 15150 kev @xmath4-radiation is emitted was @xmath42 s , although emission above the background level was detected out to @xmath43t+500 s , which is most pronounced in the soft energy band ( top panel of fig .
[ fig : batlc ] ) .
the time lag @xcite between the peak emission in the 50100 kev and the 1525 kev energy bands was @xmath44 ms , and between the 100350 kev and 2550 kev bands the lag was @xmath45 ms .
the first xrt observation was a centroiding image taken 80.4 s after the bat trigger , and a bright uncatalogued source was found at ra = @xmath46 , dec = @xmath47 , with a 90% containment radius of @xmath48 @xcite .
at 86.6 s after the bat trigger the xrt began taking data in window timing mode ( wt ) , in which data are stored with a 1.7 ms time resolution and 1-dimensional imaging . at t+2000
s the count rate fell to below @xmath49 count s@xmath35 and the automated sequence switched to photon counting ( pc ) mode , in which the time resolution is 2.5 ms and both imaging and spectroscopic information are available .
the x - ray afterglow was detected for @xmath50 s after the bat trigger before it fell below the detection threshold . during the first 2200 s of observations the count rate of the burst was high enough to cause pile - up in both wt and pc mode data . to account for this effect
, the wt data were extracted in a rectangular 40@xmath5120-pixel region with a 9@xmath5120-pixel region excluded from its centre .
the size of the exclusion region was determined following the procedure illustrated in @xcite .
the standard correction for pile - up was applied to the pc data , as described by @xcite . due to a large amount of pile - up in the first 30 s of wt observations ,
these data were not used in our spectral analysis .
the 0.310 kev light curve is well fit by a power law with decay index @xmath52 from t+80 s until at least t+@xmath21 s , when it reached the detectability threshold of xrt ( see fig .
[ fig : comblc ] ) .
we also considered the presence of a break in the light curve , resulting from either the edge of the jet coming into the line of sight or the migration of the cooling frequency , @xmath53 , through the observing energy bands .
the cooling frequency defines the frequency above which electrons no longer lose a significant fraction of their energy to radiation .
the cooling frequency is expected to evolve as @xmath54 for an ism - like circumburst environment ( sari , piran & narayan 1998 ) , and as @xmath55 for a wind - like circumburst environment @xcite , and as it migrates through the observing energy band the drop in radiation at @xmath56 produces a break in the light curve . to determine the @xmath57 lower limit on the time at which a jet or cooling break could have occurred , we applied a broken power law fit modelled on the cooling break , for which @xmath58 @xcite , and on a jet break , where @xmath59 @xcite , and in each case determined the earliest time that the break could occur within the observing time interval without deviating the goodness of fit by more than @xmath57 from the best - fit power law model .
our analysis indicates that a cooling break could not have occurred earlier than @xmath60 s after the bat trigger to 99% confidence , and any jet break after the onset of the afterglow observations must have happened at @xmath61 s. a power law fit to the early wt spectrum ( @xmath62t+600 s ) , where the galactic column density is fixed at @xmath63 @xmath64 @xcite , gives a best - fit host galaxy column density of @xmath65 @xmath64 at a redshift of @xmath66 for a spectral index @xmath67 .
spectral analysis of the pc data during the time interval t+5000 s to t+25,000 s gives a best fit host galaxy column density @xmath68 @xmath64 , and spectral index @xmath69 , consistent with the wt data , and therefore indicating that there is no spectral evolution in the x - ray data .
the uvot began observing grb 061007 71.7 s after the bat trigger , at which point it took a 9 s settling exposure in the @xmath0-filter .
this was followed by a 100 s exposure in the white light filter ( @xmath70@xmath71 ) and a 400 s exposure in the @xmath0-band .
after this the automated sequence rotated twice through the uvot filters taking a series of short exposures ( @xmath72 , white , @xmath73 ) ( 10 s for @xmath74 and white filters and 20 s for the rest ) .
the central wavelengths of the optical and uv filters are @xmath75 , @xmath76 @xmath77 , @xmath78 , @xmath79 and @xmath80 for the @xmath0 , @xmath74 , @xmath81 , @xmath82 , @xmath83 and @xmath73 , respectively .
a further 400 s @xmath0-band exposure and 100 s white exposure were taken , followed by a further series of rapid rotations around the filter wheel up to 2260 s after the trigger . with the exception of the second 100 s white band exposure ,
all observations up to this time were taken in event mode , which has a time resolution of 11 ms , and after this image mode exposures were taken .
the refined position of the afterglow is ra = @xmath84 , dec = @xmath85 to a certainty of @xmath86 ( j2000 ) .
the galactic extinction along this line of sight is @xmath87 ( schlegel , finkbeiner & davis 1998 ) . for all short exposures ( 10 s or 20 s ) and exposures taken later than t+2260 s photometric measurements were obtained from the uvot data using a circular source extraction region with a @xmath88 radius for the @xmath0 , @xmath74 and @xmath81 optical filters and a @xmath89 radius for the three uvot ultra - violet filters to remain compatible with the current effective area and zeropoint calibrations .
the background was measured in a source - free region near to the target using an extraction radius of @xmath90 .
the first @xmath0-band settling exposure and first white - band exposure were affected by coincidence loss to a degree that is outside the photometrically calibrated range of uvot ; for a @xmath0-band settling exposure complete saturation within a standard @xmath88 aperture implies that this afterglow is brighter than @xmath91 for the whole of the 9 s exposure .
during the first white band exposure the source is so bright that relative photometry can be obtained using a very large aperture , which includes the wings of the point spread function , where coincidence loss is not a problem .
the exposure was divided into 5 s bins , and source counts were obtained from an extraction region of radius @xmath92 . to calibrate this photometry , observations of the white dwarf standards gd50 and
hz2 were taken with uvot and the white filter on october 18@xmath93 2006 . from these data we determined the relationship between observed count rate within the @xmath94 aperture and corrected count rate within a standard @xmath88 uvot source aperture , which we assume to be linear . to apply this calibration to grb 061007
we apply constant offset to this relation to account for the different background count rate between the grb 061007 observations and the white dwarf standard count rate observations .
this offset was chosen to match the mean white - band count rate in the saturated exposure to the white filter light curve extrapolated from later times ; we caution that calibration of this first white exposure is crude and thus the apparent steepening of the light curve towards the end of the exposure may not be real . however , the overall behaviour of a steadily decreasing count rate over the course of the exposure is not in doubt . for the first 400 s @xmath0-band exposure a higher resolution light curve
was created by splitting it into 4 s bins , and the second into 20 s bins , consistent with the majority of the exposures during that time . in order to get the best measurement of the optical temporal decay we created a single uv / optical light curve from all the uvot filters .
to do this the light curve in each filter was individually fitted to find the corresponding normalisation , and this was then used to renormalised each light curve to the @xmath0-band .
the combined uv and optical light curve , from t+85 s up to t+@xmath95 s , is best fit with a power law decay with temporal index @xmath3 .
the combined light curve shows no apparent colour evolution , illustrated in fig .
[ fig : comblc ] .
the combined uvot light curve is shown next to the 0.310 kev and 15350 kev light curve in fig . [
fig : comblc ] , and a close up of the uvot light curve from t+1000 s to t+30,000 s is shown in the inset .
the log - log scale in this figure shows the decaying light curve in the 15 - 350 kev energy band from 80500 s much better than the linear scale in fig [ fig : batlc ] .
the magnitude along the left axis corresponds to the @xmath0-band , and the count rate along the right axis applies to the x - ray and @xmath4-ray light curves .
the prompt emission from grb 061007 also triggered konus - wind and suzaku , and the afterglow was monitored by ground based telescopes , which carried out both spectroscopic and photometric observations .
grb 061007 was measured to be at z=1.26 @xcite , for which the rest - frame isotropic energy release and maximum luminosity in the 20 kev10 mev is @xmath96 erg , and @xmath97 erg s@xmath35 @xcite .
the optical afterglow was also detected by rotse at 26.4 s after the bat trigger at a magnitude of 13.6 mag in an unfiltered exposure @xcite , and the faulkes telescope south measured an @xmath98-band magnitude of @xmath99 142 s after the trigger @xcite , consistent with the uvot observations and indicative of a significant brightening during the first 90 s. the _ swift _
observations from t+80 s show that by this time the afterglow had entered a constant power law decay ( see fig .
[ fig : comblc ] ) .
radio observations were also performed 11.24 days after the prompt emission , which resulted in no detection of a radio source , and a flux measurement of @xmath100jy at 19 ghz @xcite . at @xmath101
s , after the last flare in the bat light curve ( fig .
[ fig : batlc ] ) , the bat light curve appears to decay smoothly and monotonically out to @xmath43t+500 s , after which the flux becomes indistinguishable from the background in the bat energy range .
a power law fit to the bat light curve from t+100 s to t+500 s provides an acceptable fit ( @xmath102 for 29 dof ) with a best - fit decay index of @xmath103 , which is consistent with the decay observed in both the uv / optical and x - ray light curves .
furthermore , a power law spectral fit to this last part of the bat light curve , beyond @xmath43t+100 s , gives a best fit energy spectral index @xmath104 , in fairly good agreement with the spectral indices determined for the x - ray and optical afterglows .
this supports a scenario in which the @xmath4-emission at this epoch is generated by the same radiation mechanism as the x - ray and uv / optical afterglow .
using the full range of data available with _ swift _ an sed of the afterglow was produced using optical , x - ray and @xmath4-ray data , which were normalised to an instantaneous epoch at 600 s after the bat trigger . at this time
there are data available from all three instruments , and good colour information from uvot . for each of the uvot lenticular filters the tool uvot2pha ( v1.1 )
was used to produce spectral files compatible with xspec ( v12.2.1 ) from the exposures taken closest to t+600 s. the count rate in each band was then normalised to the estimated value at t+600 s determined from the best - fit models to the light curves .
the xrt spectrum was produced from wt mode data in the 0.510 kev energy range , omitting the first 30 s of data , which suffered from pile - up .
this was scaled to the corresponding count rate at t+600 s. for the bat spectrum , data in the 15150 kev energy range were taken from the time interval t+100 s
t+500 s , during which the bat light curve decayed at a rate consistent with x - ray and uv / optical afterglows and had a similar spectrum .
this spectrum was normalised to t+600 s in the same way as the xrt spectrum .
the sed is shown in fig .
[ fig : sed ] .
the lack of any break in the x - ray and uv / optical light curves up to @xmath60 s places the cooling frequency either above the x - ray band or below the optical at t+600 s. both energy bands will , therefore , have the same spectral index , and the sed was , thus , fit with a power law emission spectrum , and two dust and gas components were included to model the galactic and host galaxy photoelectric absorption and dust extinction .
the column density and reddening in the first absorption system were fixed at the galactic values ( see sections [ sec : xrt ] and [ sec : uvot ] ) .
the second photoelectric absorption system was set to the redshift of the grb and the neutral hydrogen column density in the host galaxy was determined assuming solar abundances .
the dependence of dust extinction on wavelength in the grb host galaxy was modelled on three extinction laws taken from observations of the milky way ( mw ) , the large magellanic cloud ( lmc ) and the small magellanic cloud ( smc ) .
the greatest differences observed in these extinction laws are in the amount of far uv ( fuv ) extinction , which is greatest in the smc and least in the mw , and in the strength of the @xmath105 absorption feature , which is most prominent in the mw and negligible in the smc . to model these extinction laws the parameterisations from @xcite were used , where @xmath106 , and 3.16 for the galactic , smc and lmc extinction laws respectively .
an important difference between the magellanic clouds and the milky way is the average metallicity of these galaxies , which is larger in the case of the milky way .
the value that we determine in our spectral analysis for @xmath107 is an equivalent neutral hydrogen column density that results from the amount of soft x - ray absorption in the spectrum , which will , therefore , depend on the abundances assumed in the spectral fit .
the smc and lmc have a metallicity that is @xmath108 and @xmath109 that of the milky way @xcite , and this will result in an underestimation of the neutral hydrogen column density by a factor of @xmath110 and @xmath111 , respectively , if solar abundances are assumed .
this is taken into account when we discuss the results from our spectral analysis in section [ sec : host ] .
[ cols= " < , < , < , < , < , < " , ] @xmath112 equivalent hydrogen column density assuming solar abundances at the grb host galaxy . at a redshift of @xmath66
the beginning of the lyman-@xmath30 forest is redshifted to an observer - frame wavelength of @xmath113 , which falls squarely on the @xmath82 filter , the reddest of the uv filters .
the unknown equivalent width of lyman-@xmath30 absorption from the host galaxy of the grb and the variation of the ly-@xmath30 forest along different lines of sight makes it difficult to model this additional source of absorption in the sed .
we , therefore , exclude the three data points affected by lyman-@xmath30 from our spectral analysis .
the results from our spectral analysis are summarised in table [ tab : sedpowfit ] .
the fitted value of the spectral index depends , for the most part , on the amount of optical / uv extinction , which in turn is affected by the dust extinction model used .
similarly , the amount of soft x - ray absorption fitted to the model will affect the best - fit spectral index . in all three cases
an absorption and extinction system is measured at the grb host galaxy , which is smallest in an smc model and largest in a mw model .
the lack of any evolution in @xmath114 or @xmath31 in the x - ray data ( see section [ sec : xrt ] ) , indicates that @xmath115 has already moved through the x - ray band , and , therefore , the neutral hydrogen column density measured is indicative of the amount of x - ray absorption at the grb host galaxy rather than an artifact of evolution in the spectral index .
the optical , x - ray and @xmath4-ray spectrum is best fit by a model that assumes the grb host galaxy to have an lmc extinction law for the grb host galaxy , which gives a @xmath116 for 308 degrees of freedom ( dof ) with a null - hypothesis probability of @xmath117 . models with an smc and mw extinction law at the host galaxy were rejected with 99.96% confidence and 100% confidence , respectively .
grb 061007 is one of the brightest bursts to be observed by _
, with an isotropic energy released in @xmath4-rays that is second only to grb 050904 @xcite and an optical flux comparable to that observed in grb 990123 @xcite .
the similarity of the afterglow of grb 061007 in the flux decay of @xmath4-ray , x - ray and uv / optical bands , and the lack of features in x - ray and optical light curves distinguishes this burst from others .
optical afterglows typically decay as power laws , although usually with decay indices @xmath118 ( zeh , klose & kann 2006 ) , and in the x - ray , the grb light curve is usually characterised by several discrete segments , with a canonical shape that is made up of three , and sometimes four , power law segments with different decay indices @xcite .
the temporal behaviour of grb 061007 is , therefore , not typical of other _ swift _ grbs , of which only @xmath119 observed promptly by the xrt ( within the first few minutes of the bat trigger ) show a constant decay with no additional features such as flaring @xcite ( e.g. grb 051117b , grb 060403 ) .
when only considering long grbs , the fraction of smooth decaying , featureless x - ray afterglows is even more infrequent , amounting to only @xmath120% of the long grb population . according to the fireball model , the temporal and spectral indices of the afterglow depend on the properties of both the grb ejecta and the circumburst environment , which determine the location of the characteristic synchrotron frequency , @xmath121 and the cooling frequency , @xmath53 , with respect to the observing band @xcite . at any one time
the location of these frequencies are dependent on the energy distribution of electrons in the circumburst environment defined by the energy index @xmath122 , the density and density profile of the circumburst medium , the amount of kinetic energy in the outflow , and on the fraction of energy in electrons and in the magnetic field , @xmath123 and @xmath124 , respectively . @xmath123 and @xmath124 are typically assumed to remain constant with time , and therefore the evolution of @xmath121 and @xmath53 , and thus the afterglow decay and spectral indices , depend on the density and density profile of the circumburst medium and on the amount of kinetic energy released by the grb .
the lack of a break in the light curves provide constraints on the spectral regime that the x - ray and uv / optical afterglow are in , and on the frequency of @xmath121 and @xmath53 , neither of which could have crossed the observing window from t+80 s to t+@xmath60 s. from the white filter we determine that around half the counts detected in the white band observations correspond to the @xmath0 and @xmath74 filter .
the lack of a break in the first white filter exposure , therefore , indicates that at t+80 s , @xmath121 was no bluer than the @xmath74-band filter .
this corresponds to an upper limit on the characteristic synchrotron frequency @xmath125 hz . from the uv / optical data we also determine that at t+80 s the flux density in the @xmath0-band , corresponding to @xmath126 hz
, is @xmath127 mjy , after correcting for absorption at the host galaxy . with these constraints in place
we investigate the various conditions that can explain the temporal and spectral properties of grb 061007 , all of which can be categorised into two parent models that differ fundamentally in the effect of collimation on the observations . in the first model we assume no jet break to have occurred before t+@xmath128
s , based on the constraints we obtained in section [ sec : xrt ] , and the observed emission is , therefore , indistinguishable from isotropic emission .
secondly , we investigate a model in which the jet break occurs before our first x - ray and uv / optical observations at t+80 s such that the assumption of isotropic emission no longer applies . the simplest model to consider is one in which the observed afterglow emission follows the flux decay and spectral evolution as would be expected from synchrotron radiation emitted from an isotropic outflow .
as already mentioned , one of the important factors that determines the evolution of the defining synchrotron frequencies is the density profile in the surrounding circumburst material , and we therefore consider both a wind - like circumburst environment , such as would be expected in the vicinity of a massive star , and a constant density profile , such as in the ism . for simplicity
, we assume the density profile of the circumburst medium to have the form @xmath129 , where @xmath130 and @xmath131 corresponds to an ism and wind - like profile , respectively .
the relation between @xmath30 and @xmath31 for synchrotron emission at different intervals in the synchrotron spectrum are defined by a set of closure relations , both for a constant density medium @xcite and a wind density profile @xcite .
assuming spherically equivalent emission , the standard closure relations do not fully satisfy the observed spectral and temporal indices of grb 061007 for either an ism or a wind - like circumburst medium , where @xmath132 and @xmath133 .
this could be the result of additional components unaccounted for in the closure relations , or an oversimplification of the assumptions made when outlying the closure relations e.g. the density profile may not be distributed as @xmath134 . for both a wind and an ism circumburst density profile our observations are in closest agreement with a model
in which the observed synchrotron emission is in the slow cooling phase ( @xmath135 ) , where the uv / optical and x - ray bands are in the same slow - cooling regime ( @xmath136 ) .
the decay index is @xmath137 steeper than expected for a constant circumburst density profile , and requires a fairly soft electron distribution index @xmath138 . in the case of a wind circumburst environment
the decay index is shallower than predicted by this model , by a difference @xmath139 .
it is possible that the density profile in the environment surrounding grb 061007 lies somewhere between the two scenarios considered here , where @xmath140 .
alternatively , the shallowness of the decay index for a wind case could be explained if energy injection is involved .
this would require a late - time source of energy into the forward shock that would slow down the cooling rate of the excited electrons and , thus , slow down the afterglow decay .
either way , the lack of a cooling break for the duration of _ swift _
observations out to @xmath21 s requires consideration , and the conditions necessary to maintain @xmath53 above or below the observing frequency range need to be investigated .
the evolution of @xmath53 with time is dependent on the density profile of the circumburst medium , whereby for @xmath141 @xmath53 will increase with time , whereas for @xmath142 it will decrease @xcite .
the conditions needed for the cooling frequency to not have crossed the observing window , therefore , differ for an ism and wind circumburst density profile . for an ism medium
the cooling frequency has to be maintained above the x - ray band for up to t+@xmath60 s , whereas in a wind scenario the cooling frequency needs to be above the x - ray band at the start of the x - ray observations , at t+80 s. either way the cooling frequency must lie above the peak x - ray frequency out to at least t+@xmath60 s for a cooling break not to cover in the x - ray observations ( see section [ sec : xrt ] ) .
this corresponds to the requirement that @xmath143 hz at t+@xmath60 s. in the next section we investigate the consistency of a spherical emission model with the data for both an ism and wind medium where @xmath144 . for an ism - like circumburst environment ,
the temporal and spectral flux decay indices are related as @xmath145 . for a value of @xmath146
, we expect @xmath147 , which is shallower than the observed decay rate of @xmath148 . however , this deviation from the relation is small , and could be accommodated if the density profile is slightly different from the uniform density we have assumed . in this regime @xmath149 , where @xmath122 is the electron energy index .
thus , for @xmath150 the electron energy index is in the range @xmath151 . we therefore take @xmath152 .
these relations hold as long as the characteristic synchrotron frequency , @xmath121 , lies below the optical band , i.e. @xmath125 hz , at t+80 s , as discussed previously . for a constant density medium
in which synchrotron radiation is the dominating emission mechanism , the characteristic synchrotron frequency , the cooling frequency and the peak density flux are given by the following equations @xmath153
@xmath155 @xcite , where @xmath124 is the fraction of energy in the circumburst magnetic field , @xmath156 is the fraction of energy in electrons , @xmath157 is the kinetic energy in the grb in units of @xmath158 , @xmath159 is the time since the prompt emission in units of days , @xmath24 is the density of the surrounding circumburst medium , @xmath160 is the compton parameter , and @xmath161 is the luminosity distance in units of @xmath162 cm .
typical values for the fraction of energy in the cicumburst magnetic field and in the electrons are @xmath163 and @xmath164 , respectively . for a spectral index @xmath165 , @xmath166 is constant for @xmath144 . at t+80 s
@xmath127 mjy at @xmath126 hz , and therefore @xmath167 . by using this together with eq .
[ eq : num ] and [ eq : fmax ] we obtain : @xmath168 in eq . [ eq : nuc ] we apply the condition that at @xmath60 s after the prompt emission @xmath53 must still be greater than @xmath169 hz , and substituting in eq .
[ eq : density ] gives @xmath170 considering now the characteristic synchrotron frequency , we know that @xmath121@xmath171 hz , and this reduces eq .
[ eq : num ] to @xmath172 this , and the inequality given in eq .
[ equal1 ] provide an upper limit to the value of @xmath173 in terms of @xmath174 , such that @xmath175 the above equation bounds us to extreme values of @xmath157 and @xmath173 .
for @xmath176 , @xmath177 , which is an unrealistically low value of @xmath124 .
even a value as small as @xmath178 requires a huge amount of kinetic energy to meet the condition that @xmath179 erg .
the amount of kinetic energy assumed is an isotropic - equivalent value which would , thus , be smaller if it were corrected for collimation .
a @xmath57 lower limit of @xmath180 s can be imposed as the earliest time that a jet break could have occurred .
@xmath22 is the time of the jet break measured from the onset of the afterglow , which we take to be equal to t. a kinetic energy @xmath181 erg and @xmath178 would require @xmath182 and @xmath183 @xmath184 for the conditions given above to be satisfied . from this
we estimate the jet opening angle using @xmath185 @xcite , where @xmath186 is in units of days .
the equation for the ic parameter is as follows : @xmath187/2\end{aligned}\ ] ] @xcite , where @xmath188 $ ] @xcite , and @xmath189 accounts for the klein - nishina correction .
for the physical parameters assumed , the ic parameter , @xmath160 , is near negligible @xmath190 , and can therefore by approximated to @xmath191 . in this case , a jet opening angle for grb 061007 would need to be @xmath192 to produce a jet break later than t+@xmath128 s. this would provide a beam - corrected kinetic energy of @xmath193 erg , which is half a solar mass and three orders of magnitude greater than usual bursts @xcite .
the energetics required by this model are unreasonably large and , therefore , an unlikely solution to explain the properties of grb 061007 . in the slow cooling regime
where @xmath144 , the spectral and temporal indices for a wind - like circumburst environment are related such that @xmath194 , which can be rejected by the observations at more than @xmath195 confidence . in order for this model to be compatible with the data an additional source of energy
is required to slow down the decay rate of the afterglow .
a continual source of energy injection could come from either the central engine itself , or late time refreshed shocks that slow down the decay of the afterglow .
the shallow decay phase observed in many xrt light curves is interpreted to be the result of energy injection , and although this phase typically lasts for @xmath196 s , there are also examples of grbs with evidence of energy injection for prolonged periods of time , and with very smooth decay slopes ( e.g. romano et al .
2006 ; huang , cheng & gao 2006 ) . taking this into account ,
an explanation whereby continual energy injection maintains the afterglow and slows down its decay over a long period of time is not unreasonable . for the purpose of this analysis the source of this additional energy
is not considered and the energy injection is simply quantified by the expression @xmath197 , where l is luminosity in units of erg s@xmath35 @xcite . a wind circumburst environment
may then be compatible with our observations if @xmath144 and @xmath198 and @xmath199 . using these best fit parameters and the known constraints provided by the data it is possible to determine the parameter space of the grb kinetic energy and the circumburst medium microphysical parameters that could reproduce the observations . for a wind environment the characteristic synchrotron frequency , @xmath121 , and flux at this frequency are given by @xmath200 @xmath201 where @xmath202 g cm@xmath35 is the density scaling such that @xmath204 @xcite .
as before the flux at @xmath205 hz is @xmath127 mjy at t+80 s , and since @xmath206 , we have @xmath207 the dependencies on the cooling frequency also vary from eq .
[ eq : nuc ] for a wind medium , where it is now defined by @xmath208 which has been modified from @xcite to include the compton parameter , @xmath160 . a new expression ,
therefore , needs to be defined regarding the conditions on the cooling frequency .
since @xmath53 increases with time , the only useful constraint is a lower limit on @xmath53 at early times .
this constraint can be determined through spectral modelling of the afterglow sed at an epoch of t+300 s , which is the weighted mid - time of the time interval over which the bat spectrum shown in fig .
[ fig : sed ] was accumulated .
this sed was fit with a broken power law with a spectral break above the bat energy range , and the change in spectral index at the break was fixed to @xmath209 to correspond to a cooling break .
taking the @xmath210 from this fit into account , the @xmath53 upper limit was then determined by shifting the spectral break to lower energies until the goodness of fit deviated by @xmath211 ( @xmath57 confidence for one interesting parameter ) .
this gave an upper limit on the cooling frequency of @xmath212 hz at t+300 s. applying this , eq .
[ eq : windnuc ] combined with eq . [ eq : winda ] reduces to @xmath213 as for an ism circumburst environment , @xmath214 , which if applied to eq .
[ eq : windnum ] gives @xmath215 and combining this expression with eq .
[ windequal ] gives @xmath216 for @xmath217 the kinetic energy in the forward shock needs to be @xmath218 erg , which when taking into account the possible jet collimation reduces to @xmath219 erg .
however , it is also necessary to include the energy injected into the fireball , which increases the total energy in the forward shock with time as @xmath220 . for @xmath221 the total energy in the forward shock at t+@xmath60 s increases by an order of magnitude to @xmath222 erg , which is similar to the amount of kinetic energy required for an ism - like circumburst environment . a wind environment
, therefore , faces the same energy budget problem as present for an ism environment in the spherical emission model discussed above the primary assumptions made in both the ism and wind scenarios are that both the x - ray and uv / optical afterglow observed are produced by the same radiation mechanism , and that the decay rate is equivalent to spherical emission . the difficulty in accommodating the observations with a standard fireball model indicate that in the case of grb 061007 either synchrotron emission is not the dominant radiation mechanism throughout the observing energy range , or that the grb outflow is collimated whereby the observed emission is no longer equivalent to isotropic emission . with this in mind
we investigate how additional radiation mechanisms and various collimation effects would alter the observed afterglow from that produced by synchrotron emission from an isotropic outflow .
the large amount of energy required to power a model with isotropic emission leads us to investigate a scenario in which the emission from the grb is highly collimated , and we once again assume the circumburst environment to be of constant density .
a very narrow jet opening angle could reduce the energy required to explain our observations by several orders of magnitude , where @xmath22 must be prior to our first afterglow observations in order to account for the lack of a break in the x - ray and uv / optical light curves at later times .
if we assume the afterglow begins no earlier than the time of trigger , this corresponds to a time @xmath223 s. within the premise that the outflow is highly collimated we consider two cases in which the cooling frequency is either above or below the observing frequencies . in these two regimes @xmath122 and @xmath31 are related by @xmath224 and @xmath225 , respectively , corresponding to @xmath226 or @xmath227 . after the jet
break the flux decays as @xmath228 @xcite , where @xmath122 is either 2 or 3 , which is steeper than the observed decay rate of @xmath229 .
however , this does not account for continual energy injection , which would reduce the afterglow decay rate . for a laterally expanding jet with constant density circumburst medium and @xmath121@xmath230 @xmath53 , the relation between the temporal decay of the grb afterglow , @xmath30 , and the spectral index , @xmath31 , post jet - break
is given by @xmath231 @xcite where @xmath232 is the energy injection parameter such that @xmath233 @xcite , and since we take the time of our jet break to be at @xmath234 s , @xmath235 , @xmath236 .
for our observed temporal and spectral indices , @xmath237 and @xmath238 , as before .
prior to the jet break eq .
[ eq : num]-[eq : fmax ] still apply , but in order take into account the effect of energy injection , the kinetic energy requires an additional factor of @xmath239 .
the time of jet break and the jet opening angle are related as @xmath240 . to keep @xmath19 as large as our observations
will allow we take @xmath234 s. at t+80 s the additional factor of @xmath239 is unity and there is no change in eq .
[ eq : num]-[eq : fmax ] . thus at t+80 s , eq .
[ eq : density ] and eq .
[ equal2 ] stand . for eq .
[ eq : nuc ] the constraint that @xmath143 at t+@xmath60 s means that the evolution of @xmath53 post jet - break has to be taken into account .
if @xmath241 is the frequency of @xmath53 at the time of the jet break , t+80 s , @xmath242 however , we are concerned with where @xmath53 lies at t+@xmath60 s. after a jet break @xmath243 , where @xmath244 is the lorentz factor of the electrons in the circumburst environment @xcite .
the cooling frequency after t+80 s is given by @xmath245 where @xmath246 is the electron lorentz factor at t+80 s. using the condition that @xmath143 hz at t+@xmath60 s , for @xmath237 eq .
[ eq : postjetnuc ] reduces to @xmath247 in the same way that we determined the inequality given in eq .
[ eq : epsilonblims ] , we combine the above equation with eq .
[ equal2 ] and determine that @xmath248 for @xmath249 erg , @xmath250 , and if we take this upper limit as the value of @xmath124 , @xmath251 .
although the assumed kinetic energy is large , this corresponds to the isotropic - equivalent value and the beam - corrected kinetic energy is a few orders of magnitude smaller than this . for @xmath238 , the inverse compton parameter @xmath160 is proportional to @xmath252^{1/2}$ ] , where @xmath253=min[@xmath254 , and for these values of @xmath123 and @xmath124 , @xmath255 .
for values , of @xmath249 , @xmath256 , and @xmath257 the particle density in the circumburst environment is @xmath258 @xmath184 , which , using eq .
[ eq : thetaj ] gives an opening angle of @xmath259 rad .
this corresponds to @xmath260 , which gives a beam - corrected kinetic energy @xmath261 erg and @xmath4-ray energy @xmath262 erg .
this would imply an efficiency of @xmath263 , which is two orders of magnitude smaller than is typically inferred @xcite .
the jet opening angle that we estimate assumes the hydrodynamics as described by @xcite , and variations on the hydrodynamics , such as that discussed in @xcite , result in a different coefficient in eq .
[ eq : thetaj ] . however , these differences are of order unity , and do not affect the validity of the model discussed .
this model is , therefore , able to satisfy our observations without the requirement of an excessively large kinetic energy .
however , the opening angle is a factor of at least 30 narrower than previously observed @xcite and the inferred @xmath4-ray energy is extremely low for a long grb , prompting us to consider other options . keeping in place the condition that @xmath265 s , we investigate a model in which the x - ray and uv / optical bands are above @xmath121 and @xmath53 . in this model our observational constraints on the values of @xmath121 , @xmath53 and @xmath266 only apply at t+80 s. since we assume a jet break to occur at @xmath234 s , the factor of @xmath239 in the kinetic energy is unity , and @xmath232 is therefore not of importance in our calculations . for @xmath53 above the observing band , @xmath225 , and therefore @xmath267 , in which case the dependence on the minimum lorentz factor of the shock accelerated electrons @xmath268 changes from @xmath269 to @xmath270 where @xmath271 and @xmath272 are the minimum and maximum energy of the shock accelerated electrons , and @xmath17 is the bulk lorentz factor in the shocked medium @xcite .
@xmath273 , and therefore the expression for @xmath121 given in eq .
[ eq : num ] has to be modified such that @xmath274 ^ 2 $ ] is replaced by [ ln@xmath275 ^ 2 $ ] . the value of ln@xmath276 is not well determined due to the unknown magnetic field in the upstream of the grb shock , although it generally ranges between 5 and 10 @xcite .
we therefore take ln@xmath277 , such that eq .
[ eq : num ] becomes @xmath278 we assume the same jet break time as before ( @xmath234 s ) , whereby eq .
[ eq : nuc ] and [ eq : modnum ] reduce to @xmath279 and @xmath280 in the slow cooling regime , where @xmath281 , @xmath282 by taking the measured value of @xmath127 at @xmath283 hz , and using eq .
[ eq : nuc ] , [ eq : fmax ] and [ eq : modnum ] , the expression for @xmath284 reduces to @xmath285 from these equalities we determine that a kinetic energy of @xmath176 erg would require @xmath286 , @xmath287 and @xmath288 . for a grb local environment with a particle density of @xmath289 @xmath184 , which is a typical density for a molecular cloud @xcite , the jet opening angle would be @xmath290 , a factor of a few smaller than that determined for some other grbs ( e.g. grb 980519 ; * ? ? ?
this gives a beam - corrected kinetic energy of @xmath291 erg .
this model provides an alternative scenario in which a highly collimated jet is the cause for the large flux observed in grb 061007 , but where a larger jet opening angle and more standard beam - corrected @xmath4-ray can satisfy our observations , with respect to the previous model . for completeness
we look into a scenario in which synchrotron self - compton ( ssc ) emission contributes to the observed emission . in the case where the x - ray emission is dominated by ssc radiation , and the uv / optical emission
is produced by synchrotron radiation , both observed bands would need to be in the fast - cooling regime , with @xmath292 and @xmath293 , in order for the decay slope and spectral index in the x - ray and uv / optical emission to be consistent . in this case
the spectral index is given by @xmath294 , which would give a value of @xmath295 .
however , the decay of the optical afterglow is expected to be @xmath296 , which is inconsistent with our observations of @xmath297 . from our spectral analysis
we determine the dust extinction in the circumburst medium to be most consistent with a wavelength dependence similar to that of the lmc .
this is consistent with previous studies of grb environments ( e.g. fruchter et al .
2006 , kann , klose & zeh 2006 ; schady et al .
2006 ) , which also find grb hosts to be irregular galaxies , and is also in agreement with the collapsar model , which requires a sub - solar metallicity progenitor star , with an upper limit of @xmath298 ( e.g. hirschi , meynet & maeder 2005 ; woosley & heger 2006 ) .
the best - fit parameters for the x - ray column density and dust extinction local to the grb are @xmath107 @xmath299 @xmath64 and @xmath300 @xmath301 , and this gives a gas - to - dust ratio of @xmath107/@xmath300 @xmath302 @xmath64 .
however , @xmath300 was determined assuming an lmc dust extinction law , whereas @xmath107 was determined assuming solar abundances , which are unlikely to be correct for an irregular , magellanic - type galaxy .
correcting the x - ray column measurement to a metallicity of 1/3 solar , which is appropriate for an ism similar to that of the lmc , we find that the gas - to - dust ratio is a factor of more than three larger than that measured in the lmc .
we note that an environment s metallicity affects the amount of dust present as well as the equivalent @xmath107 measured .
our assumptions on the metallicity , therefore , do not change the @xmath300/@xmath107 ratio as we measure it , since to first order both @xmath300 and equivalent @xmath107 scale directly with metallicity .
it is , therefore , valid to compare the dust - to - gas ratio in different environments derived this way , independent of their metallicity .
although not consistent with the lmc , the dust - to - gas ratio measured in the local environment of grb 061007 is consistent with the average gas - to - dust measured in the local environment of a sample of previous _ swift _
grbs @xcite , which was @xmath303 @xmath64 .
the lack of evolution observed in the soft x - ray absorption later than @xmath304 s after the prompt emission suggests that the burst must have completely ionised the gas in its immediate environment within this time scale . the absorbing medium that we observe must
, therefore , be far enough from the burst that it was not completely ionised in the first few 100 s of seconds after the initial explosion .
the probability of there being an absorption system in the line of sight from an unrelated , intervening galaxy that is of sufficient optical depth to contribute a measurable amount of soft x - ray absorption is very small @xcite , leading us to conclude that the absorbing material measured in the spectrum of grb 061007 is local to the burst . from the results in @xcite we estimate the absorbing medium to lie at a few tens of parsecs from the grb , and
the large value of @xmath107 suggests that grb 061007 was embedded in a dense , star - forming region , or giant molecular cloud , in common with many other grbs ( e.g. dai & lu 1999 ; wang , dai & lu 2000 ; reichart & price 2002 ) .
it is interesting to note that with @xmath305 erg in the 10 kev20 mev energy range , and a rest - frame peak energy of @xmath306 kev , grb 061007 satisfies the amati relation , given by @xmath307 @xcite .
it also satisfies a correlation between the peak isotropic luminosity , @xmath308 , and spectral lag in the prompt emission , @xmath309 , whereby @xmath310 .
both these correlations are satisfied by a large fraction of long grbs ( e.g. * ? ? ?
* ; * ? ? ?
* ) , suggesting that grb 061007 has the same class of progenitor as other long grbs .
the consistency observed in the spectral and temporal afterglow behaviour in the x - ray and uv / optical bands , and in the late time @xmath4-ray band , implies that the emission observed comes from the same spectral segment .
if we take the afterglow to be in the slow cooling regime and assume a jet break to occur at @xmath311 s a large amount of kinetic energy , of at least @xmath312 erg is required for a constant density circumburst medium , or @xmath313 erg for a wind - like circumburst medium .
the implausibly large amount of energy required by both these scenario makes them unsatisfactory models to explain the properties of grb 061007 .
we also investigate a second scenario in which a highly collimated outflow produces a jet break before t+80 s. this requires a jet angle no larger than @xmath314 , which satisfies the observational requirements with a lower energy budget and more reasonable set of parameters .
the upper limit on the jet opening angle of @xmath314 is a factor of 4 smaller than that previously determined in grbs ( berger , kulkarni & frail 2003 ) , although still physically possible .
furthermore , the need to observe the jet break in order to measure @xmath19 introduce selection effects , and the smaller and larger end of the current jet opening angle distribution will , thus , be underestimated . for the very smallest of jet opening angles to be measured ,
very early time data of the afterglow is needed , preferentially covering several energy bands
. for a jet break earlier than t+80 s our observations are valid for a spectral regime in which @xmath136 , and @xmath315 , where in the latter a larger jet opening angle is obtained by assuming a density @xmath289 @xmath184 , consistent with that observed in molecular clouds .
the dependence of the circumstellar particle density on @xmath124 , @xmath123 and @xmath157 in the former of these models does not allow for such high densities , for which we estimate @xmath258 @xmath184 .
our spectral analysis on the afterglow resulted in a large column density , which supports a scenario in which the grb is embedded within a dense molecular cloud or active star - forming region .
this therefore favours a model in which @xmath315 , where the number density is inferred to be larger . in the case of a uniform jet , where the energy and lorentz factor are constant across the jet ,
the afterglow after a jet break should decay as @xmath316 . in
both the post jet - break models that we consider the predicted decay slope is steeper than the observed value of @xmath297 .
a possible explanation is continual energy injection , which would maintain the energy in the afterglow and slow down the rate of decay .
the requirement for steady energy injection over the entire light curve is of some concern , but seems plausible on the basis of many other examples of energy injection in _ swift _ afterglow observations .
evidence of prolonged energy injection in the form of plateaus in x - ray light curves is observed in a large fraction of _ swift _ grbs , although typically up to @xmath11-@xmath196 s after the prompt emission . in the case of grb 061007 a steady energy injection
is needed for at least an order of magnitude longer than this , during which no break is observed in the light curve .
this is not to say that the central engine needs to be active for this length of time , but simply that it is active for the duration of the prompt emission and ejects shells with a large distribution in the lorentz factor .
a range in @xmath17 can provide a source of energy to the afterglow for a long time after the central engine has stopped through ongoing refreshed shocks ( e.g. * ? ? ?
* ; * ? ? ?
* ) . nevertheless , an energy injection mechanism that remains extremely constant for such a long time requires specific conditions to be in place , such as a central engine luminosity that varies smoothly with time in the case of a long - lasting engine , or a smooth distribution of lorentz factors in the case of refreshed shocks @xcite .
alternatively , the slow decay rate of the afterglow in the context of post jet - break could also be the result of a jet - edge effect in which there is no sideways expansion .
in such a case the afterglow decay index would steepen by 3/4 when @xmath317 , and would continue to decay at this rate due to the lack of lateral expansion @xcite . in such a scenario there is no need for energy injection , although additional conditions would be required to prevent the jet from expanding further , such as a high magnetic field within the jet .
a possibly more natural form of non - laterally expanding jet is a structured jet or outflow with an energy distribution that is a decreasing function of angular distance from the jet axis , @xmath318 , and a narrow , high - energy core at the axis .
a jet break is , then , observed when the edge of the uniform inner core becomes visible to the observer , and it is , therefore , the viewing angle rather than the jet half - opening angle that determines the time of the jet break . in order for a jet break to occur before t+80 s the viewing angle needs to be close to the jet axis , which is not that unreasonable given the brightness of grb 061007 . at the time of jet break the smoother distribution in energy , @xmath319 , compared to a uniform jet results in a smaller change in the decay slope that will be shallower than the 3/4 increase in decay index observed in a uniform jet break . for a power law distribution of @xmath320 ,
where @xmath321 , the smaller the value of @xmath322 the shallower the afterglow light curve will be after the jet - break . in the case of @xmath323 , where @xmath238 , a decay rate after a jet break with index @xmath324 will occur in structured jet with an @xmath320 distribution index @xmath325 , and for @xmath326 we require
@xmath327 @xcite .
a consequence of a very early jet break time is the faster evolution of @xmath17 into a non - relativistic phase , which will produce an additional break in the light curve as the decay rate slows down . for a uniform jet with no energy injection ,
this break would occur at at a time @xmath328 @xcite . for @xmath234 s and @xmath329
, the jet would , therefore , not enter the non - relativistic phase until more than @xmath330 s after the prompt emission , which is not in conflict with our observations , and in the presence of energy injection or for structured jets @xmath331 would only get larger . in the context of post - jet break emission , a decay slope of @xmath324
can , therefore , be accounted for by a non - laterally expanding jet as well as continual energy injection through refreshed shocks .
we , therefore , find an early jet break before t+80 s can satisfy the conditions imposed by our observations without requiring a large energy budget and with a reasonable set of physical parameters .
in this paper we have presented multi - wavelength spectral and temporal data taken with _ swift _ of grb 061007 , an very bright grb with unusually smooth panchromatic decay . using these data we explored the circumstances that could produce such a bright burst , and so determined the conditions that distinguish grb 061007 from other , more standard grbs . on many levels
we find grb 061007 to be typical of other grbs , suggesting that the same progenitor model for most long grbs is also applicable to grb 061007 .
the surrounding dust has properties most consistent with an lmc - like environment , and the circumburst material is likely to be dense , suggesting that it is embedded in a large molecular cloud or a region of active star formation .
furthermore , its consistency with both the amati and timelag - luminosity relations , which apply to a large fraction of long grbs , indicating that the source of energy in grb 061007 is no different than that of most other long grbs .
the enormous brightness of grb 061007 is , therefore , unlikely to be due to a different type of progenitor , but the result of extreme values for a few defining characteristics .
two inherently different models can satisfy our observations of grb 061007 , where the main difference lies in the time of the jet break ; either before or after our observations .
both require certain extreme conditions to account for the brightness of this burst , either in the form of a large energy budget or in the collimation of ejecta into a very narrow jet opening angle .
however , given the brightness of grb 061007 , it should not be a surprise that some of the defining properties are non - standard .
the large amount of kinetic energy required in a late - time jet break model lead us to favour a model in which the grb 061007 is produced by the concentration of energy into a highly narrow jet @xmath332 ; a condition that can more easily be met than the production of large amounts of energy .
the brightness of grb 061007 across the electromagnetic spectrum makes it ideal to explore the conditions surrounding the production of grbs , and its uniqueness provides further insight into the range in properties that grbs have .
the small opening angle that we determine for grb 061007 has important implications on the true range in jet opening angles , and this could provide a clue as to the cause for the large fraction of _ swift _ grbs with no clear signs of jet breaks .
we gratefully acknowledge the contributions of all members of the _ swift _ team .
sz thanks pparc for support through an advanced fellowship .
akerlof , c. , et al . 1999 ,
nature , 398 , 400 amati , l. , et al .
2002 , a&a , 390 , 81 amati , l. 2006 , mnras , 372 , 233 band , d. , et al .
1993 , apj , 413 , 281 barthelmy , s. d. , et al .
2005 , space science reviews , 120 , 143 berger , e. , kulkarni , s. r. , & frail , d. a. 2003 , apj , 590 , 379 bersier , d. , et al . 2006 , gcn , 5709 bloom , j. s. , frail , d. a. , & kulkarni , s. r. 2003 , apj , 594 , 674 burrows , d. n. , et al .
2005 , space science reviews , 120 , 165 campana , s. , et al .
2006 , a&a , 449 , 61 capalbi , m. , et al . 2007 ,
a&a , 462 , 913 chevalier , r. a. , & li , z .- y .
2000 , apj , 536 , 195 dai , z. g. , & lu , t. 1999 , apj , 519 , l155 dickey , j. m. , & lockman , f. j. 1990 , ara&a , 28 , 215 frail , d. a. , et al .
2001 , apj , 562 , l55 freedman , d. l. , & waxman , e. 2001 , apj , 547 , 922 fruchter , a. s. , et al .
2006 , nature , 441 , 463 gehrels , n. , et al .
2004 , apj , 611 , 1005 gehrels , n. , et al .
2006 , nature , 444 , 1044 ghirlanda , g. , ghisellini , g. , & lazzati , d. 2004 , apj , 616 , 331 golenetskii , s. , aptekar , r. , mazets , e. , palshin , v. , frederiks , d. , & cline , t. 2006 , gcn , 5722 groot , p. j. , et al .
1998 , apj , 502 , l123 hirschi , r. , meynet , g. , & maeder , a. 2005 , a&a , 443 , 581 hjorth , j. , et al .
2003 , nature , 423 , 847 huang , y. f. , cheng , k. s. , & gao , t. t. 2006 , apj , 637 , 873 jakobsson , p. , fynbo , j. p. u. , tanvir , n. , & rol , e. 2006 , gcn , 5716 kann , d. a. , klose , s. , & zeh , a. 2006 , apj , 641 , 993 kouveliotou , c. , et al .
1993 , apj , 413 , l101 kulkarni , s. r. , et al .
1998 , nature , 395 , 663 li , z. , & waxman , e. 2006 , apj , 651 , 328 livio , m. , & waxman , e. 2000 , apj , 538 , 187 markwardt , c. , et al . , 2006 ,
gcn , 5713 meszaros , p. , & rees , m. j. 1997 , apj , 476 , 232 mszros , p. , & rees , m. j. 1999 , mnras , 306 , l39 nicastro , l. , et al .
1999 , a&a supp . , 138 , 437 nousek , j. a. , et al .
2006 , apj , 642 , 389 norris , j. p. , & bonnell , j. t. 2006 , apj , 643 , 266 obrien , p. t. , et al .
2006 , apj , 647 , 1213 oflaherty , k. s. , & jakobsen , p. 1997 , apj , 479 , 673 osip , d. , chen , h. w. , & prochaska , j. x. 2006 , gcn , 5715 panaitescu , a. , & kumar , p. 2000 , apj , 543 , 66 panaitescu , a. 2005 , mnras , 363 , 1409 panaitescu , a. 2006 , mnras , submitted to mnras ( astro - ph/0612170 ) panaitescu , a. , mszros , p. , gehrels , n. , burrows , d. , & nousek , j. 2006 , mnras , 366 , 1357 peer , a. , & zhang , b. 2006 , apj , 653 , 454 pei , y. c. 1992 , apj , 395 , 130 perna , r. , & lazzati , d. 2002 , apj , 580 , 261 rees , m. j. , & meszaros , p. 1998 , apj , 496 , l1 reichart , d. e. , & price , p. a. 2002 , apj , 565 , 174 rhoads , j. e. 1999 , apj , 525 , 737 romano , p. , et al .
2006a , a&a , 456 , 917 romano , p. , et al . 2006b , a&a , 450 , 59 roming , p. w. a. , et al . , 2005 , space sci
rev , 120 , 95 rykoff , e. s. , & rujopakarn , w. , 2006 , gcn , 5706 sakamoto , t. , et al .
2006 , aip conf .
proc . 836 : gamma - ray bursts in the swift era , 836 , 43 sari , r. , & esin , a. a. 2001 , apj , 548 , 787 sari , r. , & mszros , p. 2000 , apj , 535 , l33 sari , r. , piran , t. , & halpern , j. p. 1999 , apj , 519 , l17 sari , r. , piran , t. , & narayan , r. 1998 , apj , 497 , l17 schady , p. , et al . , 2007 , mnras , in press tagliaferri , g. , et al .
2005 , a&a , 443 , l1 tanvir , n. r. , et al . 2004 , astronomical society of the pacific conference series , 312 , 275 van der horst , a j. , & rol .
e. 2006 , gcn , 5720 vaughan , s. , et al .
2006 , apj , 638 , 920 vetere , l. , pagani , c. , burrows , d. n. , & schady , p. 2006
, gcn , 5714 wang , x. y. , dai , z. g. , & lu , t. 2000 , mnras , 317 , 170 willingale , r. , et al .
2006 , submitted to apj ( astro - ph/0612031 ) woosley , s. e. 1993 , apj , 405 , 273 woosley , s. e. , & heger , a. 2006 , apj , 637 , 914 zeh , a. , klose , s. , & kann , d. a. 2006 , apj , 637 , 889 zhang , b. , fan , y. z. , dyks , j. , kobayashi , s. , mszros , p. , burrows , d. n. , nousek , j. a. , & gehrels , n. 2006 , apj , 642 , 354 zhang , b. , & mszros , p. 2001 , apj , 552 , l35 zhang , b. , et al . 2007a , apj , 655 , 989 zhang , b. , zhang , b .- b . ,
liang , e .- w . ,
gehrels , n. , burrows , d. n. , & mszros , p. 2007b , apj , 655 , l25 | grb 061007 was the brightest gamma - ray burst ( grb ) to be detected by _
swift _ and was accompanied by an exceptionally luminous afterglow that had a @xmath0-band magnitude @xmath1 at 80 s after the prompt emission . from the start of
the _ swift _ observations the afterglow decayed as a power law with a slope of @xmath2 in the x - ray and @xmath3 in the uv / optical , up to the point that it was no longer detected above background in the optical or x - ray bands .
the brightness of this grb and the similarity in the decay rate of the x - ray , optical and @xmath4-ray emission from 100 s after the trigger distinguish this burst from others and present a challenge to the fireball model .
the lack of a cooling or jet break in the afterglow up to @xmath5 s constrains any model that can produce the large luminosity observed in grb 061007 , which we found to require either an excessively large kinetic energy or highly collimated outflow .
analysis of the multi - wavelength spectral and high - resolution temporal data taken with _ swift _ suggest an early time jet - break to be a more plausible scenario than a highly energetic grb .
this must have occurred within 80 s of the prompt emission , which places an upper limit on the jet opening angle of @xmath6 .
such a highly collimated outflow resolves the energy budget problem presented in a spherical emission model , reducing the isotropic - equivalent energy of this burst to @xmath7 erg , consistent with other grbs .
gamma - rays : bursts - gamma - rays : observations - ism : jets and outflows |
cooperation is a key aspect in the real world , ranging from biological systems to human behavior @xcite .
therefore , people restore to the game theory to study the emergency and maintenance of cooperation in biology , psychology , computer science , and economics @xcite . especially , the prisoner s dilemma game ( pdg ) , has become a metaphor to approach the emergency of cooperation and altruism behavior . in the tradition pdg ,
each of two players chooses a strategy from cooperation ( @xmath0 ) or defection ( @xmath1 ) simultaneously and gets payoff .
they both receive @xmath2 upon mutual @xmath0 and @xmath3 upon mutual @xmath1 .
a defector gets @xmath4 when it plays game with cooperator who gets @xmath5 . in pdg
, we have @xmath6 and @xmath7 . because the mutual @xmath0 get the highest total income , @xmath1 is the better choice than @xmath0 no mater what the other player s strategy . without any mechanism for the evolution of cooperation ,
natural selection favors defection .
the other widely studied games include snowdrift game @xcite , public good game @xcite , rock - paper - scissors game @xcite , and so on .
the complex network has also attracted lots of attentions in the past few years .
the complex network is ubiquitous in nature .
the human society can also be described as the systems composed of interacting agents .
the classical social network maps the individual into the node , and the connection between individuals into the link .
the evolutionary game theory in spatial structure has became a unifying paradigm to study how cooperation may be sustained in a structured population @xcite .
it was found that the spatial extension is one of several natural mechanisms to enforce cooperation .
network structure will affect the behavior of strategy density @xcite . in lattice network ,
the cooperation is usually get together to support each other to resist the defection @xcite .
santos and pacheco found in scale - free networks the strong correlation leads to the dominating trait throughout the entire range of parameters of both games in scale - free networks @xcite . and also , there are anmount of researches on other networks , like small - world @xcite and random network @xcite
. when the player on the structure network chose the better strategy to play game , in fact , not that the players select the proper strategy , but player s strategy is determined by the network structure .
for example , in scale - free networks , the large degree nodes ( hubs ) and the nodes which connect to hubs tend to be occupied by @xmath0 @xcite .
the networks used in the most papers of this field are statistic .
the connection will never change once it is build .
it is not realistic enough , as the interactions themselves help shape the network @xcite .
what is more , in the real world , the relationship between the people is not constant .
sometimes people can not cut some relationship with their relatives , neighbors or colleagues but they can end their old relationship and build a new one . sometimes this changing is caused by the results of the game , because people would like to make friends in a reciprocal respect .
for example , people always like to make friends with rich one for a sake of pursuing fortune .
so , when we study the social model in network like pdg , the network structure should be dynamical entities @xcite .
the nodes can remove or sustain their link in network according to the game results .
till now , there are few models studied the cooperative behaviors in a groups with adaptive connections . besides some early work @xcite ,
arne build a coevolution model of strategy and structure @xcite . in this model , the probability of forming or cutting link between node @xmath8 and @xmath9 is based on their strategies .
the changing of network structure is result from the strategy changing in the network .
then it also affect the strategy density back
. however , the link could change even if the nodes strategy do not change in their model .
the rewire of link in this model is not the player s own decision .
et al . _ also build a coevolution model that the node rewire its link only for changing its strategy @xcite .
moreover , in this model , the node rewire its long range link based on the existed network structure , not the playing game results . in our opinion , a rational model for coevolution of game and
network structure should contain two features : ( 1 ) the nodes rewire their links only when agents change their status ; ( 2 ) the rewiring should be based on the playing results of game . in this paper , we will present a coevolution model of the pdg and network .
we use pdg as a metaphor to studying cooperation between unrelated individuals and consider a social networks with four fixed local links and one adjustable long - range link ( lrl ) .
the agents in the network play game with their network neighbors .
they will change their strategies and adjust lrls according to the results of game .
then the network structure changing also affect the cooperation density .
we set up a system of @xmath10 players arranged at the nodes of a ring lattice network .
each node is connected with four local nodes .
these local interactions will not change during the whole process of the evolution . besides four fixed links , every node in this lattice has an adjustable lrl which connects to another node and self - connections and the duplicate links
are excluded .
we call the lrl out - link for the node to whom it belong or in - link for the node to whom it connect
. the node can select another node to which the out - link wires , but it can not give up the lrl .
therefore , each node has at least one out - link and many possible in - links .
when node changes its strategy , it will also rewire its lrl .
we will discuss when and how lrls rewire later .
as suggested by nowak and may @xcite , we adopt @xmath11 , @xmath12 @xmath13 , and @xmath14 . then @xmath15 can be considered as the temptation to @xmath1 against @xmath0 .
every player plays the pdgs with its neighbors on network and itself and get the total payoff @xmath16 .
after each round of the game , players are allowed to inspect their neighbors total payoffs and change their strategies in the next round .
the player @xmath17 updates its strategy by selecting one of its neighbors @xmath18 with a probability @xmath19 , @xmath20 where @xmath21 is the community composed of the nearest neighbors of the player @xmath17 , and @xmath22 is the degree of node @xmath23 at time @xmath24 . in the spirit of preferential attachment proposed by a .-
barbasi and r. albert @xcite , we incorporate the preferential selection rule to model social behaviors . in eq .
[ eq1 ] , player with large degree has more probability to impact his neighbors . that is true in the society that people who have great impact often have lots of social relations and they are also focused by their friends .
node @xmath17 will follow the node @xmath18 s strategy by the probability , @xmath25 } , \label{eq2}\ ] ] where @xmath26 and @xmath27 are the total payoffs of node @xmath17 and @xmath18 , and @xmath28 indicates the noise generated by the players allowing irrational choices @xcite .
if node @xmath18 has the same strategy with @xmath17 or @xmath17 do not mimic @xmath18 s strategy , node @xmath17 will do nothing .
otherwise , it will rewire its lrl to a new one .
there are two rewiring rules in our model : random rewiring and preferential rewiring . with probability @xmath29 ,
the density of cooperation in the network , node @xmath17 will chose a new node randomly .
for the rest probability @xmath30 , node @xmath17 will chose a new node according to the node s payoff . in the preferential rewiring rule ,
the node rewires its link according to the payoff of all nodes in network , @xmath31 where @xmath32 is the probability of node @xmath17 rewiring its link to @xmath18 and @xmath33 presents all nodes in the graph .
@xmath34 is used to change the effect of payoff .
@xmath35 indicates that the payoff has no effect here and the nodes rewire their links randomly . for @xmath36
, the node will prefer to connect the node with larger payoffs .
so it also looks like a kind of preferential selection rule .
we run our simulations with varying @xmath15 and @xmath34 for fixed @xmath37 and the system size @xmath38 .
all the results in this paper are obtained from the average results with @xmath39 different monte carlo ( mc ) simulation trails .
we start with node linking its lrls to other nodes randomly with equal probability and random initial state with @xmath40 as the initial state .
the players update their strategies in random sequence . in every mc step ,
all nodes have one chance to change their strategies and rewire their links .
( color online ) frequency of cooperators @xmath29 for different @xmath34 as functions of the advantage of defectors @xmath15.,scaledwidth=50.0% ] ( color online ) frequency of cooperators @xmath29 evolve with @xmath24 for systems at different parameters on pdgs.,scaledwidth=50.0% ] figure [ fig1 ] shows the frequency of cooperators @xmath29 in our model as the functions of @xmath15 for different @xmath34 .
similar to evolutionary game in regular network @xcite , we also find two thresholds in our model .
full cooperation is achieved if @xmath15 does not exceed the threshold @xmath41 . for @xmath42
, @xmath0 can not resist the temptation of @xmath15 and can not survive in the network
. in the region of @xmath43 , @xmath0 and @xmath1 can coexist in the network .
compared with the case of @xmath35 , the position of @xmath41 does not change with @xmath34 .
however , @xmath34 affect the @xmath44 conspicuously .
the probability of node using preferential selection to rewire its lrl is @xmath30 .
therefore @xmath34 does not work at @xmath29 close to @xmath45 or @xmath15 close to @xmath41 .
when @xmath46 , the qualitative results @xmath29 remain unaffected by @xmath34 that @xmath29 decreases monotonous with @xmath15 .
when @xmath47 , there exists a region of @xmath15 promoting cooperation obviously .
this promotion starts at @xmath48 ( @xmath49 ) and this region enlarge with increasing @xmath34 .
but the effect of promotion does not increase with @xmath34 .
we observe that @xmath29 does not change at @xmath50 for @xmath51 , @xmath52 , and @xmath53 .
actually , the transition is caused by the changing of network structure .
we will discuss it in the next subsection . in order to discuss how the @xmath34 promotes @xmath29 in the promotion region
, we present the time evolutions of @xmath29 in fig .
[ fig2 ] for fixed @xmath54 with different @xmath34 values .
the red , blue , and black lines are the averages of @xmath39 trials for @xmath35 , @xmath55 , and @xmath53 respectively .
the green one is the @xmath29 time series of one trail in the black line . for @xmath35 and @xmath55 ,
@xmath29 decreases with time to its station state quickly . as shown in fig .
[ fig1 ] , @xmath29 for @xmath56 is a little higher than that of @xmath35 .
however , for @xmath57 , @xmath29 decreases like @xmath35 firstly , and then the evolution of network drives @xmath29 increasing with time to @xmath58 . considering that the black line is the average of @xmath39 trails , we believe the green line in fig .
[ fig2 ] contains more details of the evolution . in the early stage of the green line
, @xmath29 decreases to a temporary stable state in a manner similar to but a little larger than @xmath56 .
however , at @xmath59 , there is a sharp increasing in the green line from about @xmath60 to @xmath58 which is also the final level of the average result ( the black line ) .
it means that the gradually increasing of the black line is caused by the average effect of @xmath39 same sharply increase at different times . in this model
the behavior of @xmath29 and the evolution of network structure are equal important .
the evolution of network structure results in the transition of @xmath29 . in order to describe the network structure ,
we first present the degree distribution @xmath61 in fig .
panel ( a ) is @xmath61 in the case of the stable state of red line in fig .
[ fig2 ] . here
the preferential rewiring does not work and all lrls select the target nodes randomly .
considering the self - connection is forbidden , we know @xmath62 here @xmath10 usually is large enough , so one can get @xmath63 figure [ fig3](b ) is @xmath61 for the stable state of blue line in fig .
@xmath61 in ( b ) is similar to that of ( a ) but the largest degree is @xmath64 .
[ fig3](c ) is @xmath61 for the stable state of gree line in fig .
[ fig2 ] and ( d ) is for the green line after the sharp increasing . both ( c ) and ( d ) in fig .
[ fig3 ] are the degree distributions of one trial , but not the cumulative stationary degree distribution of @xmath39 different trials . by comparing ( c ) with ( d ) , it is helpful to uncover the reason of the sharp increasing in fig .
[ fig2 ] . in fig .
[ fig3 ] ( d ) , there is only one node that its degree is larger than half of the other nodes connected to it .
we name this node which has the largest degree in the network as hub node ( hn ) . as presented in fig .
[ figbc ] , the other nodes can be divided into two types : the nodes connect their lrls to hn and the nodes do not .
we name the first node as an and the second one as bn .
the number of them are @xmath65 and @xmath66 , respectively .
( color online ) the cumulative stationary degree distributions @xmath61 in pdgs.,scaledwidth=50.0% ] ( color online ) illustration of hn , an , and bn .
each node in the network has four fixed links and there are five red nodes wire their lrls to the blue one . in order to make an and hn prominent
, we do not draw the lrls of other nodes .
the blue node has the largest degree in this net , so blue node is hn , and the red one is an and the others are bn .
we draw the arrows in the figure to present these lrls are out - links for an and in - links for hn.,scaledwidth=30.0% ] now , we exam the detail of the network after the sharp increasing in the green line ( @xmath57 , @xmath54 ) of fig .
note that the strategy of hn is always @xmath0 and the strategy of most ans is also @xmath0 . before the sharp increasing or in the case of other parameters without sharp increasing ,
the hns are also prefer to @xmath0 . this phenomenon is also observed in some other networks with hub nodes @xcite .
more detailed information of our model are listed in table [ tab1 ] . in table
[ tab1 ] , @xmath67 is the cooperation density of an and @xmath68 is for bn .
almost all nodes of ans chose the strategy @xmath0 , so we do not need to present the mean payoff of an with @xmath1 .
what is more , it is found that @xmath69 is close to the case of @xmath35 ( @xmath70 for @xmath54 , @xmath71 for @xmath72 , and @xmath73 for @xmath74 ) .
it means that the existence of an does not affect the strategy density of bn . as discussed in ref
@xcite , an can resist the temptation of @xmath15 by mimicking the strategy of hn .
after the sharp increasing , the probability that an mimics the strategy of hn is much larger than that of other neighbors .
the hn s payoff is also larger because it has a lots of in - link lrls .
we will discuss the details of these probabilities in the next subsection . on the other hand
, only the node with strategy @xmath0 can grow into hn .
if hn is occupied by @xmath1 , hn will get higher payoff temporarily .
however , as we discussed above , an will follow hn s strategy and the strategy of an will be @xmath1
. then the hn can not earn payoff from its in - link lrls .
once hn can not earn enough payoff , both preferential and random rewiring will drive ans to rewire its lrl to other nodes .
then a new hn with strategy @xmath0 will appear in the network .
so it seems that strategy @xmath0 is a better choice for hn because it can earn a stable higher payoff . from table
[ tab1 ] , we also find that the bns with @xmath1 earn the most payoff and the payoff of bn with @xmath0 is close to the payoff of an .
however , although the mean payoff of bn with @xmath1 is the highest , in fact , the density of cooperator does nt decrease with time .
it shows that the probability of @xmath0 mimicking @xmath1 strategy and @xmath1 mimicking @xmath0 strategy are the same .
.the detailed information of prisoner s dilemma games ( @xmath57 ) . [ cols="^,^,^,^",options="header " , ] [ tab1 ] each horizontal line in fig .
[ fig4 ] presents a snapshot of the network .
we arrange these snapshots with time from top to the botton to show how the an evolves with time .
so we can depict every player s strategies in network and observe the evolution of these strategies .
the riht panel is @xmath65 at the same time with the left .
there is also a sharp increasing of @xmath65 at the same time like the green line in fig .
[ fig2 ] in looks . at @xmath75 ,
about @xmath76 mc steps before the transition , @xmath65 increase gradually from about @xmath77 to @xmath76 .
after the sharp increasing , @xmath65 still increase gradually to the final stable state .
moreover , before the sharp increasing happened , one can observe many black blocks ( the upper part of the left panel in figure [ fig4 ] ) .
it means the model has the similar feature of pdg in regular network that the @xmath0 node tends to get together for blocks to resist the @xmath1 .
these blocks start at a few @xmath0s , maybe three or more , and then close to each other in the network coincidentally . then a block is established and it will grow to change their neighbors strategies .
after some mc steps , the block will shrink and then disappears in the last .
after the sharp increasing of @xmath65 , there are too few red dots ( @xmath1 in an ) .
the green strip ( @xmath0 in an ) indicates that the ans or bns are very stable in the network .
the probability of an change to bn is very small and vice versa .
based on the results in the above context , the effect of @xmath34 is different from various @xmath15 and @xmath34 . after the sharp increasing
, the nodes in network can be divided into an and bn . almost all an are @xmath0 and the density of @xmath0 in bn is close to the case of @xmath35 .
so we can use the mean field theory and some basic feature of stable state to explain why the sharp increasing happened .
after the sharp increasing in fig .
[ fig3 ] , the system reaches the stable state gradually .
then we have @xmath78 or @xmath79 , where @xmath80 is
the average number of nodes changed from an to bn in one mc step and @xmath81 is that changed from bn to an .
considering that there are too few @xmath1s in an , we assume that @xmath80 is only caused by @xmath82 and random rewire . here
, we neglect the preferential rewiring . because the contribution of preferential rewiring is only about @xmath83 of random rewiring .
then we get @xmath84 here , @xmath29 means the change happened in the random rewiring , and @xmath85 is the mean degree of nodes in networks .
we neglect self - connection and multi - connection forbidden and we have @xmath86 here .
because an has the same strategy with hn , an only mimics the strategy from other @xmath87 neighbors .
the big fraction is the probability of an do not chose hn to mimic the strategy . the last @xmath88 is the probability of mimicked target with strategy @xmath1 .
we assume @xmath89 is the probability of success in the mimicking .
then @xmath81 will be more complicate .
we assume that bn change to an because they use the preferential rewiring .
the contribution of random rewiring is about @xmath90 of preferential rewiring , so we neglect it and derive the following formula , @xmath91 q_{b\rightarrow a } ( 1-p_c)n_b \nonumber \\ & & \frac{(n_a+5)^{\alpha}}{(n_a+5)^{\alpha}+n_a ( 2 + 4p_c)^{\alpha}+ p_{bc } n_b \left ( 1+(5+\frac{n - n_a}{n})p_c \right ) ^{\alpha } + n_b ( 1-p_{bc } ) \left ( b(5+\frac{n - n_a}{n})p_c \right ) ^{\alpha } } , \label{eq5}\end{aligned}\ ] ] where @xmath30 means the preferential rewiring , and @xmath92 is the probability of bn with strategy @xmath0 to mimic its @xmath1 neighbor and that @xmath1 try to mimic its @xmath0 neighbor .
the fraction here is the probability of node rewire to hn using the preferential rewiring .
we assume @xmath93 is the probability of success in the mimicking .
now , one can get @xmath68 from the simulation of @xmath94 and @xmath95 , and then we know how @xmath65 evolves with time by using @xmath96 . if @xmath97 , @xmath65 will not change with time . and
@xmath98 means @xmath65 will increase in the next mc step . however , we do not know @xmath89 and @xmath93 yet . according that the mean payoff of @xmath99 with @xmath1 is larger than mean payoff of @xmath100 in tab .
[ tab1 ] , we conjecture @xmath101 and @xmath102 .
indeed , we find @xmath103 is fit to our model
. we will take @xmath54 with @xmath57 and @xmath104 as examples . for @xmath54 and @xmath35
we get @xmath70 from the simulation . fig .
[ fig5 ] plots @xmath105 as different @xmath65 . for @xmath104
there are two stable points at @xmath106 and @xmath107 , and one unstable point at @xmath108 . for @xmath57 , there is only one stable state at @xmath109 .
the unstable point will decrease with the increasing of @xmath34 and coincident with the first stable point at @xmath110 . however , even @xmath35 , the maximal degree in the network is about @xmath111 .
so the first stable point can be discarded .
when the unstable point crosses @xmath112 or there is only one stable point , the system will reach to the second stable point .
( color online ) @xmath96 with various @xmath65 .
panel ( b ) is an enlargement of ( a).,scaledwidth=50.0% ]
the coevolution of dynamics and network structure is rapidly becoming an important field of the evolutionary game .
it contains more details about the social interaction in the real world . in this paper
, we build a co - evolution model of pdg and network structure . each node in network
has four fixed local links and one adjustable lrl .
when the node changes its strategy , it will rewire its lrl to another node according to the node s payoff and density of cooperation . and we introduce a parameter @xmath34 to denote the effect of payoff .
many early works @xcite also proved that the adaptive network can enhance the cooperation .
all these enhancements are caused by the emergency of cooperator with large degree in the network . in @xcite
, the cooperation is very sensitive to the plasticity parameter and only the adaptive network can enhance the cooperation .
in our model , the players rewire their lrls for any @xmath34 , but the cooperation is enhanced only in the case of @xmath36 that this enhancement is obvious for @xmath47 in a certain region of @xmath15 .
however , our results show that the enhancement of cooperation only happen in the case of changing the network structure property . in our model , for @xmath35 , the node will also rewire its lrl , but the network property will not change and the cooperation level will not be enhanced .
the cooperation is enhanced only when the node rewires its lrl according to the payoff .
similar phenomena was also observed in our simulations with snowdrift game ( sg ) .
we found that sg is more sensitive to @xmath34 than pdg and the obvious enhancement is for a smaller @xmath34 .
so we conjecture that the coevolution of network structure and game is an important mechanics to maintain the cooperation in the real society .
different from the results in @xcite which the cooperation always dominates in the adaptive network and the increasing of cooperation is limited . that is caused by two reasons : ( 1 ) in the probability of @xmath29 the player use the random rewiring .
( 2 ) the existance of four fixed links in network can be regarded as a noise to prevent the preferential selection . in @xcite
, authors discussed the leaders and the global cascades .
if every node could change its strategy in a smaller probability , the global cascades of coopertation is also observed in our model .
the analysis in this paper is based on the balance of an and bn . however , when the sharp increasing did nt happen , perhaps there exist more than one hns and hn is changing from one node to another frequently . because of the absense of the information about spacial structure and eq .
[ eq2 ] , the presented analysis in this model is not very precise any more .
actually , it is impossible to include all the details for the analysis .
we just hold on the main factors of the model and it works well enough to explain the main features of our model .
this work was supported by the national natural science foundation of china under grant no .
@xmath113 and by the fundamental research fund for physics and mathematic of lanzhou university .
this study is supported by the high - performance computer program in lanzhou university . | _ most papers about the evolutionary game on graph assume the statistic network structure .
however , social interaction could change the relationship of people . and the changing social structure will affect the people s strategy too . we build a coevolutionary model of prisoner s dilemma game and network structure to study the dynamic interaction in the real world . based on the asynchronous update rule and monte carlo simulation , we find that , when players prefer to rewire their links to the richer , the cooperation density will increase .
the reason of it has been analyzed . _ |
duality in the nucleon structure function , @xmath2 , was discovered before quantum chromodynamics ( qcd ) @xcite .
thus the structure function , @xmath3 , in the resonance region ( @xmath4gev ) , is approximately the same as ( duals ) the deep inelastic region ( @xmath5gev ) , when the functions are expressed in terms of the scaling variable , @xmath6 . here
@xmath7 is the final - state hadron mass .
moreover , the occurrence of duality appears to be local in the sense that it exists in each interval of @xmath8 over the prominent nucleon resonances .
an explanation of bloom - gilman duality was offered by de rujula , georgi , and politzer in 1977 @xcite . using the operator product expansion , they represented the scaling function as a sum of various twist operators , and then studied the contributions from individual twists ( moments of the scaling functions ) in the scaling variable , @xmath9 , where @xmath10 .
they argue that the @xmath11 moment , @xmath12 , of @xmath13 has the following twist expansion , @xmath14 here , @xmath15 is a mass scale @xmath16 , and @xmath17 depends logarithmically on @xmath18 , being roughly on the order of @xmath19 . according to eq .
1 , there exists a region of @xmath20 and @xmath21 where the higher twist contributions are neither large nor negligible , and where the dominant contribution to the moments comes from low - lying resonances . in this region for example , the moments defined by eq .
1 would not correspond to positive definite functions .
thus one might expect the structure functions to oscillate when large @xmath22 is important , and this would lead to the appearance of local duality .
a more recent study of parton - hadron duality by ji and unrau @xcite gives more quantitative estimations .
there is no doubt that duality is a very interesting phenomenon , and it could allow one , under certain circumstances , to bridge the gap between perturbative predictions and experimentally observed quantities in non - perturbative regions @xcite .
however , duality , as expressed in the above analyses , only reflects properties of the parton structure functions . appealing to the concept of factorization in the strong interaction ,
it is natural to ask whether there is duality in the parton fragmentation sector .
although in twist - two , perturbative calculations , parton structure functions and parton fragmentation functions have some similarities , this does not indicate that they have similar higher twist expansions . recently , an experiment at the jefferson national acceleration facility , jlab e00 - 108 , studied duality in semi - inclusive @xmath23 reactions .
semi - inclusive reactions can reflect higher twist contributions in the parton fragmentation sector , so it is worthwhile to study these contributions in more detail in order to understand how one might extract these contributions in such reactions .
we investigate these issues in this note .
in section 2 , we study the twist expansions in the fragmentation sector , and investigate twist two perturbative calculations of semi - inclusive @xmath23 reactions .
we also discuss a method to extract higher twist effects in these measurements .
the paper is summarized in section 3 .
in this section , we study the twist expansions of the parton fragmentation functions in @xmath0 and @xmath23 semi - inclusive reactions . the kinematics for @xmath24 is illustrated in fig .
[ fig : duality1 ] .
two kinematic invariants , @xmath18 and @xmath25 , define the process .
we consider scalar particles , @xmath26 .
the scattering matrix of the semi - inclusive reaction , @xmath27 , when calculated to lowest order in the electromagnetic interaction , is given by ;
= 8 cm @xmath28 \ , |e^+
e^- \rangle \nonumber \\ & = & \langle hx|{(ie)^2 \over 2 ! } \int d^4 \xi_1 \int_{t_0}^{t } d^4 \xi_2 { \cal t } [ { \cal l}_{em}^{int } ( \xi_1 ) { \cal l}_{em}^{int } ( \xi_2 ) ] \nonumber \\ & & { \cal t } exp \ , [ i \int d^4 \xi { \cal l}_{qcd}^{int } ] \ , |e^+ e^- \rangle + { \cal o } ( e^2 ) \ , , \end{aligned}\ ] ] here we have defined ; @xmath29 where @xmath30 is the charge of quark , @xmath31 , in units of the proton charge , @xmath32 . using feynman rules for qed , one then finds for the kinematics of fig .
[ fig : duality1 ] , @xmath33 hence , the cross section can be written as ; @xmath34 with ; @xmath35^\dagger \nonumber \\ & & { 1\over 2 } \
{ q_\mu q_\nu -q^2 g_{\mu\nu } - ( p_e - p_{e'})_{\mu } ( p_e - p_{e'})_{\nu } \ } \ , , \ \
\mbox{and } \nonumber \\ \hat{w}^{\mu\nu } & = & { 1\over 4\pi } \sum_x ( 2\pi)^4 \delta^4 ( p_h + p_x -q ) \langle 0|j^\mu ( 0 ) |hx \rangle \langle hx|j^\nu ( 0 ) |0 \rangle \nonumber \\ & = & { 1\over 4\pi } \int d^4 \xi e^{iq \cdot \xi } \sum_x \langle 0|j^\mu ( \xi ) |hx \rangle \langle hx|j^\nu ( 0 ) |0 \rangle \ , , \end{aligned}\ ] ] we have set the electron mass to zero .
the sum over unobserved hadrons , @xmath36 , can not be complete because the state @xmath37 depends non - trivially on the observed hadron .
therefore one does not have @xmath38 and so @xmath39 .
thus @xmath27 is not controlled by the product of two operators , the operator product expansion does not apply , and no short distance analysis can be formulated . in a reference frame
where the produced hadron , @xmath26 , is fixed @xcite , the hadron and photon momenta can be expanded as ; @xmath40 where ; @xmath41 writing ; @xmath42 one finds in the bjorken limit ; @xmath43 this result implies that as @xmath44 , then @xmath45 and @xmath46 . because @xmath47 , one finds in the bjorken limit that @xmath48 .
therefore in this limit , light - like separation occurs , @xmath49 , which dominates the integration region of eq . ( 6 ) .
fragmentation is generally a non - perturbative process .
we first consider the simplest quark fragmentation function represented diagrammatically in fig .
[ fig : duality2 ] . more complicated fragmentation processes , such as coherent fragmentation of several quarks and gluons , do not contribute until order @xmath50 .
= 8 cm in the case of fig .
[ fig : duality2 ] , the diagram of @xmath27 shown in fig .
[ fig : duality1 ] can be re - drawn as in fig . [
fig : duality3 ] .
= 8 cm the fermion field can be written as ; @xmath51 \ , , \end{aligned}\ ] ] where @xmath52 @xmath53 and @xmath54 @xmath55 are annihilation ( creation ) operators for particles and anti - particles .
assuming that there is no final state interaction between @xmath56 and @xmath57 , one can rewrite @xmath58 in eq .
( 6 ) as ; fig .
[ fig : duality2 ] as : + @xmath59 \ , .\end{aligned}\ ] ] to obtain this result we have used the fact that ; @xmath60 and due to the un - physical energy in the @xmath61 state , then ; @xmath62 in this expression , summation over color and flavor is assumed . due to the strong interaction lagrangian , @xmath63 , eq . ( 12 ) can be modified because of the higher order contributions in perturbative qcd calculations .
these interactions cause vertex corrections , gluon polarization , etc , and their contributions to the cross section give logarithmic corrections .
these contributions do not change the twists of the terms in the matrix elements . however , there are terms which represent quarks propagating in a gluon background in order to preserve color gauge invariance of the bilocal operator , @xmath58 .
these terms are essential to generate higher twist corrections , and can be included by changing the singular function of the free field theory @xcite , @xmath64 , to ; @xmath65 ) \ , .\end{aligned}\ ] ] substituting eq .
( 15 ) into eq . ( 12 ) and using the relationships ; @xmath66 \ , |0 \rangle & = & \sum_z \langle 0| { \cal p } ( exp \ , [ i \int_\infty^{\xi } d \zeta^{\mu } a_{\mu } ( \zeta ) ) ] \ ,
|z \rangle \cdot \nonumber \\ & & \langle z| { \cal p } ( exp \ , [ i \int_0^\infty d \zeta^{\mu } a_{\mu } ( \zeta ) ) ] \ , |0 \rangle \ , , \end{aligned}\ ] ] one has , @xmath67 \ , |hx \rangle { 1\over 2\pi } \partial_\rho \epsilon ( \xi_0 ) \delta ( \xi^2 ) \cdot \nonumber \\ & & ( s^{\mu\rho\nu\sigma } \gamma_\sigma - i \epsilon^{\mu\rho\nu\sigma } \gamma_\sigma
\gamma_5 ) _ { \alpha\lambda } \langle hx| { \cal p } ( exp \ , [ i \int_0^\infty d\zeta^\mu a_\mu ( \zeta ) ] \ , ) \psi_{\lambda } ( 0)|0 \rangle + \nonumber \\ & & \langle 0| \psi_\beta ( \xi ) { \cal p } ( exp \ , [ i \int_\xi^\infty d\zeta^\mu a_\mu ( \zeta ) ] \ , ) |hx
\rangle { 1\over 2\pi } [ -\partial_\rho \epsilon ( \xi_0 ) \delta ( \xi^2 ) ] \cdot \nonumber \\ & & ( s^{\nu\rho\mu\sigma } \gamma_\sigma - i \epsilon^{\nu\rho\mu\sigma } \gamma_\sigma
\gamma_5 ) _ { \delta\beta } \langle hx| { \cal p } ( exp \ , [ - i \int_0^\infty d\zeta^\mu a_\mu ( \zeta ) ] \ , ) \bar{\psi}_\delta ( 0 ) |0 \rangle \ , ] \nonumber \\ & = & { \cal p}_{h/\bar{b } } + { \cal p}_{h / b } \ , , \end{aligned}\ ] ] where @xmath68 is the fragmentation function of a quark @xmath69 fragmenting into hadron @xmath26 .
as shown in eq .
( 10 ) , light - like separation @xmath70 dominates the semi - inclusive process , @xmath27 @xcite . therefore one can write ; @xmath71 where @xmath72 includes contributions from @xmath73 and @xmath74 . both @xmath75 and @xmath74
go to zero ( @xmath76 ) as @xmath77 with eq .
( 8) and @xmath78 , one can expand eq .
( 16 ) in terms of @xmath79 .
this gives , @xmath80 \
, ) { 1\over 2\pi } \partial_\rho \epsilon ( \xi_0 ) \delta ( \xi^2 ) \cdot \nonumber \\ & & ( s^{\mu\rho\nu\sigma } \gamma_\sigma - i \epsilon^{\mu\rho\nu\sigma } \gamma_\sigma
\gamma_5 ) _ { \alpha\lambda } \langle hx| { \cal p } ( exp \ , [ i \int_0^\infty d\tau n\cdot a ( \tau n ) ] \ , ) \psi_{\lambda } ( 0)|0 \rangle + \nonumber \\ & & \langle 0| \psi_\beta ( \lambda n ) { \cal p } ( exp \ , [ i \int_\lambda^\infty d\tau n\cdot a ( \tau n ) ] \ , ) |hx
\rangle { 1\over 2\pi } \partial_\rho \epsilon ( -\xi_0 ) \delta ( \xi^2 ) \cdot \nonumber \\ & & ( s^{\nu\rho\mu\sigma } \gamma_\sigma - i \epsilon^{\nu\rho\mu\sigma } \gamma_\sigma
\gamma_5 ) _ { \delta\beta } \langle hx| { \cal p } ( exp \ , [ - i \int_0^\infty d\tau n\cdot a ( \tau n ) ] \ , ) \bar{\psi}_\delta ( 0 ) |0 \rangle \ , ] \nonumber \\ & & + { \cal o}(\hat{\xi } ) \ } \ , .\end{aligned}\ ] ] the @xmath81 term in eq .
( 19 ) is the same as the definition of parton fragmentation given in ref.@xcite .
working in the light - cone gauge , @xmath82 , explicit reference to gluons disappears , and using the definition of the twist of an invariant matrix element of a light - cone bilocal operator @xcite , ref.@xcite studied the twist expansion of the possible terms for the production of scalar hadron or hadrons whose spins are not observable , @xmath83 \ , , \nonumber \\ & & \hspace{-3.95 in } \mbox{and } \nonumber \\
z\int { d\lambda \over 2\pi } e^{-i\lambda /z } \langle 0| \psi ( 0 ) |hx \rangle \langle hx| \bar{\psi }
( \lambda n ) |0 \rangle & = & 4 m \hat{e}_1 ( z ) \ , .\end{aligned}\ ] ] here @xmath84 have twists 2 , 3 , 4 , respectively ; and @xmath85 is a generic qcd mass scale . since the twists for light - cone bilocal operators only represent the leading @xmath18 dependence , it is possible for @xmath86 and@xmath87 to include multiplicative factors of @xmath88 .
therefore eq . ( 20 ) can be rewritten as ; @xmath89 ; \nonumber \\ & & \hspace{-0.3 in } \mbox{and } \ , ; \nonumber \\ & & \hspace{-0.3 in } z\int { d\lambda \over 2\pi } e^{-i\lambda /z } \langle 0| \psi ( 0 ) |hx \rangle \langle hx| \bar{\psi } ( \lambda n ) |0 \rangle = 4 m \sum_{n=1}^\infty ( { m \over \sqrt{q^2}})^n \hat{e}_{1n } ( z ) \ , .\end{aligned}\ ] ] one can see from eq .
( 17 ) that each @xmath90 factor always has a gauge - covariant derivative as a companion in the @xmath91 terms ; i.e. @xmath92 . using the light - cone , @xmath93 as @xmath94 and the twist analysis of bilocal operators @xcite , one concludes that the @xmath95 terms have higher twists than their corresponding @xmath96 terms .
for example , a general @xmath95 term of order @xmath20 has the following twist expansion ; @xmath97 + \ , trace\,\ , terms \ , .\end{aligned}\ ] ] based on the twist analysis from bilocal operators ,
the leading twists for @xmath98 , @xmath99 , and @xmath100 are @xmath101 , @xmath102 , and @xmath103 respectively . since all the @xmath104 have higher twists , their corresponding diagonal terms , eq .
( 22 ) , can be written as ; @xmath105 \ , .\end{aligned}\ ] ] if there is final state interaction , there can be additional terms in @xmath58 .
@xcite has demonstrated the existence of such a terms , @xmath106 in more complicated fragmentation processes , more quarks and gluons are usually involved in the operators , and these lead to larger dimensions . in any case ,
twist expansions can still be carried out based on a twist analysis scheme involving bilocal operators . in conclusion
, one can carry out twist expansions for all bilocal terms in eq .
( 19 ) , to obtain the twist expansion for @xmath58 ; @xmath107 twist two contributions to @xmath108 are the major contributions as @xmath109 or @xmath110 . as @xmath111 ,
higher twist contributions become important . depending on
the sign of the coefficients of @xmath112 , the duality phenomenon in the fragmentation sector could occur .
although one can not carry out non - perturbative calculations , the determination of the sign of @xmath113 could be measured in duality experiments .
such experiments are straightforward in @xmath0 collisions , where one can compare the cross sections of hadron , @xmath26 , production at fixed @xmath114 for various @xmath18 . for a given accuracy ,
@xmath0 collisions at lower @xmath18 require higher - twist contributions than those at larger @xmath18 . however , it is meaningless to attempt a twist expansion in the non - perturbative region , @xmath115 .
thus the fragmentation functions can also be expanded in terms of contributions of operators of various twists , and the phenomenon of duality can appear if the signs of the coefficients of the higher twist contributions are not positive definite . because complete non - perturbative calculations are not presently possible , a comparison of measurements of fragmentation at various @xmath18 , having the same @xmath114 , can provide information on the coefficients of higher twist contributions .
although we have studied only the production of scalar hadrons , a richer structure occurs if one includes polarization effects .
@xcite . the scattering matrix of the process @xmath117 is given by , @xmath118 \ , |ep \rangle \ , .\end{aligned}\ ] ] the lowest order of the electromagnetic contribution
is shown in fig .
[ fig : duality4 ] . = 8 cm
the scattering matrix can be written as @xmath119 where @xmath120 is the proton electromagnetic current , and depends on the proton structure . from eq .
( 26 ) , one has , @xmath121 this is similar to @xmath27 , where @xmath122 due to the non - trivial dependence of @xmath37 on the observed hadron , @xmath26 .
one can work out a similar expansion as was done in the last subsection , to obtain the twist expansion of eq .
the major difference between eq .
( 28 ) and eq .
( 5 ) is that eq . (
28 ) contains information on the parton structure functions in addition to information on parton fragmentation functions .
the phenomenon of duality in parton structure functions has been established through inclusive @xmath23 scattering . in order to extract higher twist information in the parton fragmentation sector using semi - inclusive @xmath23 scattering , one needs to separate the contributions from the parton fragmentation and structure sectors .
if one allows factorization in the strong interaction , one should be able to separate the scale , @xmath123 , in the structure sector from the scale , @xmath124 , in the fragmentation sector . in general
, it is more advantageous to obtain higher twist information in the fragmentation sector using @xmath0 scattering . in order to extract information from @xmath23 scattering
, one needs to subtract the contributions from the structure sector .
this can be achieved by suitably choosing @xmath123 and @xmath124 in an experiment .
one way to achieve this is to keep the @xmath123 sufficiently high so that the contributions from the parton structure sectors are mainly from twist two operators which can be confidently calculated
. then one can suitably choose the @xmath124 functions in the fragmentation sector to obtain information of higher twist contributions .
therefore to quantitatively study higher - twist contributions in the parton fragmentation sector , it is critical to identify @xmath123 and @xmath124 and separate the parton fragmentation sector from the structure sector .
an experimental effort to detect higher twist effects in the fragmentation sector through semi - inclusive @xmath125 scattering has been recently undertaken at jlab .
we comment on this process in more detail here with the intent to clarify the experimental results .
the object of such an experiment would be to determine higher twist effects by comparing measurements which include effects from all twists to the calculated twist two results . as discussed previously , in order to study higher twist effects in fragmentation sectors through semi - inclusive @xmath125 scattering , it is important to subtract the contributions from the parton structure functions .
this is feasible by choosing high @xmath123 in the structure sector so that these contributions are predominantly from twist two operators which can be calculated perturbatively by using the well - known information on twist two parton structure functions .
one must have measurements in a sufficiently high region of @xmath123 , with various @xmath124 , in order to compare the measurements with calculations of twist two contributions in the same regions .
the differences between the measurements and the calculations provide information of higher twist effects in the fragmentation sector .
therefore we need to locate the regions with high @xmath123 and various @xmath124 to carry out the twist - two calculations . if @xmath123 is large enough so that the contributions from the parton structure sectors are twist two , the scattering process @xmath125 is illustrated in fig .
[ fig : duality5 ] .
= 8 cm before presenting a calculation of the process shown in fig .
[ fig : duality5 ] , we would like to make several remarks . 1 . in a momentum infinitive reference frame
where @xmath126 , parton momentum and energy are usually expressed as a fraction @xmath127 of the momentum ( energy ) of the parent proton momentum ( energy ) , where @xmath128 . in the cms or lab frame ,
a jlab beam energy ( @xmath129 ) is too low to use @xmath128 as the momentum fraction since the beam momentum @xmath130 in both cms and lab does not satisfy @xmath131 . in our approach in this section , we work in the target rest frame . while the target mass , @xmath132 , is non - negligible when compared to the beam energy , the quark masses are .
therefore , the quark energy and momentum in the target rest frame can be expressed as ; @xmath133 + in a finite momentum reference frame , parton transverse momentum can not be ignored and the parton structure function @xmath134 should depend on @xmath127 , @xmath18 , and @xmath135 , i.e. @xmath136 , where @xmath137 .
this structure function should be the same as the one measured in an infinite momentum reference frame , if one boosts the reference frame to a very high momentum so that @xmath135 is negligible . in a target - rest reference frame ,
the probability for a parton to have a fraction @xmath127 of the parent proton s energy should be the same as the corresponding parton structure function measured in momentum - infinitive reference frame because @xmath137 .
however , the direction of the momentum is completely random and satisfies @xmath138 .
2 . in @xmath0 scattering ,
the @xmath124 is defined as @xmath139 and is equal to @xmath140 in the cms .
if one concentrates on the fragmentation functions given by fig .
[ fig : duality2 ] , the energy of the produced quarks , @xmath141 , is equal to @xmath142 in the cms .
therefore @xmath124 can be written as ; @xmath143 the fraction of energy carried by the produced hadron from the parent parton is @xmath144 , which is equal to @xmath145 in the cms .
+ in semi - inclusive @xmath1 scattering , one usually defines ( see fig . [ fig : duality4 ] ) @xcite @xmath146 in order to adopt the fragmentation functions obtained from @xmath0 scattering in the @xmath1 scattering process , i.e. @xmath147 one should use the same definitions used in @xmath0 scattering .
therefore , we correspondingly redefine the @xmath124 and @xmath114 in @xmath1 scattering as @xmath148 compared to @xmath124 and @xmath114 in @xmath149 scattering , we believe that these definitions are reasonable and the new @xmath124 is at least proportional to @xmath124 in @xmath149 .
the @xmath123 in @xmath23 scattering is defined as usual as ; @xmath150 the differential cross section of the process @xmath151 is given by , @xmath152 where @xmath130 , @xmath153 , @xmath154 , and @xmath155 are four momenta of the incoming electron , scattered electron , quark within a target proton , and scattered quark , respectively .
also @xmath156 is the structure function of quark @xmath31 in a target proton with @xmath127 fraction of the proton energy , @xmath157 is the fragmentation function of quark @xmath31 fragmenting to hadron @xmath26 with @xmath114 fraction of the quark @xmath31 energy , @xmath158 is the solid angle of quark @xmath31 within the target in the lab frame , and the @xmath159-function reflects energy - momentum conservations among @xmath130 , @xmath153 , @xmath154 , and @xmath155 .
we define the electron beam direction as the positive @xmath114 direction and let ; @xmath160 eq . ( 34 ) can be rewritten as ; @xmath161 where @xmath162 using the parton structure functions from ref . @xcite and the pion fragmentation functions from ref .
@xcite , the pion production cross sections versus z at @xmath163 and @xmath164 and @xmath165 are plotted in fig .
[ fig : duality6 ] . in general
, the production cross section is small and the choices of @xmath166 that satisfy all conditions are limited .
this can be seen from the very small cross section at @xmath167 in fig .
[ fig : duality6 ] when @xmath163 .
this problem can be avoided by lowering the beam energy , which leads to lower @xmath123 and more choices of @xmath166 . however , lower @xmath123 brings more contributions from higher twist operators in the structure sector .
when the magnitude of @xmath123 is sufficient that the parton structure functions are safely represented by twist 2 operators , one can compare the hadron production data at various @xmath166 to obtain information on higher twist contributions in the fragmentation sector .
= 8 cm the expressions given in eq . ( 34 ) are given for the purpose of calculating the hadron production cross section versus @xmath114 at @xmath123 and @xmath166 . in experiments ,
the quantities that can be directly measured are @xmath168 , @xmath169 , @xmath170 , @xmath171 , and approximately @xmath172 and @xmath173 , if one takes @xmath174 and @xmath175 .
the quantities that can not be directly measured are @xmath127 , @xmath176 , @xmath177 , @xmath166 , and @xmath114 .
one can determine @xmath127 , @xmath176 , @xmath177 , and @xmath166 based on energy and momentum conservation and the measurements of @xmath168 , @xmath169 , @xmath170 , @xmath172 , and @xmath173 with the following equations , @xmath178 also @xmath114 can be determined through the above determination of @xmath166 and the measurements of @xmath171 , @xmath179 with eqs .
( 38 ) and ( 39 ) , one can determine @xmath180 , @xmath181 , and @xmath114 of the hadron production processes and then obtain the cross section versus @xmath114 . comparing the cross sections at various @xmath181 but with the same @xmath180 and @xmath114 , one can deduce the higher twist contributions from the fragmentation sector .
we have studied possible higher twist qcd contributions in the fragmentation sector . in summary
we find the following . 1 . the contributions to @xmath182 cross section can be expanded as contributions of operators with various twists . if @xmath183 , twist - two operators are the dominant contribution . as one decrease @xmath124 ,
higher - twist contributions become more and more important .
if the sign of the coefficients in higher - twist contributions are not positive definite , oscillation similar to duality in @xmath1 inclusive scattering could appear .
as @xmath124 becomes low , qcd becomes completely non - perturbative , and the twist expansion becomes meaningless .
a study of higher - twist contributions can be carried out through @xmath27 by comparing measurements of cross sections versus @xmath114 at various @xmath124 . in order to carry out the same study using @xmath125
, one needs to make sure that the higher twist contributions from the parton structure sector are well determined , so that one can isolate the contributions from the fragmentation sector .
one way to achieve this is to choose large enough @xmath123 so that the contributions to the cross section from the structure sectors are dominantly from twist - two operators , which can be calculated with well measured structure functions and perturbation qcd . to achieve higher @xmath123 but different @xmath124 , one can suitably choose the detector angles carefully in @xmath125 experiments for various beam energies .
we have not attempted to study the evolution of the ratio,@xmath184 as a function of @xmath114 .
however we suggest that this ratio contains information of the higher twist effects in the fragmentation sector and should be studied in the future .
4 . to adopt the fragmentation functions obtained in @xmath182 in the process of @xmath125 , redefinitions of @xmath124 and @xmath114 are necessary .
we currently use the definitions in eq .
( 31 ) .
e. d. bloom and f. j. gilman , phys .
d * 4 * ( 1971 ) 2901 .
a. de rujula , h. georgi , and h. d. politzer , ann
* 103 * ( 1977 ) 315 .
x. ji and p. unrau , phys .
d * 52 * ( 1995 ) 72 .
e. c. poggio , h. r. quimaand , and s. weinberg , phys .
d * 13 * ( 1976 ) 1958 .
r. ent et al .
, duality in meson electroproduction ( jlab proposal ) .
r. l. jaffe et al , lectures on qcd , applications , frieder lenz , harald grie@xmath185hammer , dieter stoll , eds . , 1997 .
j. c. collins , d. e. soper , nucl .
* b 194 * ( 1982 ) 445 .
xiangdong ji , phys .
d * 49 * ( 1994 ) 114 .
f. e. close , an introdution to quarks and partons , ( academic press inc .
( london ) ltd , 1979 ) . c. bourrely et al .
, phys . rep .
* 177 * ( 1989 ) 319 .
p. chiappetta et al . , nucl .
b * 412 * ( 1994 ) 3 . | we study twist expansions for parton fragmentation functions based on the definition of the twist as an invariant matrix element of a light - cone , bilocal operator .
the results are then applied to a method which might be used to extract higher twist effects in the fragmentation sector using both @xmath0 , and @xmath1 collisions .
we discuss how to apply the later measurements to experiments at the jefferson national acceleration facility . |
even though the presence of dark matter has been firmly established in recent years , the nature of the dark matter particle remains an outstanding problem in particle physics .
the idea that the dark matter particle may perhaps be detected indirectly via the detection of potential annihilation products is promising .
one particularly interesting aspect of indirect demodulation is the detection of @xmath0-rays from dark matter annihilations because it will be readily tested with the _ gamma - ray large area telescope _ ( glast ) in the next few years .
the presence of sub solar mass subhalos ( microhalos ) in the milky way can be probed by searching for the proper motion of @xmath0-rays emitted from dark matter annihilations @xcite .
such a detection is of profound importance .
first and foremost , it will be a detection of dark matter .
in addition , it will provide information on the survival rate of microhalos in the solar neighborhood , and will place bounds on the mass of the dark matter particle , its annihilation cross section and its kinetic decoupling temperature @xcite .
in the cold dark matter ( cdm ) paradigm , the first bound structures form at high redshifts .
the minimum mass is set by the rms dark matter particle velocities dictated at kinetic decoupling at a temperature @xmath1 @xcite .
for the particular case of supersymmetric ( susy ) dark matter , this scale is @xmath2 @xcite .
the abundance of microhalos in the solar neighborhood is still under debate @xcite . as such
, it can be paremetrized by assuming that a certain fraction of the local dark matter density ( @xmath3 ) is in microhalos in a logarithmic mass interval @xcite , @xmath4 in this parameterization , @xmath5 takes the maximum value of @xmath6 when all of the local dark matter density is in objects in the logarithmic interval @xmath7 .
we can get an estimate on the range of values it may take by considering the subhalo mass function of dark matter halos .
the subhalo mass function has been studied in numerical simulations and found to be described by @xmath8 , normalized in a way such that for a milky way size halo , 10% of the mass of the halo is in subhalos of mass greater than @xmath9 .
preliminary results from n - body simulations @xcite , as well as approximate analytical arguments find that this mass function is preserved down to microhalo scales , with the exception that on sub solar mass scales the survival probability is reduced to only @xmath10$]% due to early rapid merging processes as well as potential interactions with stars @xcite . in this case
, the value of @xmath5 as defined above reduces to @xmath11
. a microhalo at a distance @xmath12 and with a velocity @xmath13 will exhibit a proper motion of magnitude @xmath14 $ ] in a time interval @xmath15 . given
a velocity distribution function @xmath16 ( assumed to be described by a maxwell - boltzmann , see @xcite ) , the number of microhalos with a proper motion greater than some threshold @xmath17 is then given by @xmath18 \
, dv \ , dd.\ ] ] here , @xmath19 $ ] , and @xmath20 is the maximum distance at which a microhalo is detectable ( see next section ) . the expected flux of @xmath0-rays from a microhalo of mass @xmath7 at a distance @xmath12 depends on the annihilation cross section @xmath21 and mass @xmath22 of the dark matter particle , the density profile of the microhalo , as well as the redshift at which the microhalo collapsed , @xmath23 . assuming an navarro - frenk - white profile @xcite and a spectrum given in @xcite the number of photons per unit area and time with energy greater than @xmath24 can be written as : @xmath25
[ cols="^,^ " , ] in order to detect the proper motion of microhalos that may be present in the solar neighborhood , it is necessary to survey the whole sky in the @xmath0-ray part of the spectrum .
this is where glast is ideal in performing a search for the proper motion of nearby microhalos .
glast will survey the whole sky with an integral sensitivity of @xmath26 for photons with energy above a few gev .
assuming an effective area of @xmath27 , and an exposure time of 10 years ( assuming the sky will be exposed uniformly for @xmath28 seconds per year based on the orbit configuration of the spacecraft @xcite ) , the sensitivity threshold for glast is @xmath29 photons .
the left panel of fig .
1 shows the number of photons expected from microhalos as a function of distance for the same effective area and exposure time as mentioned above . for this particular example
, the mass of the dark matter particle was assumed to be @xmath30 , with an annihilation cross - section of @xmath31 .
this particular choice of parameters represents an optimistic scenario for supersymmetric dark matter . the shaded area in fig .
1 shows the radius of the volume needed in order to enclose at least 1 microhalo for different values of the parameter @xmath5 .
the lower limit comes from assuming that @xmath32 , while the upper bound corresponds to @xmath33 . as can be gleaned from fig .
1 , within reasonable distance ( e.g. within the shaded area ) , the number of photons on earth in a 10 year exposure is between 100 - 1000 for microhalos with masses @xmath34 .
this number is well above the sensitivity threshold for glast , suggesting that glast will be able to perform a search for microhalos in the solar neighborhood .
if the fraction of the local dark matter density in microhalos of mass less than @xmath35 is more than 0.2% , then there should be at least one microhalo potentially detectable with glast at a high signal - to - noise ratio .
the dot - dashed lines show the distance where the maximum proper motion is greater or equal to 1 & 9 arcminutes .
the single photon angular resolution of glast is better than @xmath36 arcminutes for photons of energy above few gev .
however , the localization of a source which emits @xmath37 photons is improved by a factor of @xmath38 , thus placing a proper motion threshold of detection of @xmath39 .
the solid line shows the threshold at which microhalos will be resolved as extended objects in glast ( assuming a psf of 9 arcminutes ) . in this case , the prospects for proper motion detection are much better because the excess photon flux from adjacent resolution bins in extended objects could be used to better localize the position of the source @xcite .
the right panel of fig .
1 shows the number of microhalos per logarithmic mass interval that will be detectable based on the integral sensitivity of glast as a function of microhalo mass and the particle physics properties of the dark matter particle . here , @xmath40 , and the quantity @xmath41 is left as a free parameter ( where @xmath42 de $ ] ) .
the particular case of an optimistic choice of parameters for supersymmetric dark matter is shown with the dashed line .
however , it should be emphasized that the quantity @xmath43 can attain higher values than the optimistic case of supersymmetry , e.g. kaluza klein dark matter @xcite .
the solid lines represent iso - number contours of the number of microhalos detected that will also exhibit proper motion greater than @xmath17 .
such a detection can place constraints on the properties of the dark matter particle .
more specifically , the detection of at least 1 microhalo with proper motion places a bound on the mass of the lightest supersymmetric particle to be less than 600 gev , and its kinetic decoupling temperature to be @xmath44 { \rm mev}$ ] ( see @xcite for more details on the range of values of @xmath1 for supersymmetric dark matter ) .
the potential detection of microhalos in the solar neighborhood has profound consequences in our understanding of the nature of the dark matter particle .
it is therefore essential that glast includes a search for proper motion of @xmath0-rays in the data analysis .
such an analysis can benefit significantly if the lifetime of glast is maximized , thus increasing the baseline for proper motion detection . | i discuss the prospects for detecting the dark matter via the proper motion of sub solar mass dark matter halos in the vicinity of the solar neighbourhood .
microhalos that survive tidal disruption could exhibit proper motion of order few arcminutes per year .
for dark matter particles that couple to photons , such as the lightest supersymmetric or kaluza - klein particles , microhalos could be detected via their @xmath0-ray photon emission from annihilations .
a detection of proper motion of a microhalo in the @xmath0-ray part of the spectrum contains not only information about the particle physics properties of the dark matter particle , but also provides an insight into hierarchical structure formation at very early times .
address = theoretical division & isr division , los alamos national laboratory , los alamos , nm 87545 , usa |
dna is not the static double helix that structural pictures show . at biological temperature
its thermal fluctuations are very large , leading to temporary openings of the base pairs , and when temperature is increased up to @xmath0 or @xmath1c , depending on the sequence , the two strands separate in a transition often called `` dna melting '' @xcite .
these phenomena have been the object of many experimental and theoretical studies in the last few decades but there are still open questions .
one of them is the understanding of the role of the base - pair sequence .
there are many aspects in this problem , depending on the scale at which it is observed , i.e. the length of the dna molecule which is considered .
most of the earlier studies investigated natural dna molecules , with thousands or hundreds of thousands of base pairs @xcite .
recording the melting curves of such dna samples , for instance through the variation of the uv absorbance of a solution , analyzes the phenomena with a resolution of hundreds of base pairs or more .
single molecule experiments observing the mechanical denaturation of a long dna segment @xcite , which is thermally assisted , find a good correlation between the richness of the sequence in gc base pairs , harder to break than at pairs , and the force necessary to open the molecule , but the resolution is again of the order of 100 base pairs .
thus all these studies are not able to analyze sequence effects at the scale of a few base pairs , which is the relevant scale for the genetic code . on another hand
it is now possible to perform a systematic study of the fluctuations and melting of short dna sequences ( typically 20 to 60 base pairs ) , which can be specifically designed to study a particular effect and synthesized .
this allows new experimental studies , which , in turn raise interesting theoretical questions .
this is the object of the present paper .
section [ sec : experiments ] presents experimental results .
some of them have been obtained by other authors but they are briefly reviewed here because they provide interesting test results for the theoretical analysis presented in sec .
[ sec : theory ] .
others are new and demonstrate that the large fluctuations of at - rich regions of dna are affecting the properties of the molecule several base pairs away .
this result must be taken into account to satisfactorily describe the melting of some dna sequences with a mesoscopic model , as discussed in sec .
[ sec : theory ] .
the study of some sequences which pose a particular challenge to the theories of dna melting confirms the need to include the non - local effects of the at - rich segments of dna to properly model the thermal denaturation .
this gives a second indication that the influence of highly fluctuating regions of dna on their neighborhood can extend to some distance .
the famous double - helix structure of dna is made of two strands that carry organic bases that are bound in pairs by hydrogen bonds , keeping the two strands together .
there are 4 types of bases , called a , t , g , c , but only at and gc pairs are part of the structure of dna .
the at pairs are bound by two hydrogen bonds , while the stronger gc pairs are bound by three hydrogen bonds .
the covalent bonds that bind the atoms in the backbones and inside the bases are very strong whereas the hydrogen bonds that connect the two bases in a pair are much weaker .
they can be broken by thermal fluctuations at biological temperature , exposing the bases to the surrounding solvent before the base pair closes again .
these large fluctuations which are called the `` breathing of dna '' are well known by biologists .
early studies of these fluctuations relied on deuterium - proton exchange .
if dna is dissolved into deuterated water , when the opening of a base pair exposes to the solvent the protons that form the hydrogen bonds , the so - called imino protons , those protons can be exchanged with deuterium from the water molecules of the solvent .
the exchange rate can be accelerated by a catalyst .
then nmr can be used to detect the deuterium atoms within the dna molecule @xcite .
kinetic experiments show that the lifetime of base pairs ( time during which they stay closed ) is in the range of milliseconds at @xmath2c and 10 times more at @xmath3c . at biological temperatures
the experiments show that _ single base pair opening events are the only mode of base pair disruption .
_ this indicates that the large amplitude conformational change that is able to break a pair and expose a base to the solvent is a _ highly localized phenomenon_. the lifetime of the open state , estimated from kinetics data , is of a few tens of nanoseconds @xcite . at higher temperatures the picture changes .
thermal fluctuations can break the base pairs in large regions of dna , leading to the so - called `` dna bubbles '' , which grow when temperature is raised , until the full separation of the strands .
this _ melting transition _ of dna is very sharp for homogeneous sequences .
it can be easily studied because the absorption of uv light at @xmath4 nm increases drastically when the base pairs are unstacked .
the melting transition shows up therefore as a sharp rise of the uv absorption of the dna solution , which only occurs within a few degrees for a short dna homopolymer . for heteropolymers , such as dna with a natural base sequence ,
the melting profile is different because at - rich regions tend to melt at lower temperature than the gc - rich regions .
this leads to melting curves which are very sensitive to the details of the sequence , and not yet fully understood although some empirical models can give rather accurate predictions , at a resolution of a few hundreds of base pairs for long sequences @xcite . for long molecules the thermal separation of the two strands
is probably well described by a zipper mechanism .
this is different for short dna sequences of less than 100 base pairs .
very short oligomers melt in a two - state process in which they are either fully closed as double helices or fully separated in individual strands .
longer molecules can exist in intermediate states in which a part of the molecule is open and the other part is closed .
for instance they can have either open ends or open `` bubbles '' in the middle .
this raises a question to experimentalists .
measuring the uv absorbance gives the fraction @xmath5 of open base pairs , but should a fraction of say @xmath6 be interpreted as meaning that 50% of the molecules are fully closed and the other 50% fully open , or does it means that all the molecules are in an intermediate state where half of their base pairs are open .
this question can only be answered by also determining the fraction @xmath7 of fully open molecules . using a clever trick , made possible by the availability of artificial dna sequences that can be tailored for a particular purpose , montrichok et al .
@xcite managed to do such a measurement .
the idea is to use sequences which are partly self - complementary .
each strand is such that a set of bases at one end are complementary of the bases pairs at the other end . as a result , when it is free , such a strand tends to fold into a hairpin , the two ends forming a short dna double helix closed by a loop made by the central section of the strand , which does not include self - complementary sequences .
when a solution containing double helices made of such strands is heated until a complete melting of all the molecules , and then cooled down , the strands can either form hairpins or get together in pairs to reform the double helix .
but the formation of the hairpins from single strands is much faster because two strands do not have to find each other by diffusing in the solution .
the experimental process used by montricok et al .
is the following .
they slowly heat a solution of the dna under study and record its uv absorbance that gives @xmath5 versus temperature . at selected temperatures
they take an aliquot and cool it down quickly .
all molecules which are fully melted lead to the formation of hairpins while those which are only partly open reclose as double helices .
this allows them to determine @xmath7 at these selected temperatures .
this study @xcite provides an interesting set of experimental results on dna melting because , from the data , it is possible to extract the average length of the denaturation bubbles as a function of temperature and the statistical weight of bubble states . as discussed in sec .
[ sec : theory ] , owing to peculiar properties of some of the investigated sequences , these experiments raise very challenging questions for the theoretical analysis of dna melting .
proton - deuterium exchange measurements , and melting studies measuring @xmath5 and @xmath7 provide a good set of experimental data of dna fluctuations and melting , but they lack an essential piece of information , the spatial information that says precisely where the fluctuations occur .
this is why we performed a set of experiments , using an original method , that can provide a mapping of the strength of the fluctuations of the double helix as a function of the position of along the sequence .
this adds a new dimension to the melting curves . usually extracting local information on dna fluctuations
relies on a dye or fluorophore probe inserted in a special molecular construct , which reports on the properties of dna in its vicinity @xcite .
this method has two major drawbacks .
first it provides information on a single point in the sequence , the place where the probe is located .
second the probe is probably not innocuous and it can perturb the dynamics of dna in its vicinity .
_ our method asks dna itself to report on its internal state .
_ it uses the propensity of the guanine bases g to be ionized by biphotonic excitation from a strong uv laser pulse .
two guanine modifications , oxazolone and 8-oxo-7,8-dihydro-2-oxoguanine ( 8-oxodg ) have been identified as the major one - electron oxidative dna lesions .
their formation depends on the local dna conformation and charge - transfer efficiency @xcite which is affected by local fluctuations .
while oxazolone is the unique product resulting from one - electron oxidation of the free @xmath8-deoxyguanosine , 8-oxodg appears as soon as the nucleoside is incorporated in a helical structure .
hence the measurement of the relative yield of these photoproducts at each g site tells us whether this g was in an helical structure ( closed ) or whether it was open when the molecule was hit by the laser pulse .
therefore it gives a snapshot of the opening of dna at each g site . as the experiment is not performed on a single molecule but in solution the results
are obtained on a statistical ensemble and they give the opening probability at each g site at the temperature of the study . to measure the relative yield of the production of oxazolone and 8-oxodg we can rely on standard biological methods .
piperidine cleaves dna at the sites where it has oxazolone , i.e. where the guanine was unstacked ( open base pair ) , whereas fpg protein cleaves it as sites which have 8-oxodg .
one end of the molecule is marked by a radioactive marker , and , after cleavage by piperidine or fpg , performed on different aliquots at the same temperature , the lengths of the fragments that carry the marker are measured by gel electrophoresis .
the labelled fragments of length @xmath9 are those which were cleaved at a guanine situated at distance @xmath9 from the marked end .
the ratio @xmath10 of the number of radioactively labeled fragments of length @xmath9 produced by fpg cleavage or by piperidine cleavage indicates the probability of closing for this particular guanine . as the gel image contains the data for fragments of all lengths , a single experiment measures the opening probability at all guanine sites . since our goal was to study sequence effects we performed these studies on two artificial sequences , specially designed to investigate the role of an at - rich region in dna .
such at - rich regions , called tata boxes , exist in the transcription - initiation regions of the genes of various species .
as the at base pairs bound by only two hydrogen bonds are weaker than the gc pairs , they are expected to exhibit large fluctuations even at biological temperature because they are closer to their melting temperature . depending on its length a poly - at dna melts around 55@xmath11c while a poly - gc melts around 80@xmath11c .
figure [ fig : sequv ] shows two sequences that we investigated .
sequence @xmath12 contains a tata box and two guanines , highlighted in the sequence , that were monitored as probes of the fluctuations of the dna molecule .
probe @xmath13 is separated from the tata box by a buffer region which contains 7 bases , while probe @xmath14 is adjacent to the tata box , in the @xmath15 direction of the dna strand . in sequence @xmath16 probe @xmath17
is surrounded by two small at - rich regions but the larger tata box and the guanine @xmath14 have been eliminated . in both cases
these short dna molecules are terminated by gc rich domains which act as clamps to prevent large fluctuations of the free ends of the molecules , and hold the two strands together even when we heat the sample up to 60@xmath11c . in principle
the guanines in these terminal parts could also be monitored in our analysis , but the spatial resolution near the end of the strands is not very good so that it is hard to separate two adjacent guanines in this region .
this is why we have not studied them .
[ cols="^,^ " , ] .
the probability that a given base pair is open at a particular temperature is shown as a function of the index of the base pair .
indices are those listed in fig .
[ fig : seq48 ] , the at - rich region starting at index 1.,width=340 ] however the comparison with the experimental results ( fig .
[ fig : montrichok48 ] shows that all theoretical approaches are _ qualitatively wrong ! _ it is important to stress that this is not a simple weakness of a particular model because all ising models , irrespectively of the parameter set that they use provided it is consistent with experiments for simple sequences , as well as our theoretical model which is very different , face the same problem .
moreover the discrepancy between theory and experiment is not a small effect .
the experiment finds that , although the opening occurs in two stages , the first stage extends well inside the gc rich region ( up to @xmath18 i.e. base pair 36 ) .
therefore , to understand this property one must look for a deep explanation rather than considering that it could be minor problem due , for instance , to incorrect parameters . this particular sequence , which seems to challenge all theoretical models is therefore particularly interesting .
a solution to this challenge can be provided by the experiments that we presented in sec .
[ sec : experiments ] .
they indicate that an at - rich region , which fluctuates widely , is able to influence the opening of base pairs which are at least 8 base pair away in the sequence . in the case of the experiment on sequence @xmath19
this effect could even be larger because the molecule opens at one end , so that the at - rich region can be expected to fluctuate even more than in our experiments , where it is clamped at both ends by one or several gc pairs .
this suggests that , in the presence of a large at - rich region _ the model should include non - local effects to describe its influence on parts of the sequence which are not right next to it .
_ in the experiments discussed in sec .
[ sec : experiments ] we noticed a lowering of the local melting temperature in the vicinity of an at - rich region , which is particularly strong right next to it and weaker further away . in the model
this effect can be described phenomenologically by a local rescaling of the parameters .
we know that the melting temperature decreases when the strength of the stacking interaction decreases @xcite .
moreover the large fluctuations in an at - rich region can be expected to disrupt the stacking in the vicinity .
this is why we decided to model the effect of an at - rich region by introducing a softening of the stacking interactions in the neighboring regions .
we selected the following rules * only at - rich regions which exceed a given size of @xmath20 consecutive at pairs are taken into account .
* inside the at - rich domain , the parameters are not modified .
* for all stacking interactions which are less than @xmath21 bonds from the at - rich region , the coupling constant @xmath22 is multiplied by a factor @xmath23 , * for all interactions which are @xmath24 bonds away from the at - rich region , @xmath25 , @xmath22 is multiplied by @xmath26 in order to reach smoothly the value 1 @xmath27 bonds away from the at - rich region .
* for all stacking interactions which are at least @xmath27 bonds away from the at - rich region , the parameters of the interaction potential are not modified . * if a stacking bond is affected by more than one at - rich region , which could be the case for the stacking interactions which are between two at - rich domains , only the largest correction is considered . from our experimental data on various sequences @xcite and from the theoretical studies of several cases , we set @xmath28 , @xmath29 , @xmath30 , @xmath31 , but these parameters could certainly be refined by more extensive experimental studies .
cc * ( a ) * & * ( b ) * + + when the effect of the large at - rich region on the stacking interactions along the sequence is taken into account . ( a ) dsdna melting curve .
the inset shows the derivative of the versus @xmath32 , with the same temperature scale as in the main figure .
( b ) the corresponding three - dimensional dsdna melting curve given by the model .
, title="fig:",height=188 ] & when the effect of the large at - rich region on the stacking interactions along the sequence is taken into account .
( a ) dsdna melting curve .
the inset shows the derivative of the versus @xmath32 , with the same temperature scale as in the main figure .
( b ) the corresponding three - dimensional dsdna melting curve given by the model . , title="fig:",height=188 ] figure [ fig : open48nonlocal ] shows the theoretical melting curve when the effect of the large at - rich region is taken into account .
the derivative of the dsdna melting curve still shows a two - stage melting but , contrary to the first calculation ( fig .
[ fig : montrichok48 ] ) , _ the first stage is now the dominant one , in agreement with experiments . _
the three - dimensional melting curve indicates that , in the first stage , the melting that starts in the at - rich end of the molecule extends inside the gc region until approximately 75% of the molecule , and the last part opens then at higher temperature .
the results are now in fairly good quantitative agreement with experiments , although the transition given by the model is still not as sharp as the experimental curve .
this could certainly be improved by optimizing the parameters , but the one - dimensional model is affected by strong finite - size effects which tend to broaden the denaturation transitions for short sequences .
nevertheless the results exhibit a qualitative change with respect to a calculation which only considered local properties of the molecule to determine the melting profile .
this is an important point because this improvement can not be achieved by adjusting parameters while simultaneously preserving the correct melting temperatures of at or gc homopolymers .
the sequence @xmath19 was particularly challenging .
most of the other short sequences are much less sensitive to subtle non - local effects , and therefore less demanding on the model to reach a satisfactory agreement between theory and experiments .
this is for instance the case for sequence @xmath33 of ref .
@xcite which has 60 base pairs .
it was designed with a 36 base pairs at - rich region in the middle , sandwiched by two gc rich domains .
the at - rich center is however interrupted by a few gc pairs , the largest consecutive span of at base pairs containing only 8 pairs . according to our experiments
@xcite a few isolated gc pairs scattered in an at domain are able to reduce significantly the fluctuations of this domain .
this is certainly the case for sequence @xmath33 , which explains why the small gc rich domains at the ends of the sequence are sufficient to stabilize the molecule and prevent a full separation of the strands in spite of the large open domain .
figure [ fig : open60 ] shows the melting profile given by the model , with the same parameters as for the sequence @xmath19 .
the three - dimensional profile clearly shows the `` bubble in the middle '' that was detected in the experiments @xcite .
cc * ( a ) * & * ( b ) * + + @xcite .
the inset shows the derivative of the versus @xmath32 , with the same temperature scale as in the main figure .
( b ) the corresponding three - dimensional dsdna melting curve given by the model .
, title="fig:",height=188 ] & @xcite .
the inset shows the derivative of the versus @xmath32 , with the same temperature scale as in the main figure .
( b ) the corresponding three - dimensional dsdna melting curve given by the model .
, title="fig:",height=188 ]
the structure of the melting curve of dna molecules of a few tens of base pairs contains information at a scale which is relevant for the genetic code , and this why its analysis is very interesting . in this work we showed that it is possible to add a spatial dimension to the melting curve through experiments that record the state of the guanines along the sequence
this is still incomplete because only one type of bases can be monitored but it nevertheless brings useful data on the interplay between different parts of the molecule during the melting . in particular the results indicate that an at - rich region , which fluctuates widely even at moderate temperature , can affect the fluctuations of the double helix several base pairs away .
this experimental result could be questioned because it is based on a measurement of the local fluctuations of the guanines which is not a direct observation , but involves complex transformations of the dna molecule , including electron transfers and further chemical reactions .
but it is important to notice that a theoretical analysis of independent measurements converges to the same conclusion . for this theoretical analysis we significantly modified the nonlinear model of dna that was introduced earlier .
adding a barrier for closing to the intra - pair potential @xmath34 is essential to get correct time scales for the lifetime of the open states . as long as one computes statistical quantities such as the mean stretching of base pairs at a given temperature , it could seem irrelevant to care about dynamics .
but this is however an important information which should not be neglected , and which is useful even for the calculation of static equilibrium properties because the model with a modified potential has different properties , such has a sharper melting transition .
the second evolution of the model was to introduce the sequence in the stacking interactions because these ineractions depend even more drastically from the sequence than the intra - base potential .
the analysis of the melting curve of the sequnce @xmath19 , which is challenging because standard ising models , as well as our model restricted to local effects , fail to give a correct melting curve , even at a qualitative level , showed the need of the introduction of the non - local effects of an at - rich region to get satisfactory results .
we described them phenomelogically by a rescaling of the stacking interactions .
this may not be the only choice , and further studies combining theory and experiments on various sequences are certainly necessary .
part of this work has been supported by rgion rne - alpes through its program cible . | understanding the melting of short dna sequences probes dna at the scale of the genetic code and raises questions which are very different from those posed by very long sequences , which have been extensively studied .
we investigate this problem by combining experiments and theory .
a new experimental method allows us to make a mapping of the opening of the guanines along the sequence as a function of temperature .
the results indicate that non - local effects may be important in dna because an at - rich region is able to influence the opening of a base pair which is about 10 base pairs away
. an earlier mesoscopic model of dna is modified to correctly describe the time scales associated to the opening of individual base pairs well below melting , and to properly take into account the sequence . using this model to analyze some characteristic sequences for which detailed experimental data on the melting is available [ montrichok et al .
2003 europhys
. lett . * 62 * 452 ] , we show that we have to introduce non - local effects of at - rich regions to get acceptable results .
this brings a second indication that the influence of these highly fluctuating regions of dna on their neighborhood can extend to some distance . |
quantum dots have often been termed artificial atoms and molecules.@xcite indeed , their tunability via electrical gates has permitted one to reach into a regime where only a single electron sits in each dot . a double dot then becomes the analogue of a hydrogen molecule .
the heisenberg exchange coupling in such a system can be manipulated by appropriately chosen electrical pulse sequences , up to a point which has not yet been achieved for its real counterpart.@xcite recently , it has been demonstrated that this technique allows the realization of the fundamental one- and two - qubit quantum gates @xcite in such a system.@xcite naturally , the question arises whether the successful schemes and techniques of an electrical control of spins in quantum dots can be transfered back to electron spins in single molecules . in view of the recent experimental progress in the field of molecular electronics , which demonstrates that magnetic transitions in atomic chains @xcite and single molecules @xcite can be resolved in transport measurements , such a goal should in principle be achievable . in particular , the field of molecular magnetism @xcite has provided a plethora of systems of almost arbitrary magnetic functionality.@xcite very recently , phase - coherence times of up to @xmath5 have been reported for such molecular nanomagnets after deuteration.@xcite yet , a suitable system for an electrical control of molecular qubits has still been missing . it is this electrical control which is crucial for scalability and which distinguishes this proposal from earlier ones also based on molecular magnets .
@xcite here we propose an experimental setup which permits the electrical switching of the exchange interactions between two electron spins .
the system we choose for illustrating our proposal is [ pmo@xmath0o@xmath1(vo)@xmath2@xmath3 , @xcite a polyoxometalate @xcite which consists of a central mixed - valence core based on the [ pmo@xmath0o@xmath1 ] keggin @xcite unit , able to act as an electron reservoir accommodating a variable number of delocalized electrons hopping over the mo centers , capped by two vanadyl groups containing two localized spins ( cf .
[ fig : model ] ) \(a ) ( b ) the spins on these two ( vo)@xmath7 units are weakly magnetically coupled via the delocalized electrons of the central core .
we shall show how this magnetic coupling can be switched in an all - electrical way and how this can be used for the implementation of a fundamental two - qubit gate ( providing entanglement ) , the so - called square - root - of - swap @xmath8 .
@xcite furthermore , we will detail how to use a variation of this procedure for the readout of the final state of the two qubits .
for the description of the low - energy states of the polyoxometalate , two cases have to be distinguished : for an _ even _ number of electrons on the mixed - valence keggin core , their spins pair antiferromagnetically to form a total spin @xmath9 state
. then the system can be modelled by the two spins @xmath10 on the vanadyl groups weakly coupled via an indirect exchange mechanism mediated by the core electrons . on the other hand ,
if the number of core electrons is _ odd _ , an unpaired spin @xmath10 on the core remains and one obtains a set of three coupled spins @xmath10 . restricting ourselves to two charge states differing by one electron , we can choose the electron number of the even state as reference and write the hamiltonian in the form @xmath11 here , the first term describes the indirect - exchange coupling between the left and right spins @xmath12 and @xmath13 with a coupling constant @xmath14 which depends on the number of electrons @xmath15measured with respect to the reference number on the central core of the molecule . in the case of an electron being localized on the core , its spin @xmath16 couples to the left and right spins with coupling constant @xmath17 . here ,
the operators @xmath18 ( @xmath19 ) destroy ( create ) and electron on the central core and @xmath20 is the vector of the three pauli matrices .
the two last terms contain the orbital energy @xmath21 of the electron on the central core , which can be shifted due to a gate potential @xmath22 , and the molecule s charging energy @xmath23 , respectively . the latter is assumed to be the largest energy scale of the problem .
we consider a situation where the central core of the molecule is tunnel - coupled to one or more metallic leads @xmath24 permitting electrons to flow on and off the molecule with tunnelling rates @xmath25 . in the stm setup depicted in fig .
[ fig : model](a ) , two leads , the tip and the metallic surface , are present .
we assume that an insulating layer which still allows electrons to tunnel through between the surface and the molecule leads to a molecule - surface coupling which is much smaller than the typical energy scale @xmath17 of the molecular hamiltonian .
an applied bias voltage @xmath26 then leads to a shift of the molecular levels which to a very good approximation is described by the linear relation @xmath27 with @xmath28 .
@xcite note that in a more sophisticated three - terminal setup , the parameters @xmath22 and @xmath29 can be controlled independently .
the @xmath30 operation is defined by @xmath31 , where @xmath32 . for electrically confined quantum dots
, it can be physically implemented by an appropriately chosen gate - voltage pulse which turns on the heisenberg exchange coupling between the qubits for a specific time ( see below).@xcite in a molecular system , however , the exchange constants are fixed by the chemical structure and can not be directly changed .
thus , a more sophisticated quantum - gate scheme has to be employed .
we propose to indirectly manipulate the coupling between the two qubits by changing the occupation @xmath15 of the central core in the molecule described above .
this change in @xmath15 can be induced by changing the gate voltage @xmath22 , such that different charge sectors of the molecule become stable ( cf .
[ fig : gating_sequence ] ) .
chemically this amounts to a change in the redox potential of the molecule .
we will first discuss the basic idea of the quantum - gate scheme , thereby assuming that the electron number @xmath15 can be changed in a deterministic , externally controllable way .
later on , this assumption will be relaxed and the full tunnelling dynamics due to the change in @xmath22 will be included . operation and exchange coupling constants during the corresponding gate phase . the gate is turned on for a time @xmath33 initially , one adjusts the gate voltage @xmath22 in such a way that the stable configuration is given by an even number of electrons on the central core ( @xmath34 ) .
we assume that the exchange coupling @xmath35 between the spins @xmath12 and @xmath13 will then be very small and can be disregarded for times much smaller than @xmath36 ( see also below ) .
the quantum - gate operation begins by changing the gate voltage such that the @xmath37 charge sector becomes stable .
then , on a time - scale of the inverse tunnelling rate , an electron enters the central core .
after this has happened , the dynamics of the three - qubit system is governed by the hamiltonian in the @xmath38 subspace .
the spin - dependent part of this hamiltonian becomes @xmath39 where @xmath40 is the total spin of the system and @xmath41 . before considering the time - evolution due to this hamiltonian
, we conclude our discussion of the gate cycle : after a specific time @xmath42 , one switches back to the initial gate voltage and the excess electron tunnels off the central core again .
after this has happened , the two outer spins are , up to times much smaller than @xmath36 , decoupled again . the only non - trivial dynamics is induced by the hamiltonian .
the first term contains an effective exchange coupling between the two spins .
up to an irrelevant phase factor , its corresponding time - evolution yields the @xmath30 gate if the gate time @xmath42 fulfils @xcite @xmath43 with @xmath44 being an arbitrary integer .
the second term in equation depends on the total spin of the three - spin-@xmath10 system .
while this quantity is unknown , we can eliminate its influence on the gate behavior if we restrict ourselves to stroboscopic gate times @xmath42 for which the contribution @xmath45 $ ] to the time - evolution operator is trivial .
evaluating the effect of the operator @xmath46 in the eigenbasis ( see also equation below ) , this is the case for gate times given by @xmath47 with the arbitrary integer @xmath48 .
it is important to note that relation is independent of the spin direction of the additional electron . equating conditions and , we obtain the following relation between @xmath49 and @xmath17 : @xmath50
in the preceding discussion we have assumed that the tunnelling process of the electron is highly controlled , i.e. , that we are able to instantaneously switch between the @xmath51 and @xmath38 states . in reality , tunnelling is a probabilistic event occurring on a mean time scale of the order of the inverse tunnelling rate . in order to investigate to
what extent this stochasticity affects our analytical findings , we compare them with numerical results obtained from a simulation of the incoherent quantum dynamics using the average gate fidelity @xmath52 as figure of merit for the gate process ( see app .
[ sec : br ] ) . in fig .
[ fig : fidelity ] , we show this quantity as a function of the gate time and the ratio @xmath53 for the case of a ferromagnetic coupling @xmath54 ( the antiferromagnetic case behaves very similarly ) . as a function of the gate time @xmath42 and the ratio @xmath53 for a ferromagnetic @xmath54 .
the conditions and are indicated by solid lines .
the parameters are : @xmath55 , @xmath56 , and @xmath57 ( measured from the charge - degeneracy point ) . ]
we focus on the strong - coupling regime @xmath58 , where tunnelling proceeds rapidly compared to the internal coherent dynamics .
as expected from our analytical considerations , the average gate fidelity assumes maxima whenever both conditions and are fulfilled .
we found the gate error @xmath59 to decrease with increasing tunnelling rate @xmath60 .
for @xmath61 , for instance , we obtain a fidelity @xmath62 at the maximum indicated by the circle in fig .
[ fig : fidelity ] .
we now show that in a molecular system described by the hamiltonian , it is possible to readout the quantum number @xmath63 of the two outer spins by measuring the sequential tunnelling current through the central dot .
the current can be calculated using the bloch - redfield equation approach ( see app . [
sec : br ] ) . for a qualitative understanding
it is sufficient to consider the allowed transitions , i.e. , non - zero matrix elements @xmath64 , between the eigenstates of the two different charge sectors @xmath51 and @xmath38 .
the former are the eigenstates @xmath65 of the total spin @xmath66 of the two outer spins .
the hamiltonian of the molecule with an additional electron can be readily diagonalized by rewriting it up to an irrelevant constant as @xmath67 thus , in the basis of the simultaneous eigenstates @xmath68 of @xmath46 , @xmath69 and @xmath70 , the hamiltonian in the @xmath38 subsector is already diagonal . for the matrix elements
@xmath71 , we then obtain the selection rules @xmath72 , @xmath73 , @xmath74 .
in particular , we note that the quantum number @xmath63 stays invariant under a complete sequential - tunnelling cycle .
since furthermore the energy differences determining the tunnelling rates and thus the current depend on the quantum number @xmath63 , we find that the ( quasi-)stationary value of the current can be strongly dependent on the initial preparation of @xmath63 . this situation is depicted in fig .
[ fig : readout ] , where the conductance peaks as a function of gate and bias voltage are shown schematically for two different initial values of @xmath63 .
( dashed line ) and @xmath75 ( solid line ) . ] from the figure we see that by starting initially at zero bias @xmath76 in the @xmath51 state , we first reach , upon increasing @xmath29 , the allowed ground - state transition and thus the onset of the sequential tunnelling current for the triplet @xmath75 . on the other hand , for a singlet preparation @xmath77 , sequential tunnelling will only be possible at a voltage which is higher by @xmath78 . at low enough tunnelling rate and temperature ,
i.e. , @xmath79 , it is possible to distinguish these two cases and hence to measure the quantum number @xmath63 .
as discussed before , for an independent control of the two parameters @xmath22 and @xmath29 in fig .
[ fig : readout ] , a three - terminal geometry is required . in the stm setup discussed here , one will only move along a line in fig .
[ fig : readout ] , which , however , is sufficient for the readout scheme , if the molecule is in the @xmath34 charge sector at @xmath76 .
note that for an initial superposition with singlet and triplet contributions , the readout process just described represents a projective measurement .
the first electron attempting to tunnel determines the final outcome , i.e. , the long - time current .
the average over many repeated measurements is described by the solution of the master equation .
we remark that the singlet - triplet readout process requires a substantially smaller tunnel coupling than the one for the exchange - controlled quantum - gate operation .
there are several ways how to achieve this : obviously , one could increase the molecule - tip distance which leads to an exponentially strong change in the tunnelling rate .
still , doing so on a very short time - scale might be technically challenging .
another possibility would be to use that the tunnel barrier and , thus , in turn the coupling strength depends on both the gate- and the bias - voltage .
@xcite depending on the details of the contact , this might be enough to achieve the required change in the tunnel coupling .
alternatively , one could transfer the readout scheme to the so - called cotunnelling regime , @xcite where transport proceeds via virtual tunnelling on and off the molecule . then
, the onset of the inelastic cotunnelling current is solely temperature - smeared , even in the presence of a strong molecule - lead coupling .
for measuring @xmath63 one would then need to apply a magnetic field , which will lead to a splitting of the triplet state @xmath75 only .
thus , one will find only for this state an inelastic cotunnelling step in the conductance - voltage characteristics at the bias voltage corresponding to the zeeman energy .
while the preceding discussion applies to any system described by a hamiltonian of the form , we now return to the specific case of the molecule [ pmo@xmath0o@xmath1(vo)@xmath2@xmath80 .
a high redox flexibility is a characteristical feature of polyoxometalates , where cyclic voltammetry experiments show one- or two - electron reversible redox peaks , depending on the system and on the conditions . in particular , four compounds based on the xmo@xmath0o@xmath1 ( x = si , ge , p ) keggin structure ( either alone or vanadyl - bicapped )
have been recently found to show reversible two - electron electrochemical processes .
@xcite we thus extract the exchange coupling strengths for different charge states @xmath81 of the molecule ( see app . [
sec : par ] ) . identifying even and odd values of @xmath81 with @xmath51 and @xmath38 , respectively ,
we obtain the parameters in the hamiltonian as given in table [ par ] for @xmath82 ; for higher electronic populations @xmath83 , @xmath49 and @xmath84 are of comparable size .
cc c|d|c @xmath81 & & @xmath85 $ ] + 0 & 0 & @xmath86 + 2 & 0.01 & @xmath87 + 4 & 0.01 & @xmath87 + 6 & 1.0 & @xmath87 + & c|d|d|d|c @xmath81 & & & & @xmath88 + 1 & 1 & 0.01 & < 0.1 & @xmath87 + 3 & 1 & 0.1 & < 0.1 & @xmath89 + 5 & 1 & 0.1 & < 0.1 & @xmath90 + it is worthwhile to consider how robust our predictions are against deviations from the ideal experimental setup .
specifically , ( i ) the orientation of the molecules may be difficult to control , ( ii ) its electronic structure , and , specifically , its ideal symmetry , may be altered upon deposition on a surface , ( iii ) the molecule might be affected by counter ions in the surrounding , and ( iv ) higher - energy states not described by the hamiltonian might be energetically accessible . with respect to ( i )
we note that regardless of the orientation of the molecule , the change in charge state is firmly expected to affect the central mo@xmath0o@xmath1 core and not the vanadyl groups , so we think orientation will only have an additional effect on ( ii ) . as an insulating thin layer separates the molecule from the metallic substrate
, the alteration in the electronic structure of the polyoxometalate is expected to be minimal ( details are discussed in the app . [
sec : par ] ) . the main effect on the low - energy physics will be to introduce an asymmetry in the couplings between the left and right vanadyl groups and the keggin structure , @xmath91 and @xmath92 , respectively . in table i
, we give bounds for this asymmetry , which can be quantified by the parameter @xmath93 .
furthermore , we have verified that an asymmetry of up to @xmath94 does still allow one to achieve a fidelity of @xmath95 in the regime @xmath96 ( for details see app .
[ sec : quant ] ) .
concerning ( iii ) , cations needed to compensate the negative charge of the polyoxometalate polyanions , as well as solvent molecules , may be still present when the cluster is deposited .
this feature may be a potentially important addition on ( ii ) which could be minimized using an electrospray system to choose the mass / charge ratio corresponding to an isolated polyoxoanion .
such techniques have succesfully been applied to polyoxometalates and the peaks of the isolated anions have been detected.@xcite finally ( iv ) , the full diagonalization of the effective hamiltonian also results in a set of excited states with energies @xmath97 above those states described by the hamiltonian : we also include these data in table i. having these points in mind , we see that different charge states are usable .
if we consider for instance @xmath98 and @xmath99 , one would need molecule - lead coupling rates @xmath100 in the range of around @xmath101 and @xmath102 , for the gate and readout process , respectively .
for the gate procedure , voltage pulses of a precision on the order of a few @xmath103 are required , which is not far from what can be achieved with present technology .
in fact , if @xmath104 and @xmath105 are found to be chemically stable in the experimental environment , their parameters will be even more suitable for our purposes by allowing @xmath106 times longer pulse times and yielding @xmath107 as well as @xmath108 .
the present scheme defines a class of molecular systems : those where two localized spins are coupled to a redox - active unit . the physically relevant figures of merit are included in table [ par ] ; the chemical desiderata are stability , facility of deposition on surfaces , and possibility of controlled oligomerization .
indeed , scalability of the present scheme requires covalent bonding and directed self - assembly of these logical building blocks .
molecules of this class can be found in several chemical families , besides bicapped polyoxometalates .
tetrathiofulvalene derivatives , polypyrrol , porphyrines / phthalocyanines , fullerene and single - wall nanotubes are a few among a plethora of organic systems which can reversibly lose or gain one electron , can be substituted with radical groups , and have a rich and well - controlled chemistry .
for example , theoretical calculations on phthalocyanines point to the possibility of the chemical tailoring of these molecules to enable exchange couplings in the range required for the gate scheme described here @xcite .
while for an initial experimental realization with a single molecule , an stm contacting scheme as discussed in the present article enables the best control and is clearly favorable , a scalable method for a molecular monolayer could be based , e.g. , on a crossbar architecture which already with current technology reaches very high densities.@xcite in conclusion , we have proposed an experimental setup for single - molecule all - electric two - qubit gate and readout which is within reach of current technology .
we have exemplified our scheme using a mixed - valence polyoxometalate , for which the model parameters have been calculated using an _ ab initio _ approach .
the general principles behind our proposal also apply to different classes of molecular systems .
the chemical design , synthesis and characterization of such systems should open new routes to molecular spin - qubit quantum computing .
we thank b. coish for discussions .
financial support by the eu rtn quemolna , the eu noe magmanet , the nccr nanoscience , the swiss nsf , the spanish mec ( mat2004 - 3849 ) and the generalitat valenciana is acknowledged .
for the description of the molecular dynamics in the presence of the molecule - lead coupling we employ a bloch - redfield - equation formalism .
this approach , which is valid in the sequential - tunnelling regime , i.e. , to lowest order in the tunnelling rates @xmath25 , includes the full coherent quantum dynamics of the spin states @xcite as described by the density matrix elements @xmath109 after tracing out the leads . in the usual born - markov approximation
, we obtain @xmath110 \varrho_{\alpha'\beta'}\\ & \quad- w^\ell_{\alpha \beta'\beta'\alpha'}\ , \varrho_{\alpha'\beta } - ( w^\ell_{\beta \alpha'\alpha'\beta'})^\ast\ , \varrho_{\alpha\beta ' } \bigg\}\ , , \end{split}\ ] ] where we have introduced the frequencies @xmath111 and the transition rates @xmath112 due to tunnelling across a molecule - lead contact @xmath24 . the rates for tunnelling of an electron on and off the molecule assume the form @xmath113 and @xmath114 \sum_{\sigma } \langle \beta'|d^\dagger_{{\mathrm{c}}\sigma}|\beta\rangle\!\langle \alpha|d_{{\mathrm{c}},\sigma}|\alpha'\rangle$ ] , respectively .
here , we assume that the lead electrons are described by a fermi distribution @xmath115\}^{-1}$ ] at temperature @xmath116 and electro - chemical potential @xmath117 .
the tunnelling current across contact @xmath24 can be determined from the density matrix elements via @xmath118 . using the bloch - redfield equation approach , we are able to simulate a realistic gate cycle .
this allows us to calculate the average gate fidelity @xcite @xmath119}$ ] as quantitative measure for the gate quality . here
, the overline indicates the average over 16 unentangled input states @xmath120 , where @xmath121 , @xmath122 , @xmath123 , and @xmath124 , @xmath125 is the ideal result of the @xmath30 operation as defined above and @xmath126 is the state of the system according to the quantum dynamics at the final time @xmath127 .
for the evaluation of the exchange coupling constants and to check the validity of our numerical assumptions , we used an effective hamiltonian approach .
we considered the main one- and two - center phenomena in mixed - valence systems , namely magnetic exchange , electron transfer , coulomb repulsion and orbital energy , which were parametrized for [ pmo@xmath0o@xmath1(vo)@xmath2@xmath3 ( details will be published elsewhere ) .
thus , we diagonalized a 14-site @xmath128@xmath129@xmath130@xmath131 effective hamiltonian for different electronic populations .
we projected this full energy level scheme into a subsystem containing only the localized spins in the [ v@xmath132o ] groups , plus , for odd values of @xmath81 , an unpaired spin in the mo@xmath0o@xmath1 keggin structure . for @xmath81
_ even _ , @xmath83 is given by the energy difference between the lowest singlet and triplet states , while @xmath133 is the gap to the next energy level .
for @xmath81 _ odd _ , the evaluation of @xmath133 is equally direct . obtaining the exchange parameters
is also straightforward when a symmetric molecular structure is assumed : @xmath17 and @xmath49 can then be readily calculated from the energy differences between the first doublet and quadruplet states .
when deviations from symmetry are considered , the determination of the three exchange parameters @xmath17 , @xmath91 and @xmath92 requires the analysis of the wavefunctions .
the spatially - dependent microscopic parameters @xmath128 , @xmath129 , @xmath130 , and @xmath131 were obtained from ab - initio calculations of molecular subclusters , which accounts for inhomogeneities in the molecule .
the experimental setup is expected to imply small variations in bond distances and angles , as well as a different charge distribution in the surrounding of the molecule . in order to assess how these perturbations affect the parameters in the effective hamiltonian
, the calculations were repeated for a reasonable range of deviations ( @xmath134 in @xmath129 and @xmath128 , and @xmath135ev in @xmath131 ) . except for the case with @xmath136 , where the excited states not contained in the hamiltonian lie close in energy ,
the parameters of our interest proved to be robust enough within the orders of magnitude indicated in table [ par ] .
besides the average gate fidelity discussed in the main text , one can consider further quantities for assessing the quality of a quantum - gate operation @xcite .
one is the so - called gate purity @xmath137}\,,\ ] ] which is a measure for the decoherence of the gate . here
, the average is over all 16 unentangled input states @xmath138 , @xmath139 ( see app . [
sec : br ] for the definition of the @xmath140 ) .
for an ideal @xmath30 operation it reaches unity .
another measure is the entanglement capability @xmath141 which is defined as the smallest eigenvalue of the partial transpose of the density matrix @xcite @xmath142 obtained for all 16 unentangled input states just defined .
ideally , it assumes a value of @xmath143 .
finally , the quantum degree of the gate is given by @xmath144 i.e. , as the maximum of the overlap between all possible output states obtained for all 16 unentangled input states and all maximally entangled states @xmath145 , viz , the four bell states latexmath:[$(|{\uparrow}\rangle_\mathrm{l}|{\uparrow}\rangle_\mathrm{r } \pm in fig .
[ fig : s1 ] , we show these four gate quantifiers as a function of the gating time @xmath42 and the ratio @xmath147 . the gate fidelity @xmath52 has already been discussed in the main text .
we observe that the gate purity @xmath148 assumes almost unity for the stroboscopic gating times ( 4 ) .
this means that at these times , the additional spin @xmath149 and the spin @xmath150 become disentangled again , such that one recovers a pure state when removing the additional electron .
furthermore , we find that all gate quantifiers except for the gate fidelity overestimate the quality of the gate : they almost assume their perfect values even for parameters where the gate fidelity is much below unity .
the quantum degree @xmath151 turns out to be completely independent of the gating time and the exchange - coupling ratio .
the situation in the presence of an asymmetry @xmath93 between the couplings @xmath91 and @xmath92 of the central keggin core to the left and right , respectively , vanadyl groups is shown in fig .
[ fig : s2 ] .
we have chosen the upper bound @xmath152 from the results of our _ ab initio _ calculations ( cf .
table i ) .
we find for the gating times shown in fig .
[ fig : s2 ] that this noticeable aberration from perfect symmetry does not strongly affect the gate quality for small and/or negative values of the ratio @xmath153 , where @xmath154 is the average coupling of the central spin to the left and right spins . in particular , we still find a fidelity better than @xmath62 at the maximum indicated by the circles in fig .
[ fig : s2 ] for a tunnel rate @xmath155 . | spin qubits offer one of the most promising routes to the implementation of quantum computers .
very recent results in semiconductor quantum dots show that electrically - controlled gating schemes are particularly well - suited for the realization of a universal set of quantum logical gates .
scalability to a larger number of qubits , however , remains an issue for such semiconductor quantum dots .
in contrast , a chemical bottom - up approach allows one to produce identical units in which localized spins represent the qubits .
molecular magnetism has produced a wide range of systems with tailored properties , but molecules permitting electrical gating have been lacking . here
we propose to use the polyoxometalate [ pmo@xmath0o@xmath1(vo)@xmath2@xmath3 , where two localized spins-@xmath4 can be coupled through the electrons of the central core . via electrical manipulation of the molecular redox potential , the charge of the core can be changed . with this setup , two - qubit gates and qubit readout can be implemented . |
the non - thermal @xmath0-ray spectrum of gamma - ray burst sources ( grbs ) is thought to arise in shocks which develop beyond the radius at which the relativistic fireball from the initial event has become optically thin to scattering .
however , the observed spectra are hard , with a significant fraction of the energy above the @xmath4 formation energy threshold , and a high compactness parameter can result in new pairs being formed outside the originally optically thin shocks responsible for the primary radiation .
new pairs can be made as some of the initial energetic photons are backscattered and interact with other incoming photons .
such effects have been considered by madau & thompson ( 2000 ) and thompson & madau ( 2000 ) , who investigated the acceleration of new pairs for a particular fireball model .
dermer & bttcher ( 2000 ) considered the effect of pair formation for an external shock model of grb , in which the compactness parameter is relatively low , while madau , blandford & rees ( 2000 ) investigated the effects of compton echos produced by pairs . here
we present a simplified discussion of the generic effects of pair formation in a specific class of models .
we suppose that the central engine gives rise to an unsteady baryonic wind , which is relativistic , carries a magnetic field , and lasts for a time @xmath5 . a fraction of the wind energy is converted into @xmath0-rays via internal shocks ; the remaining wind energy drives a blast wave which decelerates as it sweeps up the external medium , and gives rise to the afterglow emission .
the @xmath0-rays would propagate ahead of the blast wave , leading to pair production ( and an associated deposition of momentum ) into the external medium .
the pair cascades saturate after the external ( pair - enriched ) medium reaches a critical bulk lorentz factor , which is generally below that of the original relativistic wind . for external baryonic densities similar to those in molecular clouds the pairs can achieve scattering optical depths @xmath6 . even for less extreme external densities the effect of the additional pairs can be substantial , increasing the radiative efficiency of the blast wave and leading to distortions of the original spectrum .
this provides a potential tool for diagnosing the compactness parameter of the bursts and thus the radial distance at which shocks producing an observed luminosity can occur .
it also provides a tool for diagnosing the baryonic density of the external environment of the bursts , and testing the association with star - forming regions .
consider an initial input radiation spectrum produced by the fireball of the form ( ) = f m_e c^2 _ o^2 q ( /_o)^-
2si ; _ o , [ eq : spin ] where @xmath7 is the observed energy flux , @xmath8 is photon energy in electron rest - mass units , @xmath9 is the break energy above which the photon number spectral index @xmath10 , and below which the spectral slope is flatter , e.g. @xmath11 for a simple low - energy cutoff synchrotron spectrum ( although for the present purposes the exact low energy slope is unimportant ) , with @xmath12 a normalization constant . for most values of the low and high energy slopes , the majority of the photons in the spectrum are near the break frequency . for an @xmath13 moving away radially from the source of radiation with a velocity characterized by a lorentz factor @xmath14 , taking into account the klein - nishina drop - off in the scattering cross section above @xmath15 , the effective fraction of photons contributing to accelerating the electron is @xmath16 \simeq q^{-1 } \eps_o^{-1 } ( \beta-1)^{-1}$ ] for @xmath17 . at a distance @xmath18 in front of the radiation source
, an @xmath13 can be accelerated to a maximum value of @xmath14 satisfying ( /2 ) t m_e c^2 , [ eq : scatt ] where the cross section near @xmath19 is approximately @xmath20 , @xmath21 is the effective mass per scatterer ( equation [ [ eq : mu ] ] ) , and @xmath22 is the effective duration of the light pulse as seen by the electron .
the latter is either @xmath23 ( impulsive regime at small radii ) , or @xmath24 ( wind regime ) , depending on which one is smallest at a given radius , where @xmath5 is the duration of the wind , or essentially the burst duration as seen by a distant observer .
equation ( [ eq : scatt ] ) says that the total time - integrated momentum of the radiation intercepted by the electrons in the thompson limit is converted into their kinetic energy of motion , taking into account the effective mean mass per scatterer .
( this follows directly from the mean rate of momentum transfer per particle in a steady flow , e.g. , as discussed in madau & thompson , 2000 , their equation ( 2 ) , ( 10 ) and ( 23 ) , noting that our definition of @xmath25 differs from theirs . ) here we define @xmath25 as the effective mass per scattering electron or positron in units of electron mass , m_e = mass = m_p [ 1 + 2(/n_p)(m_e / m_p ) ] [ eq : mu ] defining a reference radius r_= l/16m_e c^3=510 ^ 1950 , [ eq : rast ] from equation ( [ eq : scatt ] ) the maximum @xmath13 lorentz factor is ( r ) [ eq : gammasc ] the transition from the @xmath26 to the steeper @xmath27 occurs at a critical radius @xmath28 , for which the lorentz factor is @xmath29 , r_c= & 2.210 ^ 14 ^ -2/5(q _ o ( -1))^-2/5 50 ^ 2/5t_w^3/5 + _ c= & 6.110 ^ 1 ^-1/5(q _ o ( -1))^2/15 50 ^ 1/5t_w^-1/5 .
[ eq : rcsc ] a second criterion for a maximum @xmath13 lorentz factor comes from the pair formation threshold , since the incident photon @xmath30 and the back - scattered photon @xmath31 must satisfy @xmath32 .
the back - scattered photon , has , in the reference frame of the scattering @xmath13 moving with @xmath33 , an energy of at most @xmath34 ; this photon can not give rise to a further pair unless it collides with another photon with , in the lab frame , an energy exceeding @xmath35 .
thus the fraction of incident photons able to make pairs through the two - photon mechanism against target photons backscattered from @xmath13 moving with @xmath33 is @xmath36 .
the maximum lorentz factor achievable by pairs before the two - photon cascade cuts off is that for which the compactness parameter has dropped to unity , ( r ) = l4r^2 _ o^-2 q ( -1 ) /3
m_e c^2 1(4)^-1 t 1 , [ eq : compact ] where the effective duration is as before .
the maximum lorentz factor for pair formation is then ( r ) .
[ eq : gammapp ] this yields a critical radius and lorentz factor for the transition between the impulsive and wind dominated regimes of r_c= & ( 2 ^ 7 - 3 9(-1)^2)^1 + 3 ( _ oq)^2 - 4 + 3 ( ct_w)^+1 + 3 510 ^ 1450 ^ 2/51 ^ 3/5 , + _ c= & ( 2 ^ -2(-1 ) 3(-1))^1 + 3 ( _ oq)^-2 + 3 ( ct_w)^1 + 3 310 ^ 1 50 ^ 1/51 ^ -1/5 , [ eq : rcpp ] where the numerical values in the second equation of both lines are calculated for @xmath37 , @xmath38 , @xmath39 erg s@xmath40 and @xmath41 s. ( at early times @xmath42 the radius @xmath43 is smaller , @xmath44 , since it takes @xmath5 for the entire photon energy to build up ) .
the maximum pair lorentz factor is shown in figure [ fig : fig1 ] for nominal parameter values .
the pair formation limit ( [ eq : rcpp ] ) is somewhat more restrictive than the scattering limit ( [ eq : rcsc ] ) .
also , unlike the scattering limit , the pair formation limit depends exponentially on the photon number slope @xmath45 . however , for the canonical value @xmath37 the two power law dependences are the same and the numerical values are close to each other . in what follows we shall use the pair formation limit of equations ( [ eq : gammapp ] ) and ( [ eq : rcpp ] ) .
given a certain external baryon density @xmath46 at a radius @xmath18 outside the shocks producing the grb primordial spectrum , the initial thomson scattering optical depth is @xmath47 and a fraction @xmath48 of the primordial photons will be scattered back , initiating a pair cascade . since the photon flux drops as @xmath49 , for a uniform ( or decreasing ) external ion density most of the scattering occurs between @xmath18 and @xmath50 , and the scattering and pair formation may be approximated as a local phenomenon .
this pair - dominated plasma , as long as its density @xmath51 , is initially held back by the inertia of its constituent ions component , provided that the pairs remain coupled to the baryons .
( the latter is likely to be the case in the presence of even weak magnetic fields , e.g. madau & thompson 2000 , or plasma wave scattering , e.g. lightman , 1982 .
moreover , the initial magnetic field strength can also increase as consequence of instabilities caused by an initial pair streaming relative to ions . in
what follows , we assume that this coupling is effective ) . the pairs ( together with the ions ) start acquiring a significant velocity @xmath52 only after their density @xmath53 , when the effective mass per scatterer @xmath54 .
after the pairs start being accelerated , the pair cascade can continue multiplying as long as the compactness parameter of equation ( [ eq : gammapp ] ) is @xmath55 .
both the compactness @xmath56 and the maximum @xmath33 decrease for increasing radius according to equations ( [ eq : compact ] ) and ( [ eq : gammapp ] ) , and the cascade process shuts off for @xmath57 at a radius where the maximum @xmath58 , r_~(4 c t_w /3 ) ^1/2~4 10 ^ 1550 ^ 1/21 ^ 1/2 .
[ eq : rell ] on the other hand , the blast wave producing the shock(s ) responsible for the afterglow spectrum starts decelerating ( i.e. its initial lorentz factor @xmath59 starts to decrease as a negative power law in radius ) at the deceleration radius r_d(3 l t_w/4n_p m_p c^2 ^2)^1/3 310 ^ 1550 ^ 1/31 ^ 1/3_2 ^ -2/3n_3 ^ -1/3 , [ eq : rdec ] where we normalized to an external baryon density @xmath60 .
( this expression is valid provided that the deceleration radius exceeds @xmath61 .
otherwise the deceleration starts before the outflow is over , and @xmath5 should be replaced by @xmath62 in the expression ) .
a larger value @xmath63 would give @xmath64 .
( such a higher density would , e.g. , be required for pair effects to be important in an external shock model , where the @xmath0-rays arise at @xmath65 ) . in an internal shock model
the primary @xmath0-rays themselves arise at a smaller radius r_i~c t_v ^2 ~310 ^ 13 t_v -1 _ 2 ^ 2 [ eq : rint ] where @xmath66 s is the variability timescale in the outflow .
the criterion for an @xmath13 cascade to form ahead of the fireball producing a primary spectrum ( [ eq : spin ] ) is that @xmath67 ( or @xmath65 for an external shock model ) is smaller than @xmath68 .
pair formation occurs then for @xmath69 .
the minimum number of pairs is formed just inside @xmath68 , where they are created essentially at rest , and their density can reach @xmath70 , at which point the mass per scatterer has become comparable to the electron mass .
this represents an increase in the electron ( or positron ) scattering opacity by a multiplication factor of @xmath71 .
as the density exceeds this value at @xmath68 the pairs are pushed by scattering beyond @xmath68 , where the condition @xmath72 prevents further multiplication . for radii @xmath73 , and as long as the blast wave has not reached that radius , pair formation exceeding the above value @xmath74 can continue to occur , with the pairs
being accelerated by scattering until they reach the maximum lorentz factor ( [ eq : gammapp ] ) .
pair production will be most copious at smaller radii @xmath75 , where the pairs are accelerated to a maximum lorentz factor @xmath76 . at each step of the cascade the lorentz factor of the new generation of pairs approximately doubles , and the maximum pair multiplication factor of the cascade can be estimated from the number @xmath77 of pair generations required to go from @xmath78 to @xmath79 given by equation ( [ eq : gammapp ] ) . at @xmath80 this is @xmath81 , and for @xmath82 it is limited to @xmath83 .
this represents an extra multiplication factor @xmath84 in the electron opacity for @xmath85 , in addition to the previous factor @xmath86 achieved before acceleration starts .
the critical value is k_c = k_p
k_a(r_c ) ~(m_p / m_e ) _ c ~ 510 ^ 4 50 ^ 1/51 ^ -1/5 , [ eq : kc ] with @xmath87 for @xmath88 , while @xmath89 for @xmath90 drops to @xmath91 at @xmath68 , and to zero beyond that . in a more detailed calculation ,
the number of cascades depends somewhat on the radiation spectral index @xmath45 , here set equal to 2 . for a steeper spectrum ,
the number of cascades is reduced ( e.g. ramirez - ruiz _ et al .
the maximum pair optical depth is achieved at @xmath43 , _
c ~k_c n_p r_c = . [ eq : tauc ] the pair opacity scales as @xmath92 for @xmath93 and @xmath94 for @xmath95 . for an external density
greater than n_p , c10 ^ 5 50 ^ -3/5 1 ^ -2/5 , [ eq : npc ] the pair density inside @xmath43 is prevented from multiplying beyond a value corresponding to @xmath96 , due to self - shielding .
the number of pairs created at @xmath43 is n_c~n_p k_c ( 4/3)r_c^3 ~310 ^ 49 n_p 50 ^ 7/5 1 ^ 8/5 , [ eq : npmc ] which scales as @xmath97 for @xmath18 below or above @xmath43 .
the maximum value for @xmath98 is @xmath99 at @xmath100 , and @xmath101 at @xmath68 .
the pair cloud extends over a range of radii comparable to @xmath18 , much larger than the photon pulse radial width @xmath102 .
figure [ fig : fig2 ] shows the schematic world - lines of three blast waves expanding into an external medium . at each radius
@xmath103 pairs form ahead of @xmath67 .
while for @xmath104 the pairs are sub - relativistic and do not propagate significantly , for smaller radii they move relativistically with an initial speed @xmath33 , with @xmath105 in general , although at very small radii , and for times @xmath42 , @xmath106 may be possible .
if the blast wave did not exist , then the pair - enriched material ( moving out with a lorentz factor that is larger at smaller r ) would pile up at a radius @xmath107 after a time @xmath108 .
however , the pair - rich external medium would generally carry less energy and inertia than the relativistic wind itself ; it therefore starts being decelerated by the external medium at a smaller radius than a blast wave with the same lorentz factor would be , so that the latter always overtakes it and sweeps up the pair - enriched external medium .
the resultant outer shock is more likely to be radiative , but may be weaker , because the ambient material might already be moving outward relativistically before the blast wave hits it . for a relatively low external density such as @xmath109 , as in curve ( a ) of figure [ fig : fig2 ] , the radius @xmath110 , and the pairs are swept up by a fireball expanding at its original speed @xmath111 constant , and deceleration of the fireball begins sometime after the pairs have been swept up . in this case
the main afterglow , produced by the deceleration of the pair - enriched fireball will be characterized by a larger emission measure than in the usual case , due to the extra pairs in the swept - up matter , but will not otherwise exhibit any irregularities in its time development . for a larger external density such as @xmath112 ,
corresponding to curve ( b ) of figure [ fig : fig2 ] , the radius @xmath113 and the main part of the afterglow , caused by the deceleration of the fireball baryons and the pairs formed at early stages , continues to sweep up further pairs which were formed at or beyond the deceleration radius ( in curve ( c ) , with low @xmath114 , deceleration occurs at radii @xmath115 ) .
these new pairs change the emission measure of the deceleration shock , since their optical depth @xmath96 is larger than that of the swept up baryonic electrons .
if @xmath116 is reached , the original spectrum may be modified by comptonization , and may approach a quasi - thermal shape .
we note that the optical depth can not substantially exceed unity .
this is because high - energy non - thermal photons would be comptonised by pairs , and after a few scatterings would be degraded below the pair - production threshold .
so , however intense the radiation incident on its inner edge , a shell of pair - dominated plasma can not build up an opacity much larger than unity .
it will come to a quasi - steady state in which the luminosity from its outer surface ( comptonised synchrotron radiation and annihilation radiation ) equals the non - thermal radiation shining on its inner surface .
the energy in the accelerated pairs is @xmath117 for @xmath118 and @xmath119 for @xmath120 .
the maximum number of pairs created is e_c n_m_e c^2_c = erg , [ eq : epm ] where the second number is the limit @xmath121 obtained for @xmath122 , at which the saturation value of @xmath96 is obtained .
the pair energy never exceeds more than a fraction of the initial wind energy , so the deceleration radius ( equation [ [ eq : rdec ] ] ) is unaffected , while the pairs which are produced and accelerated ahead of the blast wave , are decelerated at a ( smaller ) radius r_d~2.510 ^ 15 e_50 ^ 1/3 n_3 ^ -1/3(/30)^-2/3 cm , [ eq : rdecpm ] before the ejecta itself starts to decelerate .
the total number of protons in the ejecta wind is @xmath123 .
the number of protons in the ( pair enriched ) external medium which are swept up by the time the blast wave reaches its deceleration radius @xmath65 is @xmath124 .
the maximum number of swept - up pairs can exceed the number of baryons by a large factor . however the inertia of the pairs is less than that of the baryonic ejecta , since @xmath125 at @xmath43 , and @xmath126 at @xmath68 .
the pairs will modify the usual properties of the deceleration shocks and the afterglow emission .
the total energy available in the afterglow is not changed : it is still , essentially , the kinetic energy of the relativistic wind , minus the fraction dissipated ( and converted into prompt gamma rays ) in internal shocks .
but the lepton / proton ratio in the ejecta can be much larger than usual .
this increases the radiative efficiency significantly , since most of the particles are @xmath13 .
the resultant radiation will thus be softer than in the usual picture , because the same energy density has to be shared among a larger number of particles ( @xmath127 ) . in the reverse shock that occurs after the baryonic+pair
ejecta starts being decelerated by the external medium , the comoving frame peak random electron lorentz factor is @xmath128 . even for @xmath129
substantially less than unity , e.g. for an external density @xmath130 ( below the critical value @xmath131 ) for which the deceleration radius @xmath132 , the reverse shock random lepton lorentz factor is @xmath133 .
the pairs should not affect , however , the random turbulent magnetic field strength , which is in pressure equilibrium with the forward shock at some fraction @xmath134 of the equipartition value , @xmath135 g. thus the observed reverse synchrotron peak frequency _ , sy , r~10 ^ 6 b^2 ( 1+z)^-1 ~610 ^ 13 51 ^ -2/51 ^ -6/5_b^1/2 n_3 ^ -3/2_2 ^ -1 ( 1+z)^-1 hz [ eq : nusyr ] would be in the far ir , as opposed to the optical / uv of the usual baryon - dominated prompt reverse flash . the forward shock afterglow radiation , normally in the @xmath0/x - ray range ,
would be unaffected outside @xmath68 .
however , up to about the time when deceleration starts , there will be a
pair pick - up " photon pulse , when the protons and electrons of the relativistic wind moving with @xmath136 sweep up the pair - enriched external medium .
the protons are not decelerated by the pairs , so there is only a subsonic reverse compression wave in the proton ejecta , but there will be a mildly relativistic forward shock in the picked - up pairs .
the @xmath13 random lorentz factor will be equal to the bulk kinetic energy it has in the proton frame , e.g. @xmath137 near @xmath43 .
the comoving random magnetic field in the shocked pair fluid is @xmath138 @xmath139 g for @xmath140 , corresponding to a pair - pickup pulse synchrotron peak frequency in the observer frame of _ , sy , f ~210 ^ 10 _
b^1/250 ^ -1/51 ^ 5/10n_3 ^ 1/2_2 ^ 3 ( 1+z)^-1 hz , [ eq : nusyf ] and a synchrotron power law above this frequency .
there could also be additional pair - precursor signatures which are not associated with the ejecta blast wave , but with the dynamics of the pre - accelerated pair - enriched plasma .
the r - dependent @xmath33 of the pair - enriched external plasma will lead to internal shocks .
pair regions at @xmath141 whose @xmath33 differ by order unity will collide at radii @xmath142 , so for sufficiently short variability times these shocks would occur between @xmath67 and @xmath43 , i.e. up to an observer time @xmath143 s , independent of the wind luminosity @xmath144 .
( for @xmath145 the collision radii are @xmath146 , so the shocks would tend to pile up at @xmath68 at an observer timescale @xmath147 1 day , provided the main part of the fireball wind has not caught up with it before . this could be the case for a jet - like wind for pairs in a rim of angular width @xmath148 around the jet ) .
the pair internal shocks at @xmath149 can produce a fermi accelerated power law @xmath13 spectrum above @xmath150 and a turbulent comoving field @xmath151 g , leading to a synchrotron spectrum whose observer - frame peak frequency @xmath152 hz would be self - absorbed , with a power law extension above it which would be optically thin at higher frequencies . the above observational signatures would be present even if @xmath129 is low , as expected for external ion densities @xmath153 around the burst .
an additional effect of interest , for external densities in excess of this which lead to a pair screen of optical depth @xmath116 , is that a quasi - thermal pulse of x - rays could accompany the burst , caused by upscattering of diffuse progenitor stellar photons .
a collapsing massive progenitor leading to a grb is likely to be highly super - eddington for sometime after the grb event , e.g. @xmath154 erg / s .
the density of photons of energy @xmath155 ev near @xmath43 , which would be quasi - isotropized due to the condition @xmath116 at @xmath43 ( and beyond , where @xmath156 at @xmath68 ) , would be @xmath157 near @xmath43 .
the screen with @xmath96 moving with @xmath158 sweeps up a total number of photons @xmath159 .
the mean energy per photon and the time - integrated total energy of the upscattered precursor is _
x~ & 10 _ 10 50 ^ 2/51 ^ -2/5(1+z)^-1 kev + e_x~ & 510 ^ 49 l_43_10 ^ -150 ^ 4/5
1 ^ 1/5 ( 1+z)^-1 erg .
[ eq : xrprec ] this would last until the pairs are swept up by the ejecta , @xmath160 s.
we have discussed , in the context of a standard internal / external shock model of gamma - ray bursts ( which are normally assumed to occur after the original fireball has become optically thin ) the @xmath161 cascades triggered by the back - scattering of seed gamma - ray photons on the external medium .
this effect can modify the initial scattering optical depth of the outflow at radii @xmath162 ( equation [ [ eq : rell ] ] ) , which is comparable to the radii of external shocks of equation ( [ eq : rdec ] ) at which the afterglow begins , and is generally larger than the typical internal shock radii given by equation ( [ eq : rint ] ) .
the spectral effects of the pairs on the burst and the afterglow can be substantial , and within radii @xmath163 they can affect the dynamics . pair production can increase the optical depth outside of the shocks by up to @xmath164 ; self - shielding ensures that the maximum scattering optical depths achievable by the pairs is @xmath96 . for typical interstellar densities the pair opacity @xmath165 , which does not significantly affect the gamma - ray spectrum .
the number of pairs may nonetheless be large enough to increase the radiative efficiency and soften significantly the radiation spectrum of the afterglow reverse shock , where the same energy is shared among a number of @xmath13 which can exceed that of the original @xmath166 and @xmath167 of the ejecta .
the pair production processes themselves ( determining @xmath168 , equations [ [ eq : rcpp ] ] , [ [ eq : rell ] ] ) just depend on the seed " @xmath0-ray photon flux ( which are here postulated to come from internal shocks ) .
the manifestations depend on the external density and on the initial dimensionless entropy or bulk lorentz factor @xmath114 .
the external baryon density @xmath169 determines the optical depth that can be built up through back - scattering and pair multiplication .
this affects whether the pair optical depth gets up to unity , with smearing and reprocessing of the primordial @xmath0-ray spectrum , or whether it merely makes the blast wave more radiative .
madau and thompson ( 2000 ) have made this point , in the context of a specific fireball model .
the external density ( along with the initial lorentz factor @xmath114 ) determines when the outer shock and the reverse shock become important and whether this happens within the radius already polluted with pairs ( and pre - accelerated by radiation pressure before the shock hits ) .
there are two rather different cases depending on whether or not @xmath170 is less than @xmath171 ) . in the former case
the external shock responsible for the afterglow occurs beyond the region
polluted " by new pairs , while in the second case the afterglow shock may experience , after starting out in the canonical manner , a resurgence " or second kick as its radiative efficiency is boosted by running into an @xmath13-enriched gas .
internal shocks in the pair - dominated external plasma can lead to self - absorbed radiation at @xmath172 hz , while the swept - up pairs can also contribute a @xmath173 hz ` prompt ' signal , which precedes the onset of the standard deceleration afterglow phase .
additional effects are expected when @xmath174 .
this requires external baryon densities at radii @xmath73 of @xmath175 .
such high densities would only be expected if the burst is associated with a massive star in which prior mass loss led to a dense circumstellar envelope the pair optical depth saturates to @xmath96 and in addition to an increased efficiency and softer spectrum of the afterglow reverse shock , the original gamma - ray spectrum of the grb will be modified as well .
the specific nature of this spectral modification depends on the value of the luminosity , which influences ( equation [ [ eq : rcpp ] ] ) the bulk lorentz factor of the reprocessing pair cloud before it has been swept up by the ejecta , and also on the extent to which the outflow is beamed .
one of the consequences of such a critical external density leading to @xmath96 would be the presence of an x - ray quasi - thermal pulse , whose total energy may be a few percent of the total burst energy . in the case of even more extreme densities ,
there are other interesting possibilities .
for instance , dense blobs of fe - enriched thermal plasma would emit strong recombination features , as well as annihilation radiation features , if the normal electron density were augmented by extra pairs . even in isotropic situations ,
the spectrum would be modified by transmission through a pair plasma of optical depth unity .
the effect is maximal for photons of energy @xmath176 in the frame of the pair plasma : for higher energies , the klein - nishina cross - section is smaller ; for lower energies , the scatterings are almost elastic .
if , for instance , @xmath177 , the gamma - ray spectrum around photon energies @xmath178 mev would be depressed by a factor @xmath179 relative to its original value , smoothly rejoining its original value above and below that energy . for a beamed primary output , however , there would be a suppression at lower energies ( where the scattering is in the thompson regime ) because the scattered photons would be spread over a wider angle .
pair - induced processes would therefore yield evidence on the beaming properties of the bursts .
irrespective of the external density , the processes discussed here suggest that bursts and afterglows may have a more complex spectrum and time - structure than standard models suggest .
but the effects are especially interesting when the external density is high : they probe the environment of grbs , and thus can offer clues to the nature of the progenitor stars , and their location within the host galaxy .
for instance , a quasi - thermal x - ray pulse accompanying the gamma - ray emission could be indicative of an external circum - burst density of at least @xmath180 . while quasi - thermal x - ray pulses might also arise due to other reasons , e.g. from an underlying optically thick central engine , if the x - ray luminosity scales as equation ( [ eq : xrprec ] ) and is accompanied by radio or far - ir signals such as in equations ( [ eq : nusyr])([eq : nusyf ] ) , this could be indicative of birth in a dense environment from a massive progenitor .
detailed monte carlo simulations ( ramirez - ruiz , 2001 ) should provide a more detailed assessment of the self - consistent spectrum of a grb in the presence of self - induced pair formation .
dermer , c & bttcher , m. , 2000 apj 534 , l155 lightman , a.p , 1982 , apj 253 , 842 .
madau , p & thompson , c , 2000 apj , 534 , 239 madau , p , blandford , r & rees , m.j . , 2000 , apj , 541 , 712 .
ramirez - ruiz , e , , 2001 , in preparation thompson , c & madau , p , 2000 apj , 538 , 105 | gamma - ray burst sources with a high luminosity can produce electron - positron pair cascades in their environment as a result of back - scattering of a seed fraction of their original hard spectrum .
the resulting spectral modifications offer the possibility of diagnosing not only the compactness parameter of the @xmath0-ray emitting region but also the baryonic density of the environment external to the burst .
2si 50l_w50 51e_w51 1t_w1 3t_v-3 @xmath1institute of astronomy , university of cambridge , madingley road , cambridge cb3 0 ha , u.k
. + @xmath2dpt . of astronomy & astrophysics , pennsylvania state university , university park ,
pa 16803 + @xmath3institute for theoretical physics , university of california , santa barbara , ca 93106 - 4030 |
since the discovery of 51 peg b ( mayor & queloz 1995 ) , the number of known planets has gone up dramatically and today more than 100 such companions are known .
most of these systems have been discovered using high resolution echelle spectrographs .
these instruments are routinely achieving a radial velocity ( rv ) precision of @xmath2 = 5 - 15 m / s and even as low as 3 m / s in the best cases ( butler et al . 1996 ; baranne et al . 1996 ; vogt et al .
the echelle spectrographs themselves are large and expensive , with a complicated point spread function which has to be modeled appropriately . the high resolution along with the large wavelength coverage is paid for by the huge costs , the complexity of the instrument and the low throughput of around 1 - 4% ( vogt et al .
2000 ; dodorico et al .
the observational challenge now lies in detecting as many planetary systems as possible to understand the statistical distribution of planetary masses and distances from the host star . detecting a large sample of planets
will also help in understanding the physical processes underlying planetary formation .
a larger sample of stars needs to be surveyed with velocity precision less than 10 m / s to make this possible
. such a task will need large amounts of time on telescopes with current echelles and this is a significant challenge .
the next generation of rv survey instruments must have higher efficiencies , should be able to do multi - object rv surveys and should extend the wavelength range to near ir and near uv .
the development of our instrument , a multiple object dispersed fixed - delay interferometer , a new generation rv instrument , has been driven in part by these needs .
the use of a fixed - delay interferometer for doppler rv measurements is completely different from the current echelle approach . instead of measuring the absorption line centroid shifts in the echelle approach
, the rv is measured through monitoring interference fringe shifts ( ge 2002 ) . the original idea for using a fixed - delay interferometer for high precision doppler rv measurements
was proposed by two groups ( gorskii & lebedev 1977 ; beckers & brown 1978 ) .
this interferometer with a narrow bandpass has been successfully used for very high doppler precision measurements of the sun ( @xmath3 3 m / s , kozhevatov et al .
1995 , 1996 ; sub m / s precision for the gong measurements , harvey 2002 private communication ) .
the concept of combining of a fixed - delay interferometer with a moderate resolution spectrometer , or a post - disperser , for broad band operations for high precision stellar doppler measurements was proposed by dave erskine at llnl in 1997 . the initial lab experiments and telescope observing with a prototype demonstrated its feasibility ( erskine & ge 2000 ; ge et al .
a theory for this new instrument concept was developed by jian ge ( ge 2002 ) . in this interferometer approach ,
the instrument response , determined by the two beam interference , is a simple and well - defined sinusoidal function . for comparison ,
the echelle response is considerably more complex due to the interference among thousands of divided beams from grating grooves . in a fixed - delay interferometer ( fdi ) , a fixed optical delay , @xmath4 , is applied to one of the beams .
therefore , the interference happens at very high interference order , @xmath5 , determined by @xmath6 , where _ @xmath7 _ is the operating wavelength .
the doppler rv motion will shift the fringes of stellar absorption lines to neighboring orders .
the corresponding doppler velocity shift is @xmath8 where _
@xmath9 _ is the measured phase shift of a fringe .
if the absorption line density is constant over the observed band , then the total observed doppler error is @xmath10 where @xmath11 is the total number of absorption lines covered by the array , @xmath12 is the observed absorption line depth , @xmath13__= @xmath7__@xmath14__/@xmath15__@xmath16 is the coherence length of the interferometer beam for a bandwidth of _ _
@xmath15__@xmath16 , the intrinsic width of typical stellar absorption lines , @xmath17 is the observed flux within each observed fringe ( or absorption line ) , and @xmath18 is the intrinsic doppler precision at infinite spectral resolution of a post - disperser ( see ge 2002 for details ) .
this indicates that _ the doppler sensitivity of the fdi is independent of the spectral resolving power of the post - disperser _ , contrary to echelle spectroscopy . in the echelle ,
the total observed doppler error is @xmath19 where the observed line fwhm is @xmath20 and _ _ @xmath15__@xmath21 is the fwhm of the echelle response , @xmath22 is the intrinsic doppler precision at infinite spectral resolution of an echelle . _
the doppler error in the echelle approach strongly depends on the echelle resolving power and stellar intrinsic line width _
( bouchy et al . 2001 ; ge 2002 ) .
for a solar type star with absorption lines of fwhm @xmath3 5 km / s ( dravins 1987 ) , at moderate resolution ( such as _ _
@xmath23__@xmath24__@xmath3 10@xmath15__@xmath16 , or @xmath25 6000 )
, the echelle approach has @xmath3 10 times higher doppler error than the interferometer approach . at high resolution , such as @xmath26 60,000 , being used for planet detection ( e.g. , vogt et al .
1994 ; dodorico et al .
2000 ) , the doppler sensitivity for a solar type star is still @xmath3 1.4 times worse than the interferometer .
the independence of doppler sensitivity from the post - disperser resolving power in the interferometer approach opens up new possibilities for rv studies .
the use of low resolution but high efficiency post - dispersers can significantly boost the overall detection efficiency , dramatically reduce the instrument size and cost and allow single dispersion order operations for multiple object observations .
full sky coverage for an rv survey for planets becomes possible with wide field telescopes .
_ multiple object capability is one of the most significant advantages for this interferometer approach . _
the simple and stable response function in the interferometer approach leads to potential low systematic errors , which may allow this approach to reach sub m / s doppler precision .
another exciting possibility with this interferometer technique is to extend rv surveys to wavelengths other than the visible , previously not covered by echelle surveys . since the interferometer response is simple and stable
, there is no need to calibrate the instrument response in contrast to the echelle ; only wavelength calibration is required .
hence , reference sources with a lower line density than the iodine , which is popularly used in the echelle , can be used .
therefore , this instrument can be easily adapted to other wavelengths , in which more photon flux and stellar absorption lines are available for precision doppler rv measurements .
for instance , late m , l and t dwarfs have peak fluxes in the near - ir .
integration time can be significantly reduced if the ir spectra can be monitored .
b and a main sequence stars and white dwarfs have very broad intrinsic absorption lines dominated by the balmer series .
the intrinsic doppler error for each balmer line is about the same as that for late type stars .
since there are only a few dozen broad lines that can be used for early type stars while @xmath3 1000 lines can be used for late types , the overall observed doppler error is about @xmath3 10 times higher for the early types than late types for the same s / n data . since early type stars are usually much brighter in the near uv than late types at the same astronomical distance ( e.g. an a star is about 100 times brighter than a g type star in the visible ) , it is possible to achieve @xmath3 a few m / s doppler precision by increasing s /
n by a factor of @xmath3 10 .
in 2000 - 2001 , we developed a prototype dispersed fixed - delay interferometer , called exoplanet tracker ( et ) , similar to an earlier version built by jian ge and his collaborators at llnl ( ge et al .
the instrument consists of a michelson interferometer with 7 mm fixed optical path difference , a czerny - turner type spectrometer with a 100 mm diameter collimator beam and a 1k@xmath271k photonics ccd camera with 24 @xmath28 m pixel size .
the first light stellar observations were conducted at the hobby - eberly 9 m telescope ( het ) in october 2001 ( ge et al .
the spectral resolution is @xmath29 = 6700 .
the wavelength coverage is 140 with the ccd .
the measured doppler precision is approaching a photon - noise limit as demonstrated by a direct comparison between predicted theoretical rv errors and measured values shown in figure 1(left ) .
this figure also indicates that the iodine calibration contributes major errors in the measurements due to its very low fringe visibility compared to that for the star .
et was further tested at the palomar 5-m telescope in dec .
our results demonstrate that multiple object observing is feasible with the interferometer .
figure 1(right ) shows two adjacent fringe spectra obtained simultaneously with et .
one fringe data is from aldebaran .
the other is from a thar lamp .
the two fringe spectra occupy only @xmath0 1/8 ccd detector area .
therefore , we can simultaneously cover @xmath3 15 fringe data from 15 stars if multiple stellar beams are available .
a modified version of et was used for an engineering run at the kpno 2.1-m telescope in august 2002 before we install a permanent one in 2003 for a long - term survey .
the old @xmath30/10 spectrograph was replaced by an @xmath30/7.5 spectrograph .
the kpno 1k@xmath273k back - illuminated ccd with 15 @xmath28 m pixels was used instead of the old 1k@xmath271k ccd .
the wavelength coverage has been increased to 270 due to the faster instrument focal ratio and larger detector array .
an @xmath30/8 telescope beam is fed into a 200 @xmath28 m fiber , which matches a 2.5 arcsec stellar image . due to the focal ratio degradation , the output focal ratio of the fiber is @xmath30/6 , which is converted to @xmath30/7.5 to feed the spectrograph .
the spectrograph entrance slit width was dialed to about 180 @xmath28 m , which causes about 30% photon loss at the slit .
the fwhm of each absorption line is sampled by 12 pixels . under 1.5 arcsec seeing conditions , the total instrument throughput including the sky , telescope transmission , fiber loss , instrument and iodine cell transmission and detector quantum efficiency is 3.4% .
this was achieved by feeding only one of the two interferometer outputs into the spectrograph .
this allows us to routinely observe stars as faint as v = 7.6 during the whole run .
although the seeing conditions were never better than 1.5 arcsec , we still were able to continually monitor 6 stars ( arcturus , @xmath1 cas , @xmath31 and , 51 peg , 31 aql , and hd 209458 ) over 8 nights when sky was relatively clear .
part of the data have been reduced .
figure 2(left ) shows a velocity curve for @xmath1 cas within an hour .
the measured rv values well match the predicted ones caused by the earth s motion .
the rms residual from this measurement indicates a doppler precision of 2.9 m / s in a short period .
measurements from several other nights indicate we have reached a doppler sensitivity of @xmath0 3 - 8 m / s for @xmath1 cas .
figure 2(right ) shows a rv curve from 51 peg after earth s motion is subtracted , superimposed with a predicted curve from previous echelle measurements .
our result is consistent with previous measurements , demonstrating that et is capable of detecting extrasolar planets .
large long term rv measurement errors are mainly caused by data reduction and also instrument calibration .
a better version of data reduction software package is being developed .
better long term rv precision is expected .
observations with et have demonstrated that the dispersed fixed - delay interferometer is suitable for multiple object observing and can provide high throughput and high doppler precision .
multi - object and high throughput doppler surveys are important since they can significantly speed up the detection of extrasolar planetary systems and also reduce the cost .
currently , only about 4000 solar type stars are being searched by half a dozen telescopes using the echelles . based on current planet detection rate of 10% of solar type stars , @xmath3 400
planets will be discovered in the next @xmath3 10 years .
these surveys are time - consuming and very costly .
this is because current echelles can only measure a single object per exposure , and because only relatively bright stars can be observed since the instrument has low detection efficiency , whereas high photon flux is required for precision measurements .
therefore , a multiple object rv survey can significantly improve the current situation .
considering there are not many bright stars in the sky for the measurements , the multiple object survey must include relatively faint stars within the telescope field of view ( fov ) . based on our estimation from previous star count surveys in the visible ( bahcall & soniera 1980 ) and a dust extinction map ( schlegel et al . 1998 ) , we find there are about a half million stars from a m types brighter than v = 10 , and about 4 millions of stars are brighter than v = 12 . on average , about 100 stars with v magnitude brighter than 12 mag .
are within a 1 deg field - of - the - view .
therefore , in order to conduct an efficient all sky survey for extrasolar planets with a multiple object dispersed fixed - delay interferometer , we need to reach about v = 12 with a wide field telescope .
this all sky survey is possible with modern wide field telescopes such as the sloan 2.5-m and the wiyn 3.5 m telescopes .
the sloan has a 7 degree fov and the wiyn has a 1 degree fov . since both of the telescopes are moderate size , high throughput becomes critical to reach high doppler precision for faint stars .
high throughput can be achieved with an optimally designed instrument .
the prototype at the kpno 2.1-m has already demonstrated good throughput of 3.4% under 1.5 arcsec seeing . in the future ,
a factor of 4 times improvement should be achievable , including 2 times in the interferometer transmission by feeding both interferometer outputs to the spectrograph , 1.4 times in the tendu match through updating current spectrograph with a faster spectrograph ( e.g. , @xmath30/2 instead of @xmath30/7.5 ) , and 1.5 times in the spectrograph transmission by using a higher efficiency grating ( such as a vph grating with @xmath3 85% efficiency instead of current reflection grating with @xmath3 55% efficiency ) .
a total estimated detection efficiency of @xmath315% can be expected . with this detection efficiency ,
a doppler precision of @xmath3 15 m / s can be reached for a v = 12 star within an hour integration and @xmath3 1000 wavelength coverage .
therefore , most of the stars brighter than v = 12 can be surveyed with high doppler precision for planet candidates at 2 m class wide field telescopes with multi - object rv instrument .
the candidate systems can be further studied with a single object high throughput dispersed fixed - delay interferometer with higher precision at larger aperture telescopes to look for additional planet members . in the future , in order to take full advantage of the potential of the multi - object dispersed fixed - delay interferometer for all sky rv surveys , we need to simultaneously feed multi - object fiber beams into three instruments which have best sensitivity in the near - ir , visible and near - uv , respectively .
this design allows optimal match of the interferometer sensitivity with the peak of the spectral flux and lines .
for instance , a late m type star has about 10 times more flux in the near - ir than in the visible and has many molecular and atomic lines for precision doppler measurements ( kirkpatrick et al .
1993 ) . because the line width is very different from early type to very late types
, the optical delay will be changed accordingly to minimize doppler errors . to fully achieve high doppler precision with this instrument ,
a large waveband is very important since this will allow the capture of more photons from stars for the measurements . in the prototype ,
the wavelength coverage is about 200 . by simply increasing wavelength coverage to @xmath0 1000 , more than a factor of 2 times better doppler precision can result .
doppler precision can be possibly further reduced by using reference sources with higher visibility than the iodine absorption .
our measured fringe visibility for typical thorium lines in our palomar data is about 50% , resulting in intrinsic doppler error well below 1 m / s .
current rv measurements using thar calibration in dr .
mayor s group have already achieved @xmath3 5 m / s precision and with further improvement such as double fiber mode scrambling and vacuum operation , they believe that they can reach @xmath3 1 m / s in the new harps high resolution echelle spectrometer ( queloz 2002 private communications ; pepe et al . 2000 ) .
we believe we may be able to achieve similar calibration precision by using similar procedures . if this technique works , the observing efficiency can be tripled over that of our current iodine - based technique since there will be no photon loss due to iodine absorption ( typical loss by iodine absorption @xmath3 30 - 40% ) and also there is no need to create a separate stellar template for each observation
. this will be a big plus for the survey . in summary , to have a very successful all sky survey requires multi - object observation capability , high doppler precision and high throughput . a wide field telescope with at least three multi - object dispersed fixed - delay interferometers optimized for near - uv , visible and near - ir wavebands can provide the best sensitivity for detecting thousands of planets around stars from very early types to very late types in the near future
the authors are grateful to larry ramsey , don schneider , eric feigelson , steinn sigurdsson , tom soltysinski , ron reynolds , fred roesler , didier queloz , harvey moseley , bruce woodgate , roger angel , mike shao , chas beichman , bill cochran , wes traub , ed jenkins and jim gunn for useful discussions .
we acknowledge support from the het staff , the palomar observatory staff and kpno staff .
this work is supported by nasa jpl and penn state eberly college of science .
a portion 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 . | characterization of extrasolar planetary systems requires a radial velocity ( rv ) survey for planets around hundreds of thousands of nearby stars of all spectral types over next ten years .
this survey will be extremely difficult to conduct using current high resolution echelle spectrometer due to its single object observing mode and low instrument throughput .
here we propose to use a high throughput multi - object dispersed fixed - delay interferometer for the survey .
this instrument , a combination of a fixed - delay interferometer with a moderate resolution spectrometer , is completely different from current echelle spectrometers .
doppler rv is measured through monitoring interference fringe shifts of stellar absorption lines over a broad band .
coupling this multi - object instrument with a wide field telescope ( a few degree , such as sloan and wiyn ) and uv , visible and near - ir detectors will allow to simultaneously obtain hundreds of stellar fringing spectra for searching for planets .
the rv survey speed can be increased by more than 2 orders of magnitude over that for the echelles .
a prototype dispersed fixed - delay interferometer has been observed at the hobby - eberly 9 m and palomar 5 m telescopes in 2001 and demonstrated photo noise limited doppler precision with aldebaran .
our recent observations at the kpno 2.1 m telescope in 2002 demonstrate a short term doppler precision of @xmath0 3 m / s with @xmath1 cas ( v = 3.5 ) , a rv stable star and also obtained a rv curve for 51 peg .
( v = 5.5 ) , confirming previous planet detection with an independent rv technique .
the total measured detection efficiency including the sky , telescope and fiber transmission losses , the instrument and iodine transmission losses and detector quantum efficiency is 3.4% under 1.5 arcsec seeing conditions , which is comparable to all of the echelle spectrometers for planet detection .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
the starting point for the theory proposed here is simple but somewhat abstract . instead of the five equations and five unknowns that the navier - stokes - fourier formulation employs to describe a single - component fluid in three spatial dimensions , this formulation requires six ( although the number of equations / unknowns drops back down to five for the linearized problem ) .
the extra equation governs a new quantity , which i call the mechanical mass density and label as @xmath0 .
it is a book - keeping variable to be used when defining inertial quantities such as linear and angular momentum and kinetic energy .
the mechanical mass density is assumed to satisfy a continuity equation with nothing but convection to move it , while the ( actual ) mass density @xmath1 is governed by a balance law that includes a mass diffusion term in addition to a convection term . behind the equations of motion that follow ,
a philosophy has been adopted to keep thermodynamics and mechanics separate in some key respects . the non - convective mass flux , @xmath2 , is treated as one would the non - convective heat or internal energy flux , not , in other words , as a momentum involving a diffusion velocity .
it is assumed to be dissipative and , consequently , its form is chosen based on non - equilibrium thermodynamic principles .
on the other hand , the mechanical mass density may be viewed as the background mass density the material would have if it were somehow in a state of thermodynamic equilibrium with only the velocity field directly affecting it through the laws of continuum mechanics .
the balance laws stemming from these ideas are proposed in [ secbl ] . in them , there appears a local continuum mechanical velocity , @xmath3 , which is used in the convection terms and in the definitions of the momentum and kinetic energy densities .
also , in these inertial density definitions , the mechanical mass density is employed instead of the actual mass density . finally , when postulating a form for the total pressure tensor appearing in the momentum equation , there is no allowance for mass diffusion effects , nor are mass diffusion and external body forces allowed to affect one another directly .
the above statements warrant clarification : i am not claiming there to be an absence of inertial effects arising from the mass diffusion . after all , one may define momentum and kinetic energy densities involving the actual mass density or the total mass flux and then postulate balance laws for the time rate of change of those quantities .
i prefer not to work with these , however , as their balance laws are complicated by cross - effects . within my formulation ,
the velocity @xmath3 is not responsible for the overall mass flow . instead
, it is seen to govern , through convection , only the mechanical part of the flow , e.g. that which is caused by pressure disturbances , moving boundaries , gravity , etc . , whereas the diffusional part of the mass flow is handled separately in accordance with non - equilibrium thermodynamics . with this in mind ,
the balance laws given in [ secbl ] for the linear momentum , angular momentum , and kinetic energy densities are understood to be laws for the _ mechanical part _ of the linear momentum , angular momentum , and kinetic energy densities .
there are some interesting benefits to this approach .
( 1 ) it circumvents the angular momentum conservation issues and other concerns raised in ttinger et al .
@xcite , ( 2 ) because the diffusional part of the mass flux can be made to cancel with the convective part , it allows us to model an impermeable boundary with a non - zero normal velocity at that boundary see [ soundbc]as arising in the case of non - infinite acoustic impedance at a sound barrier
see morse and ingard @xcite , and ( 3 ) after appropriate constitutive equations and transport coefficients are chosen , it leads to the case of pure diffusion via fick s law when mass density gradients are present yet conditions have been chosen to eliminate all mechanical forces see [ diffusion ] . in [ secce ] and appendix [ applc ] ,
the principles of irreversible thermodynamics are used to select general forms for the dissipative fluxes appearing in the balance laws .
the navier - stokes - fourier formulation possesses two dissipative fluxes , the heat flux and the viscous pressure tensor , whereas mine at the outset possesses four : the aforementioned two , plus a non - convective mass flux and an energy flux due to chemical work .
expressions for the chemical energy and heat fluxes are chosen in strict analogy to the first law of equilibrium thermodynamics . with standard techniques ,
my formulation produces linear constitutive laws that are expanded in terms of the usual affinities involving gradients of the velocity and temperature , as in the case of the navier - stokes - fourier formulation .
however , there arise additional terms involving gradients of the chemical potential .
since the non - convective internal energy flux may be expressed as the sum of those due to heat and chemical work , my formulation is found to have three independent constitutive laws and six transport coefficients ( two for each law ) , five of which are independent if onsager reciprocity is enforced .
it may be difficult , if not impossible , to devise experiments for measuring each of my general transport coefficients individually .
therefore , the proposed framework would not be very convenient or useful without a transport model to guide specific choices for the formulas of these transport coefficients .
many thermodynamics textbooks , e.g. reif ( * ? ? ?
* ch.12 ) , present a simple particle diffusion model as an introduction to non - equilibrium transport processes , the general idea being that the particles of a gas diffuse via random collisions with one another , and they carry with them extensive quantities , such as momentum and energy , which are exchanged during the collisions .
these types of models , constructed for a classical monatomic ideal gas , are found to be qualitatively useful but quantitatively inaccurate , and thus they are immediately discarded for more complicated procedures based on the boltzmann equation . however , within the setting of my general formulation , it is of interest to revisit these ideas of particle diffusion .
i claim that my construction makes such an elementary picture a viable way of describing non - equilibrium transport in fluids .
moreover , the ideas behind this model are not constrained to the very restrictive case of a classical monatomic ideal gas . in [ sectc ]
, a transport model is proposed in which diffusing groups of particles have random encounters with other groups , whereby they exchange not only momentum and energy , but also mass .
this gives rise to formulas for all six transport coefficients involving just one diffusion parameter , @xmath4 .
, there is a little bit of flexibility regarding this issue as discussed in appendix [ tmd ] . ] furthermore , onsager reciprocity and the second law of thermodynamics are both automatically satisfied , the latter requiring only that the fluid have intrinsic equilibrium thermodynamic stability and @xmath5 .
the heat flux law arising from such a model turns out to be different from fourier s , and the diffusion transport assumption results in a new form for the viscous pressure , as well . of course
, it is important to be able relate @xmath4 to the typical transport parameters , the bulk and shear viscosities and heat conductivity , in problems for which the navier - stokes - fourier formulation is known to give the right answer .
this is easily done by using sound attenuation data , for example , as shown in [ sound ] .
an equally important issue is that there must be a way of experimentally distinguishing between the two theories for a problem where the navier - stokes - fourier equations can be shown to fail . light scattering , addressed in [ ls ] , may prove to be such an experiment , and it is currently being conducted . when considering the dimensionless knudsen number , @xmath6 , defined as
the ratio of the fluid s mean free path length to some characteristic length scale in a particular problem , it should be emphasized that , like navier - stokes - fourier , mine is a formulation intended to be used in the hydrodynamic , or small @xmath6 , regime .
i do not make any general claims that my continuum formulation works well into the more rarefied gas regime , although experimental data for sound propagation in the noble gases shows that mine performs significantly better than navier - stokes - fourier there see figure [ fig1 ] .
the previously mentioned light scattering experiment is being designed to study gases fully in the hydrodynamic regime which , up until now , has either been ignored or not adequately resolved in all other such experiments .
sound propagation , light scattering , and several other problems are explored in [ secprobs ] . as a general rule ,
one finds that the formulas derived from my equations tend to be , in many ways , much simpler and easier to interpret than those obtained from the navier - stokes - fourier theory .
in fact for some problems , when only approximate answers may be computed with the navier - stokes - fourier equations , exact solutions result from my theory .
another convenience is that , as demonstrated in appendix [ difp ] , it is quite easy to obtain values for the diffusion parameter @xmath4 , especially when compared to the task of measuring the three transport parameters required for the navier - stokes - fourier formulation .
despite these appealing features , there are two aspects of my formulation that may appear troubling .
( 1 ) the mechanical mass density is not a directly measurable quantity , and ( 2 ) my formulation is incompatible with the kinetic theory of gases . regarding the first item ,
it should be observed that there are many instances of well - accepted macroscopic theories that involve computationally useful , but not directly measurable , quantities .
the admitted difference here is that one is accustomed to conceiving of the mass density as an absolute concept with no ambiguity built into its definition and , therefore , seeming to present no need for alternate definitions .
it is typical for more licence to be granted when defining new velocities and energies , for example .
although the mechanical mass density is an abstract concept , my theory uses it not in place of but rather in conjunction with the actual mass density , which is a quantity that is measurable and tied , in an averaged sense , to the masses of the molecules comprising the fluid .
the second of the concerns mentioned above carries more significance .
the incompatibility is evident from the fact that the navier - stokes - fourier equations may be derived from the boltzmann equation via the chapman - enskog procedure for the special case of a classical monatomic ideal gas ( see huang ( * ? ? ?
* ch.6 ) , for example ) . with this procedure ,
the navier - stokes - fourier formulation is the first - order approximation in the parameter @xmath6 .
it must be emphasized that the difference between navier - stokes - fourier and my theory is _ not _ a higher - order @xmath6 effect .
therefore , mine may not be developed from the burnett equations or grad s moment method , for example .
the connection between navier - stokes - fourier and kinetic gas theory is generally interpreted to mean that both are correct , but perhaps both are missing the same concept .
i realize that without proof , this seems like a heretical claim .
however , if the results of the aforementioned light scattering experiment support my theory , then this would necessitate a reconception of kinetic gas theory in the hydrodynamic regime .
i have not been alone in my efforts to include an extra diffusive mechanism as part of the fluid equations of motion . for early examples ,
see slezkin @xcite , @xcite and vallander @xcite , and for later examples , see klimontovich @xcite , @xcite , brenner @xcite , @xcite , @xcite , @xcite , and dadzie et al .
my formulation proves to be significantly different from each of these treatments .
to facilitate the subsequent development , let us first mention a few notational items .
the number of lines under a symbol indicates its tensor order : a scalar ( zeroth - order tensor ) has no underlines ( e.g. temperature , @xmath7 ) , a first - order tensor has one underline ( e.g. velocity , @xmath3 ) , a second - order tensor has two underlines ( e.g. pressure tensor , @xmath8 ) , etc .
a tensor of arbitrary order is indicated in bold .
the symbol @xmath9 is used to represent the second - order identity tensor and @xmath10 and @xmath11 to denote the zero vector and zero second - order tensor , respectively .
all of the tensor operators used in this paper are defined in appendix [ appto ] .
extensive quantities , i.e. ones that are additive over composite subsystems , are denoted using capital letters .
below is a list of the extensive quantities considered here.@xmath12{lll}\overline{m}=\text{mechanical mass } & \underline{l}=\text{angular momentum } & w=\text{potential energy}\\ m=\text{mass } & e=\text{total energy } & s=\text{entropy}\\ v=\text{volume } & u=\text{internal energy } & \\ \underline{p}=\text{linear momentum } & k=\text{kinetic energy } & \end{array}\ ] ] the densities ( amounts per volume ) corresponding to the extensive quantities above are denoted using lower - case letters .
for example , @xmath13 is the angular momentum density , @xmath0 is the mechanical mass density , etc . for any extensive quantity , @xmath14 , its amount per unit mechanical mass is denoted as @xmath15 .
for example , @xmath16 is the internal energy per mechanical mass .
note that@xmath17 and@xmath18 the intensive quantities considered in this paper are@xmath12{l}t=\text{absolute temperature}\\ p=\text{thermodynamic pressure}\\ \mu=\text{chemical potential ( per mass),}\end{array}\ ] ] and the remaining symbols to be defined are@xmath12{l}\underline{x}=\text{position}\\ t=\text{time}\\ \underline{v}=\text{continuum mechanical velocity}\end{array}\ ] ] and a few others that are defined as they arise in the text and listed in appendix [ symb ] .
the general balance law for an extensive quantity , @xmath14 , is given by@xmath19 where @xmath20 is the local @xmath14-density which is assumed to depend on @xmath21 and @xmath22 , @xmath23 is the total @xmath14-flux , and @xmath24 denotes the volumetric @xmath14-production / destruction rate .
let us decompose the total flux into its non - convective and convective parts:@xmath25 in my proposed continuum formulation for a single - component fluid , i begin by assuming the local balance laws for the mechanical mass , mass , momentum , total energy , and entropy to be given respectively by@xmath26@xmath27@xmath28@xmath29 and@xmath30 note that equation ( [ bl1 ] ) is a continuity equation for the mechanical mass density and that equation ( [ bl2 ] ) governing the actual mass density contains an additional non - convective mass flux . in momentum equation ( [ bl3 ] ) , the momentum density is assumed to be@xmath31
also , @xmath8 denotes the total pressure tensor , i.e. the non - convective momentum flux , @xmath32 , and @xmath33 is defined to be the external body force per mechanical mass , assumed conservative , which in this setting means to satisfy@xmath34 with@xmath35 in the total energy equation ( [ bl4 ] ) , there appear , as usual , the heat flux , @xmath36 , and the energy flux arising from mechanical work , @xmath37 , but i have also included an energy flux due to chemical work , @xmath38 . based on equations ( [ bl1])-([bl5 ] ) ,
there are a few others which may be derived .
defining the kinetic energy density as@xmath39 it is shown in appendix [ appco ] that equations ( [ bl1 ] ) and ( [ bl3 ] ) may be used along with a few tensor identities to compute the following kinetic energy balance law:@xmath40 or using ( [ bl6 ] ) in the last term,@xmath41 also , since ( [ nt2 ] ) implies that the potential energy density may be expressed as@xmath42 one may employ ( [ bl1 ] ) and assumption ( [ bl6.5 ] ) , to obtain the following equation for the potential energy:@xmath43 therefore , by defining the internal energy density , @xmath44 , via@xmath45 one finds that ( [ bl4 ] ) , ( [ bl8 ] ) , and ( [ bl9 ] ) , together with identity ( [ to25 ] ) , yield the internal energy balance law,@xmath46 in view of the above , the non - convective internal energy flux is given by@xmath47 next , let us define the angular momentum density to be@xmath48 and use ( [ bl3 ] ) to compute the angular momentum balance as@xmath49 } { \partial t } & = \underline{x}\times\frac{\partial\left ( \overline{m}\underline{v}\right ) } { \partial t}\nonumber\\ & = -\underline{x}\times\left [ \nabla\cdot\left ( \underline{\underline{p}}+\overline{m}\underline{v}\,\underline{v}\right ) \right ] + \underline{x}\times\left ( \overline{m}\underline{f}^{\left ( \overline{m}\right ) } \right ) .\label{bl13}\ ] ] for the angular momentum to be conserved,@xmath50 } { \partial t}=-\nabla\cdot\left [ \underline{x}\times\left ( \underline{\underline{p}}+\overline{m}\underline{v}\,\underline{v}\right ) \right ] + \underline{x}\times\left ( \overline{m}\underline{f}^{\left ( \overline{m}\right ) } \right ) \label{bl14}\ ] ] must be satisfied , and in appendix [ appco ] , this is demonstrated to be true if and only if the pressure tensor is symmetric:@xmath51 henceforth , let us assume ( [ bl15 ] ) to hold . in preparation for the next section , definition ( [ to21 ] ) of the convective derivative may be used to rewrite some of these balance laws . employing the relation ( [ nt1 ] ) and identity ( [ to25 ] ) , one may write ( [ bl1 ] ) as@xmath52 also , using ( [ nt2 ] ) , ( [ bl1 ] ) , and ( [ to25 ] )
, one may express equations ( [ bl2 ] ) , ( [ bl11 ] ) , and ( [ bl5 ] ) as@xmath53@xmath54 and@xmath55 where ( [ bl15 ] ) has been assumed in equation ( [ bl19 ] ) . in order to compare the foregoing balance laws with those typically postulated when deriving the navier - stokes - fourier formulation ,
see de groot and mazur ( * ? ? ?
* ch.ii ) , for example .
let us study phenomena close enough to equilibrium so that classical irreversible thermodynamics is applicable . below , this formalism , which is detailed in de groot and mazur ( * ? ? ?
* ch.iii and iv ) , is used to obtain linear non - convective constitutive equations for the fluxes . for a single - component fluid ,
i propose an equilibrium fundamental relation for the entropy per mechanical mass of the form,@xmath56 note the following equilibrium thermodynamic relationships,@xmath57 let us assume the local equilibrium hypothesis , which implies that if @xmath58 represents any of the thermodynamic parameters mentioned above , then its corresponding local variable can be defined as@xmath59 in particular , the fundamental equation ( [ ce1 ] ) may be written locally as@xmath60 which , upon taking its total differential , yields@xmath61 provided that the differentiation is performed in a reference frame for which @xmath62 is constant . therefore , we may take the convective derivative of the above and multiply through by @xmath0 to obtain@xmath63 substituting equations ( [ bl17])-([bl20 ] ) into the above and using identity ( [ to28 ] ) , one arrives at the following expression for the volumetric rate of entropy production:@xmath64 in the differential form of the first law of equilibrium thermodynamics , the change in energy due to heat and chemical work are represented by @xmath65 and @xmath66 , respectively . with this in mind , let us take@xmath67 and@xmath68 by employing these in equation ( [ ce9 ] ) and using ( [ bl12 ] ) and ( [ to25 ] ) , one finds@xmath69 the second law of thermodynamics requires that@xmath70 is always satisfied , and as shown in appendix [ applc ] , this motivates the following linear constitutive equations:@xmath71@xmath72 and@xmath73 in which case , ( [ ce9.3 ] ) becomes@xmath74 \left ( \nabla t\right ) \cdot\left ( \nabla\mu\right ) + \frac{d_{m}}{t}\left\vert \nabla\mu\right\vert ^{2}.\label{ce15}\ ] ] also in appendix [ applc ] , it is shown that onsager reciprocity requires the coefficients to satisfy the relation,@xmath75 assuming this , one may eliminate , say , @xmath76 in equation ( [ ce15 ] ) to find@xmath77 from the above expression , it may be shown that condition ( [ ce10 ] ) for the second law of thermodynamics is always satisfied if@xmath78
and@xmath79 let us summarize these findings so far .
my general formulation , which i refer to as the @xmath62-formulation , may be written as equations ( [ bl1])-([bl3 ] ) and ( [ bl11])/([bl12]),@xmath80 with constitutive equations ( [ ce12])-([ce14]),@xmath81 where the transport coefficients , @xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath86 , and @xmath76 are assumed to satisfy ( [ ce16 ] ) , ( [ ce19 ] ) , and ( [ ce20 ] ) . in the next section ,
i introduce a simple diffusion transport model , which yields equations for all six transport coefficients that depend on only one diffusion coefficient . the @xmath62-formulation may be shown to satisfy galilean invariance and the conservation of angular momentum .
furthermore , the constitutive laws each satisfy the principle of objectivity required for material frame indifference . that is to say , all of the fluxes in ( [ smf2 ] ) transform in the expected way under time - dependent rigid body translation and rotation . using the notation from [ secnt ]
, the navier - stokes - fourier formulation may be written as@xmath87 with constitutive equations,@xmath88 where @xmath89 is the fourier heat conductivity , @xmath90 and @xmath91 are the navier - stokes bulk and shear viscosities , respectively , and @xmath92 represents the specific body force .
it is clear that if one were to assume @xmath93 , @xmath94 , @xmath95 , and @xmath96 in the @xmath62-formulation , then @xmath1 and @xmath0 would be the same quantity , resulting in the navier - stokes - fourier formulation above .
therefore , the navier - stokes - fourier formulation may be considered as a special case of the general @xmath62-formulation .
in this section , let us make the following assumption , which i refer to as the diffusion transport assumption : the extensive quantities , mass @xmath97 , momentum @xmath98 , and internal energy @xmath99 , which undergo dissipative transport , do so objectively via diffusion .
that is to say their corresponding constitutive equations are each assumed to have the form , @xmath100 or its objective part if this quantity is not material frame indifferent . for the mass , the above assumption implies@xmath101
as detailed in callen @xcite , for example , legendre transformations may be used to recast this in the grand canonical description , whereupon , ( [ tc2 ] ) becomes@xmath102 , \label{tc3}\ ] ] with specific enthalpy @xmath103 , isothermal compressibility @xmath104 , and thermal expansion coefficient @xmath105 . comparing ( [ tc3 ] ) with ( [ ce12 ] ) ,
one obtains the following expressions for the mass transport coefficients:@xmath106 and@xmath107 similarly , for the internal energy , the diffusion transport assumption yields@xmath108 which , in the grand canonical description , becomes@xmath109{c}\left [ c_{v}+\frac{m\kappa_{t}}{t}\left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}\right ) \left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}-\mu\right ) \right ] \nabla t+\\ m\kappa_{t}\left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}\right ) \nabla\mu \end{array } \right\ } , \label{tc7}\ ] ] where @xmath110 represents the isochoric specific heat per mass . comparison of ( [ tc7 ] ) and ( [ ce14 ] ) gives the internal energy transport coefficients,@xmath111 d\label{tc8}\ ] ] and@xmath112 next , let us examine the momentum under the diffusion transport assumption . using ( [ tc1 ] ) with @xmath113 , one finds@xmath114 or employing identity ( [ to25 ] ) and definitions ( [ to14 ] ) and ( [ to14.1]),@xmath115 with the same procedure as described in jou et al .
@xcite , for example , one may demonstrate that the first term on the right - hand side of ( [ tc11 ] ) is objective . however , the second and third terms are not . requiring all constitutive equations to be material frame indifferent
, one chooses the viscous pressure @xmath116 to be only the objective part of @xmath117 , i.e.@xmath118 which , upon comparison with ( [ ce13 ] ) , yields the following expressions for the viscosity coefficients:@xmath119
and@xmath120 to summarize , the diffusion transport assumption gives rise to equations ( [ tc4 ] ) , ( [ tc5 ] ) , ( [ tc13 ] ) , ( [ tc14 ] ) , ( [ tc8 ] ) , and ( [ tc9 ] ) for the transport coefficients appearing in the @xmath62-formulation:@xmath121 d\nonumber\\ d_{u } & = \frac{m^{2}\kappa_{t}}{\mu}\left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}\right ) d.\nonumber\end{aligned}\ ] ] i refer to the above as the `` diffusion transport coefficients . ''
it is easily shown that these satisfy onsager reciprocity condition ( [ ce16 ] ) and that if @xmath5 and there is intrinsic equilibrium thermodynamic stability , i.e. @xmath122 , then conditions ( [ ce19 ] ) and ( [ ce20 ] ) for the second law of thermodynamics automatically hold .
substitution of ( [ tc15 ] ) into constitutive equations ( [ ce12])-([ce14 ] ) yields@xmath123 \label{tc16}\\ \underline{\underline{p } } & = p\underline{\underline{1}}-\overline{m}d\left ( \nabla\underline{v}\right ) ^{sym}\label{tc17}\\ \underline{q}_{u } & = -d\left\ { \begin{array } [ c]{c}\left [ mc_{v}+\frac{m^{2}\kappa_{t}}{t}\left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}\right ) \left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}-\mu\right ) \right ] \nabla t+\\ m^{2}\kappa_{t}\left ( h^{\left ( m\right ) } -\frac{t\alpha_{p}}{m\kappa_{t}}\right ) \nabla\mu \end{array } \right\ } .\label{tc18}\ ] ] of course , one may use the grand canonical formalism to transform constitutive equations ( [ tc16 ] ) and ( [ tc18 ] ) back so that they are , again , written in terms of the variables @xmath1 and @xmath44 . in doing so , we have our original postulations,@xmath124 forms ( [ tc19])-([tc21 ] ) display the fact that we have modeled the dissipative fluxes via diffusing groups of particles which transfer mass , momentum , and energy .
notice that only one transport parameter @xmath4 appears . is not the self - diffusion coefficient , although these quantities are the same order of magnitude for gases ( see appendix [ difp ] ) . ]
the most prevalent transport parameter in the navier - stokes - fourier formulation is the shear viscosity @xmath91 .
therefore , for convenience , let us define a dimensionless coefficient @xmath125 ( possibly a function of thermodynamic variables ) via@xmath126 in order to match greenspan s @xcite sound attenuation data in the hydrodynamic regime for classical monatomic ideal gases at room temperature , for example , we take @xmath127 .
( see [ sound ] . ) in appendix [ difp ] , values of @xmath125 and @xmath4 for various types of gases and liquids are computed using ideas from [ sound ] , comparisons between @xmath4 and the self - diffusion coefficient @xmath128 are made , and the temperature , pressure , and mass density dependence of @xmath4 is examined .
although it upsets the underlying simplicity of the diffusion transport assumption , one may be curious to see what happens when each of the diffusion processes involves a different coefficient .
this idea is explored in appendix [ tmd ] .
we have arrived at the following equations of motion for a single - component fluid near equilibrium and in the hydrodynamic regime : the @xmath62-formulation balance laws ( [ smf1]),@xmath129 together with the diffusion transport constitutive equations ( [ tc19])-([tc21]),@xmath130 substituting the constitutive equations into the balance laws , we have a system of four partial differential equations in the four unknowns , @xmath0 , @xmath1 , @xmath3 , and @xmath44 . in the above expression for the total pressure tensor
, i have indicated that the equilibrium thermodynamic pressure , @xmath131 , should be written in terms of the variables , @xmath44 and @xmath1 .
for example , to model a classical monatomic ideal gas , one takes@xmath132 if the fluid has intrinsic equilibrium thermodynamic stability , then it is a straight - forward matter to use legendre transformations in order to replace the @xmath44 and @xmath1-dependency in the above formulation with other thermodynamic variables .
for example , using the helmholtz free energy formalism , we can recast the above in terms of the four unknowns , @xmath0 , @xmath1 , @xmath3 , and @xmath7 ; using the gibbs free energy formalism , we can recast the above in terms of @xmath0 , @xmath131 , @xmath3 , and @xmath7 ; and , as we have already seen , using the grand canonical formalism , we can recast the above in terms of @xmath0 , @xmath133 , @xmath3 , and @xmath7 . due to the nature of certain equations of state and/or boundary conditions , it is often more convenient to work with one of these other forms .
i hope to have motivated the proposed formulation by the simple transport mechanism underlying its constitutive equations , one that is easily generalized to model more complicated problems such as multicomponent fluid mixtures ( see appendix [ appmc ] ) .
of course , its utility nonetheless hinges on the answers to two important questions : ( 1 ) how adequately does this formulation perform in problems of fluid mechanics , and ( 2 ) are its predictions in the hydrodynamic regime quantifiably different from those found with the navier - stokes - fourier formulation ?
next , let us briefly highlight some of the results that have been obtained so far .
in all of the following sections but [ shock ] , problems that may be examined with low mach number linearizations are considered .
, to be the ratio of the characteristic speed in a particular problem to the fluid s sound speed , the low mach number regime is taken to be that in which @xmath134 .
] a one - dimensional linear stability analysis is carried out in [ stability ] to show that my formulation is unconditionally stable provided that the diffusion parameter , @xmath4 , is positive . in [ diffusion ] , a mass equilibration problem with no mechanical forces is studied in order to show that my formulation reduces to one of pure diffusion governed by fick s law .
next , low amplitude sound propagation , which is the subject of [ sound ] , is demonstrated to be a way of relating my diffusion parameter , @xmath4 , to the transport parameters of the navier - stokes - fourier formulation by matching attenuations in the hydrodynamic ( small @xmath6 ) regime , thereby connecting the two theories in this sense .
obviously , therefore , sound is not a good way to distinguish between my theory and navier - stokes - fourier .
however , in [ ls ] , light scattering is shown to provide a way of testing the difference . the shifted brillouin peaks , which correspond to the sound ( phonon ) part of the spectrum , are virtually indistinguishable , as one would expect , but the central rayleigh peak predictions are significantly different . for example , in a classical monatomic ideal gas in the hydrodynamic regime , my theory predicts a rayleigh peak that is about @xmath135 taller and narrower than the navier - stokes - fourier theory ( but with the same area so that the well - verified landau - placzek ratio remains intact ) .
previous experimenters , e.g. clark @xcite and fookson et al .
@xcite , did not study gases fully in the hydrodynamic regime , @xmath136 , choosing instead to focus on more rarefied gases with knudsen numbers in the range , @xmath137 , which extends into the slip flow and transition regimes .
therefore , in order to verify my theory , it is important to conduct a high - resolution light scattering experiment for a gas in the hydrodynamic regime .
another striking difference in predictions arises from the non - linear steady - state shock wave problem considered in
[ shock ] .
it is well - known that for this problem , the navier - stokes - fourier formulation and all other accepted theories , for that matter
produces normalized density , velocity , and temperature shock wave profiles that are appreciably displaced from one another with the temperature profile in the leading edge of the shock , velocity behind it , and density trailing in the back .
my formulation , however , predicts virtually no displacement between these three profiles in the leading edge of the shock and much less pronounced displacements in the middle and trailing end of the shock .
if an experiment can be devised to measure both the structure and position of the steady shock profiles of the mass density and , say , the temperature , then this would be another decisive test between my theory and all others .
thermophoresis , or the down - temperature - gradient motion of a macroscopic particle in a resting fluid , is discussed in [ therm ] . for an idealized problem
, it is shown that , through the steady - state balancing of the mechanical and diffusive terms of the mass flow , my theory provides a mechanism for thermophoresis that is not present in the navier - stokes - fourier formulation . in the latter type
, a thermal slip boundary condition at the particle s surface is needed in order to model thermophoretic motion , whereas the former type may be used to describe this motion with a no - slip boundary condition , i.e. one in which the tangential velocity of the particle at its surface equals that of the fluid .
the concept of thermal slip is based on kinetic gas theory arguments in the slip flow regime , appropriate for @xmath138 , yet thermophoresis is observed in gases for much smaller knudsen numbers in the hydrodynamic regime and also in liquids . therefore , there are obvious advantages to being able to model this problem without the thermal slip condition .
another example in which we cancel the diffusional and convective parts of the mass flux at an impermeable boundary is considered in [ soundbc ] . here
, we discuss acoustic impedance at a sound barrier , as presented in morse and ingard @xcite .
in particular , it is demonstrated that my formulation allows an impermeable boundary with a non - zero normal velocity at that boundary as that which occurs in the case of non - infinite impedance .
there is one topic that remains to be introduced . as a step towards bridging the gap between microscopic and macroscopic scales in my theory ,
the subject of fluctuating hydrodynamics is addressed in [ flh ] , where it is shown that the thermodynamic interpretation of a continuum mechanical point ( i.e. a point occupied by matter and moving with velocity @xmath3 ) stemming from my formulation is fundamentally different from that of navier - stokes - fourier . that is , in the navier - stokes - fourier theory , a continuum mechanical point is viewed as a thermodynamic subsystem in contact with its surrounding material which acts as a heat reservoir ; whereas in my theory , the surrounding material acts not only as a heat but also a particle reservoir .
although i did not explicitly demonstrate it in this paper , the correlation formulas in [ flh ] may be used to derive the light scattering spectra which follow in [ ls ] .
let us consider the cartesian one - dimensional problem in which variation is assumed to be in the @xmath139-direction only with @xmath140 as the only non - zero component of the velocity and there are assumed to be no body forces .
if we use the helmholtz free energy formalism to recast the @xmath62-formulation ( [ smf1 ] ) with diffusion transport constitutive equations ( [ tc19])-([tc21 ] ) in terms of the variables , @xmath0 , @xmath1 , @xmath140 , and @xmath7 , and then linearize about the constant equilibrium state,@xmath141 via the expansions,@xmath142 where @xmath143 is a small dimensionless parameter proportional to the mach number , then we arrive at the following system of linear equations:@xmath144 where the subscript `` @xmath145 '' indicates that the parameter is evaluated at the constant equilibrium state ( [ c1 ] ) . note that for this linearized problem , the mechanical mass equation ( [ c6 ] ) may be uncoupled from the rest of the system , ( [ c7])-([c9 ] ) . postulating a solution,@xmath146{c}m_{1}\\ v_{x,1}\\ t_{1}\end{array }
\right ] , \ ] ] to ( [ c7])-([c9 ] ) that is proportional to@xmath147 for @xmath148 real and @xmath149 complex , one obtains the dispersion relation,@xmath150 in the above , we have employed the equilibrium thermodynamic relationships,@xmath151 where @xmath152 is the adiabatic sound speed and @xmath153 represents the isobaric to isochoric ratio of specific heats . equation ( [ c11 ] ) may be solved for @xmath149 to obtain the three exact roots,@xmath154 clearly , if@xmath155 is satisfied , then the real parts of @xmath156 , @xmath157 , and @xmath158 , are negative for all @xmath148 , resulting in the unconditional stability of linearized system ( [ c7])-([c9 ] ) . for comparison ,
carrying out a similar procedure with the navier - stokes - fourier formulation ( [ smf3])/([smf4 ] ) yields the linearization,@xmath159{c}-\frac{1}{m_{0}}\left ( \frac{1}{m_{0}\kappa_{t,0}}\frac{\partial m_{1}}{\partial x}+\frac{\alpha_{p,0}}{\kappa_{t,0}}\frac{\partial t_{1}}{\partial x}\right ) + \\
\frac{1}{m_{0}}\left ( \zeta_{ns,0}+\frac{4}{3}\eta_{ns,0}\right ) \frac{\partial^{2}v_{x,1}}{\partial x^{2}}\end{array } \right ] \label{c16.4}\\ \frac{\partial t_{1}}{\partial t } & = \frac{k_{f,0}}{m_{0}c_{v,0}}\frac{\partial^{2}t_{1}}{\partial x^{2}}-\frac{t_{0}\alpha_{p,0}}{m_{0}\kappa_{t,0}c_{v,0}}\frac{\partial v_{x,1}}{\partial x},\label{c16.5}\ ] ] with the dispersion relation,@xmath160{c}\omega^{3}+\left [ \frac{\zeta_{ns,0}}{m_{0}}+\left ( \frac{4}{3}+e\!u_{0}\right ) \frac{\eta_{ns,0}}{m_{0}}\right ] \kappa^{2}\omega^{2}+\\ \left [ c_{0}^{2}+e\!u_{0}\left ( \frac{\zeta_{ns,0}}{m_{0}}+\frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}\right ) \frac{\eta_{ns,0}}{m_{0}}\kappa^{2}\right ] \kappa^{2}\omega+\\ \frac{e\!u_{0}}{\gamma_{0}}\frac{\eta_{ns,0}}{m_{0}}c_{0}^{2}\kappa_{st}^{4}\end{array } \right ) = 0,\label{c17}\ ] ] where @xmath161 is the euken ratio defined to be@xmath162 let us also define the quantities,@xmath163 and@xmath164 \frac{\eta_{ns,0}}{m_{0}}\right\ } .\label{c18.2}\ ] ] using these , equation ( [ c17 ] ) yields the following three approximate roots,@xmath165 in the@xmath166 limit , which corresponds to the low knudsen number , or hydrodynamic , regime . for an equilibrium thermodynamically stable fluid,@xmath167
are satisfied , and therefore , if @xmath168 , @xmath169 , and @xmath170 are all chosen to be non - negative , then we arrive at the well - known result that the one - dimensional linearized navier - stokes - fourier formulation does not go unstable in the hydrodynamic regime .
if we take the limit as the transport parameters go to zero ( @xmath171 in my formulation or @xmath172 in navier - stokes - fourier ) , we are left with the euler equations of pure wave motion for which , on an infinite domain , perturbations travel at the adiabatic sound speed and never decay . in the present section ,
we study a problem on the other end of the dichotomy , one of pure diffusion . for this , it is instructive to bear the following example of thermodynamic equilibration in mind .
let us consider a cartesian one - dimensional problem on the infinite domain in which the initial conditions are given by@xmath173{c}m_{0}+\delta_{m}\text { for } x<\left\vert a\right\vert \\ m_{0}\text { for } x>\left\vert a\right\vert \end{array } \right .
\label{c23.01}\\ v_{x}\left ( x,0\right ) & = 0\text { for all } x\label{c23.02}\\ t\left ( x,0\right ) & = \left\ { \begin{array } [ c]{c}t_{0}+\delta_{t}\text { for } x<\left\vert a\right\vert \\ t_{0}\text { for } x>\left\vert a\right\vert \end{array } \right .
.\label{c23.03}\ ] ] this describes a perturbed subsystem initially shaped like an infinite slab , centered at @xmath174 with width @xmath175 , in contact with two identical infinite reservoirs on either side of it .
as time approaches infinity , one expects the subsystem to equilibrate with the reservoirs until the the whole system has uniform mass density @xmath176 and temperature @xmath177 .
let us further suppose ( 1 ) that the perturbations @xmath178 and @xmath179 are small compared their respective equilibrium values so that it is appropriate to use the linearized equations presented in [ stability ] and ( 2 ) that @xmath178 and
@xmath179 are chosen to satisfy@xmath180 such that the initial pressure is in equilibrium , i.e.@xmath181 note that conditions ( [ c23.02 ] ) and ( [ c23.05 ] ) mean that there are initially no mechanical forces acting on the system , and so it is of interest to see , in a problem such as this , the consequences of assuming a @xmath182 solution . to this end , let us examine the zero velocity solution,@xmath183 to my linearized equations ( [ c7])-([c9 ] ) . substituting ( [ c23.1 ] ) into the equations yields@xmath184@xmath185 and@xmath186 equation ( [ c23.3 ] )
is a constraint that there are no pressure gradients and it forces the mass density and temperature perturbations to be related by@xmath187 as long as there exists any point at which @xmath188and @xmath7 attain their equilibrium values , @xmath176 and @xmath177 , thus causing any additive function of @xmath22 that may appear in ( [ c23.5 ] ) to be zero .
( note that at @xmath189 , ( [ c23.5 ] ) satisfies ( [ c23.04 ] ) for the problem discussed above . ) as we can see , the two diffusion equations ( [ c23.2 ] ) and ( [ c23.4 ] ) are perfectly consistent with the constraint ( [ c23.5 ] ) . furthermore , they imply that in the absence of mechanical forces , a non - equilibrium problem , such as the one described in the previous paragraph , equilibrates by pure diffusion
. equation ( [ c23.2 ] ) is fick s law which describes diffusion driven by gradients in mass density . by comparison ,
if we substitute the zero velocity solution ( [ c23.1 ] ) into the linearized navier - stokes - fourier equations ( [ c16.3])-([c16.5 ] ) , then we find only the trivial solution,@xmath190 unless we examine the special case , @xmath191 , which yields@xmath192 and@xmath193 either way , it is clear that for the initial value problem described previously , the mass density does not equilibrate via pure diffusion when the navier - stokes - fourier formulation is used . in fact , a non - zero velocity solution is required .
next , let us employ linearized system ( [ c7])-([c9 ] ) to study one - dimensional cartesian sound propagation for low amplitude sound waves of angular frequency @xmath149 . for this problem ,
one postulates a solution proportional to@xmath194 where @xmath149 is real and @xmath148 is complex , leading to the following dispersion relation:@xmath195 equation ( [ c24 ] ) may be solved to obtain six @xmath148-roots : the propagational pair,@xmath196 the thermal pair,.]@xmath197 and an extra pair,@xmath198 the thermal roots ( [ c26 ] ) are exact solutions to ( [ c24 ] ) , but ( [ c25 ] ) and ( [ c27 ] ) represent approximate solutions in the hydrodynamic regime , requiring@xmath199 note that ( [ c25 ] ) suggests that sound attenuation experiments , such as greenspan s @xcite and @xcite , may be used to measure the diffusion parameter @xmath4 of my formulation for various fluids under a wide variety of thermodynamic conditions .
commonly , one finds @xmath200 and @xmath152 data tabulated from such experiments at given temperatures and pressures , where@xmath201 \text { and } f\equiv\frac{\omega}{2\pi}.\label{c28.5}\ ] ] in the hydrodynamic regime , equation ( [ c25 ] ) implies the formula,@xmath202 and in appendix [ difp ] , this formula is used together with sound propagation data to compute the diffusion parameter for a variety of selected gases and liquids . for the navier - stokes - fourier formulation
, one may use the foregoing procedure to obtain the dispersion relation,@xmath203{c}\left [ -ic_{0}^{2}\sigma_{0}+e\!u_{0}\left ( \frac{\zeta_{ns,0}}{m_{0}}+\frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}\right ) \frac{\eta_{ns,0}}{m_{0}}\omega\right ] \kappa^{4}+\\ \left [ -i\left ( 2\gamma_{0}+\sigma_{0}\right ) \omega - c_{0}^{2}\right ] \omega\kappa^{2}-\omega^{3}\end{array } \right\ } = 0,\label{c29}\ ] ] where @xmath204 and @xmath205 are defined in equations ( [ c18.1 ] ) and ( [ c18.2 ] ) . solving equation ( [ c29 ] ) ,
one finds four @xmath206-roots : the propagational pair,@xmath207 and the thermal pair,@xmath208 , \label{c31}\ ] ] where we have defined@xmath209 , \label{c31.1}\ ] ] and all four roots are approximations in the hydrodynamic regime,@xmath210 note that a comparison of equations ( [ c25 ] ) and ( [ c30])/([c18.2 ] ) gives a convenient way of relating the diffusion coefficient , @xmath211 , of my formulation to the bulk and shear viscosities , @xmath169 and @xmath168 , and the thermal conductivity @xmath170 appearing in the navier - stokes - fourier formulation , i.e. to match these predictions for sound attenuation , one chooses@xmath212 \frac { \eta_{ns,0}}{m_{0}}\right\ } .\label{c33}\ ] ] for example , in a classical monatomic ideal gas for which one typically chooses@xmath213 equation ( [ c33 ] ) implies@xmath214 which in view of equation ( [ tc22 ] ) , written for this linearized problem as@xmath215 yields@xmath216 next , instead of the hydrodynamic regime approximations , ( [ c25 ] ) and ( [ c30 ] ) , let us consider the exact propagational roots:@xmath217 corresponding to my formulation and@xmath218{c}c_{0}^{2}+i\omega\left [ \left ( \frac{4}{3}+e\!u_{0}\right ) \frac { \eta_{ns,0}}{m_{0}}+\frac{\zeta_{ns,0}}{m_{0}}\right ] -\\ \sqrt{\begin{array } [ c]{c}\left\ { c_{0}^{2}+i\omega\left [ \left ( \frac{4}{3}+e\!u_{0}\right ) \frac{\eta_{ns,0}}{m_{0}}+\frac{\zeta_{ns,0}}{m_{0}}\right ] \right\ } ^{2}+\\ 4e\!u_{0}\frac{\eta_{ns,0}}{m_{0}}\omega\left [ -i\frac{c_{0}^{2}}{\gamma_{0}}+\left ( \frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}+\frac{\zeta_{ns,0}}{m_{0}}\right ) \omega\right ] \end{array } } \end{array } \right ) } { 2e\!u_{0}\frac{\eta_{ns,0}}{m_{0}}\left [ -i\frac{c_{0}^{2}}{\gamma_{0}\omega}+\left ( \frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}+\frac { \zeta_{ns,0}}{m_{0}}\right ) \right ] } } , \label{c38}\ ] ] corresponding to the navier - stokes - fourier formulation . figure [ fig1 ] is a plot of the dimensionless real and imaginary parts,@xmath219 and@xmath220 of the above roots versus the dimensionless parameter,@xmath221 which is inversely proportional to the knudsen number for a gas .
the @xmath222 and @xmath223 parameters characterize the sound attenuation and the inverse sound speed , respectively .
the curves ( blue for my formulation and red for navier - stokes - fourier ) correspond to a classical monatomic ideal gas for which ( [ c34 ] ) and ( [ c35 ] ) are chosen , as well as the equilibrium thermodynamic formula for the adiabatic sound speed,@xmath224 where @xmath225 is the universal gas constant and @xmath226 represents the atomic weight of the gas .
the points are greenspan s @xcite sound data for the noble gases , helium , neon , argon , krypton , and xenon , at room temperature .
both axes are logarithmic scale , and the upper and lower curves correspond to @xmath223 and @xmath222 , respectively .
fig1c.eps from figure [ fig1 ] , it is clear that my theory s sound predictions match those of the navier - stokes - fourier formulation in the hydrodynamic ( high @xmath227 , or low @xmath228 ) regime , as intended .
as i have emphasized throughout , my formulation is a continuum theory intended to be used in the @xmath229 regime , and so i make no general claims that it works well into more rarefied higher knudsen number regimes .
however , figure [ fig1 ] provides an intriguing demonstration that for this problem , my theory does significantly better than navier - stokes - fourier there . of course , until i can provide a formal explanation , this should be viewed as strictly fortuitous .
morse and ingard ( * ? ? ?
* chapter 6 ) use the navier - stokes - fourier equations in order to study low amplitude sound waves hitting a wall . for easier comparison with their treatment and for general convenience , as well
, let us use the gibbs free energy formalism to recast my formulation ( [ smf1])/([tc19])-([tc21 ] ) with no body forces in terms of the variables , @xmath0 , @xmath131 , @xmath140 , and @xmath7 , and then linearize about the constant equilibrium state,@xmath230 via the expansions,@xmath231 with @xmath143 , as before , a small dimensionless parameter proportional to the mach number .
doing so , one obtains the following system:@xmath232 where @xmath233 is the isobaric specific heat per mass , `` @xmath145 '' subscripts are used to indicate evaluation at the equilibrium state @xmath234 , and the equilibrium thermodynamic relationships ( [ c12 ] ) and@xmath235 have been used .
again , we see that the mechanical mass equation ( [ c106 ] ) may be uncoupled from the rest of the system and postulating a solution@xmath146{c}p_{1}\\ v_{x,1}\\ t_{1}\end{array } \right]\ ] ] to the remaining system ( [ c107])-([c109 ] ) that is proportional to@xmath236 with @xmath149 real and @xmath148 complex , unsurprisingly leads to the same dispersion relation ( [ c24 ] ) and the same six roots ( [ c25])-([c27 ] ) as before .
however , when my equations are cast in the present form , one notes the following interesting fact : the temperature equation ( [ c109 ] ) may be uncoupled from the pressure and velocity equations , so that the dispersion relation arising from the reduced system , ( [ c107 ] ) and ( [ c108 ] ) , yields only the propagational roots ( [ c25 ] ) and the extra roots ( [ c27 ] ) .
the thermal roots ( [ c26 ] ) arise only when the temperature equation ( [ c109 ] ) is coupled back into the system .
this means that within my formulation , thermal modes of the solution may contribute only to the temperature variable and not the pressure or velocity variables , i.e. the mechanical variables . on the other hand
, propagational and extra modes may contribute to all three .
next , let us examine the length scales associated with each of these three types of roots .
if we compute the length scale as the distance it takes for a mode to attenuate to @xmath237 of its amplitude , then we find the propagational , thermal , and extra mode lengths to be@xmath238@xmath239 and@xmath240 respectively , where ( [ c25])-([c27 ] ) and ( [ c28.5 ] ) have been used in the above .
for example , if we consider sound waves at frequency @xmath241 mhz as in greenspan @xcite and argon gas at normal temperature and pressure ( @xmath242 k and @xmath243 pa ) with @xmath244 m@xmath245/s@xmath246(see appendix [ difp ] ) and @xmath247 m / s ( from the classical monatomic ideal gas formula , @xmath248 ) , then the above length scales are computed to be@xmath249@xmath250 and@xmath251 in this case and in general , for all ideal gases in the hydrodynamic regime
the extra mode has a length scale @xmath252 roughly equal to the mean free path length of the gas in its equilibrium state @xmath234 , regardless of sound frequency , and@xmath253 away from boundaries , the propagational modes of sound waves dominate .
however , depending on the boundary conditions , the thermal and extra modes may play an important role near walls , causing the formation of boundary layers with lengths @xmath254 and @xmath252 , respectively .
when the navier - stokes - fourier equations are recast in the gibbs free energy description and linearized as above , we arrive at the following system:@xmath255 which again yields dispersion relation ( [ c29 ] ) and the four roots ( [ c30 ] ) and ( [ c31 ] ) . unlike in my formulation , none of the above equations uncouple from one another .
therefore , in the navier - stokes - fourier formulation the two possible modes , propagational and thermal , may contribute to each of the three variables : pressure , velocity , and temperature .
using ( [ c30 ] ) and ( [ c31 ] ) , the length scales associated with the propagational and thermal modes of the navier - stokes - fourier formulation are computed to be@xmath256 and@xmath257 in view of ( [ c33 ] ) , equation ( [ c123n ] ) yields the same propagation length as ( [ c111n ] ) .
also , in the hydrodynamic regime , the thermal length given by ( [ c124n ] ) is the same order of magnitude as the thermal length ( [ c112n ] ) from my formulation .
let us assume the @xmath258-plane forms a sound barrier at @xmath174 and , as in morse and ingard , define the acoustic impedance of the wall as@xmath259 infinite impedance corresponds to a perfectly reflected sound wave in which its outgoing velocity is equal in magnitude and opposite in direction to its incoming velocity , resulting in@xmath260 for non - infinite wall impedances , however , there is a non - zero normal fluid velocity at the wall . using my formulation
, we may still consider the wall to be both stationary and impermeable by enforcing the no total mass flux condition,@xmath261 which becomes@xmath262 \right\vert _ { x=0}=0\label{c118n}\ ] ] when we assume the diffusion transport constitutive law ( [ tc19 ] ) , recast in the gibbs free energy description , and linearize .
on the other hand , if ( [ c116n ] ) is not satisfied in the navier - stokes - fourier formulation , there arises the curious situation of a non - zero normal mass flux at @xmath174 .
morse and ingard @xcite explain this phenomenon as follows : `` the acoustic pressure acts on the surface and tends to make it move , or else tends to force more fluid into the pores of the surface . ''
wall movement and/or fluid lost to pores in the wall are yet never treated explicitly .
the procedure for investigating stationary gaussian - markov processes , as described by fox and uhlenbeck @xcite , may be used to derive the formulas of fluctuating hydrodynamics for my theory .
this procedure involves linearizing the three - dimensional cartesian version of the hydrodynamic equations , interpreting the variables as fluctuating variables , and introducing zero - mean fluctuating hydrodynamic forces .
this yields the following langevin type of system:@xmath263 in the above , the mechanical mass density equation has been uncoupled , and inverted hats are used to distinguish the fluctuating variables from their average macroscopic continuum counterparts , e.g. @xmath264 , where @xmath265 denotes a conditional average for a given initial condition .
also , @xmath266 , @xmath267 , and @xmath268 represent the fluctuating mass , momentum , and internal energy fluxes , respectively , with @xmath267 assumed to be symmetric and@xmath269 following the steps outlined in fox and uhlenbeck leads to a generalized fluctuation - dissipation theorem for stationary gaussian - markov processes which , when expressed for the problem at hand , yields the following correlation formulas for the cartesian components of the fluctuating fluxes:@xmath270@xmath271@xmath272@xmath273{l}2k_{b}m_{0}t_{0}d_{0}\delta\left ( \underline{x}-\underline{x}^{\prime } \right ) \delta\left ( t - t^{\prime}\right ) \text { if } i = j = k = l\\ k_{b}m_{0}t_{0}d_{0}\delta\left ( \underline{x}-\underline{x}^{\prime}\right ) \delta\left ( t - t^{\prime}\right ) \text { if } i = k\text { , } j = l\text { , } i\neq j\\ k_{b}m_{0}t_{0}d_{0}\delta\left ( \underline{x}-\underline{x}^{\prime}\right ) \delta\left ( t - t^{\prime}\right ) \text { if } i = l\text { , } j = k\text { , } i\neq j\\ 0\text { otherwise}\end{array } \right . ,
\nonumber\end{gathered}\ ] ] and@xmath274 in the above formulas , the indices , @xmath275 , @xmath276 , and @xmath277 , may equal @xmath278 , @xmath279 , or @xmath280 ( the three cartesian directions ) , @xmath281 is used to represent dirac delta distributions , @xmath282 is the kronecker delta , @xmath283 denotes the boltzmann constant , and there appear the following equilibrium thermodynamic fluctuations:@xmath284 \label{c53}\\ \left ( \delta m\delta u\right ) & = \frac{k_{b}m_{0}^{2}t_{0}\kappa_{t,0}}{v}\left ( h_{0}^{\left ( m\right ) } -\frac{t_{0}\alpha_{p,0}}{m_{0}\kappa_{t,0}}\right ) , \nonumber\end{aligned}\ ] ] which apply to a thermodynamic system of fixed volume @xmath285 in contact with a heat / particle reservoir .
note the similarity in form of equations ( [ c45])-([c50 ] ) , involving the above fluctuations . for comparison , the langevin system corresponding to the navier - stokes - fourier formulation is@xmath286{c}-\nabla\left ( \varepsilon\check{p}_{1}\right ) + \left ( \zeta_{ns,0}+\frac{\eta_{ns,0}}{m_{0}}\right ) \nabla\left [ \nabla\cdot\left ( \varepsilon\underline{\check{v}}_{1}\right ) \right ] + \\
\eta_{ns,0}\nabla^{2}\left ( \varepsilon\underline{\check{v}}_{1}\right ) + \nabla\cdot\underline{\underline{r}}_{\underline{p}}\end{array } \right ] \label{c54}\\ \frac{\partial\left ( \varepsilon\check{u}_{1}\right ) } { \partial t } & = -m_{0}h_{0}^{\left ( m\right ) } \nabla\cdot\left ( \varepsilon \underline{\check{v}}_{1}\right ) + k_{f,0}\nabla^{2}\left ( \varepsilon \check{t}_{1}\right ) + \nabla\cdot\underline{r}_{u}\nonumber\end{aligned}\ ] ] from which one may derive fox and uhlenbeck s correlations,@xmath287@xmath288@xmath289@xmath290{l}2k_{b}t_{0}\left ( \frac{4\eta_{ns,0}}{3}+\zeta_{ns,0}\right ) \delta\left ( \underline{x}-\underline{x}^{\prime}\right ) \delta\left ( t - t^{\prime } \right ) \text { if } i = j = k = l\\ 2k_{b}t_{0}\eta_{ns,0}\delta\left ( \underline{x}-\underline{x}^{\prime } \right ) \delta\left ( t - t^{\prime}\right ) \text { if } i = k\text { , } j = l\text { , } i\neq j\\ 2k_{b}t_{0}\eta_{ns,0}\delta\left ( \underline{x}-\underline{x}^{\prime } \right ) \delta\left ( t - t^{\prime}\right ) \text { if } i = l\text { , } j = k\text { , } i\neq j\\ 2k_{b}t_{0}\left ( \zeta_{ns,0}-\frac{2\eta_{ns,0}}{3}\right ) \delta\left ( \underline{x}-\underline{x}^{\prime}\right ) \delta\left ( t - t^{\prime } \right ) \text { if } i = j\text { , } k = l\text { , } i\neq k\\ 0\text { otherwise}\end{array } \right . , \nonumber\end{gathered}\ ] ]
and@xmath291 therefore , in the navier - stokes - fourier formulation,@xmath292 whereas my theory predicts a non - zero fluctuating mass flux .
this exercise illuminates certain ideas about the thermodynamic nature of a continuum mechanical point , i.e. a point occupied by matter whose motion is tracked by its local velocity , @xmath3 , and whose infinitesimal volume is determined by its continuum mechanical deformation . as per the local equilibrium hypothesis ,
near - equilibrium continuum theories are constructed under the assumption that any given continuum mechanical point may be viewed as an equilibrium thermodynamic subsystem in contact with the rest of the material which acts as a reservoir .
the question as to whether or not the fluctuating mass flux @xmath266 is zero , decides the very nature of this reservoir .
all continuum theories view the surrounding material as a reservoir for heat ( so that the point exchanges energy with the surrounding fluid to maintain a temperature determined by thermodynamics ) , but is it also a particle reservoir ( so that it exchanges mass with the surrounding fluid to maintain a thermodynamically determined chemical potential ) ? my theory , with its non - zero @xmath266 , asserts that it is , whereas the navier - stokes - fourier theory , in view of its ( [ c63 ] ) prediction , argues that it is not , meaning that each continuum mechanical point should be conceived of as having an impermeable wall surrounding it .
although subsystems and reservoirs are abstract constructs in this setting , i believe that it makes sense to allow the mass of a continuum mechanical point to fluctuate .
this view is reinforced by the natural appearance of fluctuation formulas ( [ c53 ] ) for a heat / particle reservoir in my theory s correlations , ( [ c45])-([c50 ] ) .
another conceptual difference between the two formulations is seen by comparing the @xmath293 case of equations ( [ c47 ] ) and ( [ c57 ] ) : my theory predicts no cross - correlations between the diagonal elements of the fluctuating momentum flux , whereas navier - stokes - fourier , in general , does .
furthermore , for a classical monatomic ideal gas or any other fluid for which one assumes @xmath294the navier - stokes - fourier theory predicts these cross - correlations to be negative , or in other words , for the diagonal elements of the fluctuating momentum flux to be inversely correlated with one another . one may derive the light scattering spectra corresponding to my formulation by using the procedure in berne and pecora @xcite .
the main steps involved are to ( 1 ) linearize the cartesian three - dimensional equations
expressed in terms of the variables , mechanical mass @xmath295 , particle number density @xmath296 , divergence of the velocity @xmath297 , and temperature @xmath7about the constant equilibrium state,@xmath298 ( 2 ) uncouple the mechanical mass density equation , ( 3 ) take the fourier - laplace transform of the remaining linear system , ( 4 ) solve for the fourier - laplace transformed variables , and ( 5 ) use this solution to construct time - correlation functions and their spectral densities . of particular interest is the density - density spectrum , computed for my formulation to be exactly@xmath299{c}\left ( 1-\frac{1}{\gamma_{0}}\right ) \frac{d_{0}k^{2}}{\omega^{2}+\left ( d_{0}k^{2}\right ) ^{2}}+\\ \frac{1}{2\gamma_{0}}\left [ \frac{d_{0}k^{2}}{\left ( \omega+\omega\left ( k\right ) \right ) ^{2}+\left ( d_{0}k^{2}\right ) ^{2}}+\frac{d_{0}k^{2}}{\left ( \omega-\omega\left ( k\right ) \right ) ^{2}+\left ( d_{0}k^{2}\right ) ^{2}}\right ] \end{array } \right\ } .\label{c65}\ ] ] in the above , @xmath277 is the magnitude of the wave vector , @xmath300 ; @xmath149 is the angular frequency ; @xmath301 is the structure factor defined as@xmath302 where @xmath285 represents the scattering volume ; and @xmath303 is the frequency shift of the brillouin doublets,@xmath304 the right - hand side of ( [ c65 ] ) is the sum of three lorentzian line shapes , the first of which is the central rayleigh peak , and the other two , the negatively and positively @xmath305-shifted brillouin peaks . the brillouin peaks represent light scattering due to the sound modes , or phonons , generated by pressure fluctuations , and the rayleigh peak is caused by light being scattered from specific entropy fluctuations .
the navier - stokes - fourier density - density spectrum is given by the more complicated expression,@xmath306{c}-\omega^{2}+i\omega\left [ \frac{\zeta_{ns,0}}{m_{0}}+\left ( \frac{4}{3}+e\!u_{0}\right ) \frac{\eta_{ns,0}}{m_{0}}\right ] k^{2}+\\ \left ( 1-\frac{1}{\gamma_{0}}\right ) \omega^{2}\left ( k\right ) + e\!u_{0}\left ( \frac{\zeta_{ns,0}}{m_{0}}+\frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}\right ) \frac{\eta_{ns,0}}{m_{0}}k^{4}\end{array } \right\ } } { \left\ { \begin{array } [ c]{c}-i\omega^{3}-\omega^{2}\left [ \frac{\zeta_{ns,0}}{m_{0}}+\left ( \frac{4}{3}+e\!u_{0}\right ) \frac{\eta_{ns,0}}{m_{0}}\right ] k^{2}+\\ i\omega\left [ \omega^{2}\left ( k\right ) + e\!u_{0}\left ( \frac{\zeta _ { ns,0}}{m_{0}}+\frac{4}{3}\frac{\eta_{ns,0}}{m_{0}}\right ) \frac{\eta _ { ns,0}}{m_{0}}k^{4}\right ] + \frac{e\!u_{0}}{\gamma_{0}}\frac{\eta_{ns,0}}{m_{0}}\omega^{2}\left ( k\right ) k^{2}\end{array } \right\ } } \right ) , \nonumber\end{gathered}\ ] ] where @xmath161 is the euken ratio defined in ( [ c18 ] ) .
in the hydrodynamic regime for which@xmath307 equation ( [ c68 ] ) becomes approximately@xmath299{c}\left ( 1-\frac{1}{\gamma_{0}}\right ) \frac{\sigma_{0}k^{2}}{\omega ^{2}+\left ( \sigma_{0}k^{2}\right ) ^{2}}+\\ \frac{1}{2\gamma_{0}}\left [ \frac{\gamma_{0}k^{2}}{\left ( \omega + \omega\left ( k\right ) \right ) ^{2}+\left ( \gamma_{0}k^{2}\right ) ^{2}}+\frac{\gamma_{0}k^{2}}{\left ( \omega-\omega\left ( k\right ) \right ) ^{2}+\left ( \gamma_{0}k^{2}\right ) ^{2}}\right ] + \\
\frac{b\left ( k\right ) } { 2}\left [ \frac{\left ( \omega+\omega\left ( k\right ) \right ) } { \left ( \omega+\omega\left ( k\right )
\right ) ^{2}+\left ( \gamma_{0}k^{2}\right ) ^{2}}-\frac{\left ( \omega-\omega\left ( k\right ) \right ) } { \left ( \omega-\omega\left ( k\right ) \right ) ^{2}+\left ( \gamma_{0}k^{2}\right ) ^{2}}\right ] \end{array } \right\ } , \label{c70}\ ] ] where @xmath204 and @xmath205 are defined in ( [ c18.1 ] ) and ( [ c18.2 ] ) and@xmath308 \frac{\eta_{ns,0}}{m_{0}}\right\ } .\label{c71}\ ] ] in approximate expression ( [ c70 ] ) , the first three terms represent the lorentzian rayleigh and brillouin peaks and the last two terms give a fairly small , but not negligible , non - lorentzian contribution . next , as in clark @xcite and fookson et al .
@xcite , let us define a dimensionless angular frequency,@xmath309 the dimensionless uniformity parameter,@xmath310 which is inversely proportional to the knudsen number for a gas , and the dimensionless density - density spectrum,@xmath311 which has an area of @xmath278 when written as a function of @xmath139 and @xmath312 and integrated over all @xmath139 . figure [ fig2 ] is a plot of the dimensionless density - density spectra , @xmath313 , predicted by my formulation ( blue ) and the navier - stokes - fourier formulation ( red ) for a classical monatomic ideal gas with @xmath314 and equations ( [ c34 ] ) , ( [ c35 ] ) , and ( [ c42 ] ) assumed .
the arrows indicate the widths at half - height .
( since the spectra are symmetric about the @xmath145 frequency , only the right brillouin peaks are shown . )
fig2c.eps both theories predict similar brillouin peaks , , parameters were chosen such that my formulation and the navier - stokes - fourier formulation gave the same sound predictions in the hydrodynamic regime . ] but mine has a rayleigh peak with a maximum about 29% higher and width at half - height about 29% narrower than that of navier - stokes - fourier .
on the other hand , the areas under both of the rayleigh peak predictions are the same , leading to identical landau - placzek ratios , i.e.@xmath315 which for a classical monatomic ideal gas is @xmath316 .
note that in my theory , the brillouin peaks and the rayleigh peak all have the same width at half - height , a quantity that is inversely proportional to the lifetime of the excitation corresponding to the peak .
this means that my theory predicts the same lifetime for phonons as it does for excitations caused by specific entropy fluctuations , whereas the navier - stokes - fourier formulation predicts that excitations caused by specific entropy fluctuations die out sooner than the phonons ( 29% sooner in the case of a classical monatomic ideal gas ) .
one other thing to note is that the brillouin portion of my exact spectrum ( [ c65 ] ) predicts phonons corresponding exactly to the adiabatic sound speed , @xmath317 , whereas the approximate navier - stokes - fourier spectrum ( [ c70 ] ) predicts a small amount of dispersion due to the non - lorentzian contributions . in figure
[ fig2 ] , it can be observed that for a classical monatomic ideal gas , this dispersion acts to shift the navier - stokes - fourier brillouin peaks slightly toward the center frequency .
the problem examined below strains the limits of what one should expect from the navier - stokes - fourier formulation and my theory in two ways : ( 1 ) it is not near - equilibrium , which is counter to the assumption of [ secce ] and appendix [ applc ] used to justify linear constitutive laws and ( 2 ) it is a @xmath318 problem and , therefore , not in the hydrodynamic regime .
nonetheless , the navier - stokes - fourier equations are commonly employed as a useful tool for investigating the internal structure of shock waves , and so it is of interest to see how predictions from my theory compare with these .
let us consider a cartesian one - dimensional steady - state shock wave like the one described in zeldovich and raizer @xcite and depicted schematically in figure [ mffig3 ] in a fixed reference frame .
the subscript `` @xmath275 '' is used to indicate undisturbed initial values ahead of the shock , and the subscript `` @xmath319 '' is used for final values after the shock moves through .
the initial and final fluid velocities are taken to be@xmath320 where @xmath321 is the constant speed of the piston generating the shock wave . for convenience ,
one transforms the problem into a coordinate system moving with the shock wave front as illustrated in figure [ mffig4 ] , where primes denote variables in the moving coordinate system and @xmath322 represents the constant speed of the shock front .
fig3c.eps fig4c.eps expressing balance laws ( [ bl1])-([bl4])/([bl12 ] ) with no body forces and diffusion transport constitutive equations ( [ tc19])-([tc21 ] ) in the one - dimensional moving coordinate system , making the steady - state assumption that the time derivatives vanish , using the helmholtz free energy formalism to cast the system in terms of the variables , @xmath323 , @xmath324 , @xmath325 , and @xmath326 , and employing the equilibrium thermodynamic relationships,@xmath327 for a classical monatomic ideal gas , yields the following system of ordinary differential equations:@xmath328 & = 0.\label{c81}\ ] ] for boundary conditions , let us assume that@xmath329 and@xmath330 and that all gradients , @xmath331 , vanish ahead of and behind the shock wave front , i.e.@xmath332 where @xmath333 may represent any of the variables , @xmath334 , @xmath324 , @xmath325 , or @xmath326 .
then , by integrating ( [ c78])-([c81 ] ) from @xmath335 to @xmath336 , one finds@xmath337{c}-d^{\prime}\frac{dv_{x}^{\prime}}{dx^{\prime}}+\\ \left ( \frac{5rt^{\prime}m^{\prime}}{2a\overline{m}^{\prime}}-\frac{5rt_{i}}{2a}\right ) + \frac{1}{2}\left ( v_{x}^{\prime2}-v_{s}^{2}\right ) \end{array } \right ] m_{i}v_{s},\label{c88}\ ] ] where ( [ c85 ] ) has been employed in writing equations ( [ c86])-([c88 ] ) , and by taking the limit of the above equations as @xmath338 , one arrives at the relations@xmath339 where ( [ c89 ] ) and ( [ c90 ] ) have been used to write ( [ c91 ] ) , and ( [ c90 ] ) and ( [ c91 ] ) have been used to obtain ( [ c92 ] ) .
one solves ( [ c92 ] ) for the positive root to find the following expression for the shock speed:@xmath340 therefore , if the initial mass density and temperature , @xmath341 and @xmath342 , and the piston speed , @xmath321 , are known , then all of the relevant final values , @xmath343 , @xmath344 , @xmath345 , and @xmath346 , may be computed via ( [ c93 ] ) and ( [ c89])-([c91 ] ) for a classical monatomic ideal gas . note that these are identical to the shock speed and final values found with the navier - stokes - fourier equations .
using the same assumptions and carrying out the foregoing procedure for the navier - stokes - fourier formulation , yields the following system:@xmath347{c}-\frac{4\eta_{ns}^{\prime}}{3m^{\prime}}\frac{dv_{x}^{\prime}}{dx^{\prime}}+\\ \left ( \frac{5rt^{\prime}}{2a}-\frac{5rt_{i}}{2a}\right ) + \frac{1}{2}\left ( v_{x}^{\prime2}-v_{s}^{2}\right ) \end{array } \right ] m_{i}v_{s}\label{c96}\ ] ] with the same boundary conditions as for my theory . in the above , classical monatomic ideal gas assumptions given in ( [ c34 ] ) have been used . for this non - linear problem , my diffusion coefficient , @xmath348 , and the navier - stokes shear viscosity , @xmath349 , may depend on state variables , e.g. it is common to express the viscosity with a temperature power law of the form ( [ dc1 ] ) : @xmath350 where @xmath351 is some reference temperature and @xmath352 is the viscosity measured at @xmath351 ( see appendix [ difp ] ) .
also , in appendix [ difp ] , it is shown that in a classical monatomic ideal gas one might reasonably take ( [ dc4]):@xmath353 figure [ mffig5 ] and figure [ mffig6 ] contain plots of numerical solutions to dimensionless versions of ( [ c85])-([c88 ] ) and ( [ c94])-([c96 ] ) with their appropriate boundary conditions and power laws ( [ dc1 ] ) and ( [ dc4 ] ) describing the transport coefficients .
the numerical method employed is iterative newton s method with centered differences approximating the spatial derivatives .
the dimensionless variables used are@xmath354 where @xmath355 and @xmath356 are the mean free path length and adiabatic sound speed in the initial state , and the mach number is computed as@xmath357 each of the figures corresponds to @xmath358 with parameters chosen to be those of alsmeyer @xcite for his shock wave experiments in argon gas : the undisturbed initial values,@xmath359 and the parameters appearing in ( [ dc1]),@xmath360 the curves plotted in figure [ mffig5 ] and figure [ mffig6 ] are the normalized mass density , @xmath361 ( in blue ) , the normalized fluid speed , @xmath362 ( in green ) , and the normalized temperature , @xmath363 ( in red ) versus the dimensionless position , @xmath364 . since the numerical solution sets are arbitrarily positioned on the @xmath364-axis , the @xmath365 point is chosen to be the center of the mass density profile ( the point at which the highest slope of @xmath366 occurs ) .
figure [ mffig5 ] contains the normalized shock wave profiles predicted by my theory and figure [ mffig6 ] contains those predicted by the navier - stokes - fourier formulation . comparing the curves in these two figures
, one observes that the mass density , temperature , and fluid velocity profiles are much closer together in my theory than for the navier - stokes - fourier formulation .
this difference is perhaps most obvious in the leading edge of the shock wave ( the @xmath367 side ) where my theory predicts these three profiles to be very near one another , whereas navier - stokes - fourier predicts a pronounced displacement between the three with temperature leading , velocity behind it , and mass density in the back .
i am unfamiliar with any direct physical arguments or experimental evidence that would explain why these profiles should be displaced in such a manner
. therefore , it would be very interesting to obtain measurements revealing the structure and position of these quantities .
experiments , such as alsmeyer s , which have been conducted to probe the internal structure of shock waves in a gas , measure only the mass density profile .
fig5c.eps fig6c.eps when a macroscopic particle is placed in a resting fluid with a temperature gradient , the particle tends to move in the cooler direction .
this phenomenon is known as thermophoresis , and as brenner @xcite points out , it is modelled with the navier - stokes - fourier formulation only by invoking a molecularly based thermal slip boundary condition . my formulation which features a non - convective mass flux , however , can provide a natural mechanism for thermophoresis even when a no - slip boundary condition is applied at the particle surface . the simplest problem that may be associated with thermophoresis is examined below using my theory . in it ,
gravity and possible boundary layers are neglected , and only small temperature gradients are considered . also , the thermophoretic particle is assumed to be insulated and to have a radius , @xmath225 , much larger than the fluid s mean free path length , @xmath368 , so that the knudsen number , defined for this problem as @xmath369 , is small .
let us consider a cartesian one - dimensional problem in which a fluid at average pressure , @xmath370 , lies between two parallel plates , a colder plate at @xmath174 maintained at temperature , @xmath371 , and a hotter plate at @xmath372 kept at temperature , @xmath373 , as shown in figure [ mffig7 ] .
also , assume the temperature of the fluid at the plates equals that of the plates , yielding the boundary conditions,@xmath374 and let us define the dimensionless parameter,@xmath375 where @xmath177 is the average temperature,@xmath376 the plates are assumed to be impermeable , i.e. allowing no mass to flow through them , which requires@xmath377 fig7c.eps it is most convenient to study this problem with the linearized gibbs free energy description of my formulation , i.e. equations ( [ c106])-([c109 ] ) , used in [ soundbc ] . here
, the linearization is about the constant average state,@xmath378 and @xmath143 is taken to be the parameter defined in ( [ c100.1 ] ) and assumed to be small .
note that in equations ( [ c106])-([c109 ] ) , `` @xmath145 '' subscripts are used to indicate evaluation at the average state , @xmath379 .
assuming steady - state conditions , the time derivatives appearing on the left - hand side of ( [ c106])-([c109 ] ) vanish , in which case these equations , together with the temperature boundary conditions ( [ c100.01 ] ) and definitions ( [ c100.1 ] ) and ( [ c100.2 ] ) , yield the solutions,@xmath380 one finds an expression for the constant velocity , by enforcing the linearized version of the no mass flow condition ( [ c100.7]):@xmath381 where the diffusion transport constitutive law ( [ tc19 ] ) has been assumed and recast in the gibbs free energy description . substituting ( [ c111 ] ) and ( [ c112 ] ) into the above and using ( [ c104 ] ) and ( [ c100.1 ] ) yields@xmath382
therefore , my theory predicts a constant velocity in the @xmath383 ( colder ) direction that is proportional to the fluid s diffusion coefficient and thermal expansion coefficient and the size of the temperature gradient imposed between the plates .
if an insulated macroscopic particle were present in such a velocity field with a no - slip condition applied at its surface , then the particle would simply be carried along at the fluid s velocity , i.e. the particle s thermophoretic velocity , @xmath384 , would equal the fluid s continuum velocity , @xmath140 , computed above .
in contrast , when the non - convective mass flux is assumed to be zero , as in the case of the navier - stokes - fourier formulation , the no mass flow condition ( [ c100.7 ] ) yields@xmath385 consequently , thermophoretic motion does not arise in the same sense that it does for my theory . to account for thermophoresis with the navier - stokes - fourier equations ,
one then finds it necessary to employ a thermal slip boundary condition at the particle s surface at which there is assumed to be a non - zero tangential velocity proportional to the temperature gradient and in the down - gradient direction . under this assumption ,
various researchers have been led , theoretically and experimentally , to the following expression for the thermophoretic velocity in the case of an insulated particle in a gas:@xmath386 where @xmath387 is the thermal slip coefficient , dimensionless and theoretically believed to lie in the interval , @xmath388 , with @xmath389 being a widely accepted theoretical estimate for gases . for the problem considered here , ( [ c116 ] )
becomes@xmath390 employing ( [ c35.5 ] ) and assuming an ideal gas for which@xmath391 the thermophoretic velocity predicted by my theory in equation ( [ c114 ] ) may be expressed as@xmath392 therefore , by identifying @xmath393 with @xmath387 , one finds this to be _ the same as the thermal slip formula _ ( [ c117 ] ) .
also , recall that in section [ sound ] , the value @xmath394 was obtained for a classical monatomic ideal gas . pertaining to a few other types of gases
are provided in appendix [ difp ] . ] as discussed in brenner @xcite , there is no real theory for thermal slip in a liquid .
therefore , researchers , e.g. mcnab and meisen @xcite , typically apply the gas formulas to liquids , even though they admit that there is no theoretical justification for doing so . for the idealized problem considered here ,
mcnab and meisen s data corresponds to a thermal slip coefficient for water and @xmath395-hexane near room temperature of @xmath396 , an order of magnitude smaller than that of a gas . again using ( [ c35.5 ] ) , but no longer assuming the ideal gas formula ( [ c117.5 ] ) , my theory s thermophoretic velocity ( [ c114 ] ) becomes@xmath397 which when compared with the thermal slip formula ( [ c117 ] ) , yields the relationship,@xmath398 for water at @xmath399 k , we take @xmath400 ( see appendix [ difp ] ) and @xmath401 k@xmath402 ( from the crc handbook of chemistry and physics @xcite ) , giving a slip coefficient value of @xmath403 , which is _ the same as the experimental value mentioned above_. it is interesting to note that there is , in general , no positive definite restriction on the thermal expansion coefficient @xmath404 .
there exist equilibrium thermodynamic states in water and liquid helium , for example , in which @xmath404 may vanish or become negative . in view of equation ( [ c120 ] )
, this means that such states could give rise to the occurrence of no thermophoresis or negative thermophoresis .
in this appendix , indicial notation is used in which sums are taken over repeated indices .
the indices may assume the values 1 , 2 , or 3 , representing the three spatial directions .
* dots and contraction operators:@xmath405@xmath406@xmath407 _ { j } & = d_{iji}\label{to6}\ ] ] if @xmath14 and @xmath408 are tensors of order @xmath97 and @xmath409 , respectively , with @xmath410 , then@xmath411 _ { i_{1}\ldots i_{m - n}}=a_{i_{1}\ldots i_{m - n}j_{1}\ldots j_{n}}b_{j_{1}\ldots j_{n}}.\label{to6.5}\ ] ] * norms:@xmath412 * tensor product:@xmath413 * cross products:@xmath414 where @xmath415 is the alternating symbol,@xmath416{l}1\text { if } ( i , j , k)=(1,2,3),(2,3,1),\text { or } ( 3,1,2)\\ -1\text { if } ( i , j , k)=(1,3,2),(2,1,3),\text { or } ( 3,2,1)\\ 0\text { if any of the indices are repeated}\end{array } \right\ } \label{to12}\ ] ] the transpose , symmetric part , skew - symmetric part , spherical part , and deviatoric part of a tensor are defined by@xmath417 and the second - order identity tensor is defined as@xmath418 where @xmath282 is the kronecker delta,@xmath419{l}1\text { if } i = j\\ 0\text { if } i\neq j \end{array } \right\ } .\label{to16.2}\ ] ] * gradients ( cartesian coordinate definitions):@xmath420 * divergences ( cartesian coordinate definitions):@xmath421 * convective derivative:@xmath422 * generalized partial derivative : if @xmath14 and @xmath408 are tensors of order @xmath97 and @xmath409 , respectively , then@xmath423 @xmath424 @xmath425 @xmath426 @xmath427 @xmath428 @xmath429 \label{to27}\ ] ] @xmath430 @xmath431 @xmath432
taking the partial time derivative of @xmath277 as defined by ( [ bl7 ] ) and using the product rule , we find@xmath433 or , substituting ( [ bl3 ] ) and ( [ bl1 ] ) into the above,@xmath434 + \overline{m}\underline{f}^{\left ( \overline{m}\right ) } \cdot\underline{v}+\frac{1}{2}\left\vert \underline{v}\right\vert ^{2}\nabla\cdot\left ( \overline{m}\underline{v}\right ) .\ ] ] one can then use tensor identities ( [ to22 ] ) and ( [ to25 ] ) to obtain@xmath435 + \underline{\underline{p}}^{t}\colon\nabla \underline{v}+\overline{m}\left ( \underline{v}\,\underline{v}\right ) \cdot\cdot\nabla\underline{v}+\\ & \overline{m}\underline{v}\cdot\underline{f}^{\left ( \overline{m}\right ) } + \nabla\cdot\left ( \frac{1}{2}\overline{m}\left\vert \underline{v}\right\vert ^{2}\underline{v}\right ) -\frac{1}{2}\overline{m}\underline{v}\cdot\nabla\left ( \underline{v}\cdot\underline{v}\right ) .\end{aligned}\ ] ] finally , if we employ tensor identities ( [ to23 ] ) and ( [ to26 ] ) , then we arrive at equation ( [ bl7.5 ] ) . with the aid of tensor identity ( [ to27 ] ) , the first term on the right - hand side of ( [ bl13 ] ) may be written as@xmath436 & = -\nabla \cdot\left [ \underline{x}\times\left ( \underline{\underline{p}}+\overline { m}\underline{v}\,\underline{v}\right ) \right ] + c_{1,3}\left [ \underline{\underline{1}}\times\underline{\underline{p}}\right ] + \nonumber\\ & c_{1,3}\left [ \underline{\underline{1}}\times\left ( \overline { m}\underline{v}\,\underline{v}\right ) \right ] , \label{am1}\ ] ] using the fact that@xmath437 and the linearity of the contraction operator , @xmath438 . next , by writing out its components , it is observed that@xmath439 = \underline{0}\ ] ] is satisfied if and only if @xmath440 is symmetric . therefore , the symmetry of @xmath441 causes the third term on the right - hand side of ( [ am1 ] ) to vanish .
furthermore , the second term also vanishes , implying conservation of momentum ( [ bl14 ] ) , if and only if @xmath8 is symmetric .
here , classical irreversible thermodynamics , as described in de groot and mazur ( * ? ? ?
* ch.iii and iv ) and jou et al .
@xcite , is used to derive linear constitutive laws ( [ ce12])-([ce14 ] ) .
recall equation ( [ ce9.3 ] ) for the volumetric entropy production rate and note that the second term on the right - hand side suggests the following decomposition for the pressure tensor of a fluid:@xmath442 where @xmath116 represents the viscous part of the pressure , assumed symmetric , i.e.@xmath443 so that condition ( [ bl15 ] ) is satisfied , and @xmath131 , as before , denotes the equilibrium thermodynamic pressure . consequently , equation ( [ ce9.3 ] ) becomes@xmath444 one can further decompose the viscous pressure into its spherical and deviatoric parts via@xmath445 with@xmath446 substituting ( [ l4 ] ) and ( [ l5 ] ) into ( [ l3 ] ) and using identity ( [ to28 ] ) yields@xmath447 + \nonumber\\ & \underline{\underline{p}}_{visc}^{dev}\cdot\cdot\left [ -\frac{1}{t}\left ( \nabla\underline{v}\right ) ^{sym , dev}\right ] + \underline{q}_{m}\cdot\left [ -\nabla\left ( \frac{\mu}{t}\right ) \right ] , \label{l6}\ ] ] where property ( [ l2 ] ) has been employed in conjunction with tensor identity ( [ to30 ] ) .
the terms have been grouped as in expression ( [ l6 ] ) so that @xmath448 has the form,@xmath449 where @xmath450 and @xmath451 represent fluxes and the affinities , respectively .
one requires the affinities to be objective tensor quantities and for there to exist equilibrium states for which the affinities and fluxes simultaneously vanish .
comparing ( [ l7 ] ) and ( [ l6 ] ) suggests the following identifications:@xmath452{|l|l|l|}\hline $ j$ & $ \mathbf{f}^{\left ( j\right ) } $ ( flux ) & $ \mathbf{a}^{\left ( j\right ) } $ ( affinity)\\\hline $ 1 $ & $ \underline{f}^{\left ( 1\right ) } = \underline{q}_{u}$ & $ \underline{a}^{\left ( 1\right ) } = \nabla\left ( \frac{1}{t}\right ) $ \\\hline $ 2 $ & $ f^{\left ( 2\right ) } = p_{visc}$ & $ a^{\left ( 2\right ) } = -\frac{1}{t}\nabla\cdot\underline{v}$\\\hline $ 3 $ & $ \underline{\underline{f}}^{\left ( 3\right ) } = \underline{\underline{p}}_{visc}^{dev}$ & $ \underline{\underline{a}}^{\left ( 3\right ) } = -\frac{1}{t}\left ( \nabla\underline{v}\right ) ^{sym , dev}$\\\hline $ 4 $ & $ \underline{f}^{\left ( 4\right ) } = \underline{q}_{m}$ & $ \underline{a}^{\left ( 4\right ) } = -\nabla\left ( \frac{\mu}{t}\right ) $ \\\hline \end{tabular } .\label{l8}\ ] ] note that @xmath453 , @xmath454 , @xmath455 , and @xmath456 are each indeed objective tensor quantities .
next , let us assume that the fluxes are written as@xmath457 where @xmath458 represents a list of state parameters , each evaluated at @xmath459 , e.g. @xmath460 or @xmath461 .
the assumption that the affinities and fluxes vanish in a state of equilibrium implies@xmath462 so that we obtain the following expansions for the fluxes about the equilibrium state:@xmath463 assuming the magnitudes of the affinities are small enough to neglect the higher - order terms and defining the phenomenological coefficients as@xmath464 equation ( [ l11 ] ) becomes the linear constitutive law,@xmath465 with the use of table ( [ l8 ] ) , the linear constitutive laws for our single - component viscous fluid are given by@xmath466 + \nonumber\\ & \underline{\underline{l}}^{\left ( 14\right ) } \cdot\left [ -\nabla\left ( \frac{\mu}{t}\right ) \right ] , \label{l14}\]]@xmath467 + \nonumber\\ & \underline{l}^{\left ( 24\right ) } \cdot\left [ -\nabla\left ( \frac{\mu}{t}\right ) \right ] , \label{l15}\]]@xmath468 + \nonumber\\ & \underline{\underline{\underline{l}}}^{\left ( 34\right ) } \cdot\left [ -\nabla\left ( \frac{\mu}{t}\right ) \right ] , \label{l16}\ ] ] and@xmath469 + \nonumber\\ & \underline{\underline{l}}^{\left ( 44\right ) } \cdot\left [ -\nabla\left ( \frac{\mu}{t}\right ) \right ] .\label{l17}\ ] ] as discussed in de groot and mazur ( * ? ? ?
* ch.vi ) , in the absence of external magnetic fields and coriolis forces , onsager reciprocity requires that the phenomenological coefficients appearing in the above satisfy@xmath470{c}\mathbf{l}^{\left ( kj\right ) } \text { for } \left ( j , k\right ) = ( 1,1),(1,4),(2,2),(2,3),(3,2),(3,3),(4,1),(4,4)\\ -\mathbf{l}^{\left ( kj\right ) } \text { for } \left ( j , k\right ) = ( 1,2),(1,3),(2,1),(2,4),(3,1),(3,4),(4,2),(4,3 ) \end{array } \right\ } .\label{l18}\ ] ] furthermore , as discussed in segel @xcite , isotropic materials like fluids possess isotropic tensors for each of their phenomenological coefficients , @xmath471 , and this reduces equations ( [ l14])-([l17 ] ) to@xmath472 , \label{l24}\]]@xmath473@xmath474 , \label{l26}\ ] ] and@xmath475 \label{l27}\ ] ] with@xmath476@xmath477@xmath478@xmath479 and@xmath480 \label{l32}\ ] ] for scalars , @xmath481 , @xmath482 , @xmath483 , @xmath484 , @xmath485 , and where @xmath282 represents the kronecker delta . substituting ( [ l29 ] ) and ( [ l30 ] ) into equation ( [ l24 ] ) , one finds@xmath486 whereby defining@xmath487 and@xmath488 one obtains equation ( [ ce14 ] ) .
similarly , if one employs ( [ l30.2 ] ) and ( [ l31 ] ) , then equation ( [ l27 ] ) becomes@xmath489 and , therefore , defining@xmath490 and@xmath491 yields equation ( [ ce12 ] ) .
note that equations ( [ l35.5 ] ) , ( [ l38 ] ) , and ( [ l39 ] ) imply the relation,@xmath492 which is a consequence of the onsager reciprocity .
next , let us substitute ( [ l32 ] ) into ( [ l26 ] ) to find@xmath493 \left ( \nabla\underline{v}\right ) _ { kl}^{sym , dev}\\ & = \frac{2\mathcal{l}^{\left ( 33\right ) } } { 3t}tr\left [ \left ( \nabla\underline{v}\right ) ^{sym , dev}\right ] \delta_{ij}-\frac { \mathcal{l}^{\left ( 33\right ) } } { t}\left [ \left ( \nabla\underline{v}\right ) _ { ij}^{sym , dev}+\left ( \nabla\underline{v}\right ) _ { ji}^{sym , dev}\right ] , \end{aligned}\ ] ] which implies@xmath494 since the trace of any deviatoric tensor is zero . if one defines@xmath495 and@xmath496 then ( [ l43 ] ) and ( [ l25 ] ) become@xmath497 and@xmath498 finally , using the two equations above , together with ( [ l5 ] ) , ( [ l4 ] ) , and ( [ l1 ] ) , one arrives at equation ( [ ce13 ] ) for the total pressure .
it is easily shown that if the mass and internal energy fluxes are chosen as@xmath499 then onsager reciprocity condition ( [ ce16 ] ) can not be satisfied unless@xmath500 furthermore , constitutive equations ( [ tc16 ] ) and ( [ tc18 ] ) may be used to prove that similar flux laws hold for other thermodynamic extensive quantities , as well , e.g. the entropy , for which@xmath501 consequently , i call ( [ tc4 ] ) , ( [ tc5 ] ) , ( [ tc8 ] ) , and ( [ tc9 ] ) , `` thermodynamic diffusion transport coefficients , '' and ( [ tc16 ] ) and ( [ tc18 ] ) , or the more compactly expressed , ( [ tc19 ] ) and ( [ tc21 ] ) , `` thermodynamic diffusion flux laws . '' on the other hand , there is nothing obvious to prevent the choice of a different diffusion coefficient , @xmath502 , to be associated with the viscous pressure . in this case , the `` momentum diffusion transport coefficients , '' ( [ tc13 ] ) and ( [ tc14 ] ) , may be written as@xmath503 and@xmath504 and the `` momentum diffusion flux law '' ( [ tc20 ] ) as@xmath505 of course , the diffusion transport fluxes are found by choosing thermodynamic and momentum diffusion fluxes with @xmath506 .
to model a multicomponent mixture of fluids , one may use the same ideas as those used to obtain ( [ smf1])/([tc19])-([tc21 ] ) . let @xmath507 represent the mass of component @xmath275 in an @xmath508-component mixture . for an @xmath508-component fluid near equilibrium and in the continuum regime ,
i propose the @xmath62-formulation balance laws,@xmath509 with diffusion transport constitutive equations,@xmath510 notice that there still appears only one transport coefficient , @xmath4 .
also , note that since@xmath511 one of the partial mass equations in the above may be replaced by the total mass equation,@xmath512 with@xmath513 in this setting , the continuum velocity @xmath3 is interpreted as a center of mechanical mass velocity , and in this center of mechanical mass frame , there is a non - zero total mass flux due to diffusion . also , note that if the equilibrium thermodynamic pressure , @xmath131 , may be expressed as a function of @xmath44 and @xmath1 only as in the case of a homogeneous mixture the @xmath514 partial mass density equations may be uncoupled from the rest of the system , leaving formulation ( [ smf1])/([tc19])-([tc21 ] ) derived for a single - component fluid .
this means that homogeneous mixtures such as air may be treated in the same way as a single - component fluid .
here , measured sound propagation data is used together with formulas from [ sound ] to compute the diffusion coefficient , @xmath4 , of my formulation for various types of fluids .
the temperature , pressure , and mass density dependence of this parameter is explored where data is available .
table [ tab1 ] provides values of the diffusion coefficient , @xmath211 , and the self - diffusion coefficient , @xmath515 , ( where available ) for several gases and liquids at @xmath242 k ( unless indicated otherwise ) and @xmath243 pa . [ c]|llll|**fluid * * & & & + * gases * & & & + helium & & & + neon & & & + argon & & & + krypton & & & + xenon & & & + nitrogen & & & + oxygen & & & + carbon dioxide & & & + methane & & & + air & & & + * liquids * & & & + water & & & + mercury & & & + glycerol & & & + benzene & & & + ethyl alcohol & & & + castor oil & & & + [ tab1 ] for the noble gases , helium , neon , argon , krypton , and xenon , @xmath393 is given by ( [ c36 ] ) and the diffusion coefficient is computed via ( [ c35 ] ) with shear viscosities from kestin et al .
@xcite and mass densities determined by the classical ideal gas relationship,@xmath516 which is appropriate for each of the table [ tab1 ] gases in the normal temperature and pressure regime . for the diatomic gases , nitrogen and oxygen , and the polyatomic gases ,
carbon dioxide and methane , @xmath211 and @xmath393 are computed by equations ( [ c33 ] ) and ( [ c35.5 ] ) , respectively , using data presented in marques @xcite , viscosity data from cole and wakeham @xcite and trengrove and wakeham @xcite , and densities computed via relation ( [ dc.1 ] ) . for air ) . ] and the liquids , @xmath211 is computed via equation ( [ c28.6 ] ) and @xmath393 by relation ( [ c35.5 ] ) . the sound propagation quantities , @xmath200 and @xmath317 , are calculated for air using @xcite with @xmath241 mhz and @xmath517 humidity , given for mercury in hunter et al .
@xcite , and tabulated for the rest of the liquids in @xcite .
the viscosities and mass densities are taken from @xcite for air , @xcite for water , @xcite for mercury , @xcite for glycerol , @xcite and @xcite for benzene , @xcite for ethyl alcohol , and @xcite and @xcite for castor oil .
the self - diffusion coefficients are taken from kestin et al .
@xcite for the noble gases , from winn @xcite for the diatomic and polyatomic gases , from holz et al .
@xcite for water , from nachtrieb and petit @xcite for mercury , from tomlinson @xcite for glycerol , from kim and lee @xcite for benzene , and from meckl and zeidler @xcite for ethyl alcohol . from the values presented in table [ tab1 ]
, one makes the following observations . * in each of the gases , the diffusion and self - diffusion coefficients are observed to be the same order of magnitude .
however in liquids , the self - diffusion coefficients are several orders of magnitude smaller than the @xmath211 values .
this indicates that although in gases the diffusion coefficient may be roughly approximated by self - diffusion , the same may not be said of liquids . * the diffusion parameters of the noble gases
are seen to decrease with increasing mass density . *
the value for @xmath211 in air is approximately a weighted average of the @xmath211 values for nitrogen and oxygen based on their fractional compositions in air ( @xmath518 nitrogen and @xmath519 oxygen ) .
* with the exception of very light gases like helium and very heavy gases like krypton and xenon , diffusion coefficients for gases at normal temperature and pressure are typically on the order of @xmath520 m@xmath521s .
* the @xmath211 values of water and ethyl alcohol which fall into the category of medium viscosity , medium density liquids are similar and on the order of @xmath522 m@xmath521s , an order of magnitude smaller than that of a typical gas .
the high viscosity , medium density liquids , glycerol and castor oil , possess similar @xmath211 values on the higher order of @xmath523 m@xmath521s .
mercury s viscosity is medium range like water and ethyl alcohol , but its higher mass density results in a smaller @xmath211 on the order of @xmath524 m@xmath521s .
[ [ gases ] ] gases + + + + + in general , it is observed from sound attenuation data for gases in the hydrodynamic regime that the product , @xmath525 , may vary with temperature but has negligible pressure or density dependence at a fixed temperature . using this observation ,
together with relationship ( [ tc22 ] ) and the fact that the navier - stokes shear viscosity @xmath91 also has negligible pressure or density dependence , leads us to conclude that although @xmath125 is possibly a function of @xmath7 , it does not tend to vary much with @xmath131 or @xmath1 for gases .
typically , @xmath91 may be expressed as a temperature power law of the form,@xmath526 where @xmath351 is some reference temperature and @xmath352 is the viscosity measured at @xmath351 . using this in equation ( [ tc22 ] ) , one finds@xmath527 equations ( [ c33 ] ) and ( [ c35.5 ] ) together imply@xmath528 \right\ } .\label{dc3}\ ] ] for noble gases , such as argon , one expects the values for @xmath169 , @xmath529 , and @xmath530 given in ( [ c34 ] ) to hold over a broad temperature range . therefore ,
in the noble gases , we take @xmath127 to be constant and , substituting this into ( [ dc2]),@xmath531 on the other hand , as observed in tables [ tab2]-[tab3.5 ] , for non - monatomic gases such as air , nitrogen , and methane , @xmath125 displays a tendency to increase with temperature . in tables
[ tab2]-[tab3.5 ] , the @xmath211 values are computed by equation ( [ c28.6 ] ) with @xmath200 calculated using @xcite for air and taken from @xcite for nitrogen and methane and with @xmath317 calculated from the classical ideal gas formula,@xmath532 where the @xmath529 values are taken from @xcite for air and approximated to be constant over the studied temperature range for nitrogen and methane with respective values @xmath533 and @xmath534 .
the @xmath393 values appearing in tables [ tab2]-[tab3.5 ] are computed via equation ( [ c35.5 ] ) with navier - stokes shear viscosities taken from @xcite for air and from @xcite for nitrogen and methane and mass densities taken from @xcite for air and computed by ideal gas formula ( [ dc.1 ] ) for nitrogen and methane .
[ c]|l|l|l|@xmath177 ( k ) & @xmath211 ( m@xmath521s ) & @xmath393 + @xmath535 & @xmath536 & @xmath537 + @xmath538 & @xmath539 & @xmath537 + @xmath540 & @xmath541 & @xmath542 + @xmath543 & @xmath544 & @xmath542 + @xmath545 & @xmath546 & @xmath547 + @xmath548 & @xmath549 & @xmath550 + @xmath551 & @xmath552 & @xmath553 + @xmath554 & @xmath555 & @xmath556 + [ tab2 ] [ c]|l|l|l|@xmath177 ( k ) & @xmath211 ( m@xmath521s ) & @xmath393 + @xmath557 & @xmath558 & @xmath559 + @xmath560 & @xmath561 & @xmath537 + @xmath562 & @xmath563 & @xmath564 + @xmath545 & @xmath565 & @xmath566 + [ tab3 ] [ c]|l|l|l|@xmath177 ( k ) & @xmath211 ( m@xmath521s ) & @xmath393 + @xmath557 & @xmath567 & @xmath550 + @xmath560 & @xmath568 & @xmath569 + @xmath562 & @xmath570 & @xmath571 + @xmath545 & @xmath572 & @xmath573 + [ tab3.5 ] [ [ liquids ] ] liquids + + + + + + + using equations ( [ c28.6 ] ) and ( [ c35.5 ] ) with sound speed and attenuation data from @xcite and mass density and shear viscosity data from @xcite , one obtains the estimates of @xmath211 and @xmath393 for water at atmospheric pressure ( @xmath243 pa ) and various temperatures @xmath177 between the freezing and boiling point appearing in table [ tab4 ] .
[ c]|c|c|c|c|@xmath177 ( k ) & @xmath211 ( m@xmath245/s ) & @xmath515 ( m@xmath245/s ) & @xmath393 + @xmath540 & @xmath574 & @xmath575 & @xmath576 + @xmath543 & @xmath577 & @xmath578 & @xmath579 + @xmath545 & @xmath580 & @xmath581 & @xmath579 + @xmath548 & @xmath582 & @xmath583 & @xmath584 + @xmath551 & @xmath585 & @xmath586 & @xmath584 + @xmath554 & @xmath587 & @xmath588 & @xmath584 + @xmath589 & @xmath590 & @xmath591 & @xmath584 + @xmath592 & @xmath593 & @xmath594 & @xmath584 + @xmath595 & @xmath596 & @xmath597 & @xmath579 + @xmath598 & @xmath599 & @xmath600 & @xmath579 + @xmath601 & @xmath602 & @xmath603 & @xmath576 + [ tab4 ] as we can see , the parameter @xmath393 does not vary much over the entire temperature range , but the diffusion coefficient @xmath211 decreases with temperature .
an excellent least squares fit is made to the table [ tab4 ] data with@xmath604 in the third column of table [ tab4 ] , measured self - diffusion coefficients from holz et al .
@xcite are tabulated for water at atmospheric pressure . unlike our observation for gases ,
these self - diffusions are two to three orders of magnitude less than the corresponding @xmath211 values .
also , unlike @xmath211 , the self - diffusion _ increases _ with temperature . in table
[ tab5 ] , @xmath211 and @xmath393 are computed for water at two fixed temperatures , @xmath540 k and @xmath548 k , and various pressures roughly between @xmath278 and @xmath605 atmospheres .
the values are obtained using equations ( [ c28.6 ] ) and ( [ c35.5 ] ) together with data presented in litovitz and carnevale @xcite .
[ c]|llll|@xmath370 ( pa ) & & & + @xmath606 k & & & + @xmath607 & & & + @xmath608 & & & + @xmath609 & & & + @xmath610 & & & + @xmath611 & & & + @xmath612 k & & & + @xmath607 & & & + @xmath608 & & & + @xmath609 & & & + @xmath610 & & & + @xmath611 & & & + [ tab5 ] as we can see in the fourth column of table [ tab5 ] , the product of the mass density and the diffusion coefficient does not vary much over the studied pressure range ( @xmath613 for @xmath540 k and @xmath614 for @xmath548 k ) .
therefore , as in gases , it is reasonable to assume that the quantity @xmath525 is essentially pressure independent for water and perhaps for other liquids , as well , although this remains to be experimentally verified . using equations ( [ c28.6 ] ) and ( [ c35.5 ] ) with data from hunter et al .
@xcite yields the estimates of @xmath211 and @xmath393 for liquid mercury at atmospheric pressure ( @xmath243 pa ) and various temperatures @xmath177 between @xmath615 k and @xmath616 k appearing in table [ tab6 ] .
[ c]|l|l|l|@xmath177 ( k ) & @xmath211 ( m@xmath521s ) & @xmath393 + @xmath615 & @xmath617 & @xmath618 + @xmath548 & @xmath619 & @xmath620 + @xmath551 & @xmath621 & @xmath622 + @xmath623 & @xmath624 & @xmath625 + @xmath589 & @xmath626 & @xmath627 + @xmath628 & @xmath629 & @xmath630 + @xmath592 & @xmath631 & @xmath632 + @xmath595 & @xmath633 & @xmath634 + @xmath635 & @xmath636 & @xmath637 + @xmath616 & @xmath638 & @xmath639 + [ tab6 ] unlike our observation for water in table [ tab4 ] , the @xmath211 values for mercury are seen to increase with increasing temperature .
a fairly good fit to the table [ tab6 ] data is given by@xmath640
in addition to those introduced in [ secnt ] , this paper utilizes the following symbols and notational conventions : the constants,@xmath452{ll}$r$ & universal gas constant\\ $ k_{b}$ & boltzmann constant\\ $ a$ & atomic weight\\ $ n_{a}$ & avagadro 's number , \end{tabular}\ ] ] the equilibrium thermodynamic quantities,@xmath452{ll}$\widetilde{s}^{\left ( \overline{m}\right ) } $ & fundamental relation for entropy per mechanical mass\\ $ n_{p}$ & particle number density\\ $ h^{\left ( m\right ) } $ & specific enthalpy\\ $ \alpha_{p}$ & coefficient of thermal expansion\\ $ \kappa_{t}$ & isothermal compressibility\\ $ c_{v}$ & isochoric specific heat per mass\\ $ c_{p}$ & isobaric specific heat per mass\\ $ \gamma$ & ratio of specific heats , $ c_{p}/c_{v}$\\ $ c$ & adiabatic sound speed , \end{tabular}\ ] ] the non - equilibrium thermodynamic quantities,@xmath452{ll}$\mathbf{j}_{\mathbf{a}}$ & total $ \mathbf{a}$-flux\\ $ \mathbf{q}_{\mathbf{a}}$ & non - convective $ \mathbf{a}$-flux\\ $ \mathbf{q}_{\mathbf{a}}^{dif}$ & diffusion transport $ \mathbf{a}$-flux\\ $ \mathbf{r}_{\mathbf{a}}$ & volumetric $ \mathbf{a}$-production / destruction rate \end{tabular}\ ] ] and@xmath452{ll}$\underline{\underline{p}}$ & total pressure tensor\\ $ \underline{\underline{p}}_{visc}$ & viscous part of the pressure tensor\\ $ p_{visc}\underline{\underline{1}}$ & spherical part of $ \underline{\underline{p}}_{visc}$\\ $ \underline{f}^{\left ( \overline{m}\right ) } $ & external body force per mechanical mass\\ $ \underline{q}_{heat}$ & heat flux\\ $ \underline{q}_{chem}$ & energy flux due to chemical work , \end{tabular}\ ] ] the transport coefficients,@xmath452{ll}$d_{m},d_{u}$ & coefficients in general mass flux law\\ $ \eta,\zeta$ & coefficients in general momentum flux law\\ $ k_{m},k_{u}$ & coefficients in general internal energy flux law \end{tabular}\ ] ] and@xmath452{ll}$\eta_{ns}$ & navier - stokes shear viscosity\\ $ \zeta_{ns}$ & navier - stokes bulk viscosity\\ $ k_{f}$ & fourier heat conductivity\\ $ d$ & my diffusion transport coefficient\\ $ d_{sd}$ & self - diffusion coefficient , \end{tabular}\ ] ] and the dimensionless parameters,@xmath452{ll}$k\!n$ & knudsen number\\ $ m\!a$ & mach number\\ $ \varepsilon$ & linearization parameter $ \propto m\!a$\\ $ e\!u$ & euken ratio\\ $ c$ & coefficient relating $ d$ to $ \eta_{ns}/m.$\end{tabular}\ ] ] furthermore , the subscript `` @xmath145 '' is used to denote evaluation at the equilibrium state and the subscript `` @xmath278 '' to indicate first - order ( linear ) terms in the small @xmath143 expansion about equilibrium . | the continuum equations of fluid mechanics are rederived with the intention of keeping certain mechanical and thermodynamic concepts separate .
a new `` mechanical '' mass density is created to be used in computing inertial quantities , whereas the actual mass density is treated as a thermodynamic variable .
a new set of balance laws is proposed , including a mass balance equation with a non - convective flux .
the basic principles of irreversible thermodynamics are used to obtain linear constitutive equations that are expansions of not only the usual affinities involving gradients of temperature and velocity but also the gradient of the chemical potential .
transport coefficients are then chosen based on an elementary diffusion model , which yields simple constitutive laws featuring just one diffusion transport parameter .
the resulting formulation differs from the navier - stokes - fourier equations of fluid motion . in order to highlight key similarities and differences between the two formulations ,
several fluid mechanics problems are examined , including sound propagation , light scattering , steady - state shock waves , and thermophoresis . |
the ads / cft correspondence introduced by maldacena @xcite has helped shed light on behaviors of systems otherwise far too complicated to tackle . specifically , it has given us a method by which strongly interacting regimes of theories can be studied quite easily through considering the dual gravity theory which is mostly described by geometry .
also , it allows us to study responses to the field theory to background electromagnetic fields by introducing gauge fields in the bulk , which is another simplifying feature .
many ads / cft systems with a background magnetic field have been studied , especially in the case of the ads@xmath7 magnetic brane used to investigate 2 + 1 gauge theories in magnetic fields @xcite .
also , ads@xmath5 case has been tackled in @xcite , which was motivated by rhic physics which involves strongly coupled gauge theories in magnetic fields @xcite .
a slight generalization of the configuration of the magnetic fields can prove to be useful in shedding light on more exotic features of the system .
another interesting aspect of ads / cft is the correspondence in 2 dimensions , i.e. ads@xmath8/cft@xmath9 . as opposed to common belief that smaller dimension means simpler physics , this case of the duality has proven to be the least understood @xcite .
this is partly due to the disconnected feature of the boundary of the ads@xmath1 space , giving us two disconnected dual descriptions commonly understood to be systems described by conformal quantum mechanics .
recent interest stems from the emergence of quantum critical behavior in this duality . such behavior , involving non - fermi liquid behavior , was studied in @xcite .
a new setup showing the the emergence of the ads@xmath1 space - time will prove useful to the cause of understanding the duality in 2 dimensions . driven by these motivations we consider bulk solutions that interpolate between ads@xmath5 in the uv regime and ads@xmath0@xmath10t@xmath11 in the ir along with the presence of a magnetic field .
a similar system was studied in @xcite but we generalize the problem by considering a magnetic field that is comprised of a linear combination of three @xmath3 s of @xmath4 . in this generalization
, the linear combination is parameterized by elements of the unimodular tensor , @xmath12 , that is in the 20 representation of @xmath4 which emerges from the deformations of the 5-sphere in the ads@xmath5@xmath13 after truncation @xcite .
we consider a simple diagonal form for the tensor with three parameters @xmath14 . at finite temperature
, the large coupling limit is the product of a btz black hole and a t@xmath11 .
the entropy is found to scale as @xmath15 , temperature , just as in @xcite .
the added feature in our result is that the entropy is also a function of the linear combination parameters , @xmath16 .
we also consider the effect of this generalization on the @xmath17 sym theory .
the charged particles are now charged under a linear combination of the @xmath18 magnetic fields .
it s entropy too depends on the linear combination parameters .
the relation between the entropies takes the form @xmath19 where the peak of the function @xmath20 , occurring at @xmath21 , corresponds to the result @xmath22 obtained in @xcite .
we also find solutions for arbitrary @xmath23 below the peak value .
this is an indication of the influence of the chosen linear combination on the physics of the problem .
next , we look at the other system of interest which interpolates between ads@xmath5 and ads@xmath1@xmath10t@xmath24
. the magnetic field considered has the added feature that each @xmath3 points in a different direction in physical space ; one along each direction of the torus .
we find that this system can be solved exactly , and we obtain the interpolating solution for all @xmath25 .
we find that the solution interpolates between ads@xmath5 at large @xmath25 and the product of an extremal two dimensional black hole with t@xmath24 .
the black hole thus has finite entropy at zero temperature hinting at a 1 dimensional dual field theory with a degenerate ground state . across the duality ,
we show that ads@xmath1 magnetic field configuration does not alter the temperature scaling of the entropy .
we then show that we still have the regular free @xmath17 sym theory , with zero entropy at zero temperature .
finally , we investigate the stability properties of the solutions at hand , as this is often a problem for condensed matter applications of ads / cft duality .
we consider small perturbations of the unimodular tensor away from the identity .
we find two classes of solutions having different @xmath26 .
we find that for both space - times the @xmath26 s are well above the respective @xmath27 .
thus both the space - times are stable . in this paper
, we start in section 2 by constructing the space - time that interpolates between ads@xmath5 and ads@xmath28t@xmath11 and we compute the relevant entropy . in section 3
we discuss the @xmath17 sym dual field theory and compare the entropies .
then we go on to investigate the thermodynamics of the ads@xmath5/ads@xmath29t@xmath11 space - time in section 4 . in section 5
we present an argument showing that the entropy of the free field limit of @xmath17 sym still scales as temperature for the ads@xmath1 magnetic field configuration . in section 6
we discuss the stability of the space - times considered by computing the @xmath26 of the unimodular tensor fields .
we use the results obtained by @xcite in truncating the 10 dimensional iib theory to a 5 dimensional einstein - maxwell theory .
the kaluza - klein reduction ansatz is given by @xmath30 where @xmath31 where @xmath12 is a symmetric 6@xmath106 unimodular tensor used to represent the 20 scalars in the 20 representation of @xmath4 .
this also represents the l @xmath32 deformation of the 5-sphere .
the @xmath33 are the 1-form potentials , antisymmetric in @xmath34 and @xmath35 , that represent the 15 @xmath4 yang - mills gauge fields .
the radius of the five compact dimensions is given by the inverse of @xmath36 .
also just to define some notation , here are a few important relations @xmath37 where @xmath38 is the usual covariant derivative .
the first two equations are in form notation while the third is in terms of tensors . + the truncated lagrangian has the form @xmath39 where @xmath40 is a potential given by @xmath41
also we have omitted chern - simons terms that will not contribute to the physical system of interest to us .
the unimodular tensor field equation that one obtains from this lagrangian is @xmath42 \label{tmotion}\end{aligned}\ ] ] the solutions we are looking for are those that interpolate between ads@xmath5 at high energies and ads@xmath0@xmath10t@xmath11 and low energies .
we are also interested in incorporating a magnetic field that is tangent to the boundary directions .
we also assume lorentz invariance in the boundary spatial directions orthogonal to the magnetic field .
the metric ansatz then takes the form @xmath43 we are interested in a magnetic field that is a linear combination of three different @xmath3 s taken out from the group @xmath4 .
we can obtain this via the following choice of @xmath12 and the maxwell stress tensor @xmath44 @xmath45 where @xmath46 and the unimodular condition becomes @xmath47 . with our choice of @xmath12 and maxwell tensor @xmath44
, we go ahead to solve the tensor field equations ( [ tmotion ] ) .
we do this with the assumption that @xmath12 is a constant in the near horizon regime . due to complexity of solving for the @xmath48 we instead solve for @xmath49 to get the relations @xmath50 plus the cyclic permutations in 1 , 2 , and 3 .
next , the langrangian for the truncated theory becomes @xmath51 where the @xmath52 is now @xmath53 implementing our ansatz for the metric , we obtain the following einstein s equations @xmath54 where @xmath55 we now look for the required solutions . for @xmath56 , we have the solution representing ads@xmath5 given by the usual metric @xmath57 this is the expected uv solution , located at @xmath58 , of our theory . in this region
the magnetic field becomes negligible and can be ignored .
the next case is with @xmath59 ; this will represent the field theory at longer length scales , and thus is expected to be located at finite @xmath25 .
the field equations produce the following solution at the ir @xmath60 this ir solution represents the product between a btz black hole and a torus , t@xmath11 .
we see that at zero temperature we obtain ads@xmath61@xmath10t@xmath11 .
the entropy of the btz black hole is given by @xmath62 , where @xmath15 is temperature , @xmath63 is the length of the @xmath64 direction , and @xmath65 is the central charge .
we obtain the central charge using the brown - henneaux formula @xcite , @xmath66 , where @xmath67 is the ads@xmath0 radius given by @xmath68 , and @xmath69 is the 3 dimensional gravitational constant . we can write @xmath69 as @xmath70 , where @xmath71 is the volume of the two compact dimensions @xmath72 . simplifying the expression further , we write @xmath73 in terms of relevant string theory constants to get @xmath74 where @xmath75 unit volume of the @xmath76 dimensions @xmath77 , and @xmath78 comes from defining the integral of the 5-form as @xmath79 , and @xmath80 is the number of branes , which is given by @xmath81 we note that @xmath82 is a function of the charges , and thus a function of the elements of the unimodular tensor , @xmath12 . now that we have the central charge on the gravity side , we proceed to obtain the central charge across the duality and compare .
we go on now to calculate the entropy of the free field limit of @xmath17 sym theory in an external magnetic field .
we consider the @xmath17 theory in @xmath84 terms , with a @xmath3 r - symmetry . in this theory
, we have @xmath85 complex scalars of each charge , @xmath86 , @xmath87 , and @xmath88 .
these are the scalars in the chiral multiplet .
we also have @xmath85 vector fields of charge 0 , these are the gauge fields . from the charges of the complex scalars we obtain the charges of the fermions of the chiral multiplet and the gauginos to be @xmath89
so we have @xmath85 weyl spinors of each charge @xmath90 , @xmath91 , and @xmath92 .
these are the fermions of the chiral multiplet . and finally we have @xmath85 weyl spinors of charge @xmath93 , these are the gauginos . + next we write down the logarithm of the partition function for each type of field . for the details of obtaining these we refer the reader to @xcite .
the expressions are @xmath94 @xmath10t@xmath11 theory and the @xmath17 sym free field theory . on the right
we have the ratio as a function of @xmath95 . on the left we have a slice through @xmath96 .
[ cc],width=264 ] @xmath10t@xmath11 theory and the @xmath17 sym free field theory . on the right
we have the ratio as a function of @xmath95 . on the left we have a slice through @xmath96 .
[ cc],width=264 ] we are interested in calculating the entropy in the extremely low temperature limit .
we use the standard formula , @xmath97 . in the required limit , the only contribution we have is the fermionic term with @xmath98 which gives @xmath99 we compare the central charges of the two theories across the duality .
we consider the ratio @xmath100 . simplifying this , we find it to be a function of any two of the set ( @xmath101 , @xmath102 , @xmath103 ) .
figure ( [ cc ] ) shows a plot of @xmath100 as a function of @xmath101 and @xmath102 , along with a slice through the @xmath96 plane .
the result of @xcite is reproduced at the point @xmath104 , which corresponds to the peak of @xmath100 where @xmath105 .
we also note the wide range of values of @xmath100 as we vary the linear combination of our gauge fields in group space .
in this section we search for a different type of solution motivated by finding a field theory dual that lives in one dimension .
we thus look for a solution to the truncated iib theory that interpolates between ads@xmath5 at large @xmath25 and ads@xmath106t@xmath24 at small @xmath25 .
the ansatz we choose has both translational and rotational symmetry in the toroidal directions .
this is manifest in the following metric @xmath107 the magnetic field we consider in this section is a linear combination of fields that point in different directions in actual space ; one along each toroidal direction .
also , each magnetic field will be chosen to be part of a different @xmath3 as before .
while the unimodular tensor , @xmath12 , remains the same , the maxwell tensor will then be chosen as @xmath108 where we set the magnetic fields to point in orthogonal directions , as shown below @xmath109 upon solving the tensor field equations for @xmath49 we find @xmath110 plus the cyclic permutation in 1 , 2 , and 3 .
the lagrangian is slightly modified from the previous case @xmath111 where @xmath52 is given in eq .
( [ potential ] ) .
varying the action we obtain the following einstein s field equations @xmath112 with @xmath113 these field equations , along with the scalar field equation result for @xmath114 , [ ads2tmotion ] , constrain our choice of the magnitudes of the magnetic fields , charges , and the @xmath12 entries .
we are forced to set @xmath115 the field equations then take a simpler form and can be solved exactly . at the uv limit
we impose @xmath56 .
the result , as expected , is the usual ads@xmath5 solution given by @xmath116 next , we turn on the magnetic field and solve the field equations exactly for @xmath117 and @xmath118 .
we note that there are two types of solutions to the einstein s field equation ; one where @xmath119 is taken to be a constant , and the other is where @xmath118 varies with @xmath25 .
we expect that the constant @xmath120 solution be a limiting case of the varying solution .
we choose to place the constant value at @xmath121 which will be the location of the horizon of the black hole . solving for @xmath117
we find that it does indeed have a solution that vanishes at an @xmath121 giving us a horizon .
we also find that when we place the @xmath120 at @xmath121 the solution of the black hole becomes an extremal one .
it s metric is given by @xmath122 which is the product between a two - dimensional black hole and a three torus , t@xmath24 as required .
calculating the temperature of this black hole we find it to be @xmath123 as expected of a black hole at extremality .
the interesting aspect of this black hole is that we can still go ahead and calculate an entropy at zero temperature .
we do so by using the standard formula , @xmath124 , with the area , @xmath125 , being @xmath126 as for the gravitational constant , we have the usual ads / cft relation , @xmath127
. we note that the relations used here are much simpler than those used in in section 2 because of the simpler form of @xmath128 .
this means that the @xmath129 is not deformed which simplifies many of the equations in the truncation process .
finally we compute the entropy to be @xmath130 we emphasize again that this entropy is finite at zero temperature .
this implies that our gravity theory is dual to a field theory with a degenerate ground state at large coupling .
we proceed now to evaluate the entropy of the free field limit of the dual theory .
the particle content here is identical to the one in section 3 .
the difference comes in the value of the assigned charges .
the complex scalars each are assigned the charge @xmath131 .
three of the weyl fermions are assigned the charge @xmath132 and the gaugino is assigned the charge @xmath133 .
the claim now is that not much has changed from the previous case in section 3 .
since the scalars transform under the @xmath134 of @xmath4 each complex scalar component will be charged under only one gauge field , thus its energy eigenstates will not change and the partition function will not be altered . since the gauge field is not charged , we expect that it also will attain the same partition function . as for the weyl fermions
, they transform as a @xmath135 of @xmath4 .
thus each component will be charged under a linear combination of the different @xmath3 s and hence of the different magnetic fields pointing in the different directions .
however , we will show with an example that this does not alter the answer except by a change in the magnitude of the magnetic field it feels .
here is a simple heuristic argument : since the fermions are charged equally under all three fields , then it will feel a resultant field equal to the square root of the sum of each magnetic field squared , @xmath136 . as for the direction of @xmath137 ,
this depends on the details of @xmath3 linear combination for each fermion .
+ here is the example : consider a fermion , @xmath138 , charged under the three @xmath3 s in the following way @xmath139 where @xmath140 the covariant derivative then becomes @xmath141 under a simple change of basis , we can write this in a much more transparent form @xmath142 thus in this basis , the fermion , @xmath143 , is charged under the gauge field @xmath144 .
the conclusion is then that the entropy for this system will have the same behavior as was found previously ; it will vanish as @xmath145 .
explicitly , the entropy will be @xmath146 we remind ourselves with the result of the previous section where we found the entropy to be finite at zero temperature .
thus we have found a conflict in the behaviors of the systems across the duality .
the weakly coupled picture suggests a new mechanism for local quantum criticality ( a. almuhairi and j. polchinski , work in progress ) .
a charge in a magnetic field can only move parallel to field lines . for a composite operator
whose constituents move in different directions , the correlator will be local in position space , and exhibit critical behavior only in time .
in this section we investigate the stability properties of the two space - times ads@xmath28t@xmath11 and ads@xmath29t@xmath24 with respect to perturbations of the unimodular tensor field .
we consider the perturbation as @xmath147 , where @xmath148 = 0 $ ] because @xmath149 = 1 $ ] . as usual
, we are looking for the field equations of the perturbations from which we can extract the @xmath26 ; we use the @xmath12 field equations given in above in eq .
[ tmotion ] . also across the duality
, we expect @xmath12 to correspond to a dimension 2 operator , and thus its @xmath150 in the case of no magnetic field .
this is already saturating the bf - bound .
thus we hope that the magnetic field stabilizes this and does not tip the mass in the wrong direction .
the linearized matrix equation of motion is @xmath151_{i j } + g^2 [ a_\mu , [ a^\mu , t]]_{i j } = -4g^2 t_{i j } - \frac{1}{2}(t f^2)_{i j } - \frac{1}{2}(f_{\mu \nu}tf^{\mu \nu})_{i j}\ ] ] these set of equations become diagonal when we construct the linear combination of @xmath152 s : @xmath153 and @xmath154 , where @xmath155 . due to the traceless property of @xmath152 , the solutions for @xmath156 mix together and we are forced to pick the combinations @xmath157 , along with the other two cyclic terms . carrying out the algebra one finds the following @xmath26 . for ads@xmath0 @xmath158 for ads@xmath1 @xmath159 , \
for\ all \ i\\ & m^2_{z_1 \bar{z_1 } + z_2 \bar{z_2 } - 2 z_3 \bar{z_3 } } = 12 g^2 , \ also \ for \ cyclic \ in \ ( 1,2,3)\end{aligned}\ ] ] where @xmath160 designates the landau level excitation .
we find that all the @xmath26 s are positive and thus safely above the @xmath161 for both space - times .
hence the magnetic field acts as a stabilizing mechanism in this problem .
in this paper we have generalized solutions pertaining to magnetic branes in ads .
we have shown that a more general configuration of magnetic fields gives us control over tuning the ratio between the entropies across the duality in the ads@xmath0 case .
there are solutions that can fix the entropies to be identical .
this manipulation shows how dependent the physics at large coupling is on the configuration of the magnetic field .
+ we also investigated the ads@xmath1 case , where we found that the black hole solution was an extremal one .
this gave us a finite entropy at zero temperature .
upon comparing this with the field theory , we observe a conflict , since the field theory entropy scales with temperature .
this indicates the presence of a phase transition when turning up the coupling . this is a problem worth tackling .
+ finally , we studied the stability properties of both space - times and showed that they are stable .
in fact we find that the usual dimension two tachyon operator gets a boost in its @xmath26 due to the presence of the magnetic field .
thus the magnetic field acts as a stabilizing agent .
j. m. maldacena , `` the large n limit of superconformal field theories and supergravity , '' adv .
* 2 * , 231 ( 1998 ) [ int .
j. theor .
phys . * 38 * , 1113 ( 1999 ) ] [ arxiv : hep - th/9711200 ] .
s. a. hartnoll and p. kovtun , `` hall conductivity from dyonic black holes , '' phys .
d * 76 * , 066001 ( 2007 ) [ arxiv:0704.1160 [ hep - th ] ] .
s. a. hartnoll , p. k. kovtun , m. muller and s. sachdev , `` theory of the nernst effect near quantum phase transitions in condensed matter , and in dyonic black holes , '' phys .
b * 76 * , 144502 ( 2007 ) [ arxiv:0706.3215 [ cond-mat.str-el ] ] .
e. dhoker and p. kraus , `` magnetic brane solutions in ads , '' jhep * 0910 * , 088 ( 2009 ) [ arxiv:0908.3875 [ hep - th ] ] .
d. t. son and p. surowka , `` hydrodynamics with triangle anomalies , '' phys . rev . lett . *
103 * , 191601 ( 2009 ) [ arxiv:0906.5044 [ hep - th ] ] .
a. strominger , jhep * 9901 * , 007 ( 1999 ) [ arxiv : hep - th/9809027 ]
. t. faulkner , h. liu , j. mcgreevy and d. vegh , `` emergent quantum criticality , fermi surfaces , and ads2 , '' arxiv:0907.2694 [ hep - th ] .
t. faulkner and j. polchinski , `` semi - holographic fermi liquids , '' arxiv:1001.5049 [ hep - th ] .
m. cvetic , h. lu , c. n. pope , a. sadrzadeh and t. a. tran , `` consistent so(6 ) reduction of type iib supergravity on s(5 ) , '' nucl . phys .
b * 586 * , 275 ( 2000 ) [ arxiv : hep - th/0003103 ] .
j. d. brown and m. henneaux , `` central charges in the canonical realization of asymptotic symmetries : an example from three - dimensional gravity , '' commun .
phys . * 104 * , 207 ( 1986 ) . | we investigate ads@xmath0 and ads@xmath1 magnetic brane solutions within the consistent truncation of ads@xmath2 supergravity .
the ads@xmath0 solution extends earlier work by allowing a general embedding of the magnetic @xmath3 in @xmath4 .
we determine the ratio of strong- and weak - coupling entropies as a function of this embedding .
further , by considering crossed magnetic fields in different @xmath3 s we are able to construct a solution that runs from ads@xmath5 in the uv to ads@xmath6 in the ir .
we find the notable result that there is a zero temperature entropy at strong coupling but not at weak coupling .
we also show that the ads@xmath1 solution and at least some of the ads@xmath0 solutions are stable within the truncation . * * * ads@xmath0 and ads@xmath1 magnetic brane solutions * * ahmed almuhairi * _ department of physics , university of california , + santa barbara , california 93106 , usa _ |
the asteroid belt is a window open on the past of the solar system , as it contains records of both its formation process and its secular evolution .
however , disentangling the sequence of events that characterized its history is a difficult task , since the more ancient features have been erased or masked by the more recent ones ( see e.g. @xcite and @xcite for an in - depth discussion ) .
we know from meteoritic constrains ( see e.g. @xcite for a review ) and theoretical models ( see e.g. @xcite ) that the progenitors of the present - day asteroids formed and , in some cases , differentiated in the solar nebula on a @xmath3 ma time - scale , soon followed by jupiter and the other giant planets ( see e.g. @xcite ) . however , the actual scenario for the formation of the planetesimals , its efficiency and the resulting initial size - frequency distribution are still poorly constrained .
the proposed formation scenarios differ in the assumptions on the formation environment , i.e. a quiescent or a turbulent solar nebula , and produce very diverse size - frequency distributions ( sfds in the following ) of the primordial planetesimals ( see e.g. @xcite and @xcite and references therein ) .
the number of planetesimals populating the asteroid belt at this early time was likely @xmath4 orders of magnitude higher than the present population of asteroids @xcite , so the collisional evolution of the primordial asteroid belt was characterized by a more rapid pace than the present one @xcite .
according to recent theoretical results @xcite , in a few ma the size distribution of the planetesimals in the asteroid belt spanned over @xmath5 orders of magnitude , covering the range between @xmath6 km and @xmath7 km independently on the formation mechanism .
+ the present view of the evolution of the asteroid belt assumes that it underwent a major phase of depletion and enhanced collisional evolution after the dispersal of the nebular gas due to the combined perturbations of planetary embryos in the inner solar system and of the giant planets in the outer solar system @xcite .
the later migration of the giant planets , suggested to be responsible for the late heavy bombardment , would have caused a second phase of depletion and possibly of enhanced collisional evolution of the asteroid belt @xcite .
the violent depletion processes acting in the early solar system left then place to slower , secular depletion mechanisms .
chaotic diffusion into the resonances with the giant planets has been suggested to have caused a depletion of a factor 2 of the population of large ( d @xmath8 km ) asteroids @xcite .
previous studies , however , always assumed the formation time of the giant planets as negligible and the giant planets ( generally jupiter and saturn ) were introduced instantaneously and all at the same time @xcite.in our previous work ( @xcite , hereafter paper i ) , we suggested instead that jupiter could have been the first giant planet to form and that its formation triggered a phase of primordial bombardment due to its rapid mass increase and its likely inward radial migration . this jovian early bombardment ( jeb in the following ) preceded the depletion phases discussed before and its duration was estimated of the order of @xmath9 ma ( see paper i ) , but it was suggested @xcite that the later formation of saturn could act to extend the duration of the jeb .
after the end of the jeb and the formation of saturn , the solar system and the asteroid belt are expected to resume the evolutionary path described by the generally accepted dynamical and collisional scenarios ( see @xcite for the dynamical scenario and @xcite for the unified dynamical and collisional scenario ) .
the effects of the jeb , which in paper i we investigated using vesta and ceres as case studies in view of the arrival of the nasa dawn mission at vesta , depend on the extent and the time - scale of the jovian migration but also on the sfd of the primordial planetesimals . in paper
i we suggested that a solar nebula whose population of planetesimals was dominated by large ( i.e. 100 - 1000 km in diameter ) planetesimals or where jupiter s radial migration exceeded a few tenths of au would prove an extremely hostile environment for the survival of vesta and ceres .
it must be noted , however , that the assumptions on the erosion of vesta and ceres due to the jeb did not account for the re - accretion of the excavated material and thus overestimated the implications of the bombardment for the survival of the two target bodies . in this paper
we address the implications of the jeb for the global evolution of the asteroid belt using a more complete set of target bodies and a more realistic evaluation of the cumulative collisional erosion .
in our simulations we reproduced the evolution of a template of the forming solar system across a temporal window of @xmath10 ma , located during the solar nebula phase ( see e.g. @xcite and references therein ) and centred on the moment the planetary core of jupiter reached its critical mass and the planet started to accrete its gaseous envelope .
our template of the solar system is composed by the sun , the forming jupiter , a disk of planetesimals and a set of target bodies whose perturbations on the nearby planetesimals are considered negligible . + the numerical model we used in our simulations is based on the set of equations we described in paper i :
as we anticipated , in this paper we improved the model by introducing a more realistic estimate of the cumulative erosive effects of the jeb . in the following
we will summarise the set of equations and the methods we used to evaluate the intensity of the jeb : for the full derivation of the equations we refer the interested readers to paper i. in place of vesta and ceres we used as our case study in paper i , in this work we used two sets of synthetic bodies of different sizes as the targets of the jeb , to obtain a clearer picture of its effects on the planetesimals populating the asteroid belt .
+ all members of each set were characterised by the same orbit : the first set of bodies was located on circular , planar orbits at @xmath11 au from the sun ( i.e. midway between the inner edge of the asteroid belt and the @xmath12 resonance with jupiter , region a in the following ) while the second set was located on circular , planar orbits at @xmath13 au from the sun ( i.e. midway between the @xmath12 and the @xmath14 resonances with jupiter , region b in the following ) .
+ each set was composed by five different target bodies , characterised respectively by a diameter of @xmath0 km , @xmath1 km , @xmath2 km , @xmath15 km and @xmath16 km .
an average density @xmath17 was assumed for all target bodies in order to evaluate their masses and escape velocities from the surface .
the evolution of jupiter during our simulations is divided into two phases : the growth of the planetary core to its critical mass and the accretion of the nebular gas to form a massive envelope .
the phase of gas accretion can also be divided into two sub - phases , a first one where the jovian mass increases exponentially and a second one where the jovian mass asymptotically approaches its final value ( see e.g. @xcite ) .
+ at the beginning of our simulations , jupiter is a planetary embryo with mass @xmath18 that grows to the critical mass @xmath19 in @xmath20 years .
since the total accretion time of jupiter s core is the sum of our @xmath21 to the time needed to form the initial mars - sized planetary embryo , our choice of @xmath21 is consistent with the lower limits indicated by theoretical works for the formation of jupiter s core ( a few ma , see e.g. @xcite and references therein ) . during this first phase , the mass growth of jupiter is governed by the equation @xmath22 upon the reaching of the critical mass value of @xmath23 , the nebular gas surrounding jupiter is assumed to become gravitationally unstable and to be rapidly accreted by the planet to form its massive envelope . during this phase
, the mass growth of jupiter is instead governed by the equation @xmath24 where @xmath25 is the final mass of the giant planet .
the e - folding time @xmath26 years is derived from the hydrodynamical simulations described in @xcite , @xcite and @xcite .
+ as we anticipated , we followed the evolution of our template of the solar system across the gas accretion phase for @xmath27 years , i.e. for @xmath28 . + while jupiter starts on a circular orbit in all our simulations , theoretical models indicate that forming giant planets should undergo type i and ii migrations and drift inward due to their interactions with the protoplanetary disk ( see e.g. @xcite and references therein ) . the time - scale @xmath29 of migration is a non - linear function of the mass and the heliocentric distance of the forming planet : for a planet at @xmath30 au , @xmath29 varies between @xmath31 years for type i migration and is of the order of @xmath32 years for type ii migration @xcite . as an approximation
, we ignored the distinction between type i and ii migrations and started the migration of jupiter as soon as its mass reaches @xmath23 , which is equivalent to say that the value of @xmath29 becomes of the order of @xmath32 years ( ibid ) .
as a consequence , jupiter moves on a circular orbit while its planetary core is growing to the critical mass and starts to spiral inward once the phase of gas accretion begins .
the equation governing the migration of jupiter is functionally similar to eq .
[ gasgrowth ] we used to describe the growth of the gaseous envelope : @xmath33 where @xmath34 is the orbital radius of jupiter at the beginning of the simulation , @xmath35 is the final radius and @xmath36 years .
we assumed jupiter s final semi - major axis equal to the present one , an assumption consistent with both the standard model of planetary formation and the scenario described by the first formulation of the nice model @xcite . in our simulations we considered four different migration scenarios : @xmath37 au ( no displacement ) , @xmath38 au , @xmath39 au and @xmath3 au . since the parameter @xmath40 we used in eq .
[ radiallaw ] is an e - folding time , displacements of @xmath38 au , @xmath39 au and @xmath3 au with the assumed value of @xmath40 are equivalent to assuming values of @xmath29 at @xmath30 au respectively of @xmath41 years , @xmath42 years and @xmath43 years , consistently with the results of theoretical studies @xcite .
+ the disk of planetesimals populating our template of the forming solar system is simulated through a swarm of @xmath44 massless particles .
the disk of planetesimals extend between @xmath10 au and @xmath45 au ( @xmath46 ) and is assumed to be dynamically cold , i.e. the orbits of the planetesimals are characterised by low eccentricity ( @xmath47 ) and inclination ( @xmath48 ) values .
the values of eccentricity and inclination associated to each massless particle were chosen randomly as @xmath49 where @xmath50 , @xmath51 rad and @xmath52 is a number extracted from a uniform distribution in the range @xmath53 $ ] .
+ the dynamical evolution of the disk of massless particles was computed through a fourth order runge - kutta integrator with a self adjusting time - step .
the time - step is chosen by evaluating at each given time the smallest time - scale @xmath54 between the orbital periods of jupiter , of the target bodies and of the massless particles and the free - fall time of jupiter - particle pairs considered as isolated systems .
the time - step is then computed as @xmath55 where @xmath56 in our simulations .
+ while from a dynamical point of view the planetesimals populating the disk are treated as massless particles , we associated to each of them mass and density values in order to model the effects of their impacts on the target bodies .
the mass values were obtained through the monte carlo methods we will describe in sect .
[ sfds ] , while for the density values we considered two compositional classes .
planetesimals formed in the inner solar system ( * iss * in the following ) were considered rocky bodies with mean density @xmath57 , while planetesimals formed in the outer solar system ( * oss * in the following ) were considered volatile - rich bodies with mean density @xmath58 .
the change between the inner and the outer solar system is assumed to coincide with the location of the snow line . when not explicitly stated otherwise , the snow line was assumed to be at @xmath59 ( see e.g. @xcite and references therein ) .
+ finally , we treated each massless particle as a swarm of real planetesimals by associating to each of them a normalization factor @xmath60 , described in the sect .
[ sfds ] and used to estimated a realistic number of impacts during the jeb basing on the method detailed in sect .
[ impacts ] .
we studied the effects of the jeb in disks characterised by different sfds , which in turn link the jeb scenario to the different scenarios proposed for the formation and early collisional evolution of the primordial planetesimals .
we used these sfds to associated mass values to the test particles through the monte carlo approaches we will now describe .
the first sfd we considered is that of a disk of planetesimals formed by gravitational instability of the dust in the mid - plane of a non turbulent protoplanetary nebula @xcite .
the protoplanetary nebula is assumed to have a mass of @xmath61 m@xmath62 distributed between @xmath63 au with dust - to - gas ratio @xmath64 and density profile @xmath65 , where @xmath66 g @xmath67 is the surface density at @xmath3 au and @xmath68 .
for such a nebula it can be showed @xcite that the average diameters of the planetesimals would follow the semi - empirical relationship @xmath69 where @xmath70 and @xmath71 are expressed in @xmath72 , @xmath73 is expressed in @xmath74 and @xmath75 .
the value @xmath71 is the average mass of a planetesimal at @xmath76 , i.e. @xmath77 g @xcite .
if we assume that the mass dispersion of the planetesimals about the average values of eq .
[ masslaw ] is governed by a maxwell - boltzmann distribution , in paper i we showed that we can associate a mass value to each test particle by means of a monte carlo method where the uniform random variable @xmath78 varying in the range @xmath79 $ ] is @xmath80 where @xmath81 is the lower incomplete gamma ratio .
the inverse of the lower incomplete gamma ratio can be computed numerically and , by substituting @xmath82 back with @xmath83 we obtain @xmath84 since the use of massless particles assures the linearity of the processes investigated over the number of considered bodies , we can extrapolate the number of impacts expected in a disk of planetesimals by multiplying the number of impacts recorded in our simulations by a factor @xmath60 where @xmath85 where @xmath86 and @xmath87 is given by @xmath88 where @xmath89 au , @xmath90 au and the symbol @xmath76 indicate the value of the astronomical unit expressed in cm , i.e. @xmath91 cm .
+ the second sfd we considered is that of planetesimals formed by concentration of dust particles in low vorticity regions in a turbulent protoplanetary nebula @xcite .
following @xcite , we assumed the protoplanetary nebula as characterized by a surface density @xmath92 g @xmath67 at @xmath3 au , a nebula density profile with exponent @xmath93 , a snow line placed at @xmath94 au ( see paper i ) and a dust - to - gas ratio @xmath95 beyond the snow line and @xmath96 inside the snow line ( see fig .
@xmath97 , gray dot - dashed line , ibid ) .
the results of @xcite supply the average diameter of planetesimals as a function of heliocentric distance ( see fig . @xmath97 , gray dot - dashed line , ibid ) , from which we derived the following semi - empirical relationship analogous to eq .
[ masslaw ] @xmath98 where @xmath99 , @xmath100 km is the average diameter of the planetesimals at @xmath3 au . by substituting the primed quantities to the original ones in eqs .
[ massval ] and [ ntot ] , we can thus obtain the mass and the normalisation factor for each massless particle through the same approach we described previously .
+ the third and the fourth sfds we considered are derived by the results of @xcite .
@xcite did not explore a specific model of planetesimal formation in quiescent or turbulent disks but instead tried to constrain the initial size - frequency distribution of planetesimals in the orbital region of the main asteroid belt .
their results suggest that the best match with the present - day sfd of the main asteroid belt is obtained for planetesimal sizes initially spanning @xmath101 km ( see fig .
@xmath45 , ibid ) , a range consistent with their formation in a turbulent nebula .
@xcite supplies two sfds associated to this case : a first one describing the primordial sfd of the planetesimals , which spans @xmath101 km ( see fig .
@xmath102 , black dots , ibid ) , and a second , collisionally evolved one where accretion and break - up of the primordial planetesimals extended the size distribution between @xmath103 km ( see fig .
@xmath102 , black solid line , ibid ) . for each iss impact event in our simulations
, we then estimated the mass of the impacting planetesimal through a simple monte carlo extraction based on the cumulative probability distributions of the two sfds supplied by @xcite .
the normalization factor is estimated through eq .
[ ratio ] using the total number of planetesimals in the asteroid belt supplied by the cumulative sfds from fig .
@xmath102 in @xcite . in using the sfds from @xcite we did not considered the contribution of oss impactors , since the authors only investigated the primordial sfd of planetesimals in the orbital region of the asteroid belt . to reproduce the collisional histories of the target bodies we opted for a statistical approach based on solving the ray
torus intersection problem between the orbital torus of a target body and the linearised path of a massless particle across a time step ( see paper i ) .
our method is similar to the analytical method developed by @xcite , but given that the jovian perturbations may significantly change the orbits of the massless particles on time - scales analogous to their precession time - scales , we preferred the use of a numerical algorithm that did not require averaging over orbital angles other than the mean anomaly .
+ the orbital torus representing the spatial probability density of a target body is characterised by a mean radius @xmath104 and a section @xmath105 where @xmath106 is the semi - major axis of the considered target body , @xmath107 is its physical diameter and @xmath108 is the gravitational focusing factor @xmath109 with @xmath110 being the escape velocity from the target body and @xmath111 its relative velocity respect to the massless particle .
+ when a massless particle crosses a torus , the impact probability is the probability that both the particle and the target body will occupy the same spatial region at the same time .
this is equivalent to writing @xmath112 where @xmath113 is the orbital period of the target body and @xmath114 and @xmath115 are respectively the time spent by the target body and the massless particle into the crossed region of the torus .
once the crossed region is identified by solving the ray
torus intersection problem ( see paper i ) , @xmath114 and @xmath115 can be derived in a straightforward way since the orbital velocities of the two bodies are known .
+ for the sizes and the densities we are considering , the escape velocity from the surface of the target bodies would be of the same order of magnitude of the ejection velocity of the excavated material . as a consequence , re - accumulation of the excavated material should be taken into account and the simple cratering approximation used in paper i is not adequate to estimate the net erosion of the target bodies .
+ to evaluate the effective erosion cumulatively caused by the impacts , we need to estimate the amount of material excavated and ejected with velocities greater than the escape velocity from the surface of the target bodies . to estimate the net mass loss , we used the scaling law for rocky targets by @xcite : @xmath116 where @xmath117 is the escaped mass
, @xmath118 is the mass of the impactor , @xmath119 is the impact velocity , @xmath120 is the impact angle relative to the normal to the target surface , @xmath121 is the escape velocity from the surface of the target body and @xmath122 and @xmath123 are the densities respectively of the impactor and the target body .
+ since we are treating the collisional evolution of the target bodies in a statistical way , we are interested in the average erosion efficiency : therefore , we need to remove the dependency from the impact angle in eq .
[ h&hlaw ] . following @xcite
, we can average eq .
[ h&hlaw ] over @xmath124 to obtain @xmath125 to get a more complete picture of the implications of the jeb for the collisional evolution of the asteroid belt , for every impact event we also computed the specific impact energy @xmath126 expressed in units of the specific dispersion energy @xmath127 of the target body .
we evaluated the catastrophic disruption threshold @xmath127 of the target bodies using eq .
@xmath128 from @xcite and the coefficients for basaltic targets computed by these authors ( see table @xmath129 , ibid ) .
we used the coefficients of the case @xmath130 from @xcite for all impact events with a velocity greater or equal than @xmath131 and those of the @xmath132 for all the other impact events .
for those impacts where @xmath133 , we substituted eq .
[ svelaw ] with eq .
@xmath45 from @xcite expressed in terms of the eroded mass : @xmath134 where @xmath135 for @xmath136 and @xmath137 for @xmath138 .
the effects of impacts with @xmath139 are not accounted for in our evaluation of the eroded mass .
+ we used the number of cumulative impacts with @xmath140 ( high energy impacts ) and @xmath139 ( critical energy impacts ) together with the cumulative eroded mass estimated with eqs . [ svelaw ] and [ benzlaw ] as the main parameters for our analysis .
on the time - scales considered in our simulations , the flux of impactors associated to the jeb is dominated by the iss planetesimals , in particular those affected by the @xmath12 and the @xmath141 resonances with jupiter .
the flux of oss impactors is more erratic and , as we show in figs .
[ fig1 ] and [ fig2 ] , their contribution to the total flux is relevant mainly in the scenario where jupiter does not migrate . as we show in table [ fluxtable ] for target bodies with radius of @xmath0 km , @xmath1 km and @xmath2 km , the location of a body with respect to the @xmath12 and the @xmath141 resonances with jupiter and the extent of the jovian displacement are the key factors in determining the intensity of the jeb for a given sfd of the planetesimals .
+ when jupiter does not migrate the target bodies located in region b receive a flux of impactors that is @xmath142 than that received by target bodies in region a. when jupiter migrates inward by @xmath38 or @xmath39 au , target bodies located in region b receive a flux that is a factor @xmath143 than that received by target bodies in region a. in the case of planetesimals formed in a turbulent disk and characterized by the sfd from @xcite , the ratio increases to about @xmath144 ( see table [ fluxtable ] ) .
when jupiter migrates by @xmath3 au , however , the sweeping resonances are more efficient in enhancing the bombardment in region a respect to the previous case . as a consequence ,
the flux of impactors in region b `` drops '' again to @xmath145 respect to that in region a. +
the implications of the phase of enhanced collisional activity associated to the jeb for the evolution of the asteroid belt strongly depend on the sfd of the planetesimals in the solar nebula at the time of the formation of jupiter .
globally , the jeb causes the destruction of planetesimals up to a certain threshold size , whose exact value is a function of the extent of jupiter s migration and of the abundance of large ( i.e. @xmath146 km ) planetesimals in the solar nebula . differently from what was generally assumed in previous studies of the collisional evolution of the asteroid belt @xcite , our results show that during the jeb the cumulative erosion due to non - critical impacts plays a role as important as , or possibly more important than , the one of critical impacts .
+ if the first planetesimals formed in a quiescent solar nebula and the protoplanetary disk was still governed by its primordial sfd at the time jupiter formed , the jeb would have important implications for the asteroids whose size varied between @xmath147 km ( see table [ quiescentdisk ] ) . according to our results , critical impacts
would cause the catastrophic disruption of planetesimals with size of about @xmath0 km in diameter in those scenarios where jupiter migrated significantly ( i.e. @xmath3 au for target bodies in region a and @xmath9 au for target bodies in region b ) .
moreover , due to the higher flux of oss impactors , target bodies of about @xmath0 km in diameter would likely be shattered also in those scenarios where jupiter did not migrate while forming ( see table [ quiescentdisk ] ) .
planetesimals of @xmath1 km in diameter or bigger would not be undergo catastrophic collisions .
+ the cumulative erosion due to normal impacts ( @xmath148 ) and high - energy impacts ( @xmath133 ) , however , would cause the disruption of those bodies between @xmath147 km in diameter both in region a and b in almost all migration scenarios ( see table [ quiescentdisk ] ) . even in the least erosive scenarios ( jupiter migrating by @xmath149 au and target bodies located in region a )
, planetesimals of @xmath1 km in diameter would lose between @xmath150 of their original mass due to collisional erosion .
collisional erosion would affect also planetesimals of about @xmath2 km in diameter in those scenarios where jupiter migrates significantly ( i.e. @xmath3 au for region a and @xmath9 au for region b ) , causing them to lose between @xmath151 of their original mass ( see table [ quiescentdisk ] ) .
for smaller displacements of the giant planet , planetesimals of about @xmath2 km in diameter would only lose a few per cent of their mass due to erosion .
+ planetary bodies ranging @xmath152 km would be mostly unaffected by the jeb in such a protoplanetary disk .
if planetesimal formation in the solar nebula was instead driven by turbulence and the protoplanetary disk was still characterized by its primordial sfd when jupiter formed , critical impacts would dominate the collisional evolution of bodies as big as @xmath1 km in diameter ( see table [ turbulentdisk ] ) , which would be destroyed both in region a and b. + planetary bodies of @xmath2 km in diameter or larger , however , would not undergo to critical impacts and their collisional evolution would be dominated by erosion .
planetesimals of @xmath2 km in diameter would be completely eroded both in region a and b in almost all scenarios . even in the only case where these planetesimals could survive the jeb ( jupiter migrating by @xmath38 au and target bodies located in region a )
, they would be stripped by about @xmath153 of their original mass . also planetary bodies of @xmath15 km in diameter located either in region
a or region b would be completely eroded in those scenarios where jupiter does not migrate or migrates by @xmath3 au . in the other scenarios ,
the same bodies would be stripped of about @xmath154 of their original mass in region a while in region b the mass loss would rise to @xmath155 ( see table [ turbulentdisk ] ) .
+ finally , planetary embryos of about @xmath16 km in diameter would lose only a few per cent of their original mass if the migration of jupiter was limited ( i.e. less than @xmath3 au for region a and less than @xmath39 au for region b ) but their mass loss would rise to @xmath156 for higher values of the jovian displacement ( see table [ turbulentdisk ] ) . in the most plausible scenario , i.e.
a solar nebula populated by collisionally evolved planetesimals , the survival of the primordial planetesimals would be in an intermediate position between the previous two cases , as is shown in table [ evolveddisk ] .
+ planetesimals of about @xmath0 km in diameter would lose @xmath157 of their original mass both in region a and region b due to collisional erosion in all the considered migration scenarios .
also planetesimals of @xmath1 km in diameter would be completely eroded by the jeb for high values of the jovian migration ( i.e. @xmath3 au for region a and @xmath9 au for region b ) . in almost all the other scenarios ,
moreover , these planetesimals would be stripped by a significant fraction of their original mass due to collisional erosion ( @xmath158 in region a and about @xmath159 in region b , see table [ evolveddisk ] ) .
+ planetary bodies of @xmath2 km in diameter would be stripped by about @xmath160 of their original mass only in those scenarios where jupiter migrates significantly ( i.e. @xmath3 au for region a and @xmath9 au for region b ) .
planetesimals of @xmath152 km in diameter would instead survive the jeb without any significant mass loss . + finally , the scenario described by the primordial sfd by @xcite is shown in table [ primordialdisk ] and is qualitatively analogous to the case of the jeb in a turbulent disk shown in table [ turbulentdisk ] .
planetesimals of @xmath2 km in diameter or smaller are either catastrophically disrupted or eroded by the jeb .
planetesimals of @xmath15 km undergo a significant erosion ( @xmath161 of their original mass ) if the migration of jupiter was limited ( i.e. less than @xmath3 au for region a and less than @xmath39 au for region b ) while are collisionally eroded in the other cases .
similarly , planetary bodies of @xmath16 km in diameter suffer a limited erosion ( a few per cent , see table [ primordialdisk ] ) in most scenarios , but are stripped of @xmath162 of their original mass when jupiter migrates significantly ( i.e. @xmath3 au for region a and @xmath9 au for region b ) .
+ before proceeding further , we want to stress once again that this estimate of the effects of the jeb in the `` asteroids were born big '' scenario by @xcite takes into account only the contribution of iss planetesimals . by comparison with the cases of planetesimals formed in quiescent and turbulent disks and taking into account figs .
[ fig1 ] and [ fig2 ] , we can estimate that the inclusion of oss impactors would likely imply the collisional erosion of bodies of @xmath1 km ( collisionally evolved sfd ) and @xmath2 km ( primordial sfd ) also in those scenarios were jupiter does not migrate while forming .
in applying the results we just described to assess the evolution of the early solar system , one should be aware of the underlying assumptions and the present limitations of the model .
first , we assume that jupiter was the first giant planet to form : this is equivalent to assume that the formation of saturn took place a few @xmath163 years after that of jupiter .
second , we assume that the gravitational perturbations of jupiter are the dominant effect on the dynamical evolution of the planetesimals across the jeb : we ignored the effects of gas drag and of the planetary embryos already formed in the solar nebula ( see e.g. @xcite ) .
the choice of not including neither gas drag nor planetary embryos is mostly due to the limitation of the numerical implementation of our model . as we stated previously ,
we use a fourth - order runge - kutta algorithm with a self - adapting time - step to reproduce the dynamical evolution of the disk of planetesimals .
as we described in paper i , we split the simulations into sub - simulations , each of them focusing on the dynamical evolution of a @xmath3 au - wide ring of @xmath7 test particles .
each sub - simulation requires about @xmath3 month of computational time and we run a total of 48 simulations to assemble the dataset analysed in this work . the inclusion of gas drag , in particular , causes a significant increase in the impact events recorded in our simulations ( please note we are not referring to the normalized flux of impactors ) and the use of smaller time - steps due to the inward migration it forces on the planetesimals . with our choice of the numerical integrator , this translates in computational times in excess of @xmath10 months for some of the sub - simulations . we plan to explore the implications of the inclusion of planetary embryos and gas drag in future works , using a more detailed version of the model and a more efficient n - body algorithm .
+ finally , the exclusion of the gravitational perturbations due to the presence of planetary embryos in the solar nebula implies that we do not take into account the depletion process of the asteroid belt @xcite across the jeb . according to @xcite , however
, the presence of planetary embryos in the region of the asteroid belt does not cause a significant depletion of the planetesimals if jupiter and saturn are not present .
this implies that the mass depletion process should be active only during the last @xmath164 years of our simulations .
according to the results of @xcite for the case where jupiter and saturn are initially on circular orbits , the depletion of the asteroid belt on this temporal interval should amount to about @xmath165 of the original population of planetesimals . since the bulk of the jeb takes place in a few @xmath163 years
( see paper i ) , the depletion of the asteroid belt should not affect our results significantly .
the results we presented here have some important implications , since they add new pieces to the mosaic of our understanding of the history of the early solar system .
+ first , our results show that the formation of jupiter triggers a short but intense primordial bombardment across the asteroid belt .
the migration of jupiter can act to enhance its intensity but is not necessary to start the jovian early bombardment .
+ second , our results clearly highlight the fact that , due to the more abundant population of the asteroid belt at the time of the jovian early bombardment , cumulative erosion plays a more important role than that of critical impacts in determining the fate of the planetesimals .
such effects , not included in previous studies of the collisional history of the asteroid belt ( e.g. @xcite ) , could help to explain the long equivalent time - scale ( @xmath166 ga instead of the real @xmath167 ga , ibid ) needed to achieve the degree of collisional evolution of the present - day asteroid belt .
+ third , our results suggest that the generally accepted view that most asteroids of about @xmath0 km in diameter or larger are primordial may not be correct .
we showed that the exact threshold size depends on the considered region of the asteroid belt , on the extent of jupiter s migration , and on the size - frequency distribution of the planetesimals at the time of the jovian early bombardment .
we can generally state that , if the population of planetesimals in the solar nebula was dominated by objects smaller than @xmath0 km , the threshold size can be of the order of @xmath1 km . if , instead , the solar nebula was populated by planetesimals larger than @xmath0 km , the threshold size can rise up to about @xmath2 km .
planetesimals in the @xmath147 km size range are large enough to differentiate due to the energy released by the decay of short - lived radionuclides ( see e.g. @xcite and references therein ) : the largest planetesimals disrupted by the jovian early bombardment may therefore represent the parent bodies of the most ancient differentiated meteorites .
+ fourth , our results support the claim we made in paper i that the survival of vesta to the jovian early bombardment could be used to constrain the dynamical evolution of jupiter .
our improved estimates of the collisional effects of the jovian early bombardment indicate that , even in the least erosive scenarios , planetesimals of about @xmath2 km in diameter would be stripped of @xmath168 of their original mass if jupiter migrated by about @xmath3 au . therefore
, if vesta formed near its present orbit its survival to the jovian early bombardment suggests that jupiter s migration was less than @xmath3 au ( see paper i for a discussion ) .
+ fifth , as we discussed in @xcite our results suggest that the formation of saturn would trigger a second phase of bombardment , possibly less intense than the jovian early bombardment due to the lower mass of saturn and the depletion caused in the solar nebula by the formation of jupiter .
if the formation of saturn took place only a few @xmath163 years after that of jupiter , these two phases of bombardment would overlap into a single event , which in analogy to the late heavy bombardment we called the primordial heavy bombardment @xcite .
the enhanced collisional activity of the inner solar system caused by either the jovian early bombardment or the primordial heavy bombardment could prove extremely important to discriminate between the standard scenario for the evolution of the solar nebula we assumed in this work or alternative scenarios like the one recently proposed by @xcite .
+ finally , since the physical processes considered here are general to planetary systems harbouring forming giant planets , our results indicate that the formation of an extrasolar giant planet in a circumstellar disk would trigger a phase of bombardment whose duration ( i.e. about @xmath9 ma ) would be a significant fraction of the lifetime of the disk itself ( i.e. about @xmath166 ma , see e.g. @xcite ) .
the erosive effects of such an extrasolar primordial heavy bombardment on already formed planetesimals would cause a sustained production of dust characterized by a large grain size ( from tens of microns to cm , see e.g. @xcite ) .
if the amount , the physical properties or the location of the dust produced by giant planet - induced bombardments differ significantly from those of the dust due to the collisional cascade caused by the presence of planetary embryos as observed in debris disks ( see e.g. @xcite for a review ) , the enhancements in the dust population observed in circumstellar disks ( see e.g. @xcite for a review ) could in principle be used to identify those systems were the formation of extrasolar giant planets is taking place .
the exploration of this possibility is the subject of ongoing research and will be discussed in a forthcoming paper .
diego turrini and gianfranco magni wish to dedicate this work to their friend and colleague angioletta coradini , who passed away after a long illness on september 5th , 2011 .
the authors also wish to thank the reviewer , john chambers , for his comments on the manuscript and the results .
this research has been supported by the italian space agency ( asi ) through the asi - inaf contracts i/015/07/0 and i/010/10/0 .
the computational resources used in this research have been supplied by inaf - iaps through the project `` hpp - high performance planetology '' .
ccccc & & @xmath0 km target & & + @xmath169 au & @xmath170 & @xmath171 & @xmath172 & @xmath173 + @xmath38 au & @xmath174 & @xmath175 & @xmath176 & @xmath177 + @xmath178 au & @xmath179 & @xmath180 & @xmath181 & @xmath182 + @xmath183 au & @xmath184 & @xmath185 & @xmath186 & @xmath187 + & & @xmath1 km target & & + @xmath169 au & @xmath188 & @xmath189 & @xmath190 & @xmath191 + @xmath38 au & @xmath192 & @xmath193 & @xmath194 & @xmath195 + @xmath178 au & @xmath196 & @xmath197 & @xmath198 & @xmath199 + @xmath183 au & @xmath200 & @xmath201 & @xmath202 & @xmath203 + & & @xmath2 km target & & + @xmath169 au & @xmath204 & @xmath205 & @xmath206 & @xmath207 + @xmath38 au & @xmath208 & @xmath209 & @xmath210 & @xmath211 + @xmath178 au & @xmath212 & @xmath213 & @xmath214 & @xmath199 + @xmath183 au & @xmath215 & @xmath216 & @xmath217 & @xmath218 + ccccccccccc & & & & @xmath0 km & target & & & + @xmath169 au & @xmath219 & @xmath220 & @xmath221 & @xmath222 & @xmath223 & @xmath224 & @xmath225 & @xmath226 + @xmath38 au & @xmath227 & @xmath228 & @xmath229 & @xmath230 & @xmath231 & @xmath232 & @xmath233 & @xmath234 + @xmath178 au & @xmath235 & @xmath236 & @xmath229 & @xmath237 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + @xmath183 au & @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 & @xmath247 & @xmath248 & @xmath249 + & & & & @xmath1 km & target & & & + @xmath169 au & @xmath250 & @xmath251 & @xmath169 & @xmath252 & @xmath253 & @xmath254 & @xmath169 & @xmath183 + @xmath38 au & @xmath255 & @xmath256 & @xmath169 & @xmath38 & @xmath257 & @xmath258 & @xmath169 & @xmath259 + @xmath178 au & @xmath260 & @xmath261 & @xmath169 & @xmath262 & @xmath263 & @xmath264 & @xmath169 & @xmath265 + @xmath183 au & @xmath266 & @xmath267 & @xmath169 & @xmath268 & @xmath269 & @xmath270 & @xmath169 & @xmath271 + & & & & @xmath2 km & target & & & + @xmath169 au & @xmath272 & @xmath169 & @xmath169 & @xmath273 & @xmath274 & @xmath169 & @xmath169 & @xmath273 + @xmath38 au & @xmath275 & @xmath169 & @xmath169 & @xmath229 & @xmath276 & @xmath169 & @xmath169 & @xmath277 + @xmath178 au & @xmath278 & @xmath169 & @xmath169 & @xmath279 & @xmath280 & @xmath169 & @xmath169 & @xmath281 + @xmath183 au & @xmath282 & @xmath169 & @xmath169 & @xmath283 & @xmath284 & @xmath169 & @xmath169 & @xmath285 + ccccccccccc & & & & @xmath0 km & target & & & + @xmath169 au & @xmath286 & @xmath287 & @xmath226 & @xmath288 & @xmath289 & @xmath290 & @xmath291 & @xmath292 + @xmath38 au & @xmath293 & @xmath294 & @xmath295 & @xmath296 & @xmath297 & @xmath298 & @xmath299 & @xmath300 + @xmath178 au & @xmath301 & @xmath302 & @xmath303 & @xmath292 & @xmath304 & @xmath305 & @xmath306 & @xmath307 + @xmath183 au & @xmath308 & @xmath309 & @xmath310 & @xmath311 & @xmath312 & @xmath313 & @xmath314 & @xmath315 + & & & & @xmath1 km & target & & & + @xmath169 au & @xmath316 & @xmath317 & @xmath318 & @xmath319 & @xmath320 & @xmath321 & @xmath322 & @xmath323 + @xmath38 au & @xmath324 & @xmath325 & @xmath326 & @xmath327 & @xmath328 & @xmath329 & @xmath330 & @xmath331 + @xmath178 au & @xmath332 & @xmath333 & @xmath334 & @xmath335 & @xmath336 & @xmath337 & @xmath338 & @xmath339 + @xmath183 au & @xmath340 & @xmath341 & @xmath342 & @xmath343 & @xmath344 & @xmath345 & @xmath346 & @xmath347 + & & & & @xmath2 km & target & & & + @xmath169 au & @xmath348 & @xmath349 & @xmath350 & @xmath351 & @xmath352 & @xmath353 & @xmath279 & @xmath354 + @xmath38 au & @xmath355 & @xmath356 & @xmath169 & @xmath357 & @xmath358 & @xmath359 & @xmath229 & @xmath360 + @xmath178 au & @xmath361 & @xmath362 & @xmath229 & @xmath363 & @xmath364 & @xmath365 & @xmath229 & @xmath366 + @xmath183 au & @xmath367 & @xmath368 & @xmath169 & @xmath369 & @xmath370 & @xmath371 & @xmath229 & @xmath372 + ccccccccccc & & & & @xmath0 km & target & & & + @xmath169 au & @xmath373 & @xmath374 & @xmath277 & @xmath375 & @xmath376 & @xmath377 & @xmath378 & @xmath379 + @xmath38 au & @xmath380 & @xmath381 & @xmath350 & @xmath382 & @xmath383 & @xmath384 & @xmath385 & @xmath309 + @xmath178 au & @xmath386 & @xmath387 & @xmath283 & @xmath388 & @xmath389 & @xmath390 & @xmath300 & @xmath391 + @xmath183 au & @xmath392 & @xmath393 & @xmath394 & @xmath395 & @xmath396 & @xmath397 & @xmath398 & @xmath399 + & & & & @xmath1 km & target & & & + @xmath169 au & @xmath400 & @xmath401 & @xmath169 & @xmath378 & @xmath402 & @xmath403 & @xmath404 & @xmath277 + @xmath38 au & @xmath405 & @xmath406 & @xmath233 & @xmath407 & @xmath408 & @xmath409 & @xmath296 & @xmath379 + @xmath178 au & @xmath410 & @xmath411 & @xmath412 & @xmath283 & @xmath413 & @xmath414 & @xmath415 & @xmath416 + @xmath183 au & @xmath417 & @xmath418 & @xmath419 & @xmath11 & @xmath420 & @xmath421 & @xmath422 & @xmath423 + & & & & @xmath2 km & target & & & + @xmath169 au & @xmath424 & @xmath425 & @xmath169 & @xmath229 & @xmath426 & @xmath279 & @xmath169 & @xmath229 + @xmath38 au & @xmath427 & @xmath412 & @xmath169 & @xmath229 & @xmath428 & @xmath429 & @xmath169 & @xmath277 + @xmath178 au & @xmath430 & @xmath431 & @xmath169 & @xmath233 & @xmath432 & @xmath433 & @xmath169 & @xmath434 + @xmath183 au & @xmath435 & @xmath436 & @xmath229 & @xmath437 & @xmath438 & @xmath439 & @xmath229 & @xmath440 + ccccccccccc & & & & @xmath0 km & target & & & + @xmath169 au & @xmath441 & @xmath442 & @xmath443 & @xmath444 & @xmath445 & @xmath446 & @xmath447 & @xmath419 + @xmath38 au & @xmath448 & @xmath449 & @xmath450 & @xmath451 & @xmath452 & @xmath453 & @xmath454 & @xmath455 + @xmath178 au & @xmath456 & @xmath457 & @xmath458 & @xmath459 & @xmath460 & @xmath461 & @xmath462 & @xmath463 + @xmath183 au & @xmath464 & @xmath465 & @xmath466 & @xmath467 & @xmath468 & @xmath469 & @xmath470 & @xmath471 + & & & & @xmath1 km & target & & & + @xmath169 au & @xmath472 & @xmath473 & @xmath474 & @xmath326 & @xmath475 & @xmath476 & @xmath477 & @xmath478 + @xmath38 au & @xmath479 & @xmath480 & @xmath481 & @xmath482 & @xmath483 & @xmath484 & @xmath485 & @xmath486 + @xmath178 au & @xmath487 & @xmath488 & @xmath489 & @xmath298 & @xmath490 & @xmath491 & @xmath492 & @xmath442 + @xmath183 au & @xmath493 & @xmath494 & @xmath495 & @xmath496 & @xmath497 & @xmath498 & @xmath499 & @xmath500 + & & & & @xmath2 km & target & & & + @xmath169 au & @xmath501 & @xmath502 & @xmath503 & @xmath504 & @xmath505 & @xmath506 & @xmath507 & @xmath508 + @xmath38 au & @xmath509 & @xmath510 & @xmath277 & @xmath511 & @xmath512 & @xmath513 & @xmath285 & @xmath514 + @xmath178 au & @xmath515 & @xmath516 & @xmath517 & @xmath518 & @xmath519 & @xmath520 & @xmath259 & @xmath521 + @xmath183 au & @xmath522 & @xmath523 & @xmath287 & @xmath524 & @xmath525 & @xmath526 & @xmath527 & @xmath528 + 100 benz w. , asphaug e. , `` catastrophic disruptions revisited '' , 1999 , icarus , 142 , 5 - 20 bottke w. f. , durda d. d. , nesvorny d. , jedicke r. , mordibelli a. , vokrouhlicky d. , levison h. , `` the fossilized size distribution of the main asteroid belt '' , 2005 , icarus , 175 , 111 - 140 bottke w. f. , durda d. d. , nesvorny d. , jedicke r. , mordibelli a. , vokrouhlicky d. , levison h. , `` linking the collisional history of the main asteroid belt to its dynamical excitation and depletion '' , 2005 , icarus , 179 , 63 - 94 chambers j. e. , `` planetesimal formation by turbulent concentration '' , 2010 , icarus , 208 , 505 - 517 chambers j. e. , wetherill g. w. , `` planets in the asteroid belt '' , 2001 , meteoritics and planetary science , 36 , 381 - 399 coradini a. , federico c. , magni g. , `` formation of planetesimals in an evolving protoplanetary disk '' , 1981 , astronomy and astrophysics , 98 , 173 - 185 coradini a. , magni g. , turrini d. , `` from gas to satellitesimals : disk formation and evolution '' , 2010 , space science reviews , 153 , 411 - 429 coradini a. , turrini d. , federico c. , magni g. , `` vesta and ceres : crossing the history of the solar system '' , 2011 , space science reviews , doi:10.1007/s11214 - 011 - 9792-x cuzzi j. n. , hogan r. c. , shariff k. , `` toward planetesimals : dense chondrule clumps in the protoplanetary nebula . '' , 2008 , the astophysical journal , 687 , 14321447 cuzzi j. n. , hogan r. c. , bottke w. f. , `` towards initial mass functions for asteroids and kuiper belt objects '' , 2010 , icarus , 208 , 518 - 538 dangelo g. , kley w. , henning t. , `` orbital migration and mass accretion of protoplanets in three - dimensional global computations with nested grids '' , 2003 , the astrophysical journal , 586 , 540 - 561 encrenaz t. , `` water in the solar system '' , 2008 , annual review of astronomy and astrophysics , 46 , 5787 goldreich p. , ward w. r. , `` the formation of planetesimals '' , 1973 , the astrophysical journal , 183 , 10511062 gomes r. , levison h.f .
, tsiganis k. , morbidelli a. , `` origin of the cataclysmic late heavy bombardment period of the terrestrial planets '' , nature , 2005 , 435,466 - 469 holsapple k. a. , housen k. r. , `` a crater and its ejecta : an interpretation of deep impact '' , 2007 , icarus , 187 , 345 - 356 jorda l. , lamy p. , faury g. , et al .
`` properties of the dust cloud caused by the deep impact experiment '' , 2007 , icarus , 187 , 208 - 219 krivov a. , `` debris disks : seeing dust , thinking of planetesimals and planets '' , 2010 , research in astronomy and astrophysics , 2010 , 10 , 383414 lissauer j. j. , hubickyi o. , dangelo g. , bodenheimer p. , `` models of jupiter s growth incorporating thermal and hydrodynamics constraints '' , 2009 , icarus , 199 , 338 - 350 magni g. , coradini a. , `` formation of jupiter by nucleated instability '' , 2004 , planetary and space science , 52 , 343 - 360 meyer m. r. , `` circumstellar disk evolution : constraining theories of planet formation '' , 2008 , proceedings of the international astronomical union , 4 , 111 - 122 , doi : 10.1017/s1743921309031767 minton d. a. , malhotra r. , `` dynamical erosion of the asteroid belt and implications for large impacts in the inner solar system '' , 2010 , icarus , 207 , 744757 morbidelli a. , levison h.f . , tsiganis k. , gomes r. , `` chaotic capture of jupiter s trojan asteroids in the early solar system '' , nature , 2005 , 435 , 462 - 465 morbidelli a. , bottke w. f. , nesvorny d. , levison h. f. , `` asteroids were born big '' , 2009 , icarus , 204 , 558 - 573 nagasawa m. , thommes e. w. , kenyon s. j. , bromley b. c. , lin d. n. c. , `` the diverse origins of terrestrial - planet systems '' , 2007 , protostars and planets v , eds .
b. reipurth , d. jewitt , and k. keil , university of arizona press , tucson , 639 - 654 natta a. , testi l. , calvet n. , henning th .
, waters r. , wilner d .. `` dust in protoplanetary disks : properties and evolution '' , 2007 , protostars and planets v , eds .
b. reipurth , d. jewitt , and k. keil , university of arizona press , tucson , 767 - 781 obrien d. p. , morbidelli a. , bottke w. f. , `` the primordial excitation and clearing of the asteroid belt revisited '' , 2007 , icarus , 191 , 434 - 452 obrien d. p. , sykes m. v. , `` the origin and evolution of the asteroid belt
implications for vesta and ceres '' , 2011 , space science reviews , doi:10.1007/s11214 - 011 - 9808 - 6 opik e. j. , `` interplanetary encounters '' , 1976 , eds .
kopal z. & cameron a. g. w. , elsevier , isbn 0 - 444 - 41371 - 5 papaloizou j. c. b. , nelson r. p. , kley w. , masset f. s. , artymowicz p. , `` disk - planet interactions during planet formation '' , 2007 ,
protostars and planets v , eds .
b. reipurth , d. jewitt , and k. keil , university of arizona press , tucson , 655 - 668 petit j. , morbidelli a. , chambers j. , `` the primordial excitation and clearing of the asteroid belt '' , 2001 , icarus , 153 , 338 - 347 safronov v. s. , `` evolution of the protoplanetary cloud and formation of the earth and planets '' , 1969 , nauka press , moscow , english translation : nasa ttf-677.7 scott e. r. d. , `` chondrites and the protoplanetary disk '' , 2007 , annual reviews of earth and planetary sciences , 35 , 577 - 620 scott e. r. d. , `` meteoritics and dynamical constrains on the growth mechanisms and formation times of asteroids and jupiter '' , 2006 , icarus , 185 , 72 - 82 svetsov v. , cratering erosion of planetary embryos " , 2011 , icarus , 214 , 316 - 326 tsiganis k. , gomes r. , morbidelli a. , levison h.f .
, `` origin of the orbital architecture of the giant planets of the solar system '' , nature , 2005 , 435,459 - 461 turrini d. , magni g. , coradini a.,``probing the history of solar system through the cratering records on vesta and ceres '' , 2011 , monthly notices of the royal astronomical society , 413 , 2439 - 2466 k.j .
walsh , a. morbidelli , s.n .
raymond , d.p .
obrien , a. m. mandell , `` a low mass for mars from jupiter s early gas - driven migration '' .
nature , 2011 , doi:10.1038/nature10201 weidenschilling s. j. , `` the distribution of mass in the planetary system and solar nebula '' , 1977 , astrophysics and space science , 51 , 153 - 158 . weidenschilling s. j. , `` dust to planetesimals - settling and coagulation in the solar nebula '' , 1980 , icarus , 44 , 172 - 189 weidenschilling s. j. , `` initial sizes of planetesimals and accretion of the asteroids '' , 2011 , icarus , 214 , 671 - 684 wetherill g. w. , `` an alternative model for the formation of asteroids '' , 1992 , icarus , 100 , 307 - 325 | the asteroid belt is an open window on the history of the solar system , as it preserves records of both its formation process and its secular evolution .
the progenitors of the present - day asteroids formed in the solar nebula almost contemporary to the giant planets .
the actual process producing the first generation of asteroids is uncertain , strongly depending on the physical characteristics of the solar nebula , and the different scenarios produce very diverse initial size - frequency distributions . in this work
we investigate the implications of the formation of jupiter , plausibly the first giant planet to form , on the evolution of the primordial asteroid belt .
the formation of jupiter triggered a short but intense period of primordial bombardment , previously unaccounted for , which caused an early phase of enhanced collisional evolution in the asteroid belt .
our results indicate that this jovian early bombardment caused the erosion or the disruption of bodies smaller than a threshold size , which strongly depends on the size - frequency distribution of the primordial planetesimals .
if the asteroid belt was dominated by planetesimals less than @xmath0 km in diameter , the primordial bombardment would have caused the erosion of bodies smaller than @xmath1 km in diameter .
if the asteroid belt was instead dominated by larger planetesimals , the bombardment would have resulted in the destruction of bodies as big as @xmath2 km . |
as discussed in the main text , we find compelling evidence for the existence of the symmetry - protected topological haldane insulator phase in the one - dimensional ( 1d ) extended anyon - hubbard model ( eahm ) , by calculating an order parameter from the largest eigenvalues of the generalized transfer matrix within an infinite density - matrix renormalization group ( idmrg ) scheme . here
we show that for the eahm ( with maximum number of particles per site @xmath30 ) further quantities can be exploited in order to determine and characterize the phase boundaries and quantum phase transitions with high precision .
the entanglement analysis in particular provides us with valuable information about the existence of a symmetry protected haldane insulator ( hi ) in the eahm .
furthermore , it allows to determine the phase boundaries between the hi and other insulating phases . dividing a system into two subblocks , @xmath117 , and considering the reduced density matrix @xmath118 $ ]
, the entanglement spectra @xcite can be extracted from the singular values @xmath119 of @xmath120 as @xmath121 .
in addition , the correlation length @xmath89 can be obtained from the second largest eigenvalue of the transfer matrix for some fixed bond dimension @xmath90 in an idmrg simulation @xcite .
while @xmath89 stays finite as a consequence of the fixed bond dimension @xmath90 , the physical correlation length will diverge at the critical point .
nevertheless , @xmath89 is useful to pinpoint a phase boundary because it rapidly increases with @xmath90 close to the quantum phase transition point
. figures [ xi - es - cstar](a ) and ( b ) show @xmath89 and @xmath122 as functions of @xmath85 for fixed @xmath47 .
the strong upturn of @xmath89 indicates the formation of a hi phase in the eahm for @xmath88 .
we find distinct peaks at @xmath123 and @xmath124 , which become more pronounced as @xmath90 grows from 100 to 200 , signaling a divergence of @xmath125 as @xmath126 . at the same time
, the entanglement spectra develops a characteristic double degeneracy in all entanglement levels for @xmath127 , indicating a symmetry - protected topological phase between mi and dw states .
the universality class of these quantum phase transitions can be explored by calculating the central charge numerically , just as in case of the ebhm @xcite .
when the system gets critical the central charge can be determined very accurately by dmrg , utilizing the relation @xcite @xmath128}{\ln[\cos(\pi / l)]}\;. \label{cstar}\end{aligned}\ ] ] in this way , the mi - sf transition in the bhm @xcite , and especially , the university class of the mi - hi and hi - dw quantum phase transitions in the ebhm have been determined in the past @xcite .
figure [ xi - es - cstar](c ) displays @xmath129 for the 1d eahm , where the model parameters are the same as in figs .
[ xi - es - cstar](a ) and ( b ) .
running the dmrg we adopt periodic boundary conditions for system sizes up to @xmath130 . for @xmath47 and @xmath131 [ @xmath132
, we find @xmath133 [ @xmath134 , which points to the universality class of the luttinger liquid ( ising ) model , in accordance with what was obtained for the corresponding quantum phase transitions in the ebhm ( @xmath93 ) .
[ panel ( a ) ] and entanglement spectrum @xmath122 [ panel ( b ) ] as a function of @xmath85 for @xmath47 and @xmath48 from idmrg .
panel ( c ) displays the central charge @xmath129 for the same parameter set , signaling a mi - hi ( hi - dw ) quantum phase transition with @xmath91 ( @xmath92 ) . here
data obtained by the finite - system dmrg with periodic boundary conditions . ]
6 ifxundefined [ 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/physrevlett.101.010504 [ * * , ( ) ] http://arxiv.org/abs/0804.2509 [ ] link:\doibase 10.1016/j.aop.2010.09.012 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.113.020401 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.195108 [ * * , ( ) ]
link:\doibase 10.1103/physreva.85.053644 [ * * , ( ) ] | we demonstrate numerically the existence of a nontrivial topological haldane phase for the one - dimensional extended ( @xmath0-@xmath1 ) hubbard model with a mean density of one particle per site , not only for bosons but also for anyons , despite a broken reflection parity symmetry .
the haldane insulator , surrounded by superfluid , mott insulator and density - wave phases in the @xmath1-@xmath0 parameter plane , is protected by combined ( modified ) spatial - inversion and time - reversal symmetries , which is verified within our matrix - product - state based infinite density - matrix renormalization group scheme by analyzing generalized transfer matrices . with regard to an experimental verification of the anyonic haldane insulator state
the calculated asymmetry of the dynamical density structure factor should be of particular importance .
anyons represent a third fundamental class of particles with fractional exchange statistics that interpolates , to some degree , between those of bosons and fermions having symmetric or antisymmetric wave functions under exchange @xcite .
by contrast , the exchange of two anyons creates a phase factor @xmath2 in the many - body wave function , where the statistical parameter @xmath3 can be of any value in the interval @xmath4 . in the beginning anyons
were thought to be relevant solely for two - dimensional systems .
describing the fractional quantum hall effect experiments in particular , the quasiparticles could be viewed as anyons with @xmath3 fixed by the filling factor @xcite .
with haldane s generalized pauli principle and definition of fractional statistics , however , the concept of anyons becomes important in arbitrary dimensions @xcite . in one dimension , the physics of anyons might be studied successfully with ultra - cold atoms in optical lattices @xcite .
for example , keilmann _ et al .
_ proposed that anyon statistics in one - dimensional ( 1d ) lattices can be implemented by bosons with occupation - dependent hopping amplitudes generated by assisted raman tunneling @xcite , an approach that subsequently was refined by greschner and santos @xcite .
an alternative route to create 1d anyons in an optical lattice has been put forward by strter _
et al . _
@xcite quite recently .
their scheme exploits lattice - shaking - assisted tunneling against potential offsets generated by a combination of a static potential tilt and strong on - site interactions .
thereby , advantageously , no additional lasers are required , except for those employed on creating optical lattices . however , in spite of the huge experimental efforts , a conclusive detection of 1d anyons in optical lattices has not yet been achieved . notwithstanding , from a theoretical point of view , anyons in one dimension received a continuous and legitimate interest on account of their intriguing physical properties . using
the bethe ansatz technique kundu obtained the exact solution of an 1d anyon gas with delta - function potential @xcite . in the framework of the anyon - hubbard model ( ahm ) on a 1d lattice , the possibility of a statistically induced quantum phase transition between mott - insulator ( mi ) and superfluid ( sf ) phases @xcite and the asymmetry of the momentum distribution for hard - core @xcite and soft - core anyons @xcite have been addressed so far .
since the ahm is equivalent to a variant of the bose - hubbard model ( bhm ) with state - dependent bosonic hopping amplitudes @xcite , the next very interesting question might be whether the symmetry - protected topological haldane state @xcite , observed in the extended bhm ( ebhm ) with an additional nearest - neighbor particle repulsion @xcite , also shows up in the extended ahm ( eahm ) .
as the hopping phase factor breaks reflection parity in the system @xcite , at first glance one would expect the haldane state to disappear in the eahm for any finite fractional phase @xmath3 .
perhaps surprisingly , this does not necessarily need to happen . to examine the possibility of a topological haldane state in one dimension , in this letter
, we study the eahm using the matrix - product - state ( mps ) based density - matrix renormalization - group ( dmrg ) technique @xcite .
considering the invariance of the density - dependent hopping amplitudes and calculating generalized transfer matrices @xcite from the infinite mps ( imps ) of the infinite dmrg ( idmrg ) simulations @xcite , we can determine the protecting symmetry of the haldane state and derive the complete ground - state phase diagram of this paradigmatic anyonic model hamiltonian . in order to discriminate the topological haldane insulator ( hi ) from the other , more conventional mott ( mi ) and density - wave ( dw ) phases in possible future experiments
, we will also analyze the dynamical density response of the system , showing a characteristic asymmetry in the brillouin zone , which can be attributed to the fractional phase factor of the anyons .
the hamiltonian of the 1d eahm consists of three terms , @xmath5 , with @xmath6 @xmath7 and @xmath8 , describing the nearest - neighbor anyon transfer ( @xmath9 ) , as well as the repulsive on - site ( @xmath10 ) and nearest - neighbor ( @xmath11 ) particle interaction , respectively .
the anyon creation ( @xmath12 ) , annihilation ( @xmath13 ) and particle number ( @xmath14 ) operators at lattice site @xmath15 are defined by the generalized commutation relations @xcite : @xmath16 where the sign function @xmath17 for @xmath18 is mandatory , since two anyons on the same site behave as ordinary bosons .
anyons with @xmath19 represent so - called `` pseudofermions '' , namely , they are fermions off - site , while being bosons on - site . implementing a fractional jordan wigner transformation @xcite , @xmath20 where @xmath21 ( @xmath22 ) is a boson creation ( annihilation ) operator , @xmath23 with density - dependent hopping amplitudes , @xmath24 that is when a boson hops to the left from site @xmath25 to site @xmath15 it acquires a occupation dependent phase @xmath26 .
of course , @xmath27 which means that @xmath28 and @xmath29 are form - invariant under the anyon - boson mapping .
if we limit the maximum number of particles per site as @xmath30 , the ebhm , resulting in the limit @xmath31 from @xmath32 , maps to an effective @xmath33 spin-1 chain model @xcite : @xmath34 with the pseudospin operator @xmath35 . here , we have neglected terms which break the particle - hole symmetry .
note the negative sign of the first term compared to the usual @xmath33 spin - chain hamiltonian .
this leads to a protecting modified inversion symmetry ( @xmath36 ) for the haldane state of the ebhm @xcite : @xmath37 owing to the occupation - dependent hopping in the hi phase in the eahm seems not be protected by the modified inversion symmetry @xmath36 . to clarify whether @xmath38 is invariant under certain symmetry operations ,
let us first consider the inversion symmetry operator @xmath39 , acting on @xmath40 with @xmath41 applying next a time - reversal transformation @xmath42 , @xmath43 , we obtain @xmath44 in order that @xmath38 stays invariant under the combined symmetry operations , we take the following transformation into account : @xmath45 since the second term of eq .
transforms as @xmath46 , it is equal to the first term of @xmath38 . , defined by , selecting the topological state in the eahm at fixed @xmath47 and @xmath48 for different @xmath49 [ panel ( a ) ] , and at fixed @xmath50 and @xmath51 for @xmath30 [ panel ( b ) ] .
data obtained by idmrg using with a ( relatively small ) bond dimension @xmath52 . ]
the hamiltonian @xmath32 is therefore invariant under the transformation @xmath53 we now show that the combination of @xmath54 and @xmath55 is related to a symmetry - protected topological phase in the eahm , and define a corresponding topological order parameter . following ref .
@xcite , we use the imps representation formed by @xmath56 complex matrices @xmath57 and positive , real , diagonal matrices @xmath58 .
the imps is assumed to be in the canonical form : @xmath59 .
if a state @xmath60 is invariant under an internal symmetry which is represented by a unitary matrix @xmath61 , then the transformed @xmath62 matrices can be shown to satisfy @xcite @xmath63 where @xmath0 is a unitary matrix that commutes with the @xmath58 matrices , and @xmath64 is a phase factor .
similar relations hold for time reversal symmetry , inversion symmetry , and a combination of both . in those cases
@xmath57 on the left - hand side is replaced by its complex conjugate @xmath65 , its transpose @xmath66 and its hermitian transpose @xmath67 , respectively .
the properties of the matrices @xmath0 can be used to classify symmetry - protected topological phases @xcite .
for instance , in the case of time reversal or ( modified ) inversion symmetry the matrices satisfy @xmath68 or @xmath69 , and the sign distinguishes between two symmetric phases . in the eahm , the situation is slightly different because time reversal and inversion are not symmetries of the system , only a combination @xmath55 of them is . for @xmath70 and @xmath55
we have @xmath71 and @xmath72 .
from this we can derive a symmetry - protected topological phase similar to the case of the @xmath73 spin rotation symmetry of @xmath70 and @xmath74 in the spin-1 @xmath33 chain @xcite . since the phase factors @xmath75 and @xmath76 can be removed by absorbing them into the corresponding matrices @xmath77 and @xmath78 they have no physical meaning .
however , if both @xmath70 and @xmath55 are preserved , the combination @xmath79 is a symmetry as well and its phase factor is not arbitrary if @xmath77 and @xmath78 have been fixed .
indeed one can show that @xmath80 which defines two different phases . to verify that the eahm has a nontrivial topological phase protected by @xmath70 and @xmath55
, we calculate the order parameter @xcite @xmath81
if the state is symmetric under both @xmath55 and @xmath82 . otherwise , if one of the symmetries is broken , the order parameter is zero .
the idmrg results for the order parameter are shown in fig .
[ topo - order ] .
if @xmath78 and @xmath77 commute ( @xmath83 ) , the system is in a trivial phase , i.e. , a site - factorizable mi state , whereas if they anticommute ( @xmath84 ) , the system realizes a nontrivial haldane phase . increasing the number of particles per site @xmath49 at fixed @xmath47 , the hi phase ( @xmath84 ) slightly shifts to larger value of @xmath85 but , most notably , the haldane phase still occupies a solid parameter region , see data for @xmath86 and @xmath87 in fig .
[ topo - order](a ) . increasing the fractional angle @xmath3 for @xmath30 , the haldane state region narrows and finally disappears for @xmath19 ( compare the data presented in fig .
[ topo - order](b ) for @xmath50 ) .
we like to emphasize that the hi sector marked in fig .
[ topo - order ] by the gray area agrees with that extracted from the correlation length , the entanglement spectrum , and the numerically obtained central charge @xcite . , @xmath30 , and @xmath48 .
most notably the haldane insulator ( hi ) , located between mott insulator ( mi ) and density wave ( dw ) insulating phases in the ebhm , survives for any @xmath88 , i.e. , in the anyonic case .
likewise the superfluid ( sf ) appears in the very weak - coupling regime .
the mi - hi ( squares ) and hi - dw ( circles ) transition points can be determined by a divergent correlation length @xmath89 as @xmath90 increases , i.e. , the model becomes critical with the central charge @xmath91 and @xmath92 , respectively ( see @xcite ) . for comparison
, the dotted ( dashed ) line marks the mi - hi ( hi - dw ) transition in the ebhm ( @xmath93 ) @xcite .
the dash - dotted line with triangles up denotes the first - order mi - dw phase transition for @xmath19 . ]
figure [ pd ] represents the ground - state phase diagram of the 1d eahm in the @xmath1-@xmath0 plane , as obtained from large - scale ( i)dmrg calculations for @xmath48 and @xmath30 .
the phase boundaries are determined simulating the order parameter @xmath94 , as well as the correlation length and the entanglement spectrum @xcite .
the eahm exhibits three different insulating phases ( mi , dw , and hi ) and a superfluid state in the weak interaction regime , just as for the ebhm @xcite but with the addition that the region of the intervening anyonic hi phase at @xmath48 is slightly reduced . the hi will disappears in the pseudofermionic case ( @xmath19 ) .
the mi - hi and hi - dw quantum phase transitions belong to the universality class of tomonaga - luttinger liquid and ising model , with central charge @xmath91 and @xmath95 , respectively , see fig . [ xi - es - cstar](c ) in @xcite .
perhaps the most striking feature of the ahm is the asymmetry of the momentum distribution function in @xmath96-space @xcite .
the position of the maximum strongly depends on the fractional phase @xmath3 [ remind that the momentum distribution diverges at @xmath97 in the bose - hubbard model ( @xmath93 ) ] .
we expect that this asymmetry can also be observed in dynamical quantities such as the dynamical structure factor @xmath98 .
hence , if an anyonic system will be realized in optical lattices , @xmath98 might be one of the best physical quantities to look at , comparing theoretical predictions with real experiments , like for 1d bose hubbard type models @xcite .
@xmath98 should be easily accessible by momentum resolved bragg spectroscopy @xcite .
furthermore , it has been recently demonstrated that @xmath98 can also be used to distinguish the topological hi from the conventional mi and dw states @xcite , in analogy to exploiting the dynamical spin - spin structure factor in the spin-1 @xmath33 chain @xcite .
the dynamical density structure factor is defined as @xmath99 where @xmath100 ( @xmath101 ) denotes the ground ( @xmath102th excited ) state and @xmath103 . to compute this quantity ,
we follow ref .
@xcite and first determine the two - point correlation function @xmath104 by real - time evolution of the imps @xmath100 .
fourier transformation then provides us with accurate numerical results of the dynamical structure factor in the eahm .
figure [ skw ] compares the intensity of the dynamical wave - vector - resolved density response in the ebhm ( @xmath93 ) with those in the eahm for @xmath48 , for @xmath47 , at five characteristic @xmath85-values .
one point worthy of remark is that each of the phases and phase transitions can be distinguished by looking at @xmath98 . in the mi , at @xmath105 [ fig .
[ skw](a ) and ( f ) ] , the excitation gap appears at @xmath106 . with increasing @xmath85 ,
the mi - hi transition occurs at @xmath107 , where the excitation gap closes at @xmath97 as shown in fig .
[ skw](b ) and ( g ) .
deep in the hi phase , @xmath108 [ fig . [ skw](c ) and ( h ) ] , the spectral weight exclusively concentrates at @xmath109 and there are finite excitation gaps at @xmath97 and @xmath51 .
it is of particular interest to see whether the gap @xmath98 closes at the hi - dw transition point .
indeed , the excitation gap at @xmath110 closes , but at momentum @xmath111 , reflecting the lattice - period doubling in the dw phase .
moreover , in the dw phase [ fig .
[ skw](e ) and ( j ) ] , we find a large excitation gap at @xmath111 and two dispersive branches , where a changeover of the intensity maximum occurs at @xmath112 ( @xmath113 ) for @xmath93 ( @xmath48 ) .
interestingly , the influence of the occupation - dependent phase of @xmath38 in shows up in @xmath98 as well , which helps to differentiate the results from those of the ebhm .
@xmath98 of the eahm is asymmetric for any @xmath114 , while @xmath98 in the ebhm is always symmetric about @xmath111 .
to summarize , we carried out an unbiased numerical investigation of the extended anyon - hubbard model in one dimension and determined its ground - state phase diagram with high precision exploiting the behavior of correlation lengths and entanglement spectra . defining an order parameter that distinguishes trivial and nontrivial topological phases
, we were able to show that the eahm possesses an anyonic haldane insulator state sandwiched between superfluid , mott insulator and density - wave phases .
both the hi - mi and hi - dw quantum phase transitions are critical with central charge 1 and 1/2 , respectively . while the hi state survives the ebhm limit @xmath115 ) , it vanishes when the system is composed of pseudofermions @xmath116 ) . if an 1d interacting anyonic system could be realized experimentally in the future , maybe in an optical - lattice setup with ultracold atoms , we suggest performing momentum - resolved bragg spectroscopy to look for the pronounced asymmetry of the density response spectra in momentum space which we have demonstrated in our model calculation theoretically .
the idmrg simulations were performed using the itensor library @xcite .
this work was supported by deutsche forschungsgemeinschaft ( germany ) , sfb 652 , project b5 .
33ifxundefined [ 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 * * , ( ) link:\doibase 10.1103/physrevlett.49.957 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.48.1559 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.50.1395 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.67.937 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.80.885 [ * * , ( ) ] link:\doibase 10.1038/ncomms1353 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.115.053002 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.117.205303 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.83.1275 [ * * , ( ) ] link:\doibase 10.1103/physreva.94.013611 [ * * , ( ) ] link:\doibase 10.1103/physreva.79.043633 [ * * , ( ) ] http://stacks.iop.org/1367-2630/17/i=12/a=123016 [ * * , ( ) ] link:\doibase 10.1103/physrevb.80.155131 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.064439 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.97.260401 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.245119 [ * * , ( ) ] link:\doibase 10.1142/0961 [ _ _ ] ( , ) link:\doibase 10.1103/physrevlett.69.2863 [ * * , ( ) ] link:\doibase 10.1016/j.aop.2010.09.012 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.125441 [ * * , ( ) ] http://arxiv.org/abs/0804.2509 [ ] link:\doibase 10.1103/physrevlett.98.070201 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.100.167202 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.235128 [ * * , ( ) ] @noop link:\doibase 10.1103/physrevlett.113.020401 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1038/nphys1476 [ * * , ( ) ] http://stacks.iop.org/1742-6596/592/i=1/a=012134 [ * * , ( ) ] link:\doibase 10.1103/physrevb.91.045121 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.245107 [ * * , ( ) ] @noop |
in this paper we develop a model for high energy cosmic rays ( hecrs ; here defined as @xmath8ev / nucleon crs ) , based on the underlying assumption that crs are accelerated in the relativistic and nonrelativistic shocks found in grbs and their attendant supernovae ( sne ) .
this model extends the work of vietri @xcite and waxman @xcite , who proposed that uhecrs originate from grbs ( see also ref.@xcite ) , to include cosmic ray production from sne and grb sources in our galaxy @xcite . to test the model ,
kascade @xcite and hires - i and hires - ii monocular data @xcite are fit over the energy range @xmath9 ev to @xmath10 ev .
a good fit to the entire data set is possible if crs are injected with a power - law spectrum with number index @xmath11 , as could be expected in a scenario where particles are accelerated by relativistic shocks @xcite .
crs that are injected with energies @xmath12 ev diffuse through and escape from their host galaxy .
the sources of high - energy crs , namely grbs , are located in star - forming regions found in the galaxy s disk .
cr transport in the disk and halo of the milky way is modeled using a time - dependent , spherically - symmetric propagation model that employs an energy - dependent , spatially - independent diffusion coefficient .
the random - walk pathlengths are assumed to arise from gyroresonant , pitch - angle scattering of crs with a magnetohydrodynamic ( mhd ) turbulence spectra that can be decomposed into two components reflecting different power - law distributions of turbulence over different wavelength ranges .
the spectral break at the cr knee energy is explained in an impulsive , single - source model for hecrs if the turbulence spectral index changes from kraichnan to kolmogorov turbulence near the wavenumber resonant with crs at the knee of the cr spectrum .
we model the preliminary kascade data reported in 2001 by assuming that all ionic species have the same injection index , with the compositions of the ions adjusted to fit the observed cr spectrum near the knee .
additional hardenings of the low - energy cr spectrum from a single grb source of hecrs can result both from energy - dependent diffusion and from a low - energy cutoff in the cr injection spectrum .
superposition of the contributions from many sne are assumed to accelerate the bulk of the gev / nuc tev / nuc crs , as in the conventional scenario @xcite .
uhecrs have such large gyroradii and diffusion mean free paths that they are assumed to escape directly from the halo of the grb host galaxy and stream into metagalactic space .
the uhecr energy spectrum evolves in response to photo - pair and photo - pion losses on the redshift - dependent cosmic microwave background radiation ( cmbr ) , which we treat in a continuous energy - loss formalism .
uhecrs in intercluster space also lose energy adiabatically due to cosmic expansion .
the measured uhecr spectrum arises from the contributions of sources throughout the universe , with an intensity that depends on the local luminosity density of grb sources and the evolution of the grb luminosity density with redshift ( see refs .
@xcite and @xcite for recent reviews of uhecrs and grbs , respectively ) .
the uhecr spectral model assumes that many sources produce the measured ultra - high energy and super - gzk ( @xmath13 ev ) crs .
our model for uhecrs from grbs implies a local time- and space - averaged cr luminosity density @xmath14 of crs . the local ( @xmath15 ) luminosity density @xmath16 ergs mpc@xmath17 yr@xmath18 is inferred from batse observations of the hard x - ray / soft @xmath7-ray ( x/@xmath7 ) emission from grbs @xcite .
the value of @xmath19 depends sensitively on the minimum energy @xmath20 of cr injection for soft injection spectral indices @xmath4 . for @xmath21
if @xmath22 gev then @xmath23 more energy must be injected in nonthermal hadrons than is observed as x/@xmath7 emission from grbs .
if @xmath24 tev , then @xmath25 more energy is required .
such large baryon loads would provide a bright cascade emission signature in grb spectra at mev
gev energies through photopion or hadron synchrotron processes @xcite .
the large nonthermal baryon load , @xmath26 , required to fit the uhecr data assuming that the grbs comoving luminosity density traces the star formation rate ( sfr ) history , implies that grbs can be much more luminous neutrino sources than predicted under the standard assumption that the energy injected into a grb blast wave in the form of crs is equal to the energy inferred from x/@xmath7 fluence measurements of grbs @xcite .
we therefore predict that if this model is correct , and crs are accelerated and injected in the form of soft power - laws with index @xmath4 , then icecube should detect up to several neutrinos from the brightest grbs with total x/@xmath7 radiation fluence at levels @xmath27 erg @xmath28 .
this prediction holds both in a collapsar scenario when the lorentz factors @xmath29 of the relativistic outflows are @xmath30 , or if the grb takes place in an intense external radiation background for a wide range of @xmath29 .
lower limits to @xmath29 can be inferred from @xmath7-ray transparency arguments applied to observations of grbs with the _ gamma - ray large area space telescope_. we also consider whether this model can explain the agasa data @xcite for the uhecrs .
poor fits are found if grbs inject soft cr spectra with @xmath4 .
however , if grbs inject hard spectra with @xmath31 , for example , through a second - order relativistic shock - fermi process @xcite or through the converter mechanism @xcite , then the highest - energy agasa data can be fit , though the reduced @xmath32 of our best fits are not compelling . because the injection spectrum is so hard , most of the produced high - energy neutrinos are too energetic and the flux too weak to be detected with icecube , though other telescope arrays , such as the extreme universe space observatory ( euso ) , could be sensitive to these gzk neutrinos . to explain the crs below @xmath33 ev , an additional component of crs
would still be required either from grbs or another class of sources , and these would make an additional contribution to high - energy neutrino production . section 2 gives a discussion of the cr and photon luminosity density of grbs and their event rate .
our propagation model describing cr diffusion in the disk and halo of the galaxy is presented in section 3 , where we fit the kascade data between @xmath34 and 200 pev with a single galactic grb source @xmath35 pc away that took place around 200,000 yrs ago . in section 4 , we describe our calculation of the uhecr flux , including energy losses from cosmic expansion , and photo - pair and photo - pion production .
we present minimum @xmath36 fits to the high - energy kascade , hires - i and hires - ii monocular data covering the energy range @xmath37 ev to @xmath38 ev .
we also fit the agasa data from @xmath39 ev to @xmath38 ev for hard cr injection spectra .
section 5 presents new high - energy neutrino calculations from hadronically dominated grbs , and our predictions for km - scale high - energy neutrino telescopes such as icecube or a deep underwater , northern hemisphere array .
discussion of the results and conclusions are given in section 6 .
our treatment of the uhecr attenuation and flux calculation is described in appendix a.
the progenitor sources of hecrs are likely to be grbs , as suggested by the following observations : 1 .
the evidence from kascade @xcite that the break energies of the different cr ionic species are proportional to rigidity , and that the mean atomic mass increases with @xmath40 through the knee region , admits a propagation solution to understand the energies of the spectral breaks of different cr ions given a galactic source that injects power - law crs to the highest energies .
composition changes above the second knee " at @xmath41 ev , and possible gzk features in the uhecr spectrum @xcite , implies that uhecrs and their sources are metagalactic .
the cr all - particle spectrum breaks at @xmath42 pev by @xmath43 units and then extends without spectral change , other than for a possible weak softening above the second knee , to the ankle at @xmath44 ev @xcite . a single power - law injection source , modified by acceleration and transport effects ,
provides the simplest solutions .
a comprehensive model of hecrs therefore seems achievable if a source type that injected power - law distributions of relativistic crs were found in star - forming @xmath45 galaxies such as the milky way , with injection episodes frequent and energetic enough to power the hecrs . for reasons reviewed in refs .
@xcite and @xcite , grbs offer the most likely solution .
we first address the question of the local ( @xmath46 ) time- and space - averaged grb luminosity density ( or emissivity ) @xmath47 ( ergs mpc@xmath17 yr@xmath18 ) and the local luminosity density @xmath48 required to power the super - gzk ( @xmath13 ev ) crs .
pre - beppo sax estimates @xcite concerning the x/@xmath7 emission from grbs found @xmath49 with @xmath50 .
this exceeds the local ( @xmath51 ) gzk emissivity @xmath52 ergs mpc@xmath17 yr@xmath18 required to power @xmath53 ev crs , where @xmath54 ergs cm@xmath17 is the observed energy density in @xmath53 ev crs ( @xmath56 for hires and @xmath57 for agasa ) , and the photo - pion energy - loss timescale for a @xmath58 ev proton is @xmath59 mpc/@xmath60 @xcite .
a better estimate that corrects for the energy - dependence of @xmath61 implies @xmath62 ergs mpc@xmath17 yr@xmath18 , comparable with the grb emissivity @xmath47 , so that the required luminosity density to power the super - gzk crs is in coincidence with that available from grbs .
based on the study by bttcher and dermer @xcite of batse statistics in the external shock model , a value of @xmath63 was derived @xcite using an analytic fit to the sfr derived from hubble deep field measurements @xcite .
this emissivity includes , in addition to the photon luminosity that is dominated by hard x - ray and @xmath7-ray emission , an inefficiency factor for production of photons due to the radiative regime implied by the fits to the data @xcite , and a further correction for dirty and clean fireballs @xcite , which have since been discovered @xcite . removing the inefficiency factor from the estimate of ref .
@xcite gives the local grb emissivity in the form of nonthermal photon radiation to be @xmath64 .
vietri , de marco , and guetta @xcite argue for a value of @xmath65 .
given the various uncertainties , we take the local grb emissivity in the form of hard x - rays and 100 kev
mev @xmath7-rays from grbs to be @xmath66 a second issue to be considered is the source rate of grbs in our galaxy . beaming breaks in the optical light curves of grbs indicate that the most apparently luminous grbs are highly collimated .
frail et al .
@xcite argue that a typical grb has a beaming factor of 1/500@xmath67 of the full sky , so that grbs are in actual fact 500 times more numerous , and 1/500@xmath67 as energetic on average , as the measured energy releases imply .
this means that most grbs typically release @xmath68 ergs in x/@xmath7 emission .
the local density of @xmath45 spiral galaxies , of which the milky way is representative , can be derived from the schechter luminosity function , and is @xmath69(200500 mpc@xmath70 ) @xcite .
the batse observations imply @xmath71 grbs / day over the full sky @xcite . due to beaming , this rate is increased by a factor of @xmath72 , where @xmath73 .
given that the volume of the universe is @xmath74 mpc)@xmath75 , this implies a rate per @xmath45 galaxy of @xmath76 @xmath77 @xmath78 the factor @xmath79 corrects for the star - formation activity at the present epoch [ @xmath80 ( see section 4 ) , and the factor @xmath81 accounts for dirty and clean fireball transients that are not detected as grbs @xcite .
this estimate is in agreement with the result of ref .
@xcite that excludes short grbs .
thus a grb occurs about once every 310 millennia throughout the milky way , or at about 10% of the rate of type ib / c sne .
we write @xmath82 for the local luminosity density of crs injected by grbs , where @xmath83 is the nonthermal baryon - loading factor by which the emissivity injected by grbs in the form of hadronic crs exceeds the emissivity inferred from direct observations of the x/@xmath7 emission from grbs . in the following ,
we take @xmath84 ( eq . [ l44 ] ) . from eqs .
( [ rhol * ] ) and ( [ elldot ] ) , we find that the average apparent isotropic energy release per grb is @xmath85 ergs , and the smaller actual mean energy release per grb due to beaming and the consequent larger number of grb sources is @xmath86 ergs .
for our canonical model grb used to fit cr data near the knee , we take @xmath87 ergs , corresponding to @xmath88 . a value of @xmath89 is implied by the data fits if this model for hecrs is correct ( section 5 ) .
hence , in a unified model for hecrs from grbs , the grb blast wave must be strongly baryon - loaded with @xmath90 , at least during the prompt phase of the grb when the acceleration is most rapid .
observations indicate that grb sources are located in the disks and star - forming regions of galaxies undergoing active star formation , such as the milky way .
the relativistic ejecta in the grb explosions accelerate and inject crs into the ism of the grb host galaxy in the form of a power law to the highest energies , though possibly with a low - energy cutoff to the accelerated proton spectrum . in relativistic blast waves ,
this cutoff is expected at @xmath91 gev energies , which could easily reach tev
pev energies for typical grb blast waves with @xmath92 300 .
uhecr acceleration and injection probably occurs during the @xmath7-ray luminous phase of a grb , which is on the order of minutes to hours .
acceleration and injection of lower energy crs might operate on the sedov time scale , which could exceed thousands of years .
these acceleration times are still short compared to the times for particles with the corresponding energies to diffuse a distance comparable to the disk scale height , as can be shown from the diffusion properties of the galaxy derived below .
a grb source can therefore be treated as an impulsive source of cr injection . in this section
we present a simplified propagation model for crs in the disk and halo of the galaxy .
we assume that the galaxy s disk and halo magnetic field consists of a large - scale field of mean strength @xmath93
@xmath94 g on which is superposed a spectrum of mhd turbulence .
the diffusion mean - free - path @xmath95 for crs diffusing through this field is a function of the larmor radius @xmath96 where @xmath97 is the lorentz factor of the relativistic cr proton or ion with atomic mass @xmath98 , @xmath99 , and the pitch angle @xmath100 is set equal to @xmath101 .
we assume isotropic turbulence , though it is straightforward to generalize the treatment for particle - scattering properties that differ in directions parallel and transverse to the plane of the disk .
we consider a very simplified approach to cr transport where particles diffuse via pitch - angle scattering with resonant mhd turbulence ; see , e.g. , refs .
@xcite , and ref .
@xcite for a detailed treatment .
let @xmath102 represent the differential energy density of isotropic mhd turbulence with wavenumbers between @xmath103 and @xmath104 .
first consider the case where the mhd turbulence spectrum is described in the inertial range by a single power - law function @xmath105 @xmath106 , where @xmath107 when @xmath108 and @xmath109 otherwise , and @xmath110 is the spectral index of the wave spectrum .
the index @xmath111 for a kolmogorov spectrum of turbulence , @xmath112 for a kraichnan spectrum of turbulence , and @xmath113 is the total energy density of mhd turbulence when @xmath114 and @xmath115 .
note that this description of the wave turbulence is an extreme oversimplification , insofar as we do not consider different wave modes and helicities , or distinguish between forward- and backward - propagating waves .
the ratio of the mhd wave energy @xmath116 to the energy density @xmath117 of the large - scale component of the galactic magnetic field is denoted by @xmath118 , so that @xmath119 for @xmath120 .
the value of @xmath121 corresponds to the inverse of the largest size scale on which turbulence is injected .
we make the ansatz that the diffusion coefficient of a cr with larmor radius @xmath122 is inversely proportional to the the energy density in gyroresonant waves with wavenumber @xmath123 @xcite .
we therefore have @xmath124 , giving @xmath125 where @xmath126 pc@xmath18 .
equation ( [ lambda ] ) gives the mean - free path @xmath95 of relativistic ( @xmath127 ) crs that diffuse due to gyroresonant interactions with large - wavenumber ( @xmath128 ) turbulence .
when @xmath129 , or @xmath130 pc , crs gyroresonate with small - wavenumber ( @xmath131 ) turbulence . if the knee in the cr all - particle spectrum at @xmath132 pev is due to propagation effects , then @xmath133 when @xmath134 for the dominant cr proton species .
the break energies in the spectra of other cr ions are also given by the relation @xmath135 , so that the knee energies for the different cr ions occur at energies @xmath136 the fit to the data described below uses @xmath137 and @xmath134 .
if the value of @xmath118 , corresponding to the ratio of the energy density @xmath138 turbulence to @xmath139 is a few percent , then @xmath140 for @xmath141 pev cr protons , where @xmath142 is the scale height of molecular gas or massive stars in the disk of the milky way .
a break in the all - particle spectrum due to a change in the propagation mode occurs at @xmath141 pev , where @xmath143 , with a pattern of break energies @xmath144 for the ionic constituents of the cr all - particle spectrum . although a cr proton with energy @xmath145 has @xmath146 pc @xmath147 , its mean - free - path @xmath148 because pitch - angle scattering results only from gyroresonant interactions with mhd waves that constitute a small fraction of the total magnetic field energy density .
crs injected with @xmath149 will diffuse through the galaxy s disk and halo through gyroresonant pitch - angle scattering with mhd turbulence carried by waves with wavenumbers @xmath150 that gyroresonate with particles with @xmath151 .
this suggests that we consider a more general wave spectrum over a larger range of @xmath103 given by the expression @xmath152\ ; , \label{kwk}\ ] ] where @xmath153 is the index of the small wavenumber mhd turbulence , and @xmath154 normalizes the energy densities of the small ( @xmath131 ) and large ( @xmath128 ) wavenumber turbulence spectra at @xmath155 .
the normalization condition @xmath156 implies @xmath157 , from which the normalization coefficient @xmath158 can be easily derived . the wave spectrum from eq .
( [ kwk ] ) is plotted in fig .
[ fig : kwk ] by the solid lines .
the parameters are @xmath159 pc@xmath18 , @xmath160 pc@xmath18 , @xmath161 , @xmath162 , @xmath163 , @xmath164 , and @xmath165 g
. sixty - five percent of the turbulence energy density is in the small wavenumber component , so that @xmath166 .
the sharp edges in this wave spectrum introduce unphysically sharp features in the spectra of particles for this simple model of transport .
a more physically plausible turbulence spectrum that retains the features of eq .
( [ kwk ] ) but smooths the particle distribution function , as would be physically expected from 3-wave interactions and turbulence - energy diffusion and cascading @xcite , is given by @xmath167 + a_0 ( { k\over k_1})^{1-q}[1-\lambda_\delta({k\over k_1 } ) ] , \label{kwknew}\ ] ] where the smoothing function @xmath168 , and we let @xmath169 . eq.([kwknew ] ) reduces to eq.([kwk ] ) in the limit @xmath170 .
[ fig : kwk ] shows a plot of eq .
( [ kwknew ] ) with @xmath171 and , for comparison , @xmath172 .
the subsequent fits to the data use the mhd wave spectrum , eq .
( [ kwknew ] ) , with @xmath171 . the ansatz @xmath173 $ ] with @xmath174 therefore implies @xmath175 + ( r _ { \rm l}k_1)^{q-1}[1-\lambda_\delta({r_{\rm l}k_1})]}.\ ; \label{lambdanew}\ ] ] in fig .
[ fig : lambda ] , we plot @xmath95 and @xmath122 as a function of total energy @xmath40 for cr protons and fe nuclei .
the turbulence spectrum is given by fig .
[ fig : kwk ] with @xmath176 .
note that @xmath177 kpc when the cr proton energy @xmath178 ev .
this energy is near the onset of quasi - rectilinear propagation in a galactic halo of size @xmath179 kpc , and we expect a cutoff in the cr proton flux from a recent grb in this energy range whenever we are not in the beam of a grb jet ( which is nearly always true because of the extreme improbability of such an occurrence ) .
the energy - dependent cutoff will be @xmath180 , as observed .
the actual situation is of course more complicated due to the magnetic field gradient from the disk to the halo and into metagalactic space .
we assume that a grb injects a total energy @xmath181 ergs in the form of cosmic rays . for impulsive injection
, the number injection spectrum of crs with atomic charge @xmath182 and atomic mass @xmath98 is assumed to be given by @xmath183 for @xmath184 , and we assume that @xmath185 is the same for all ( @xmath186 ) .
when @xmath187 , @xmath188 , and @xmath189 , @xmath190^{-1}\ ; { ( p-2)e_{cr } \over m_pc^2 \sum_{z , a } a c_{z , a}}\;. \label{ka}\ ] ] if @xmath191 , @xmath189 , @xmath192 cosmic rays are assumed to diffuse isotropically in the interstellar medium , with the initial anisotropy from injection by the grb jet quickly washed out . for energy - dependent diffusive propagation , the spectral number density of relativistic cr ions with charge @xmath182 and mass @xmath98 measured at a distance @xmath193 pc away from an impulsive source of crs from a grb that occurred a time @xmath194 earlier is @xmath195\;.\ ; \label{nza}\ ] ] here the diffusion radius @xmath196 , and eq .
( [ nza ] ) assumes that crs suffer no significant energy losses during transport ( see ref .
@xcite for the more general case ) . inspection of eq .
( [ nza ] ) shows that when the observer is within the diffusion radius , that is , when @xmath197 , then @xmath198 the measured spectrum from a burstlike source is therefore steepened by @xmath199 units , where the diffusion coefficient @xmath200 @xcite .
( by comparison , the spectral index is steepened by @xmath201 units for continuous injection sources of crs . ) eq .
( [ napprox ] ) shows that energy - dependent diffusion due to pitch - angle scattering with a kolmogorov and kraichnan spectrum of mhd turbulence steepens the injection spectrum of an impulsive source by 0.5 and 0.75 units , respectively .
if the disk and halo magnetic fields of the galaxy support a two - component mhd turbulence spectrum with indices @xmath111 and @xmath163 , then an injection spectrum with @xmath11 will be steepened to a measured spectrum @xmath202 , with @xmath203 at @xmath204 and @xmath205 at @xmath151 . because these indices are similar to the measured cr indices below and above the knee energy , we adopt this simplified model for cr transport .
speculations concerning the origin of such a turbulence spectrum are deferred to section 7 .
[ fig : evolve ] shows the cr proton flux measured at earth at different times after a grb explosion located 500 pc from earth .
the low - energy cutoff energy is given by @xmath206 , so that most of the energy is deposited by @xmath207 gev / nuc crs .
when @xmath208 , the cr intensity is increased at energies above the low - energy cutoff by the factor @xmath209^{p-2}$ ] .
the flux of relativistic crs is calculated from the expression @xmath210 where @xmath211 and @xmath212 is in cgs units .
propagation effects arising from the combined disk and halo mhd turbulence spectrum produce a break in the spectrum at @xmath213 pev .
the index of the cr number fluxes above the knee is @xmath214 , for the reasons given above , until propagation effects at the second knee " at @xmath215 ev start to soften the spectrum so that the metagalactic component begins to make a dominant contribution .
the observer is within the diffusion radius when @xmath216 , or when @xmath217 .
if we are considering crs with energies below the knee , then these particles scatter with the large wavenumber @xmath128 , @xmath111 turbulence , so that @xmath218 .
we write @xmath219 pc , where the observer is irradiated by a cr flux from a grb that took place @xmath220 myr ago . at early times
following the grb , the low - energy portion of the measured cr spectrum from a grb source is exponentially attenuated due to the slower diffusion of lower energy crs .
the observer will see the maximum flux from crs with energies @xmath221 when @xmath222 this expression uses the parameters for the model fits discussed below , with @xmath162 and @xmath223 , thus showing that the maximum flux of cr protons near the knee energy reaches an observer 500 pc away @xmath224 years after the grb , in accord with fig .
[ fig : evolve ] .
( [ tmyr ] ) shows that the low - energy cutoff @xmath225 of the cr proton flux in fig .
[ fig : evolve ] evolves according to the relation @xmath226 .
note that eq.([tmyr ] ) is only valid when @xmath227 .
fits to the cr data near the knee are compatible with a 500 pc distant grb releasing @xmath228 ergs in cosmic rays if the grb took place @xmath229 yrs ago .
this implies that the 0.1 100 pev cr fluxes were up to 100 times brighter one or two hundred thousand years ago than they are today , though no test of this implication suggests itself .
we now apply this model to fit the preliminary kascade data announced in 2001 @xcite . the points in fig .
[ fig : kascade ] show kascade data and error bars for cr protons , he , carbon , fe , and the all - particle spectrum in panels ( a ) ( e ) , respectively .
[ fig : kascade](d ) also shows the reported 1@xmath230 boundary to the estimated systematic errors for cr fe .
the fits to the kascade data shown in fig .
[ fig : kascade ] use the propagation model previously described , with a grb source of crs at a distance @xmath231 pc that exploded 210,000 yrs ago .
the change in the diffusion and propagation properties leads to a break in the cr particle spectrum at break energies @xmath180 ( eqs . [
[ lambda ] ] and [ [ ek ] ] ) .
because of the preliminary nature of the data and the potentially significant systematic errors that could still remain in the early analyses , we did not perform a rigorous fit to the data , but instead adjusted the wavenumber @xmath232 and the compositions of the different ionic species until a reasonable fit to the data was obtained . with @xmath233 g , the best value for the energy - dependent break was obtained with @xmath234 pc in the spectrum of turbulence . the anders - grevesse solar photospheric composition @xcite , and the ( energy - independent ) metallicity enhancements , compared to the anders - grevesse compositions , that were used to fit the data , are listed in table 1 .
the strong enhancements by a factor of 50 and 20 for c and fe , respectively , may be possible from highly - enriched winds that could be peculiar to a grb stellar progenitor , or if there was an earlier supernova as in the supranova model .
note that o can not be strongly enhanced if we are to maintain a good fit to the all - particle spectrum .
a more detailed discussion about implications from composition will depend on fits to final analysis of kascade data @xcite . [
cols="<,<,<",options="header " , ] the reported agasa flux above @xmath58 ev provides a clue that uhecrs may violate the gzk - cutoff .
we explore this possibility within the context of an astrophysical source with a hard ( @xmath235 ) spectrum and cutoff @xmath236 ev sufficient to produce a super - gzk flux .
the results of five models are presented in figure [ fig : cragasa ] with fits to agasa s highest nine energy bins . for hard cr - spectra
grb sources , the cr energetics are not a problem because the high - energy portion of the spectrum receives most of the nonthermal cr energy .
however , it is clear from our fits that if a super - gzk flux is verified by observations with the auger observatory , it calls for either a multi - component model of astrophysical sources ( including nearby sources ) or new physics in the form of top - down ( for recent reviews , see refs .
@xcite ) or hybrid ( for some examples , see refs .
@xcite ) scenarios .
the prediction of our model that the energy output in relativistic protons and nuclei accelerated by relativistic shocks of grbs can be much larger than the total power inferred from the observed x - ray / mev radiation , @xmath237 , suggests a very important conclusion concerning the question of detectability of grbs by forthcoming km - scale neutrino detectors like icecube .
a recent study @xcite of this question shows that , at best , only from the brightest fraction of grbs with neutrino fluence @xmath238 in the prompt phase exceeding @xmath239 erg @xmath28 , would it be possible to detect multi - tev neutrinos by icecube .
this is a general model - independent result which only assumes that the radiation fluence derived from the x - ray / mev @xmath7-ray measurements in the prompt phase of grbs does not strongly underestimate the total radiation output of grbs ( that could be missed at higher energies ) .
this result follows from the direct calculation of the number of muon neutrinos @xmath240 to be expected for a detector like icecube , with an area @xmath241 , assuming a spectral fluence of the neutrinos @xmath242 erg @xmath28 . for a spectrum of @xmath243 with the spectral index @xmath244
( i.e. , with index @xmath245 for the differential number fluence @xmath246 ) , a simple derivation give @xmath247 for @xmath248 , and @xmath249 for @xmath250 ( see @xcite ) .
thus if the neutrino fluence is comparable to the x/@xmath7 photon fluence , detection of @xmath251 neutrinos could be expected only from very powerful grbs , at a fluence level @xmath252 .
note that @xmath253 corresponds to the neutrino fluence @xmath254 erg @xmath28 integrated per each decade of energy .
note also the decline in the number of neutrinos to be expected per equal energy fluences per decade when @xmath255 .
this is explained by the change in the detection efficiency @xmath256 of upward - going muon neutrinos at this energy ( see @xcite ) , which can be approximated as @xmath257 , where @xmath258 for @xmath248 , and @xmath259 for @xmath260 .
therefore , since a typical spectrum of neutrinos expected from grbs due to photomeson interactions of hecrs extends well beyond 100 tev , detection of even a single neutrino at multi - tev energies would become likely only if the total fluence in the @xmath251 would significantly exceed @xmath239 erg @xmath28 .
meanwhile , the total amount of energy released in the neutrinos ( including @xmath261 ) is only about @xmath262-@xmath263 of the energy of secondaries resulting from photomeson interactions .
therefore at least the same amount of energy is released in the gamma - rays and electrons of multi - tev energies produced in the secondary pion decays . for a blastwave lorentz factor @xmath264 - 300 ,
the grb radiation fields in the prompt phase are typically optically thick to photo - absorption for gamma - rays with energies above the gev domain as is apparent from the curves in fig .
even if one assumes that the radiation in the prompt phase of a grb is contributed only by several individual spikes , with a characteristic duration as large as @xmath265 - 10s , the source becomes transparent to gamma - rays below the gev domain only in case of a collapsar grb with doppler - factor @xmath266 .
therefore most of the energy injected in multi - tev gamma - rays will be efficiently converted to hard radiation in the x - ray to sub - gev gamma - ray domains through the synchro - compton photon - pair cascade developing in the relativistic shock / grb source .
our knowledge of high - energy ( @xmath267 mev ) emission from grbs is limited to 7 grbs detected with the spark chamber on egret @xcite and the milagrito detection of grb 970417a @xcite .
the average spectrum of four grbs detected simultaneously with batse and the egret spark chamber from @xmath268 mev to @xmath179 gev is consistent with a @xmath269 photon number spectrum , implying that in these cases the @xmath267 mev fluence does not exceed the batse fluence by more than a factor of a few .
however , the examples of the prompt and delayed high - energy emission from grb 940217 @xcite , the anomalous hard component detected with batse and the egret tasc from grb 941017 @xcite , and grb 970417a , which required @xmath179 times more energy in the tev range than in the batse range , show that the high - energy behavior of grbs is yet poorly measured and even more poorly understood .
the baseline assumption made here is to require that the energy fluence in neutrinos not exceed the fluence observed in electromagnetic radiation at hard x - ray and soft @xmath7-ray energies .
calculations of the neutrino fluxes expected from grbs in case of baryon - loading factor @xmath270 , i.e. , assuming equal energies for relativistic hadrons and directly accelerated electrons ( which are assumed to produce the observed radiation ) , show that even in the case of very powerful bursts with radiation fluence at the level @xmath271 ergs @xmath28 , which happen only few / several times per year , the probability of detection of neutrinos by a km - scale detector would remain hopelessly small in the framework of the collapsar model unless the doppler factor @xmath272 or the variability time scale @xmath273 s. a significant contribution to the target photon field density from a radiation component external to the grb source / blob , as in the plerionic emission in the `` supranova '' scenario @xcite , or from radiation scattered by the progenitor material in the circumburst environment in the collapsar model , improves chances to detect neutrinos from grbs ( see ref .
@xcite for details ) .
this is because for large doppler factors and @xmath274 ergs @xmath28 , only a small fraction of the injected proton energy can be converted into secondaries through photomeson interactions of hecrs with internal photons .
the current model for the origin of galactic high and ultra - high energy crs from grbs requires that the nonthermal baryon loading factor @xmath26 , and this prediction will be tested by the icecube class neutrino detectors . in fig .
[ neu]a we show the neutrino fluxes expected in the collapsar scenario from a burst at the radiation fluence level @xmath275 , calculated for 3 values of the doppler factor @xmath276 assuming a grb source at redshift @xmath277 ( for @xmath278 ) . in order to demonstrate clearly the dependence of the neutrino fluxes on @xmath276 , here we fixed the baryon - loading factor at a value @xmath279 for 3 values of @xmath276 . as in fig .
[ tau ] , for calculations in fig .
[ neu]a we assume that the prompt emission is contributed by @xmath280 spikes with characteristic timescales @xmath281 each , which define the characteristic size ( in the proper frame ) of individual spikes through @xmath282 .
the numbers of ( muon ) neutrinos to be expected by icecube for @xmath283 and 300 are @xmath284 and 0.016 , respectively .
we should note , however , that for the assumed value of @xmath83 , the calculated total fluence of neutrinos ( both @xmath251 and @xmath261 ) produced when @xmath285 is @xmath286 erg @xmath28 , i.e. , by a factor @xmath287 _ larger _ than the assumed radiation fluence .
it means that in the light of the arguments given above , the maximum value of the baryon loading that could be allowed if the high - energy radiation fluence is less than the x/@xmath7 fluence for this particular case should be about 8 - 10 , instead of 20 .
consequently , the expected number of neutrinos for @xmath288 should be reduced to @xmath289 . on the other hand , in fig .
[ neu]a the neutrino fluence for the case @xmath290 is equal to @xmath291 , so this accommodates a baryon - loading increased from @xmath292 up to @xmath293 , with the expected number of neutrinos observed by icecube @xmath294 . if the radiation fluence at mev
gev energies is allowed to exceed the x/@xmath7 fluence by an order of magnitude , a possibility that glast will resolve , then the expected number of detected neutrinos could be increased correspondingly . in order to understand the degree of stability of the predicted numbers of neutrinos against the model assumptions , and most importantly with respect to the assumed number of spikes which determine the characteristic size / compactness of individual `` contributing blobs , '' in fig .
[ neu]b we present the results of calculations of the @xmath251 fluxes assuming @xmath295 , corresponding to @xmath296 . in this case
we chose a higher @xmath297 , anticipating a lower than previously considered fraction of energy extraction from the relativistic protons .
the heavy solid and dashed curves show in fig .
[ neu]b show the fluxes calculated for @xmath285 and 300 , respectively , in the collapsar scenario .
the corresponding total neutrino fluences are equal to @xmath298 and @xmath299 , and the relevant numbers of @xmath251 to be expected for icecube are @xmath300 and @xmath301 . correcting these numbers for the maximum acceptable level of neutrino fluence assuming @xmath302 , results in @xmath303 and @xmath304 for @xmath305 and 300 , respectively , which are indeed close to the numbers derived above in fig .
note that for the latter case , the implied baryon - loading factor would be as high as @xmath306 , which could still be marginally acceptable , as discussed in section 5.1 above .
the two thin lines in fig .
[ neu]b are calculated in the framework of a `` supranova '' scenario for the same values of @xmath285 and 300 , assuming an external radiation field density in the `` plerionic nebula '' with age @xmath307 ( see ref .
@xcite for details of calculations ) . the number of neutrinos to be expected when @xmath285 would be practically the same as for the collapsar model , with @xmath308 .
it is noteworthy , however , that for @xmath309 , the predicted probability of neutrino detection would be greatly increased relative to the collapsar model prediction , namely @xmath310 , which would at the same time demand a significantly smaller value for the baryon - loading factor , with @xmath311 only .
although these numbers are still smaller than 1 , these calculations leave a non - negligible probability for detection of 1 2 , and hopefully up to a few , neutrinos from some of most powerful grbs during a reasonable observation time even in the case of a `` collapsar '' grb with @xmath309 .
note that these neutrinos would be mostly at energies @xmath312 where almost no background neutrinos are to be expected from a given direction to any single grb @xcite .
such a detection would be a confirmation of high baryon - loading in grb blast waves , and would also suggest a very significant contribution of the accelerated hadrons in the observed hard radiation through secondaries produced in the photomeson interactions .
another prediction of a hadronically dominated grb model is the possible detection of gamma - rays in the multi - gev energy domain by glast at a stage near the prompt phase when the density of radiation fields would still be sufficient for extraction of a small fraction of the energy of accelerated protons , but when the secondary gamma - rays would not be suppressed by the photo - absorption process . in case of relatively close grbs , with @xmath313 , these gamma - rays
could then be detectable also by ground - based gamma - ray detectors like veritas and hess with energy thresholds @xmath314 gev .
we have proposed a model where hecrs originate from grbs .
the cr flux near the knee is assumed to result from crs produced by a single grb which has occurred relatively recently and not very far from us in the milky way .
these crs propagate diffusively in the galactic disk and halo .
the simple diffusive propagation model developed in section 3 implies that the measured cr flux results from the modification of the injection spectrum of an impulsive cr source due to transport through a magnetic field with a given turbulence spectrum . using a turbulence spectrum harder at smaller wavenumbers and steeper at larger wavenumebrs , we have fit the 2001 kascade data for cr ion spectra between @xmath315 and 100 pev , and explained the change in the all - particle spectra from @xmath316 to @xmath317 .
a grb releasing @xmath318 ergs in hecrs , located @xmath35 pc away , and occurring @xmath319 years ago , provides reasonable fits to the kascade data . our model of a single grb source making crs at energies through
the knee of the cr spectrum bears some similarity to the single - source model proposed by erlykin and wolfendale @xcite to fit data near the knee .
these authors argue , however , that propagation ( galactic modulation " ) effects can not explain the constant rigidity break of the cr ionic species , whereas we employ a propagation model to produce that break . in this respect
our propagation model treats rigidity - dependent transport as in the model of swordy @xcite and builds upon the detailed study of atoyan , aharonian , and vlk @xcite for the spectral modification effects due to energy - dependent diffusive propagation of crs from a single source .
our model explains , moreover , cr data not only through the first and second knees but also at the highest energies . the turbulence spectrum that fits the cr spectrum near the knee employs a kraichnan spectrum at small wavenumbers and a kolmogorov spectrum at large wavenumbers .
turbulence is thought to be generated at the smallest wavenumbers or largest size scales , with subsequent energy cascading to smaller size scales @xcite .
it is interesting to note that there are two crucial length scales in our turbulence spectrum , namely @xmath320 pc , and @xmath321 pc .
the generation of turbulence in the disk and halo of the galaxy at the larger size scale could be associated with halo - disk interactions ( e.g. , through the interactions of high - velocity clouds with the galactic disk ) , which would deposit turbulence throughout the disk and the halo of the galaxy on a size scale @xmath322 pc .
the smaller length scale is typical of the sedov length scale for supernova explosions in the disk of the galaxy .
indeed , sne would generate a large amount of turbulence energy which could make a distinct contribution to the turbulence spectrum in this wavenumber range .
the origin of the different indices of the two components of the mhd turbulence spectra at small and large wavenumbers could be related to the time available for the turbulence energy injected at the different size scales to cascade to larger wavenumbers .
medium - energy crs with energies between @xmath323
@xmath324 ev / nuc will diffuse by gyroresonant pitch - angle scattering in response to mhd waves with @xmath325/pc .
the model turbulence spectrum at large values of @xmath103 is given by a kolmogorov spectrum with index @xmath111 , as seen in fig.([fig : kwk ] ) . because medium - energy crs are thought to arise from a superposition of many sne , we can treat their transport in the framework of continuous injection . as noted previously , the measured index of crs from continuous sources is steepened by a factor @xmath326 .
if medium - energy crs , whose measured number intensity index is @xmath327 @xcite , result from many sne that produce crs which diffuse through pitch - angle scattering in a spectrum of mhd turbulence with index @xmath111 , then it is necessary that the injection indices of these medium - energy crs lie between @xmath328 and 2.4 .
this is a surprising result , because it is generally thought that the strong shocks in sne accelerate and inject crs with an injection closer to 2.0 than 2.4 @xcite .
such a soft injection spectrum could be avoided if medium - energy crs diffuse in a turbulence spectrum with index @xmath112 . in this case , the measured index is steepened by 0.5 units compared to the injection index . if the large-@xmath103 component with the steeper index is superposed on the extrapolation of the small-@xmath103 component to large wavenmbers , then at sufficiently large values of @xmath103 , the turbulence spectrum will change from a kolmogorov to a kraichnan spectrum , as illustrated by the long - dashed line in fig .
[ fig : kwk ] . in this case
, a continuous injection scenario implies that an injection index of 2.2 yields an observed cr spectrum with a 2.7 number index .
a @xmath11 injection index is in accord with expectations from nonrelativistic shock acceleration .
note that wavenumbers of 10@xmath329 pc@xmath18 are gyroresonant with @xmath330 tev cr protons .
an important feature of recent grb studies is their association with sne .
cr acceleration at sn or grb shocks is crucial for the production of crs from the lowest energies , @xmath331 gev / nuc , to the highest energies @xmath1 ev . the different speeds of the sn shocks in the different types of sne ejecta ranging from relatively slow type ii ( @xmath332 10000 km s@xmath18 ) to marginally relativistic type ib / c is important to produce the full cosmic ray spectrum @xcite .
we therefore predict that hess and veritas , with their improved sensitivity and imaging , will detect @xmath7-ray emission from supernova remnants at a low level unless that sn has also hosted a grb . at energies
@xmath333 ev , crs stream out of galaxies to form the metagalactic cr component .
we assume that grbs evolve with cosmic epoch according to the sfr history of the universe , so that most of the uhecrs are produced at redshift @xmath334 when the sfr is greatest .
the uhecrs have their spectrum attenuated by photo - pion , photo - pair , as well as adiabatic energy losses which become important for particle energies below @xmath335 ev .
our best fit was found to have spectral index @xmath336 and @xmath337 ev .
for these parameters , a slightly better fit was found for the stronger grb redshift evolution ( `` upper sfr '' ) .
stronger grb evolution contributes more of the extragalactic cr flux over the range @xmath338 ev to @xmath339 ev than the lower sfr case , giving a best - fit value @xmath340 ev .
this is an effect of cr attenuation from photopion and photopair production processes where all of the cr flux produced at @xmath341 ev ( and @xmath334 ) are swept to lower energies .
crucial for deriving these conclusions are the relative calibrations of the kascade and high - res experiments .
our best fit to the uhecr data gives a measure of the local grb luminosity density @xmath19 required in crs .
we find @xmath342 for @xmath343 ev and @xmath344 .
this implies that grbs must be baryon - loaded by a factor @xmath345 if this model for hecrs is correct .
the precise value of @xmath83 varies with doppler factor according to @xmath346 , for a @xmath11 injection spectrum .
nonthermal baryon - loading factors @xmath347 implies that grbs should also be luminous in high - energy neutrinos . in section 6 we calculate the fluence in neutrinos and predict that @xmath348 few neutrinos of energy @xmath349
tev@xmath349 pev may be observable in icecube a few times per year from individual grb explosions , which depends on the radiation fluence measured from grbs at @xmath267 mev gev energies , which is not yet well known . in a collapsar model calculation @xcite
, the predicted number of neutrinos depends sensitively on @xmath276 , because the density of internal synchrotron photons is much larger for smaller values of @xmath276 .
the x - ray flashes @xcite may be dirty fireball " grbs , which are similar to classical long - duration grbs , though with larger ( thermal ) baryon loading and smaller average values of @xmath276 @xcite .
this suggests that the x - ray flashes may be more promising candidates for neutrino detection than grbs with peak photon energies of the @xmath350 flux at several hundred kev energies .
the low statistics of current data at the highest energies make unclear the presence or absence of a super gzk uhecr component .
the hires data appears to be consistent with the gzk effect , thereby favoring an origin of uhecrs in conventional " astrophysical sources .
if the uhecr flux is found to exceed predictions of a gzk cutoff , then this could be evidence of new physics , could indicate the existence of super - gzk sources of cosmic rays within @xmath351 mpc from earth , or both .
our current interpretation of the data is made uncertain by results from the agasa experiment @xcite which reveals a super - gzk flux , in conflict with the hires and fly s eye data @xcite ( see , however , ref .
@xcite , who argue that the discrepancy is not serious given the uncertainties in calibration and the statistical variance of cr fluxes measured at such high energies ) .
the possibility of new physics associated with a super - gzk cr flux has generated much interest .
scenarios for a super - gzk flux include the z - burst scenario @xcite , magnetic monopole primaries @xcite , and the decay of supermassive relic particles@xcite .
( for a recent review , see ref.@xcite . )
another approach to generate super gzk fluxes is from astrophysical sources distributed within a gzk distance from earth , @xmath352 mpc .
the auger detector should resolve this question by measuring @xmath353 events per year above @xmath58 ev . in conclusion
, we have investigated the hypothesis that grbs are the sources of hecrs .
our model provides a unified source - type to fit all the cr spectri , from some minimum cr energy produced in grbs , which could be somewhere @xmath354 ev , to a maximum energy @xmath1 ev .
the total cr flux that we calculate is a superposition of crs originating from a past grb in our galaxy in whose hecr halo we inhabit ( with a gradual transition at lower energies to crs contributed from ordinary sn events with sub - relativistic blast waves ) , and @xmath355 ev crs and uhecrs orginating in extragalactic grbs and grbs at cosmological distances .
the hecr all - particle spectrum and our model fit to this data are shown in fig .
[ fig : craps ] .
a grb model for hecr production requires that grb blast waves are hadronically dominated by nearly 2 orders of magnitude .
this could be related to the injection process in relativistic shocks and the large proton / electron mass ratio .
this has important consequences for 100 mev
tev @xmath7-ray emission and high - energy ( @xmath267 tev ) neutrino detection
. allowed hadronic production in grb blast waves will make @xmath7 rays of comparable fluence as the neutrino fluence .
insisting that the high - energy radiation fluence from a photomeson cascade can only reach the level of the x/@xmath7 emission measured with batse ( unless it is missed by detectors operating at larger energies ) , we have shown that @xmath356 neutrino could be detected coincident with a grb at the fluence level @xmath357 ergs cm@xmath18 by icecube .
if glast shows that grbs are much more fluent at @xmath267 mev energies than at x/@xmath7 energies , then in these grbs one could expect a few high - energy neutrinos .
detection of even 1 or 2 neutrinos from grbs with icecube or a northern hemisphere neutrino telescope would unambiguously demonstrate the high nonthermal baryon load in grbs .
high - energy neutrino detection , especially from grbs with bright 100 mev
gev emission components , or from grbs with small lorentz factors or large baryon - loading such as the x - ray bright grbs , would provide compelling support for this scenario for the origin of the cosmic rays . 0.2 in we thank the referee for a helpful report .
the work of s.d.w . was performed while he held a national research council research associateship award at the naval research laboratory ( washington , d.c ) .
the work of c.d.d . is supported by the office of naval research and nasa _ glast _ science investigation grant dpr # s-15634-y .
a.a . acknowledges support and hospitality during visits to the high energy space environment branch of the naval research laboratory .
the larmor radius of a uhe proton is @xmath359 kpc for a proton of energy @xmath360 ev propagating through a magnetic field of strength @xmath361 .
we approximate the uhecrs as propagating rectilinearly where their spectrum is modified by energy losses . here
we describe the three attenuation processes included in our propagation code : 1 ) red - shifting of the proton momentum , 2 ) pion production from proton - cmbr scattering ( @xmath362 ) , and 3 ) electron - pair production from proton scattering on the cmbr ( @xmath363 ) .
energy loss from proton scattering on cmbr is given by -= _ _ th/2_p^d _ _ th^2_pd_r _ p(_r)k_p(_r)_r [ eq : eloss ] @xcite , where @xmath365 is the black - body photon distribution of the cmbr with red - shift dependent temperature @xmath366 , @xmath367 is the cross - section for the process , and @xmath368 is the inelasticity .
the mean - free - path between scatterings is found from eq .
( [ eq : eloss ] ) by setting @xmath369 for all @xmath370 we approximate photo - pion production following ref .
@xcite , modified to include three step cross - section and inelasticity functions _ p(_r)= and k_p(_r)= with increasing @xmath371 , the three regions correspond to 1 ) direct pion production at threshold , 2 ) the enhanced cross - section at the @xmath372-resonance , and 3 ) multi - pion production in the diffractive scattering regime where photon coupling to the vector mesons @xmath373 and @xmath374 dominate . in comparison with the attenuation calculations of mcke et al .
@xcite and stanev et al .
@xcite , we agree to @xmath375 2 ( 4)% at @xmath376 ev .
our treatment of the photo - pair production losses follows that of chodorowski et al .
the greatest photopair energy - loss rate is found at @xmath377 ev when @xmath378 , and pair - production dominates other energy losses over the range @xmath379 our uhecr flux calculation follows the formalism of berezinsky and grigoreva @xcite for diffuse sources .
the flux from a volume element @xmath380 observed at earth is dj(e)de=. where @xmath381 is the comoving source density described in eq .
( [ eq : evol ] ) and fig .
[ fig : madau ] , @xmath382 is the number of crs produced per unit energy per unit time for an injection spectrum with power - law index @xmath185 and exponential cutoff energy @xmath383 the hubble length @xmath384 mpc , and @xmath385 is the proton energy at generation which is subsequently measured with an energy @xmath386 .
we use @xmath387 and @xmath388 to obtain j(e)= f(e , e_max)_0^z_maxdz n(z ) ( ) ^-p || , [ eq : flux ] where the term @xmath389 transforms the energy interval from propagation losses and we take @xmath390 .
results of calculating eq .
( [ eq : flux ] ) for several different spectra and evolution models are found in figs .
[ fig : cr2 ] , [ fig : cr3 ] , [ fig : cr4 ] , and [ fig : cragasa ] .
a. w. blain et al . , mon . not .
r. astron .
soc . 309 ( 1999 ) 715 .
c. gronwall , in t. x. thuan et al . , eds . , dwarf galaxies and cosmology , proc
xviii moriond meeting ( astro - ph/9806240 ) .
m. a. treyer et al . ,
mon . not .
r. astron .
soc . 300 ( 1998 ) 303 .
l. tresse and s. j. maddox , astrophys .
j. 459 ( 1998 ) 691 .
s. j. lilly , o. lefevre , f. hammer , and d. crampton , astrophys .
j. 460 ( 1996 ) l1 .
a. j. connolly et al .
, astrophys .
j. 486 ( 1997 ) l11 .
p. madau et al .
, mon . not .
r. astron .
soc . 283 ( 1996 ) 1388 . m. pettini et al .
, astrophys .
j. 508 ( 1998 ) 539 .
s. t. scully and f. w. stecker , astropart .
phys . 16 ( 2002 ) 271 .
v. s. berezinskii , a. z. gazizov , and s. i. grigorieva , ( hep - ph/0204357 ) .
m. drees , pascos 03 ( hep - ph/0304030 ) . l. anchordoqui et al .
j. mod .
a18 ( 2003 ) 2229 .
t. j. weiler , astropart .
phys . 11 ( 1999 ) 303 .
t. w. kephart and t. j. weiler , astropart .
phys . 4 ( 1996 ) 271
. s. d. wick , t. w. kephart , t. j. weiler , and p. l. biermann , astropart .
phys . 18 ( 2003 ) 663 .
v. s. berezinskii , s. i. grigoreva , astron .
astrophys . 199 ( 1988 ) 1 ; ( e ) 210 ( 1989 ) 462 . a. m. atoyan and c. d. dermer , astrophys . j. 586 ( 2003 ) 79 .
a. mucke , j. p. rachen , r. engel , r. j. protheroe , and t. stanev , pub .
australia 16 ( 1999 ) 160 .
m. j. chodorowski , a. a. zdziarski , and m. sikora , astrophys .
j. 400 ( 1992 ) 181 . | a model is proposed for the origin of cosmic rays ( crs ) from @xmath0 ev / nucleon to the highest energies ( @xmath1 ev ) .
grbs are assumed to inject cr protons and ions into the interstellar medium of star - forming galaxies including the milky way with a power - law spectrum extending to a maximum energy @xmath2 ev .
the cr spectrum near the knee is fit with crs trapped in the galactic halo that were accelerated and injected by an earlier galactic grb .
these crs diffuse in the disk and halo of the galaxy due to gyroresonant pitch - angle scattering with mhd turbulence in the galaxy s magnetic field .
the preliminary ( 2001 ) kascade data through the knee of the cr spectrum are fit by a model with energy - dependent propagation of cr ions from a single galactic grb .
ultra - high energy crs ( uhecrs ) , with energies above the ankle energy at @xmath3 ev , are assumed to propagate rectilinearly with their spectrum modified by photo - pion , photo - pair , and expansion losses .
we fit the measured uhecr spectrum assuming comoving luminosity densities of grb sources consistent with possible star formation rate histories of the universe . for power - law cr proton injection spectra with injection number index @xmath4 and low and high - energy cutoffs , normalization to the local time- and
space - averaged grb luminosity density implies that if this model is correct , the nonthermal content in grb blast waves is hadronically dominated by a factor @xmath5 - 200 , limited in its upper value by energetic and spectral considerations .
calculations show that 100 tev 100 pev neutrinos could be detected several times per year from all grbs with kilometer - scale neutrino detectors such as icecube , for grb blast - wave doppler factors @xmath6 .
glast measurements of @xmath7-ray components and cutoffs will constrain the product of the nonthermal baryon loading and radiative efficiency , limit the doppler factor , and test this scenario .
gamma ray bursts , cosmic rays , gamma rays , neutrinos |
image potential states are single - electron excitations trapped outside the crystal surface of a metal by the attractive interaction between the electron and the polarizable bulk @xcite . at large distances
@xmath0 from the surface the interaction is governed by the classical electrostatic image potential which scales in the limit of an infinitely extended flat surface as @xmath1 .
in contrast , at short distances ( a few bohr radii ) the specific microscopic material properties and many - body interactions become important @xcite . on a structural as well as on a formal level ,
this situation is reminiscent of the physics of the outer valence electron in an alkali atom .
accordingly , the spectrum of image states above flat surfaces can be described very accurately by a rydberg series with adapted quantum defects @xcite .
experimentally observed image states localize typically 10 to 100 bohr radii above the surface with lifetimes that may extend to several ps @xcite .
studying the dynamics of these states via time - resolved two - photon photoemission spectroscopy has led to an improved understanding of scattering processes occurring at surfaces @xcite .
in contrast to image states above flat surfaces , tubular image states ( tis ) occur around curved surfaces like metallic nanotubes and are characterized by an angular momentum number @xmath2 @xcite . for high angular momenta ( typically @xmath3 ) the interplay between the repulsive centrifugal barrier and the long - range part of the image potential forms stable radial potential wells that support detached image states localized at distances of 10 - 50 nm from the surface .
this trapping mechanism leads to enhanced lifetimes of these states which have been estimated to 1 ns-1 @xmath4s @xcite . since their prediction in 2002 @xcite tis
have attracted much interest which led to numerous subsequent studies .
this includes on the one hand explorations of novel setups like tis at finite segmented nanowires @xcite , parallel nanowires @xcite , nanowire lattices @xcite and fullerene buckyballs @xcite .
on the other hand , this comprises also refined studies of the image force at the short - range scale for particular microscopically modeled semiconducting or conducting surfaces @xcite . while low-@xmath2 tis have been observed experimentally via two - photon photoemission at free standing multi - walled carbon nanotubes @xcite and via scanning tunneling microscopy / spectroscopy at supported double - walled carbon nanotubes @xcite , an experimental detection of high-@xmath2 tis has not yet been reported . in the present paper
we analyze high-@xmath2 tis around nanorings .
these tis resemble those in ref .
@xcite but are trapped at curved nanowires having both ends connected . to this aim we construct an exact expression for the electrostatic image potential and analyze subsequently electronic eigenstates bound by the competition between the image force and a repulsive centrifugal force .
although the angular momentum around the wire is not conserved we show that characteristic properties of these states can be derived from an effective radial potential .
the paper is organized as follows . in sec .
[ sec : image_potential ] we present the underlying electrostatic problem . in sec .
[ sec : electrostatic_problem ] the image potential is set up in toroidal coordinates . in sec .
[ sec : image_pot_analysis ] we analyze an exemplary nanoring and point out general scaling properties of the image potential .
[ sec : potential_states ] deals with image states trapped in the potential . in sec .
[ sec : finite_diff ] we describe our numerical approach which is employed in sec .
[ sec : image_states_results ] to compute image states .
finally , the states are analyzed and interpreted by means of an effective radial potential in sec .
[ sec : effective_radial_potential ] .
our conclusions are provided in sec .
[ sec : conclusions ] .
we consider an electron at position @xmath5 in the proximity of a conducting nanoring whose surface is parametrized as a torus with major radius @xmath6 and minor radius @xmath7 for @xmath8 , see fig .
[ fig : torus ] .
if the position of the electron is fixed , the total electrostatic potential at any point @xmath9 outside the ring can be expressed as @xmath10 where @xmath11 is the coulomb potential and @xmath12 is the induced potential which results from the polarization of the nanoring ( for convenience we work here in atomic units ) . the boundary conditions for an ideal conductor require that the total electrostatic potential is constant for all points @xmath13 lying on the surface of the conductor , i.e. @xmath14 . for neutral , i.e. non - charged , conductors this constant @xmath15 is a priori not known and needs to be determined by employing the additional condition that the induced potential @xmath12 is source - free , in contrast to grounded nanostructures ( @xmath16 ) that have in general non - source - free @xmath12 . here
we focus exclusively on neutral nanorings . in this case
one can show that the energy needed to separate the electron and the nanostructure is given by @xmath17 , which is called the image potential @xcite . as for metallic cylinders @xcite , the image potential for tori
can be expressed analytically , too . to this aim
we introduce toroidal coordinates @xmath18 , @xmath19 , @xmath20 which are linked to cylindrical coordinates @xmath21 , @xmath0 , @xmath20 via @xmath22 and @xmath23 with @xmath24 . in the new toroidal
coordinates the torus surface is defined by @xmath25 with @xmath26 . employing these relations ,
the image potential of the neutral torus can be constructed by combining classical electrostatic results obtained for conducting tori in different external field configurations @xcite and reads @xmath27 \label{eqn : vim_torus_neutral}\ ] ] where @xmath28 and @xmath29 ^ 2 .
\label{eqn : sum2}\ ] ] here @xmath30 , @xmath31 , @xmath32 and @xmath33 are associated legendre functions and @xmath34 with the kronecker delta @xmath35 .
the detailed derivation of @xmath36 is given in appendix [ sec : imagepotential_derivation ] . while the sum @xmath37 converges for all @xmath38 , the second sum @xmath39 diverges for @xmath40 .
this agrees well with the fact that the image potential of any ideal conductor will become arbitrarily large for a decreasing distance to its surface . to evaluate ( [ eqn : vim_torus_neutral ] )
numerically we check the convergence of the sums @xmath37 and @xmath41 at each coordinate @xmath42 by increasing the range of summation until a desired precision is achieved . as an illustrative example we consider the image potential of a ring with @xmath43 ( @xmath44 ) and @xmath45 ( @xmath46 ) .
this length @xmath7 as well as the circumference @xmath47 are comparable to the dimensions of single and multi - walled nanotubes with radii ranging from 0.68 nm to 7.1 nm and lengths ranging from 200 nm to 1600 nm which have been studied in the context of tis @xcite . in fig .
[ fig : vim_torus_3d ] we depict the resulting image potential in cylindrical coordinates .
the plot illustrates the attraction of the electron towards the torus surface which is a circle of radius @xmath7 in the @xmath21-@xmath0 plane .
close to the surface the potential becomes nearly isotropic with respect to the tube center @xmath48 due to its divergent behavior . at larger distances from the surface , however , the anisotropy is clearly visible .
general properties of nanorings with different parameters @xmath6 and @xmath7 can be illustrated by defining the rescaled quantities @xmath49 , @xmath50 and @xmath51 .
the resulting rescaled potential @xmath52 depends then only on the aspect ratio @xmath53 .
radial cuts of this rescaled potential are presented in fig .
[ fig : vrescale ] for several @xmath53 .
as can be seen the ratio @xmath53 , firstly , determines the positions of the torus surface @xmath54 where the potential curves diverge , and , secondly , it controls the asymmetry around @xmath55 , which becomes clearly visible for the largest @xmath56 . what is not obvious from fig .
[ fig : vrescale ] is that also the scaling properties of the image potential depend significantly on @xmath53 . as we will discuss in the next section ,
the scaling is crucial for the capability of nanorings to support detached tis . to analyze this property we compare the nanoring potential to a power - law with exponent @xmath57 which is typical for many image potentials , e.g. at flat surfaces ( @xmath58 ) or at grounded metallic spheres ( @xmath59 ) . for simplicity
we restrict this analysis here only to a radial cut of the potential in the region @xmath60 . by introducing the rescaled radial distance of the electron to the tube center
@xmath61 the potential can be rewritten as @xmath62 .
the local exponent of the potential is then defined by @xmath63 which is the slope in a double logarithmic plot of @xmath64 .
this exponent is depicted in fig .
[ fig : exponent ] for different ratios @xmath53 .
all curves converge for large @xmath65 to @xmath66 , which is the expected long - range scaling for the interaction of a point charge and a polarizable neutral finite object ( monopole - induced - dipole - interaction ) . at short distances
@xmath67 diverges as expected for @xmath68 . in between these limits
there is an intermediate regime where the exponent approaches a maximal value .
the smaller the ratio @xmath53 the larger is this maximal value , e.g. one has @xmath69 if @xmath70 .
in other words , only sufficiently thin nanorings have spatial regions where the radial image potential decreases slower than @xmath71 .
although electronic excitations at metal surfaces are in general a complex many - body phenomenon , image potential states have been described in many situations very successfully within a simple one - electron picture @xcite . in this model one
solves the schrdinger equation of the single outer electron subject to the image potential .
as the image potential expresses only the electrostatic energy of the system it is clear that dynamical corrections are required when retardation effects or collective excitations of the metal ( e.g. plasma oscillations ) become important @xcite .
the nanoring image potential @xmath72 is rotationally symmetric around the @xmath0-axis and we solve the schrdinger equation in cylindrical coordinates .
the stationary solution with energy @xmath73 can be written as an eigenstate of the angular momentum operator @xmath74 @xmath75 with angular momentum @xmath76 .
the separation of the angular degree of freedom leads then to the two - dimensional stationary schrdinger equation @xmath77u_{m , n}(\rho , z)=0 \label{eqn : schrod_eqn_2d}\ ] ] for the wave function @xmath78 with an effective potential @xmath79 that includes a centrifugal barrier depending on @xmath76 .
the second quantum number @xmath80 thus labels the different eigenstates and energies for fixed @xmath76 .
( [ eqn : schrod_eqn_2d ] ) formally resembles a two - dimensional schrdinger equation in cartesian coordinates with the additional restriction that @xmath21 needs to be positive .
we compute the energetically lowest eigenstates by representing @xmath78 as a vector on an equidistant rectangular grid with @xmath81 gridpoints @xmath82 for @xmath83 . in this representation @xmath84
becomes a diagonal matrix and the derivative operators @xmath85 and @xmath86 can be expressed as sparse matrices by employing the finite - difference approximation ( for our calculations we use a 7-point - stencil for each dimension ) . to compute detached tis with this method ,
two modifications of the original physical system are necessary .
firstly , we impose hard - wall boundary conditions at the borders of the grid which have no effect on bound states if the grid is chosen large enough ( see below ) .
secondly , we introduce a cut - off energy @xmath87 to regularize the divergence at the surface and replace @xmath72 by @xmath87 if @xmath88 . among all
computed states @xmath89 we then select states which are insensitive to the cut - off as well as to the boundary conditions .
this method works well for wave functions localized in regions with @xmath90 but it should be mentioned for completeness that hard - wall boundary conditions may in general be inappropriate for states having @xmath91 which are allowed to have finite values @xmath92 for @xmath93 and which are therefore not part of our analysis . to illustratively exemplify the existence of tis at nanorings we employ our computational approach to determine image potential states at the nanoring with @xmath43 and @xmath94 .
we choose @xmath95 , the grid is built up by @xmath96 grid points and we set @xmath97 such that the image potential is cut off only in regions closer than 2 bohr radii to the surface . in fig .
[ fig : denplots_torus ] we show the ten states having the lowest integrated probability to be found in the inner torus region ( @xmath98 ) .
this probability is on the order of @xmath99 .
these states have a pronounced circular symmetry around the tube center @xmath48 and resemble in the transversal @xmath21-@xmath0-plane strongly high angular momentum states around nanotubes @xcite .
the main differences are , however , that the circular symmetry is not perfect ( as can be clearly seen from the higher excited states in fig .
[ fig : denplots_torus ] ) and that the tis are periodically extended along the toroidal @xmath20-direction ( not visible in fig .
[ fig : denplots_torus ] ) . to quantify the circular motion we define the angular momentum operator @xmath100 $ ] which is the generator of rotations along the so - called poloidal direction ,
[ fig : torus ] . for each state fig .
[ fig : denplots_torus ] contains the expectation value of the squared angular momentum in terms of an effective angular momentum quantum number @xmath101 .
the numerical accuracy of @xmath102 depends on our spatial resolution ( here @xmath103 ) and can for this example be estimated to @xmath104 . for the presented states the number @xmath102 lies always close to an integer value counting the number of maxima of the electronic wave function in the @xmath21-@xmath0-plane , i.e. half of the maxima in the density plots .
states having the same number of radial and angular nodes appear as energetically nearly degenerate pairs ( differences below 1@xmath105ev ) and have a different parity along the @xmath0-axis . in the limit of an exact circular symmetry
the degeneracy would become exact and @xmath102 would become an integer . the above analyzed states
correspond hence to a regime where the image potential experienced by the electron is nearly circularly symmetric .
this situation is comparable to the case of two parallel ( infinitely extended ) nanotubes sharing an electronic image state which has been analyzed in @xcite . for sufficiently large tube separations
all image states are gerade or ungerade superpositions of single - tube states and can be labeled by a good angular momentum quantum number
. the situation changes , however , for smaller tube separations where an increasing asymmetric distortion leads to a collapse of the electronic states onto the tubes .
a similar effect can be observed in the nanoring system for an increasing toroidal angular momentum @xmath76 .
this is illustrated in fig .
[ fig : m_spec ] where we follow the energies and densities of the six energetically highest tis presented in fig .
[ fig : denplots_torus ] .
the densities can be compared to fig .
[ fig : denplots_torus ] and illustrate the impact of the additional centrifugal potential . as can be seen an increasing @xmath76
distorts the circular symmetry of the tis . for high @xmath76
this distortion destabilizes the system such that the wave functions are less well - separated from the surface but also less well bound to the nanoring , see fig .
[ fig : m_spec ] @xmath106 . to a good approximation
the energy shift with respect to @xmath76 of the detached tis @xmath78 can be described by the parabola @xmath107 which is shown in fig .
[ fig : m_spec ] @xmath108 .
this is the expected behavior based on eq .
( [ eqn : v_eff ] ) in the limit that the tis localizes sufficiently well at a mean distance @xmath6 from the coordinate center and that the state dependent offset @xmath109 can be approximately replaced by @xmath110 . in a regime where the image potential is almost circularly symmetric around the tube center @xmath111 , it is beneficial to analyze instead of the full image potential
the effective radial potential @xmath112 @xmath113 includes an additional centrifugal barrier associated to the angular momentum @xmath2 .
it can be viewed as an approximation for the potential energy associated to the degree of freedom describing radial motion with respect to the tube center .
this approximation becomes accurate for states localized sufficiently close to the tube surface where the potential @xmath114 is circularly symmetric . in a more sophisticated treatment small anisotropies
could be taken into account by separating the circular and radial motion adiabatically and analyzing instead of ( [ eqn : veff_rad ] ) a resulting set of appropriately circularly averaged adiabatic potential curves . in fig .
[ fig : veff ] the radial potential from ( [ eqn : veff_rad ] ) is depicted for the exemplary nanoring ( @xmath43 , @xmath115 ) for an angular momentum @xmath95 and effective angular momenta @xmath2 ranging from 5 to 12 . in the figure
all curves with @xmath116 possess potential wells that are separated by centrifugal barriers from the nanoring surface . as expected from the ring geometry these wells
are non - symmetric with respect to reflections at the axis @xmath117 .
further asymmetry stems from the centrifugal potential associated to @xmath76 which leads to the divergent behavior for @xmath93 .
the properties of the wells can be compared to the exact solutions from fig .
[ fig : denplots_torus ] . for example , the curve with @xmath118 possesses potential wells at @xmath119 and @xmath120 with depths of approximately 1 mev .
these values agree well with the energies and the radial extent of the four tis possessing @xmath121 in fig .
[ fig : denplots_torus ] .
their radial wave functions can be interpreted as the lowest and first exited radial states in these wells .
similar agreement can be found for the other states in fig .
[ fig : denplots_torus ] .
furthermore , the radial potential predicts that tis with @xmath122 and @xmath123 are either not bound to the ring or are not detached from its surface which agrees with our numerical results . in a regime where the circular and the radial motion are approximately separable
, the effective potential @xmath113 is consequently a useful tool to identify angular momenta @xmath102 leading to detached tis and to specify their properties . in the following
we will employ @xmath113 to estimate general criteria for the existence of detached tis at nanorings in a regime where the impact of @xmath76 is small . a general framework to study the conditions for which infinite metal cylinders support
detached tis has been provided in @xcite . in particular
it has been shown that an infinite cylinder of radius @xmath7 possesses only detached effective potential wells if the dimensionless parameter @xmath124 lies close to @xmath125 , where @xmath126 denotes the bohr radius .
this is a useful criterion to estimate properties for cylinders having different geometries . by performing a similar scaling analysis for the potential ( [ eqn : v_eff ] ) one can derive analogous criteria for the nanoring .
for the exemplary nanoring one finds e.g. that potential wells exists for @xmath127 and one can conclude that this is also true for all other nanorings having the same ratio @xmath128 .
we computed critical @xmath129 also for nanorings with other ratios @xmath130 and found @xmath129 lying always between 0.7 and 1.1 . for larger
@xmath53 , however , the potential wells become so broad , that their identification becomes increasingly difficult .
this behavior can be understood by means of the local power - law scaling of the nanoring image potential that depends on the ratio @xmath53 , see sec .
[ sec : image_pot_analysis ] .
a basic calculation shows that an effective potential of the form @xmath131 , where @xmath132 , @xmath133 and @xmath57 are constants , possesses minimal extremal values only for @xmath134 .
this is for example the reason why there exist no high angular momentum image states around spherical graphite particles @xcite .
this criterion can also be applied locally and we conclude from fig . [
fig : exponent ] for the nanoring that potential wells supporting detached tis exist only if @xmath135 , i.e. for sufficiently thin tori .
in the present work we constructed an analytical expression for the electrostatic image potential between a point charge and a neutral metallic ring .
we analyzed this potential and demonstrated that it supports image states `` circulating '' the torus which can be viewed as a curvilinear generalization of tubular image states around nanocylinders @xcite .
in contrast to the latter the angular momentum @xmath2 shielding the states from the torus surface is not conserved and the states are not circularly symmetric . the second angular momentum @xmath76 ( associated to the motion along the nanoring ) is conserved and can be used to tune the energy and also to some extent the shape of the image states .
too high @xmath76 will , however , couple the different @xmath2 states too strongly and destabilize the system . the main features of tis can be explained by means of an effective radial potential including a centrifugal barrier associated to @xmath2 . by analyzing the scaling properties of this potential we argued that centrifugally detached high angular momentum image states exist only around sufficiently thin nanorings having aspect ratios smaller than @xmath136 .
however , it has to be expected that nanorings ( like nanotubes ) also support low angular momentum tis states probing the surface , whose investigation might be an interesting subject for future studies . in the last decades
there has been a tremendous progress in the development of nanofabrication techniques @xcite and the production of almost perfectly ring - shaped silver nanowires with dimensions @xmath137 and @xmath138 is already possible @xcite .
compared to experimentally realized tis around straight cylindrical structures the additional curvature of nanorings could offer interesting new applications .
for example the repulsive mutual coulomb interaction of two or more simultaneously excited electrons at a nanoring could lead to special equilibrium configurations where each electron resides at a particular angle along the ring .
the here presented results are a first exploration of tis at curved nanowires and should be of relevance for all future studies and applications of image states at even more complex - shaped nanostructures .
c.f . gratefully acknowledges a scholarship by the studienstiftung des deutschen volkes .
our aim is to construct the electrostatic potential @xmath139 outside a neutral metallic torus in the presence of an electron at position @xmath5 .
this setup is very similar to the problem of a point charge in the proximity of an isolated , neutral and conducting sphere @xcite and can be solved by a related approach . to this aim
we construct @xmath140 as a linear superposition of the form @xmath141 where @xmath142 is the potential around a grounded metallic torus in the presence of the electron and @xmath143 is the particular potential of a charged metallic torus with the constant surface potential @xmath144 .
the factor @xmath145 needs to be determined from the condition that the total charge in the inner torus region vanishes , i.e. that the induced potential is source free .
the solution @xmath142 has been presented in @xcite .
it can be written as @xmath146 where the induced potential reads in toroidal coordinates @xmath147 \cos \left[n(\eta-\eta_e)\right]}{p_{n-1/2}^m(\cosh \xi_0 ) } \nonumber \\ & \cdot
q_{n-1/2}^m(\cosh \xi_0 ) p_{n-1/2}^m(\cosh \xi_e ) p_{n-1/2}^m(\cosh \xi ) .\end{aligned}\ ] ] here @xmath148 , @xmath149 , @xmath150 are the toroidal coordinates of the electron .
for large @xmath151 , i.e. @xmath152 , one finds @xmath153 by comparing this expression to the multipole expansion @xmath154 and by substituting ( [ eqn : sum1 ] ) one can identify the induced positive monopole @xmath155 this monopole needs now to be compensated by the potential @xmath156 .
the therefore required solution @xmath157 is stated in @xcite @xcite . by employing whipple s transformation for legendre functions and ( [ eqn : sum1 ] )
it can be expressed in toroidal coordinates as @xmath158 from the long - range behavior of ( [ eqn : potential_insulated_charged_torus ] ) one can derive the capacitance of the torus @xmath159 which depends only on the geometry of the torus and links the induced charge to the surface potential @xcite . due to the linearity of the laplace equation
the solution @xmath156 corresponds to a torus with surface potential @xmath145 carrying a charge @xmath160 .
consequently , by setting @xmath161 one obtains with ( [ eqn : pot_torus_ansatz ] ) and ( [ eqn : potential_insulated_charged_torus ] ) the total potential @xmath162 the image potential is consequently given by @xmath163 \ .\end{aligned}\ ] ] | we study electronic image states around a metallic nanoring and show that the interplay between the attractive polarization force and a repulsive centrifugal force gives rise to rydberg - like image states trapped several nanometers away from the surface .
the nanoring is modeled as a perfectly conducting isolated torus whose electrostatic image potential is derived analytically .
the image states are computed via a two - dimensional finite - difference scheme as solutions of the effective schrdinger equation describing the outer electron subject to this image potential .
these findings demonstrate not only the existence of detached image states around nanorings but allow us also to provide general criteria on the ring geometry , i.e. the aspect ratio of the torus , that need to be fulfilled in order to support such states . |
since the first detection of pre - main sequence binaries by joy & van biesbroeck ( 1944 ) and herbig ( 1962 ) , major advances have been gained in our understanding of the formation and early evolution of binaries .
one of the key results is that binaries are about twice as common in t tauri associations as among field stars ( e.g. , reipurth & zinnecker 1993 , simon et al .
1995 , duchne 1999 , ratzka et al .
studies of embedded sources suggest that multiple systems may possibly be more prevalent among class i sources than among the more evolved t tauri stars ( e.g. , duchne et al . 2004 ) .
dynamical evolution among higher - order multiples is likely to be important during the embedded phase , leading to the occasional ejection of lower - mass members and the formation of powerful jets ( reipurth 2000 ) .
the orion nebula cluster ( onc ) is the nearest ( d@xmath5450 pc ) star forming cluster , with an age of less than 1 myr , and it has been extensively studied at multiple wavelengths ( e.g. , herbig & terndrup 1986 , hillenbrand 1997 , getman et al .
2005 and references therein ; for earlier work see the review by herbig 1982 ) .
it is well known that the massive stars in the trapezium have a very high binary frequency ( e.g. , preibisch et al .
binarity of the low - mass population of the central part of the onc has been studied extensively , and it has been suggested that the binary frequency is lower than for the dispersed low - mass young population in associations and comparable to the binary frequency in the field ( prosser et al .
1994 , padgett et al .
1997 , petr et al .
1998 , simon et al . 1999 ,
khler et al . 2006 ) . in this paper
, we report the results of a major binary survey of an extensive region centered on the onc , based on a large h@xmath0 imaging survey using the hubble space telescope .
we have detected 78 multiple systems , of which 55 are new discoveries , in an area of 412 arcmin@xmath6 that excludes a circular area with radius of 60@xmath1 centered on @xmath2 ori c. we discuss the membership of these binaries in the onc , analyze their binary properties , and discuss the binary separation distribution function and binary formation in the onc .
the present survey is based on our recent study of the orion nebula using the wide field camera of the advanced camera for surveys onboard hst . during program go-9825
, we observed 26 acs fields with the f658n ( h@xmath0+[nii ] ) filter and with an exposure time of 500 sec per pointing . in total
our h@xmath0 survey covers an area of 415 arcmin@xmath6 of the central orion nebula . due to the high stellar density near the trapezium
we have excluded a circular area with radius of 60@xmath1 centered on @xmath2 ori c , so the area investigated is 412 arcmin@xmath6 .
the exquisite resolution of the hst combined with the 0@xmath305 pixel - size of the acs offers a unique opportunity to detect close binaries .
we examined all stellar sources in the images by eye at a range of contrast levels , allowing us to pick up faint as well as bright companions .
thanks to the very stable point - spread function across the images , this could be done in a rather homogeneous fashion .
we discuss the completeness of this procedure in sect .
we used reduced images that were combined through multidrizzle .
the faintest star we measured had a magnitude of 21.5 in the f658 filter .
pixel coordinates for both primaries and secondaries were determined with 2-dimensional gaussian functions , and separations and position angles calculated from these coordinates .
relative fluxes were determined using aperture photometry with corrections for the sky background . for further details of the observations , see bally et al .
the region of the onc we have studied is shown in figure 1 , which outlines the mosaic of 26 acs fields .
one field had to be slightly turned to acquire a guidestar .
we examined all stars ( 1051 ) in these fields with the exception of stars within an exclusion zone of 60@xmath1 around @xmath2 ori c. all previous hst studies of binarity in the onc were done in the region around the trapezium , where line - of - sight pairs due to the high stellar density is a major issue , and thus focused mostly on sub - arcsec binaries .
in contrast , our study covers a much larger area including the outskirts of the onc .
we have also excluded the inner region with radius of 60@xmath1 , so contamination is much less of an issue .
based on our observed star density , we can calculate the fraction of our binaries that are not physical pairs for different separation limits . from this
we have chosen an upper separation limit of 1@xmath35 , which implies an 11% contamination , as described below .
we have tested our completeness of companions by blindly `` observing '' real stars with artificially added companions ( scaled from real stars in the images ) at random separations , position angles , and flux ratios .
we find that our detections are essentially complete down to separations of 0@xmath31 , a result made possible by the very sharp point spread function of the hst / acs system .
figure 2 shows two panels with plots of @xmath7 m vs. separation in arcseconds . the left panel shows the actual observations of the binaries we detected , while the right panel shows our artificial binaries .
filled circles indicate companions we were able to identify , and empty circles indicate those we failed to see .
our detection limit is essentially complete to 0@xmath31 except for the largest brightness differences , but very few binaries are found with large flux differences ( see sect .
4.5 ) . to be certain that we are complete
, we considered binary companions in the more limited separation range from 0@xmath315 to 1@xmath35 .
however , a much more severe problem with completeness is due to the strong and highly structured emission from the surrounding hii region , a problem that is compounded by our use of images taken through an h@xmath0 filter . in some areas , especially near the trapezium
, our ability to detect faint stars is diminished , an effect that is unquantifiable but should be kept in mind .
most binaries that have been found prior to this study were identified using infrared techniques . whereas all binaries we have detected have been checked against lists of previously discovered binaries , we have not made any systematic effort to check if all these earlier binaries are also detectable in the optical . from random checks
it appears that many of these infrared - detected binaries are not visible in our optical images .
the main source for membership information in the onc are astrometric studies , principally the study by jones & walker ( 1988 ) , which covers the inner 20@xmath8@xmath930@xmath8 of the onc and thus includes our area of study .
it is based on photographic i - band plates , which favors red and reddened stars . of the 1051 stars in our images ,
655 stars are listed in the jones & walker catalog , and of these 596 have membership probabilities higher than 93% .
the large majority of our binaries are not resolved in the jw survey , and thus the proper motions can be affected by relative brightness variations of the components that can shift the photocenter .
we have seen such variability in a binary lying in the overlap region between two fields , and thus observed twice .
we have also found stars with clear pre - main sequence characteristics that have membership probability of 0% , indicating that astrometric information can be used to _ support _ membership , but should not be used to _ exclude _ membership .
we are therefore complementing the proper motion membership analysis with other sources , such as x - ray emission , h@xmath0 emission , and variability .
getman et al .
( 2005 ) has determined that 1373 stars from the 1616 sources in the coup survey , which largely overlaps with our field ( see figure 1 ) , are orion members . among our stars
, we find 658 which are detected by coup and also are classified as onc members .
finally , we use the presence of h@xmath0 emission ( partly from the literature and from our own unpublished deep survey ) , as well as irregular variability as listed in the general catalogue of variable stars . a total of 99 stars were found to have h@xmath0 emission .
note that we do not use near - infrared excess as a membership criterion , since our binaries by default are likely to have infrared excesses due to their mostly lower - mass companions . in total
, we find 781 stars that have at least one ( and commonly several ) of these membership characteristics . among the 781 known onc members detected in our hst images
, we found 72 binaries and 3 triples in the range from 0@xmath315 to 1@xmath35 .
these are listed in table 1 , where each triple is listed as two binaries , in the manner of counting of kuiper ( 1942 ) .
the first and second columns list the jones & walker i d and the probability of the star to be a member of the onc based on its proper motion , respectively .
the third column lists the parenago number , the fourth column lists the name in the general catalog of variable stars , the fifth column lists the coup number , and the sixth column lists the number given by hillenbrand ( 1997 ) . the right ascension and declination for equinox 2000 are listed in the seventh and eighth columns .
the ninth column lists the position angle of the system and the tenth column lists its separation in arcsec . the apparent magnitude in the i band ( from hillenbrand 1997 )
is listed in the eleventh column and the difference in magnitudes , in the h@xmath0 filter , between the primary and secondary is listed in the twelfth column .
the position angle towards and separation ( in arcsec ) to @xmath10ori c are listed in columns 13 and 14 , respectively . then follows the spectral type as listed by simbad .
the sixteenth column lists the character of membership for each binary ( p : proper motion ; x : coup onc source ; h : h@xmath0 emission ; v : irregular variable ) .
the last column lists the discovery paper if a binary was known prior to this work .
additionally , we have found 3 binaries among the onc members outside the 60@xmath1 exclusion zone with separations between 0@xmath311 and 0@xmath315 .
they are listed in table 2 .
we have furthermore examined the optically visible onc members within the exclusion zone , and have found another 3 binaries with separations less than 0@xmath34 , an ( arbitrarily chosen ) upper limit we imposed due to the high stellar density in that region .
two of these are new discoveries .
many more binaries have been found by earlier studies of the central region ( e.g. , prosser et al .
1994 , padgett et al .
1997 , petr et al .
1998 , simon et al . 1999 ,
khler et al .
2006 ) , but most are only detectable in the infrared , since most cluster members are hidden by extinction ; these three binaries are the only we could detect in our images in the range 0@xmath31 - 0@xmath34
. we do not include the abovementioned binaries in our statistical analysis , but list them in table
finally , we have searched the 270 stars outside the exclusion zone for which we have no evidence for membership , and have found another 7 binaries . some of these
may be foreground or background stars , but others may turn out to be young stars , and we also list them in table 2 .
the columns in table 2 are the same as in table 1 , except we have added an additional column , which characterizes the binaries .
all of our binaries with known spectral types are late - type stars , presumably classical and weak - line t tauri stars , with the exception of parenago 2149 ( jw 945 ) , which appears to be a herbig ae / be star . in figure 3a ,
b we show figures of the 78 multiple systems we have identified among onc members outside the exclusion zone , as well as the three close binaries ( @xmath110@xmath34 ) found inside the exclusion zone .
each stamp is 2@xmath1 on a side .
some of the binaries are particularly interesting for a variety of reasons , e.g. v1118 ori is a famous exor , other binaries or companions are likely substellar objects , and some are associated with jets or proplyds .
in the appendix , we provide more details about these cases .
could some of the wider binaries in the onc be due to contamination by line - of - sight pairs ?
the surface density of stars in the onc is a steeply declining function of distance from the center of the cluster , with only relatively minor fluctuations due to subclustering ( either real or caused by extinction variations , figure 4 ) .
we have determined the stellar density @xmath12 in an area with radius 30@xmath1 around each of the 781 onc members , and have determined the probability @xmath13 of finding an unrelated star within a distance @xmath14 from each primary using the expression @xmath15 ( correia et al .
2006 ) , where we have set @xmath14 to 1@xmath35 .
the probability @xmath13 is principally a function of the distance from @xmath2 ori c , with smaller variations due to local inhomogeneities in the cluster density .
figure 5 shows the distribution of probability of a line - of - sight association for each of the 781 onc members as a function of distance from @xmath2 ori c. the figure mostly follows the radial stellar density curve in figure 4 , although the effect of local small - scale groupings is visible .
the sum of contamination probabilities for the 781 stars is 911% . in other words , among the binaries detected we are likely to have roughly 9 line - of - sight doubles . in the following we correct for these false binaries .
the detection of 72 binaries and 3 triples with separations in the range 0@xmath315 - 1@xmath35 among the 781 onc members must be corrected for the estimated 9 binaries that are likely to be due to line - of - sight pairing .
counting the three triples as six binaries , we then have 69 physical binaries , which implies a 8.8%@xmath41.1% binary fraction @xmath16 in the interval from 67.5 to 675 au , or a multiplicity frequency ( where the 3 triples are counted as 1 system each ) of 8.5%@xmath41.1% ( here and in the following the error estimates are 1@xmath17 and are derived using poisson statistics ) .
petr et al . ( 1998 ) found 4 binaries in the separation range 0@xmath314 to 0@xmath35 in the inner 40@xmath1@xmath940@xmath1 of the trapezium , corresponding to 5.9%@xmath44.0% . in the same limited range of separations
we find 50 binaries , corresponding to 6.4%@xmath40.9% .
the numbers are consistent within their errors .
reipurth & zinnecker ( 1993 ) observed nearby t tauri associations and found 38 binaries ( of a sample of 238 stars ) in the range 150 - 1800 au .
the range we study statistically is 67.5 - 675 au .
we have derived the number of binaries in the _ common _ range 150 - 675 au , and find 11.8%@xmath42.2% for the associations and 5.3%@xmath40.8% in the onc ( for simplicity we here count a triple as two binaries ) .
both surveys represent _ observed _ binary fractions , but the reipurth & zinnecker survey has negligible incompleteness correction in the chosen range . for our present survey
, we have 9 line - of - sight pairs among our 78 binaries , and when we analyze their statistical distribution for different separations , we find that all 9 are likely to be found in the range 0@xmath333 to 1@xmath35 . of the 50 binaries in the range 150 - 675 au we thus should count only 41 as physical binaries , leading to the abovementioned binary percentage of 5.3% .
we thus find that the binary fraction in associations is higher by a factor of 2.2 compared to the onc , in qualitative agreement with the study of petr et al .
( 1998 ) , who found a difference of almost a factor of 3 based on small - number statistics . in figure 6
we show the separation distribution of binaries towards the onc on an angular scale and in bins 0@xmath31 wide .
the first bin from 0@xmath31 to 0@xmath32 is incomplete .
it is striking that there is clear evidence for structure in the separation distribution function , with a sudden decrease in the number of binaries as the separations increase beyond 0@xmath35 , corresponding to 225 au . at larger separations
, the distribution is quite flat .
we return to the physical interpretation of this `` wall '' in subsequent sections .
an early precursor to this result may have been seen by simon ( 1997 ) , who studied the two - point correlation function in the onc based on the results of prosser et al .
( 1994 ) and noted a transition from binary companions to the large - scale cluster at projected separations around 400 au .
we have made the same plot on a logarithmic scale in figure 7 .
the data are essentially complete over the whole separation range displayed .
we have corrected the distribution for the 9 line - of - sight pairs by calculating the probability of finding another star within a circle with radius corresponding to the separations of each binary , and then adding these probabilities up for each bin .
while such a procedure is meaningless for an individual object , it can be used to distribute the 9 false binaries across the 5 bins .
binaries with the largest separations are most likely to be line - of - sight pairs , and the last bin has a large correction of 6 stars , while three of the four remaining bins each is corrected for one star .
error bars are given for each corrected column .
figure 7 additionally shows the distribution of field stars in the same interval from duquennoy & mayor ( 1991 ) .
the dot - dash curve represents the gaussian that duquennoy & mayor fitted to their entire data set , while the two dashed crosses represent their actual data points within our range .
it is evident that the gaussian is a poor fit to these data points in this specific separation interval .
the duquennoy & mayor data are given as a function of period , and we have converted these to separations by assuming a mean total binary mass of 1.2 m@xmath18 , appropriate for their sample of f7 to g9 binaries .
we have also assumed , as is commonly done , that the projected separation represents the semimajor axis , although statistically there is a small difference ( kuiper 1935 , couteau 1960 ) . as noted above
, we find a binary frequency ( after correcting for the 9 line - of - sight pairs and treating the three triples as six binaries ) of 8.8%@xmath41.1% in the interval from 67.5 au to 675 au .
if we calculate the binary frequency from the duquennoy & mayor data in the same range , we find a binary frequency of 13.7% using a simple trapezoidal approximation to their log - normal curve , and 12.4% using linear interpolation of their data points .
it follows that , in the specific range from 67.5 au to 675 au , there is approximately 1.5 times more binaries in the field than in the onc .
we discuss this result further below .
we have explored whether there is a difference in the number of wide binaries relative to close binaries as we move from the dense inner cluster regions to the outer reaches of the cluster .
such a difference has been searched for but not found in earlier studies ( khler et al .
2006 ) . in view of the dramatic change in the separation distribution function at 0@xmath35
, we have chosen this separation as a dividing point , such that we consider binaries between 0@xmath315 and 0@xmath35 as ` close ' and binaries between 0@xmath35 and 1@xmath35 as ` wide ' .
figure 8 shows the cumulative distributions of these close and wide binaries as function of distance from @xmath2 ori c in the trapezium .
it is evident from the figure that the two sets of binaries do not have the same radial distribution with increasing distance from the cluster core . to explore this further
, we show in figure 9 the ratio of wide to close binaries as a continuous cumulative distribution , with the first point calculated 30@xmath1 outside the exclusion zone , that is , at a distance of 90@xmath1 from @xmath2 ori c. for each distance , the curve gives the ratio for all binaries from the edge of the exclusion zone to that particular distance .
in other words , as the curve moves away from the trapezium it accounts for more and more binaries , until the last points on the curve represent the mean ratio of wide - to - close binaries for the entire onc .
the dashed lines indicate the 1@xmath17 errors on the numbers .
we have furthermore calculated the same ratio for the duquennoy & mayor ( 1991 ) binaries , and the two dotted lines indicate the values calculated from their gaussian fit ( lower line ) and their actual data points ( upper line ) .
these lines thus represent the ratio for field stars .
the curve does not take into account that 9 of the binaries are likely to be line - of - sight associations .
given that the probability of being a line - of - sight pair increases with separation and with proximity to the center of the onc , it follows that the curve would dip even deeper down in the inner region if we could remove the 9 non - physical pairs .
it is evident that there is a very pronounced and almost monotonic change in the ratio of wide to close binaries as one moves away from the core of the onc until a distance of about 460@xmath1 , at which point the ratio becomes flat .
it is also clear that the mean ratio of wide to close binaries for the whole onc is lower than the duquennoy & mayor values .
we discuss the implications of these findings in section 5 .
in addition to the 72 binaries in the separation range from 0@xmath315 to 1@xmath35 , we have found 3 triple systems with separations in the same range .
tokovinin ( 2001 ) suggests that the ratio of triples to binaries is [email protected] in general , whereas after subtracting the 9 line - of - sight pairs we find [email protected] in the onc . within the considerable errors the numbers are close to being consistent , but it should also be recalled that our survey is restricted to the limited range between the 0@xmath315 completeness limit and the 1@xmath35 confusion limit .
it is entirely possible that some close pairs in triple systems have already evolved to become spectroscopic binaries and have thus become unobservable with our detection method .
nor can it be excluded that some wide pairs exist in triple systems with separations larger than 1@xmath35 , although such wide systems can not be common ( scally et al . 1999 ) .
much better statistics is required to settle the question whether the onc might be deficient in triple systems .
we have determined the flux - ratios of those of our binaries that are not saturated in our images .
figure 10 shows the resulting histogram as a function of @xmath7mag .
as is well established from other binary studies , the majority of binaries in the onc also have unequal components , although the bin with equal components ( @xmath7 m @xmath11 0.5 mag . )
is the most populated .
only very few of the companions have a magnitude difference to their primary of more than 2.5 magnitudes . in the absence of spectroscopic information ,
we have made a crude attempt to investigate the nature of the companions based on their measured flux ratios . in order to do that we have made some assumptions .
first , we assume that the observed flux ratios reflect the photospheric fluxes , in other words that h@xmath0 emission line fluxes are not seriously affecting the ratios .
this may not always be a good assumption , since mid - to - late m dwarfs often are very active , and their photospheric fluxes are so low that h@xmath0 line emission could be a significant contribution .
if we mistake h@xmath0 line emission for photospheric flux from the primary , our estimate of a spectral type for a companion will be earlier than it is in reality .
second , most published photometry of late - type dwarfs is broad - band , whereas we have observed in the narrow - band h@xmath0 filter , so we assume that our observed magnitude differences can be compared to r - band photometry .
since we are dealing with the difference between two stars , this is probably not a bad assumption , at least when the flux ratio is not large .
third , we assume that the observed flux ratios are not affected by differences in extinction between the components .
given the young age and occasional association with molecular clouds , this may not in all cases hold true . with
these caveats spelled out , we have used the m@xmath19 vs spectral type relation and the r - i colors for m dwarfs ( e.g. , dahn et al .
2002 ) to derive the difference in r - band magnitudes as a function of spectral type for m - dwarfs .
the relation turns out to be essentially linear , with a mean drop of 0.56 magnitudes per spectral subtype throughout the m spectral range .
we then used the spectral classifications for the primaries provided by simbad , of mixed provenance , with our flux ratios to estimate a spectral type for the secondaries . for a cluster with an age between 0.5 and 3 myr
, the substellar limit is around spectral type m6.25 following the models of baraffe et al .
( 1998 ) and chabrier et al .
( 2000 ) and using the temperature scale of luhman et al .
( 2003 ) , see also luhman et al .
( 2006 ) . to our surprise
, quite a number of the secondaries appear to be brown dwarfs , in at least one case forming a wide bd - bd binary .
we comment on selected cases in the appendix . given the assumptions
involved and the simplistic nature of these spectral type estimates , it is obvious that spectroscopy is required to establish the true nature of the secondaries .
the separation distribution function of field binaries has been approximated by various functions .
pik ( 1924 ) suggested that binaries with separations larger than @xmath560 au follow a _
f(a ) @xmath5 1/a _ distribution , whereas kuiper ( 1942 ) proposed that the overall distribution could be represented by a gaussian .
the latter distribution was adopted as a good fit to their data in the influential study by duquennoy & mayor ( 1991 ) , although it is evident from figure 7 that in the specific separation range discussed in the present paper , the gaussian is a poor approximation to the actual data . integrating the pik relation over logarithmic bins results in a straight horizontal line to describe the logarithmic separation distribution , and
this is clearly a better fit to the two data points of duquennoy & mayor seen in figure 7 .
the pik relation has also been supported by numerous studies summarized by poveda & allen ( 2004 ) . for the separation distribution of onc binaries ( corrected for line - of - sight associations ) both a gaussian and a horizontal line provide acceptable fits to the data when considering the uncertainties of the data points .
it is thus not possible to classify the structure of the onc separation distribution function with the available data .
the separation distribution function in the onc most probably has not yet found its final shape .
two basic mechanisms operate that can affect the orbit of a young binary : rapid dynamical decay in small - n clusters ( e.g. , sterzik & durisen 1998 , reipurth & clarke 2001 , bate et al . 2002 ) and the passage of a binary through a dense cluster ( e.g. , kroupa 1995a , b , kroupa et al . 1999 ) .
the former occurs primarily during the class 0 phase ( reipurth 2000 ) , and is no longer relevant for stars in the onc .
the latter , however , may be essential for understanding the binary population of the onc .
a wide binary falling through the potential well of a cluster will gain kinetic energy through encounters , and binaries with weak binding energies are eventually disrupted ( heggie 1975 ) .
a prediction of this scenario is that binaries at distances from the cluster center larger than the corresponding crossing time should not be showing any dynamical alterations due to encounters with other cluster members ( kroupa et al . 1999 , 2001 ) .
the crossing time of a star through a cluster is @xmath20 , where @xmath21 is the cluster radius and @xmath17 is the mean one - dimensional velocity dispersion in the cluster . assuming that this velocity dispersion in the onc is of the order of 2 km s@xmath22 ( e.g. , jones & walker 1988 ) , we have in figure 9 indicated the crossing time for different distances to the cluster center .
the figure suggests that for distances larger than roughly 460 arcsec there is no longer a measurable change in the ratio of wide - to - close binaries .
given that the wide - to - close binary ratio changes by a factor of 4 - 5 from the inner to the outer regions , this suggests that many , and perhaps most , of the wide binaries are disrupted after only a few passages of the cluster center .
the variation of the ratio of wide - to - close binaries from the inner to the outer regions of the onc , seen in figure 9 , offers the first compelling observational evidence that dynamical interactions in the dense central region of the onc have taken place . in principle , the diagram in figure 9 allows us to determine the age of the onc .
an angular distance of @xmath5460 arcsec corresponds to a crossing time of about 1 million years .
however , an age determined in this manner is directly dependent on the velocity dispersion assumed .
therefore , all we can say about the age of 1 million years we estimate for the onc is that it is consistent with other onc age estimates ( e.g. , hillenbrand 1997 )
. the significantly lower number of binaries that we find in the onc compared to associations thus appears to be due , at least in part , dissolution of wide binaries .
however , under certain circumstances an encounter could lead to hardening of the binary , making it closer than our resolution limit , so it is not lost to the overall binary budget . even in its outermost regions ,
the onc shows a smaller ratio of wide - to - close binaries than seen in the field by duquennoy & mayor ( 1991 ) . considering that the majority of field stars are likely to have been formed in a cluster
, it follows that many of the _ wide _ binaries in the field must have formed in the gentler environment of a loose t association .
durisen & sterzik ( 1994 ) found , on theoretical grounds , that binaries are more likely to form in clouds with lower temperatures .
reipurth & zinnecker ( 1993 ) found observational evidence that clouds with more stars have relatively fewer binaries in the separation range under study ( mostly wide visual binaries ) .
both these results could indicate that loose t associations produce or retain more wide binaries than do clusters . however , brandeker et al .
( 2006 ) noted that the young sparse @xmath23 chamaeleontis cluster has a deficit of wide binaries .
unless this small cluster is the remnant of a much denser cluster , then this result would seem to be in contradiction to the notion that wide binaries are preferentially formed / preserved in loose associations .
in any case , there is no question that the field binary population is a mix of binaries formed in clusters and in associations . we can attempt to calculate the fraction of stars that originate in clusters and in associations . in sect .
4.2 we showed that , in the limited separation range 67.5 - 675 au , the binary fraction of the field stars of duquennoy & mayor ( 1991 ) is about a factor 1.5 higher than in the onc .
we also found that in the range 150 - 675 au associations have a factor 2.2 more binaries than the onc .
in other words , in these limited ranges associations have about 1.5 times as many binaries as the field population , and the onc has only about 2/3 as many binaries as the field population .
it follows that the field star binary population in principle can be produced if only 1/3 of binaries come from the binary - rich associations and 2/3 originate in binary - poor onc - type clusters , in excellent agreement with the results of patience et al .
( 2002 ) and in approximate agreement with findings that 70 - 90% of all stars may be formed in clusters ( e.g. , lada & lada 2003 ) .
of course , this assumes that the binary ratios between field stars , associations , and the onc derived here within narrow separation ranges can be extrapolated to all separations , an assumption that may not hold at all . as noted in sect .
4.5 , spectroscopy is required to determine the true nature of the companions we have found . however , unless significant extinction differences are common among our binaries , then the difference in component brightness , combined with spectral type estimates for the primaries when available , are indicative of the spectral types of the secondaries .
while this is highly uncertain in individual cases , overall it is likely to be indicative of the properties of the companion population .
table 1 lists the spectral types for 46 primaries , and of these 38 , or 83% , are m - type stars .
we find that more than half ( 25 ) of these 46 binary stars have secondaries with spectral types of m5 or later , and about one third ( 17 ) could be substellar .
these numbers are upper limits , since we can not correct for the 9 false binaries that we expect to be present in our sample of 78 multiple systems , but they are probably more likely to be found among pairs with faint , widely separated companions . however , even if we subtracted all 9 objects from the 17 that could be substellar , this would still indicate that 8 , or about 17% , of the 46 binaries with spectral type estimates are likely to have substellar companions .
selected cases are discussed in the appendix .
the closest binaries we were able to resolve with our imaging technique have projected separations of 50 au . all of the binaries with likely substellar companions are thus quite wide , which is of interest for formation theories ( see , e.g. , lucas et al . 2005 , and reviews by burgasser et al . 2007 , luhman et al .
2007 , whitworth et al .
2007 ) .
it is likely that there are several brown dwarf brown dwarf binaries in our binary sample , but the only one that we are certain of is coup 1061 .
infrared spectroscopy of this source indicates a spectral type of m9-l0 ( meeus & mccaughrean 2005 ) .
we have resolved this object into two components with almost equal brightness and a projected separation of 100 au .
brown dwarf binaries with such large separations are rare ( e.g. , lucas et al . 2005 , allen et al .
2007 ) , although a few very wide pairs are known ( e.g. , artigau et al .
2007 , barrado y navascus et al . 2007 ) .
we have analyzed a large set of h@xmath0 images of the orion nebula cluster acquired with the hubble space telescope and the advanced camera for surveys with the goal of detecting binaries among a sample of 781 onc members .
the following results were obtained : \1 .
a total of 75 binaries and 3 triple systems were detected , of which 55 are new discoveries . \2 . within the limited angular range of 0@xmath315 to 1@xmath35 , corresponding to projected separations of 67.5 au to 675 au , we have found a binary fraction of 8.8%@xmath41.1% after correcting for a statistically determined contamination of 9 line - of - sight binaries .
the field binary fraction for solar type stars in the same separation range is 1.5 times larger , and for t tauri associations it is 2.2 times larger than in the onc , confirming earlier results that the onc is deficient in binaries , now with statistically significant data .
the separation distribution function for young binaries in the onc shows a dramatic decrease in binaries for angular separations larger than 0.5@xmath1 , corresponding to projected separations of 225 au .
the ratio of cumulative distributions of wide ( 0@xmath35 to 1@xmath35 ) to close ( 0@xmath315 to 0@xmath35 ) binaries show an increase out to a distance of about 460@xmath1 from the center of the onc , after which it levels out . we interpret this as clear observational evidence for dynamical evolution of the binary population as a result of passages through the potential well of the onc .
these results are consistent with an age of the onc of about 1 million yr .
it appears that much of the deficiency of binaries in the onc compared to the field star population can be understood , at least in part , in terms of the destruction of wide binaries combined with a secondary effect from orbital evolution of binaries towards closer separations that are unobservable in direct imaging surveys .
limited spectral information about the primary stars indicate that they are low mass t tauri stars ( except for one herbig ae / be star ) . assuming that most binaries are not affected by differential extinction
, we find that possibly as many as 50% of the binaries have companions with spectral types later than m5 , and from 1/6 to 1/3 of the binaries may have substellar companions , all of which with separations of at least 50 au .
this large number of wide substellar companions is of interest for theories of brown dwarf formation .
we thank rainer khler , pavel kroupa , luiz paulo vaz , and adam burgasser for valuable comments , and an anonymous referee for a very helpful referee report .
we are grateful to burton jones for providing us an electronic version of the jones & walker ( 1988 ) proper motion table .
marcelo guimares acknowledges financial support by capes , cnpq and fapemig .
based on observations taken under program go-9825 with the nasa / esa hubble space telescope obtained at the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
this material is based upon work supported by the national aeronautics and space administration through the nasa astrobiology institute under cooperative agreement no .
nna04cc08a issued through the office of space science .
this research has made use of the simbad database , operated at cds , strasbourg , france , and nasa s astrophysics data system bibliographic services .
in the following , comments are provided on binaries of particular interest , especially regarding possible substellar companions . * v1438 ori * ( jw 39 ) .
the spectral type of m3 is due to hillenbrand ( 1997 ) .
if the 2.6 mag brightness difference is not affected by extinction differences , then the companion is an m8 brown dwarf at a projected separation of 170 au .
stassun et al .
( 1999 ) detected lithium in the primary . *
jw 71*. the spectral type of m4 is due to hillenbrand ( 1997 ) .
if the 1.4 mag brightness difference is not affected by extinction differences , then the companion is an m7 brown dwarf with a projected separation of 90 au .
* v1118 ori * ( jw 73 ) .
this is a member of the class of exor s ( herbig 1989 ) .
the star , which is also known as chanal s object , has had 5 outbursts since its discovery in 1984 , and typically varies within the range 14.5 @xmath11v@xmath1118 , with a rise time of less than a year and a declining phase about twice as long .
optical and infrared spectroscopy of v1118 ori is reported by parsamian et al .
( 2002 ) and lorenzetti et al .
( 2006 ) , and x - ray observations are analyzed by audard et al .
the discovery that v1118 ori has a companion only 0.4 magnitude fainter in the h@xmath0 filter raises interesting questions about which of the components that may drive the variability ( herbig 2007 ) . *
jw 121*. the spectral type of m3 is due to hillenbrand ( 1997 ) .
if the 3.0 mag brightness difference is not affected by extinction differences , then the companion is an m8 brown dwarf with a projected separation of 153 au . *
jw 147*. the spectral type of m5 is due to hillenbrand ( 1997 ) .
if the 1.4 mag brightness difference is not affected by extinction differences , then the companion is an m8 brown dwarf with a projected separation of 135 au . *
jw 152*. the spectral type of m3 is due to hillenbrand ( 1997 ) .
if the 2.5 mag brightness difference is not affected by extinction differences , then the companion is an m7 brown dwarf with a projected separation of 81 au . *
jw 235*. jw 235 is associated with the hh 504 object , discovered by bally & reipurth ( 2001 ) , see also odell ( 2001 ) .
the binarity of the star was discovered by khler et al .
its proper motion in the jones & walker catalog implies that it has 0% probability of being a member of the onc .
however , given the similar brightness of the two components , it is likely that brightness variations of the stars shifted the photocenter used for astrometry .
evidence for onc membership is strong and is based on the presence of h@xmath0 emission , detection in x - rays , optical variability , and association with an hh object . * v1274 ori * ( jw 248 ) .
the spectral type of the primary is somewhat in dispute .
hillenbrand ( 1997 ) suggests m0.5-m2 ( with the ca ii lines in emission ) , edwards et al .
( 1993 ) suggest m3 , and meeus & mccaughrean ( 2005 ) suggest m0-m4 .
the binarity of the system was found by the coup survey .
our primary is the secondary in the coup catalog , indicating that our secondary is highly x - ray active .
sicilia - aguilar et al . (
2005 ) detected lithium in the ( optical ) primary , and found an inverse p cygni profile at h@xmath0 . if there is not an extinction difference between the primary and secondary , then the large flux difference ( @xmath243.7 mag ) suggests that the companion could be an l0 brown dwarf .
however , the system is only 3 arcmin from @xmath2 ori c , so there is a 1% chance of a line - of - sight association . * jw 296*. jw 296 is the primary of a hierarchical triple system . the secondary and tertiary are very faint stars , surrounded by a common proplyd - like envelope known as 066 - 652 ( odell & wong 1996 ) .
the primary is of spectral type m4.5 according to hillenbrand ( 1997 ) .
if the brightness difference is taken at face value , the secondary and tertiary should be l0 brown dwarfs , but given the obvious association with nebulosity it is likely that their faintness is merely due to extinction . *
jw 355*. the primary of this system is located at the edge of a proplyd known as 109 - 246 ( odell & wong 1996 ) .
the proper motion of the star suggests it is not an onc member , but it is found to be a member by the coup project .
hillenbrand ( 1997 ) suggests a mid - k spectral type .
the system is only 90@xmath1 from @xmath2 ori c , and the star density is so high that the possibility of a line - of - sight association is several percent . *
jw 370*. there is a silhouette disk close ( @xmath53@xmath34 ) to this binary . given the local density of stars around this system , the probability that such a silhouette disk is just a chance alignment is @xmath57% .
the chance of a line - of - sight alignment for the binary itself is 1.5% .
hillenbrand ( 1997 ) suggests a spectral type of k0-k3 . * v1492 ori * ( jw 383 ) .
the primary is of spectral type m3 according to hillenbrand ( 1997 ) .
if the 2.1 mag brightness difference is not affected by extinction differences , then the companion is an m7 brown dwarf with a projected separation of 86 au .
stassun et al .
( 1999 ) derive a rotation period of 7.11 days for the primary , and notes the presence of lithium in the spectrum .
* parenago 1806 * ( jw 391 ) .
the primary is of spectral type m1 according to edwards et al .
( 1993 ) and hillenbrand ( 1997 ) , who also notes the presence of ca ii emission .
the primary is saturated in our images , but with a magnitude difference of at least 3.3 mag , the companion would be an m7 brown dwarf if there is no difference in extinction .
however , the pair is located in a region of high stellar density , so the propability of a chance alignment is more than 5% . *
jw 509*. this binary was first detected by prosser et al .
( 1994 ) and subsequently by padgett et al .
( 1997 ) and lucas et al .
although it is a well known binary it does not have a spectral type in the literature . since this system is located in a crowded region , the probability of a chance alignment is @xmath5 2.7% , but the reality of the binary is supported by an interaction zone between the stars , visible in our image ( fig .
3b ) . * parenago 1914 * ( jw 551 ) .
the primary of this hierarchical triple system has a spectral type m1 according to hillenbrand ( 1997 ) .
the tertiary is not detectable either in the optical or in x - rays , and since its separation from the primary is as large as 1@xmath327 it could be a background object .
the system seems to be part of a small subcluster , which gives it a high probability of a chance alignment ( @xmath5 2.2% ) . given the primary spectral type , the difference in magnitudes to the tertiary ( 3.6 ) and assuming that the tertiary is not a background object , the tertiary could have a spectral type m7 or m8 and thus would be a brown dwarf candidate .
* coup 967*. this star was catalogued as a binary by prosser et al .
( 1994 ) , padgett et al .
( 1997 ) and lucas et al .
our image shows that the primary is associated with a proplyd ( 184 - 427 in odell & wong 1996 ) and the secondary is a faint object ( @xmath7m@xmath25 = 2.4 ) .
the spectral type for the primary is m2.5 according to hillenbrand ( 1997 ) .
the system is very close to @xmath2 ori c ( @xmath5 70@xmath1 ) and the probability of a chance alignment is 3.5% .
if the components have the same extinction , then the secondary would be of spectral type m6.5 , and thus at the hydrogen burning limit . *
jw 592*. the spectral type m2.5 is due to hillenbrand ( 1997 ) and m4 according to edwards et al .
the brightness difference ( 4 magnitudes ) would imply an m9.5 spectral type or later for the secondary , making it a brown dwarf with a projected separation of @xmath5 275 au . * coup 1061*. this very close system ( @xmath5 100
au ) has a combined spectral type m9-l0 according to meeus & mccaughrean ( 2005 ) , who did not resolve it , and since the components have virtually the same brightness , this is therefore a _ bona fide _ brown dwarf - brown dwarf binary with a projected separation of 100 au .
* v1524 ori * ( jw 681 ) .
this binary was first catalogued by prosser et al .
( 1994 ) and has a spectral type k7 according to prosser & stauffer ( unpublished ) .
the secondary is associated with the proplyd 213 - 346 ( odell & wong 1996 ) . in the hillenbrand ( 1997 )
catalog this system has two entries , both with number 681 but with different coordinates .
simbad identifies three different objects : h97b-681 , h97b-681a , and h97b-681b .
the first is jw 681 , also known as v1524 ori , and is the primary in this double .
the second is the unrelated object mlla 312 , and the third is the x - ray source coup 1149 , which forms the secondary component , associated with the proplyd . because this system is located near to @xmath2 ori c ( @xmath26 ) the probability of a chance alignment is 3.5% . *
v1528 ori * ( jw 727 ) .
the spectral type m2 is due to hillenbrand ( 1997 ) and the difference in brightness suggests a spectral type of l0 for the companion if the components have the same extinction . if so , the secondary is a brown dwarf with projected separation of 329 au .
the primary shows h@xmath0 emission according to our unpublished survey . *
jw 748*. this binary was discovered by prosser et al .
( 1994 ) , and the spectral type of the primary is m3 according to hillenbrand ( 1997 ) . given the brightness diffference and if the components have the same extinction , the secondary could be a brown dwarf with spectral type m7 and a projected separation of 158 au . *
jw 767*. first discovered by khler et al .
( 2006 ) this system has a spectral type m2.5 due to hillenbrand ( 1997 ) .
the secondary has a silhoutte disk and is associated with the hh 668 object ( bally et al .
the extinction caused by the silhoutte disk probably is responsible for the large difference in brightness by more than 4 magnitudes between the primary and secondary .
* parenago 2075 * ( jw 867 ) .
this object was observed by khler et al .
( 2006 ) but despite its brightness and separation ( 0@xmath329 ) they did not resolve it , suggesting possible major variability of the secondary .
the system presents evidence of h@xmath0 emission , x - rays and optical variability .
a spectral type of m1 was suggested by blanco ( 1963 ) , and more recently duncan ( 1993 ) assigned a spectral type of k8ve .
sicilia - aguilar et al .
( 2005 ) reported lithium in the spectrum of the primary .
* jw 906*. the primary is of spectral type m3 according to hillenbrand ( 1997 ) , and if the difference in brightness ( 2.7 mag ) is not affected by extinction then the secondary is a brown dwarf with spectral type m8 and a large projected separation of 625 au .
sicilia - aguilar et al .
( 2005 ) reported lithium in the spectrum of the primary .
allen , p.r . ,
koerner , d.w .
, mcelwain , m.w . ,
cruz , k.l . , & reid , i.n .
2007 , , 133 , 971 artigau , e. , lafrenire , d. , doyon , r. , albert , l. , nadeau , d. , & robert , j. 2007 , , 659 , l49 audard , m. , gdel , m. , skinner , s.l . ,
briggs , k.r . , walter , f.m .
2005 , , 635 , l81 bally , j. & reipurth , b. 2001 , , 546 , 299 bally , j. , licht , d. , smith , n. , & walawender , j. 2006 , , 131 , 473 baraffe , i. , chabrier , g. allard , f. , & hauschildt , p.h .
1998 , , 337 , 403 barrado y navascus , d. , bayo , a. , morales - caldern , m. , hulamo , n. , stauffer , j.r . , & bouy , h. 2007 , , in press bate , m.r . ,
bonnell , i.a .
, & bromm , v. 2002 , , 336 , 705 blanco , v.m .
1963 , apj , 137 , 513 brandeker , a. , jayawardhana , r. , khavari , p. , haisch , k.e . , & mardones , d. 2006 , , 652 , 1572 burgasser , a.j . ,
reid , i.n . ,
siegler , n. , close , l. , allen , p. , lowrance , p. , & gizis , j. 2007 , in _ protostars and planets v _ , eds . b. reipurth , d. jewitt , & k. keil , univ . of arizona press , 427 chabrier , g. , baraffe , i. , allard , f. , & hauschildt , p.h .
2000 , , 542 , 464 choi , p.i . &
herbst , w. 1996 , , 111 , 283 correia , s. , zinnecker , h. , ratzka , th . , &
sterzik , m.f .
2006 , , 459 , 909 couteau , p. 1960 , journ .
des observateurs , 43 , 41 dahn , c.c . ,
harris , h.c . ,
vrba , f.j . ,
guetter , h.h . ,
canzian , b. , henden , a.a .
2002 , , 124 , 1170 duchne , g. 1999 , , 341 , 547 duchne , g. , bouvier , j. , bontemps , s. , andr , p. , & motte , f. 2004 , , 427 , 651 duncan , d.k .
1993 , apj , 406 , 172 duquennoy , a. & mayor , m. 1991 , , 248 , 485 durisen , r.h . & sterzik , m.f .
1994 , , 286 , 84 edwards , s. , strom , s.e . ,
hartigan , p. , strom , k.m . ,
hillenbrand , l.a .
1993 , , 106 , 372 getman , k.v . ,
feigelson , e.d . ,
grosso , n. , mccaughrean , m.j . , micela , g. et al .
2005 , , 160 , 353 heggie , d.c .
1975 , , 173 , 729 herbig , g.h . 1962 , _ adv .
_ , 1 , 47 herbig , g.h .
1982 , in _ symp . on the orion nebula to honor henry draper _ ,
glassgold & p.j .
huggins , new york acad .
, 64 herbig , g.h .
1989 , in eso workshop on _ low mass star formation and pre - main sequence objects _ , ed .
bo reipurth , p.233 herbig , g.h .
2007 , in prep .
herbig , g.h . & terndrup , d.m .
1986 , , 307 , 609 hillenbrand , l.a .
1997 , , 113 , 1733 jones , b.f . & walker , m.f .
1988 , , 95 , 1755 joy , a.h . &
van biesbroeck , g. 1944 , pasp , 56 , 123 kroupa , p. 1995a
, , 277 , 1491 kroupa , p. 1995b
, , 277 , 1507 kroupa , p. 2000 , new astronomy , 4 , 615 kroupa , p. , petr , m.g . , & mccaughrean , m.j . 1999 , new astronomy , 4 , 495 kroupa , p. , aarseth , s. , & hurley , j. 2001 , , 321 , 699 kuiper , g.p . 1935 , , 47 , 121 kuiper , g.p . 1942 , , 95 ,
201 khler , r. , petr - gotzens , m. g. , mccaughrean , m. j. , bouvier , j. , duchne , g. et al .
2006 , , 458 , 461 lada , c.j . & lada , e.a .
2003 , ann .
41 , 57 lorenzetti , d. , giannini , t. , calzoletti , l. , puccetti , s. , antoniucci , s. et al . 2006 , , 453 , 579 lucas , p.w . , roche , p.f . , & tamura , m. 2005 , , 361 , 211 luhman , k.l . ,
stauffer , j.r . ,
muench , a.a . ,
rieke , g.h . ,
lada , e.a . ,
bouvier , j. , & lada , c.j .
2003 , , 593 , 1093 luhman , k.l . ,
whitney , b.a . ,
meade , m.r . ,
babler , b.l . ,
indebetouw , r. , bracker , s. , & churchwell , e.b .
2006 , , 647 , 1180 luhman , k.l . , joergens , v. , lada , c. , muzerolle , j. , pascucci , i. , & white , r. 2007 , in _ protostars and planets v _
, eds . b. reipurth , d. jewitt , & k. keil , univ . of arizona press , 443 meeus , g. & mccaughrean , m.j .
2005 , an , 326 , 977 odell , c.r .
2001 , , 122 , 2662 odell , c.r . &
wong , k. 1996 , , 111 , 846 padgett , d.l . ,
strom , s.e . , & ghez , a. 1997 , , 477 , 705 parsamian , e.s .
, mujica , r. , & corral , l. 2002 , astrophysics , 45 , 393 patience , j. , ghez , a.m. , reid , i.n . , & matthews , k. 2002 , , 123 , 1570 petr , m.g . , coud du foresto , v. , beckwith , s. , richichi , a. , & mccaughrean , m.j .
1998 , , 500 , 825 poveda , a. & allen , c. 2004 , rev .
21 , 49 preibisch , t. , balega , y. , hofmann , k .- h . , weigelt , g. , & zinnecker , h. 1999 , new astron . , 4 , 531 prosser , c.f . ,
stauffer , j.r . , hartmann , l. , soderblom , d.r . ,
jones , b.f .
1994 , , 421 , 517 ratzka , t. , khler , r. , & leinert , ch .
2005 , , 437 , 611 reipurth , b. 2000 , , 120 , 3177 reipurth , b. & clarke , c.j .
2001 , , 122 , 432 reipurth , b. & zinnecker , h. 1993 , , 278 , 81 scally , a. , clarke , c.j . , & mccaughrean , m.j .
1999 , , 306 , 253 sicilia - aguilar , a. , hartmann , l.w .
, szentgyorgyi a.h . ,
fabricant , d.g . ,
frsz , g. et al .
2005 , aj , 129 , 363 simon , m. 1997 , , 482 , l81 simon , m. , ghez , a. m. , leinert , ch . , cassar , l. , chen , w. p. et al .
1995 , , 443 , 625 simon , m. , close , l.m . , beck , t.l . 1999 , , 117 , 1375 stassun , k.g . , mathieu , r.d . ,
mazeh , tsevi , & vrba , f.j .
1999 , , 117 , 2941 sterzik , m.f . &
durisen , r.h .
1998 , , 339 , 95 tokovinin , a. 2001 , in iau symp .
no.200 _ the formation of binary stars _ , eds . h. zinnecker & r.d .
mathieu , p.84 pik , e. 1924 , tartu obs .
publ . , 25 , no .
6 whitworth , a. , bate , m.r . ,
nordlund , a. , reipurth , b. , & zinnecker , h. 2007 , in _ protostars and planets v _
, eds . b. reipurth , d. jewitt , & k. keil , univ . of arizona press , 459 ccccccccccccccclc 39 & 99 & & v1438 & & 39 & 5:34:38.1 & -5:27:41 & 21 & 0.38 & 14.29 & 2.6 & 246 & 628 & m3 & pv & + 52 & 98 & & & & 52 & 5:34:40.8 & -5:28:09 & 231 & 0.20 & 14.58 & 0.1 & 242 & 604 & m5 & p & + 63 & 99 & 1569 & v1441 & 17 & 63 & 5:34:43.0 & -5:20:07 & 0 & 0.27 & 12.75 & 1.5 & 291 & 537 & k6 & pxv & + 71 & 99 & & & 21 & 71 & 5:34:44.5 & -5:24:38 & 105 & 0.20 & 15.61 & 1.4 & 261 & 483 & m4 & px & + 73 & 99 & & v1118 & & 73 & 5:34:44.7 & -5:33:42 & 329 & 0.18 & 14.04 & 0.4 & 217 & 779 & m1 & phv & + 81 & 97 & 1600 & & 28 & 81 & 5:34:46.4 & -5:24:32 & 272 & 1.34 & 12.91 & 1.9 & 261 & 454 & m0 & pxv & + 121 & 99 & & & & 121 & 5:34:51.2 & -5:16:55 & 200 & 0.34 & 14.90 & 3.0 & 316 & 541 & m3 & pv & + 124 & 99 & & & 64 & 124 & 5:34:51.8 & -5:21:39 & 200 & 0.48 & 13.94 & 0.0 & 286 & 382 & m3.5 & pxv & + 127 & 99 & & & 66 & 127 & 5:34:52.1 & -5:24:43 & 254 & 0.48 & 14.28 & 0.6 & 258 & 373 & m3.5 & pxhv & + 128 & 97 & & v1458 & 67 & 128 & 5:34:52.2 & -5:22:32 & 215 & 0.39 & 12.64 & @xmath240.6 & 278 & 366 & m2.5 & pxhv & + 135 & 99 & & v1460 & 72 & 135 & 5:34:52.7 & -5:29:46 & 11 & 0.40 & 14.45 & 0.1 & 223 & 522 & m3 & pxv & + 147 & 99 & & & & 147 & 5:34:54.4 & -5:17:21 & 26 & 0.30 & 14.80 & 1.4 & 318 & 489 & m5 & p & + 151 & 99 & & & 95 & 151 & 5:34:54.8 & -5:25:13 & 332 & 0.15 & 15.44 & 1.6 & 251 & 341 & m3 & px & + 152 & 99 & & & 96 & 152 & 5:34:55.1 & -5:25:30 & 51 & 0.18 & 14.12 & 2.5 & 248 & 343 & m3 & px & + 176 & 99 & 1673 & kq & 123 & 176 & 5:34:57.8 & -5:23:53 & 336 & 1.28 & 11.58 & @xmath241.1 & 264 & 280 & k8 & pxv & e + 190 & 99 & & & 134 & 190 & 5:34:59.3 & -5:23:33 & 355 & 0.56 & 15.33 & 0.3 & 268 & 256 & m6 & px & f + 201 & 99 & & & 150 & 201 & 5:35:01.0 & -5:24:10 & 246 & 1.09 & 13.41 & 0.8 & 258 & 235 & m2.5 & pxh & + 222 & 98 & & v1320 & 174 & 222 & 5:35:02.2 & -5:29:10 & 89 & 1.07 & 14.14 & 1.4 & 212 & 407 & m2 & pxv & + 223 & 99 & & & 177 & 223a & 5:35:02.4 & -5:20:47 & 130 & 0.34 & 13.64 & 1.3 & 307 & 261 & & px & + 224 & 99 & & & 180 & 224 & 5:35:02.7 & -5:19:45 & 164 & 0.27 & 15.38 & 2.4 & 317 & 299 & m1 & pxv & + 235 & 0 & & & 197 & 235 & 5:35:03.6 & -5:29:27 & 346 & 0.54 & 13.69 & 0.4 & 208 & 411 & & pxhv & d + 248 & 99 & & v1274 & 214 & 248 & 5:35:04.4 & -5:23:14 & 320 & 0.90 & 12.73 & @xmath243.7 & 273 & 180 & m3 & pxv & e + - & - & & & 260 & 3064 & 5:35:06.2 & -5:22:13 & 284 & 0.40 & 16.20 & 0.3 & 295 & 169 & m4 & x & f + 296 & 99 & & & 275 & 296 & 5:35:06.6 & -5:26:51 & 196 & 0.89 & 14.30 & 3.2 & 215 & 255 & m4.5 & px & + 296 & 99 & & & 275 & 296 & 5:35:06.6 & -5:26:52 & 158 & 0.27 & 14.30 & 0.4 & 215 & 256 & m4.5 & px & + 305 & 99 & & & & 305 & 5:35:07.6 & -5:24:01 & 302 & 0.45 & 14.58 & 0.3 & 254 & 137 & m3 & pv & f + 355 & 0 & & & 403 & 355 & 5:35:10.9 & -5:22:46 & 197 & 0.42 & 15.03 & 3.5 & 294 & 90 & k : & x & + 370 & 99 & & & 452 & 370 & 5:35:11.9 & -5:19:26 & 122 & 0.73 & 13.86 & 3.5 & 344 & 246 & k1.5 & px & + 383 & 99 & & v1492 & 489 & 383 & 5:35:12.7 & -5:16:14 & 280 & 0.19 & 14.57 & 2.1 & 353 & 433 & m3 & pxv & + 392 & 99 & & & 498 & 392 & 5:35:12.7 & -5:27:11 & 185 & 0.23 & 15.21 & 0.2 & 194 & 234 & m6 & px & + 391 & 99 & 1806 & & 501 & 391 & 5:35:12.8 & -5:20:44 & 94 & 0.29 & 12.97 & @xmath243.3 & 341 & 168 & m1 & pxv & + 399 & 99 & & & 523 & 399b & 5:35:13.2 & -5:22:21 & 345 & 0.22 & 16.77 & 0.1 & 322 & 78 & & px & a , c + 410 & 99 & & & & 410 & 5:35:13.2 & -5:36:18 & 167 & 1.19 & 16.00 & 1.3 & 184 & 777 & & pv & + 406 & 99 & & v1327 & 543 & 406 & 5:35:13.5 & -5:17:10 & 271 & 0.95 & 13.88 & 2.6 & 353 & 375 & m1 & pxv & d + - & - & & & 562 & 9048 & 5:35:13.6 & -5:21:21 & 62 & 0.24 & 16.27 & 1.5 & 341 & 129 & m3 & x & a + 422 & 99 & & v1495 & 566 & 422 & 5:35:13.7 & -5:28:46 & 74 & 0.31 & 14.43 & 0.0 & 187 & 326 & & pxv & + 436 & 0 & & & 620 & 436a & 5:35:14.3 & -5:22:04 & 238 & 0.32 & 16.97 & 1.0 & 338 & 85 & & x & a , c + 439 & 99 & & v1329 & 626 & 439 & 5:35:14.5 & -5:17:25 & 150 & 0.30 & 14.97 & 0.1 & 355 & 359 & m1 & pxv & + 439 & 99 & & v1329 & 626 & 439 & 5:35:14.5 & -5:17:25 & 78 & 1.21 & 14.97 & 4.1 & 355 & 359 & m1 & pxv & + 444 & 99 & & & 651 & 444 & 5:35:14.6 & -5:16:46 & 190 & 0.18 & 15.53 & 1.0 & 356 & 398 & & px & + 445 & 26 & & & 645 & 445a & 5:35:14.7 & -5:20:42 & 191 & 0.44 & 14.67 & 1.8 & 351 & 163 & & x & a + - & - & & v1500 & & 466 & 5:35:14.9 & -5:36:39 & 82 & 0.38 & 14.45 & 2.2 & 182 & 797 & & v & + 465 & 97 & 1875 & v409 & & 465 & 5:35:14.9 & -5:38:06 & 271 & 0.43 & 14.71 & 4.1 & 182 & 883 & & pv & + 498 & 99 & 1873 & v1504 & & 498 & 5:35:15.8 & -5:32:59 & 122 & 0.70 & 13.81 & 0.2 & 181 & 576 & & phv & + 509 & 99 & & & 789 & 509a & 5:35:16.2 & -5:24:56 & 46 & 0.49 & 15.06 & 0.1 & 182 & 93 & & pxv & a , b , f + 511 & 99 & & & & 511a & 5:35:16.3 & -5:22:10 & 241 & 0.41 & 15.67 & 2.4 & 358 & 73 & m1 & px & a , c , e , f + - & - & & & 822 & 3031 & 5:35:16.8 & -5:17:17 & 84 & 0.26 & 16.37 & 0.5 & 1 & 366 & & x & + 551 & 99 & 1914 & & 881 & 551a & 5:35:17.4 & -5:25:45 & 265 & 0.14 & 13.87 & 0.0 & 174 & 142 & m1 & px & a + 551 & 99 & 1914 & & 881 & 551a & 5:35:17.4 & -5:25:45 & 262 & 1.27 & 13.87 & 3.6 & 174 & 143 & m1 & px & + 552 & 99 & 1908 & v410 & 897 & 552a & 5:35:17.5 & -5:21:46 & 156 & 0.46 & 14.50 & @xmath241.8 & 9 & 99 & & pxv & a , c + 560 & 98 & & v1334 & 927 & 560 & 5:35:17.9 & -5:15:33 & 235 & 0.60 & 14.17 & 3.2 & 3 & 471 & & pxv & + 570 & 99 & & & 937 & 570a & 5:35:17.9 & -5:25:34 & 0 & 0.15 & 14.73 & 0.8 & 170 & 133 & & px & a , b + 566 & 0 & & & 939 & 566 & 5:35:18.0 & -5:16:13 & 214 & 0.86 & 14.93 & 0.0 & 3 & 430 & & x & d + - & - & & & 967 & 9224 & 5:35:18.4 & -5:24:27 & 78 & 0.42 & 15.16 & 2.4 & 155 & 70 & m2.5 & x & a , b , f + 592 & 99 & & & 974 & 592 & 5:35:18.5 & -5:18:21 & 285 & 0.61 & 14.37 & 4.0 & 6 & 304 & m2.5 & px & + 597 & 0 & & & 994 & 597 & 5:35:18.8 & -5:14:46 & 307 & 0.17 & 14.02 & 0.1 & 4 & 519 & & xhv & + - & - & & & 998 & 9239 & 5:35:18.8 & -5:22:23 & 314 & 0.20 & 17.54 & 0.3 & 31 & 70 & & x & a + - & - & & & 1061 & 5066 & 5:35:20.0 & -5:18:47 & 71 & 0.22 & 17.31 & 0.1 & 11 & 281 & m9 & 0x00 & + 638 & 99 & & & 1077 & 638 & 5:35:20.0 & -5:29:12 & 354 & 1.01 & 17.03 & 3.2 & 171 & 353 & & pxv & + 681 & 99 & & v1524 & & 681a & 5:35:21.4 & -5:23:45 & 232 & 1.09 & 16.39 & 0.9 & 107 & 77 & k7 & px0v & a + 687 & 54 & & & 1158 & 687a & 5:35:21.7 & -5:21:47 & 231 & 0.49 & 14.66 & 2.1 & 39 & 124 & & xv & a , c + 709 & 99 & 1994 & & 1202 & 709 & 5:35:22.2 & -5:26:37 & 309 & 0.22 & 12.92 & 0.5 & 156 & 213 & m0.5 & px & + 722 & 98 & & & 1208 & 722 & 5:35:22.3 & -5:33:56 & 216 & 0.36 & 14.51 & 0.2 & 172 & 639 & m4.5 & px & + 727 & 99 & & v1528 & 1233 & 727 & 5:35:22.8 & -5:31:37 & 285 & 0.73 & 13.87 & 4.7 & 169 & 503 & m2 & pxhv & + 748 & 99 & & & 1279 & 748a & 5:35:24.1 & -5:21:33 & 277 & 0.35 & 14.77 & 2.0 & 46 & 159 & m3 & pxv & a + 767 & 99 & & & 1316 & 767 & 5:35:25.2 & -5:15:36 & 75 & 1.13 & 13.89 & @xmath244.1 & 16 & 485 & m2.5 & pxv & d + 776 & 99 & & v496 & 1328 & 776a & 5:35:25.4 & -5:21:52 & 73 & 0.50 & 14.20 & 1.8 & 56 & 162 & & pxv & a + 777 & 80 & & & 1327 & 777 & 5:35:25.5 & -5:21:36 & 347 & 1.19 & 15.15 & 1.2 & 52 & 173 & k6 & x & + 783 & 95 & & & & 783 & 5:35:25.5 & -5:34:03 & 283 & 0.40 & 14.17 & 0.8 & 168 & 656 & & px & + 797 & 99 & & & 1363 & 797 & 5:35:26.6 & -5:17:53 & 323 & 1.42 & 16.30 & 2.8 & 25 & 363 & & px & + - & - & & & 1425 & 3042 & 5:35:29.5 & -5:18:46 & 329 & 0.23 & 16.06 & 2.6 & 35 & 338 & & xv & + 841 & 99 & & & & 841 & 5:35:30.0 & -5:12:28 & 16 & 0.41 & 14.41 & 0.1 & 17 & 686 & m4 & phv & + 867 & 99 & 2075 & & 1463 & 867 & 5:35:31.3 & -5:18:56 & 212 & 0.29 & 12.19 & @xmath240.3 & 40 & 347 & k8 & pxhv & + 884 & 99 & & & & 884 & 5:35:32.4 & -5:14:25 & 287 & 1.32 & 14.91 & 2.8 & 24 & 589 & & p & + 893 & 99 & & & & 893 & 5:35:33.2 & -5:14:11 & 208 & 0.15 & 14.07 & 0.0 & 24 & 606 & & pv & + 906 & 99 & & & & 906 & 5:35:34.7 & -5:34:38 & 293 & 1.39 & 14.17 & 2.7 & 158 & 728 & m3 & pv & + 924 & 99 & & & & 924 & 5:35:36.5 & -5:34:19 & 73 & 1.04 & 16.63 & 1.3 & 156 & 722 & & p & + 945 & 99 & 2149 & & & 945 & 5:35:40.2 & -5:17:29 & 46 & 1.41 & 12.51 & 3.6 & 45 & 501 & b6 & p & + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a - prosser et al .
( 1994 ) + b - padgett et al .
( 1997 ) + c - simon et al .
( 1999 ) + d - koehler et al .
( 2006 ) + e - getman et al .
( 2005 ) + f - lucas et al .
( 2005 ) + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ccccccccccccccclcc 553 & 99 & 1911 & v1510 & 899 & 553a & 5:35:17.6 & -5:22:57 & 113 & 0.36 & 12.41 & 1.4 & 34 & 31 & k3.5 & pxv & a & 1 + 596 & 99 & 1927 & af & 986 & 596 & 5:35:18.7 & -5:23:14 & 78 & 0.30 & 13.06 & 2.2 & 75 & 34 & k3.5 & pxv & & 1 + & & & & 1085 & 3075 & 5:35:20.2 & -5:23:09 & 104 & 0.21 & & 0.4 & 76 & 57 & & x & & 1 + 182 & 48 & & & 127 & 182 & 5:34:58.0 & -5:29:41 & 77 & 0.11 & 16.56 & 0.2 & 216 & 467 & & x & & 2 + 290 & 99 & & & 266 & 290 & 5:35:06.4 & -5:27:05 & 84 & 0.11 & 15.48 & 1.3 & 214 & 268 & & px & & 2 + - & - & & & 1195 & 3034 & 5:35:22.3 & -5:18:09 & 227 & 0.11 & 15.67 & 0.8 & 15 & 326 & & x & & 2 + & & 1423 & & & 3114 & 5:34:19.5 & -5:27:12 & 168 & 0.56 & 10.52 & 0.1 & 255 & 881 & g5 & & & 3 + 58 & 0 & & & & 58 & 5:34:41.8 & -5:34:30 & 282 & 1.43 & 16.13 & 2.2 & 218 & 844 & & & & 3 + 61 & 46 & & & & 61 & 5:34:42.7 & -5:28:37 & 153 & 0.45 & 13.52 & 0.9 & 238 & 594 & & & & 3 + & & & & & & 5:34:45.8 & -5:30:58 & 324 & 0.46 & & 0.1 & 225 & 645 & & & & 3 + & & & & & 10343 & 5:35:01.4 & -5:24:13 & 292 & 0.28 & & 0.6 & 257 & 231 & & & & 3 + & & & & & 9221 & 5:35:18.3 & -5:24:39 & 96 & 0.67 & 16.72 & 2.2 & 160 & 81 & & & & 3 + & & & & & & 5:35:30.0 & -5:34:31 & 237 & 1.29 & & 0.3 & 163 & 699 & & & & 3 + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a - prosser et al .
( 1994 ) + 1 - binaries with angular separation @xmath110@xmath34 inside the exclusion zone + 2 - binaries with angular separation @xmath110@xmath315 outside the exclusion zone + 3 - binaries with angular separation @xmath111@xmath35 outside the exclusion zone but with no evidence of onc membership _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ | we have carried out a major survey for visual binaries towards the orion nebula cluster using hst images obtained with an h@xmath0 filter . among 781 likely onc members
more than 60@xmath1 from @xmath2 ori c , we find 78 multiple systems ( 75 binaries and 3 triples ) , of which 55 are new discoveries , in the range from 0@xmath31 to 1@xmath35 .
about 9 binaries are likely line - of - sight associations .
we find a binary fraction of 8.8%@xmath41.1% within the limited separation range from 67.5 to 675 au .
the field binary fraction in the same range is a factor 1.5 higher . within the range 150 au to 675 au
we find that t tauri associations have a factor 2.2 more binaries than the onc .
the binary separation distribution function of the onc shows unusual structure , with a sudden steep decrease in the number of binaries as the separation increases beyond 0@xmath35 , corresponding to 225 au .
we have measured the ratio of binaries wider than 0@xmath35 to binaries closer than 0@xmath35 as a function of distance from the trapezium , and find that this ratio is significantly depressed in the inner region of the onc .
the deficit of wide binaries in the central part of the cluster is likely due to dissolution or orbital change during their passage through the potential well of the inner cluster region .
many of the companions are likely to be brown dwarfs . |
quantum optical systems are ideal for the experimental test of the foundation of quantum mechanics @xcite as well as for the experimental implementation of quantum information protocols like quantum cryptography @xcite , quantum teleportation @xcite , quantum dense coding @xcite and quantum computation @xcite . in most of the above experiments the key point is the generation and the detection of entangled states .
while the generation of various kind of entangled states is now part of the daily routine of a good laboratory the detection can be a surprisingly difficult task .
the most typical example is probably the detection of bell states @xcite , for which it has been shown the impossibility to build a setup able to discriminate with 100% efficiency all the four bell states using only linear optical devices @xcite .
such impossibility to discriminate the states of an orthogonal basis is by no means restricted to entangled systems .
we will show that this difficulty is present also in the case of an orthogonal basis of a bipartite system which has been introduced in connection with non locality without entanglement .
non locality has always been associated with quantum entanglement . in a recent article
however @xcite bennett et al have provided a counterexample by showing the existence of an orthogonal set of states of a bipartite system which , although not entangled , are not distinguishable by means of bilocal measurements .
given two particles , each of which described by a three dimensional hilbert space , they construct the following orthogonal basis : @xmath0 where a , b label the two particles and @xmath1 are three orthogonal states for each particle .
the peculiar property of states ( [ primi ] ) is that they can not be reliably distinguished by two separate observers by means of any sequence of local operations even if they are allowed to exchange classical communication . in this paper
we propose an optical realization of states ( [ primi ] ) and investigate the possibility to fully discriminate them with a _ global measurement by means of linear elements_. a related problem has been investigated in connection with the possibility to discriminate bell states .
it has been shown @xcitethat it is not possible to perform a complete bell measurement on a product hilbert space of two two - level bosonic systems states by means of purely linear optical elements .
one might expect that this is due to the entangled nature of the bell states .
however , following the line of @xcite , we will show that also states ( [ primi ] ) are not fully distinguishable by a global measurement using only linear elements , even though they are not entangled .
[ figdomino ]
in our optical setup the three dimensional hilbert space of each subsystem is mapped into the single photon state of three different modes of the electromagnetic field .
the basis states @xmath1 for each of the two subsystems will therefore be of the form @xmath2 where @xmath3 ( 1=1,2,3 ) are bosonic creation operators of three orthogonal modes and @xmath4 is the vacuum state . in this
notation states ( [ primi ] ) are written as follows @xmath5 the impossibility to distinguish states ( [ stati ] ) by means of bilocal measurements implies that they are not distinguishable by measuring directly the photon number of each individual mode .
a first attempt to implement a collective measurement could be to mix the modes by means of linear devices and than to measure the output modes of such device .
however , following @xcite we will adopt a more general strategy .
we will assume to have at our disposal a set of as many additional modes as we like , here indicated with bosonic creation operators @xmath6 , with any number of photons we like and we will assume that these auxiliary modes can be mixed with modes @xmath7 in a black box .
the output modes of this box are linked to the input ones by a unitary transformation @xmath8 .
it has been shown @xcite that any such unitary transformations of modes can be obtained by means of linear optical devices , like beam splitters and phase shifters . to ensure the largest possible generality in our measurement apparatus we will assume the possibility to perform conditional measurements . in practice
this means what follows : assume that a measurement is made on one selected output mode while the other are kept in a delay loop and that , according to the outcome of the measurement , these modes are fed into a selected further black box , in a cascade setup ( see figure [ cascade ] ) .
the final assumption we will make is that our detectors have the ability to discriminate the number of incident photons .
this assumption is clearly unrealistic .
we will show , however , that even if such detectors were available , the measurement setup described above can not discriminate states ( [ stati ] ) . )
are mixed in a first `` box '' with auxiliary modes .
selected output mode is then measured and depending on its outcome the remaining output modes are fed in a new box . the process can be repeated over and over again , width=302 ]
in this section we will describe some symmetry properties of states ( [ stati ] ) which are not only interesting per se but also will turn out useful in the following .
consider the following transformation @xmath9 which permutes the modes of photon @xmath10 with the ones of photon @xmath11 : @xmath12 this is obviously a linear transformation . in the basis states @xmath13,@xmath14,@xmath15,@xmath16,@xmath17,@xmath18 @xmath19 takes the following matrix form @xmath20 the set of states ( [ stati ] ) is globally invariant under the action of @xmath19 since @xmath21 furthermore it is straightforward to verify that @xmath22 .
another linear transformation we will use in the following is the one which introduces a phase change of @xmath23 on states @xmath14 and @xmath17 leaving unaltered all the others . in matrix form @xmath24 the action of @xmath25 on states ( [ stati ] ) is simply @xmath26 with @xmath27 form a group which leaves states ( [ stati ] ) invariant . furthermore , by repeated action of @xmath28 and @xmath19 , it is possible to transform any @xmath29 into any other @xmath30 , with the exception of @xmath31 which is mapped onto itself .
for instance we can transform @xmath32 into a generic @xmath33 ( with k=1 ... 4 ) by acting with the operator @xmath34 this implies that the problem of how to generate the states ( [ stati ] ) reduces to the problem of how to generate one of them as the others can be obtained by repeated action of @xmath35 and @xmath19 and , as we have said already , this can be achieved by linear optical devices .
we will now show that the use of auxiliary photons in the measurement setup described in section [ setup ] does not help in increasing the distinguishability of states ( [ stati ] ) .
the argument is a generalization to our more complex set of states of the one used in @xcite in connection with the problem of distinguishing bell states with an analogous setup . in this section
we will outline the proof , leaving the details to appendix [ appa ] . as described already our measuring apparatus
consists of a cascade of `` black boxes '' , in which modes are linearly mixed , and partial measurements , which determine the sequence of unitary mixing .
the first of such black box , denoted by @xmath36 , is made out of linear optical elements and its input and output are a set of bosonic modes .
the joint input modes consist of our six `` system '' modes @xmath37 and an arbitrary number of auxiliary modes @xmath38 .
these input modes are unitarily mixed in the box into a set of output modes @xmath39 where the @xmath40 mode is the one on which a measurement will be performed
. the measurement outcome will determine the specific unitary mixing that will be performed in next step of the measurement , consisting of a second box @xmath41 .
while the measurement on mode @xmath40 is performed the photons in the remaining @xmath42 modes are kept in a waiting loop .
the whole measurement procedure consists of a cascade of conditional measurements as described above .
let s now look more in detail at the first block of the apparatus .
the input state of @xmath36 can be written as @xmath43 where @xmath44 is a polynomial of degree 2 and @xmath45 is a polynomial of arbitrary degree in the @xmath46 the corresponding output state is @xmath47 where @xmath48 and @xmath49 are nothing but @xmath45 e @xmath50 written in terms of the creation and annihilation operators at the output of @xmath36 .
we can expand @xmath51 in terms of decreasing powers of @xmath40 as follows @xmath52 in ( [ espan ] ) @xmath53 is the largest order in @xmath40 for the nine @xmath54 and by definition is independent on index @xmath55 ( @xmath56 can be zero for some @xmath55 ) .
analogously @xmath57 is defined as the order in @xmath58 of polynomial @xmath59 .
we can therefore rewrite ( [ output ] ) as @xmath60 out of the possible outcomes of the measurement of the number @xmath61 of photons in mode @xmath62 we will concentrate on two particular outcomes , namely those resulting in the highest number , @xmath63 and @xmath64 where @xmath65 .
the reason of this particular choice will be shortly evident .
let s suppose now that the number of photons on the selected mode @xmath62 is measured
. if @xmath61 is the outcome of such measurement the ( unormalised ) conditional state of the remaining modes can be we written as @xmath66 if the input states are to be distinguishable the conditional states @xmath67 _ must be orthogonal for each possible value of @xmath61 _ ,
i.e. @xmath68 in appendix [ appa ] we will show that the two conditions @xmath69 can be simultaneously satisfied if and only if the two conditions @xmath70 are simultaneously satisfied .
the important point is that ( [ without ] ) _ do not depend on the auxiliary input states_. it is easy to convince oneself that this is the case since from ( [ stcond ] ) follows that @xmath71 where @xmath72 is the conditional output state obtained from @xmath73 in the absence of auxiliary photons when @xmath61 photons are measured in mode @xmath62 .
the central point of this section is that the fact that condition ( [ withaux ] ) implies condition ( [ without ] ) is equivalent to say that any pair of states @xmath73 , @xmath74 are distinguishable in the presence of auxiliary photons only if they are distinguishable in the absence of auxiliary photons . in other words
auxiliary photons do not improve complete distinguishability .
we will now show that it is impossible for states ( [ stati ] ) to satisfy @xmath75 for all @xmath55 , @xmath76 ( @xmath77 ) . in the absence of auxiliary photons
states @xmath29 can be written in terms of a polynomial of creation operators as @xmath78 let s now define the creation operator vector as @xmath79 where the @xmath80 are a possible set of ( empty ) auxiliary modes . since the @xmath73 are two photon states they can be written in terms of a real symmetric matrix @xmath81 as follows @xmath82 where the exact form of @xmath81 can be obtained from ( [ stati ] ) .
if @xmath83 is a generic unitary matrix transforming the input modes into the output ones of our apparatus than @xmath84 where @xmath85 and @xmath86 with @xmath87 corresponding to the detected output mode .
states ( [ psi ] ) can than be written as @xmath88 where @xmath89 is the generic matrix element of @xmath90 whose dimension @xmath91 corresponds to the number of output modes involved .
let s write @xmath83 as @xmath92 where @xmath93 are the element of the first column of the matrix and @xmath94 ( with @xmath95 ) are vectors of dimension @xmath96 representing the remaining elements of row @xmath97 . as a consequence of the unitarity of @xmath83 we have @xmath98
we define the columns vector @xmath99 whose elements are the first 6 elements of the @xmath100 column of @xmath83 : @xmath101 we recall that @xmath53 is the highest degree of @xmath87 in polynomials @xmath102 for all values of @xmath55 , in other words the maximum number of photons which can be detected in @xmath62 for all possible input states @xmath103
. obviously @xmath53 can assume only values @xmath104 , depending on the specific choice of @xmath83 and @xmath62 .
we will now show that for all possible value of @xmath53 it is impossible to satisfy simultaneously ( [ condo1 ] ) and ( [ condo2 ] ) . _
@xmath105 _ : this corresponds to a bad choice of mode @xmath62 , as the monitored mode would be decoupled from the input ones for all possible input state . _
@xmath106 _ : this corresponds to @xmath107 for all value of @xmath55 ( see ( [ svlp ] ) ) .
this implies that @xmath108 the above relation is a constrain on @xmath99 which we will now show to be incompatible with ( [ condo1 ] ) . to this end
we note that from ( [ svlp ] ) follows that after the detection of one photon in mode @xmath62 the remaining modes are left in the ( unormalised ) conditional state @xmath109 where for convenience of notation we have introduced the vector @xmath110 from ( [ statocondns1 ] ) and ( [ vettm0 ] ) follows that the trivial solution @xmath111 implies @xmath112 @xmath113 , i.e. @xmath105 .
we must therefore look for possible nontrivial solutions of eq .
( [ ns1 ] ) compatible with ( [ condo1 ] ) , which in this particular case reads @xmath114 however , as shown in appendix [ appb ] conditions ( [ ns1 ] ) and ( [ condo1ns1 ] ) are compatible only with the trivial solution .
this implies that it is not possible to build a complete discriminator for @xmath106 . _
@xmath115 _ this corresponds a non zero probability to measure two photons in mode @xmath62 for some @xmath73 which implies @xmath116 for at least one value of @xmath55 . on the other hand condition ( [ condo1 ] ) can be satisfied in this specific case if and only if @xmath116 for at most one value of @xmath55 , which we will denote by @xmath117 .
condition ( [ condo1 ] ) than becomes @xmath118 and ( [ condo2 ] ) becomes equivalent to condition ( [ condo1ns1 ] ) . in order to complete our proof
it will therefore be enough to show that whatever the value of @xmath117 conditions ( [ ns2 ] ) and ( [ condo1ns1 ] ) can not be simultaneously satisfied .
suppose in particular that they are not satisfied for @xmath119 , the symmetry analysis carried out in the previous section immediately lead to the conclusion that they can not be satisfied by any other value @xmath117 ( apart from @xmath120 ) .
we have shown that it is always possible to build a linear operator @xmath121 which transforms @xmath32 into @xmath73 ( @xmath122 ) and leaves the set of states @xmath123 globally invariant . if there were a linear operator @xmath124 such to satisfy conditions ( [ ns2 ] ) and ( [ condo1ns1 ] ) for any value of @xmath125 than @xmath126 would satisfy the same conditions for @xmath127 , which contradicts our initial assumption .
the problem than reduces to the analysis of the cases @xmath128 and @xmath127 . such analysis , straightforward but tedious ( see appendix [ appb ] ) , leads to the result that indeed for both values of @xmath117 conditions ( [ ns2 ] ) and ( [ condo1ns1 ] ) are incompatible .
in this paper we have proposed an optical realization of states ( [ primi ] ) .
bennett et al .
@xcite have shown that they can not be discriminated by means of local action and classical communication .
we have demonstrated that to add the possibility of global interference it is still not sufficient . in other words
we have shown the impossibility to fully discriminate them by means of a global measurement using linear elements , like beam splitters and phase shifters , delay lines and electronically switched linear elements , photodetectors and auxiliary photons . the impossibility to implement such a measurement
has already been shown for the set of maximally entangled bell states .
we have proved an analogous no - go theorem for a set of states which , although non local , are not entangled .
this opens new questions on which class of photon states can be in general be fully discriminated by means of linear optical systems .
we would like to thank c.h .
bennett , j.calsamiglia , d.divincenzo , a.k.ekert , i.jex , n.ltkenhaus for helpful discussions .
this work was supported in part by the eu under grants tmr - erb fmr xct 96 - 0087 - `` the physics of quantum information '' ist - 1999 - 11053 - equip,``entanglement in quantum information processing and communication '' .
are simultaneously satisfied . from ( [ stcond ] )
follows that the scalar product between the ( unormalised ) states @xmath131 obtained after the measurement of @xmath61 photons in mode @xmath62 is @xmath132 above we have used the fact that @xmath136\!\!=\!0 $ ] @xcite and introduced the completeness relation @xmath137 , where @xmath138 is a fock states of the relevant modes .
note that only the term corresponding to @xmath139 survives .
[ regoledicomm ] @xmath144=\left[{\tilde{q}}_a^{(m)},{\tilde{q}}_{\psi_i}^{(n_s)\dag}\right]=0\quad\forall\;m , n\ ] ] and that : @xmath145= n_a n_s\cdot{\tilde{q}}_a^{(n_a)}{\tilde{q}}_{\psi_i}^{(n_s)\dag}\label{comm1}\ ] ] as all the states @xmath73 contain a definite number of photons , namely @xmath149 , @xmath150 is a homogeneous polynomial of degree @xmath151 in @xmath87 and @xmath152 and therefore the generic @xmath153 is a homogeneous polynomial of degree @xmath154 in @xmath152 . as a consequence @xmath155 . from this
follows that the first term at the right hand side of ( [ c10comm ] ) is equal to zero . in this appendix
we will prove that condition ( [ ns2 ] ) is incompatible with ( [ condo1ns1 ] ) for both @xmath161 and @xmath162 . to this goal it is helpful to define a matrix @xmath163 , linear combination of the @xmath81 : @xmath164 we now impose condition ( [ ns2 ] ) with @xmath161 on vector @xmath99 , i.e. we impose that @xmath168 is the only nonzero coefficient .
we have therefore to equal to zero all coefficients in ( [ m00 ] ) except the one multiplying @xmath169 .
the only solution compatible with this condition is @xmath170 we will now show that conditions ( [ condo1ns1 ] ) , ( [ ns2 ] ) can not be simultaneously fulfilled with @xmath162 , i.e that the only non zero coefficient of @xmath175 is the one multiplying the coefficient @xmath176 .
and @xmath178 reduces to @xmath179+\nonumber\\ & + & \mu_{-1 } u_0({\mathbf{\b{r}}}_3-{\mathbf{\b{r}}}_4 ) + ( \mu_2-\mu_{-2 } ) u{\mathbf{\b{r}}}_2+\nonumber\\ & + & ( \mu_4+\mu_{-4 } ) u_0{\mathbf{\b{r}}}_5 + \mu_3 u({\mathbf{\b{r}}}_1+{\mathbf{\b{r}}}_2)+\\ & + & \mu_{-3 } u({\mathbf{\b{r}}}_1-{\mathbf{\b{r}}}_2)\}\nonumber\end{aligned}\ ] ] 99 c. h. bennett , d. divincenzo , c.a .
fuchs , t. mor , e. rain , p.w.shor , j.a .
smolin , w.k .
wootters , phys .
rev.a * 59 * , 1070 ( 1999 ) .
a.zeilinger , rev .
phys . * 71 * , 288 ( 1999 ) .
t. jennewein , c. simon , g. weihs , h. weinfurter , and a. zeilinger , phys .
lett . * 84 * , 4729 ( 2000 ) d. bouwmeester , j .- w .
pan , k. mattle , m. eibl , h. weinfurter and a. zeilinger , nature * 390 * , 575 ( 1997 ) .
k. mattle , h. weinfurter , p.g .
kwiat and a. zeilinger , phys .
lett . * 76 * , 4656 ( 1997 ) .
e. knill , r. laflamme and g.j .
milburn , nature * 409 * , 49 ( 2001 ) .
m. michler , k. mattle , h. weinfurter and a. zeilinger , phys .
a * 53 * , r1209 ( 1996 ) ; h. weinfurter europhys .
* 25 * 559 ( 1994 ) ; s. l. bernstein and a. mann , phys . rev . a * 51 * , r1727 ( 1995 ) .
l.vaidman and n.yoran , phys .
rev . a * 59 * , 116 ( 1999 ) .
n.ltkenhaus , j. calsamiglia , k .- a .
suominen , phys .
rev . a * 59 * , 3295 ( 1999 ) . j.calsamiglia and n.ltkenhaus , app .
b , 67 - 71 ( 2001 ) . m.zukowski , a. zeilinger , m.a.horne , phys . rev . a * 55 * , 2564 ( 1997 ) .
m.reck , a.zeilinger,h.j.bernstein and p.bertani phys.rev.lett . *
73 * , 58 ( 1994 ) . to demonstrate these commutation relations write @xmath143=0 $ ] as a power series of normally ordered @xmath87 , @xmath62 operators , and
impose that each term of this power series must vanish .
the coefficients with power @xmath57 , @xmath185 , @xmath53 , and @xmath186 include the commutators we want to calculate . | in a recent paper bennett et al .
@xcite have shown the existence of a basis of product states of a bipartite system with manifest non - local properties .
in particular these states can not be completely discriminated by means of bilocal measurements . in this paper
we propose an optical realization of these states and we will show that they can not be completely discriminate by means of a global measurement using only optical linear elements , conditional transformation and auxiliary photons . |
as exposed in detail by dr .
denig in his very nice and introductory talk on the kloe status and first results ( see a. g. denig s talk in these proceedings ) , @xmath6 radiative decays into two pseudoscalar mesons and a photon are very interesting processes to be measured .
their accurate measurements could offer us the possibility of shedding some light on the presently unclear nature of scalars ( in particular , in those channels involving the @xmath7 and @xmath2 resonances ) .
these decays , and in general all the associated @xmath8 radiative decays with @xmath9 and @xmath10 , are not only a challenge for experimentalists but also for theorists .
the reason for that is at least twofold .
first , the center of mass energy of these processes is around 1 gev , an energy region that is , on the one side , too low for perturbative qcd to be applicable and , on the other , too high for chiral perturbation theory ( chpt ) predictions to be reliable .
accordingly , these processes provide us with an excellent laboratory for testing our knowledge of hadron physics in the 1 gev energy domain .
second , due to the quantum numbers of the initial vector and of the final photon , both with @xmath11 , the remaining two pseudoscalar meson system is in a @xmath12 configuration states are very suppressed because the nearest tensorial resonances , @xmath13 and @xmath14 , are well above the lightest vector masses @xcite . ] , whose quantum numbers correspond precisely to those of a scalar state .
so then , these vector radiative decays may help us to establish the nature of these scalar resonances , nowadays very controversial molecules @xcite or ordinary @xmath15 mesons @xcite . ]
@xcite , and their poorly known properties ( masses and decay widths ) @xcite , thus adding further interest but also complexity to this 1 gev energy region . in this presentation
, we propose the linear sigma model ( l@xmath0 m ) as an appropriate framework for incorporating the lightest scalar resonances and their pole effects into a chpt inspired context .
this will allow us to benefit from the common origin of chpt and the l@xmath0 m to improve the chiral loop predictions for the @xmath8 decays exploiting the complementarity of both approaches . as well known , chpt is the established theory of the pseudoscalar interactions at low energy .
however , it is not reliable at energies of a typical vector meson mass and scalar resonance poles are not explicitly included . as a result
, chpt inspired loop models can give rough estimates for @xmath16 but will hardly be able to reproduce the observed enhancements in the invariant mass spectra .
the l@xmath0 m is a much simpler model dealing similarly with pseudoscalar interactions but taking into account scalar resonances in a systematic and definite way .
consequently , the l@xmath0 m should be able to reproduce the resonance peaks in the spectra associated to the @xmath17 states and , although it does not provide a complete framework for the pseudoscalar meson physics , this model might be of relevance in describing the scalar resonance effects when linked to a well established chpt context .
the novosibirsk cmd-2 and snd collaborations have reported very recently the branching ratio and the @xmath3 invariant mass distribution for the @xmath1 decay . for the branching ratio ,
the cmd-2 collaboration reports @xmath18 @xcite , while the snd result is , consistently , @xmath19 @xcite . the observed invariant mass distribution shows a significant enhancement at large @xmath3 invariant mass that , according to refs .
@xcite , could be interpreted as a manifestation of a sizeable contribution of the @xmath20 intermediate state .
the last issue of the review of particle physics , including the previous cmd-2 value together with an older snd measurement @xcite announces @xmath21 @xcite , while the preliminary result presented in this conference by the kloe collaboration is @xmath22 ( see a.g .
denig s contribution ) .
this and other radiative @xmath6 decays are intensively investigated at the frascati @xmath6-factory da@xmath23ne @xcite . on the theoretical side ,
the @xmath24 decays have been extensively studied @xcite .
specifically , it has been shown that the intermediate vector meson contributions to @xmath1 lead to a small @xmath25 @xcite , whereas a chiral loop model closely linked to standard chpt predicts @xmath26 @xcite .
needless to say , the scalar resonance effects and , in particular , the resonance pole associated to the @xmath27(980 ) were not contemplated in these two schemes .
the recent experimental data from novosibirsk and frascati for both the branching ratio and the @xmath3 invariant mass spectrum seem therefore to disfavour these predictions based on vector meson exchange and/or a simple extrapolation of chpt ideas .
if we rely on the resonance picture , it is clear that the @xmath2 scalar meson lying just below the @xmath6 mass and having the appropriate quantum numbers should play an important rle in the @xmath1 decay .
several theoretical attempts to describe the effects of scalars in @xmath6 radiative decays have appeared so far : the no structure " model @xcite , the @xmath28 model @xcite , where the @xmath29 amplitude is generated through a loop of charged kaons , and the chiral unitary approach @xmath30 @xcite , where the decay @xmath1 occurs through a loop of charged kaons that subsequently annihilate into @xmath31 . in the two former cases the scalar resonances are included _
ad hoc _ while in the latter they are generated dynamically by unitarizing the one - loop amplitudes . as an example of our work
, we discuss the @xmath1 decay mode .
this process is the easiest to calculate since it only involves the @xmath2 scalar resonance whose properties are quite well known .
this circumstance reduces to a minimum the uncertainties of the calculation and will render our theoretical prediction more definite and solid .
this presentation is based on a more extensive analysis published in ref .
the vector meson initiated @xmath8 decays can not be treated in strict chpt .
this theory has to be extended to incorporate on - shell vector meson fields . at lowest order ,
this may be easily achieved by means of the @xmath32 chpt lagrangian @xmath33 where @xmath34 mev at this order , @xmath35 with @xmath36 being the usual pseudoscalar nonet matrix , and @xmath37 in the isospin limit .
the covariant derivative , now enlarged to include vector mesons , is defined as @xmath38 - i g [ v_\mu , u]$ ] , with @xmath39 being the quark charge matrix and @xmath40 the additional matrix containing the nonet of ideally mixed vector meson fields .
the diagonal elements of @xmath41 are @xmath42 and @xmath6 .
there is no tree - level contribution from this lagrangian to the @xmath1 amplitude and at the one - loop level one needs to compute the set of diagrams shown in ref .
a straightforward calculation leads to the following finite amplitude for @xmath43 : @xmath44 where @xmath45 is the loop integral function and @xmath46 is the invariant mass of the final pseudoscalar system ( see ref .
@xcite for further details ) .
the coupling constant @xmath47 is defined through the strong amplitude @xmath48 and is the part beyond standard chpt which is fixed from phenomenology .
the four - pseudoscalar amplitude is instead a standard chpt amplitude if only the @xmath49 contribution is taken into account as in ref .
@xcite . ] which is found to be @xmath50 in eqs .
( [ aphichpt ] , [ a4pchptphys ] ) , a value of @xmath51 is taken for the @xmath52-@xmath53 mixing angle .
this value , based on classical arguments of nonet symmetry , is in fairly agreement with recent phenomenological estimates @xcite . as a function of the @xmath54 invariant mass in a chiral loop model .
experimental data are taken from ref .
@xcite . ] the invariant mass distribution for the @xmath1 decay is then given by the spectrum ( see fig .
[ figchpt ] ) @xmath55 & \times\ |l(m^2_{\pi^0\eta})|^2 \end{array}\ ] ] integrating eq .
( [ dgchpt ] ) over the whole physical region one obtains for the branching ratio @xmath56 as expected , fig .
[ figchpt ] shows that our chiral loop approach gives a reasonable prediction for the lower part of the spectrum but fails to reproduce the observed enhancement in its higher part , where @xmath2-resonance effects should manifest . as a consequence
, the branching ratio is below the experimental value by about a factor of 2 .
to analyze the scalar resonance effects in the @xmath8 decay amplitudes , we use the linear sigma model ( l@xmath0 m ) @xcite as an adequate framework for describing such effects .
it is a well - defined @xmath57 chiral model which incorporates _ ab initio _ both the nonet of pseudoscalar mesons together with its chiral partner , the scalar mesons nonet .
recently , this model has shown to be rather phenomenologically successful in studying the implications of chiral symmetry for the controversial scalar sector of qcd @xcite . in this context
, the @xmath8 decays proceed through a loop of charged pseudoscalar mesons emitted by the initial vector that , due to the additional emission of a photon , can rescatter into pairs of neutral pseudoscalars with the quantum numbers of a scalar state .
the scalar resonances are then expected to play an essential rle in this rescattering process and the l@xmath0 m will be shown particularly appropriate to fix the corresponding amplitudes . for the case of @xmath1 ,
the dominant contribution arises exclusively from a loop of charged kaons coupling . ]
that subsequently rescatters into the final @xmath58 @xmath3 state . in the l@xmath0 m
, the @xmath59 rescattering amplitude is driven by a contact term , a term with an @xmath27 exchanged in the @xmath60-channel , and two terms with a @xmath61 ( _ i.e . _ the strange @xmath62 scalar resonance ) exchanged in the @xmath63- and @xmath64-channels .
however , the latter @xmath61-exchange contributions are absent for an @xmath52-@xmath53 mixing angle @xmath51 since the @xmath65 coupling constant appearing in one of the @xmath61 vertices vanishes contributions . ] .
a straightforward calculation of the @xmath1 decay amplitude leads to an expression identical to that in eq .
( [ aphichpt ] ) but with the four - pseudoscalar amplitude now computed in a l@xmath0 m context . in this case , the amplitude is just @xmath66 where @xmath67 is the @xmath27 propagator .
it is worth mentioning a few remarks on the four - pseudoscalar l@xmath0 m amplitude in eq .
( [ a4plsigmamphys ] ) and its chpt counterpart in eq .
( [ a4pchptphys ] ) : _ i ) _ for @xmath68 and ignoring @xmath69-breaking in the pseudoscalar masses and decay constants , the l@xmath0 m amplitude reduces to the chpt one .
this means that the large scalar mass limit of the l@xmath0 m mimics perfectly ( in the @xmath69 limit ) the contributions from the derivative and massive terms of the chpt lagrangian ( [ chptlag ] ) .
this , we believe , is the main virtue of our approach and makes the use of the l@xmath0 m reliable ; _ ii ) _ the reason for the l@xmath0 m and chpt yielding slightly different amplitudes in the @xmath68 limit is because of the way @xmath69-symmetry is broken in the two approaches .
while in the l@xmath0 m a non @xmath69 symmetric choice of the vacuum expectation values makes simultaneously @xmath70 and @xmath71 @xcite , in chpt @xmath70 is already present at tree level whereas @xmath71 is only achieved at next - to - leading order ; _ iii ) _ the vicinity of the @xmath2 mass to the @xmath3 production threshold makes unavoidable the presence of the @xmath27 propagator to take into account the pole effects . its inclusion should guarantee the proper behavior of the higher part of the @xmath3 invariant mass spectrum ; _ iv ) _ concerning the @xmath27 propagator , the opening of the @xmath72 channel close to the @xmath2 mass generates some uncertainties about which precise form for that propagator should be used . a first possibility consists in using a breit - wigner propagator with an energy dependent width to incorporate the known kinematic corrections : @xmath73 , where @xmath74 and @xmath75 \left[1-\frac{(m_a - m_b)^2}{s}\right]}\ , \theta ( \sqrt{s}-(m_a+m_b))\ , \ ] ] for @xmath76 ( see ref .
@xcite for the coupling constants @xmath77 ) .
another interesting and widely accepted option was proposed by flatt time ago specifically to the two - channel @xmath27 resonance @xcite : @xmath78 , where @xmath79 and @xmath80 i\sqrt{\frac{4m^2_k}{s}-1 } & \ \mbox{for $ \sqrt{s } < 2m_k$ } \end{array } \right.\ ] ] for @xmath81 .
the relative narrowness of the observed @xmath82 peak around 980 mev is then explained by the action of unitarity and analyticity at the @xmath72 threshold . due to these distinct possibilities to deal with the @xmath27 propagator , as well as other differences introduced when implementing and fitting the basic l@xmath0 m lagrangian by several authors , a set of predictions can be obtained for the four - pseudoscalar amplitude ( [ a4plsigmamphys ] ) . in turn , these various amplitudes have to substitute the four - pseudoscalar chpt amplitude in eq .
( [ dgchpt ] ) to finally obtain the corresponding invariant mass distributions of the @xmath1 decay mode . as a function of the @xmath54 invariant mass in the l@xmath0 m .
the dotted , dashed and solid lines correspond to the versions of the l@xmath0 m proposed by refs .
@xcite respectively .
experimental data are taken from ref .
@xcite . ]
we start our discussion along the lines of ref .
@xcite taking for the @xmath27 propagator the simple breit - wigner prescription . the use of this propagator for the l@xmath0 m amplitude eq .
( [ a4plsigmamphys ] ) and its insertion in eq .
( [ dgchpt ] ) predicts the @xmath54 invariant mass spectrum shown by the dotted line in fig .
[ figlsigmam ] .
integrating over the whole physical region leads to the branching ratio @xmath83 in agreement with the experimental branching ratio .
however , since the simple expression used for the @xmath27 propagator implies a large @xmath27-width ( @xmath84 mev @xcite ) , the desired enhancement in the invariant mass spectrum appears in its central part rather than around the @xmath27 peak .
this unpleasant feature is partially corrected when turning to the proposal by trnqvist @xcite , where a gaussian form factor related to the finite size of physical mesons and depending on the final cm - momentum is introduced to describe the decays of scalar resonances in this approach . as a result , the decay width of @xmath27(980 ) into @xmath3
is reduced ( @xmath85 mev @xcite ) without affecting that of @xmath27(980 ) into @xmath72 .
this fact produces an enhancement in the spectrum for the higher values of the @xmath54 invariant mass , as shown by the dashed line in fig .
[ figlsigmam ] .
the integrated branching ratio is then predicted to be @xmath86 none of these drawbacks are encountered in the treatment proposed by shabalin @xcite , where the flatt corrections ( indeed , a more precise form of them ) are introduced in the @xmath27 propagator .
the @xmath27 width is then drastically reduced to a more acceptable visible width of @xmath87 mev . within this approach ,
our prediction for the @xmath54 invariant mass is shown by the solid line in fig .
[ figlsigmam ] and the integrated branching ratio is @xmath88 both the spectrum and the branching ratio are in nice agreement with the experimental data @xcite . the fact that shabalin s treatment incorporates the flatt corrections to the @xmath27 resonant shape has played a major rle in this achievement .
in this presentation , i have discussed a new amplitude for the @xmath1 process that includes the effects of the @xmath2 scalar resonance .
the l@xmath0 m has shown to be a very suitable framework for including the @xmath27 pole effects in a systematic and definite way .
an explanation of the higher part of the @xmath3 invariant mass spectrum is then achieved . in the low invariant mass region ,
the result coincides with that coming from a chiral loop model calculation , thus making our whole approach reliable . as a result of our analysis
, it is safely concluded that the @xmath1 branching ratio is in the range @xmath4@xmath5 , a prediction that is compatible with the present experimental data .
nevertheless , the uncertainties affecting these predictions suggest that more refined analyses are needed , particularly when the higher accuracy data from da@xmath23ne will be available .
i would like to express my gratitude to the hadron structure2000 organizing committee for the opportunity of presenting this contribution , and for the pleasant and interesting conference we have enjoyed . | a prediction for the contributions of chiral loops and the l@xmath0 m model to the radiative @xmath1 decay mode is presented .
the l@xmath0 m is used as an appropriate framework for describing the pole effects of the @xmath2 scalar resonance . as a result , a better agreement with present available data is achieved for the higher part of the @xmath3 invariant mass spectrum . for the branching ratio , a value of @xmath4@xmath5 is found .
lnf00/032(p ) + hep - ph/0012050 + december 2000 * chiral loop and l@xmath0 m predictions for @xmath1 * + r. escribano + _ infn - laboratori nazionali di frascati , p.o .
box 13 , i-00044 frascati , italy _ |
neutrino oscillations were reported last year by the super - kamiokande collaboration @xcite to have been convincingly seen by the super - kamiokande detector in kamioka , japan , in the study of the atmospheric neutrino problem .
this has rekindled interest to explain also other deviations from expected results in neutrino physics in terms of neutrino oscillations .
there are at present essentially four different types of experiments looking for neutrino oscillations : solar neutrino experiments , atmospheric neutrino experiments , accelerator neutrino experiments , and reactor neutrino experiments .
most analyses of neutrino data so far have been performed in two - flavor models .
although this is quite illustrative , it can lead to wrong conclusions concerning the necessary number of degrees of freedom to explain data .
it is , e.g. , possible that the four neutrino scenario , advocated by several authors , is an artifact of this simplification .
recently , several papers @xcite have been published pertaining at demonstrating the possibility to obtain an overall description in terms of a three - flavor model . most of these investigations have focused on demonstrating the plausibility of the scenario , without truly fitting all the data . here
we would like to report on the first stage of a global analysis of all the neutrino oscillation data in a three - flavor model .
we will assume that @xmath5 nonconservation is negligible at the present level of experimental accuracy .
thus , the cabibbo - kobayashi - maskawa ( ckm ) mixing matrix for the neutrinos is real .
there are then in principle five parameters that can be fitted in this model : three mixing angles and two mass squared differences .
( this is in contrast with two two - flavor models , which give only four parameters : two mixing angles and two mass squared differences . )
the two mass squared differences in the three - flavor model enter the argument of the sinodial function in combination with the ratio @xmath6 , where @xmath7 is the source - detector distance and @xmath8 is the neutrino energy .
the argument of the sinodial functions related to the energy dependence can then be treated as yet another angle . in this case , the probability function can be seen as given by a rotation of the flavor states by five angles .
this is in analogy to the two - flavor case , which can be seen as a rotation by two angles , one of which is given by @xmath9 , @xmath10 being the mass squared difference of the two neutrino masses . in the present analysis
, we will make the simplifying assumption that the relevant ranges of mass squared differences are such that the mikheyev - smirnov - wolfenstein ( msw ) effect @xcite in the sun is negligible .
this excludes essentially the sensitivity region @xmath11@xmath12 ev@xmath13 .
the absence of any day - night variation in the super - kamiokande and kamiokande data is also consistent with neglecting the msw resonance effect in the earth for the electron - neutrino data .
the differing results for the solar neutrino experiments present another problem . since the super - kamiokande and kamiokande experiments can resolve the energy for the solar neutrinos , these experiments are of a rather different and more detailed kind than the radio - chemical experiments .
it is possible to vary the disappearance rate by going to a mass squared realm , where the arguments , i.e. , the angles that depend on the mass squared differences , are sensitive to the energy of the neutrinos , since this is different for the different detection methods .
however , in this case , the energy domains for the different experiments do not vary in the way expected , making the cl experiment the smallest ( @xmath14 ) and the ga experiments the largest ( @xmath15 ) of the expected rate .
one would , in fact , expect the super - kamiokande experiment , with the highest mean energy for the neutrinos , to give the smallest result , whereas in reality it is in between . confronting this with the apparent lack of variation in the experimental solar neutrino detection probability with energy , at least up to approximately 13 mev , in the super - kamiokande data
, we will adopt the point of view that the mass squared differences are in a range that leads to the solar neutrino oscillations being averaged out to their classical values . by fitting the three mixing angles
, we then account for the weighted mean of all the solar neutrino experiments . by nature of the sensitivity level of the accelerator and atmospheric neutrino experiments
, we will choose the mass squared difference ranges as @xmath16 and @xmath17 ev@xmath13 . here
@xmath18 will regulate the accelerator experiments and the multi - gev atmospheric neutrino data , whereas @xmath10 regulates the low energy atmospheric experiments .
this choice is thus not inherited from two - flavor model analyses , but arises from the sensitivity of the different types of experiments .
it is in principle possible that there would be a resonance effect for the atmospheric electron - neutrinos passing through earth .
our chosen ranges for the mass squared differences exclude this effect in the present fit , since @xmath19 ev@xmath13 . since the reactor experiments are disappearance experiments with no signal seen , they can at best exclude certain ranges for the mass squared differences due to their sensitivity level .
the two ranges of mass squared differences chosen here respect this sensitivity level .
the paper is organized as follows . in sec .
[ sec : form ] , we briefly discuss the formalism used in our analysis including notations , gaussian averaging , and minimization . in sec .
[ sec : cod ] , we discuss the choice of data from different experiments used in the minimization procedure .
[ sec : na ] describes the minimization procedure and includes the results thereof .
we also show that our results are consistent with the experimental data , assuming that all of the deviations from the expected values are indeed related to the physics of neutrino oscillations .
finally , sec .
[ sec : dc ] presents a discussion of the results and also our main conclusions .
in the present analysis , we will use the plane - wave approximation to describe neutrino oscillations . in this approximation ,
a neutrino state @xmath20 with flavor @xmath21 is a linear combination of neutrino mass eigenstates @xmath22 such that @xmath23 where the @xmath24 s are entries in a unitary @xmath25 matrix @xmath26 .
the unitary matrix @xmath26 is given by @xmath27 a convenient parametrization for @xmath26 is @xcite @xmath28 where @xmath29 and @xmath30 for @xmath31 .
this is the so called standard representation of the ckm mixing matrix .
we have here put the @xmath5 phase equal to zero in the ckm matrix .
this means that @xmath32 for @xmath33 and @xmath34 .
the probability of transition from a neutrino flavor state @xmath21 to a neutrino flavor state @xmath35 is given by @xmath36 where @xmath37 is kronecker s delta and @xmath38 .
the three mass squared differences @xmath39 , @xmath40 , and @xmath41 are not linearly independent , since they satisfy the relation @xmath42 therefore , we will define @xmath43 and thus @xmath44 . from eq .
( [ eq : p_ab ] ) it is clear that @xmath45 and since we have assumed that @xmath5 is conserved , i.e. , the ckm mixing matrix @xmath26 is real , it holds that @xmath46 .
this implies that @xmath47 since in practice the neutrino wave is neither detected nor produced with sharp energy or with well defined propagation length , we have to average over the @xmath6 dependence and other uncertainties in the detection and emission of the neutrino wave
. we will here use the gaussian average , which is defined by @xmath48 where @xmath49 here @xmath50 and @xmath51 are the expectation value and standard deviation , respectively . by taking the gaussian average of eq .
( [ eq : p_ab ] ) , we obtain the average transition probabilities from flavor @xmath52 to flavor @xmath53 as @xmath54 , \nonumber \\ & & \alpha,\beta = e , \mu , \tau .
\label{eq:<p_ab>}\end{aligned}\ ] ] the physical interpretations of the parameters @xmath50 and @xmath51 are the following : the parameter @xmath50 deals with the sensitivity of an experiment and is given by @xmath55 , where @xmath56 should be given in km / gev or m / mev , and @xmath57 should be measured in @xmath58 . note that we will here use @xmath59 .
the parameter @xmath51 is a so called damping factor . for large values of @xmath51 , the dependence on the mass squared differences
will be completely washed out , since @xmath60 when @xmath61 , and the transition probabilities @xmath62 will just be dependent on the mixing angles @xmath63 , where @xmath64 .
this corresponds to the classical limit . in the other limit , @xmath65 ,
we will just regain eq .
( [ eq : p_ab ] ) from eq .
( [ eq:<p_ab > ] ) , i.e. , @xmath66 . in order to obtain the mixing angles @xmath63 , where @xmath67 , and the mass squared differences @xmath68 and @xmath69 ,
we minimize the following object function @xmath70 where @xmath71 and @xmath72 is the number of neutrino oscillation experiments .
the function @xmath73 is a least square method error function for the relative errors of the neutrino oscillation experiments with weights @xmath74 , where @xmath75 .
the quantities @xmath76 and @xmath77 are the theoretical ( model ) and experimental transition probabilities for the @xmath78th experiment , respectively .
the reason why we are using relative errors in the error function and not absolute errors ( as is usually done ) is due to the fact that the different categories of experiments have rather different numerical ranges for the experimental transition probabilities . in the rest of this paper
, we will write @xmath79 , @xmath7 , and @xmath8 instead of @xmath80 , @xmath81 , and @xmath82 , respectively , to keep the notations simpler . in accordance with the discussion in the introduction
, the following constraints will be used in the minimization procedure @xmath83 the weights in the object function ( [ eq : obj ] ) will be chosen as @xmath84 where @xmath85 is the number of experiments used in the minimization of the same category ( solar , atmospheric , accelerator , or reactor ) and @xmath86 is the relative error in the @xmath78th experiment .
our choice for the weights is motivated by our wish to treat all categories of experiments on the same footing ( the factor @xmath87 ) , and enhance experiments with small relative errors ( the factor @xmath88 ) . here again , relative errors are used , since the range of values of the probabilities are rather different . however , in all our cases , @xmath89 .
higher powers of @xmath88 would enhance the good experiments too much . since @xmath90 is a sum of the statistical and systematical errors , an underestimation of the latter will , of course , also overemphasize the relative weight of a certain experiment .
for the damping factor of an experiment , we will use @xmath91 where @xmath92 is a combination of the uncertainties coming from ( a ) the length of the detector and ( b ) the distance between the detector and the neutrino source and @xmath93 is the uncertainty in the mean neutrino energy . in table
[ tab : experiments ] , we have listed all the experiments , which we have used in the minimization procedure .
most of the data used in our analysis are taken directly from the quoted publications .
( see table [ tab : experiments ] for the references to each experiment . )
we have chosen in this first analysis not to discriminate between the various experiments in any other way than by their experimental uncertainties .
for the chlorine ( cl ) and gallium ( ga ) experiments , we will define the mean neutrino energy as @xmath94 where @xmath95 .
the index @xmath96 labels the different solar neutrino sources . in the case of a cl experiment ( homestake )
, @xmath96 can be _ pep _ , @xmath97be , @xmath98b , _ hep _ , @xmath99o , or @xmath100f , and similarly for a ga experiment ( sage or gallex ) , @xmath96 can be _ pp _ , _ pep _ , @xmath97be , @xmath98b , _ hep _ , @xmath101n , @xmath99o , or @xmath100f .
the quantities @xmath102 and @xmath103 are the average neutrino energy and standard solar model ( ssm ) prediction for the neutrino capture rates of the solar neutrino source @xmath96 , respectively . using the values given by the bahcall - basu - pinsonneault 1998 ( bbp98 ) ssm @xcite , we obtain the mean neutrino energy to be @xmath104 for the sage and gallex experiments and @xmath105 for the homestake experiment . as for the water - cherenkov detector experiments ( super - kamiokande and kamiokande ) , we have chosen the mean energy of the neutrinos to be @xmath106 .
we will assume that the uncertainty in the energy is @xmath107 , respecting the threshold energy for detection .
for the super - kamiokande and kamiokande experiments , the energy resolution is taken to be @xmath108 mev . for solar neutrino experiments ,
the path length for the neutrinos is given by the mean sun - earth distance .
the length of the detector , the deviation in the earth s orbit around the sun , and the solar radius are all small in this connection , giving @xmath109 for solar neutrinos .
the path length of the atmospheric neutrinos is given by @xmath110 where @xmath111 is the so called zenith angle , @xmath112 is the radius of the earth , and @xmath113 is the altitude of the production point of atmospheric neutrinos .
we have taken @xmath114 and @xmath115 , which gives @xmath116 for the bin mean value @xmath117 ( macro and kamiokande ) and @xmath118 for bin mean value @xmath119 ( soudan , super - kamiokande , and imb ) .
the bins @xmath120 and @xmath121 have @xmath122 and @xmath123 , respectively
. the atmospheric neutrino experiments present a special problem , since they are not entirely comparable to each other .
the super - kamiokande experiment measures several parameters and has presented the `` ratio of ratios '' integrated over the energy ( actually @xmath6 ) as does also the kamiokande experiment .
other experiments can only measure the ratio of observed neutrino flux and expected neutrino flux .
the double ratio eliminates most of the uncertainty in the muon - neutrino flux .
however , this function is also more complicated to fit , and has lost much of the information of path length and energy of the neutrinos .
we have therefore used the directly measured disappearance rates evaluated from the data for upward - going events given in the tables and figures of refs .
this makes it possible to use several of the other atmospheric neutrino experiments on the same footing .
the disadvantage is the uncertainty in the expected neutrino flux .
this uncertainty is claimed to be of the order of @xmath124 @xcite .
the main uncertainty of the neutrino flux comes from the uncertainty in the primary cosmic radiation flux giving rise to the pions and kaons responsible for the muon and ( later ) electron neutrinos .
let the overall factor of flux uncertainty be denoted by @xmath125 for neutrino flavor @xmath126 .
thus , e.g. , the actual muon - neutrino flux is @xmath127 , where @xmath128 is the theoretical muon - neutrino flux calculated in ref . @xcite .
then , in the absence of neutrino oscillations , the ratio of measured and theoretical flux is just @xmath129 . in the presence of neutrino oscillations , on the other hand ,
we instead obtain @xmath130 for any flavor @xmath33 . here
@xmath79 is the probability of oscillation from flavor @xmath21 to flavor @xmath35 .
since the ratio @xmath131 is considered to be quite well determined theoretically , perhaps even to @xmath132 , due to known branching ratios of the pions @xcite and their decays , we conclude that the various factors @xmath125 are essentially the same for @xmath133 .
this is consistent with the picture , which says that the uncertainty is due to uncertainty in the primary cosmic ray flux .
we will further also neglect contributions from the @xmath134 leptons in what follows .
they should come from production of @xmath135 or @xmath136 mesons , but the relevant cross sections are small compared to the pion cross sections . for the muon - neutrino ratio measured by several experiments , we obtain @xmath137 at @xmath138 gev ( super - kamiokande ) ,
we have @xmath139 @xcite , and the last term can be neglected if @xmath140 is small , which turns out to be the case .
thus , we obtain @xmath141 .
we will therefore calculate the solution with the overall factor @xmath142 varying from @xmath143 to @xmath144 in the experimental values for @xmath145 of upward - going events with @xmath146 .
for the electron - neutrino ratio @xmath147 , the term @xmath148 is not negligable compared to the term @xmath149 , and can therefore not be neglected . instead of using @xmath150 as input
, we will analyze @xmath150 as an outcome of the optimization process , i.e. , in terms of the fitted parameters .
finally , the sensitivity for the atmospheric neutrinos can be divided into two groups .
the first group ( soudan and super - kamiokande ) is sensitive to the small mass squared difference only , whereas the second group ( macro , kamiokande , and imb ) is sensitive both to the large and ( slightly ) to the small mass squared differences .
changing the ratio from @xmath143 to @xmath144 times the quoted ones ( table [ tab : experiments ] ) , will therefore influence the determination of the mass squared differences .
the mean energies and path lengths for the accelerator neutrino experiments are all given in refs .
@xcite . in most cases ,
a measure of @xmath92 is given by half the length of the detector , giving @xmath151 in the range @xmath152@xmath153 .
for the transition probabilities , we have used the presented @xmath154 c.l .
upper bounds , except for the liquid scintillator neutrino detector ( lsnd ) experiment , for which we have used their quoted @xmath155 transition probability .
the reactor neutrino experiments are all disappearance experiments with no signal detected .
we have used the published @xmath154 c.l .
bounds on disappearance signals mainly to limit the mass squared ranges .
once these are outside the sensitivities of the reactor experiments , the solution will automatically reproduce these .
for the reactor experiments , the uncertainties in @xmath7 are all small compared to the uncertainties in energy determination .
to be able to minimize eq .
( [ eq : obj ] ) we have to use numerical methods , since there is no successful way to do this with analytical methods , the present function being to large and we have far too many variables .
our procedure will be as follows .
first , we will use a very simple stochastic procedure .
we choose some number ( @xmath85 ) of random points in the specified domain and evaluate the object function ( [ eq : obj ] ) .
we simply take the point which gives the smallest value of this function . to obtain good statistics
we repeat this procedure @xmath156 times . including all 16 experiments , discussed in the previous section , and choosing @xmath157 and @xmath158 , we obtained @xmath159 where we have used @xmath160 the quantities @xmath161 and @xmath162 are the sample mean value and sample standard deviation , respectively .
the point which generated the smallest value of the object function ( [ eq : obj ] ) in our simulation was @xmath163 having these @xmath164 data points , we can now use them as initial points in a deterministic minimization procedure .
we will use a sequential quadratic programming ( sqp ) method .
the results obtained are @xmath165 for the mean value and @xmath166 for the best point .
finally , we have evaluated the theoretically expected probabilities for the different experiments , using the best point solution , to demonstrate that they are indeed consistent with experimental data .
this is shown in table [ tab : probabilities ] . upon varying @xmath142 ,
we have made an analysis to see the effect on the different parameters in the model .
the parameters then vary as is shown in fig .
[ fig : eta_depend ] .
the stability of @xmath167 is to be noted .
the mixing angle @xmath168 varies only slowly around @xmath169 .
the mixing angle @xmath170 varies from @xmath171 up to @xmath172 in the interval of @xmath142 investigated .
just as expected , the mass squared differences change slowly to compensate for this change in the atmospheric values . as mentioned earlier , the reason for this is that the two first of the atmospheric experiments are sensitive to the small mass squared difference and the three others to the large mass squared difference .
from the results in table [ tab : probabilities ] , we see that the parameter set given as best point is indeed consistent with most of the experiments .
it is clear from the beginning that the solar neutrino experiments can not all be reproduced with the parameter range used here , as was mentioned in the introduction .
the more detailed super - kamiokande and kamiokande results , that include energy resolution of the solar neutrinos , are very well fitted by this solution .
the gallex and homestake experiments are not so well fitted . in the case of gallex ,
the fit is within two standard deviations , whereas the homestake result is several standard deviations off .
the accelerator neutrino experiments are all well accounted for .
it remains to be seen if further sampling of statistics in the lsnd experiment will confirm its present value for @xmath173 .
finally , both reactor experiments are very well reproduced by the best point solution .
the solution to the experimental constraints all give @xmath174 and @xmath175 .
the mixing angle @xmath170 varies from @xmath171 to @xmath172 as the flux factor @xmath142 varies from @xmath143 to @xmath144 times its nominal value of 1 .
since the ckm matrix can be written as @xmath176 , where @xmath177 is a rotation by an angle @xmath178 in the @xmath179 plane , it is clear that @xmath180 corresponds to maximal mixing between the two lightest neutrinos , @xmath181 and @xmath182 .
there are two particularly simple `` solutions '' more or less in the ranges of the mixing angles : ( a ) bimaximal mixing @xcite with @xmath183 , @xmath184 , and @xmath167 and ( b ) single maximal mixing with @xmath185 , @xmath184 , and @xmath186 .
these two `` solutions '' both reproduce data quite well ( apart from the lsnd experiment ) the atmospheric ones with the appropriate factor @xmath142 included .
the ckm matrices for the simple `` solutions '' are given by @xmath187 and @xmath188 to be compared with the ckm matrix for the best point solution @xmath189 note that the matrix element @xmath190 is small compared to all the other matrix elements , in agreement with earlier suggestions @xcite .
for the simple `` solutions , '' it becomes identically equal to zero . for the double ratio ,
we obtain approximately @xmath191 in our case , the ckm matrix is real so @xmath192 . using the published @xcite values for the theoretical @xmath193 flux ratio for the upward - going muon - neutrinos @xmath194 , we have calculated the ratio of ratios for the solution obtained with @xmath195 .
this ratio also eliminates the uncertainty in the muon - neutrino flux from the atmosphere .
the result is @xmath196 , in good agreement with the value given by the super - kamiokande collaboration of @xmath197 for multi - gev data .
this indicates that the electron - neutrino ratio @xmath150 is also in agreement with data .
see fig .
[ fig : r_l ] for the @xmath6 dependence of @xmath198 .
our solution is also consistent with the @xmath199 asymmetry measured by the super - kamiokande collaboration @xcite , as well as with the corresponding @xmath200 asymmetry . with our approximations ,
the @xmath199 asymmetry is given by @xmath201 where @xmath202 is the number of @xmath203-like upward - going events and @xmath204 is the number of @xmath203-like downward - going events .
the measured value for multi - gev events is @xmath205 @xcite , which is consistent with the value @xmath206 obtained from our analysis when @xmath207 . in the present analysis ,
we have deliberately considered mass ranges for the solution that avoid the region where the solar ( and atmospheric ) neutrinos could be affected by the msw effect .
we plan to address this question specifically in a future analysis . the mass range for the solar neutrino problem will then interfere with the analysis of atmospheric neutrinos .
this work was supported by the swedish natural science research council ( nfr ) , contract no .
f - aa / fu03281 - 312 . support for this work was also provided by the engineer ernst johnson foundation ( t.o . ) .
t.o . would like to thank samoil m. bilenky , manfred lindner , and martin freund for stimulating discussions on neutrino physics .
super - kamiokande and kamiokande collaborations , t. kajita , in _ proceedings of the xviiith international conference on neutrino physics and astrophysics ( neutrino 98 ) _ , takayama , japan , 1998 [ nucl . phys .
b ( proc . suppl . ) * 77 * , 123 ( 1999 ) ] . j.n . bahcall , s. basu , and m.h .
pinsonneault , phys .
b * 433 * , 1 ( 1998 ) .
bilenky , c. giunti , and w. grimus , report no .
uwthph-1998 - 61 , dftt 69/98 , kias - p98045 , sfb 375 - 310 , tum - hep 340/98 , hep - ph/9812360 , prog .
( to be published ) .
m. honda , t. kajita , k. kasahara , and s. midorikawa , phys .
d * 52 * , 4985 ( 1995 ) ; v. agrawal , t.k .
gaisser , p. lipari , and t. stanev , _ ibid . _ * 53 * , 1314 ( 1996 ) ; t.k .
_ , _ ibid .
_ * 54 * , 5578 ( 1996 ) ; t.k . gaisser and t. stanev , _ ibid . _ * 57 * , 1977 ( 1998 ) . lsnd collaboration , c. athanassopoulos _ et al .
lett . * 77 * , 3082 ( 1995 ) ; lsnd collaboration , c. athanassopoulos _ et al .
_ , _ ibid .
_ * 81 * , 1774 ( 1998 ) ; lsnd www page url http://www.neutrino.lanl.gov/lsnd/ karmen collaboration , k. eitel and b. zeitnitz , in proceedings of the xviiith international conference on neutrino physics and astrophysics ( neutrino 98 ) @xcite [ nucl .
b ( proc . suppl . ) * 77 * , 212 ( 1999 ) ] ; karmen www page url http://www-ik1.fzk.de/www/karmen/karmen_e.html v. barger , s. pakvasa , t.j .
weiler , and k. whisnant , phys .
b * 437 * , 107 ( 1998 ) ; a.j .
baltz , a.s .
goldhaber , and m. goldhaber , phys .
lett . * 81 * , 5730 ( 1998 ) ; m. tanimoto , phys .
d * 59 * , 017304 ( 1999 ) ; y. nomura and t. yanagida , _ ibid . _ * 59 * , 017303 ( 1999 ) ; g. altarelli and f. feruglio , phys . lett .
b * 439 * , 112 ( 1998 ) ; e. ma , _ ibid .
_ * 442 * , 238 ( 1998 ) ; n. haba , phys . rev .
d * 59 * , 035011 ( 1999 ) ; h. fritzsch and z.z .
xing , phys . lett .
b * 440 * , 313 ( 1998 ) ; h. georgi and s.l .
glashow , report no .
hutp-98/a060 , hep - ph/9808293 ; s. davidson and s.f .
king , phys .
b * 445 * , 191 ( 1998 ) ; r.n .
mohapatra and s. nussinov , _ ibid .
_ * 441 * , 299 ( 1998 ) ; r.n .
mohapatra and s. nussinov , phys .
d * 60 * , 013002 ( 1999 ) ; g. altarelli and f. feruglio , j. high energy phys . * 11 * , 021 ( 1998 ) ; c. giunti , phys .
d * 59 * , 077301 ( 1999 ) ; s.k .
kang and c.s .
kim , _ ibid .
_ * 59 * , 091302 ( 1999 ) ; c. jarlskog , m. matsuda , s. skadhauge , and m. tanimoto , phys .
b * 449 * , 240 ( 1999 ) ; y .- l .
wu , report no . as - itp-99 - 01 , hep - ph/9901245 , eur .
j. c ( to be published ) .
gallex collaboration , p. anselmann _
et al . _ ,
b * 342 * , 440 ( 1995 ) ; gallex collaboration , t. kirsten , in proceedings of the xviiith international conference on neutrino physics and astrophysics ( neutrino 98 ) @xcite [ nucl .
b ( proc . suppl . )
* 77 * , 26 ( 1999 ) ] .
super - kamiokande collaboration , y. fukuda _ et al .
* 81 * , 1158 ( 1998 ) ; super - kamiokande collaboration , y. fukuda _ et al .
_ , _ ibid .
_ * 81 * , 4279 ( 1998 ) ; super - kamiokande www page url http://www-sk.icrr.u-tokyo.ac.jp/doc/sk/ kamiokande collaboration , y. fukuda _ et al .
lett . * 77 * , 1683 ( 1996 ) ; kamiokande
www page url http://www-sk.icrr.u-tokyo.ac.jp/doc/kam/ soudan collaboration , w.w.m .
b * 391 * , 491 ( 1997 ) ; soudan www page url http://hepwww.rl.ac.uk/soudan2/ super - kamiokande collaboration , y. fukuda _ et al .
b * 433 * , 9 ( 1998 ) ; super - kamiokande collaboration , y. fukuda _ et al .
_ , _ ibid .
_ * 436 * , 33 ( 1998 ) ; super - kamiokande www page url http://www-sk.icrr.u-tokyo.ac.jp/doc/sk/ super - kamiokande collaboration , y. fukuda _ et al . _ ,
lett . * 81 * , 1562 ( 1998 ) .
macro collaboration , s. ahlen _ et al .
b * 357 * , 481 ( 1995 ) ; macro collaboration , m. ambrosio _
et al . _ , _ ibid . _ * 434 * , 451 ( 1998 ) ; macro collaboration , d. michael , talk given at the xxixth international conference on high - energy physics ( ichep 98 ) , vancouver , british columbia , canada , 1998 ( unpublished ) ; macro www page url http://www.lngs.infn.it/lngs/htexts/macro.htmlx imb collaboration , d. casper _
et al . _ ,
66 * , 2561 ( 1991 ) ; imb collaboration , r. becker - szendy _ et al .
_ , _ ibid .
_ * 69 * , 1010 ( 1992 ) ; imb collaboration , r. becker - szendy _ et al .
d * 46 * , 3720 ( 1992 ) .
chooz collaboration , m. apollonio _
et al . _ ,
b * 420 * , 397 ( 1998 ) ; chooz collaboration , c. bemporad , in proceedings of the xviiith international conference on neutrino physics and astrophysics ( neutrino 98 ) @xcite [ nucl .
b ( proc . suppl . )
* 77 * 159 ( 1999 ) ] ; chooz www page url http://duphy4.physics.drexel.edu/chooz_pub/index.htmlx | global fits to all data of candidates for neutrino oscillations are presented in the framework of a three - flavor model .
the analysis excludes mass regions where the msw effect is important for the solar neutrino problem .
the best fit gives @xmath0 , @xmath1 , @xmath2 , @xmath3 , and @xmath4 indicating essentially maximal mixing between the two lightest neutrino mass eigenstates . |
this paper reports on some recent progress that has been made in the analytical modeling of defect formation , far from threshold , in pattern forming physical systems .
we will take a moment here to very briefly sketch the physical and mathematical background that motivates what is done in this paper . the relevant class of pattern - forming physical systems to consider are those in which the spatial physical field can be described as planar and the first bifurcation from a homogeneous state , having arbitrary translational symmetry in the plane , produces a striped pattern which has only a discrete periodic symmetry in one direction .
this _ symmetry - breaking _ occurs at a critical threshold ; above this threshold the pattern can deform and , further away , _ defects _ can form .
it is the desire to understand and model this process of defect formation that motivates our study .
a good particular example of these kinds of physical systems is a high prandtl number rayleigh - bnard convection experiment .
the critical threshold in this example is the critical rayleigh number at which fluid convection is initiated from the sub - threshold homogeneous conducting state .
the `` striped pattern '' here can be taken to be the horizontal cross - section of the temperature field at the vertical midpoint of the experimental cell in which _
convection rolls _ have formed .
because of its periodic structure , the striped pattern can be described in terms of a periodic form function of a _ phase _ , @xmath0 , where the magnitude of @xmath1 is the wavenumber of the pattern and the orientation of @xmath1 is perpendicular to the stripes . here
@xmath2 is a physical point in the plane .
even though the striped pattern will deform far from threshold , over most of the field ( and in particular away from defects ) it can be locally approximated as a function of a well - defined phase , @xmath3 , for which a local wavevector can be defined as @xmath4 which differs little from a constant vector unless one varies over distances on the order of many stripes in the pattern .
this slowly - varying feature of pattern formation far from threshold motivates the introduction of a _ modulational _ ansatz in the microscopic equations describing these physical systems from which an order parameter equation for the behavior of the phase can be formally derived .
this was originally done by cross and newell @xcite .
these equations are variational and from our perspective it is advantageous to study their solutions by studying the behavior of the minimizers of the variational problem . the version of the variational problem that we study corresponds to the following energy functional on a given domain @xmath5 with specified dirichlet boundary values .
@xmath6 which is expressed in terms of _ slow _ variables stemming from the modulational ansatz mentioned above : @xmath7 .
we refer to this functional as the _ regularized cross - newell _ ( rcn ) _
energy_. it consists of two parts : a non - convex functional of the gradient ( the cn part ) plus a quadratic functional of the hessian matrix of @xmath8 , which is the regularizing singular perturbation . without this regularization ,
the cn variational equations admit non - physical caustic formation . instead , by studying the limit of minimizers of @xmath9 as @xmath10 , one may be able to identify the formation of a physical defect as a limiting jump discontinuity or other kind of singularity in the wavevector field associated to the @xmath11-indexed family of minimizing phase fields . for more details on
what has been rather tersely outlined above , we refer the reader to @xcite where analytical results on the asymptotic limit of minimizers for rcn and their defects in certain geometries are also derived .
see also @xcite where further refinements and generalizations are developed .
we further mention that the variational problem associated to ( [ eq : gl ] ) also arises in other physical contexts ( unrelated to pattern formation ) where it is known as the _ aviles - giga energy _ @xcite .
we now turn to the focus of this paper .
the kind of defects that are seen to arise far from threshold are not supported by asymptotic minimizers of ( [ eq : gl ] ) if the class of functions over which one is varying is restricted to be single - valued phases . in particular
, one can see for purely topological reasons that this restriction rules out _ disclinations _ @xcite . in @xcite , physical , numerical and experimental arguments
are developed which make a strong case in support of the hypothesis that the correct order parameter model for the phase in pattern forming systems far from threshold should come from a variational problem admitting test functions which are _ multi - valued _ and in particular _ two - valued_. in physical parlance this is often expressed by saying that the wavefield @xmath1 should be allowed to be a _ director field _ ; i.e. an unoriented vector field .
one figure ( see fig . [
fig : sh - zip ] below ) from @xcite will help to crystallize the issue and the focus of this paper .
this figure shows seven numerical simulations , each done in a horizontal strip , of a solution to the _ swift - hohenberg equation _ which is a generic model of microscopic equations for a pattern forming system .
each of these is run far from theshold but with differing boundary conditions imposed at the edges . in each case
the boundary conditions impose a constant orientation of the stripe at the edges such that the normal to the stripe is @xmath12 along the top edge and @xmath13 along the bottom edge .
the only thing that changes from one simulation to the next is the value of @xmath14 which in the figure is recorded on the left in each respective cell .
the results of @xcite together with symmetry considerations establish that for an analogous domain and boundary values , the asymptotic minimizers of ( [ eq : gl ] ) , within the class of single - valued phases , should have the form shown in the bottom - most cell of figure ( [ fig : sh - zip ] ) .
that is , they should have wavevectors very close to @xmath12 in the upper region of the cell and very close to @xmath15 in the lower region of the cell with a boundary layer around the mid - line in which the wavevector transitions smoothly but rapidly from one state to the other .
these minimizers are dubbed _ knee solutions _ in @xcite and in the limit as @xmath10 , they tend to a configuration in which there is a sharp jump in the wavevector along the mid - line .
this kind of defect is called a _
grain boundary_. in other words , the theory for ( [ eq : gl ] ) with single - valued phases predicts that the grain boundary should be the limiting defect independent of the value of @xmath14 .
the different result appearing in figure ( [ fig : sh - zip ] ) was one of the pieces of evidence sited in @xcite to argue the necessity for the larger variational class of multi - valued phases , even in such simple geometries as those of figure ( [ fig : sh - zip ] ) . in this paper
we are going to carry out a careful analytical study of the rcn variational problem in exactly this geometry but within a larger class of two - valued phases .
we will firmly establish that the form of the asymptotic minimizers in this more general model does in fact depend non - trivially on @xmath14 .
in addition , the construction of test functions in section [ sec : u_bound ] and the numerical simulations in section [ sec : results ] gives some intuitive and experimental support to the belief that the stable solutions of the rcn equations qualitatively resemble what is seen in the swift - hohenberg simulations . in @xcite ,
the term swift - hohenberg `` zippers '' was coined to refer to the problem studied in figure ( [ fig : sh - zip ] ) . in this paper
we will be studying _
cross - newell zippers_.
we are given an angle @xmath14 that determines the boundary conditions on the pattern as @xmath16 by @xmath17 note that this differs from the setup underlying the swift - hohenberg zippers in that the boundary conditions are placed at @xmath18 in the @xmath19-direction rather than at finite values of @xmath19 .
this simplifies our technical considerations in that we do nt need to worry about adjusting the location of these boundaries as @xmath14 changes .
also , all of the patterns we want to consider here are _
shift - periodic _ in the @xmath20-direction .
this allows us to reduce our study to domains that are periodic in @xmath20 .
we introduce the ( small ) parameter @xmath21 and we define the period @xmath22 .
we consider the following variational problem on the strip @xmath23 : minimize @xmath24 $ ] given by @xmath25 = \iint_{{\mathcal}{s}^\epsilon } \left\ { [ \delta \theta]^2 + ( 1 - |\nabla \theta|^2)^2 \right\ } dx dy\ ] ] over all @xmath26 , \delta \in \mathbb{r}$ ] and @xmath27 satisfying the boundary conditions @xmath28 \in h^2({\mathcal}{s}^\epsilon ) .
\nonumber\end{gathered}\ ] ] we take a moment here to explain the considerations that have motivated the mixed dirichlet - neumann boundary conditions here , the first two boundary conditions in ( [ eq : bc ] ) above .
we argued in the introduction that in order to capture the physically relevant minimizers , the rcn variational problem needed to allow for multi - valued phases in its admissible class of test functions .
however , the numerical results on the swift - hohenberg zippers suggest that in certain symmetrical geometries the appropriate multi - valuedness can be introduced in a tractable fashion . indeed in the case of the sh zippers we see that the symmetry of the boundary conditions between the upper and lower edges of the domain is preserved in the symmetry of all of the exhibited solutions about the middle horizontal axis ; i.e. , the reflection in @xmath19 about the @xmath29 axis .
this suggests that a single - valued phase could describe the solution in the upper half - plane with the solution in the lower half - plane given as a symmetric reflection of that in the upper half - plane about @xmath29 .
figure [ fig : bcs ] illustrates two instances of the form that we expect these zippers to take in the infinite ( in @xmath19 ) geometry .
the figure on the left illustrates level curves ( _ stripes _ in the parlance of the introduction ) of what we will shortly define to be a _ self - dual knee solution_. this is indeed symmetric about the mid - axis , which we will take to be the @xmath29 axis ; moreover , one can see that its gradient field along @xmath29 is tangential to this axis .
thus the gradient field in the upper half - plane is completely symmetrical to that in the lower half - plane under reflection about @xmath29 .
however , for the striped pattern on the right in figure [ fig : bcs ] , this is not the case .
there are regions , illustrated for example by the darkened interval along @xmath29 , where the gradient field is tangential to this axis ; but , there are other regions , illustrated for example by the lightened interval along @xmath29 , where the gradient field needs to be perpendicular to this axis . by reflection symmetry this field will point upwards in the upper half - plane and downward in the lower half - plane
. this can not be supported by a vector field but it is allowable for a director field .
this indicates that in this region a two - valued phase is required . to get at the conditions on the phase itself we observe that patterns of the type illustrated here are analytically given in terms of a form function @xmath30 of the phase @xmath31 such that @xmath30 is locally periodic of period @xmath32 in @xmath27 and such that @xmath33 is even in @xmath19 and smooth in @xmath34 . in order to allow
@xmath27 to be two - valued we also require @xmath30 to be even in @xmath27 .
( an example of a global form function having these properties is @xmath35 . )
it follows from these requirements that either @xmath36 is even in @xmath19 , in which case @xmath37 , a neumann boundary condition ; or , @xmath36 is an odd function of @xmath19 modulo @xmath38 , in which case @xmath39 for some integer @xmath38 , a dirichlet boundary condition .
thus to realize the pattern on the right in figure [ fig : bcs ] in terms of a single - valued phase in the upper half - plane , we would need to take the neumann boundary condition on the darkened interval and the dirichlet boundary condition on the lightened interval .
this is what we have done in ( [ eq : bc ] ) . for the self - dual knee pattern on the left we would take the entire boundary condition to be neumann .
the functional @xmath40 is the rcn energy functional but with the scaling @xmath11 removed .
it is appropriate to do this because the demonstration that the nature of the rcn minimizers depends on @xmath14 is independent of this scaling .
the first and the second conditions impose a mixed dirichlet / neumann boundary condition at @xmath41 , the third condition imposes ( shifted- ) periodicity in @xmath20 and the last condition ensures that the test functions @xmath27 approach the straight parallel roll patterns @xmath42 as @xmath43 .
note that the dependence of the functional @xmath40 on the parameter @xmath44 is through the dependence of the domain @xmath45 and the boundary conditions on @xmath44 .
the parameters @xmath46 and @xmath47 are determined by minimization .
the parameter @xmath46 is a measure of the fraction of the boundary at @xmath48 that has a dirichlet boundary condition , and @xmath47 represents the _ asymptotic phase shift _ , that is the difference in phases between the test function @xmath27 and the roll pattern @xmath49 which satisfies @xmath50 .
the case where @xmath46 is set to zero is considered in earlier references @xcite .
the test functions @xmath36 satisfy a pure neumann boundary condition at @xmath41 and the minimizers in this case are the self - dual knee solutions @xmath51 these solutions have an asymptotic phase shift of @xmath52 and the energy of the minimizers in the strip @xmath45 is given by @xmath53 = \frac { 4 \pi \sqrt{1-\epsilon^2}}{3 \epsilon}. \label{eq : chevrons}\ ] ] the existence of @xmath54 minimizing @xmath40 can be shown from the direct method in the calculus of variations .
we also prove the following results about the minimizers , and their energy _ upper bound _ there is a constant @xmath55 such that @xmath56 \leq e_0 $ ] for all @xmath57 $ ] .
[ thm : u_bound ] we prove this result in sec . [
sec : u_bound ] by exhibiting an explicit test function satisfying this bound .
note the implication that the minimizers for sufficiently small @xmath44 can not be the self - dual solutions , since the energy in eq .
( [ eq : chevrons ] ) diverges as @xmath58 .
consequently , @xmath59 for sufficiently small @xmath44 .
_ lower bound _ there are constants @xmath60 and @xmath61 such that , even for the optimal test function @xmath62 and the optimal parameter values @xmath63 and @xmath64 , we have @xmath65 \geq e_1 $ ] for all @xmath66 .
further , there are constants @xmath67 such that @xmath68 for sufficiently small @xmath44 .
[ thm : l_bound ] we prove this result in sec . [
sec : l_bound ] . combining this result with the preceding theorem
, we obtain a rigorous scaling law for the energy of the minimizer , and for the quantity @xmath69 as @xmath58 . as a corollary to theorem [ thm : l_bound ]
, we find that an @xmath70 part of the energy of the minimizer concentrates on the set , @xmath71 .
this can be interpreted as saying that a nontrivial part of the energy of the minimizer lives in the region of the convex - concave disclination pair @xcite .
we will first show an upper bound for the enrgy functional @xmath72 , uniform in @xmath44 , by constructing a family of explicit test function whose energy is uniformly bounded .
the idea for the construction of these test functions comes from the self - dual ansatz @xcite which requires that the energy density of the functional @xmath73 should be _ equi - partitioned _ between its two terms .
functions satisfying this ansatz solve the self - dual ( resp .
, anti - self - dual ) equation : @xmath74 solutions of this equation can be constructed via the logarithmic transform @xmath75 which reduces ( [ eq : selfdual ] ) to the linear helmholtz equation ( [ selfdual ] ) .
we refer the reader to @xcite for more background on self - dual reduction .
we consider the helmholtz equation in the upper half - plane , @xmath76 subject to the mixed boundary conditions @xmath77 and with asymptotic behavior for large @xmath19 given by const .
@xmath78 where @xmath79 and @xmath80 .
we seek a shift - periodic solution , meaning that we change variables to @xmath81 and look for periodic solutions of @xmath82 with boundary conditions of periodicity in @xmath20 of period @xmath83 ; mixed boundary conditions at @xmath29 , @xmath84 and with asymptotic behavior for large @xmath19 given by const .
given such a @xmath86 , @xmath87 would satisfy the boundary conditions ( [ eq : bc ] ) .
( however , for notational simplicity , in the remainder of this section we will set @xmath88 . )
we now let @xmath45 denote the half - cylindrical domain , @xmath83-periodic in @xmath20 and with @xmath89 .
the existence of a weak solution to ( [ shiftper ] ) satisfying the above boundary conditions can be established via the lax - milgram theorem with appropriate energy estimates .
however , in order to derive uniform asymptotic energy estimates ( as @xmath90 ) for the cn zipper problem we need to go beyond existence results and try to construct a more explicit representation of the solution to ( [ selfdual ] ) .
unfortunately , at present , the solutions one can construct using greens function methods and the like do not yield sufficient a priori boundary regularity near @xmath29 to control the asymptotic behavior of the energy in this _ finite _ part of @xmath45 . we will therefore instead study solutions of a self - dual problem with modified boundary conditions ( more precisely , with pure dirichlet boundary conditions ) .
subsequently we will make a local modification of these solutions near the boundary to produce functions ( no longer global self - dual solutions ) whose asymptotic energy we can control _ and _ which are valid test functions for the cross - newell zipper problem .
the modified boundary value problem we consider is ( [ shiftper ] ) with ( [ 5]-[6 ] ) replaced by the pure dirichlet boundary condition @xmath91 where @xmath92 is a function which smoothly interpolates , up through second derivatives , between @xmath93 at @xmath94 on the left and @xmath95 at @xmath96 on the right .
there are clearly many choices for such a function ; the precise choice for our purposes will be made later at the end of subsection [ energy ] . by elliptic regularity @xcite , the solution to this boundary value problem satisfies @xmath97 . in the following sections we will construct the solutions to this problem and study its asymptotics relative to the rcn energy @xmath98 .
the whole plane green s function for the helmholtz equation ( [ selfdual ] ) is explicitly given in terms of the bessel potential @xcite : @xmath99 in terms of this green s function we can then represent a solution to ( [ selfdual ] ) , with asymptotic behavior for large @xmath19 given by const .
@xmath100 , as @xmath101 note that @xmath102 solves ( [ selfdual ] ) with respect to the standard dirichlet boundary condition which equals minus the jump of @xmath103 along the @xmath20-axis .
one may check directly ( see ( [ ft ] ) ) that in fact @xmath104 almost everywhere . integrating ( [ dirbase ] ) with respect to @xmath19
gives @xmath105 since both @xmath106 and the rhs of ( [ intdirbase ] ) decay as @xmath107 , it follows that @xmath108 .
this is consistent with the ansatz ( [ soln ] ) , taking @xmath109 to be the jump in the normal derivative of @xmath110 along @xmath29 .
we make the following shift - periodic ansatz for @xmath111 , @xmath112 with this one can expand out ( [ soln ] ) more explicitly as @xmath113 @xmath114 in ( [ 11 ] ) we have interchanged the order of integration which is justified by tonelli s theorem ; in ( [ 12 ] ) we ve made the substitution @xmath115 and in ( [ 13 ] ) we ve commuted the sum past the integrals which is justified by monotone convergence all terms in the series are positive and hence the partial sums are monotonic .
in ( [ 14 ] ) we write each summand as a single exponential and then appropriately complete the square in each exponent . finally in ( [ 15 ] ) we apply jacobi s identity @xcite .
here @xmath116 is one of the jacobi theta functions , in this setting explicitly given as @xmath117 finally , from ( [ 15 ] ) we can express our candidate for the solution to ( [ shiftper ] ) , ( @xmath118 ) as @xmath119 where @xmath120 . from the previous sections we have that @xmath121 is periodic of period @xmath83 ; also @xmath122 is periodic in @xmath20 of period @xmath83 and @xmath123 along @xmath124 when @xmath29 .
moreover , taking the fourier transform of ( [ shiftpersoln ] ) one finds that the fourier coefficients , in @xmath20 , must satisfy @xmath125 for each value of @xmath19 .
taking the limit as @xmath126 on both sides of ( [ ft ] ) gives @xmath127 this is a determining conditions for @xmath121 .
we note that differentiating ( [ ft ] ) with respect to @xmath19 and setting @xmath128 demonstrates that @xmath129 , at least in the @xmath130 sense . since @xmath131 , it follows , by comparison , that @xmath132 .
given this we can now define @xmath133 which characterizes @xmath134 as an element of @xmath135 .
it follows from sobolev s lemma @xcite that @xmath134 can be taken to be continuous .
this last observation also justifies the existence of the fourier coefficients @xmath136 that were formally introduced in ( [ ft ] ) .
we now determine the large @xmath19 asymptotics of ( [ soln ] ) . by ( [ ft ] ) , @xmath137 has a fourier representation given by @xmath138 ( we note that for large @xmath19 this series converges uniformly to a smooth , in fact real - analytic , function of @xmath20 . )
moreover , @xmath139 is non - zero since by the maximum principle @xcite applied to the elliptic pde ( [ shiftper ] ) on the cylinder @xmath140 \times \left(-\infty , \infty\right)$ ] , the integrand , @xmath141 , is non - negative and in fact , by ( @xmath142 ) , non - vanishing on @xmath140 $ ] ( the definition of @xmath143 which we give later will insure that this is so ) .
we will now try to show that the regularized cross - newell energy of @xmath144 is uniformly bounded in @xmath44 .
this would establish a uniform ( in @xmath44 ) upper bound for the energy minimizers . recall that the energy is calculated by integrating the energy density over the domain @xmath45 . making
this estimate breaks naturally into the consideration of two regions : @xmath145 \times \{y\geq m_\epsilon\}$ ] and @xmath146 \times \{y < m_\epsilon\}$ ] where @xmath147 is to be determined .
we remark that the so - called `` knee solution '' of the self - dual equation provides an upper bound for the energy for values of @xmath44 bounded away from zero .
so we only need to be concerned with small values of @xmath44 .
since @xmath148 solves the helmholtz equation , it will suffice to bound the density @xmath149 ( since the integral of this density equals that of @xmath150 for self - dual solutions ) .
we begin by considering the domain for large @xmath19 . since @xmath151 we may reduce our considerations to studying the asymptotics of @xmath137 and its first derivatives .
it will be convenient to replace the convolution integral in ( [ shiftpersoln ] ) by the fourier series whose coefficients are the product of the fourier coefficients of @xmath134 and the @xmath116 series .
this results in the following alternative representation of @xmath137 : @xmath152 with the change of variables , @xmath153 this representation takes the form @xmath154 where @xmath155 .
the critical point of the exponent is @xmath156 and the expansion of the exponent in the @xmath157 term of the series near this critical point has the form @xmath158 an asymptotic expansion in large @xmath19 may be developed for the integral in each term of the series ( [ asympw ] ) by the method of laplace . by the uniform convergence of the series ( for large @xmath19 ) , the asymptotic expansion of the series is equivalent to the sum of the asymptotic expansions from each term .
we implement this strategy to find the leading order , large @xmath19 behavior , and next corrections , for @xmath137 , @xmath159 and @xmath160 : @xmath161 where in the @xmath162 term of each series , @xmath163 , respectively .
we can now apply laplace s method to each term and then observe that the dominant contributions for large @xmath19 come from the @xmath164 fourier modes .
retaining just these we derive the following asymptotic behavior for @xmath165 : @xmath166 based on these asymptotics we can now estimate the energy in the large @xmath19 domain .
@xmath167 from which it follows that @xmath168 and so the energy density @xmath169 > from this it follows that the `` large @xmath19 '' part of the total energy is bounded as @xmath170 thus , if we take @xmath171 , this part of the total energy will remain finite as @xmath90 .
we next turn to consideration of the energy density in the _ finite _ part of the domain . to facilitate this consideration
we will sometimes make the uniformizing change of variables @xmath172 and @xmath173 in the jacobi theta function ( [ jacobi ] ) : @xmath174 in what follows we will assume that @xmath46 is chosen to depend on @xmath44 in such a way that @xmath175 .
we will make use here of the following single - layer potential counterpart of the double - layer potential representation ( [ shiftpersoln ] ) , which in fact can be deduced directly from a change of variables in ( [ asympw ] ) : @xmath176 @xmath177 we study the asymptotic behavior of the convolution integral in ( [ shiftpersoln2 ] ) for @xmath178 and for _ times _ @xmath179 of order less than @xmath180 : @xmath181 where in the third line above , the form of the integrals follows from making the change of variables as in ( [ perjacobi ] ) . in the second integral
we smoothly extend @xmath182 to be zero on @xmath183 .
the final line follows for @xmath179 of order less than @xmath180 because in this regime the jacobi theta function inside the convolution behaves as a _ dirac comb _ as @xmath90 .
the second term has this asymptotic behavior because @xmath184 and the support of @xmath185 is complementary to this interval , so that this integral decays exponentially to zero with @xmath44 , as with a dirac sequence away form its support .
based on ( [ convoest ] ) we can estimate @xmath137 as @xmath186 + \mathcal{o}\left(e^{-1/\epsilon}\right)\\ & = & e^{\epsilon x } e^{-\sqrt{1-\epsilon^2}y}+ o(\epsilon).\end{aligned}\ ] ] the evaluation of the previous integral may be deduced from a basic bessel identity ( see @xcite 9.6.23 ) . in order to estimate @xmath187 , we also need to estimate the @xmath20 and @xmath19 derivatives of @xmath188 .
to this end we first consider the @xmath20-derivative of the internal convolution integral which equals @xmath189 integrating by parts , the above derivative may be rewritten as @xmath190 as for the second integral in the last line of ( [ convoest ] ) .
thus , @xmath191 for the @xmath19 logarithmic derivative we have @xmath192 the last equivalence follows from a bessel recurrence identity @xcite , formula 9.6.26 , together with formula 9.6.6 .
thus we finally have @xmath193 and hence @xmath194 . since the domain @xmath195 \times \left\{y < m_\epsilon\right\}$ ] has dimensions @xmath196 , the total energy in this region is also asymptotically finite .
it remains to estimate the energy in the region @xmath197
\times \left\{y < m_\epsilon\right\}$ ] which has dimensions @xmath198 .
the question of the finiteness of the energy of the @xmath62 we have been considering in this region is beside the point for general purpose of but , this self - dual solution does not satisfy the boundary condition ( [ 6 ] ) in this region .
as stated earlier we are going to modify the self - dual test function in this region so that the boundary condition ( [ 6 ] ) is satisfied . to that end
we fix a small value of @xmath47 and let @xmath199 denote the @xmath47-neighborhood of @xmath200 $ ] in @xmath201 .
we modify @xmath137 in this neighborhood as follows .
define @xmath202 where @xmath203 is a partition of unity subordinate to the cover of @xmath201 given by @xmath204 and @xmath205 , where @xmath141 here is defined as in @xmath206 .
one has @xmath207 and @xmath208 .
it is straightforward to check that @xmath209 satisfies the boundary conditions ( [ 5 ] ) and ( [ 6 ] ) : @xmath210 for @xmath211 $ ] . for @xmath212 $ ]
, @xmath213 thus , @xmath214 is an admissible test function for the regularized cross - newell variational problem .
we can now estimate the energy of this test function in @xmath199 . the energy density in this region
is bounded and therefore the energy in @xmath199 is finite .
it remains to estimate the energy in @xmath216 \times \left\{y <
m_\epsilon\right\}\right ) \backslash \mathbf{b}(\delta ) $ ] . to proceed with this
we will need a more specific definition of @xmath143 which we now give .
note first that by our assumption that @xmath217 , the interval @xmath218 $ ] remains of size @xmath219 for arbitrarily small values of @xmath44 .
we will now further pin this down by setting @xmath220 for a value of @xmath221 that is fixed , independent of @xmath44 . consequently , @xmath218 $ ] is now an interval of fixed length @xmath222 which can therefore also be represented as @xmath223 $ ] . recall that @xmath143 needs to be built so that on this interval it matches , through second order , to @xmath224 at the left endpoint and similarly to @xmath95 at the right endpoint . toward this end
we observe that the required leading order value on the right is @xmath225 , independent of @xmath44 while on the left the leading order value limits to the stable value of @xmath226 as @xmath90 . choosing a value @xmath227 that is small with respect to @xmath222
, we define a compressed _ tanh_-profile that interpolates between the point @xmath228 and the point @xmath229 and where @xmath230 is another chosen value required to be smaller than @xmath225 .
( this last requirement will insure that the positivity claim made at the end of subsection [ zipper ] holds . ) explicitly this tanh - profile is given by @xmath231 note that the profile of @xmath232 is independent of @xmath44 .
the only way in which @xmath233 depends on @xmath44 is that this profile translates uniformly with @xmath83 as @xmath44 changes .
we will define @xmath234 on the subinterval @xmath235 = [ \ell - c\pi + \nu,\ell - \nu]$ ] of @xmath223 $ ] .
next we will define the piece of @xmath92 on the left that interpolates between the point @xmath236 and the point@xmath237 . choose a value @xmath238 that is small with respect to @xmath239 .
consider the covering of @xmath240 $ ] by the two sets @xmath241 and @xmath242 $ ] and let @xmath243 be a partition of unity subordinate to this cover which means , in particular , that @xmath244 \\ \end{array}\right .
\\ \psi_2 & = & \left\{\begin{array}{cc } 1 & ( x_0 - \sigma , x_0 ] \\
0 & [ \ell - c\pi , \ell - c\pi + \sigma ) .
\\ \end{array } \right.\end{aligned}\ ] ] on @xmath245 $ ] we define @xmath246 it is straightforward to see that with these choices @xmath92 is smooth throughout @xmath247 and satisfies the smooth matching conditions on the left .
moreover , it is clear from the functions comprising ( [ qa ] ) that @xmath143 and its derivatives remain bounded on @xmath247 as @xmath90 .
a similar construction may be made on the right ; i.e. , on @xmath248 $ ] .
this completes our description of @xmath92 .
the study of the convolution integral in ( [ shiftpersoln2 ] ) in the region where @xmath212 $ ] now proceeds similarly to what was done in ( [ convoest ] ) and subsequent formulae . in particular
, the analogous result to ( [ convoest ] ) is that for @xmath249 and for _ times _ @xmath179 of order less than @xmath180 : @xmath250 with @xmath92 here bounded away from zero , independent of @xmath44 , by our earlier choice of @xmath251 . hence the denominators in the estimates analogous to ( [ xlogd ] ) and ( [ ylogd ] ) are under control . in the subsequent formulae
the roles of @xmath93 and @xmath92 are effectively interchanged as above and all proceeds as before .
the result is that the energy in @xmath216 \times \left\{y < m_\epsilon\right\}\right)\backslash \mathbf{b}(\delta)$ ] is asymptotically bounded like @xmath252 .
it thus follows that the total energy of our family of test functions is uniformly bounded in @xmath44 .
following the ideas of jin and kohn @xcite , we will prove _ ansatz - free _ lower bounds for the functional @xmath72 by identifying vector fields @xmath253 such that @xmath254 \geq c^{-1 } \left|\iint_{{\mathcal}{s}^\epsilon } \nabla \cdot \sigma(\nabla \theta ) dx dy \right|.\ ] ] this allows us to obtain information about the energy @xmath40 purely in terms of the boundary conditions on @xmath27 . to avoid the proliferation of symbols , here and henceforth
, @xmath255 , _ etc _ denote ( finite ) constants whose precise value is unimportant , and different occurrences of the same symbol might denote different values of the constants .
@xmath256 , _ etc _ denote constants that have the same value in all their occurrences .
a smooth vector function @xmath257 is _ subordinate to the energy _ if @xmath258 for some @xmath259 .
if @xmath260 is subordinate to the energy , it follows that @xmath261 where we use the identification @xmath262 and @xmath263 .
consequently , @xmath264 & = \iint_{{\mathcal}{s}^\epsilon } \left\{[\nabla \nabla \theta]^2 + ( 1 - |\nabla \theta|^2)^2\right\ } dx dy - 2 \int_{\partial { \mathcal}{s}^\epsilon } \theta_x d \theta_y \\ & \geq 2 \iint_{{\mathcal}{s}^\epsilon } |\nabla \nabla \theta| |1 - |\nabla \theta|^2| dx dy - 2 \int_{\partial { \mathcal}{s}^\epsilon } \theta_x d \theta_y \\ & \geq \frac{2}{c } \left|\iint_{{\mathcal}{s}^\epsilon } \nabla \cdot \sigma(\nabla \theta ) dx dy \right| - 2 \int_{\partial { \mathcal}{s}^\epsilon } \theta_x d \theta_y \\ & \geq c^{-1 } \left|\iint_{{\mathcal}{s}^\epsilon } \nabla \cdot \sigma(\nabla \theta ) dx dy \right|.\end{aligned}\ ] ] in obtaining the last equation , we use the fact that @xmath265 for the boundary conditions in ( [ eq : bc ] ) .
there are constants @xmath266 such that for all @xmath267$],@xmath26 $ ] and @xmath268 , we have @xmath25 \geq \frac{e_1 \epsilon^2}{(1-a)^2 } - k_1\epsilon^2.\ ] ] [ lem : squeeze ] let @xmath269 be a smooth , compactly supported function such that @xmath270 an explicit example of a function @xmath271 whith these properties is @xmath272 & p \in ( -1,2 ) \\ 0 & \text { otherwise } \end{cases}\ ] ] let @xmath273 .
define the vector field @xmath274 by @xmath275 d\eta.\end{aligned}\ ] ] since @xmath271 has compact support , it follows that @xmath260 is bounded on @xmath276 .
an explicit calculation shows that the quantities @xmath277 and @xmath278 are zero .
also , @xmath279 where @xmath280 is clearly finite since @xmath271 is compactly supported and twice differentiable .
this proves that @xmath260 is subordinate to the energy .
since @xmath281 we obtain @xmath282 ^ 2\ , dx dy \right ] \nonumber \\ & = \frac{cb^2}{2 } { \mathcal}{f}^\epsilon[\theta;a,\delta ] \label{eq : lbnd1}\end{aligned}\ ] ] integrating by parts , we have @xmath283 where the contributions from the boundaries at @xmath284 and @xmath285 cancel due to the periodicity . by construction
, @xmath286 and @xmath287 has a maximum value at @xmath288 .
consequently , @xmath289 also , @xmath290 and @xmath271 is lipschitz so that @xmath291 for some finite @xmath292 . combining these estimates with ( [ eq : lbnd1 ] )
, we obtain @xmath25 \geq \frac{2 \pi}{c b^2}\left[\phi(0 ) - \phi(1 ) - c'(1-a)^2 \right],\ ] ] and rewriting @xmath293 in terms of @xmath46 and @xmath44 yields the desired conclusion .
the above lemma shows that the energy grows without bound as the quantity @xmath69 becomes small .
however , we do not have _ a priori _ control on the size of @xmath69 . consequently , to obtain a lower bound for the energy , we need a complementary estimate which shows that the energy grows as the quantity @xmath69 becomes large . to prove this result
, we first construct a vector field @xmath260 subordinate to the energy functional as follows : let @xmath294 be a smooth , compactly supported function such that @xmath295 we can always construct such a function , given @xmath296 , a compactly supported function with @xmath297 .
observe that @xmath298 where @xmath299 .
consequently , by an appropriate choice of @xmath300 , we get @xmath301 with the required properties .
we define the functions @xmath302 and @xmath303 by @xmath304 note that the functions @xmath305 and @xmath306 are well defined for positive values of their arguments , that is the expressions on the right hand sides of the above equations are even functions of @xmath307 . from these expressions
, we have @xmath308 we will also use the same letters @xmath305 and @xmath306 to denote smooth extensions of the functions defined above to all of @xmath309 .
we will pick extensions such that the supports of @xmath305 and @xmath306 are contained in @xmath310 .
let @xmath311 be a compactly supported function and set @xmath312 \nonumber \\ & - \int_0^p ( p-\xi)\left\ { \sigma(1-\xi^2 ) \left [ \frac{\partial^2}{\partial \xi^2 } \varphi(\xi^2 ) \right ] - \zeta(1-\xi^2 ) \left [ \frac{\partial^2}{\partial \xi^2 } ( \xi^2 \varphi(\xi^2 ) ) \right ] \right\}\ , d\xi .
\label{eq : defnv}\end{aligned}\ ] ] @xmath313 is now an even function of @xmath314 and @xmath307 . define the vector field @xmath260 by @xmath315 from ( [ eq : def ] ) it follows that @xmath316 also , @xmath317 . finally , an explicit calculation yields @xmath318\left ( \sigma(q^2 ) - \sigma(1-p^2 ) \right ) - \left [ \frac{\partial^2}{\partial p^2 } p^2 \varphi(p^2 ) \right]\left ( \zeta(q^2 ) - \zeta(1-p^2 ) \right)\right| \nonumber \\ & \leq c_2 |1 - p^2 - q^2|\end{aligned}\ ] ] where @xmath319 can be bounded in terms of the support of @xmath320 and the maximum values of @xmath321 and @xmath322 . clearly @xmath320 and all it s derivatives
are uniformly bounded since it is smooth and compactly supported . from ( [ eq : zeta_sigma ] ) , we have @xmath323 since @xmath324 is compactly supported , these derivatives vanish as @xmath325 , implying that @xmath326 and @xmath327 are bounded for all positive values of the argument .
since @xmath306 and @xmath305 are smooth and are identically zero if their arguments are sufficiently negative , it follows that @xmath328 .
it thus follows that the vector field @xmath260 is subordinate to the energy functional . for future use , let us record a few observations that follow directly from the construction : @xmath286 since @xmath329 is an odd function of @xmath307 .
@xmath330 consequently , the non - degenerate critical points are at @xmath331 . differentiating in @xmath307 ,
we get @xmath332 so that @xmath333 has a maximum at @xmath334 and a minimum at @xmath335 .
finally , @xmath336 so that @xmath337 and @xmath338 as @xmath339
. @xmath340 will denote the maximum value of @xmath333 .
@xmath341 for a constant @xmath342 . in obtaining the last line
, we use @xmath343 for some bounded constants @xmath344 .
[ obs : estimate ] there are constants @xmath345 such that for all @xmath267 $ ] , @xmath26 $ ] and @xmath268 , we have @xmath25 \geq \frac{e_2(1-a)}{\epsilon } - k_2 \epsilon.\ ] ] [ lem : extend ] for the case @xmath346 , corresponding to the self - dual minimizers , this estimate captures the right scaling of the minimum energy as @xmath58 .
the proof of the lemma follows from estimating a lower bound for the functional @xmath24 $ ] using the vector field @xmath260 that we constructed above . for the vector field @xmath260 we have @xmath347 as before
, the contributions from the boundaries at @xmath284 and @xmath348 cancel due to the periodicity . by construction
, @xmath349 and @xmath333 has a maximum value @xmath350 at @xmath351 .
consequently , @xmath352 from observation [ obs : estimate ] , we obtain @xmath353 since @xmath260 is subordinate to the energy , @xmath25 \geq \frac{m \pi}{c \epsilon}\left[1 - a - k \epsilon^2 \right],\ ] ] which yields the desired conclusion .
we can now prove theorem [ thm : l_bound ] using lemma [ lem : squeeze ] and lemma [ lem : extend ] .
let @xmath293 denote the quantity @xmath354 . from lemma [ lem : squeeze ]
we get @xmath25 \geq \frac{e_1}{b^2 } - k_1 \epsilon^2 \geq 3 \left(e_1 e_2 ^ 2\right)^{1/3 } - 2 e_2 b - k_1 \epsilon^2\ ] ] where the last inequality comes from linearizing the convex function @xmath355 at @xmath356 . combining this estimate with the conclusion of lemma [ lem : extend ]
, we get @xmath25 \geq \max\left(3 \left(e_1 e_2 ^ 2\right)^{1/3 } - 2 e_2 b - k_1 \epsilon^2 , e_2 b - k_2 \epsilon \right ) \geq ( e_1 e_2 ^ 2)^{1/3 } - \frac{2 k_2 \epsilon + k_1 \epsilon^2 } { 3}.\ ] ] if we set @xmath357 for all @xmath358 , all @xmath26 $ ] and all @xmath27 satisfying the boundary conditions in ( [ eq : bc ] ) , we have @xmath254 \geq \frac{(e_1 e_2 ^ 2)^{1/3}}{3 } \equiv e_1.\ ] ] combining the upper bound @xmath359
\leq e_0 $ ] in theorem [ thm : u_bound ] with the lower bounds for @xmath40 in lemma [ lem : squeeze ] and lemma [ lem : extend ] , it follows that for @xmath360 we have @xmath361 consequently , @xmath362 for sufficiently small @xmath44 with @xmath363 and @xmath364 .
in this section , we will present the results of numerical simulations that illustrate and clarify our analysis of the energy and also the structure of the minimizers for the regularized cross - newell energy @xmath98 within the class of functions given by ( [ eq : bc ] ) . for our numerical simulations , we restrict ourself to the finite domain , @xmath365 , where @xmath366 is a length scale much larger than the typical wavelength of the pattern .
the boundary conditions in ( [ eq : bc ] ) which are appropriate for the semi - infinite strip @xmath45 are modified for the finite domain as follows
@xmath367 ; \nonumber \\
\theta(x , l ) = \left[\epsilon x + \sqrt{1 - \epsilon^2 } l + \delta\right ] \label{eq : num_bc}\end{gathered}\ ] ] it is rather straightforward to show that there exist @xmath368 satisfying ( [ eq : num_bc ] ) for an @xmath369 and @xmath370 minimizing the functional @xmath25 = \iint_{{\mathcal}{r}^\epsilon } \left\ { [ \delta \theta]^2 + ( 1 - |\nabla \theta|^2)^2 \right\ } dx dy.\ ] ] let @xmath371 , and @xmath268 be given .
@xmath372 is a sequence of functions that converges weakly to zero .
@xmath373 denotes the completion of periodic ( in @xmath20 ) functions on @xmath374 with respect to the @xmath375 norm .
if @xmath376 is a sequence satisfying ( in the sense of trace ) @xmath377 it follows that , up to extraction of a subsequence and relabelling , we have @xmath378 in @xmath379 . if @xmath383 is an infimizing sequence for @xmath24 $ ] subject to the boundary conditions in ( [ eq : num_bc ] ) , then let @xmath384 , where @xmath320 is a smooth function on @xmath374 satisfying the boundary conditions in ( [ eq : num_bc ] ) .
it then follows from the form of @xmath40 and the fact that @xmath383 is infimizing that @xmath385 is a bounded sequence in @xmath130 , and so converges weakly to a limit @xmath386 . applying the compactness result of the preceding lemma with reference to the sequence @xmath387
, we obtain the existence of a minimizier for the functional @xmath24 $ ] for a fixed @xmath46 and @xmath47 .
note that , for a given @xmath46 it is easy to construct smooth transformations @xmath388 such that @xmath389 is the identity , if @xmath27 satisfies the boundary conditions in ( [ eq : num_bc ] ) , then @xmath390 satisfies the same boundary conditions with the fraction of the boundary with a dirichlet boundary condition equaling @xmath391 .
further , the energy @xmath392 $ ] is a smooth function of @xmath179 for sufficiently small @xmath179 .
a standard argument now implies that , for a given @xmath47 the map @xmath393\ ] ] is continuous for @xmath80 .
a similar argument shows that the map is also continuous at @xmath346 . in lemma
[ lem : squeeze ] we showed that @xmath394 = \infty$ ] . combining these results , we see that @xmath395 = \inf_a \left[\inf_{\theta } { \mathcal}{f}^\epsilon[\theta;a,\delta]\right].\ ] ] we now consider variations @xmath396 , where @xmath397 is a smooth function vanishing identically on @xmath398 $ ] and equal to 1 on @xmath399 $ ] .
the functions @xmath400 satisfy the boundary conditions in ( [ eq : num_bc ] ) , except the asymptotic phase shift is given by @xmath401 .
a similar argument as above shows that the map @xmath402\ ] ] is continuous and it is easy to see that @xmath403 \rightarrow \infty$ ] as @xmath404 . in particular
, this proves the existence of an optimal @xmath47 , and combining with the results from above , we see that the minimizer @xmath405 , can be obtained by successive minimization in each of the factors . this suggests the following discretization for the functional @xmath40 , which should converge as the grid spacings @xmath406 .
we define a grid by @xmath407 , where @xmath408 .
the discretization of the test function @xmath36 is @xmath409 we define the difference operator @xmath410 by @xmath411 with similar definitions for @xmath412 . in terms of the discretization ,
the boundary conditions are @xmath413 and the energy functional is discretized as @xmath414 ^ 2 + \left [ \frac{(\delta_x^+ \theta)_{i , j}^2 + ( \delta_x^- \theta)_{i , j}^2 + ( \delta_y^+ \theta)_{i , j}^2 + ( \delta_y^- \theta)_{i , j}^2 } { 2 } - 1 \right]^2\ ] ] computing this functional requires assigning values for @xmath415 with @xmath416 and @xmath417 . the values for @xmath418 are obtained form the shift - periodicity of @xmath27 by @xmath419 .
the values for t @xmath417 are assigned using the dirichlet boundary condition @xmath420 .
this functional is minimized using matlab s conjugate - gradient minimization .
[ fig : hyster ] shows the results from minimizing the rcn energy over the pattern @xmath27 and also the phase shift @xmath47 , for different values of @xmath44 , and for a range of values of @xmath421 .
the results do indeed suggest that the ( partial ) minimization with respect to the pattern and the asymptotic phase yields a functional that _ depends continuously _ on @xmath46 .
further , this functional has first - order ( discontinuous ) phase transition at a bifurcation value @xmath422 , below which the global minimizer has @xmath423 .
[ fig : energy ] shows the energy of the minimizer ( minimizing over the pattern , asymptotic phase and the parameter @xmath46 ) as a function of @xmath44 .
note that the minimum energy is a non - differentiable function of @xmath44 , as one would expect for a first - order phase transition .
figure [ fig : patterns ] show the numerically obtained minimizing patterns at various values of @xmath44 .
note that , for sufficiently large @xmath44 , the minimizers are the knee - solutions ( [ eq : chevrons ] ) with @xmath346 , whereas for sufficiently small @xmath44 , the minimizers have convex - concave disclination pairs , and have @xmath423 . 7 m. abramowitz and i. stegun .
handbook of mathematical functions , u.s .
printing office , washington , d.c . , 1972 .
p. aviles , y. giga , on lower semicontinuity of a defect energy obtained by a singular limit of the ginzburg - landau type energy for gradient fields , proc .
edinburgh , * 129a * ( 1999 ) 1 - 17 .
cross , a.c .
newell , convection patterns in large aspect ratio systems , physica d * 10 * ( 1984 ) 299 - 328 .
ercolani , r. indik , a.c .
newell , t. passot , the geometry of the phase diffusion equation , j. nonlinear sci .
* 10 * ( 2000 ) 223 - 274 .
ercolani , r. indik , a.c .
newell , t. passot , global description of patterns far from onset : a case study , physica d * 184 * ( 2003 ) 127 - 140 .
ercolani , m. taylor , the dirichlet - to - neumann map , viscosity solutions to eikonal equations , and the self - dual equations of pattern formation , physica d * 196 * ( 2004 ) 205 - 223 .
l. c. evans .
partial differential equations , american mathematical society , providence , ri , 2002 .
w. jin , r. kohn , singular perturbation and the energy of folds , j. nonlinear sci . * 10 * ( 2000 ) 355 - 390 .
whittaker and g.n .
a course in modern analysis , cambridge university press , cambridge , england , 1990 . | we derive a rigorous scaling law for minimizers in a natural version of the regularized cross - newell model for pattern formation far from threshold .
these energy - minimizing solutions support defects having the same character as what is seen in experimental studies of the corresponding physical systems and in numerical simulations of the microscopic equations that describe these systems . |
it has long been known that source - free maxwell s equations for an electro - magnetic ( em ) field show an internal symmetry , known as the duality transformation . in a nutshell ,
maxwell s equations are invariant when the electric and the magnetic fields @xmath0 and @xmath1 are mixed by a rotation of an arbitrary angle @xmath2 , as @xmath3 this symmetry can be extended to maxwell s equations in the presence of sources , provided that additional magnetic charges and currents are introduced in the theory .
the duality symmetry has been proven to be associated with a set of conserved currents , as first discussed by lipkin @xcite ( see also refs .
@xcite @xcite @xcite @xcite ) . whenever a pseudoscalar axion - like field @xmath4 is introduced in the theory ,
the dual symmetry is spontaneously and explicitly broken .
axion - like fields and their interactions with the em field have been intensively studied @xcite @xcite @xcite @xcite @xcite @xcite @xcite , and they have recently received attention due to the possible role they might play in building topological insulators . however , the set of equations used in the axion electrodynamics literature does not preserve the duality relation in eq .
( [ rotation_em ] ) . this fact has not been previously noticed , and it might be due to the fact that the introduction of an axion - like interaction with the electromagnetic field only modifies two of the four maxwell s equations , namely gauss and ampere laws ; the remaining two equations ( faraday law and gauss law for @xmath1 ) are usually coupled to the new gauss and ampere laws without modifications from the original maxwell set . in this paper , we derive maxwell s equations in the presence of an axion - like field , with the requirement that the electric and magnetic fields satisfy the duality relation in eq .
( [ rotation_em ] ) . along with gauss and ampere laws for axion electrodynamics , we obtain new terms that also modify faraday law and gauss law for @xmath1 .
we show that , when the electric charges and currents are set to zero , maxwell s equations for axion electrodynamics allow for a solution in which coupled axion and electromagnetic waves propagate . in order to fix our notation , we first review classical electromagnetism in the presence of an electric charge distribution @xmath5 and an electric current density @xmath6 .
this paper is organized as follows . in sec .
[ review of axion electrodynamics ] , we review maxwell s formulation of classical em , the duality symmetry within classical em , and the inclusion of a pseudo - scalar axion - like field to the maxwell s lagrangian for electromagnetism . in sec . [ imposing the duality symmetry ]
, we obtain a new set of equations describing the em field configuration when both electric and magnetic sources , together with an axion - like field , are considered .
this set of equations is derived by imposing that the equations for the axion electrodynamics be invariant under the duality transformation in eq .
( [ rotation_em ] ) .
[ propagation of waves in the axion electrodynamics ] is devoted to expressing a solution in free space for the axion electrodynamics .
we obtain one explicit solution in which the axion field is described by a klein - gordon equation of motion , with the electric and magnetic fields propagating orthogonally and coupling to the axion field . in sec .
[ gauge transformations in the axion electrodynamics ] , we derive the expression for the gauge potentials in the theory , and we express the equations for the axion electrodynamics in terms of these potentials . finally , in sec .
( [ conservation of currents ] ) , we show that an internal symmetry between the axion field and the gauge potential exists , and we derive a set of conserved current related to this symmetry .
in the vacuum , maxwell s equations for an em field with electric field @xmath0 , magnetic field @xmath1 , and the sources @xmath5 , @xmath7 are @xmath8 where @xmath9 and @xmath10 are the electric polarizability and the magnetic permeability of the vacuum , respectively . the speed of light in the vacuum is @xmath11 .
thanks to the third and the fourth line in eq .
( [ maxwell_equations ] ) , it is possible to write the electric and the magnetic fields in terms of a scalar potential @xmath12 and a vector potential @xmath13 , as @xmath14 maxwell s eqs .
( [ maxwell_equations ] ) can be reformulated in an elegant form that is explicitly covariant .
firstly , we notice that the source @xmath15 and the potential @xmath16 transform as four - vectors under a lorentz transformation .
once the antisymmetric field tensor and its dual are being introduced as @xmath17 the electric and magnetic fields read @xmath18 with this notation , maxwell s equations are rewritten as @xmath19 the first line of maxwell s eqs .
( [ covariant_eq ] ) can be obtained from euler s equation of motion , @xmath20 using the em lagrangian @xmath21 . however , the second line in eq .
( [ covariant_eq ] ) does not directly derive from the lagrangian principle , being motivated empirically by the non - discovery of magnetic monopoles ( @xmath22 ) , and the absence of magnetic currents . when a magnetic charge density @xmath23 and a magnetic current @xmath24 are included in the theory , maxwell s eqs .
( [ maxwell_equations ] ) appear in the symmetric form @xmath25 similarly to what shown in eq .
( [ covariant_eq ] ) , this last set of equations can be expressed in a covariant form .
in fact , defining @xmath26 , we have @xmath27 taking the derivative of each line in eq .
( [ covariant_eq_monopole ] ) in @xmath28 yields to the conservation of the electric and magnetic currents , @xmath29 this result is obtained by noticing that , since @xmath30 is a symmetric tensor operator , we have @xmath31 . introducing the so(2 ) matrix @xmath32 the field strength and its dual in eq .
( [ covariant_eq_monopole ] ) are invariant under a rotation by an angle @xmath2 , @xmath33 when the sources are rotated by @xmath34 notice that the transformations in eqs .
( [ rotation_em ] ) and ( [ rotation_fields ] ) are equivalent . since magnetic monopoles and currents are not observed in nature , we choose the angle @xmath35 for which , given a certain set of currents @xmath36 and @xmath37 , makes @xmath38 disappear in the rotated configuration .
using eq .
( [ rotation_sources ] ) , this angle is defined by @xmath39 to sum up , although the theory is invariant under the transformations in eqs . ( [ rotation_fields ] ) and ( [ rotation_sources ] )
, empirical results force the physical choice @xmath40 .
nevertheless , we wish to keep the so(2 ) symmetry even when an axion - like term is added to the electromagnetic lagrangian .
maxwell s eqs .
( [ maxwell_equations ] ) modify if the additional axion - like term @xmath41 is added to the em lagrangian @xmath42 . in eq .
( [ axion_lagrangian ] ) , @xmath4 is a pseudo - scalar field known in the particle physics literature as the axion - like field , and @xmath43 is a coupling constant ; the factor @xmath44 assures that @xmath43 is dimensionless .
the effects to maxwell s equations , of the axion - like term above , have been long discussed @xcite @xcite @xcite @xcite @xcite .
notice that the appearance of the axion - like term in eq .
( [ axion_lagrangian ] ) explicitly breaks the so(2 ) symmetry in eq .
( [ rotation_em ] ) , unless we are at the cp - preserving configuration @xmath45 .
when we consider the em lagrangian with the term in eq .
( [ axion_lagrangian ] ) added , the resulting equations of motion for the @xmath0 and @xmath1 fields are maxwell s eqs .
( [ maxwell_equations ] ) , with the electric charge and current densities replaced by @xmath46.\ ] ] the axion - like term in eq .
( [ axion_lagrangian ] ) can be written in a manifestly covariant form as @xmath47 and the lagrangian for the axion electrodynamics is @xmath48 where the axion lagrangian with a potential @xmath49 is @xmath50 using the euler - lagrange equations for @xmath51 , @xmath52 we obtain the dynamical equations of motion for the electromagnetic field @xmath53 where @xmath54 .
we now derive the set of maxwell s equations that takes into account the axion - like term introduced in sec .
[ adding an axion - like term ] , and is invariant under the duality transformation discussed in sec . [ duality symmetry and magnetic sources ] .
in fact , the lagrangian for the axion electrodynamics in eq .
( [ lagrangian_axion ] ) allows us to derive the set of eqs .
( [ axion_field_equations ] ) , but does not provide any information on the additional equation that the dual tensor @xmath55 satisfies . usually , in the axion electrodynamics literature , it is assumed the constraint @xmath56 which is the same as the one in maxwell s eq .
( [ covariant_eq ] ) in the absence of an axion term and magnetic sources .
however , this choice makes the set of eqs .
( [ axion_field_equations ] ) and ( [ additional_constraint ] ) not invariant under the so(2 ) duality symmetry . here
, we impose the duality symmetry to the in the axion - like theory as well , and we replace the condition in eq .
( [ additional_constraint ] ) with a dynamical equation for @xmath57 .
we start by noticing that , in the symmetric eq .
( [ covariant_eq_monopole ] ) , the second line is obtained from the first line by using the transformations in eq .
( [ rotation_fields ] ) and ( [ rotation_sources ] ) with @xmath58 .
applying the same transformations to the first line of eq .
( [ axion_field_equations ] ) , we obtain @xmath59 together , eqs .
( [ axion_field_equations ] ) and ( [ additional_field_eq ] ) relate the field strengths @xmath55 and @xmath60 to the electric and magnetic currents , @xmath36 and @xmath37 , and to the axion - like field @xmath61 . summing up , the complete set of equations for the axion - electromagnetic field is @xmath62 in terms of the electric and magnetic fields @xmath0 and @xmath1 , eq . ( [ maxwell_axion_complete ] ) rewrites as @xmath63 rotating the fields and sources by an angle @xmath35 , so that @xmath23 and @xmath64 vanishes , we finally obtain maxwell s equations in the presence of an axion - like field @xmath65 this set of equations for the axion electrodynamics was first derived in refs .
@xcite @xcite using a four - vector lagrangian . instead ,
here we have avoided complications due to a vector lagrangian , and we have derived this set of equations by using the gauge symmetry in eq .
( [ rotation_em ] ) .
from here to the end of the paper , we switch to natural units . in order to consider the propagation of axion - electromagnetic waves in free space , we fix the external sources @xmath66 . assuming an axion potential of the form @xmath67 with @xmath68 the axion mass , the set of wave eq .
( [ maxwell_axion1 ] ) in free space reads @xmath69 taking the curl of the third and fourth lines in eq .
( [ maxwell_axion_free_wave ] ) and rearranging , we obtain @xmath70 setting @xmath71 one solution to the set of wave equations for the fields @xmath72 and @xmath73 is @xmath74 this solution describes transverse plane waves propagating along the @xmath75-direction with momentum @xmath76 , frequency @xmath77 , and amplitude governed by @xmath78 and @xmath79 . the wave equations for @xmath72 and @xmath73 force @xmath80 .
the corresponding values of the fields @xmath0 and @xmath1 are found by inverting eq .
( [ definition_hateb ] ) , obtaining @xmath81 since this solution in eq .
( [ solution_hateb ] ) yields @xmath82 , the electric and magnetic fields propagate orthogonally when in the absence of external sources also in the axion electrodynamics .
however , the solution for the electric and magnetic fields is no longer expressed in terms of plane waves , because of the presence of the axion field @xmath61 in eq .
( [ solution_eb ] ) .
in fact , the equation for the axion field reduces to the usual klein - gordon equation , @xmath83 a solution of which is a wave of momentum @xmath84 , frequency @xmath85 , and amplitude @xmath86 , @xmath87 notice that , because of the orthogonality between the electric and magnetic fields , the equation of motion for the axion field describes the propagation of a free scalar massive particle along the direction @xmath84 .
in writing the set of eqs .
( [ maxwell_axion_complete_1 ] ) , we have expressed the electric and magnetic fields in terms of the vector potential @xmath88 and the scalar potential @xmath12 as in eq .
( [ gauge_classic ] ) , @xmath89 however , in the axion electrodynamics theory , these potentials do not directly satisfy the gauge symmetry in the new set of maxwells eq .
( [ maxwell_axion1 ] ) , and new gauge potentials have to be introduced . to find the gauge field equations in the axion electrodynamics theory
, we write the third line in eq .
( [ maxwell_axion1 ] ) as @xmath90 where @xmath91 is a new vector gauge field , and we use the identity @xmath92 . using this definition for @xmath91 in the fourth line in eq .
( [ maxwell_axion1 ] ) , and switching the order of the curl and the time derivative , we obtain @xmath93 thanks to the identity @xmath94 , valid for any scalar field @xmath95 , we obtain @xmath96 rearranging eqs .
( [ gauge_1 ] ) and ( [ gauge_2 ] ) , we finally obtain @xmath97 these gauge conditions define the electric and magnetic fields in terms of the gauge potentials @xmath91 and @xmath95 , and reduce to the ordinary electrodynamics potentials in eqs .
( [ gauge_classic1 ] ) when either @xmath98 or @xmath99 .
axion electrodynamics can be reformulated in terms of the gauge potentials . inserting the gauge eqs .
( [ gauge_1 ] ) and ( [ gauge_2 ] ) in the first two lines in eq .
( [ maxwell_axion1 ] ) and rearranging , we obtain the set of wave equations for the axion - electromagnetic field @xmath100 the first two lines in eq .
( [ axion_em_wave ] ) resemble the wave equations found in the classical electrodynamics theory , with the only difference lying in the definition of the gauge potentials @xmath91 and @xmath95 given in eqs .
( [ eb_gauge ] ) . once the gauge has been fixed so that to fulfill the lorenz - lorentz condition @xmath101 the coupled set of eqs .
( [ axion_em_wave ] ) in this gauge reads @xmath102 eq .
( [ axion_em_wave1 ] ) represents the set of equations for coupled axion - electromagnetic waves in the presence of sources @xmath5 and @xmath6 .
when switching back to the description of the dynamics in terms of the potentials @xmath88 and @xmath12 , the analogous of the wave eqs .
( [ axion_em_wave ] ) are found by arranging maxwell s source equations in eq .
( [ maxwell_axion1 ] ) , giving @xmath103 + \partial_t\,\left({{\bf \nabla}}\cdot { { \bf a}}+ \partial_t \phi\right ) , \\
\box\,{{\bf a}}= \left[{\bf j}_e - \kappa \,(\partial_t\,\theta)\,{{\bf b}}- \kappa\,{{\bf \nabla}}\theta\times{{\bf e}}\right ] - { { \bf \nabla}}\,\left({{\bf \nabla}}\cdot { { \bf a}}+ \partial_t \phi\right ) , \\ \box\,\theta = -\kappa\,{{\bf e}}\cdot{{\bf b}}- \frac{\partial u(\theta)}{\partial\theta}. \end{array}\ ] ] the terms in square brackets in eq .
( [ axion_em_wave_1 ] ) correspond to the new source terms in the presence of the axion field .
in fact , eq . ( [ axion_em_wave_1 ] ) can be obtained directly from eq .
( [ axion_em_wave ] ) by substituting the sources as in eq .
( [ substitution ] ) , while changing @xmath104 and @xmath105 .
we consider the temporal gauge @xmath106 , in which @xmath107 , and the last two lines in eq .
( [ axion_em_wave_1 ] ) read @xmath108 we assume that the vector @xmath88 is of the form @xmath109 where @xmath110 depends on time only , while @xmath111 is constant and equal to the magnitude of the magnetic field , @xmath112 . in this condition , the equations for the @xmath113-component of @xmath88 and for @xmath61 with the axion potential in eq .
( [ quadratic_potential ] ) read @xmath114 where a dot indicates a partial derivation with respect to time .
the equations above show a symmetry between the axion and the electric fields when either @xmath115 or when @xmath116 .
the first case corresponds to introducing a massive photon term @xmath117 in the lagrangian , with the photon mass being equal to the axion mass ; the second case corresponds to a massless axion . here
, we discuss the first and more general case , which leads to @xmath118 this set of equations is invariant under the symmetry @xmath119 where @xmath120 is a dimensionless parameter describing the o(2 ) rotational symmetry between @xmath110 and @xmath61 . following the work by guendelman @xcite
, we write the interaction part of the action as @xmath121 so that , once the total derivative term is dropped , the action describing the axion electrodynamics for @xmath61 and the field @xmath110 is @xmath122\,d^4x.\ ] ] the lagrangian term in the equation above is symmetric with respect to the transformation in eq .
( [ o2_symmetry ] ) . given a generic field @xmath123 and a lagrangian @xmath124 invariant under an internal symmetry @xmath125 ,
noether theorem states that a set of conserved currents is given by @xmath126 for the axion electromagnetic action in eq .
( [ axion_em_action ] ) , we obtain @xmath127 imposing the conservation of the current , @xmath128 , gives @xmath129 which can also be obtained by combining the lines in eq .
( [ axion_em_wave4 ] ) .
in eq . ( [ maxwell_axion1 ] ) , we have presented a new set of equations describing the evolution of an electromagnetic field in the presence of an axion - like field @xmath61 .
this set of equations is a generalization of maxwell s equations when an axion term is present , and it has been derived by requiring its invariance under the dual transformation in eq .
( [ rotation_em ] ) .
this latter request makes this work distinct from previous literature on the subject , in which the dual transformation has not been invoked .
we showed that this set of equations allows for the propagation of a coupled axion - electromagnetic wave , and in sec .
[ propagation of waves in the axion electrodynamics ] we have given a particular solution in which the axion field propagates according to a klein - gordon equation , see eq .
( [ solution_axion ] ) , with the orthogonal electric and magnetic fields being expressed in eq .
( [ solution_eb ] ) . similarly to the classical electrodynamics , the electromagnetic field propagates as a transverse wave ; on the contrary , the axion wave behaves as a longitudinal wave .
we have derived the new set of gauge symmetries in the axion electrodynamics , and we have discussed the wave equations in terms of the gauge potentials in eq .
( [ axion_em_wave1 ] ) . using this set of equations ,
we have derived a new internal symmetry of the axion - electromagnetic field lagrangian , between the axion field and the gauge potential @xmath88 , which leads to a conserved current thanks to noether theorem .
the author would like to thank paolo gondolo ( u. of utah ) for useful discussions .
d. m. lipkin , j. math .
phys . * 5 * , 696 ( 1964 ) .
t. a. morgan , j. math .
phys . * 5 * , 1659 ( 1964 ) . t. w. kibble , j. math . phys .
* 6 * , 1022 ( 1965 ) .
d. m. fradkin , j. math . phys .
* 6 * , 879 ( 1964 ) .
a. karch , phys .
lett . * 103 * , 171601 ( 2009 ) .
n. anderson and a. m. arthurs , int .
j. electron .
* 45 * , 333 ( 1978 ) .
j. rosen , am . j. phys . * 48 * , 1071 ( 1980 ) .
p. sikivie , phys .
* 137b * , 353 ( 1984 ) . m. c. huang and p. sikivie , phys .
d * 32 * , 1560 ( 1985 ) .
f. wilczek , phys .
lett . * 58 * , 1799 ( 1987 ) .
m. gasperini , phys .
lett . * 59 * , 396 ( 1987 ) .
l. jacobs , physica b * 152 * , 288 ( 1998 ) .
a. sudbery , j. phys .
a math . gen . * 19 * , l33 ( 1986 ) .
s. c. tiwari , physics essays * 4 * , 212 ( 1991 ) .
e. i. guendelman , mod .
a * 23 * , 191 - 196 ( 2008 ) . | we extend the duality symmetry between the electric and the magnetic fields to the case in which an additional axion - like term is present , and we derive the set of maxwell s equations that preserves this symmetry .
this new set of equations allows for a gauge symmetry extending the ordinary symmetry in the classical electrodynamics .
we obtain explicit solutions for the new set of equations in the absence of external sources , and we discuss the implications of a new internal symmetry between the axion field and the electromagnetic gauge potential . |
agent - based modeling can be used to study systems exhibiting emergent properties which can not be explained by aggregating the properties of the system s components.@xcite statistical mechanics and economics share the property to analyze big ensembles where the collective behaviour is found out as a result of interactions at the microscopic level and where agent - based simulations can be applied .
many systems are studied in terms of the nature that defines their inner components while others are considered from the point of view of the interactions among the agents that can be pictured through a complex network .
plenty of information is encoded in connectivity patterns .
hierarchical structures appear in a natural way when we study societies and , according to many authors,@xcite one of the milestones is to understand why and how from individuals with initial identical status , inequalities emerge .
this is related to the question of hierarchy formation as a self - organization phenomenon due to social dynamics.@xcite elitarian distributions can arise starting from a society where people initially own an equal share of economic resources , _ e.g. : _ the exponential distribution for the low and medium income classes in western societies.@xcite we consider dragulescu - yakovenko gas - like models in economic exchanges@xcite so , let our system be composed of @xmath0 economic agents , being @xmath1 and constant .
each agent @xmath2 owns an amount of money @xmath3 so the state of the system at a given time is defined by the values that every variable @xmath3 takes at that moment , @xmath4 money distribution among the agents should never be confused with the notion of wealth distribution .
money is only one part inside the whole concept of wealth .
transfer of money represents payment for goods and services in a market economy .
we study simplified models which keep track of that money flux but do not keep track of what goods or services are delivered . at each interacting step
, agents trade by pairs and local conservation of money is sustained , @xmath5 transactions result in some part of the money involved in the interaction changing its owner .
for simplicity , we do not consider models where debts are allowed .
it is deeply established in the common knowledge that highly - ranked individuals in societies have easier access to resources and better chances to compete .
this is a motivation to look for internal correlations between money and surrounding environment .
we wonder if the exchange rules that define simple gas - like models for random markets , when implemented on networks , are capable of depicting correlations between purchasing power of an agent inside a social network and the influence of the agent on the rest of the system .
we associate the purchasing power concept to the mean money per economic agent computed as a function of the connectivity degree of each agent in a network . at this level ,
influence of an agent is only related to the degree of the node representing the agent .
we implement the exchange rules on two type of networks : uniform random spatial graph and barabsi - albert model , and then examine the relationship between the former econo - social agent indicators for the different underlying architectures . in section [
sec : randommarkets ] , we review two well - known random undirected exchange rules : general and uniform savings models . in section [ sec : rde ] , we introduce a new family of interactions : random but directed ones .
the main property of this simple exchange rule is that it is a real inspired model where social inequalities in money distribution emerge in a natural way . in section [ sec : ns ] , we show the relation between mean money per economic agent and the connectivity degree of the agent . for the models with undirected exchange rules , we observe no correlation between money and the degree of the nodes .
linear dependence is found for the new random exchange model we propose .
section [ sec : conc ] is devoted to gather the most relevant conclusions .
for some random economic systems where money is a conserved quantity , the asymptotic distribution of money among the agents is given by the boltzmann - gibbs distribution ( bg ) , @xmath6 where the role of the effective temperature is played by the average amount of money per agent , @xmath7 this feature was first shown by dragulescu and yakovenko in 2000 by means of numerical simulations.@xcite subsequently , analytical justification was given by lpez - ruiz _ et al .
_ in 2008 and 2012 .
bg can be geometrically deduced@xcite under the assumption of equiprobability of the possible economic microstates .
we also know that an asymptotic evolution towards bg is obtained regardless of the initial distribution for those systems with total money fixed and when considering random symmetric interactions between pairs of components.@xcite this comes from bg being the stable fixed point of the distributions space , @xmath8 under the iterated action @xmath9 of the integral operator @xmath10 given by @xmath11(z)=\iint_{s(z)}\frac{p(x)\,p(y)}{x+y}\;dx\,dy,\label{integral}\ ] ] where @xmath12 is the integration domain .
+ let us now consider the gas - like model originally proposed by dragulescu and yakovenko so , at each computational step , we randomly choose a pair of agents and then , one -the labeled as @xmath2- is chosen to be the winner in the interaction process and the other one -labeled as @xmath13- becomes the loser and , according to the previously stated rule ( [ transactiondelta ] ) , an amount of money @xmath14 is transferred from the loser to the winner . assuming @xmath15 , it is obvious
that if the loser does not have enough money to pay , which is nothing but the local condition @xmath16 , the transaction is forbidden and we should proceed with a different pair of agents . instead of considering the restriction in the interaction , we state the exchange rule considering @xmath17 that gives rise to the completely random case given by @xmath18 where @xmath19 $ ] is a uniform random number which is refreshed at every computational step .
observe that both agents can be winner or loser in a symmetric way , depending on the random number @xmath20 at each step .
this approach also ensures that no agent will evolve to own a negative amount of money or , in other words , debts are not allowed .
the condition @xmath21 for every agent @xmath2 in the system is accomplished in a natural way .
this exchange rule ( [ transactionrandom ] ) is a very rough macroeconomic model where individuals or corporations raise their money for a venture and then , the market effect or their mutual interaction determines the final distribution .
see figure [ rg ] .
the concept of savings arises in an obvious way from observing human behaviour when people are inmerse in a market economy.@xcite this feature is introduced through a parameter , @xmath22 $ ] , which is called a propensity factor.@xcite this means that each agent saves a fraction @xmath23 of its money when an interaction occurs and trades randomly with the other part : @xmath24 we consider the model with uniform savings which means that @xmath23 is fixed to be constant among the agents and with no dependence on the time .
the statistically stationary distribution @xmath25 decays rapidly on both sides of the most probable value for the money per agent which , in this case , is shifted from the poorest part of the system to @xmath26 when @xmath27 see figure [ sg ] .
this behaviour was already described as a self - organising feature of the market induced by self - interest of saving by each agent without any global perspective in an analogous way to the self - organisation in markets with restricted comodities.@xcite first attempt to give a quantitative description for the steady distribution towards this model evolves is due to patriarca _
et al . _ in 2004 .
they stated that numerical simulations of ( [ transactionsavings ] ) could be fitted to a standard gamma distribution.@xcite subsequently ( 2007 ) , chatterjee and chakrabarti offered a brief study of the consequences that this modelization implies and stated that as @xmath23 increases , effectively the agents retain more of its money in any trading , which can be taken as implying that with increasing @xmath23 , temperature of the scattering process changes.@xcite according to their study , fourth and higher order moments of the distributions are in discrepancy with those of the gamma family so , the actual form of the distribution for this model still remains to be found out .
( 2011 ) gave an iterative recipe to derive an analytical expression solving an integral equation.@xcite a similar expression was derived in a different way by lallouache _
we have shown how the propensity factor @xmath23 is introduced ( [ transactionsavings ] ) as a variation for the general random undirected exchange rule ( [ transactionrandom ] ) showing how , from individual responsible decissions -such as saving a fraction of your money when entering an exchange market- , self - organized distributions where the mean and the mode are close arise so , richness would be quite balanced distributed among the group .
a completely different scheme@xcite that modifies the general rule ( [ transactiondelta ] ) proposes a random sharing of an amount @xmath28 ( instead of @xmath29 ) only when @xmath30 , trading at the level of the lowest economic class in the trade .
this model leads to an extreme situation in which all the money in the market drifts to one agent and the rest become truely pauper .
from this idea , we give a new and more general exchange rule reflecting this directed or biased orientation for the interaction and including this particular result .
we propose an integral operator which is the analytical approach to this new rule in the mean - field or gas - like case and , in section [ sec : ns ] , we implement this rule on networks to study how it is affected when we mix directed interactions with undirected networks .
the directed exchange can be understood as a first approach to microeconomic activities where money is transferred only in one direction , similar to payments for goods .
we consider the most general family of interations , @xmath31 where @xmath19 $ ] is a random number chosen with uniform probability . at each time
, the system is described by the probability distribution function of money when we choose one of the agents randomly , @xmath32 in the continuous limit , we can picture the system to be the combination of two identical copies , @xmath33 and @xmath34 of the original system itself , @xmath35 for each interaction we can consider that the two different agents @xmath2 and @xmath13 , with money values being @xmath3 and @xmath36 , are two realizations of picking up randomly one agent from each copy , @xmath37 and @xmath38 respectively . in every interaction step @xmath39
there is a transaction of the agents conforming the distribution @xmath25 to a new configuration given by @xmath40 where the agents @xmath41 will conform @xmath42 and @xmath43 will conform @xmath44 let us now consider the probability of a randomly chosen agent among the first copy @xmath45 owning an amount of money @xmath46 after the interaction happened . from (
[ transactiondirected ] ) , it is clear that @xmath47 , and as the result @xmath41 is uniformly distributed in @xmath48 $ ] so , the probability of obtaining a certain value @xmath49 is given by @xmath50 .
the interaction of pairs @xmath51 in the first configuration of the system gives rise to the evolution of @xmath33 to the following probability @xmath52 of obtaining @xmath46 : @xmath53 for the second copy , @xmath54 we should consider that money of the second agent after the interaction , @xmath55 should be a value between the initial money it has , @xmath56 , and the maximum possible amount of money it can have after the interaction , which will be associated to get all the money from the first agent
so , @xmath57 again , the length of the segment @xmath58 $ ] is @xmath59 so , the probability of @xmath43 is uniformly distributed in that segment and so , the probability to have @xmath60 will be @xmath50 .
the expression for the probability @xmath61 of having @xmath60 in the second copy of the system results in @xmath62 we explicitly compute the norm , @xmath63 and the expected value , @xmath64=\frac{3}{2}\langle u\rangle.\nonumber\end{aligned}\ ] ] from results ( [ normop ] ) and ( [ expop ] ) , together with definition ( [ op1 ] ) , the norm and mean value of the distributions are conserved when we define the operator for the directed random interaction by @xmath65(x)=\frac{1}{2}\int_{u > x}\frac{p(u)}{u}\,du+\frac{1}{2}\iint_{v < x < u+v}\frac{p(u)\,p(v)}{u}\,du\,dv.\label{opdef}\ ] ] although we can not give a proof for the infinite iteration of this operator , we see that it piles up the distribution at the lower values for @xmath66 which gives rise to a very impoverished population and a very slightest fraction of too opulent agents .
even with such a very narrow initial distribution we choose , the effect of this operator is very strong in only a couple of iterations .
this result is in perfect agreement with the first model of biased interaction@xcite we mentioned in our previous section .
other models with asymetric rules that establish a transition between boltzmann - gibbs and pareto distributions can be found in the literature.@xcite see figure [ operador ] .
we stated in the introduction of this work the desire of finding a first model which , when implemented on networks , is able to show a relation between money and influence of an agent on the system . for this purpose
, we simulate the two models we have already studied in section [ sec : randommarkets ] and the new one we have presented in section [ sec : rde ] .
we choose two representative cases : the random uniform spatial network ( sp ) and the barabsi - albert model ( ba ) .
sp is an easy way to build networks with poissonian degree distributions@xcite and ba is an algorithm for generating random scale - free networks following the preferential attachment prescription.@xcite at the beginning , every agent is given the same amount of money so , the initial distribution of money among the agents is written @xmath67 and we obtain the steady state distribution for the exchange rules ( [ transactionrandom ] ) , ( [ transactionsavings ] ) and ( [ transactiondirected ] ) implemented on these two different topologies . from the histogram related to the distribution of money among the agents , we can consider the entropy associated to that distribution in a discrete form , @xmath68 where @xmath69 just recalls the coarse graining when computig @xmath25 as a discrete histogram . for simplicity
, we measure @xmath70 the entropy as a fraction of the maximum which is given for the exponential distribution . if the money distribution is affected by the topology of the underlying network , it should show some kind of dependence on the degree of the agents . for our purposes
it is enough for the reader to associate the degree of a node to its number of neighbouring agents it can interact with . as we will only use for now undirected and static networks ,
the degree of an agent will be a distinctive feature .
we define the _ mean money per economic agent as a function of k _ ,
@xmath71 given by @xmath72 note that we compute the mean money of the nodes inside each class of connectivity at every step of our simulation and then , we consider the time - averaged mean money per node according to the different degrees .
@xmath73 denotes how many nodes have degree @xmath74 and our simulations run for @xmath75 steps .
there is no need to be worried about the transitory regime disturbing the computed results for @xmath76 because it is negligible : less than @xmath77 of total computed steps in the worst case ( see figure [ dnet ] ) .
we also compute the standard deviation , @xmath78 , given by @xmath79 ^ 2}\label{sigmamk}\\\nonumber\;\;\forall\,i\;:\;k_i = k.\end{aligned}\ ] ] in figures [ rnet ] , [ snet ] and [ dnet ] we show , for the rules ( [ transactionrandom ] ) , ( [ transactionsavings ] ) and ( [ transactiondirected ] ) , respectively , the stationary probability distribution @xmath25 with entropy - evolution of @xmath80 for the whole simulation with inset detail of the transitory regime when it is required . when we plot the distributions @xmath81 and @xmath78 , it is also shown the characteristic degree distribution of the networks we implement .
we clearly see in these figures how rules ( [ transactionrandom ] ) and ( [ transactionsavings ] ) are transparent to the underlying topology , provoking always the decay of the economic system to the bg or gamma - like distributions , respectively , indistinctly of the different connectivities of the agents .
we can say that for these types of interactions the economic classes are blind respect to the social influence of the agents . by contrary , the new directed rule ( [ transactiondirected ] ) separates the agents in economic classes correlated with their connectivities , in such a way that more connected agents show a bigger propensity to accumulate more money , in this case with a linear relationship between money and connectivity .
this is a characteristic that , in general and independently of the political system installed in the power , seems to be more likely to be found in the reality .
\(a ) spatial network : \(b ) barabsi - albert model : \(a ) spatial network : \(b ) barabsi - albert model : \(a ) spatial network : \(b ) barabsi - albert model :
we have found that topology does not determine the final equilibrium distribution for the family of undirected random markets , both with or without uniform savings , as can be seen by comparing the distributions @xmath25 from figures [ rnet ] and [ snet ] to those from figures [ rg ] and [ sg ] .
thus , from the uniform value of @xmath81 and @xmath78 , we can conclude that the connectivity of an agent immersed in undirected random markets does not determine if the agent will have more or less money , that is , the degree @xmath74 of an agent does not decide its richness .
when we consider the directed random market given by the operator ( [ opdef ] ) , simulations plotted in the figure [ dnet ] suggest that this model reproduces very clearly the real insight based on the impression that in certain societies the richness is owned by a very small fraction of the population and poverty is extended among the majority of the agents . for this type of economies
, we also discover that the underlying topology determines the stationary distribution of money among the agents , @xmath25 , and that the connectivity of each agent proportionally determines its average richness . assuming that we can apply this rule for any quantity that can be exchanged , and
not only money , this result introduces the interesting idea of how we can create systems with a property shared by its inner components with an steady distribution essentially determined by the topology defined by the connections between the agents , although evidently this statistical equilibrium is dynamical and presents a continuous flow between those agents .
it is also interesting to highlight how this system evolves from the initial zero - entropy state to a state with maximum entropy and then how it relaxes towards the asymptotic equilibrium state .
let us conclude by saying that we have introduced a new directed random market model in the context of economic gas - like models , which can be understood as a first and simple model characterized by _ more connectivity implies more money_. 00 a. namatame , t. kaizouji and y. aruka , editors , _ the complex networks of economic interactions .
essays in agent - based economics and econophysics _
( springer , 2006 ) . c. castellano , s. fortunato and v. loreto , _ rev .
* 81 * , 591 ( 2009 ) .
c.tovey , d. spangler - martin , i.d . chase and m.
manfredonia , _ pnas _ * 99 * , 5744 ( 2002 ) .
a. dragulescu and v.m .
yakovenko , _ eur .
j. b _ * 20 * , 585 ( 2001 ) .
a. dragulescu and v.m .
yakovenko , _ physica a _ * 299 * , 213 ( 2001 ) .
a.c . silva and v.m .
yakovenko , _ europhys .
lett . _ * 69 * , 304 ( 2005 ) .
yakovenko , in _ encyclopedia of complexity and system science _ , r. a. meyers ( ed . ) ( springer 2009 ) .
a. dragulescu and v.m .
yakovenko , _ eur .
j. b _ * 17 * , 723 ( 2000 ) .
r. lpez - ruiz , j. saudo and x. calbet , _ am .
j. phys . _
* 76 * , 780 ( 2008 ) .
lpez , r. lpez - ruiz and x. calbet , _ j. math
appl . _ * 386 * , 195 ( 2012 ) . r.h .
frank , _ microeconomics and behavior _ ( mcgraw - hill / irwin , 2009 ) .
a. chakraborti and b.k .
chakrabarti , _ eur .
phys . j. b _ * 17 * , 167 ( 2000 ) .
a. chakraborti , s. pradham and b.k .
chakrabarti , _ physica a _ * 297 * , 253 ( 2001 ) .
m. patriarca , a. chakraborti and k. kaski , _ phys .
e _ * 70 * , 016104 ( 2004 ) .
a. chatterjee and b.k .
chakrabarti , _ eur .
j. b _ * 60 * , 135 ( 2007 ) .
x. calbet , j .- l .
lpez and r. lpez - ruiz , _ phys .
e _ * 83 * , 036108 ( 2011 ) .
m. lallouache , a. jedidi and a. chakraborti , _ science and culture _ * 76 * , 478 ( 2010 ) .
a. chakraborti , _ int .
c _ * 13 * , 1315 ( 2002 ) . c. pellicer - lostao and r. lpez - ruiz , _ int .
c _ * 22 * , 21 ( 2011 ) .
m. barthelemy , _ physics reports _ * 499 * , 1 ( 2011 ) .
r. albert and a .-
barabsi , _ rev
phys . _ * 74 * , 47 ( 2002 ) . | boltzmann - gibbs distribution arises as the statistical equilibrium probability distribution of money among the agents of a closed economic system where random and undirected exchanges are allowed .
when considering a model with uniform savings in the exchanges , the final distribution is close to the gamma family . in this work , we implement these exchange rules on networks and we find that these stationary probability distributions are robust and they are not affected by the topology of the underlying network .
we introduce a new family of interactions : random but directed ones . in this case
, it is found the topology to be determinant and the mean money per economic agent is related to the degree of the node representing the agent in the network .
the relation between the mean money per economic agent and its degree is shown to be linear . |
confining a binary mixture one can profoundly alter its miscibility behavior.@xcite if a mixture is confined into a quasi one dimensional ( e.g. , cylindrical ) pore no true phase transition occurs , unlike the prediction of the mean field theory .
fluctuations destroy long range order and only a pronounced maximum of the susceptibility remains in the vicinity of the unmixing transition in the bulk . in a two dimensional system ( e.g. , a slit like pore or a film ) with identical surfaces a true phase transition occurs ( capillary condensation ) and the shift of the critical point away from its bulk value has been much investigated.@xcite the confinement changes the universality class of the transition from 3d ising critical behavior in the bulk to 2d ising critical behavior in the film .
the latter manifests itself in much flatter binodals in a film close to the unmixing transition than in the bulk .
no such change of the critical exponents is observed in mean field theory .
the phase behavior of symmetric mixtures in a thin film with antisymmetric surface interactions has attracted abiding interest recently.@xcite the right surface attracts one species with exactly the same strength as the opposite surface attracts the other species .
in contrast to capillary condensation , the phase transition does not occur close to the unmixing transition in the bulk , but rather in the vicinity of the wetting transition .
close to the unmixing transition in the bulk , enrichment layers at the surfaces are gradually built up and an interface is stabilized in the middle of the film . in this `` soft mode ''
phase the system is laterally homogenous no spontaneous breaking of the symmetry occurs .
if the wetting transition of the semi infinite system is of second order one encounters a second order localisation
delocalisation transition slightly below the wetting transition temperature .
the system phase separates laterally into regions where the interface is located close to one surface ( localized state ) .
the order parameter , i.e. , the distance between the interface and the center of the film , grows continously .
this prediction of phenomenological theories has been corroborated by detailed simulation studies@xcite and it is also in accord with experimental findings.@xcite if the wetting transition is of first order and the thickness of the film not too small , mean field calculations@xcite predict the occurance of two critical points which correspond to the prewetting critical point of the semi infinite system . unlike the wetting transition,@xcite the prewetting transition can produce a critical ( singular ) behavior in a thin film , because only the lateral correlation length diverges at the prewetting critical point ; the thickness of the enrichment layers at the surfaces remains finite .
the mean field treatment invokes approximations and it can not be expected to capture the subtle interplay between 2d ising fluctuations at the critical points , `` bulk''like composition fluctuations , and interface fluctuations typical for the wetting transition.@xcite consequently , a detailed test of the mean field predictions via monte carlo simulations is certainly warranted and elucidates the role of fluctuations .
using monte carlo simulations of the ising model ferrenberg _
_ studied the interface localisation
delocalisation transition also for the case that the wetting transition of the semi infinite system is of first order.@xcite the simulation study was centered on the dependence on the film thickness , which is a convenient parameter to be varied in experiments .
however , the study was restricted to the coexistence between strictly symmetric phases and many questions remained open .
the general features of the phase behavior are shared by all binary mixtures . here , we present large scale monte carlo simulations aiming at investigating the phase behavior of a symmetric binary polymer blend confined between antisymmetric walls .
computationally , simulations of a polymer blend@xcite are much more demanding than studying simple fluids ( e.g. , the ising model ) , but recent mean field calculations made detailed predictions for the phase behavior of confined polymer mixtures@xcite and serve as guidance for choosing the model parameters in the simulations .
simulating polymer blends , we can , at least in principle , control the importance of fluctuations by varying the degree of interdigitation , i.e. , the chain length.@xcite the mean field theory is expected to become accurate in the limit of infinite interdigitation . in a binary polymer blend
the wetting transition occurs at much lower temperatures than the critical temperature of the unmixing transition in bulk.@xcite hence , `` bulk''like composition fluctuations are not important in the vicinity of the wetting transition temperature and we can isolate the effect of interface fluctuation . moreover , these systems are also suitable candidates to examine the phase behavior experimentally .
indeed , one of the first studies of the `` soft mode '' phase has employed a binary polymer blend.@xcite our paper is broadly arranged as follows : first , we present a phenomenological description of the phase behavior in a film with antisymmetric short range surface interactions . using a standard model for the effective interface potential
we calculate the phase behavior in mean field approximation , discuss the regime of validity of the mean field approach , and consider the crossover between the different critical behaviors .
second , we briefly describe our coarse grained lattice model for a binary polymer mixture .
then , we present our monte carlo results : we obtain the phase diagram for film thicknesses ranging from @xmath5 to @xmath6 , where @xmath7 denotes the radius of gyration of the polymer chains , investigate the critical behavior and present evidence that interface fluctuations renormalize the effective interface potential . we close with a comparison of the phase diagram to the behavior of the bulk and of films with symmetric boundary conditions .
rather than describing the configuration of the system by the detailed composition profile across the film , much qualitative insight into the thermodynamics can be deduced from the effective interface potential . below the bulk critical temperature enrichment layers of the prefered components form at the surfaces and stabilize an @xmath8 interface which runs parallel to the walls . the effective interface potential
@xmath9 describes the free energy per unit area as a function of the distance @xmath10 between this @xmath8 interface and a wall . in the case of short range forces between the monomers and the walls ,
the interface profile becomes distorted in the vicinity of the walls and this gives rise to an interaction which decays exponentially as a function of the distance l between the @xmath8 interface and a single wall : @xmath11 this expression retains only the lowest powers of @xmath12 , which are necessary to bring about a first order wetting transition of the semi infinite system .
the coefficient @xmath13 is explicitely temperature dependent , while the temperature dependence of @xmath14 and @xmath15 is neglected .
@xmath16 is assumed throughout the discussion .
all coefficients are of the same magnitude as the interfacial tension @xmath17 between the coexisting bulk phases . for polymer blends this quantity scales with chain length @xmath18 and monomer number density @xmath19 like @xmath20 .
@xmath21 measures the degree of interdigitation .
@xmath22 denotes the spatial range of the interactions and it is of the order @xmath7 .
@xmath23 gives rise to a second order wetting transition at @xmath24 ; and @xmath25 to a tricritical transition .
for @xmath26 one encounters a first order wetting transition at @xmath27 where the thickness of the enrichment layer jumps discontinuously from @xmath28 to a macroscopic value.@xcite the wetting spinodals take the values @xmath29 ( from the wet phase ) and @xmath30 ( from the non wet phase ) .
the concomitant prewetting line terminates at the prewetting critical point @xmath31 and @xmath32 .
we approximate the effective interface potential in a film to be the linear superposition of the interactions originating at each wall and analyze the behavior .
scf calculations@xcite lend support to this approximation .
the interface potential in a film of thickness @xmath0 takes the form : @xmath33 \right)-1\right ) - 2 b \exp(-\lambda d ) \left(\cosh \left ( 2 \lambda [ l - d/2 ] \right)-1\right ) \nonumber \\ & & + 2 c \exp(-3 \lambda d/2 ) \left(\cosh \left ( 3 \lambda [ l - d/2 ] \right)-1\right)\end{aligned}\ ] ] in general , the phase boundaries depend on the variables @xmath34 and @xmath35 .
if we proceeded as in the ref@xcite by expanding the @xmath36 in powers of @xmath37 $ ] , the further analysis would be rather cumbersome .
a more transparent procedure employs the variable @xmath38 \right)-1\right ) = \left ( \exp(-\lambda d/4 )
\lambda[l - d/2 ] \right)^2 + \mbox { higher orders of}\quad [ l - d/2]\ ] ] to rewrite the interface potential in the form @xmath39
\qquad \mbox{with } \qquad r=\frac{b-6c\exp(-\lambda d/2)}{2c } \qquad \mbox{and } \qquad t = \frac{a - a_{\rm wet}-b\exp(-\lambda d/2)}{c } \label{eqn : landau}\ ] ] the qualitative form of the effective interface potential has been inferred previously on the basis of a landau expansion,@xcite here it is derived explicitely from the standard form of the interface potential ( [ eqn : sw ] ) for a first order wetting transition in the semi infinite system .
negative values of @xmath40 correspond to second order localization delocalization transitions , @xmath41 to a tricritical one , and positive values of @xmath40 give rise to first order transitions .
@xmath42 measures the distance from the tricritical transition temperature ( for @xmath43 ) and denotes the triple temperature in the case of a first order interface localization delocalization transition ( cf .
below ) . for @xmath43
the phase boundaries depend only on the two parameter combinations @xmath40 and @xmath42 . in these variables
the limit @xmath44 is particularly transparent : @xmath45 , @xmath46 and @xmath47
. a second order interface localization delocalization transition ( i.e. @xmath49 ) will occur either if the wetting transition is second order ( i.e. , @xmath23 ) or if the wetting transition is first order but the film thickness @xmath0 small enough to comply with @xmath50 .
this behavior is in accord with previous findings@xcite and we shall corroborate this further by our present simulations . since the coexisting phases are symmetric with respect to exchanging @xmath10 and @xmath51 , phase coexistence occurs at @xmath52 or @xmath53 . from this condition
we obtain for the binodals : @xmath54 the critical temperature is given by @xmath55 and @xmath56 denotes the distance from the critical temperature at fixed @xmath40 . for @xmath49 the binodals at the critical point open with the mean field exponent @xmath57 .
this corresponds to mean field critical behavior ( 2dmf ) of a system with a single scalar order parameter , i.e. , @xmath58 $ ] . at larger distance
the order parameter grows like @xmath59 with @xmath60 .
the latter exponent is characteristic for the mean field behavior at a tricritical point ( 2dtmf ) .
the crossover between mean field critical and tricritical behavior occurs around @xmath61 . as we decrease the magnitude of @xmath62 we approach the tricritical point and the regime where mean field critical behavior is observable shrinks . at the tricritical point
only the tricritical regime ( 2dtmf ) exists , i.e. , @xmath63 , and the binodals take the particularly simple form @xmath64 .
the crossover in the binodals for @xmath65 is illustrated in the inset of fig.[fig : cross](*a * ) .
of course , the above considerations neglect fluctuations and the behavior close to the transition is governed by ising critical exponents and two dimensional tricritical exponents , respectively .
the crossover between ising critical behavior ( 2di ) and tricritical behavior ( 2dt ) occurs at @xmath66 , where the crossover critical exponent is not @xmath67 ( as for the crossover between the mean field regimes ) but rather @xmath68.@xcite following ref@xcite we calculate the critical amplitudes and estimate the location of the crossover between mean field critical behavior and the region where fluctuations dominate the qualitative behavior . for small values of the order parameter
@xmath69 $ ] we approximate @xmath70 and obtain for the mean field critical amplitudes : @xmath71 the susceptibility of the order parameter above the critical temperature is related to the inverse curvature of the interface potential in the middle of the film @xmath72 .
using eq([eqn : landau ] ) we obtain for critical and tricritical mean field transitions : @xmath73 the ratio @xmath74 of the critical amplitudes above and below the critical point is universal and takes the mean field value 2 at the critical point and 4 at the tricritical point . at the transition the correlation length @xmath75 diverges .
this lateral length is associated with fluctuations of the local interface position , i.e. , capillary waves . in mean field approximation
the parallel correlation length takes the form : @xmath76 and @xmath77 and @xmath78 , respectively . knowing the critical amplitudes we can estimate the importance of fluctuations via the ginzburg criterium:@xcite as it is well known , mean field theory is self
consistent if the fluctuations of the order parameter in a volume of linear dimension @xmath75 are small in comparison to the mean value of the order parameter . for our quasi
two dimensional system @xmath79 we obtain : @xmath80 this result is as expected : for our quasi to dimensional system we obtain for a second order interface localization delocalization transition @xmath81 in accord with ref@xcite , while we obtain @xmath82 upon approaching the tricritical point . for bulk ( @xmath83 ) tricritical phenomena landau theory is marginally correct .
combining the above results we find the following behavior upon approaching the critical temperature : far away from the tricritical point , i.e. , @xmath84 we find mean field tricritical behavior ( 2dtmf ) for @xmath85 , mean field critical behavior ( 2dmd ) for @xmath86 , and finally two dimensional ising critical behavior ( 2di ) for @xmath87 .
closer to the tricritical point , i.e. @xmath88 , we find mean field tricritical behavior ( 2dtmf ) for @xmath89 , two dimensional tricritical behavior ( 2dt ) for @xmath90 , and ising critical behavior ( 2di ) for @xmath91 .
the prefactor @xmath92 must be chose such that all crossover lines ( 2di @xmath93 2dt , 2dt @xmath93 2dtmf , 2dtmf @xmath93 2dmf , and 2dmf @xmath93 2di ) intersect in a common point .
this yields @xmath94 .
of course , the term `` crossover line '' is not meant as a sharp division between different behaviors , but should be understood rather as a center of a smooth crossover region .
likewise , the above constant @xmath92 may involve a constant of order unity which has been suppressed for simplicity .
the two different sequences can be clearly distinguished in the monte carlo simulations , because the probability distribution of the order parameter exhibits a three peak structure@xcite only close to the tricritical point ( 2dt ) .
we shall use this property to distinguish between the two different sequences in our mc simulations .
the anticipated behavior is summarized in fig.[fig : cross](*a * ) . in the monte carlo simulation
this rich crossover scenario is further complicated by finite size rounding .
the monte carlo results are subjected to pronounced finite size effects whenever the correlation length becomes of the order of the lateral system size . in the mean field regime
the correlation length scales like @xmath95 . knowing the ginzburg number for the crossover from 2dmf to 2di behavior we estimate the correlation length in the ising critical regime:@xcite @xmath96 where we have assumed that the scaling function @xmath97 assumes a power law behavior for small and large arguments and we have used the value @xmath98 appropriate for the divergence of the correlation length in the 2di regime .
similarly , we determine the correlation length in the 2dt regime : @xmath99 @xmath100 denotes the exponent of the correlation length in the 2dt universality class.@xcite the correlation lengths at the various crossovers are compiled in tab.1 .
the largest correlation length occurs at the crossover from 2dt to 2di behavior @xmath101 in order to observe the true ising critical behavior for negative values of r , the system size @xmath102 has to exceed this correlation length . in the vicinity of the tricritical point (
i.e. , for small negative values of @xmath40 ) this requirement is very difficult to be met in computer simulations . for positive values of @xmath40
the interface potential exhibits a three valley structure .
the three minima at @xmath104 and @xmath105 have equal free energy at @xmath106 .
this corresponds to the triple point . at lower temperatures an @xmath1rich phase
coexist with a @xmath2rich phase , and since the two phases are symmetrical the coexistence occurs at @xmath107 .
the binodals below the triple point take the form : @xmath108 above the triple temperature @xmath109 there are 2 two phase coexistence regions symmetrically located around @xmath105 .
these phase coexistences terminate in two critical points .
since the coexisting phases correspond to a thick and a thin enrichment layer of the prefered phase at each wall , there is no symmetry between the coexisting phases , and the exchange potential @xmath110 at coexistence differs from zero .
unfortunately , the phase boundaries for @xmath109 and @xmath103 depend not only on @xmath40 and @xmath42 but also on @xmath35 explicitely , and we have determined them numerically . the dependence of the critical temperature @xmath111 on @xmath40 for several values of @xmath35 is presented in fig.[fig : cross](*b * ) .
the coexistence curve for @xmath112 and various values of @xmath35 are presented in the inset of fig.[fig : cross](*b * ) . as the film thickness
is decreased the critical temperature decreases and the critical points move closer to the symmetry axis of the phase diagram .
they are determined by the condition : @xmath113 in two limiting cases a simple behavior emerges : \(i ) if @xmath114 we can replace the derivative w.r.t .
@xmath10 by derivatives w.r.t .
@xmath115 and obtain:@xcite @xmath116 and @xmath117 .
this approximation holds for @xmath118 . expanding @xmath119 in powers of @xmath120 we obtain ( omitting an irrelevant term linear in @xmath121 ) : @xmath122 this allows us to calculate the binodals in the vicinity of the critical points , the susceptibility and parallel correlation length .
the presence of a 5th order term in @xmath121 in the expansion ( [ eqn:15 ] ) is a manifestation of the fact that the phase boundaries of the prewetting transitions are not symmetric around @xmath123 .
this lack of symmetry is also evident from the numerical results in fig.[fig : cross](*b * ) ( insert ) .
the critical amplitudes scale in the same way with @xmath40 , @xmath124 , and @xmath35 as for @xmath49 .
in particular we find for the crossover between 2dmf behavior and 2di behavior @xmath125 .
\(ii ) in the limit of large film thickness @xmath44 , the critical point tends towards the prewetting critical point at @xmath126 . in this limit confinement effects
are negligible and the coexistence curves in the vicinity of the critical points corresponds to the prewetting lines at the corresponding surfaces .
we expect the same critical behavior as at the prewetting critical point . in this case
the ginzburg number does not depend on the film thickness .
for @xmath44 we employ the interface potential at a single wall , and find for the validity of the mean field description : @xmath127
modeling polymeric composites from the chemical details of the macromolecular repeat units to the morphology of the phase separated blend within a single model is not feasible today even with state - of - the - art supercomputers . yet
, there is ample evidence that by a careful choice of simulation and analysis techniques , coarse grained models of flexible polymers like the bond fluctuation model@xcite provide useful insights into the universal polymeric features . in the framework of the bond fluctuation model
each effective monomer blocks a cube of 8 neighboring sites from further occupancy on a simple cubic lattice in three dimensions .
effective monomers are connected by bond vectors of length @xmath128 , or @xmath129 in units of the lattice spacing .
the bond vectors are chosen such that the excluded volume condition guarantees that chains do not cross during their motion.@xcite each effective bond represents a group of @xmath130 subsequent @xmath131-bonds along the backbone of the chain.@xcite hence , the chain length @xmath132 employed in the present simulations corresponds to a degree of polymerization of @xmath133 in a real polymer . if we increased the chain length @xmath18 , the mean field theories would yield a better description of the equilibrium thermodynamics ( self consistent field theory is believed to be quantitatively accurate in the limit @xmath134 ) but the length scale of the ordering phenomena would be larger . hence , our choice of @xmath18 is a compromise determined by the computational resources .
the statistical segment length @xmath14 in the relation for the radius of gyration @xmath135 is @xmath136 ( i.e. , @xmath137 for @xmath132 ) .
we study thin films of geometry @xmath138 .
periodic boundary conditions are applied in the two lateral directions , while there are hard impenetrable walls at @xmath139 and @xmath140 modeling a film of thickness @xmath0 .
the average number density in the film is @xmath141 , i.e. , half of the lattice sites are occupied by corners of monomers .
this density corresponds to a melt or concentrated solution .
the density profile of occupied lattice sites , normalized by the bulk value , is presented in fig.[fig : dens ] for film thickness @xmath142 and @xmath143 . for this choice of temperature and monomer wall interaction an interface
is stabilized in the center of the film . due to the extended shape of the monomers and the compressibility of the fluid there are packing effects at the walls.@xcite overall the walls are repulsive and the monomer density is slightly reduced in the boundary region
. the spatial extension of this region is independent of the film thickness .
moreover , the density is reduced at the center of the interface as to reduce the energetically unfavorable contacts between unlike species.@xcite both effects are not incorporated into the mean field calculations@xcite and cause the density in the `` bulk''like region of the film to be slightly larger for thinner films than for thicker ones . in the following
we employ the density of occupied lattice sites in the layers @xmath144 as a measure of the density of the film . for large @xmath0
the data are compatible with a behavior of the form @xmath145 .
the film thickness ranges from @xmath146 to @xmath147 and we vary the lateral extension over a wide range @xmath148 to analyze finite size effects . in the two layers nearest to the walls , monomers experience a monomer wall interaction .
an @xmath1monomer is attracted by the left wall and repelled by the right wall , the interaction between @xmath2monomers and the walls is exactly opposite .
each monomer wall interaction changes the energy by an amount @xmath149 in units of @xmath150 . for these parameters
the wetting transition and the phase diagram of a blend confined between symmetric walls has been investigated previously.@xcite binary interactions between monomers are catered for by a short ranged square well potential @xmath151 which is extended up to a distance @xmath152 .
the phase separation is brought about by the repulsion between the unlike species .
the flory - huggins parameter is @xmath153 where @xmath154 denotes the effective coordination number in the bulk@xcite at @xmath141 . for @xmath149
previous simulations find a strong first order wetting transition at @xmath155.@xcite this value corresponds to @xmath156 , which is well inside the strong segregation limit .
the polymer conformations are updated via a combination of random monomer displacements and slithering snake like movements .
the latter relax the chain conformations about a factor of @xmath18 faster than the local displacements.@xcite we work in the semi
grandcanonical ensemble,@xcite i.e. , we control the temperature @xmath157 and the exchange potential @xmath158 between the two species , and the concentration fluctuates .
this semi
grandcanonical ensemble is realized in the monte carlo simulations via switching the polymer identity @xmath159 at fixed chain conformation .
the different monte carlo moves are applied in the ratio : slithering snake : local displacements : semi grandcanonical identity switches = 12:4:1 . during production runs , we record all 150 slithering snake steps the composition , the energy , and the surface energy and obtained the joint probability distribution in form of a histogram .
we use the semi grandcanonical identity switches in junction with a reweighting scheme,@xcite i.e. , we add to the hamiltonian of the system a reweighting function @xmath160 , which depends only on the overall composition @xmath161 .
@xmath162 and @xmath163 denote the number of @xmath1 and @xmath2 polymers in the simulation cell , respectively .
the choice @xmath164 , where @xmath165 denotes the probability distribution of the composition in the semi grandcanonical ensemble , encourages the system to explore configurations in which both phases coexist in the simulation cell .
otherwise these configurations would be severely suppressed due to the free energy cost of interfaces . in the framework of this reweighting scheme the system `` tunnels '' often from one phase to the other and
this allows us to locate the phase coexistence accurately and measure the free energy of the mixture as a function of the composition @xmath166 .
use of histogram extrapolation technique@xcite permits histograms obtained at one set of model parameters to be reweighted to yield estimates appropriate to another set of model parameters .
we employ this analysis technique to obtain estimates for the reweighting function @xmath167 .
firstly , we locate the critical points of the phase diagrams . for very small film thickness
we find a second order localisation
delocalisation transition even though the wetting transition is of first order .
et al _ have predicted this behavior in the framework of a square gradient theory@xcite and such a behavior is also born out in our self consistent field calculations for polymer blends@xcite and simulations of the ising model.@xcite upon increasing the film thickness we encounter a nearly tricritical transition .
a truly tricritical transition can not be achieved by tuning the film thickness only , because of the discreteness of the lattice , but it could be brought about by varying the monomer wall interaction . in an experiment using real materials , of course , the film thickness can be varied continously , and a truly tricritical transition is in principle always accessible . for even larger film thickness
the interface localisation
delocalisation is first order and we find two critical points at @xmath168 . secondly , we locate the triple line for the two largest values of the film thicknesses and discuss how capillary waves lead to a strong dependence of the effective interface potential on the lateral system size .
thirdly , we detail our results on the thickness dependence of the phase diagram and relate our findings to the binodals of the bulk and the mixture confined into a film with symmetric boundaries . for film thicknesses which are comparable to the radius of gyration of the molecules ,
the effective interface potential originating from the two surfaces strongly interfere .
this might change the order of the interface localisation
delocalisation transition from first to second . in this case ,
a single critical point occurs on the symmetry axis @xmath3 of the phase diagram .
the transition is thought to belong to the 2d ising universality class . in fig.[fig:2di ] ( * a * ) we present the probability distribution of the composition for various inverse temperatures @xmath171 , film thickness @xmath172 and lateral film extension @xmath173 . upon increasing the monomer
monomer interaction @xmath171 the probability distribution @xmath165 changes from single peaked to bimodal , which indicates that a phase transition occurs in this temperature range .
no signature of the trimodal distribution occurs and , hence , we conclude that the system is far away from the tricritical point , i.e. , @xmath174 . in this case
, we expect a crossover from 2dmf to 2di behavior . along the coexistence curve @xmath175 and its extension to higher temperatures we use the cumulant intersection method to locate the critical point.@xcite in the vicinity of the critical point the probability distribution of the order parameter @xmath176 scales to leading order like:@xcite @xmath177 where @xmath178 denotes the distance from the critical point along the coexistence curve and @xmath179 and @xmath180 are the critical exponents of the order parameter and the correlation length .
@xmath181 is characteristic of the universality class and has been obtained from simulations of the ising model@xcite at the critical temperature @xmath106 .
cumulants of the form @xmath182 are expected to exhibit a common intersection point for different system sizes @xmath102 at the critical temperature.@xcite the value of the cumulant at the intersection point is universal .
our simulation data are presented in panel ( * b * ) and exhibit some corrections to scaling due to the crossover 2dmf to 2di behavior .
similar corrections were observed in simulations of a second order interface localization / delocalization transition in the ising model.@xcite from the intersection points of neighboring system sizes and from the intersection of the cumulant with the universal value of the ising model we estimate the critical temperature to be @xmath183 . in the inset of fig.[fig:2di](*b * ) we show the probability distribution normalized to unit variance and norm at our estimate of the critical temperature @xmath184 , and compare the distribution to universal scaling curve of the 2d ising universality class .
the probability distributions for the smaller system sizes are slightly broader than the universal scaling curve , but the deviations decrease as we increase the system size .
the simulation data for @xmath185 are presented in fig.[fig:2di](*c * ) and ( * d * ) .
as we lower the temperature the probability distribution of the composition for @xmath186 changes from single peaked to bimodal . at intermediate values of @xmath171 , however , a three peak structure is clearly discernable .
this is characteristic of the 2dt regime and indicates the vicinity of the tricritical point . in the phenomenological considerations this regime occurs only for @xmath187 .
we note that the distribution for that small lateral system sizes resembles at no value of @xmath171 the universal shape of the order parameter distribution of the 2d ising model .
we conclude that the finite size rounding for this lateral system size sets in before we observe the crossover from 2dt to 2di behavior , i.e. , the correlation length @xmath188 in eq ( [ eqn : xl ] ) exceeds the lateral systems size @xmath102 . for such small lateral extensions
the universal properties of the transition are completely masked .
larger system sizes and a careful finite size scaling analysis is indispensable to determine the type of transition and accurately locate the transition temperature . the temperature dependence of the cumulant is presented in fig.[fig:2di](*d * ) .
there is no unique intersection point and the value of the cumulants at the crossing is larger than the universal value of the cumulant of the ising class .
this behavior indicates pronounced corrections to scaling due to the crossover from 2dt behavior away from the critical point to 2di behavior at the critical point . from the intersection points of neighboring system sizes and from the intersection of the cumulant with the universal value of the 2d ising model we estimate the critical temperature to be @xmath189 .
the inset of panel ( * d * ) compares the distribution of the order parameter at our estimate of the critical temperature and the ising scaling function .
as we increase the lateral system size the `` third '' peak in the distribution vanishes and @xmath165 gradually approaches the universal scaling curve .
this indicates that our largest system sizes exceed the correlation length at the crossover from 2dt to 2di behavior .
the comparison of @xmath165 with the universal scaling curve for several system sizes accurately locates the critical point and gives evidence that the transition belongs to the 2d ising universality class . for @xmath190
we find a single interface localisation
delocalisation transition of second order at @xmath3 .
the three
peak structure in the probability distribution for @xmath185 and small lateral extensions @xmath102 has indicated the vicinity of the tricritical interface localisation delocalisation transition . increasing the film thickness we need larger and larger lateral extensions to observe the 2di behavior as @xmath188 diverges .
right at the tricritical point the distribution of the composition is expected to exhibit a three peak structure for all lateral system sizes and the distribution , when scaled to unit variance and norm , coincides with a universal scaling function . wilding and
nielaba@xcite have obtained this scaling function via simulations at the tricritical point of the spin-1 blume
capel model@xcite in two dimensions .
assuming that the tricritical interface localisation
delocalisation transition belongs to the same universality class , we vary the film thickness @xmath0 and the interaction strength @xmath171 as to match the probability distribution of the composition onto the predetermined scaling function of the tricritical universality class . this strategy largely facilitates the search of the tricritical interface localisation
delocalisation transition .
fig.[fig : tri](*a * ) displays the probability distribution of the composition for film thicknesses ranging from @xmath185 to @xmath192 and the universal scaling curve .
the temperature was adjusted for each film thickness such that the relative heights of the central and outer peaks correspond to the ratio of the universal scaling curve . for small @xmath193
the `` valley '' between the peaks is too shallow and for @xmath194 the probability between the peaks is too small .
for @xmath195 this situation corresponds to the triple point ( cf .
below ) and the probability of finding a system between the peaks is suppressed by the free energy cost of the interface between the phases with composition close to @xmath196 or @xmath197 and the `` soft mode '' phase with composition @xmath3 . as we increase the film thickness the temperature at which the ratio between the peak height equals @xmath198 shifts towards lower temperatures and approaches the wetting transition temperature from above .
panel ( * a * ) of fig.[fig : tri ] suggests that the tricritical transition occurs close to the film thickness @xmath199 .
this is further corroborated in fig.[fig : tri](*b * ) , where we show the distribution function at @xmath200 for various system sizes . within the statistical accuracy of our data the distribution functions for the larger systems sizes collapse well onto the universal scaling curve . for smaller systems the outer peaks are slightly sharper and centered at smaller values of the order parameter .
of course , no perfect data collapse can be expected because we can tune the film thickness only in units of the lattice spacing . in view of the statistical accuracy and possible systematic corrections to scaling ,
however , we did not attempt to vary the monomer
wall interaction @xmath201 as to achieve a better collapse . for @xmath199
the system is very close to the tricritical transition .
though the system is strictly symmetric the critical points for larger film thickness ( @xmath194 ) do not occur at @xmath3 but rather there are two critical points at critical compositions @xmath204 and @xmath205 .
these critical points are the finite film thickness analogs of the prewetting critical points , which occur in the limit @xmath206.@xcite below the critical temperature the phase diagram comprises two miscibility gaps .
the coexisting phases correspond to surfaces with a thin and a thick enrichment layer of the preferred component . due to
the missing symmetry between the coexisting phases the coexistence value of the chemical potential @xmath110 differs from zero .
we determine @xmath110 via the equal weight rule,@xcite i.e. , we adjust @xmath158 such that @xmath207 along this coexistence curve and its finite size extension to higher temperatures we use the cumulant intersection to locate the critical temperature .
this is shown in fig.[fig : tcd48](*a * ) for the film thickness @xmath142 .
for the system sizes accessible in the simulations the intersection points between cumulants of neighboring systems sizes systematically shift to lower temperatures and the value of the cumulant at the intersection point gradually approaches the value of the 2d ising universality class from above .
the latter is indicated in the figure by the horizontal line . from these data we estimate the critical parameters to be @xmath208 , and @xmath209 and @xmath210 respectively .
this corresponds to a critical thickness @xmath211 of the enrichment layer .
a similar procedure has been employed to locate the critical temperature in the film of thickness @xmath212 .
the temperature and system size dependence of the cumulants are displayed in fig.[fig : tcd48](*c * ) . from this
we extract the estimate @xmath213 for the critical temperature and @xmath214 and @xmath215 for the critical compositions .
this value corresponds to a distance between then wall and the interface of @xmath216 . since increasing the film thickness from @xmath217 to @xmath218
does not change @xmath219 or @xmath220 substantially , we are in the regime @xmath221 and the critical behavior is characteristic of the prewetting critical point in the semi infinite system . the behavior of the cumulants and the very gradual approach of the probability distribution towards the ising curve indicate pronounced corrections to scaling . for the simulation of the bulk phase diagram@xcite
a nice cumulant intersection has been obtained with system sizes in the range @xmath222 to @xmath223 . in the present study
we employ systems with about an order of magnitude more polymers and obtain no clear intersection of the cumulants !
there are three reasons for strong corrections to the leading 2d ising scaling behavior : ( i ) the aspect ratio @xmath224 of our simulation cell is always finite .
truly two dimensional behavior can only be observed for vanishing aspect ratio , and our data might fall into the broad crossover region between three dimensional critical behavior and the two dimensional critical behavior .
such a crossover has been studied in our polymer model for neutral walls@xcite and walls , which attract both the same species ( i.e. , capillary condensation).@xcite however , we note that unlike these situations there is no three dimensional critical point in the vicinity for antisymmetric boundary conditions .
the temperature of the unmixing transition in the bulk is a factor 4 higher than the critical point in a thin film . since the critical point in a thin film
is related to the prewetting transition of the semi infinite system , i.e. , a transition with no three dimensional analogy , we expect the corrections to be qualitatively different from the case of neutral or symmetric boundaries .
( ii ) unlike the situation for small film thickness @xmath185 the probability distribution of the order parameter is asymmetric , because the critical point does not lay on the symmetry axis of the phase diagram .
this missing symmetry between the two phases gives rise to field mixing effects,@xcite which manifest themselves in corrections of relative order @xmath225 .
these corrections are antisymmetric to leading order and , hence , are not expected to influence even moments ( like the cumulants ) of the order parameter distribution profoundly .
the effects are , however , detectable in the order parameter distribution which we present in fig.[fig : tcd48](*b * ) and ( * d * ) .
the distribution functions at our estimate of the critical temperature clearly lack symmetry and approach very gradually the symmetric scaling curve of the 2d ising universality class .
( iii ) additionally , there are corrections to scaling by non singular background terms .
one source of ( non critical ) composition fluctuations are `` bulk like '' fluctuations in the @xmath1rich and @xmath2rich domains . in a bulk system ,
i.e. , with periodic boundary conditions in all directions , the susceptibility is rather small . at @xmath226
it takes the value @xmath227 with @xmath228 . in a system of size @xmath229
this susceptibility corresponds to composition fluctuations of the order @xmath230 .
therefore , we believe that `` bulk like '' composition fluctuations are not the major source of background terms . however , we can not rule out that the presence of an @xmath8 interface gives rise to enhanced composition fluctuations .
another source of corrections to scaling stems from the fluctuations in the average interface position itself .
since the effective interaction between the interface and the wall is rather weak , they give rise to a finite but large susceptibility away from the critical point .
we have estimated the susceptibility from the curvature of @xmath231 close to the triple point ( i.e. , @xmath232 ) , and we have obtained values of the order @xmath233 ( and a smaller value is obtained if the interface is close to a wall . ) for the same system size as above , this yields composition fluctuations of the order @xmath234 ( a value which should be compared to @xmath235 ) .
this observation partially rationalizes why the peak in the probability distribution of the composition close to @xmath3 is always broader than the peak which corresponds to the phase in which the interface is close to the wall .
as we approach the critical temperature composition fluctuations grow . at the critical point
the typical composition fluctuations are of the order @xmath236 , where we have used the critical exponents for the susceptibility @xmath237 and the correlation length @xmath238 appropriate for the 2d ising universality class .
hence , for small system sizes typical fluctuations yield compositions which differ substantially from the critical composition ; only for very large sizes the composition fluctuates in the vicinity of the critical value .
moreover , the critical density is much displaced from the symmetry axis @xmath3 and typical fluctuations in a finite system are cut - off by the constraint @xmath239 or @xmath240 . therefore , the susceptibility of a small system is reduced compared to the value expected from the leading scaling behavior .
this observation is in accord with our monte carlo data , and a similar reasoning has been used by bruce and wilding@xcite in discussing background terms to the specific heat and the concomitant corrections to scaling in the energy distribution . for the largest two film thicknesses @xmath142 and
@xmath212 the interface localisation
delocalisation transition is first order and the concomitant two miscibility gaps join in a triple point . at this temperature
an @xmath1-rich phase , a @xmath2rich phase and a phase where the interface is located in the middle of the film ( @xmath3 ) coexist .
the coexisting phases correspond to three peaks in the distribution of the composition . upon increasing the lateral system size
the peak positions do not shift ( as opposed to the behavior at the tricritical point ) , the peaks become more pronounced and configurations with intermediate compositions are more and more suppressed , because of the presence of interfaces between the coexisting phases .
the composition of the system and the average interface position are related via @xmath241 ( integral criterium ) , where we assume that the coexisting bulk phases are almost pure , i.e. , @xmath242 or @xmath197 . from the probability distribution
we then calculate the effective interface potential @xmath243 : @xmath244 in principle , not only fluctuations of the interface position @xmath245 but also `` bulk''like fluctuations contribute to composition fluctuations @xmath246 . since the wetting transition in a binary polymer blend occurs far below the critical point of the bulk , the bulk susceptibility is very small , and the latter contribution can be neglected .
the dependence of the free energy per unit area on the position of the interface is a key ingredient into the theory of wetting.@xcite the interface interacts with the boundaries and the ( bare ) interface potential exhibits three minima .
these correspond to the three coexisting phases .
in the two phases with @xmath166 close to @xmath196 and @xmath197 , the interface is localized close to the wall , the interaction between the wall and the interface is rather strong , and the effective interface potential possesses a deep minimum . in the `` soft mode ''
phase the interface is only weakly bound to the center of the film and the minimum is much broader . in fig.[fig : rend ] we present the effective interface potentials for film thicknesses @xmath142 ( * a * ) and @xmath212 ( * b * ) and various lateral system sizes in the vicinity of the triple temperature .
the three minima are clearly visible , however , the shape of the interface potential and the value of the minima depend on the lateral system size @xmath102
. moreover , the minima which correspond to the localized states broaden and ( slightly ) shift to larger distances between wall and interface upon increasing @xmath102 ( cf .
inset ) .
fluctuations of the local interface position , i.e. , capillary waves , lead to a renormalization of the effective interface potential @xmath243 and cause the dependence of @xmath243 on the lateral system size , which we observed in a microscopic model of a polymer mixture . describing the configuration of the system only via the local position @xmath247 of the interface ( sharp kink approximation ) we write
the coarse grained free energy in form of the capillary wave hamiltonian:@xcite @xmath248 = \int { \rm d}^2x\ ; \left\ { \frac{\sigma}{2 } ( \nabla l)^2 + g(l ) \right\ } \label{eqn : eih}\ ] ] where @xmath17 approaches the @xmath8 interface tension between the coexisting bulk phases for large separations between the wall and the interface .
an increase of @xmath17 at smaller distances @xmath10 as revealed by previous mc simulations is neglected.@xcite in the vicinity of a minimum of @xmath243 we may approximate the interface potential by a parabola .
@xmath249 @xmath250 denotes the deviation of the local interface position from the position where the @xmath243 attains its minimum .
@xmath251 is the parallel correlation length of interface fluctuations . for lateral distances much smaller than @xmath75
the fluctuations of the local interface position are hardly perturbed by the interaction between the interface and the wall ; the interface behaves like a free interface . for lateral distances which exceed @xmath75
capillary waves are strongly suppressed .
@xmath75 is larger for the minimum of @xmath243 in the center of the film than for the minima , in which the interface is localized at a wall . from the curvature of the effective interface potential @xmath243 for film thickness
@xmath142 we estimate @xmath252 and @xmath253 , where we have used the bulk value @xmath254 for the interfacial tension at @xmath255 . for the thicker film
we obtain @xmath256 , but the curvature in the middle of the film could not be accurately estimated .
the value is of the order @xmath257 , and we expect this value to decrease exponentially with the film thickness . hence , this fluctuation effect is the stronger the larger the film thickness . for the system sizes employed in the mc simulations @xmath258
is of order unity . in our monte carlo simulations
the finite lateral system size @xmath102 acts as an additional cut off for the spectrum of interface fluctuations@xcite and upon increasing @xmath102 we extend the spectrum of interface fluctuations . allowing for interface fluctuations we decrease the free energy of the system .
therefore , we expect the free energy density of the system to decrease when we increase the lateral system size , and we expect the effect to be the stronger the larger @xmath75 . therefore , the free energy of the `` soft mode '' phase becomes smaller compared to the free energy of the phase , where the interface is located close to a wall when we increase @xmath102 .
this effect is clearly observed in the mc simulations . to be more quantitative ,
we consider a system where the laterally averaged interface position is at the minimum of @xmath243 , and we expand the deviation @xmath259 from the minimum in a fourier series @xmath260 with @xmath261 .
the coefficients @xmath262 vanish identically , all other coefficients can take any real value . using this expansion ( [ eqn : exp ] ) and the effective interface hamiltonian ( [ eqn : eih ] ) , we calculate the average size of fluctuations @xmath263^{-1}\ ] ] and the free energy @xmath264 \exp\left(\frac{-{\cal h}[l]}{k_bt}\right ) \nonumber \\ & = & { \rm const } + \frac{2}{l^2 } \sum_{nm=0}^\infty \eta_{nm } \ln \left\{\frac{\sigma}{k_bt } \left [ k_\|^2 + \left(\frac{2\pi}{l}\right)^2(n^2+m^2 ) \right ] \right\ } \end{aligned}\ ] ] where the factor @xmath265 takes the values @xmath266 , @xmath267 and @xmath268 for @xmath269 and @xmath270 in order to account for the restriction on the coefficients @xmath271 .
the additive constant is independent of the wavevector cut - off @xmath272 .
the dependence of the free energy on the system size is dominated by the small @xmath273 behavior . in this regime
the discrete nature of the wavevector space matters and , hence , we do not replace the sum over @xmath273 by integrals . using the measured values of the wavevector cut - offs
we calculate the lateral system size dependence of the free energy difference between the `` soft mode''phase and the delocalized state .
the results are compared to the mc data in fig.[fig : d24 ] .
good agreement is found for large @xmath102 , whereas there are deviations for smaller @xmath102 . for small @xmath102
the amplitude of the fluctuations becomes large and the a parabolic interface potential is no longer a good approximation especially for the localized state where the interface is located very close to the walls .
we have used histogram extrapolation to adjust the temperature such that the difference @xmath274 of the minima vanishes .
this corresponds to the equal height criterium for the triple point .
the equal weight condition , which we have applied to determine the binodals close to the critical points , would require : @xmath275 .
both conditions agree , of course , when we extrapolate our results to @xmath276 . from this procedure
we obtain the following estimates for the triple point : @xmath277 and @xmath278 for @xmath142 and @xmath279 and @xmath280 for @xmath212 . the thickness of the microscopic enrichment layer at the wetting transition temperature is of the order @xmath281 ; a value which is consistent with expectation for strong first order wetting transitions .
the dependence of the critical temperature and the triple temperature on the film thickness is summarized in fig.[fig : tc ] .
when we increase the film thickness the critical temperature @xmath282 shows a non monotonic dependence . at @xmath199 the tricritical point ( where the critical temperature and the triple temperature merge ) occurs at @xmath283 , for film thickness @xmath142 we find @xmath284 and at @xmath212 @xmath213
this effect is rooted in two opposing effects . on the one hand ,
the self consistent field calculations predict the flory
huggins parameter @xmath285 to decrease upon increasing the film thickness @xmath0 for an incompressible fluid .
this shift in temperature decreases exponentially with the film thickness .
one the other hand , packing effects , which are not incorporated in the self consistent field calculations , increase the density in the `` bulk''like portion of the film when we decrease the film thickness .
these packing effects at the walls depend strongly on the computational model , but qualitatively similar effects might occur in experimental systems as well .
this thickness dependence of the density in the middle of the film modifies the relation between the depth of the square well potential and the @xmath286parameter .
this leads to a behavior of the form @xmath287 , where we use the dependence of the density profile ( cf .
fig.[fig : dens ] ) on the film thickness as obtained by direct measurement in the monte carlo simulations .
a dependence of the fluid packing structure on the density is neglected . a similar @xmath288 correction to the difference in surface free energies between the @xmath1 and
the @xmath2rich phase has been observed in previous simulations.@xcite . attempting to separate these two effects we also present @xmath289^{-1}$ ] which corresponds to the inverse flory
huggins parameter . within
the error bars the behavior of this quantity is consistent with the mean field prediction .
the critical value of the inverse flory
huggins parameter increases and the triple value decreases as we increase the film thickness .
the latter approaches the wetting transition temperature@xcite @xmath290 from above . for film thickness
@xmath212 we have determined the complete phase diagram .
close to the critical point we assume 2di behavior with an exponent @xmath291 for the order parameter and employ finite size scaling to estimate the critical amplitude . outside the critical region but above the triple temperature we have estimated the location of the binodals via the equal weight criterium in a system of size @xmath292 ; but no finite size analysis has been applied .
the phase diagram for a blend confined into a film with antisymmetric walls is presented in fig.[fig : phasen ] .
( * a * ) confinement into a film with antisymmetric boundary conditions enlarges the one phase region up to the prewetting critical temperature .
since the wetting transition in binary polymer blends occurs far below the unmixing critical temperature in the bulk the effect is quite pronounced .
the temperature region between the prewetting critical point and the triple point is about @xmath293 of the wetting transition temperature .
this value strongly depends on the details of the structure at the walls .
the stronger the wetting transition the larger are the prewetting lines and the more extended is the region of the two miscibility gaps .
the phase diagram of the bulk and a film with symmetric walls are displayed for comparison in fig.[fig : phasen ] .
the symmetric film has the same thickness as the antisymmetric film and the monomer wall interactions at one wall are identical and attract the @xmath1component . while the prewetting at the wall which prefers the @xmath1component leads to a two phase region in the antisymmetric case
there is only a change in curvature of the binodal detectable in the symmetric case .
panel ( * b * ) of fig.[fig : phasen ] presents the phase diagram as a function of temperature and exchange chemical potential . in the antisymmetric case
@xmath294 up to the triple temperature .
there , two coexistence lines emerge which are the thin film analogies of the prewetting lines at the two walls .
since the monomer wall interactions are short ranged the prewetting line in the bulk and the coexistence curves in the film deviate from the bulk coexistence value linearly ( up to logarithmic corrections).@xcite they end in two critical points . though the system is strictly symmetric with respect to exchanging @xmath159 phase coexistence is not restricted to @xmath175 , and the coexisting phases are not related by the symmetry of the hamiltonian . the coexistence curve of the symmetric film is shown for comparison . the coexistence value of the chemical potential is shifted to values disfavoring the component attracted by both walls . there is a change in the temperature dependence of the coexistence curve close to the wetting transition temperature , but the coexistence curve stays far away from the prewetting line .
if the two lines intersected there would also be a triple point in the symmetric case.@xcite since , the shift of the chemical potential @xmath158 is roughly proportional to the inverse film thickness ( kelvin equation ) we expect a triple point to occur only for much larger film thicknesses .
this is in accord with self consistent field calculations.@xcite the typical distance @xmath10 between the interface and the wall at coexistence is of order @xmath295 in the antisymmetric case , while it is only of the order @xmath296 in the symmetric case .
hence , smaller film thicknesses are sufficient to study the interaction between the interface and the wall , and antisymmetric boundary conditions are computationally more efficient to investigate the wetting behavior .
we have studied the phase diagram of a symmetric polymer mixture in a thin film with antisymmetric boundary conditions via large scale monte carlo simulations .
the walls interact with monomers via a short range potential ; the one wall attracts the @xmath1 component and repels the @xmath2 component while the interaction at the opposite wall is exactly reversed .
the salient features of the phase diagram and its dependence on the film thickness as obtained by our mc simulations are in accord with the results of mean field theory.@xcite fluctuations , which are neglected in the mean field calculations , do not modify the qualitative phase behavior .
however , they give rise to a rich crossover behavior between ising critical behavior , tricritical behavior and their mean field counterparts .
this has been elucidated by phenomenological considerations and is qualitatively consistent with our simulation results . since the critical point of the thin binary polymer film occurs at much lower temperature than the unmixing transition in the bulk , `` bulk like '' composition fluctuations are only of minor importance .
the dominant fluctuations of the composition of the film arise from capillary waves at the interface between the @xmath1rich and @xmath2rich regions in the film .
the interaction between the walls and the interface is rather small , because it is mediated via the distortion of the interface profiles at the walls and the strength of the interaction decreases exponentially with the distance .
hence , the interface is only very weakly bound to the minimum of the effective interface potential .
these large fluctuations give rise to rather pronounced corrections to scaling in our systems of limited size .
however , using the cumulant intersection method@xcite and the matching of the order parameter distribution onto the predetermined universal scaling function,@xcite we give evidence for the 2d ising universal character of the critical points .
the same strategy has proven computationally very convenient to locate the tricritical point as a function of the film thickness.@xcite this technique allows us to locate the critical points of the confined complex fluid mixture with an accuracy of a few percent .
interface fluctuation do not only impart 2d ising critical behavior onto the critical points , but they are important in the whole temperature range .
monitoring the probability distribution of the laterally averaged interface position we extract the effective interface potential @xmath243 .
its dependence on the lateral system size yields direct evidence for the renormalization of the interface potential by interface fluctuations .
interface fluctuations lead to a broadening of the minima in the interface potential , a shift of the minima towards the center of the film , and to a relative reduction of the free energy of the broader minimum .
this leads to a systematic overestimation of the triple temperature by the mean field calculations .
moreover , our simulations indicate that packing effects in thin films result in corrections of the order @xmath288 to the density of the film or to the effective flory
huggins parameter .
such corrections are likely to mask completely the subtle thickness dependence of the triple temperature and the triple temperature predicted by the mean field calculations . for short range interactions between walls and monomers
the predicted shifts decrease exponentially with the film thickness @xmath0 . however , power law dependencies are expected for the case of long range ( i.e. , van der waals ) interactions between walls and monomers .
the gross features of the phase diagram as well as our simulation and analysis techniques are not restricted to binary polymer fluids but generally apply to binary liquid mixtures in confined geometries . moreover , mean field calculations@xcite indicate that for small deviation from perfectly antisymmetric boundary conditions a qualitatively similar phase behavior emerges .
the stronger the first order wetting transitions at the boundaries , the larger deviations from antisymmetry are permissible without alternating the topology of the phase diagram .
hence , a thin binary film on a substrate against air / vacuum , where the substrate energetically favors one component of the mixture while the other component has an affinity to the air surface , is an experimental realization of the boundary conditions discussed here .
our findings also imply that ultrathin enrichment layers at one surface are unstable in the temperature range @xmath297 .
such effects have been observed experimentally@xcite in polymeric films , although for a liquid vapor transition instead of a liquid liquid demixing .
however , recent experiments have observed the wetting transition in binary polymer blends.@xcite it is a great pleasure to thank n.b .
wilding for stimulating discussions and for providing the universal probability distributions of the order parameter at the 2d ising critical and 2d tricritical point .
we have benefitted also from discussions with e.v .
albano and a. de virgiliis . financial support by the dfg under grant bi 314/17 in the priority program `` wetting and structure formation at interfaces '' and the daad / proalar2000 as well as generous grants of computing time at the nic jlich , the hlr stuttgart and the computing center in mainz are gratefully acknowledged .
99 r. evans , j.phys.cond.matter * 2 * , 8989 ( 1990 ) .
a.o . parry , j.phys.condens.matter * 8 * , 10761 ( 1996 ) .
s. dietrich , phase transitions and critical phenomena , vol 12 , c. domb and j. lebowitz ( eds ) academic press , london ( 1988 ) .
de gennes , wetting statics and dynamics , rev.mod.phys . * 57 * , 827 ( 1985 ) .
a. budkowski , adv .
polymer sci . * 148 * , 1 ( 1999 ) .
k. binder , adv .
polymer sci . * 138 * , 1 ( 1999 ) .
fisher and h. nakanishi , j.chem.phys . *
75 * , 5857 ( 1981 ) .
h. nakanishi and m.e .
fisher , j.chem.phys .
* 78 * , 3279 ( 1983 ) .
f. brochard wyart and p .-
g . de gennes , acad.sci .
paris * 297 * , 223 ( 1983 ) .
a.o . parry and r. evans , phys.rev.lett . * 64 * , 439 ( 1990 ) , physica * a 181 * , 250 ( 1992 ) .
swift , a.l .
owczarek , and j.o .
indekeu , europhys.lett .
* 14 * , 475 ( 1991 ) .
k. binder , d.p .
landau , and a.m. ferrenberg , phys.rev.lett . * 74 * , 298 ( 1995 ) ; phys.rev . *
e 51 * , 2823 ( 1995 ) .
albano , k. binder , d.w .
heermann , and w. paul , surf.sci .
* 233 * , 151 ( 1989 ) .
indekeu , a.l .
owczarek , and m.r .
swift , phys.rev.lett . *
66 * , 2174 ( 1991 ) .
a.o . parry and r. evans , phys.rev.lett . *
66 * , 2175 ( 1991 ) .
j. rogiers and j.o .
indekeu , europhys.lett .
* 24 * , 21 ( 1993 ) .
e. carlon and a. drzewinski , phys.rev.lett .
* 79 * , 1591 ( 1997 ) .
k. binder , r. evans , d.p .
landau , and a.m. ferrenberg , phys.rev . *
e 53 * , 5023 ( 1996 ) .
a. werner , f. schmid , m. mller , and k. binder , j.chem.phys . * 107 * , 8175 ( 1997 ) .
t. kerle , j. klein , and k. binder , phys.rev.lett .
* 77 * , 1318 ( 1996 ) , europ .
j. b*7 * , 401 ( 1999 ) .
m. sferrazza , m. heppenstall butler , r. cubitt , d. bucknall , j. webster , r.a.l .
jones , phys.rev.lett . *
81 * , 5173 ( 1998 ) ; m. sferrazza , c. xiao , r.a.l .
jones , g.d .
bucknall , j. webster , j. penfold , phys.rev.lett . * 78 * , 3693 ( 1997 ) .
m. mller , k. binder , and e.v .
albano , europhys.lett . * 49 * , 724 ( 2000 ) ; m. mller , k. binder , and e.v .
albano , physica * a 279 * , 188 ( 2000 ) ; m. mller , e.v .
albano , and k. binder , phys.rev . * e * ( in press ) .
a.m. ferrenberg , d.p .
landau , and k. binder , phys.rev .
* e 58 * , 3353 ( 1998 ) .
m. mller , macromol .
theory simul .
( feature article ) * 8 * , 343 ( 1999 ) .
m. mller and k. binder , macromolecules * 31 * , 8323 ( 1998 ) .
b. nienhuis , a.n .
nerker , e.k .
riedel , and m. schick , phys.rev.lett .
* 43 * , 737 ( 1979 ) .
den nijs , j.phys . * a 12 * , 1857 ( 1979 ) .
rearson , phys.rev .
* b 22 * , 2579 ( 1980 ) .
ginzburg , sov.phys.solid state * 1 * , 1824 ( 1960 ) ; p.g .
de gennes , j.phys.lett .
( paris ) * 38 * , l-441 ( 1977 ) ; j.f .
joanny , j.phys.a * 11 * , l-117 ( 1978 ) ; k. binder , phys.rev.a * 29 * , 341 ( 1984 ) .
i. carmesin and k. kremer , macromolecules * 21 * , 2819 ( 1988 )
deutsch and k. binder , j. chem .
phys . * 94 * , 2294 ( 1991 ) .
j. baschnagel _
adv.polym.sci . * 152 * , 41 ( 2000 ) .
v. tries , w. paul , j. baschnagel , and k. binder , j.chem.phys . * 106 * , 738 ( 1997 ) .
m. mller , k. binder , and w.oed , j.chem.soc .
faraday trans .
91 , 2369 ( 1995 ) .
m. mller and k. binder , macromolecules * 28 * , 1825 ( 1995 ) .
a. sariban and k. binder , j. chem .
* 86 * , 5859 ( 1987 ) .
berg and t. neuhaus , phys.rev.lett * 68 * , 9 ( 1992 ) ; j. lee , phys.rev.lett * 71 * , 221 ( 1993 ) ; e. marinari and g. parisi , europhys.lett . *
19 * , 451 ( 1992 ) .
a.m. ferrenberg and r.h .
swendsen , phys .
* 61 * , 2635 ( 1988);. ibid * 63 * , 1195 ( 1989 ) ; a. bennett , j.comp.phys * 22 * 245 ( 1972 ) .
bruce and n.b .
wilding , phys.rev.lett . *
68 * , 193 ( 1992 ) ; n.b .
wilding , phys.rev .
* e 52 * , 602 ( 1995 ) . k. binder , z.phys * b 43 * , 119 ( 1981 ) ; phys.rev.lett * 47 * , 693 ( 1981 ) ; rep.prog.phys . * 60 * , 487 ( 1997 ) .
n.b . wilding and p. nielaba , phys .
e 53 * , 926 ( 1996 ) .
m. blume , phys.rev .
* 141 * , 517 ( 1966 ) ; h.w .
capel , physica * 32 * , 966 ( 1966 ) . c. borgs and k. kotecky , j.stat.phys . *
61 * , 79 ( 1990 ) ; phys.rev.lett . *
68 * , 1734 ( 1992 ) .
deutsch and k. binder , macromolecules * 25 * , 6214 ( 1992 ) .
y. rouault , j. baschnagel , and k. binder , j.stat.phys .
* 80 * , 1009 ( 1995 ) .
bruce and n.b .
wilding , phys.rev .
* e 60 * , 3748 ( 1999 ) .
cahn , j.chem.phys .
* 66 * , 3667 ( 1977 ) .
m. schick , liquids at interfaces , les houches , session xlviii , j. charvolin , j.f .
joanny , and j. zinn - justin ( eds ) , elsevier , amsterdam ( 1990 ) .
r. lipowski , d.m .
kroll , and r.k.p .
zia , phys.rev .
* b 27 * , 499 ( 1983 ) .
e. brezin , b.i .
halperin , and s. leibler , phys.rev.lett .
* 50 * , 1387 ( 1983 ) .
d.s . fisher and d.a .
huse , phys.rev .
* b 32 * , 247 ( 1985 ) .
fisher and h. wen , phys.rev.lett . *
68 * , 3654 ( 1992 ) .
huse , w. van saarloos , and j.d .
weeks , phys.rev .
* b 32 * , 233 ( 1985 ) .
e.h . hauge and m. schick ,
* b 27 * , 4788 ( 1983 ) .
d. nicolaides and r. evans , phys.rev .
* b 39 * , 9336 ( 1989 ) .
r. evans and u. marini bettolo marconi , phys . rev . * a 32 * , 3817 ( 1985 ) .
w.zhao , m.h .
rafailovich , j. sokolov , l.j .
fetters , r. plano , m.k .
sanyal , and s.k .
sinha , phys.rev.lett .
* 70 * , 1453 ( 1993 ) .
j. rysz , a. budkowski , a. bernasik , j. klein , k. kowalski , j. jedlinski , and l.j .
fetters , europhys.lett . *
50 * , 35 ( 2000 ) .
m. geoghegan , h. ermer , g. jngst , g. krausch , and r. brenn , phys.rev .
* e 62 * , 940 ( 2000 ) . .
[ tab:1 ] compilation of the boundaries of the different regimes in the vicinity of the tricritical point and the correlation lengths at the crossover .
the latter quantity gives an estimate of the system size required to observe the crossover in the monte carlo simulations . [ cols="<,<,<",options="header " , ] | using extensive monte carlo simulations we study the phase diagram of a symmetric binary ( ab ) polymer blend confined into a thin film as a function of the film thickness @xmath0 .
the monomer wall interactions are short ranged and antisymmetric , i.e , the left wall attracts the @xmath1component of the mixture with the same strength as the right wall the @xmath2component , and give rise to a first order wetting transition in a semi
infinite geometry . the phase diagram and the crossover between different critical behaviors is explored . for large film thicknesses
we find a first order interface localisation / delocalisation transition and the phase diagram comprises two critical points , which are the finite film width analogies of the prewetting critical point .
using finite size scaling techniques we locate these critical points and present evidence of 2d ising critical behavior .
when we reduce the film width the two critical points approach the symmetry axis @xmath3 of the phase diagram and for @xmath4 we encounter a tricritical point . for even smaller film thickness the interface localisation / delocalisation transition is second order and we find a single critical point at @xmath3 .
measuring the probability distribution of the interface position we determine the effective interaction between the wall and the interface .
this effective interface potential depends on the lateral system size even away from the critical points .
its system size dependence stems from the large but finite correlation length of capillary waves .
this finding gives direct evidence for a renormalization of the interface potential by capillary waves in the framework of a microscopic model . |
the relation between the morphologies of host galaxies and the presence of nuclear activity has been investigated for many years .
one of the first studies arguing for a difference between active and inactive galaxy hosts is the one of @xcite .
these authors have claimed that seyfert galaxies hosts are more distorted than inactive galaxies , showing an excess of bars , rings and tails .
more recent studies did not confirm the excess of bars in seyferts ( e.g. @xcite ) while others have found a small excess ( e.g. @xcite ) .
a number of studies using hubble space telescope images of nearby galaxies have revealed a trend for active galaxies always showing a lot of dust structure in the nuclear region . @xcite
have found that radio - loud early - type galaxies have more dust than radio - quiet ones . @xcite and later @xcite found that seyfert galaxies almost always present dusty filaments and spirals in the nuclear region , while @xcite found that , among early hubble types , active galaxies present more dust structure than inactive galaxies .
recently , @xcite have also argued that dust in early - type galaxies is correlated with nuclear activity .
the goal of the present paper is to discuss the results of a recent study ( @xcite ) which provides a robust analysis of the relation between nuclear dust structures and activity in galaxies on the basis of optical hst images .
the particular case of ngc1097 is then discussed , for which we have obtained , in adition , kinematics of the gas associated with the nuclear dust structures @xcite which reveal streaming motions towards the nucleus .
we then use results of previous studies ( @xcite , @xcite ) to relate the mass accretion rate reaching the accretion disk to the larger scale mass flow rate derived from the observed streaming motions .
we ( @xcite ) have recently selected a sample comprising all active galaxies from the palomar survey @xcite which have optical images in the hst archive , as well as a pair - matched sample of inactive galaxies , and constructed structure maps " , a technique proposed by @xcite to enhance both absorption and emission structures in the images .
the total sample comprises 34 matched pairs of early - type galaxies ( t@xmath20 ) and 31 pairs of late - type galaxies ( t@xmath30 ) .
the results for the 34 early - type pairs of our sample are illustrated in fig . 1 for a subsample of 10 pairs
: while all ( 100% ) active galaxies present some dust structure , only 26% ( 9 of 34 ) of the inactive galaxies possess some dust .
another difference is the finding that at least 13 of the 34 ( 38% ) early - type inactive galaxies present bright _ stellar disks _ , while only one of the active galaxies show such disks . most dust structures and stellar disks
are seen within a few hundred parsecs from the nucleus .
a different result was obtained for the late - type pairs : all galaxies show circumnuclear dust regardless the presence or absence of nuclear activity , and stellar disks were found only in a couple of galaxies . the above findings imply a strong correlation between circumnuclear dust and nuclear activity , indicating that the dust is connected to the material which is currently feeding the active nucleus , and is probably tracing this material on its way to the nucleus .
the morphologies of the dust structures range from chaotic filaments to regular dusty spirals and disks , suggestive of a `` setling scenario '' for the dust , as proposed by @xcite .
our finding of nuclear stellar disks in inactive galaxies suggests that there is one further step in this evolutionary scenario : the stellar disk , which shows up in the inactive phase of the galaxy .
the evolutionary scenario can be described as follows : externally acquired matter is traced by the chaotic filamentary structure which gradually settles into more regular nuclear spirals and disks .
stars then form in the dusty disks , and when the gas and dust are fed to the black hole , nuclear activity ceases and and the stellar disks are unveiled @xcite . as stellar disks should be longer lived than the gasesous dusty disks , the stellar disks are probably present also in the active phase , but are obscured by dust . in the evolutionary scenario proposed above , the episodic accretion of matter then replenishes the disk , which grows together with the mass of the nuclear black hole . on the theoretical side , support for the evolutionary scenario includes the work of @xcite who demonstrated that , if a central smbh is present , nuclear disks of gas and dust can develop spiral shocks and generate gas inflow compatible to the accretion rates observed in local active galaxies .
on the observational side , kinematic evidence for inflow along nuclear dusty spiral arms has been found so far in one case : ngc1097 @xcite .
the observations of ngc1097 were obtained with the integral - field unit of the gemini multi - object spectrograph and allowed the maping of the gas kinematics in the inner few hundred parsecs @xcite . after subtracting a circular velocity model , streaming motions along spiral arms with inward velocities of up to 50kms@xmath1 were found .
another relevant finding on this galaxy is the young obscured starburst recently discovered @xcite very close to the nucleus ( within @xmath09pc ) , in agreement with the suggestion that inflowing gas and dust gives birth to stars in the nuclear dusty spiral or disk @xcite .
the velocity observed for the streaming motions along the nuclear spiral allows an estimate of a few myr for the gas to flow from a few hundred parsecs to the nucleus .
for an estimated gas density of @xmath0500protons@xmath4 , and estimated circular cross - section of 3 spiral arms at 100pc from the nucleus ( opening angle 20@xmath5 ) , we conclude that the mass flow rate along the nuclear spiral arms in ngc1097 is @xmath035 times the one at the accretion disk of dm / dt=1.1@xmath610@xmath7gs@xmath1 ( @xcite ) , and allows the accumulation in the nuclear region of @xmath010@xmath8m@xmath9 in a few myr , which can provide the necessary matter to give birth to the nuclear starburst @xcite .
we have found a strong correlation between circumnuclear dust and nuclear activity , and between nuclear stellar disks and absence of nuclear activity in early - type galaxies . as both the active and inactive galaxies
can be thought of as the `` same galaxy '' observed in different phases , our findings suggest an evolutionary scenario in which the nuclear activity is triggered by capture of dusty gas to the nuclear region .
the origin of the gas and dust is still not clear , but the absence of dust in the inactive phase suggests that in can not originate from continuous mass loss from stars and is most probably external . in our proposed evolutionary scenario ,
once the gas and dust are captured to the nuclear region they gradually settle into a nuclear spiral or disk , where new stars are born while the excess gas and dust is accreted by the nuclear smbh .
replenishment of a nuclear stellar disk ( observed in the inactive galaxies ) is the final product after activity ceases . as a result ,
both the stellar component of the galaxy and its smbh at the nucleus grow after each activity cycle , as implied by the m@xmath10 relation ( e.g. @xcite ) .
the evolutionary scenario is supported by recent kinematic observations of the nuclear spiral in ngc1097 , which reveal streaming motions along the spiral arms , providing an accretion flow rate which is enough to both form the starburst recent observed surrounding the nucleus @xcite in a few myr and feed the nuclear smbh . as similar dusty structures
are observed in most active galaxies ( @xcite ) , streaming motions along nuclear spirals may be the main mechanism for black hole feeding in active galaxies . | a recent analysis of hst optical images of 34 nearby early - type active galaxies and of a matched sample of 34 inactive galaxies both drawn from the palomar survey shows a clear excess of nuclear dusty structures ( filaments , spirals and disks ) in the active galaxies .
this result supports the association of the dusty structures with the material which feeds the supermassive black hole ( hereafter smbh ) . among the inactive galaxies there
is instead an excess of nuclear stellar disks .
as the active and inactive galaxies can be considered two phases of the `` same '' galaxy , the above findings and dust morphologies suggest an evolutionary scenario in which external material ( gas and dust ) is captured to the nuclear region where it settles and ends up feeding the active nucleus and replenishing the stellar disk which is hidden by the dust in the active galaxies with new stars .
this evolutionary scenario is supported by recent gas kinematics of the inner few hundred parsecs of ngc1097 , which shows streaming motions ( with velocities @xmath050kms@xmath1 ) towards the nucleus along spiral arms .
the implied large scale mass accretion rate is much larger than the one derived in previous studies for the nuclear accretion disk , but is just enough to accumulate one million solar masses over a few million years in the nuclear region , thus consistent with the recent finding of a young circumnuclear starburst of one million solar masses within 9 parsecs from the nucleus in this galaxy . |
the experimental progam in gravitational wave detection that is currently under way @xcite has brought renewed attention to the problem of obtaining high accuracy predictions for the evolution and gravitational radiation power spectra of binary star systems with neutron star or black hole constituents . the conventional approach to calculating the initial inspiral of the system as it slowly loses energy to gravitational radiation employs the post - newtonian ( pn ) approximation to general relativity . essentially , this formalism consists of systematically solving the einstein equations with non - relativistic ( nr ) sources as a power series expansion in @xmath6 , where @xmath3 is a typical three - velocity of the system under consideration @xcite . for binary systems that can be detected by ligo , which have binary constituents with masses in the range @xmath7 ,
the inspiral phase spans roughly @xmath8 , corresponding to orbital separations from about @xmath9 to a few times @xmath10 .
a detector such as ligo is particularly sensitive to the time varying phase @xmath11 of the gravitational waves emitted during the evolution of the binary system @xcite . in the pn formalism , this observable can be obtained by calculating the mechanical energy @xmath12 as well as the total power @xmath13 emitted in gravitational radiation as functions of the orbital parameters of the binary system . taking for illustration the extreme mass ratio limit and using conservation of energy , @xmath14 , for circular orbits the gravitational wave phase seen by the detector
is then given by @xmath15 where @xmath16 is the wave s frequency and @xmath17 the mass of the heavy component star . because a typical inspiral event will sweep a large number of radians of phase as it scans the detector frequency band ( e.g , for ligo , with a frequency range @xmath18 hz
, the number of cycles for a neutron star / neutron star binary is of order @xmath19 ) , to compare the predictions of gr with the experimental data requires computing the observable @xmath11 to extremely high order in the velocity expansion .
for instance , for binary neutron star inspirals in the ligo band , one must know @xmath11 to @xmath20 ( or 3pn " ) beyond the leading order quadrupole radiation results , and higher order terms in @xmath3 may be necessary for tracking the phase variation of the gravitational waves emitted by more massive objects @xcite .
several independent research groups have applied variants of the pn formalism to the computation of high order corrections to the gravitational wave observables during the binary inspiral phase @xcite .
heroic calculational efforts have yielded in the last few years the @xmath4 analytic results necessary to fully interpret the ligo data @xcite .
while the independent calculations are in agreement including terms up to @xmath21 ( or @xmath22pn " ) beyond the predictions of newtonian gravity , the @xmath4 corrections do not give well defined results . at this order
, there seems to be an ambiguity in the pn expansions for the binary system , stemming from scheme dependence in the regularization of singular integrals that arise in the perturbative series .
recently , refs .
@xcite have proposed a resolution of this ambiguity purely within the framework of the pn expansion .
the singular integrals that cause trouble at high orders in @xmath3 can be seen to arise from treating the binary star constituents as point particle sources coupled to general relativity .
it is then clear that these singularities are simply the usual ultraviolet ( uv ) divergences that appear in any field theory coupled to point sources .
their presence is merely a sign that the formalism has limited predictive power and needs to be supplemented by a more complete model that accounts for the fact that the sources are not really pointlike . while in principle one could obtain finite , unambiguous results for experimental observables in this more complete theory
, the price would be much greater computational complexity . even then
, it would be difficult to disentangle the effects that depend upon short distance physics from those that are solely a consequence of the gravitational dynamics . since the goal of the gravitational wave experiments is both to probe the internal structure of strongly gravitating astrophysical systems as well as to test the dynamics of general relativity ( or possible deviations from it )
, it is critical to have a computational approach that can cleanly separate these two aspects of the binary problem . in this paper , we propose a lagrangian framework for systematically calculating to any order in the pn expansion . within our theory the ambiguities " that plague the conventional pn techniques at order @xmath4 and beyond have a natural place in the renormalization program .
our formalism is based on effective field theory ( eft ) reasoning , and it is motivated by the eft approach to nr bound state problems in quantum field theories such as qed and qcd @xcite . rather than trying to resolve the point particle singularities by resorting to a specific model of the short distance physics , in an eft framework @xcite we systematically parametrize our ignorance of this structure by including in the effective point particle lagrangian the most general set of operators consistent with the symmetries of the well understood , long wavelength physics . in our case , this long wavelength physics is general relativity , and the symmetry that constrains the dynamics is just general coordinate invariance .
the operators in the point particle action have coefficients which encapsulate the properties of the structure of the extended sources .
uv divergences encountered in calculations in the eft can then be dealt with using any convenient regulator , and renormalized by adjusting the operator coefficients scheme ) .
this has the advantage of not introducing spurious length scales that could spoil dimensional analysis arguments within the eft .
see @xcite for a discussion of dimensional regularization in the context of conventional pn formalisms ] . given a model of stellar structure , the precise value of these coefficients can be obtained by a uv matching calculation that we describe below in sec .
[ sec : matching ] .
note that because these operators are intrinsic to the structure of the sources , the matching procedure can be carried out for a single source in isolation , independently of the complicated binary star dynamics . in this way the eft manages to disentangle the model dependent aspects of the gravitational wave signals from the features that arise as unambiguous predictions of general relativity .
an eft also has the advantage of having manifest power counting in the expansion parameter of the theory , in our case the velocity @xmath3 .
this means that in the eft it is possible to calculate , using simple scaling arguments , the order in @xmath3 which a given term in the perturbative series ( i.e , a typical feynman diagram obtained from the vertices of the effective lagrangian or an operator in the effective theory ) first contributes to a given physical observable .
in particular , one can determine ahead of any calculation the number of terms in the point particle lagrangian that will appear at any order in @xmath3 . as an example , using our formalism we will see that for binaries with non - spinning black hole or neutron star constituents , the proper time term is sufficient to describe the physics up to @xmath21 .
the apearance of logarithmic divergences in the pn expansion that first appear at order @xmath4 indicates that two new non - minimal terms whose form is dictated by general covariance must be included at this order to consistently renormalize the point particle theory .
however , the coefficients of these operators can be shifted arbitrarily by making field redefinitions of the metric , which indicates that the @xmath4 operators can not contribute to observables .
this gives a clean argument for the absence of @xmath4 finite size effects in the predictions for gravitational wave templates .
ignoring the effects of absorption @xcite .
we will show how absorptive effects are incorporated into the eft in a future paper .
] , we also find that the next set of point particle operators , including for instance operators quadratic in the riemann curvature , do not contribute until order @xmath5 .
these operators , which can not be removed by field redefinitions , are associated with the leading effects of tidal forces exerted on an isolated neutron star or black hole by an external gravitational field .
the fact that such tidal operators do not arise until @xmath5 is consistent with purely newtonian estimates which also conclude that tidal forces are @xmath5 effects . finally , an eft has the advantage that it is possible to use renormalization group ( rg ) methods to understand in a simple way logarithms of scale ratios that may arise in the perturbative series for a given observable .
for example , in the nr expansion of the binary star gravitational wave observables , one encounters logarithmic terms of the form @xmath23 where @xmath24 is the ratio of a distance scale @xmath0 that characterizes the size of the binary constituents and @xmath1 is a distance of order the orbital radius ( note that for black hole binaries , the virial theorem relates @xmath25 ) .
we will show that in our eft , the coefficients of operators in the effective point particle action undergo non - trivial rg flows , so that the logarithms of @xmath26 ( tidal logarithms ) can be obtained by rg running from a renormalization scale @xmath27 where the point particle theory is initially defined down to a scale @xmath28 where the binary dynamics takes place .
moreover , the complete structure of the rg flows is encoded in the black hole metric . as an explicit example of the rg method
, we will give a simplified derivation of the terms at @xmath20 in the two - body interaction lagrangian ( or equivalently the binary gravitational binding energy ) that are enhanced by @xmath29 .
we then show explicitly that the effect of these logarithms do not contribute to physical observables .
this result must follow since , as we already mentioned above , the @xmath4 worldline operators can be removed by redefining the metric . however , this calculation is still useful as a template for determining the logarithmically enhanced terms at higher orders in @xmath3 that stem from true finite size effects .
the outline of the paper is as follows . in the next section
we formulate our eft .
the starting point is a theory of relativistic point particles coupled to gravity .
this theory is appropriate for discussing the dynamics of extended sources at inverse momentum transfers larger than their size @xmath0 .
we then describe the general procedure for matching this theory at the mass scale . ]
@xmath28 , where the binary orbital dynamics takes place , onto an eft that is optimized for the nr limit ( nrgr ) .
the resulting eft has manifest power counting in the typical velocity @xmath3 of the sources , and governs the dynamics of the long wavelength
radiation " gravitons that propagate out towards the gravitational wave detector . to illustrate our formalism , in this section we also give two classic examples of how the matching procedure and the eft velocity power counting rules are applied to reproduce the leading non - trivial consequences of general relativity in the context of the nr two - body problem : the lagrangian that describes the gravitational interactions of two nr objects at @xmath30 beyond newtonian gravity @xcite , and the interactions of gravitational radiation with the binary star at leading order in the velocity . in sec .
[ sec : rg ] we return to the relativistic point particle effective lagrangian , and show how all point particle divergences can be absorbed into worldline operators . in sec .
[ sec : matching ] , we discuss how , given a more complete stellar model one can determine the coefficients of operators in the point particle eft at the scale @xmath27 . finally , in sec .
[ sec : v6log ] we show how the rg running of the eft coefficients at order @xmath4 can be exploited to yield a simplified calculation of the terms in the nr expansion of the two - body gravitational lagrangian that scale as @xmath31 .
this calculation serves as a prototype for the rg running of worldline operators .
we have also included an appendix which illustrates how the feynman rules of the eft are derived and used .
to construct an eft that has manifest velocity power counting , we must first carefully analyze the kinematic configurations that arise in the binary problem . consider a binary system composed of slowly moving neutron stars or black holes .
let @xmath32 characterize the typical mass of the constituents ( taking them , in this discussion , to be roughly the same size ) , @xmath1 the typical separation between them and @xmath3 their relative three - velocity . in a system
whose evolution is determined to leading order by a @xmath33 newtonian interaction , these parameters are not independent , since the virial theorem relates @xmath34 the eft for the binary system will contain degrees of freedom representing the point particle binary constituents coupled to gravitons . while the typical particle momenta are of order @xmath35 , the gravitons appearing in a generic feynman diagram have momenta that can be divided into two classes . gravitons with momentum scaling
as @xmath36 mediate the forces responsible for binding the 2-body system .
these potential gravitons can never go on shell and thus will not appear as propagating degrees of freedom . on the other hand , radiation gravitons with momentum @xmath37 can appear on shell and must be kept in the eft to reproduce the correct long distance physics .
note that the interaction of an nr particle with a single potential or radiation graviton causes the particle to recoil by a fractional amount @xmath38 , where @xmath39 is the typical orbital angular momentum of the system .
thus as far as the graviton dynamics is concerned , the nr particles can be treated as background non - dynamical sources .
consequently , in this regard the eft that describes the binary configuration has more in common with heavy particle effective field theories ( eg . , hqet @xcite ) , in which light modes interact with a static , non - propagating source than a theory such as nrqcd @xcite , where the quark and gluon degrees of freedom have momentum components which differ only by powers of the velocity .
however , unlike hqet our eft is not an expansion in inverse powers of the mass .
as we will see , treating the stellar objects as non - propagating sources of gravitons also resolves a long standing formal problem in the treatment of general relativity as an eft @xcite . for astrophysical systems with masses @xmath40
, one would naively conclude from the conventional feynman rules of the eft that the expansion parameter is @xmath41 and therefore the energy expansion implicit in the interpretation of gravity as an eft seems to be spoiled @xcite .
in fact this obstruction to the eft interpretation of general relativity is only an artifact of treating the point sources as dynamical degrees of freedom , and disappears when the theory is formulated in terms of worldlines interacting with gravitons .
once this is done , the power counting in @xmath3 becomes manifest at the level of the action .
given these remarks , the starting point of our eft formulation consists of a theory of relativistic point particles coupled to gravity @xmath42 where the einstein - hilbert term @xmath43 describes the graviton dynamics and signature @xmath44 .
@xmath45 and @xmath46 determines the motion of the 2-body system ( @xmath47 labels the different particles ) . in this equation , @xmath48 is the proper time along the worldline @xmath49 of the @xmath50 particle
. we will ignore in this paper additional degrees of freedom that describe the spin or any additional multipole moments carried by each particle .
this is justified for an extended object in which the energy cost of exciting internal degrees of freedom is large compared to the typical frequency of the gravitons which it interacts with .
thus strictly speaking the formalism presented here can be used only to describe the dynamics of spinless black holes or spherically symmetric compact objects .
we will save the problem of including multipole moments in the eft for future work .
the first term in @xmath51 generates geodesic motion about the metric @xmath52 . besides this term , we have also explicitly shown the first two of an infinite set of possible non - minimal couplings of the point objects to the spacetime metric . since these non - minimal couplings cause deviations from pure geodesic motion , one would expect them to be associated with the effects of finite spatial extent of the sources .
thus by including the most general set of such operators , we systematically take into account all possible corrections due to the finite size @xmath0 of the object . as we discuss later , the operators linear in the curvature appearing in eq .
( [ eq : pp ] ) will not contribute to any physical observable .
this should be contrasted with higher dimension operators in the point particle action involving higher powers of the curvature tensor , which encode tidal deformations of the binary constituents .
we will discuss in a sec .
[ sec : matching ] how one goes about determining the precise relation between the coefficients of these operators and the microscopic physics ( ie , the stellar structure model ) that determines the internal structure .
a discussion of the laws of motion for extended objects in general relativity can be found in @xcite . a gravitational wave detector such as ligo measures indirectly the power emitted in gravitational radiation from the binary system ( see eq .
( [ eq : phase ] ) ) . given the action eq .
( [ eq : zoso ] ) , it is in principle possible to calculate this quantity by expanding the metric around flat space , @xmath53 , and integrating out the graviton field @xmath54 to obtain an effective action for the particle coordinates alone @xmath55 = \int{\cal d}h_{\mu\nu } \exp[i s_{eh } + is_{pp}].\ ] ] the effective action @xmath56 $ ] has a real part which generates the coupled equations of motion for the 2-body system , and consequently its mechanical binding energy .
it also has an imaginary part that measures the total number of gravitons emitted by a fixed two - particle configuration @xmath57 over an arbitarily large time @xmath58 @xmath59 = { 1\over 2 } \int de d\omega { d^2\gamma\over de d\omega},\ ] ] where @xmath60 is the differential rate for graviton emission from the binary system . from this quantity
we obtain the classical power spectrum @xmath61 , and therefore the gravitational wave phase seen by the detector . in principle
, one could directly evaluate eq .
( [ eq : zosopi ] ) using the lorentz covariant feynman rules generated by the einstein - hilbert lagrangian ( the feynman diagram expansion for perturbative gravity is reviewed , for example , in @xcite ) , taking the nr limit at the level of each amplitude .
however , the perturbative series generated in this way is not optimal for taking the limit @xmath62 .
for example , the one - graviton exchange term in @xmath63 $ ] , @xmath64 d_f(x_a - x_b)\ ] ] contains an infinite number of terms from the nr expansion of the proper times @xmath65 and the feynman propagator @xmath66 . while for a term like this it is easy to do the expansion in small velocity , in a more complicated diagram with multiple internal graviton lines it becomes more cumbersome to keep track of all the necessary terms at a given order in @xmath3 .
furthermore suppose one is interested , because of the limited sensitivity of the detector , in computing an observable only to some fixed order in @xmath3
. then how many diagrams do we need to keep ?
it is not possible to answer this question with the feynman rules generated by eq .
( [ eq : zoso ] ) as it stands .
likewise , given the non - minimal tidal couplings in eq .
( [ eq : pp ] ) it is not possible to determine at what order in @xmath3 they may first contribute .
since from the point of view of the point particle theory , eq . ( [ eq : zoso ] ) ,
the nr dynamics occurs in the far infrared , to formulate an eft that is tailored to the limit @xmath62 , we must integrate out all field degrees of freedom with wavelengths shorter than the orbital distance scale @xmath1 .
we now turn to a formulation of this eft .
the eft that systematically describes the nr two - body problem , which we call nrgr , has manifest power counting in the velocity @xmath3 . to construct nrgr
, we proceed in several steps .
we work in a preferred set of frames in which the relative velocity is small , allowing us to expand in @xmath3 . then in order to have interaction vertices with homogeneous velocity scaling , we simply expand @xmath51 in powers of the particle three - velocities . ignoring for now the effects of the higher - dimension tidal operators
, this leads to a worldline lagrangian @xmath67,\end{aligned}\ ] ] where @xmath68 are evaluated on the point particle worldline @xmath69 ( we use the euclidean metric @xmath70 to raise / lower spatial indices @xmath71 ) .
the propagator for the field @xmath54 appearing here is still fully relativistic , so it does not distinguish between potential gravitons , which have no right to appear as degrees of freedom in nrgr , and radiation gravitons . because of this , the feynman rules still do not scale homogeneously with @xmath3 . to deal with this problem , it is convenient to decompose the graviton as @xmath72 where @xmath73 represents the potential gravitons , with @xmath74 and @xmath75 describes a long - wavelength radiation field @xmath76 actually , it is better to further decompose @xmath73 by removing from it the large momentum fluctuations @xmath77 . to do this
we simply work with the fourier transformed field @xcite @xmath78 @xmath79 where @xmath80 is shorthand for @xmath81 .
the advantage of this redefinition is that now derivatives acting on any field in the eft scale in the same way , @xmath82 , so it is easy to count powers of @xmath3 coming from derivative interactions .
the nrgr lagrangian can then be derived by computing the functional integral @xmath83 = \int { \cal d}h_{\mu\nu } \exp[i s[{\bar h}+h , x_a ] + i s_{gf}],\ ] ] where @xmath84 is a suitable gauge fixing term .
we have not included ghost terms because as we will see , they are not needed in any nrgr computation .
( [ eq : nrgrpi ] ) indicates that as far as the potential modes @xmath73 are concerned @xmath75 is just a slowly varying background field . to preserve gauge invariance of the nrgr action , we will chose @xmath84 to be invariant under general coordinate transformations of the background metric @xmath85 . to be definite , we take the gauge @xmath86 with @xmath87 , where @xmath88 is the covariant derivative derived from the background metric @xmath89 . in this gauge ,
the @xmath90 terms in the action are ( after performing a rescaling of @xmath73 to obtain a canonically normalized kinetic term ) @xmath91,\end{aligned}\ ] ] where @xmath92 .
the terms in the second line of this equation are suppressed relative to the first line by a power of @xmath93 , and are treated perturbatively , as operator insertions , in correlation functions .
thus , the propagator for @xmath94 is @xmath95 where @xmath96 $ ] with @xmath97 the spacetime dimension .
if we assign the scaling @xmath98 , @xmath99 , we learn from this that a potential graviton scales as @xmath100 .
likewise , because in mometum space the radiation graviton propagator scales as @xmath101 , the coordinate space radiation graviton field is taken to scale as @xmath102 . just as we had to expand @xmath73 in terms of momentum labels to obtain manifest power counting @xcite ,
it is also necessary to perform a multipole expansion of the radiation field @xmath75 whenever it couples to particle worldlines or to potential modes @xcite . to see the necessity for this ,
consider the absorbtion of a radiation graviton from a potential graviton as shown in fig . [ fig : mult ] .
the outgoing potential graviton has a propagator of the form @xmath103 which does not scale homogenously in @xmath3 since @xmath104 and @xmath105 .
the propagator must be expanded in powers of @xmath106 , which is achieved by multipole expanding the interactions of the radiation graviton field with the potentials at the level of the action , absorbs a radiation graviton with momentum @xmath107.,width=151 ] @xmath108 where @xmath109 , and @xmath110 is some reference point in the vicinity of the point particle ensemble , for instance the center of mass .
likewise , consider the amplitude for the point particle system to absorb a @xmath111-polarized radiation graviton with 4-momentum @xmath107 @xmath112 a graviton emitted during the binary inspiral phase has 4-momentum @xmath113 , so if the coordinates @xmath114 are measured relative to , for instance , the center of mass , then we have @xmath115 and consequently this amplitude contains an infinite number of terms each with a different velocity scaling .
this situation is also remedied by the multipole expansion in eq .
( [ eq : mult ] ) .
@xmath116 these field redefinitions are enough to recast the original point particle theory of eq .
( [ eq : zoso ] ) into a form that is better suited to a small velocity expansion .
the power counting rules in table [ tab ] allow one to determine the order in the velocity at which any diagram in the calculation of the functional integral in eq .
( [ eq : nrgrpi ] ) , as well as in feynman diagrams computed in nrgr .
however the simplest example of the power counting rules , the counting of the particle nr kinetic term , reveals a slight subtlety that we must deal with .
given that @xmath117 , we have @xmath118 thus this term scales as the orbital angular momentum @xmath119 .
this is not the only term that scales in this way .
the exchange of a single @xmath120 potential graviton between two particles , responsible for the newton force , involves the computation of the diagram of fig .
[ fig : newt]a , which is given by @xmath121 ( @xmath122 ) .
this scales as @xmath123 ^ 2\sim mvr,\ ] ] by virtue of the virial theorem ( the power divergent self - energy graphs fig .
[ fig : newt](b ) , fig .
[ fig : newt](c ) , which renormalize the particle masses , vanish when evaluated in dimensional regularization ) .
it can be shown in general that all other contributions to @xmath56 $ ] are down by powers of @xmath3 relative to the particle kinetic terms or the newton potential ( see the appendix ) .
it is then easy to deal with the presence of the large parameter @xmath124 in the power counting .
we must treat the operators that scale as @xmath125 non - perturbatively , or in other words , we must expand about background fields @xmath114 that to leading order in @xmath3 satisfy the equations of motion of newtonian gravity , treating operators which scale as @xmath126 perturbatively .
this is the resolution of the problem raised in @xcite regarding the apparent breakdown of a manifest perturbative expansion when the particle masses are larger than @xmath127 .
one can also show that diagrams that contain graviton loops are supressed by powers of @xmath128 relative to the tree diagrams ( after particle worldlines are stripped off ) .
this is in accord with the intuition that loop diagrams represent quantum corrections which should not be kept in the perturbative expansion of classical field theory observables .
the fact that loop corrections are irrelevant to the quantities of interest also explains why we did not need to keep ghost terms in the functional integral of eq .
( [ eq : nrgrpi ] ) . what is the resulting theory after all potential gravitons are integrated out ?
it is a theory of radiation gravitons coupled to moments of the two - particle distribution in a gauge invariant manner .
working in the cm frame , @xmath129 , the leading order contributions to processes in nrgr can be calculated from @xmath130\\ \nonumber & & -{1\over 2 m_{pl } } \epsilon_{ijk } { \bf l}_k\partial_j { \bar h}_{0i } + { 1\over 2 m_{pl}}\sum_a m_a { \bf x}_{ai}{\bf x}_{aj } r_{0i0j},\end{aligned}\ ] ] where we suppress the argument @xmath131 of the canonically normalized radiation field @xmath75 , and write @xmath132 @xmath133 , @xmath134 .
note that after integrating out the potential modes , the only dependence on the orbital scale @xmath1 appears in the wilson coefficients ( the moments ) of the theory .
thus in our theory , the decoupling between near - zone nr physics and far - zone radiation is manifest at the level of the nrgr lagrangian .
this explicit separation of scales , which is what motivates the eft approach , should simplify the evaluation of higher order radiative tail effects within our formalism .
of the terms shown here , only the coupling of the ( linearized ) riemann tensor to the moment @xmath135 can source gravitational radiation , since all other terms in this lagrangian couple to conserved quantities and therefore can not radiate , up to higher order radiation reaction effects .
in fact , the trace of the linearized riemann tensor @xmath136 vanishes for on - shell graviton matrix elements , so that radiation only couples to the traceless quadrupole moment of the source @xmath137 in the next section we will show more explicitly how to derive eq .
( [ eq : nrgrl ] ) . given eq .
( [ eq : nrgrl ] ) , how do we obtain the power radiated in gravitational waves from this formalism ?
we simply calculate the imaginary part of the self - energy diagram of fig .
[ fig : selfrad ] , @xmath138 ( we have only kept the part of the diagram that arises from the coupling @xmath139 , since it is the only term at this order which has a non - zero imaginary part ) .
computing the two - point function of the linearized @xmath140 one finds that it is proportional to the projection operator onto symmetric and traceless two - index spatial tensors , @xmath141 . writing the above equation in terms of the fourier transform of the quadrupole moment @xmath142 we end up with @xmath143
this diagram then gives rise to the following contribution to @xmath144 $ ] @xmath145 the integrand in this equation is proportional to the differential graviton emission rate over the history of the two - particle system . to get the differential power ,
multiply the integrand by an additional factor of the graviton energy @xmath146 .
in particular , including this extra factor of the energy and integrating over all momenta , we obtain the total power radiated in gravitational waves @xmath147 where the dots denote time derivatives , and the brackets denote a time average .
of course , this equation is just the usual quadrupole radiation formula .
the construction of the eft in the previous section was perhaps somewhat formal , so it is helpful to see examples of how to explicitly calculate the terms in the nrgr action .
we will work out two textbook examples in this section : the lagrangian that leads to the leading order predictions for the power radiated in gravitational waves , eq .
( [ eq : nrgrl ] ) , and the eih lagrangian @xcite that describes the @xmath30 general relativity corrections to the gravitational interaction of two point masses .
first consider @xmath148 .
it arises from the diagrams shown in fig .
[ fig : eih1 ] and fig .
[ fig : eih2 ] . using the nrgr feynman rules , the sum of the diagrams with a single potential graviton exchange are given by @xmath149,\end{aligned}\ ] ] with @xmath150 , @xmath151 and @xmath152 \int_{\bf k } { 1\over { \bf k}^2 } e^{-i{\bf k}\cdot({\bf x}_1 -{\bf x}_2 ) } \\ & = & { 3i\over 2}\int dt { g_n m_1 m_2\over |{\bf x}_1 -{\bf x}_2| } { \bf v}^2_1.\end{aligned}\ ] ] here , we have used the formula @xmath153 however , at this order in @xmath3 , the single exchange diagrams are not sufficient .
the nrgr power counting tells us that we also get contributions from diagrams with two and three graviton propagators .
for example , the graph in fig .
[ fig : eih2]a scales as @xmath154 ^ 3 \left[dx^0 ( d^3{\bf k})^3 \delta^3({\bf k } ) { { \bf k}^2 h^3_{00}\over m_{pl}}\right ] \sim \left[{r\over v}{1\over r^3 } ( l v)^{1/2 } r^2 v^{1/2}\right]^3\left[{1\over m_{pl}}{r\over v } { 1\over r^8 } \left(r^2 v^{1/2}\right)^3\right ] \sim l v^2.\ ] ] the first factor stems from three insertions of leading order vertices while the second bracket arises from an insertion of a three - graviton interaction . expanding the einstein - hilbert lagrangian to order @xmath155 , it is straightforward to show @xmath156 so that @xmath157 where we have dropped linearly uv divergent self - energy contributions which can be absorbed into the particle masses , and which vanish when evaluated in dimensional regularization . finally , the seagull diagram , fig .
[ fig : eih2]b , also contributes at order @xmath158 , and is given by @xmath159 ^ 2 = { i\over 2}\int dt { g^2_n m_1 m^2_2\over |{\bf x}_1(t ) -{\bf x}_2(t)|^2}\ ] ] putting together eq .
( [ eq : eih1 ] ) , eq .
( [ eq : eih2 ] ) , eq .
( [ eq : eih3 ] ) , eq .
( [ eq : eih4 ] ) , and eq . ( [ eq : eih5 ] ) together with their mirror images under the interchange of the particle labels , we end up with @xmath160- { g^2_n m_1 m_2 ( m_1+m_2)\over 2 |{\bf x}_1 -{\bf x}_2|^2 } , \end{aligned}\ ] ] where we have also included the relativistic corrections to the kinetic energy of the point particles , which contribute to the effective action at order @xmath158 .
this lagrangian was first derived by einstein , infeld and hoffmann @xcite , by different methods .
a diagrammatic derivation of the static @xmath161 part of this lagrangian , using lorentz covariant worldline methods @xcite , was given in ref .
a quantum field theoretic derivation of the @xmath161 potential , treating the source particles as dynamical fields , was presented in @xcite .
it is also instructive to use the formalism to derive the leading order lagrangian that describes the interactions of radiation gravitons with nr sources , eq ( [ eq : nrgrl ] ) .
the derivation involves all aspects of the construction of nrgr ( the multipole expansion for radiation , integrating out potential gravitons , velocity power counting rules ) . at the lowest two orders in the expansion , the interaction of radiation with the nr sources follows simply from the multipole expansion of the radiation field . expanding the proper time time term in powers of @xmath3 , using the counting rules in table i , leads to @xmath162 in the cm frame .
note that to preserve manifest scaling in @xmath3 we must multipole expand about the center of mass , as defined below eq .
( [ eq : nrgrl ] ) .
these terms ( when integrated over time to form an action ) scale as @xmath163 , and @xmath164 respectively according to the power counting rules of the eft , and generate diagrams which ( i.e. terms in @xmath56 $ ] ) that are proportional to a single power of @xmath124 .
at the next order in the velocity , the situation is not as straightforward .
keeping terms with one more power of @xmath3 ( either from the multipole expansion or from explicit powers of the particle velocities ) one finds , in the cm frame , @xmath165.\end{aligned}\ ] ] this lagrangian has two problems .
first , it is not gauge invariant under infinitesimal coordinate transformations .
second , it seems to predict that unphysical @xmath166 graviton polarizations can be emitted by the two - particle bound state .
of course , these two problems are related .
the resolution to this is that at this order in the expansion there are additional contributions to the effective lagrangian , from diagrams with both external radiation gravitons and internal potentials , such as those of fig .
[ fig : jeebus ] .
the diagram in fig .
[ fig : jeebus]b is given by @xmath167 for the diagram in fig .
[ fig : jeebus]c , we need the @xmath168 , vertex
. it can be obtained from @xmath169 + { \bf k}^2 { \bar h}_{0i}\left[2 h^{00}_{\bf k } { h_{-\bf k}}_{0i } - { h_{\bf k } } _ { 0i } h_{-\bf k}\right]\\ \nonumber & & + { \bar h}_{ij } \left[{1\over 2 } { \bf k}_i { \bf k}_j h_{\bf k}^{\mu\nu } { h_{-{\bf k}}}_{\mu\nu } + { \bf k}^2 { h_{\bf k}}_{i \mu } { h_{-{\bf k}}}_j^\mu - { 1\over
2}{\bf k}^2 { h_{\bf k}}_{ij } h_{\bf -k } - { 1\over 4 } { \bf k}_i { \bf k}_j h_{\bf k } h_{-{\bf k } } -\delta_{ij}{\bf k}^2 \left(-{1\over 4 } h_{\bf k}^{\mu\nu } { h_{-{\bf k}}}_{\mu\nu } + { 1\over 8 } h_{\bf
k } h_{-{\bf k}}\right ) \right],\end{aligned}\ ] ] where the integral over momentum @xmath170 has been suppressed , and because of the multipole expansion , @xmath75 is evaluated at @xmath131 .
given this term , it is easy to show that @xmath171.\end{aligned}\ ] ] using @xmath172,\ ] ] as well as the equations of motion for @xmath173 from the order @xmath174 effective action , we find @xmath175.\end{aligned}\ ] ] adding together the result of eq .
( [ eq : jeebus2 ] ) plus its mirror image under @xmath176 and eq .
( [ eq : jeebus3 ] ) to eq .
( [ eq : jeebus1 ] ) we find the complete nrgr action at order @xmath177 , @xmath178\\ \nonumber & & -{1\over 2 m_{pl}}\epsilon_{ijk } { \bf l}_k \partial_j { \bar h}_{i0 } + { 1\over 2 m_{pl}}\sum_a m_a { { \bf x}_a}_i { { \bf x}_a}_j r_{0i0j}.\end{aligned}\ ] ] there are several things to notice about this equation .
first , we find that @xmath179 couples at this order in @xmath3 to the newtonian gravitational energy of the bound system . at this order in the expansion
this is a constant of the motion , as is the orbital angular momentum , which couples to @xmath180 .
consequently , neither of these quantities can radiate .
this is in accord with the expectation that @xmath181 do not correspond to physical graviton polarizations , and should not be emitted by a graviton source . it should be possible to give a general diagrammatic proof that the diagrams in the matching involving one external @xmath179 are related in a simple way to the diagrams with no external radiation gravitons , i.e , the diagrams that contribute to the total energy , as required by general covariance .
also , note that this lagrangian is manisfestly gauge invariant under infinitesimal gauge transformations .
this is obvious for the last term in eq .
( [ eq : nrgrv2 ] ) , which involves the ( linearized ) riemann tensor . for
the first term use the fact that the newtonian energy is ( at this order ) time independent , so that under the transformation @xmath182 , this term in the lagrangian transforms into a total time derivative .
likewise , under @xmath183 , the coupling of @xmath180 to the angular momentum transforms into a total time derivative .
we expect that the requirement of gauge invariance should provide strong constraints on the form of the nrgr lagrangian .
in particular , the derivation of the @xmath30 radiation couplings presented here could have been significantly shortened had we imposed gauge invariance . in practice , one can reduce the amount of work by keeping only the terms from the multipole expansion and fixing the remaining terms by demanding gauge invariance of the radiation couplings .
it is interesting to note that in order for the couplings of the radiation gravitons to preserve gauge invariance , it is crucial that we include the diagrams of fig .
[ fig : jeebus]b and fig .
[ fig : jeebus]c . these diagrams , which arise due to the non - abelian nature of the gravitational interactions ,
represent the fact that in gravity it is not just the ensemble of nr particles that acts as a source of gravitons , but also the energy stored in the gravitational field that it sets up . to obtain consistent results such non - linear effects must be included . without the velocity power counting rules of the eft
, it would have been difficult to ascertain whether such non - linear diagrams contribute already at this order in the velocity expansion . in the classical general relativity approach to calculating the power emitted in gravitational waves by a nr source
, this non - linearity appears in the expression for the far zone gravitational field as an expansion in moments not over the stress tensor @xmath184 of the matter alone , but also over a suitable ( non - unique ) stress - energy
pseudotensor " @xmath185 for the gravitational field itself .
this pseudotensor is defined in such a way that @xmath186 , and physically plays the same role as the diagrams of fig .
[ fig : jeebus]b and fig .
[ fig : jeebus]c do in our formalism . as an aside
, we also note that the necessity to include diagrams involving gauge particle self - interactions in order to derive the couplings of radiation to sources has an analog in the worldline approach to nr bound states in qcd @xcite .
for instance , to show that the color dipole moment of a @xmath187 bound state couples to the full color electric field , it is not enough to keep diagrams with gluons emitted from the quark worldlines .
one also needs to keep diagrams involving the 3-gluon vertex , which can be shown to contribute at the same order in the velocity @xcite .
the coefficients of the relativistic point particle theory of eq .
( [ eq : pp ] ) exhibit non - trivial rg flows , even at the classical level ( rg flows in classical field theories coupled to extended spatial sources have been analyzed in @xcite ) .
the non - trivial rg structure of the theory is associated with logarithmic uv divergences which arise due to the singular nature of the point particle limit . to renormalize the theory we consider a single particle at rest at @xmath188 and compute the diagrams that contribute to the terms in the effective action for a background gravitational field that are linear in the background @xmath189 . if we ignore for now the non - minimal curvature couplings of eq .
( [ eq : pp ] ) , this is given by diagrams such as those of fig .
[ fig : tadpoles ] with a single external graviton and any number of insertions of the worldline proper time coupling , each proportional to the particle mass ( we restrict ourselves to tree diagrams , which are the ones that remain in the limit @xmath190 , with @xmath191 ) . while the diagram with a single mass insertion is finite , diagrams with more mass insertions have uv divergences . for instance
the diagram with two powers of @xmath32 in fig .
[ fig : tadpoles](b ) contains a piece which has a power divergence of the form @xmath192 , as can be seen by counting powers of momentum from the internal propagators and the graviton self - interaction vertex .
an explicit calculation shows that this linearly divergent piece , which vanishes in dimensional regularization , just renormalizes the bare mass parameter @xmath32 appearing in the point particle action . besides this power divergent piece , the diagram of fig .
[ fig : tadpoles](b ) also gives rise to a finite , non - analytic term in the background field effective action @xcite of the form @xmath193={1\over
m_{pl}}\int { d^4 k\over ( 2\pi)^4}h_{\mu\nu}(-k ) t_{(2)}^{\mu\nu}(k)+\cdots,\ ] ] where , in background field gauge with a gauge fixing term for the quantum " graviton field @xmath73 of the form @xmath194 , @xmath195 one finds @xmath196\int { d^3{\bf q}\over ( 2\pi)^3 } { 1\over { \bf q}^2 ( { \bf q } + { \bf k})^2 } \\
\nonumber & = & ( 2\pi)\delta(k^0 ) { m^2\over 16 m^2_{pl } |{\bf k}|}\left[-{7\over 32}(\eta_{\mu\nu } k^2 -k_\mu k_\nu)\right.\\ & & { } \left .
+ { 1\over 32 } k^2 v_\mu v_\nu\right]\end{aligned}\ ] ] where @xmath197 is the particle four - velocity .
as expected by background field gauge invariance , we have @xmath198 .
in fact , since @xmath199 $ ] must be covariant with respect to the background gravitational field , the above result is sufficient to completely fix all terms in the effective action that are proportional to two powers of @xmath32 and any number of powers of the background field @xmath54 .
although the power divergences that arise in the calculation of @xmath200 have no physical consequences , logarithmic divergences can give rise to measurable effects since by dimensional analysis , in a massless field theory the logarithm of the cutoff is always accompanied by a logarithm of the external momentum .
consider the diagrams with three insertions of the particle mass .
they give rise to a term in the effective action of the form @xmath193={1\over m_{pl}}\int { d^4 k\over ( 2\pi)^4}h_{\mu\nu}(-k ) t_{(3)}^{\mu\nu}(k)+\cdots,\ ] ] where we find , from fig .
[ fig : tadpoles](c ) and fig .
[ fig : tadpoles](d ) @xmath201.\ ] ] as a non - trivial check of this result , we note that gauge invariance with respect to the background field , @xmath202 , does not hold diagram by diagram , but is only manifest after the terms in fig .
[ fig : tadpoles](c ) and fig .
[ fig : tadpoles](d ) are combined . to obtain this equation , we worked out the necessary background field gauge feynman rules @xcite in @xmath97 dimensions , and performed an expansion in @xmath203 , keeping only the terms that do not vanish as @xmath204 .
@xmath205 contains a uv logarithmic divergence , as can be seen by power counting the integral @xmath206 which evaluates to @xmath207+{\cal o}(\epsilon).\\\end{aligned}\ ] ] in the second line we have expanded about @xmath208 and introduced an arbitrary subtraction scale @xmath209 . inserting this expression back into eq .
( [ eq : t3 ] ) , we see that the set of diagram with three mass insertions has a @xmath210 pole which can not be absorbed by a shift in the mass parameter @xmath32 of the minimal point particle action , whose feynman rule is , for a particle moving on a straight line trajectory with constant four - velocity @xmath211 , @xmath212 however , the non - minimal operators of eq .
( [ eq : pp ] ) contribute terms to @xmath199 $ ] of the form @xmath213 , with @xmath214,\ ] ] where we have dropped terms proportional to @xmath215 which vanish due to the delta function . because the divergent terms in eq .
( [ eq : t3 ] ) have the same dependence on @xmath107 and @xmath3 as the contribution from the curvature couplings , we see that the @xmath210 poles can be subtracted away by a proper choice of coefficients @xmath216 . in the @xmath217 scheme ,
the renormalized @xmath218 term in the effective action now becomes @xmath219,\ ] ] where @xmath220 are the renormalized couplings .
since the renormalization scale @xmath209 introduced above is arbitrary , we must have @xmath221/d\mu=0 $ ] ( since the classical field @xmath54 does not pick up an anomalous dimension ) .
thus the explicit dependence of the logarithm in eq .
( [ eq : renormt3 ] ) on the subtraction scale @xmath209 must be cancelled by allowing the coefficients @xmath216 to vary with scale .
our theory therefore exhibits non - trivial _ classical _ rg scaling , @xmath222 as a consistency check , we may use the linear term in @xmath199 $ ] to calculate the spacetime metric generated by a point mass at rest , up to order @xmath223 . fixing the gauge @xmath224 for the background field @xmath54 ,
this is simply given by @xmath225 , where , in momentum space @xmath226\over \delta h_{\alpha\beta}(-k)}\right|_{h=0}.\ ] ] using our results for @xmath227 , we find , dropping delta function contributions which first arise at order @xmath228 , @xmath229\\ & & { } + { x_i x_j\over r^2}\left[7\left({g_n m\over r}\right)^2 -{4\over 3 } \left ( { g_n m\over r}\right)^3 + 24 \left ( { g_n m\over r}\right)^3\left\{{c_r(\mu)\over g_n^2 m^3 } + { 1\over 6}\ln\mu r + { \gamma\over 6}\right\}\right]+\cdots , \end{aligned}\ ] ] and @xmath230 it is straightforward to show that this metric satistfies the vacuum einstein equations @xmath231 away from the point @xmath188 where the point particle source is located , independently of the value of the renormalized parameter @xmath232 .
this is consistent with birkhoff s theorem , which states that the space of static , rotation invariant solutions of the vacuum einstein equations is one - dimensional , parametrized by the mass @xmath32 .
alternatively , one may check the validity of our result for @xmath52 by constructing a coordinate transformation to the more familiar harmonic coordinates @xmath233 with @xmath234 , which in terms of our coordinates @xmath235 are given by , up to @xmath236 terms , @xmath237.\end{aligned}\ ] ] thus , as required by general arguments , the parameters @xmath220 are not observable in the long range field of isolated static sources . as we argue in the next section ,
it is likely that the value of the matching coefficients @xmath238 are themselves of order @xmath239 . in this case
, it is easy to see using the nrgr power counting rules that diagrams with one insertion of the non - minimal operators in eq .
( [ eq : pp ] ) , for instance the diagram of fig .
[ fig : v6logv1](a ) lead to corrections to the two - body interaction lagrangian which are suppressed by a power of @xmath4 relative to the leading newton interaction .
however , it can be shown that the worldline operators @xmath240 can be removed by a local redefinition of the metric tensor ( that is , the coefficients @xmath216 can be shifted arbitrarily by these transformations ) .
given that any possible signature of internal structure that may arise at @xmath4 must be encoded in these two operators withing a point particle description , we conclude for spinless objects _ there are no finite size effects up to order @xmath4_. even though all values of @xmath209 give rise to the same physics , in practice certain choices may be more convenient when doing actual computations . in general
, this leads to renormalization group equations which can be used to simplify the calculation of logarithmically enhanced terms in the perturbative expansion .
we will give an example in section [ sec : v6log ] of how to exploit the freedom in choosing @xmath209 for the unphysical couplings @xmath241 to simplify the calculation of certain contributions to the @xmath4 two - body potentials that are enhanced by a logarithm of @xmath24 , where @xmath0 is a length scale that characterizes the size of the extended object .
the calculation of logarithms for physical tidal couplings , which involve the riemann tensor , is completely analogous to this case .
any choice for the couplings @xmath220 can be trivially expressed in terms of the parameters at some fixed scale @xmath242 by solving the above rg equations @xmath243 ( as we will see later , at higher order in the velocity expansion , the coupling constant flows show more interesting behavior than is depicted in these equations ) .
if the couplings @xmath238 had any physical content , in order to fix them say at a scale of order the characteristic size @xmath0 , requires model - dependent information about the details of the short wavelength physics that resolves the point particle divergences .
this will be true for the physical tidal couplings as well .
this physics is independent of the nr two - body dynamics , so that no pn formalism based on point particles can , by itself , give rise to a full prediction for the gravitational wave observables . in the next section
we will explain how , given this short distance model , it is possible to perform a matching calculation that fixes the values of the worldline operator coefficients in terms of the parameters ( size , equation of state , @xmath244 ) that characterize the extended object .
an interesting question arises as to the size of the matching coefficient @xmath238 , or their more physical tidal counterparts . if it turned out that these coefficients were anomalously large , then this would put the systematics of the pn approximation in jeopardy .
for instance if the coefficients in eq .
( [ eq : pp ] ) were much larger than @xmath239 ( no explicit powers of @xmath3 , a dynamical variable , can appear in operator coefficients ) , the newtonian result would no longer be leading order , and the pn expansion would cease to be valid .
although it seems rather unlikely that this constraint is violated , it is interesting to note that within the eft formalism here , we would retain calculability given a model for the short distance physics , but we would have to change our power counting scheme in order to accomodate the lack of a newtonian gravity limit .
if one works at high enough of an order for the worldline operators to become relevant , i.e. beyond @xmath4 , then given that pn theory has nothing to say about the actual values of the worldline operator coefficients , one needs a model of the internal structure .
otherwise the best one can do is to express the predictions for the gravitational wave observables in terms of the parameters of the eft and use the gravitational wave data to obtain information about the magnitude of the coefficients . alternatively ,
if one does have information about the structure of the binary constituents , one can fix the values of the couplings in the effective action at a renormalization point @xmath27 . in this case
the data can , in principle , be used to test one s assumptions regarding the structure of the binary star sources . in this paper
we will not perform the matching calculation , but instead just sketch how such a calculation would lead to an unambiguous prediction for the gravitational wave observables .
we will illustrate how this works within the context of a specific toy model which captures the essential physics .
the matching procedure is standard within effective theories @xcite , but to keep the discussion self - contained we will review here the basic concepts .
the idea is simply to compare observables such as green s functions in the full " ( valid at short distances ) and effective theories and adjust the value of the eft couplings in such a way that the effective theory reproduces the results of the full theory .
the effective theory is designed to reproduce the long distance physics , characterized by non - analytic behavior about zero external momentum , but it may get the short distance , which is analytic in the momentum ( that is , local in coordinate space ) , wrong ( hence the divergences ) .
consider for instance the matching of one - point functions in the full theory ( consisting of an extended source interacting with gravitons ) and the eft . in the full theory ,
the one point - function of the graviton are of the form @xmath245 where @xmath246 is in general a complicated function of the external momentum @xmath107 as well as some dimensionful parameters @xmath247 which describe internal structure . for simplicity
we assume in the discussion that the only relevant scale is the size of the object @xmath0 .
the eft is valid in the kinematic regime @xmath248 , so in order to match , one expands the function @xmath249 about this point . in the simplest situations ,
this expansion takes the form @xmath250,\ ] ] where @xmath251 are polynomials . in general , the non - analytic dependence on the momentum may be more complicated , involving polynomials times more general functions of @xmath252 , however , we focus on this case for simplicity . in practice , since we are interested in a low energy expansion , we need to keep only the first few terms in the polynomials @xmath251 .
provided that the point particle coupled to gravity properly reproduces the ir ( long distance ) physics , the eft result must be of the same general form as the above @xmath253,\ ] ] where the parameters @xmath254 are the coupling constants of the eft defined by some suitable renormalization prescription , which in the point particle theory are the particle mass and the curvature couplings .
the functions @xmath255 are polynomials of @xmath256 whose coefficients are products of the @xmath257 .
note that the log divergence arises because we have taken the singular limit @xmath258 . since this logarithm arises in the full theory from the light degrees of freedom which the eft must also contain , it is crucial that the non - analytic momentum dependence in eq .
( [ eq : geft ] ) coincides with that of eq .
( [ eq : gfull ] ) . in order for the eft to reproduce the results of the full theory ( at least to some fixed order in the small parameter @xmath259 )
, we then tune the coefficients @xmath220 so that @xmath260 in the limit @xmath261 .
for instance , in the previous section , we had @xmath262 and @xmath263 , hence in the full theory that resolves the point particle singularities , we expect @xmath264 for @xmath265 ( we are ignoring finite , non - analytic power corrections of the form @xmath266 from fig [ fig : tadpoles]b , which are not relevant to this discussion ) . comparing eq .
( [ eq : gfull ] ) and eq .
( [ eq : geft ] ) in this case we obtain a relation @xmath267 where @xmath268 is a full theory parameter obtained from the expansion of @xmath269 to linear order . to avoid large logs which could hinder our perturbative expansion , we choose @xmath270 .
this is the boundary value of the solution to the rg equation eq .
( [ eq : rgsol ] ) .
actually , this equation ignores the presence of scheme dependent constants ( which are not relevant to physical predictions ) which are of the same magnitude as the coefficient of the logarithm .
thus even if the short distance parameter @xmath268 were much smaller than @xmath271 , we would find @xmath272 .
in this section we illustrate the general remarks above in the context of two simple toy models .
the matching of the worldline operator coefficients which carry information about the static properties is conceptually similar to the matching in a model consisting of a self - interacting scalar field @xmath273 coupled to a point - particle source . to match point particle couplings which carry dynamical information about the binary star constituents , for instance a tidal operator of the form @xmath274 that may appear in the point particle eft , requires matching amplitudes with more than one external graviton .
we will consider an analog of this in electrodynamics , where we will describe the matching onto the point particle operators @xmath275 , @xmath276 which encode the dynamical repsonse of a charge distribution to applied low frequency electromagnetic fields .
explicit calculations of the counterterms need to fix the operators coefficients for the point particle eft a gravitating compact object are currently in progress @xcite .
first consider a self - interacting massless scalar field @xmath277 in four dimensions coupled to a localized source @xmath278 , @xmath279 and imagine that through some unspecified dynamics , there is a frame in which the source @xmath278 acquires a profile of the form @xmath280,\ ] ] where @xmath281 , and @xmath0 are constants that parametrize the short distance physics .
although this model is unrealistic in the sense that the profile @xmath278 does not arise dynamically but is rather put in by hand , it is adequate for our purposes here . in the limit where only long wavelength modes of the field @xmath277 are accessible , the detailed structure of the source @xmath278 becomes irrelevant .
as far as long distance modes are concerned , one may replace the localized source with a point particle that has suitable local couplings to the field @xmath277 .
@xmath282 - m\int d\tau -g \int d\tau \phi + \cdots,\ ] ] where @xmath283 $ ] is the bulk scalar field action , and we have suppressed higher dimension worldline couplings involving more powers of @xmath273 or its derivatives . the coupling @xmath284 appearing in this equation is a function of the paramaters @xmath281 , @xmath0 of the short distance source model , which we determine by matching amplitudes in the low energy eft with those obtained from eq .
( [ eq : uvmodel ] ) .
at lowest order in @xmath285 , this is trivial .
consider for instance the amplitude for @xmath286 . in the full theory @xmath287
so that @xmath288,\ ] ] which essentially is the form factor associated with the source distribution @xmath278 . in the eft , @xmath289
the two amplitudes agree in the limit @xmath290 if we set @xmath291 at higher order in the momentum expansion we would also need operators of the form @xmath292 in the low energy theory to reproduce the full theory amplitude .
similarly , by matching the vacuum to vacuum amplitudes , we find @xmath293
. note that in the point particle eft , there are logarithmic divergences in the @xmath286 amplitude at @xmath294 , from a diagram analogous to the one in fig . [
fig : tadpoles](d ) , @xmath295,\end{aligned}\ ] ] where @xmath296 was defined in the previous section , and as before we have used dimensional regularization plus the @xmath217 scheme to renormalize the divergent amplitude .
if the eft is to be a correct description of the low energy physics , the non - analytic dependence on the momentum in this equation should reproduce a similar log in the full theory amplitude . indeed , calculating in the full theory and including insertions of the @xmath297 interaction we find , dropping terms subleading in @xmath298 @xmath299 which agrees with the claimed universal result eq .
( [ eq : gfull ] ) . therefore @xmath300,\ ] ] which indicates that to avoid large logarithms that may render perturbation theory invalid , we should match at a scale @xmath27 , so that @xmath301 . then we can use the rg equation @xmath302 to sum logarithms of @xmath303 .
even though we phrased the matching problem using quantum field theory language , we stress that our discussion here applies also in the classical regime .
this can be seen by matching not scattering amplitudes , but the 1-point functions @xmath304 , in the presence of the point particle source , which have an obvious classical interpretation .
how does this model bear on the matching problem for the point particle coupled to gravity ? as in the gravitational case , our toy scalar field model exhibits uv divergences in the point particle limit . in both cases , we expect those divergences to be resolved by the finite extent of the distribution which sources the field .
we saw this explicitly in the scalar field theory , where we learned that the matching procedure fixes the couplings of the point particle at a length scale of order the size of the source .
we expect that the same will occur in the gravitational problem .
however , there are several aspects of the gravity problem that are not captured by this simple model . in the scalar field theory ,
the uv divergences associated with the point particle singularities affect the leading order coupling @xmath305 itself , and thus the point @xmath306 is a ( trivial ) fixed point in the rg sense . in the gravitational case ,
on the other hand , the structure of the rg flow is such that even if the curvature couplings were equal to zero at @xmath27 , non - zero values would get induced at longer wavelengths .
second , in the scalar model , the order @xmath307 correction to the long range field encodes information about the internal structure of the extended object that couples to @xmath273 . for
a source in isolation coupled to gravity , birkhoff s theorem states that the curvature couplings can not generate long range static forces that could carry information about the short distance dynamics .
nevertheless , birkhoff s theorem does not prohibit the momentum space off - shell graviton one - point function from having sensitivity to the internal structure .
thus in principle one could still match the graviton one - point functions ( or the quantities @xmath308 appearing in the background field action ) , although in practice it may be difficult to compare in a gauge invariant manner .
because of this , it may be simpler to match by comparing gauge invariant quantities , such as a the amplitudes for fields to scatter off the geometry of the extended source versus the scattering amplitude in the point particle theory .
fixing the coefficients of the operators that measure tidal effects in the point particle eft will require matching amplitudes with more external gravitons .
a similar situation arises in electrodynamics , where dynamical effects such as induced dipole moments are encoded in the photon two - point function ( equivalently the response to an external field ) .
consider for instance the effective point particle action that describes the interactions of the electromagnetic field with a perfectly conducting grounded sphere of radius @xmath0 .
since by definition the times scales associated with the dynamics of a perfect conductor are short compared to those of the electromagnetic fields it interacts with , the electrodynamic analog of eq .
( [ eq : pp ] ) is then @xmath309 imagine putting this system in a constant background electromagnetic field @xmath310 .
expanding the field about @xmath311 in the point particle eft , the fluctuating photon field @xmath312 acquires a vacuum expectation value which is given , in the feynman gauge , by @xmath313 imposing the appropriate boundary conditions for a perfect conductor , the microscopic " model predicts @xmath314 so matching fixes the gauge invariant eft coefficients to be @xmath315 , @xmath316 in this particular model , all higher dimension operators built out of more derivatives or powers of @xmath317 have vanishing coefficients .
although the operators @xmath318 and @xmath319 can be removed by a redefinition of the metric and thus never contribute to physical observables , it is instructive to use these operators to illustrate the general procedure for calculating the contributions to the two - body potential which are enhanced by a logarithm of @xmath24 .
such terms are independent of the details of the short distace dynamics . although these terms have been obtained in a different gauge by conventional pn methods in refs .
@xcite , here we give a derivation that exploits the rg scaling of the coefficients in the eft to simplify the calculations .
the idea is that because the amplitudes in the eft are independent of the subtraction scale @xmath209 we can choose any convenient value for it . in particular , if one chooses @xmath209 to be of order the orbital scale @xmath33 , then the terms in the two - body potential that involve logarithms of @xmath320 ( which arise from taking the fourier transform of diagrams involving subgraphs such as those in fig .
[ fig : tadpoles]c , d ) are minimized . in this case the log enhanced contribution to any amplitude can be obtained entirely from diagrams with insertions of the local operators @xmath318 , @xmath321 , with coefficients @xmath216 evaluated at a scale @xmath322 .
this procedure amounts essentially to removing the short distance modes of the gravitational field with wavelengths between the internal distance @xmath0 and the orbital scale @xmath1 . .
all particle vertices are the leading order mass @xmath323 mass insertion .
the @xmath324 denotes an insertion of the potential graviton kinetic term.,width=151 ] for example , consider the diagram shown in fig .
[ fig : v6small ] , which contributes to the @xmath31 terms in the potential .
schematically , it contains terms of the form @xmath325,\end{aligned}\ ] ] where @xmath326 denotes the contributions that arise due to insertions of the non - minimal curvature couplings and the logarithmic piece arises from a calculation similar to that of the @xmath327 terms of the background graviton effective action in section [ sec : rg ] .
although it is hard in general to calculate those logarithmic pieces , their argument is invariably of the form @xmath320 , so by choosing @xmath322 we may eliminate these pieces in favor of the effective couplings @xmath328 .
these in turn may be obtained in terms of the bare parameters @xmath329 by using the rg equations for the couplings in @xmath216 in eq .
( [ eq : rgcr ] ) , eq .
( [ eq : rgcv ] ) , @xmath330 since the validity of the velocity expansion requires @xmath331 , the logarithmic term gives the dominant contribution to the two - body potential at this order in @xmath3 .
following this line of reasoning , we can obtain the model independent @xmath31 terms by calculating the diagrams of fig .
[ fig : v6logv1 ] , fig .
[ fig : v6logv2 ] and evaluating the coefficients @xmath220 at the scale @xmath322 .
we find @xmath332\\ & & { } + \left\{4 g_n m_2 c^{(1)}_r(|{\bf x}_{12}|)\left[{{\bf v}^2_2\over |{\bf x}_{12}|^3}- { 3 ( { \bf v}_2\cdot { \bf x}_{12})^2 \over |{\bf x}_{12}|^5}\right ] + ( 1\leftrightarrow 2)\right\ } - { 4g^2_n m_1 m_2\over |{\bf x}_{12}|^4 } \left(c^{(1)}_r(|{\bf x}_{12}| ) + c^{(2)}_r(|{\bf x}_{12}|)\right ) .
\end{aligned}\ ] ] in this equation , the first and second terms follow from the diagrams with a single graviton exchange in fig .
[ fig : v6logv1 ] , while the last term comes from the diagrams with two and three graviton propagators , fig .
[ fig : v6logv2 ] . for the diagrams involving insertions of the operator
@xmath333 one finds that the one - graviton exchange diagrams vanish identically while the diagrams from fig .
[ fig : v6logv2 ] cancel , so that the two - body potential is independent of the coefficients @xmath334 .
we have dropped parts of diagrams that give rise to a contact delta function potential that does not generate long range forces , as well as terms that can be expressed as a total time derivative and therefore do not contribute to the equations of motion .
note in particular that had we not chosen the subtraction scale @xmath209 to cancel the logarithms coming from the divergent integrals we would have needed , in order to obtain eq .
( [ eq : v6logpot ] ) , a contribution from a diagram involving the quintic vertex in @xmath73 from the einstein - hilbert lagrangian . by choosing @xmath322
, this piece of the potential arises instead from the diagram in fig .
[ fig : v6logv2]f , which is much easier to calculate ) , eq .
( [ eq : rgcv ] ) are calculated in background field gauge , as in the previous section . ] . to see this note that the insertion of the part of the operator @xmath318 that is quadratic in @xmath94 in the diagram of fig .
[ fig : v6logv2]f is related by gauge invariance to the short distance piece ( the divergent log ) that is generated in fig .
[ fig : tadpoles ] . defining the binding energy of the binary system in terms of the two - body lagrangian as @xmath335 we find for a circular orbit of radius @xmath1 in the cm frame @xmath129 , @xmath336 where @xmath337 , @xmath338 , and @xmath339 is a length scale of order the radii of the binary star constituents , which we assume for simplicity to be of similar size .
one can check that this result agrees with the energy of a test particle of mass @xmath209 in circular motion around a schwarzschild black hole of mass @xmath32 ( in the coordinates used in section [ sec : rg ] ) in the limit @xmath340 .
this requires knowledge of the logarithmic term in @xmath341 proportional to @xmath342 @xmath343 . because @xmath1 is not a physical quantity , it is more natural to express the binding energy @xmath12 in terms of the orbital frequency @xmath344 which can be measured by observers at infinity .
a short calculation shows that the contribution to @xmath345 due to the coefficients @xmath220 exactly cancels , which by rg invariance also implies that the logarithmic dependence on @xmath209 must disappear .
this follows from the metric field redefinitions alluded to earlier .
if we write the metric @xmath52 in terms of a new metric @xmath89 as @xmath346,\ ] ] we find that to linear order in @xmath347 @xmath348=-2m^2_{pl}\int d^4 x \sqrt{{\bar g } } r[{\bar g } ] + \xi\int d\tau r,\ ] ] so that the theory with curvature coupling @xmath349 is equivalent to a theory with coupling @xmath350 .
since @xmath347 can be chosen arbitrarily without affecting physical predictions , it can be set to @xmath351 in which case the effects of the operator @xmath318 are completely removable ) induces in the two - body sector contact interactions of the form @xmath352 which are physical , although irrelevant in classical processes . ] .
similar field redefinitions can be performed to remove @xmath353 .
this conclusion can also be reached at the level of eq .
( [ eq : v6logpot ] ) . since to derive eq .
( [ eq : v6logpot ] ) we had to fix a choice of gauge , a shift in the metric becomes equivalent to a gauge transformation .
it is then easy to see that if we let @xmath354 , ( @xmath47 ) with @xmath355 and @xmath356 the shift in the leading order newtonian lagrangian cancels the @xmath4 potentials calculated above . beyond these spurious operators , any other operator involving
the ricci tensor is also removable by suitable metric redefinitions .
this means that only operators built from the weyl tensor have any physical content .
the simplest physical operators that can be written down in a theory of spinless particles are then @xmath357 where we did not build any operators using the four - acceleration @xmath358 , since such terms can be set to zero by a redefinition of the point particle worldline .
the nrgr power counting rules then indicate that these operators do not give finite size contributions to the gravitational wave observables until at least order @xmath5 .
this is consistent with newtonian expectations .
a complete account of finite size effects in the point particle eft must also involve a discussion of absorptive effects , which for spinless black hole binaries in the extreme mass limit are known to arise first at order @xmath359 @xcite .
such effects are also encoded in point particle operators , although there are subtleties involved in the construction which we will discuss in a separate paper .
in this paper , we have reformulated the nr expansion for binary star systems , an important class of signals for the gravitational wave detection experiments , in the language of effective field theory .
the eft approach is useful for the binary problem because it allows the different length scales that arise in the binary star to be treated independently . in particular
, it is possible to separate the model dependent aspects of the problem which involve the internal structure of the binary constituents from the gravitational dynamics that governs the orbital evolution and radiation emission . using the eft formulation , it is easy to show that the divergences which arise at @xmath4 in the pn expansion can be attributed to the existence of new operators in the effective point particle description .
however , these operators can be removed via a point transformation of the metric tensor and thus never contribute to physical quantities .
this leads to the conclusion that there are no finite size effects at order @xmath4 .
practically , this means that _ whenever one encounters a log divergent integral at order @xmath4 in the potential , one may simply set it to zero . its value can not affect physical predictions .
_ beyond order @xmath4 , a new set of operators , which are quadratic in the curvature and whose role is to describe tidal deformations ( induced moments ) , contribute at @xmath5 .
the magnitude of such tidal forces is consistent with conventional newtonian estimates of their size .
however , in the eft these conclusions follow from general principles ( dimensional analysis and general covariance ) without the need to carry out any calculations . as will be discussed in a future publication , the calculations necessary to determine the coefficient of these operators ( which encode the finite size effects ) are relatively straightforward , since they can be done for a compact object in isolation without the need to include superfluous information about the complete two - body problem .
this is the beauty of working within an eft formalism .
several aspects of the formalism remain to be worked out . here
we have not discussed logarithmic divergences that may arise in nrgr due to graphs with internal radiation graviton lines .
such divergences can be treated by the same renormalization and matching procedure described in this paper to deal with divergences in the calculation of potential terms .
in fact , we believe that the explicit decoupling of the orbital scale @xmath1 and the radiation scale @xmath2 will make the evaluation of effects such as radiative tail terms , which is difficult within conventional pn approaches @xcite , simpler within our approach . also , in our analysis so far , we have neglected the role of spins or higher rank tensors carried by the point particles .
such permanent moments encode the intrinsic , rather than the tidally induced , shape of the binary star constituents .
it is important to extend the formalism , in particular the velocity power counting rules , to systematically include the effects of these additional worldline degrees of freedom on the pattern of gravitational waves seen by the detector .
we expect also that the inclusion of these dynamical quantities will allow the construction of a vastly larger set of coordinate invariant worldline non - minimal couplings than those considered in this paper , and lead to an even richer structure of coupling constant rg flows
. finally it should be interesting to apply the formalism to work out the phenomenological signatures of theories of gravity that predict deviations from general relativity , including simple extensions with additional fields or more speculative models that deviate from conventional gravity in the far infrared @xcite .
we thank aneesh manohar and kip thorne for helpful discussions .
wg is supported in part by the department of energy under grant de - fg03 - 97er40546 .
wg carried out part of this research as a postdoctoral fellow at uc berkeley and thanks the financial support of the berkeley theory group .
izr would like to thank the theory groups at ucsd and uc berkeley for their hospitality .
izr is supported in part by the department of energy under grants doe - er-40682 - 143 and deac02 - 6ch03000 .
* note added : * while this paper was being reposted , ref .
@xcite appeared which discusses the same field redefinitions for removing the operators @xmath318 , @xmath360 mentioned above .
in this appendix we show that any diagram contributing to @xmath56 $ ] scales at most as @xmath361 , up to powers of the velocity . consider a vacuum diagram with @xmath362 graviton self - interaction vertices from the einstein - hilbert action , @xmath363 insertions of the point particle worldline , and @xmath13 graviton propagators .
writing the graviton kinetic term as @xmath364 , each graviton self - interaction scales as @xmath365 , a coupling to the worldlines scales as the particle mass @xmath32 , and each graviton propagator scales as @xmath366 . up to powers of @xmath3 ,
the only other scale that can appear in a feynman diagram is the orbital distance @xmath1 .
thus a typical diagram in the expansion of @xmath56 $ ] has a magnitude @xmath367 since @xmath56 $ ] is dimensionless . using the virial theorem , @xmath368 , so up to powers of @xmath3
, this equation implies that each diagram scales as @xmath369 where @xmath370 is the total number of vertices in each connected vacuum diagram that contributes to @xmath56 $ ] . starting with the trivial diagram with @xmath371 ( corresponding to a free worldline ) one can attach to it enough vertices and graviton lines to build up a diagram with arbitrary @xmath372 .
however , since the diagram obtained in this way must be connected , for each new vertex we must include at least one graviton line , so that @xmath373 for this new graph is bounded by unity . from this
we conclude that any diagram contributing to @xmath56 $ ] scales as @xmath374 , with integer @xmath375
in this appendix we review the derivation of the feynman rules used to obtain the results of this paper .
as explained in the text , in our formalism all observables are derivable from either the quantity @xmath56 $ ] defined by eq .
( [ eq : zosopi ] ) or from the effective lagrangian for the radiation graviton mode , @xmath376 $ ] .
note that because there is no propagator for the fields " @xmath377 ( as far as the gravitons are concerned , these degrees of freedom are treated as a fixed background ) we have the rule thus diagrams like the newton exchange term in fig .
[ fig : newt]a are kept in the expansion , but diagrams such as a box " diagram with a pair of worldlines interchanging two gravitons do not arise in the expansion of @xmath56 $ ] .
in fact , in our formalism the box diagram corresponds to the product of two single exchange terms , and is already accounted for by keeping just fig .
[ fig : newt]a . the procedure for integrating out modes with wavelengths shorter than the radiation scale @xmath2 involves computing diagrams with one or more external radiation gravitons and any number of internal potential gravitons .
this gives rise to the quantity @xmath378 $ ] defined in eq .
( [ eq : nrgrpi ] ) . the rule for calculating
this is due to the multipole expansion , this quantity is guaranteed to be of the general form @xmath381=\int dt l[{\bar h}({\bf x}_{cm},t),{\bf x}_a]$ ] , as we explicitly saw in the derivation of the low order terms in @xmath382 $ ] .
if we expand the metric perturbation @xmath383 in terms of potential and radiation graviton modes as @xmath384 then the propagators for @xmath385 and @xmath75 can be read off the terms in the gravitational action that are quadratic in the fields . in order to get a well defined propagator one needs to fix the gauge , which is achieved by adding a term to the action that explicitly breaks diffeomorphisms this procedure is standard for gauge theories and
can be found for instance in chapter 16 of ref .
@xcite . for gravity
this is explained in the reviews @xcite .
given the choice of gauge fixing term for potential gravitons used in the text , eq .
( [ eq : pgauge ] ) the terms quadratic in the potential mode can be written as @xmath386 with @xmath387 .
the propagator is then defined as a formal inverse inverse to this equation @xmath388 with @xmath389 the identity on symmetric two - index tensors .
inverting we find eq .
( [ eq : potprop ] ) @xmath390 the propagator for the radiation graviton also depends on the choice of gauge fixing term in the action . for instance , choosing a gauge fixing term of the form @xmath391=
m^2_{pl}\int d^d x \eta^{\mu\nu } { \bar\gamma}_\mu { \bar\gamma}_\nu,\ ] ] with @xmath392 the propagator is given by @xmath393 where the feynman propagator @xmath66 is given by @xmath394 in this equation , @xmath395 defines a choice of contour for the @xmath215 integral .
given these results , the general rule for a given feynman diagram is then : * an internal potential graviton line with label momentum @xmath170 and connecting points at times @xmath396 and @xmath397 corresponds to an insertion of eq .
( [ eq : pprop ] ) . *
an internal radiation line connecting two points @xmath398 and @xmath399 denotes a factor of eq .
( [ eq : rprop ] ) .
they feynman rules for general relativity contain an infinite number of interaction vertices .
this is because the einstein - hilbert lagrangian @xmath402 is an infinite series in the graviton @xmath54 , so that the set of terms in this expansion with @xmath379 factors of @xmath54 gives rise to an @xmath379-graviton vertex in feynman diagrams . in our theory , the expansion of the point particle couplings also generates an infinite set of vertices that can be used to build up feynman diagrams .
the method for deriving the vertex rules associated with these interactions is straightforward , although in the case of graviton self - interactions arising from the gravitational action the procedure gets messy due to the large number of terms at a given order in @xmath54 . in this section we will illustrate the procedure by deriving some of the vertex rules used in the text . the derivation of more complicated feynman rules is directly analogous .
consider first feynman rules arising from insertions of point particle interactions .
the simplest non - trivial term is the vertex that arises in the calculation of the two - body newton interaction . expanding the @xmath403 term in the action
, there is a term of the form @xmath404 this leads to a feynman rule for the vertex shown in fig .
11 @xmath405 then for instance the diagram of fig .
[ fig : newt]a is given by @xmath406 \left[-{i m_2\over 2 m_{pl}}\int dx'^0 \int_{\bf q } e^{i{\bf q}\cdot { \bf x}_2(x'^0)}\eta_{\alpha0}\eta_{\beta0}\right ] \langle h^{\mu\nu}_{\bf k}(x^0 ) h^{\alpha\beta}_{\bf q}(x'^0)\rangle,\ ] ] which reduces to eq .
( [ eq : newta ] ) . likewise , the term in the action @xmath407 leads to the feynman rule @xmath408 which can be used to calculate fig . [ fig : eih1]b .
the point particle action has also terms that are non - linear in the graviton and lead to vertices with more graviton lines .
for instance , the vertex in fig .
[ fig : eih2]b involving two graviton lines arises from a term in the action @xmath409 the feynman rule for this term contains two identical terms , which arise from the two independent ways of contracting the two fields in the interaction with the external graviton lines ( in other words , of pairing the two graviton lines with index pairs @xmath410 , @xmath411 with the two fields in the interaction ) @xmath412 using this vertex , we can then calculate the graph in fig .
[ fig : eih2]b . including the proper symmetry factors for the diagram
( see for instance chapter 4 of ref .
@xcite ) leads to eq .
( [ eq : eih5 ] ) . besides the interactions vertices involving point particle sources
, we also need to calculate the @xmath379-graviton vertices .
consider for instance the 3-point vertex for radiation gravitons . due to the large number of terms each with a complicated tensor index structure
, we do not display the full feynman rule here ( it is more efficient to do the calculations using an algebra manipulation computer package ) .
consider , however one of the terms that gives rise to the 3-graviton feynman rule .
in the gauge of eq .
( [ eq : rgauge ] ) all terms in this vertex derive from the cubic terms in the expansion of @xmath402 .
for instance , using @xmath413\eta^{\mu\nu}\\ & & -{1\over 2 m^2_{pl } } { \bar h } { \bar h}^{\mu\nu}+\cdots,\end{aligned}\ ] ] there is a term in the einstein - hilbert lagrangian of the form the tensor @xmath418 has been defined below eq .
( [ eq : propeqn ] ) . in this vertex , each incoming graviton
is labeled 1 , 2 , or 3 .
the partial derivatives in this particular term act on the propagator of the graviton labeled 3 coming into the vertex . to get the full feynman rule from the term in eq .
( [ eq : part ] ) , bose statistics also demands that we include all permutations of the terms shown in the above expression under the interchange of the graviton labels .
it is also useful to work in momentum space , in which case the expression for this piece of the 3-graviton feynman rule becomes ( taking all graviton momenta to be flowing into the vertex ) a. einstein , l. infeld and b. hoffmann , annals math .
* 39 * , 65 ( 1938 ) .
r. v. wagoner and c. m. will , astrophys .
j. * 210 * , 764 ( 1976 ) [ erratum - ibid .
* 215 * , 984 ( 1977 ) ] ; m. walker and c. m. will , phys .
* 45 * , 1741 ( 1980 ) ; l. blanchet , t. damour and g. schafer , mon . not .
* 242 * , 289 ( 1990 ) . c. cutler _ et al .
_ , phys .
* 70 * , 2984 ( 1993 ) [ arxiv : astro - ph/9208005 ] ; c. cutler and e. e. flanagan , phys .
d * 49 * , 2658 ( 1994 ) [ arxiv : gr - qc/9402014 ] ; l. s. finn , phys .
d * 46 * , 5236 ( 1992 ) [ arxiv : gr - qc/9209010 ] ; l. s. finn and d. f. chernoff , phys .
d * 47 * , 2198 ( 1993 ) [ arxiv : gr - qc/9301003 ] ; e. poisson and c. m. will , phys .
d * 52 * , 848 ( 1995 ) [ arxiv : gr - qc/9502040 ] .
a. g. wiseman , phys .
d * 48 * , 4757 ( 1993 ) ; l. blanchet and g. schafer , class .
* 10 * , 2699 ( 1993 ) ; l. blanchet and t. damour , phys .
d * 46 * , 4304 ( 1992 ) .
t. damour and n. deruelle , phys .
a * 87 * , 81 ( 1981 ) ; c.r .
paris * 293 * , 537 ( 1981 ) ; c.r .
acad . sc .
paris * 294 * , 1355 ( 1982 ) ; l. bel , t. damour , n. deroulle , j. ibanez , and j. martin , gen .
. grav . * 13 * , 963 ( 1981 ) ; t. damour and g. schafer , gen .
* 17 * , 879 ( 1985 ) .
blanchet , t. damour and b. r. iyer , phys .
d * 51 * , 5360 ( 1995 ) [ erratum - ibid .
d * 54 * , 1860 ( 1996 ) ] [ arxiv : gr - qc/9501029 ] ; c. m. will and a. g. wiseman , phys .
d * 54 * , 4813 ( 1996 ) [ arxiv : gr - qc/9608012 ] ; l. blanchet , b. r. iyer , c. m. will and a. g. wiseman , phys .
lett . * 74 * , 3515 ( 1995 ) [ arxiv : gr - qc/9501027 ] ; class .
* 13 * , 575 ( 1996 ) [ arxiv : gr - qc/9602024 ] .
kopejkin , astron .
62 * , 889 ( 1985 ) ; l. blanchet , g. faye and b. ponsot , phys . rev .
d * 58 * , 124002 ( 1998 ) [ arxiv : gr - qc/9804079 ] ; m. e. pati and c. m. will , phys .
d * 62 * , 124015 ( 2000 ) [ arxiv : gr - qc/0007087 ] ; phys .
d * 65 * , 104008 ( 2002 ) [ arxiv : gr - qc/0201001 ] ; y. itoh , t. futamase and h. asada , phys .
d * 62 * , 064002 ( 2000 ) [ arxiv : gr - qc/9910052 ] ; phys .
d * 63 * , 064038 ( 2001 ) [ arxiv : gr - qc/0101114 ] ; l. blanchet , phys .
d * 54 * , 1417 ( 1996 ) [ arxiv : gr - qc/9603048 ] ; l. blanchet , b. r. iyer and b. joguet , phys .
d * 65 * , 064005 ( 2002 ) [ arxiv : gr - qc/0105098 ] ; l. blanchet , g. faye , b. r. iyer and b. joguet , phys .
d * 65 * , 061501 ( 2002 ) [ arxiv : gr - qc/0105099 ] .
p. jaranowski and g. schafer , phys .
d * 57 * , 7274 ( 1998 ) [ erratum - ibid .
d * 63 * , 029902 ( 2001 ) ] [ arxiv : gr - qc/9712075 ] ; phys .
d * 60 * , 124003 ( 1999 ) [ arxiv : gr - qc/9906092 ] ; t. damour , p. jaranowski and g. schafer , phys .
d * 62 * , 021501 ( 2000 ) [ erratum - ibid .
d * 63 * , 029903 ( 2001 ) ] [ arxiv : gr - qc/0003051 ] ; phys .
d * 63 * , 044023 ( 2001 ) [ arxiv : gr - qc/0010009 ] ; l. blanchet and g. faye , phys .
a * 271 * , 58 ( 2000 ) ; [ arxiv : gr - qc/0004009 ] .
d * 63 * , 062005 ( 2001 ) [ arxiv : gr - qc/0007051 ] . v. c. de andrade , l. blanchet and g. faye , class .
grav . * 18 * , 753 ( 2001 ) [ arxiv : gr - qc/0011063 ] .
l. blanchet and b. r. iyer , class .
grav . * 20 * , 755 ( 2003 ) [ arxiv : gr - qc/0209089 ] .
t. damour , p. jaranowski and g. schafer , phys .
d * 63 * , 044021 ( 2001 ) [ erratum - ibid .
d * 66 * , 029901 ( 2002 ) ] [ arxiv : gr - qc/0010040 ] ; t. damour , p. jaranowski and g. schafer , phys .
d * 62 * , 021501 ( 2000 ) [ erratum - ibid .
d * 63 * , 029903 ( 2001 ) ] [ arxiv : gr - qc/0003051 ] ; l. blanchet , t. damour , g. esposito - farese and b. r. iyer , arxiv : gr - qc/0406012 .
t. damour , p. jaranowski and g. schafer , phys .
b * 513 * , 147 ( 2001 ) [ arxiv : gr - qc/0105038 ] .
l. blanchet , t. damour and g. esposito - farese , phys . rev .
d * 69 * , 124007 ( 2004 ) [ arxiv : gr - qc/0311052 ] ; w. e. caswell and g. p. lepage , phys .
b * 167 * , 437 ( 1986 ) ; m. e. luke , a. v. manohar and i. z. rothstein , phys .
d * 61 * , 074025 ( 2000 ) [ arxiv : hep - ph/9910209 ] .
for a recent review , see i. z. rothstein , arxiv : hep - ph/0308266 .
e. poisson and m. sasaki , phys .
d * 51 * , 5753 ( 1995 ) [ arxiv : gr - qc/9412027 ] . for reviews , see j. f. donoghue , arxiv : gr - qc/9512024 ; c. p. burgess , arxiv : gr - qc/0311082 . j. f. donoghue and t. torma , phys .
d * 54 * , 4963 ( 1996 ) [ arxiv : hep - th/9602121 ] .
t. damour and g. esposito - farese , phys .
d * 58 * , 042001 ( 1998 ) [ arxiv : gr - qc/9803031 ] .
k. s. thorne and j. b. hartle , phys .
d * 31 * , 1815 ( 1984 ) .
m. veltman , in _ methods in field theory , proceedings of the les houches summer school , 1975 , _ eds .
r. balian and j. zinn - justin , north holland , 1976 .
b. grinstein and i. z. rothstein , phys .
d * 57 * , 78 ( 1998 ) [ arxiv : hep - ph/9703298 ] .
g. modanese , nucl .
b * 434 * , 697 ( 1995 ) [ arxiv : hep - th/9408103 ] ; i. j. muzinich and s. vokos , phys .
d * 52 * , 3472 ( 1995 ) [ arxiv : hep - th/9501083 ] .
n. e. j. bjerrum - bohr , j. f. donoghue and b. r. holstein , phys .
d * 67 * , 084033 ( 2003 ) [ arxiv : hep - th/0211072 ] ; h. w. hamber and s. liu , phys .
b * 357 * , 51 ( 1995 ) [ arxiv : hep - th/9505182 ] ; n. e. j. bjerrum - bohr , j. f. donoghue and b. r. holstein , phys .
d * 68 * , 084005 ( 2003 ) [ arxiv : hep - th/0211071 ]
. w. fischler , nucl .
b * 129 * , 157 ( 1977 ) ; t. appelquist , m. dine and i. j. muzinich , phys .
b * 69 * , 231 ( 1977 ) ; phys .
d * 17 * , 2074 ( 1978 ) .
m. e. peskin , nucl .
b * 156 * , 365 ( 1979 ) ; for a review see m. e. peskin , slac - pub-3273 , _ presented at 11th int .
slac summer inst . on particle physics , stanford , ca , jul 18 - 26 , 1983 .
_ s. n. solodukhin , nucl .
b * 541 * , 461 ( 1999 ) [ arxiv : hep - th/9801054 ] ; w. d. goldberger and m. b. wise , phys .
d * 65 * , 025011 ( 2002 ) [ arxiv : hep - th/0104170 ] .
b. s. dewitt , phys .
162 * , 1195 ( 1967 ) ; l. f. abbott , nucl .
b * 185 * , 189 ( 1981 ) .
g. r. dvali , g. gabadadze and m. porrati , phys .
b * 485 * , 208 ( 2000 ) [ arxiv : hep - th/0005016 ] .
n. arkani - hamed , h. c. cheng , m. a. luty and s. mukohyama , jhep * 0405 * , 074 ( 2004 ) [ arxiv : hep - th/0312099 ] .
m. e. peskin and d. v. schroeder , _ an introduction to quantum field theory , _ perseus books , 1995 . | using effective field theory ( eft ) methods we present a lagrangian formalism which describes the dynamics of non - relativistic extended objects coupled to gravity .
the formalism is relevant to understanding the gravitational radiation power spectra emitted by binary star systems , an important class of candidate signals for gravitational wave observatories such as ligo or virgo .
the eft allows for a clean separation of the three relevant scales : @xmath0 , the size of the compact objects , @xmath1 the orbital radius and @xmath2 , the wavelength of the physical radiation ( where the velocity @xmath3 is the expansion parameter ) . in the eft radiation
is systematically included in the @xmath3 expansion without need to separate integrals into near zones and radiation zones . using the eft
, we show that the renormalization of ultraviolet divergences which arise at @xmath4 in post - newtonian ( pn ) calculations requires the presence of two non - minimal worldline gravitational couplings linear in the ricci curvature .
however , these operators can be removed by a redefinition of the metric tensor , so that the divergences at arising at @xmath4 have no physically observable effect .
because in the eft finite size features are encoded in the coefficients of non - minimal couplings , this implies a simple proof of the decoupling of internal structure for spinless objects to at least order @xmath4 . neglecting absorptive effects
, we find that the power counting rules of the eft indicate that the next set of short distance operators , which are quadratic in the curvature and are associated with tidal deformations , do not play a role until order @xmath5 .
these operators , which encapsulate finite size properties of the sources , have coefficients that can be fixed by a matching calculation . by including the most general set of such operators , the eft allows one to work within a point particle theory to arbitrary orders in @xmath3 . |
wetting and spreading@xcite are very common everyday phenomena , but there are still a lot of open problems in this field .
a simple example is - what happens when a fluid is forced to spread under an externally impressed force ?
there is no fully satisfactory answer to be found in the otherwise comprehensive literature on the topic , though such experiments are routinely used to characterize viscoelastic materials . in the present paper ,
we report a set of very simple experiments on the so - called squeeze film and try to find a theoretical framework to explain our results .
the problem is briefly described as follows - a small droplet of a liquid , placed on a solid plate will take the shape of a section of a sphere . under equilibrium
, the well known young - laplace relation holds , involving an equilibrium angle of contact @xmath4.the dynamic condition while the fluid is spreading , i.e. the area of contact of the fluid and solid is increasing in time , is also well studied .
there are a number of excellent reviews summarising studies on - the rate of spreading , the forces responsible and other interesting features @xcite .
we now look at a simple extension of the problem - what happens if the fluid is forced to spread beyond its equilibrium area , under a constant weight placed on top of it .
a preliminary report was published by our group in @xcite . in this paper
we study four different fluids on two different solids , so as to have different combinations of viscosities of fluids as well as different degrees of hydrophilicity of the fluids and solids .
the results are analyzed on the basis of viscous energy dissipation with and without slip@xcite .
we find that while the no - slip result gives a fair quantitative estimate of the contact area as function of time for only one case , it is possible to get a better agreement in 5 of the 8 cases , on introducing a slip coefficient .
of course , this involves introduction of a free parameter ( the slip coefficient ) whereas the no - slip analysis did not .
the slip coefficient appears to depend more on the solid surface than the fluid , and is an order of magnitude lower for the perspex plate .
the results for spreading of the same fluid on different surfaces is quite different .
there are 3 cases where the agreement between theory and experiment is unsatisfactory and can not be improved by introducing slip , so other effects are also at play .
we describe the experimental setup followed by the theory and present detailed results in the third section .
finally we compare experimental results with theory and discuss the implications .
the experimental fluids are castor oil , linseed oil , glycerol and ethylene glycol .
the choice of fluids is motivated by the following idea . while castor oil and linseed oil are insoluble in water , with a low dielectric constant @xmath5 the other two fluids ethylene glycol and glycerol having high dielectric constants , are soluble in water .
( @xmath5 = 4.67 , 3.35 , 47.70 and 43.0 respectively for castor oil , linseed oil , glycol and glycerol ) . again
, one fluid belonging to each type according to hydrophilicity , has high viscosity(@xmath6 ) , namely castor oil and glycerol(@xmath6 for castor oil=986 - 451 cp for temp 20 - 30c , glycerol is 1490 - 629 cp for temp 20 - 30c ) , while the others have much lower viscosity(@xmath6 for glycol is 19.9 - 9.13 cp for temp 20 - 40c , linseed oil=33.1 - 17.6 cp for temp 20 - 50c ) . a constant volume @xmath7 of the fluid is measured with a micropipette and placed on the lower plate .
the upper plate is then placed on the drop and a load of m kg .
( m = 1 to 5 ) is placed on it .
the pair of plates are @xmath8 1 cm thick , one pair is made of float glass and the other of perspex . a video camera , placed below the lower glass plate records the slow increase in area .
the video clips are analyzed using image pro - plus software .
we find that the area of the drop in contact with the plate increases with time at first .
then after some time it reaches a saturation value and does not increase significantly with time any more .
the time of saturation and area of saturation are different for different fluid - solid combinations .
we plotted the area vs. time graph for all fluid - solid combinations .
fig([tacp ] ) and fig([tagp ] ) show respectively area vs. time graphs for castor oil and glycol respectively on perspex for loads from 1 - 5 kg .
interestingly when the data for area are divided by the load mass m , the curves come very close to each other , almost collapsing to a single master curve . again , for castor oil on perspex area increases for a long time .
it takes several minutes to saturate approximately and continue to increase very slowly even after 5 minutes.for glycol on both the substrates , on the other hand , the area increases very fast and saturates in a few seconds .
video photos are taken initially , at the rate of 10 frames / sec , in this case .
we discuss in this section standard formulations for this kind of problem .
the squeeze flow usually discussed in chemical engineering literature is different from our experiment , ours is a constant volume squeeze , where whole mass of fluid taken is always confined within the two plates . in the usual experiment a _ constant area _
setup is used , where both plates are immersed in the fluid , the distance between the plates as function of time is measured as a force is applied on the upper plate .
first we outline the steps worked out for the no - slip case , i.e. assuming that the layer of fluid next to the static plate has zero velocity .
the analysis roughly follows @xcite .
we assume the fluid to be newtonian and incompressible with constant volume @xmath9 .
ours is a constant force set up , the force @xmath10 acts on the fluid layer
. the velocity first builds up , reaches a maximum and then starts to reduce as it squeezes the liquid film .
the details are presented in @xcite , here we consider the velocity decreasing regime only .
we equate the loss of potential energy of the loaded plate with the work done against the viscous resistance . with @xmath11 and @xmath12 representing the instantaneous thickness and radius of the film respectively , @xmath13 the pressure and @xmath14 the atmospheric pressure . in moving over a distance dh
the work done is given by @xmath15 the loss of potential energy is @xmath16 now the pressure p at any point r from the center of the upper plate for any given distance h , where the velocity of the upper plate @xmath17 can be calculated as @xmath18 therefore , we can write , @xmath19 substituting for @xmath20 , we get , @xmath21 which , on integration gives , @xmath22 @xmath23 @xmath24\ ] ] @xmath25 the right hand expression depends on the impressed force @xmath10 through the area of contact @xmath26 .
if we divide both sides by @xmath27 , the lhs is independent of @xmath27 .
the rhs is now a function of @xmath28 and @xmath29 .
we call this function @xmath30 .
we calculate the rhs from experimental data and compare with the lhs .
if the principles on which the derivation is based are correct , the rhs calculated from experimental results should be independent of @xmath27 .
that is the curves @xmath30 vs. @xmath29 should collapse to a single curve and be numerically equal @xmath31 .
so , we test the result @xmath32 by plotting the lhs and rhs separately on the same graph and check how well they agree .
of course the agreement can be expected only for the time interval before the area of contact saturates , since the lhs goes on increasing linearly .
the results for glycerol on glass are shown in fig([gmt_glcrg ] ) .
we see that for this combination assuming no slip gives an acceptable quantitative agreement between the lhs and rhs of eq([gmt ] ) .
this calculation involves no adjustable parameter .
it may further be noted that the data for different loads collapse almost to a single line , on scaling by @xmath27 .
however , for the other fluid - solid combinations , such quantitative agreement is not found .
we try to see now if this discrepancy may be attributed to the effect of slip at the fluid and solid interface .
we follow the equation given by @xcite listed in @xcite for this situation .
relative motion between a solid surface and material contacting it ( wall slip ) is a phenomenon observed with many materials @xcite . in the last section we assumed that the layer of the liquid touching the lower fixed plate has zero velocity . so that the applied force is entirely carried away by viscous dissipation . but experiments show that there may be cases where the layer of the liquid having contact with the substrate may have finite velocities , which may affect the spreading of the liquid drop . in these cases , shear stresses at the sample - plate interfaces may vanish ( for perfect slip ) or reduce ( for partial slip ) .
the amount of slip depends not only on the liquid tested , but also on the material and roughness of the substrates .
we introduce partial slip in the analysis as follows .
no slip and perfect slip are two limiting cases of slip phenomenon .
to explain our results , we take partial slip into consideration . stefan@xcite gave a slip law of the form , @xmath33 where , @xmath34 is the slip coefficient and @xmath35 is the wall shear stress .
the value of @xmath34 is zero for no slip and is infinity for perfect slip cases .
it may have any value in between , for partial slip cases , depending on the liquid and the substrate .
with the introduction of @xmath34,the force equation becomes , @xmath36 @xmath37 @xmath38 dividing by @xmath27 .
@xmath39 where , the integral can be evaluated as @xmath40_{h1}^{h2}\ ] ] with @xmath41 the above force equation , on integration , gives us the function @xmath42 similar to @xmath30 for the no - slip case .
we fit the slip coefficient @xmath34 so as to get the best fit between the left and right sides of equation [ rmt ] .
the results are shown in the next section .
table([tabl ] ) shows values of @xmath34 for different fluid - solid combinations .
we find that the values of @xmath34 ranges from @xmath43 to @xmath44 for glass and from @xmath44 to @xmath45 for perspex .
so glass has a larger @xmath34 than perspex . for a definite substrate
the values of @xmath34 are different for different liquids . in table 1
we also show the initial contact area @xmath46 ( before loading ) and the viscosities and dielectric constants of the fluids and substrates . figures ( [ rmt_gg ] ) and ( [ rmt_cp ] ) show respectively the graphs comparing @xmath42 and @xmath31 for ethylene glycol on glass and for castor oil on perspex .
the coefficients @xmath34 for the best fit are given in table 1 .
while five of the eight curves can be fit quite well with a finite @xmath34 , in the remaining three cases a discrepancy remains . here
the fit becomes worse for any finite @xmath34 .
this is shown for glycol on perspex in figure([rmt_gp ] ) , the other two cases are glycol on glass and linseed oil on perspex .
out of eight cases studied , we find that the theoretical framework considered here can account for five cases fairly well with one free parameter , the slip coefficient @xmath34 .
this is not bad considering the complexity of the problem and the simplifications introduced .
scrutiny of the table 1 , reveals several interesting features .
let us look at @xmath46 , the area of contact with only the plate on the fluid drop and no extra weights .
we consider this as the reference condition .
the mass of the perspex plate is 250 gm and the glass plate is about 700 gms .
it is seen from spreading vs. load data ( see @xcite ) that extrapolating to zero load does not change this reference point significantly .
we see that for glass @xmath46 ranges from 2 - 5 cm@xmath1 , while for perspex we have values less than 1 and the highest is 1.4 cm@xmath1 .
so equilibrium spreading on glass is higher in all cases .
however , _ under load _ fluids spread to a much greater extent on perspex , compared to glass .
looking at the percentage ( of @xmath46 ) spreading for the 5 kg .
load , we see that for glass it is always less than 50 % , whatever the fluid , while for perspex the area increases by 100 to 400 % .
so , where the initial spreading is large , the effect of load is less and vice versa .
the highest spreading with load is observed for castor oil on perspex , moreover castor oil shows a creep like behavior and continues to spread for very long times , though the rate slows down .
possibly visco - elastic effects also play important role and should be taken into consideration .
the three cases which can not be explained by the present theory are all for the low viscosity fluids ethylene glycol and linseed oil .
this treatment focussing on the viscous effect works better in case of high viscosity .
we have tabulated the dielectric constants along with other data in table 1 , because earlier work @xcite showed that the polarizabilties of the fluid and substrate are important in spreading phenomena .
but here we find no significant correlation except that the maximum spreading ( under load ) is for the low dielectric fluid on the low dielectric substrate . to conclude , spreading of different types of fluids on different solids is a very complex and interesting problem , still open for exploration .
very simple experimental apparatus provides a lot of food for thought in this area .
the spreading and flow characteristics are routinely used for classifying soft materials in food technology and chemical engineering , so a deeper understanding will be useful .
further experiments and theoretical study taking into account effects of rheology , surface roughness and ambient conditions is required .the table shows properties of the fluids and substrates used , @xmath6 represents the viscosity , @xmath48 and @xmath49 , the solid and liquid dielectric constants , s the surface tension .
results from the experiment are - the initial area of contact @xmath46 , the percentile spreading of area under loads of 1 and 5 kg . , the slip coefficient @xmath34 which gives a best fit for the spreading master curve . [ cols="^,^,^,^,^,^,^,^,^,^,^ " , ]
this work is supported by ugc , govt . of india .
ss is grateful to ugc for award of a research fellowship .
99 p. g. de gennes , wetting : statics and dynamics , rev .
57(1985 ) 827 - 862 william a. zisman , elaine g. shaffin , j.phys.chem.,1960,64(5),pp 519 - 524 .
d. bonn , j. eggers , j. meunier , e. rolley , wetting and spreading , to appear in rev .
mod . phys .
2009 j. engmann , c. servais , a.s .
burbridge , squeeze flow theory and applications to rheometry : a review , j. non - newt .
fluid mech .
132(2005)1 - 27 r. asthana , n. sobczak , wettability , spreading and interfacial phenomena in high - temperature coatings , jom - e 52(2000 ) soma nag , suparna dutta , sujata tarafdar , applied surface science 256 ( 2009 ) 353 - 355 f.r .
eirich , d. tabor , collisions through liquid films , proc .
( 1948 ) 566 - 581 j stefan , akad.wiss.math.nat.wien 69(2)(1874)713 - 735 s. sinha , t. dutta , s. tarafdar , eur .
j. e 25(2008 ) 267 - 275 | spreading of different types of fluid on substrates under an impressed force is an interesting problem . here
we study spreading of four fluids , having different hydrophilicity and viscosity on two substrates - glass and perspex , under an external force .
the area of contact of fluid and solid is video - photographed and its increase with time is measured .
the results for different external forces can be scaled onto a common curve .
we try to explain the nature of this curve on the basis of existing theoretical treatment where either the no - slip condition is used or slip between fluid and substrate is introduced .
we find that of the eight cases under study , in five cases quantitative agreement is obtained using a slip coefficient .
+ @xmath0condensed matter physics research centre , physics department + jadavpur university , kolkata 700032 , india + @xmath1physics department , st .
xavier s college , kolkata 700016 , india + @xmath2 corresponding author : email : sujata@[email protected] , + phone:+913324146666(extn .
2760 ) * keywords * : spreading , squeeze film , viscous flow , slip + * pacs nos . * : wetting in liquid - solid interfaces - 68.08.bc + : viscosity , experimental study - 66.20.ej |
@xmath5 let @xmath0 be an integer , such that @xmath6 .
the weak symmetry breaking task for @xmath0 processes is an inputless task , where the possible outputs are @xmath7 and @xmath8 .
a distributed protocol is said to solve the weak symmetry breaking task if in any execution without failed processes , there exists at least one process which has value @xmath7 as well as at least one process which has value @xmath8 . in the classical setting ,
the processes know their i d s , and are allowed to compare them .
it is however not allowed that any other information about i d s is used .
the protocols with this property are called _ comparison - based_. in practice this means that behavior of each process only depends on the relative position of its i d among the i d s of the processes it witnesses and not on its actual numerical value . as a special case , we note that each process must output the same value in case he does not witness other processes at all .
weak symmetry breaking is a standard task in theoretical distributed computing , and its solvability in the standard computational models is a sophisticated question which has been extensively studied . for the rest of this paper we shall fix the computational model to be the layered immediate snapshot model , see @xcite .
in this model the processes use two atomic operations being performed on shared memory .
these operations are : _ write _ into the register assigned to that process , and _ snapshot read _ , which reads entire memory in one atomic step . furthermore , it is assumed that the executions are well - structured in the sense that they must satisfy the two following conditions .
first , it is only allowed that at each time a group of processes gets active , these processes perform a write operation together , and then they perform a snapshot read operation together ; no other interleaving in time of the write and read operations is permitted .
such executions are called _ immediate snapshot _ executions .
second , each execution can be broken up in rounds , where in every round each non - faulty process gets activated precisely once .
even though this computational model is seemingly quite restrictive , it has been proved , see , e.g. , @xcite , that most of the commonly used shared memory computational models are equivalent , in the sense of which of the tasks are solvable , to this one .
as we will see , this model has the major advantage that the protocol complexes have a comparatively simple topological structure .
@xmath5 several groups of researchers have studied the solvability of the weak symmetry breaking by means of comparison - based layered immediate snapshot protocols . due to primarily work of castaeda and rajsbaum , @xcite , it is known that the weak symmetry breaking is solvable if and only if the number of processes is not a prime power ; see also @xcite for a counting - based argument for the impossibility part .
this makes @xmath9 the smallest number of processes for which this task is solvable .
the combinatorial structures arising in related questions on subdivisions of simplex paths have been studies in @xcite .
the specific case @xmath9 has been studied in @xcite , who has proved the existence of the distributed protocol which solves the weak symmetry breaking task in 17 rounds .
we refer to the classical textbook @xcite , as well as the more recent monograph @xcite , for general background on theoretical distributed computing , and specifically on impossibility results .
a general reference for topological methods in distributed computing is @xcite . a general reference for combinatorial topology is @xcite .
finally , we recommend the survey @xcite for background on symmetry breaking tasks .
@xmath5 our main mathematical result is the following theorem .
[ thm : main - math ] there exists a compliant binary labeling @xmath10 of the vertices of @xmath11 , such that the restriction of @xmath10 to any @xmath12-simplex of @xmath11 is surjective .
this immediately implies the following theorem in theoretical distributed computing .
[ thm : main - cs ] there exists a comparison - based layered immediate snapshot distributed protocol solving the weak symmetry breaking task for @xmath3 processes in @xmath4 rounds .
we also present the associated explicit distributed protocol .
because of the standard reduction to a mathematical question , see theorem [ thm : red ] , we can derive theorem [ thm : main - cs ] as a direct corollary of our main mathematical statement given in theorem [ thm : main - math ] .
our central method is the in - depth analysis of the simplicial structure of the second chromatic subdivision of a simplex . to this end
, we need to develop the apropriate combinatorial language , in order to formalize and to work with standard chromatic subdivisions .
this is done in sections [ sect:2 ] and [ sect:3 ] , yielding as a byproduct the combinatorial framework which is applicable in a much bigger generality .
once the combinatorics is clear , the plan of our proof is as follows .
first , in section [ sect:4 ] , we define some specially designed binary labeling on the vertices of the second chromatic subdivision of @xmath13 , calling this the _ initial labeling_. this labeling is compliant , but it has many 1-monochromatic 5-simplices . in the next step , we find a perfect matching on the set of 1-monochromatic 5-simplices , such that whenever two 5-simplices are matched , they share a 4-simplex . in section [ sect:4 ] we describe the standard matching , which is easy to define , but which is not perfect . in section [ sect:5 ]
we modify this standard matching by connecting unmatched simplices by augmenting paths .
this is a standard technique from matching theory , and it yields a perfect matching .
once we have a perfect matching , it is easy to use that to achieve a binary compliant labeling of the vertices of @xmath11 without any monochromatic simplices .
this is done in the end of section [ sect:5 ] , and it proves our main theorem [ thm : main - math ] .
the remaining two sections are dealing with further specific issues . the explicit protocol for solving weak symmetry breaking task for @xmath3 processes in @xmath4 rounds
is given in section [ sect:5 ] .
we finish with section [ sect:6 ] , where we give a very short argument for impossibility of solving weak symmetry breaking task in @xmath8 round for any number of processes .
while the focus of this paper has been on the case of 6 processes , it is possible to refine our analysis to a much more general setting , yielding new bounds for the symmetry breaking function , as well as producing distributed protocols for other numbers of processes .
however , there is a number of technical issues which need to be resolved , and there is a number of new ideas which need to be introduced , since giving the simplex paths explicitely , as was done in table [ table:21paths ] , is prohibitive for higher values of @xmath0 .
for this reason , the extension of the techniques from this paper to higher values of @xmath0 will appear in a separate article , see @xcite .
@xmath5 for an arbitrary nonnegative integer @xmath0 , we set @xmath14:=\{0,\dots , n\}$ ] .
we let @xmath15 denote the standard @xmath0-simplex .
note that @xmath15 has @xmath16 vertices , which we index by the set @xmath14 $ ] . for brevity
, we shall skip the curly brackets for the sets consisting of a single element , and write expressions like @xmath17 and @xmath18 , rather than @xmath19 and @xmath20 .
given a set @xmath21 and an element @xmath22 , we shall call the pair @xmath23 a * node*. we shall say that @xmath24 is the _ color _ of this node , and write @xmath25 .
when @xmath26 is any set of nodes , we set @xmath27 . the reasons for our terminology will become apparent soon .
[ df : osp ] an * ordered set partition * @xmath28 of a set @xmath21 is an ordered tuple @xmath29 of nonempty subsets of @xmath21 such that @xmath21 is a disjoint union of @xmath30 .
we shall use the notation @xmath31 . assume now that we are given an ordered set partition @xmath31 of a set @xmath21 .
we shall call the set @xmath32 the * set of nodes of @xmath28 * , where @xmath33 denotes the unique index such that @xmath34 .
note that we have @xmath35 .
if @xmath36 is a node of an ordered set partition @xmath28 of a set @xmath21 , such that @xmath37 , then we say that @xmath24 is * almost maximal * with respect to @xmath28 .
[ rem : amax ] for every ordered set partition @xmath28 of a set @xmath21 , there exist at least two elements of @xmath21 , which are almost maximal with respect to @xmath28 . indeed ,
if @xmath31 , then all elements of @xmath38 are almost maximal . if @xmath39 , then also all elements of @xmath40 are almost maximal .
given a permutation @xmath41 of @xmath14 $ ] , there is a natural ordered set partition of the set @xmath14 $ ] associated to it , namely @xmath42 ; we shall call it @xmath43 . this way
, we obtain precisely all ordered set partitions consisting of singletons .
there is a standard mathematical reformulation of the solvability of the weak symmetry breaking task which we now proceed to describe .
the central role in this description is played by the following abstract simplicial complex , called _ standard chromatic subdivision of a simplex_. [ df : chi1 ] let @xmath0 be a nonnegative integer .
the abstract simplicial complex @xmath44 is given as follows : * the set of vertices @xmath45 consists of all nodes @xmath46 , such that @xmath47 $ ] ; * the maximal simplices of @xmath48 are indexed by ordered set partitions of @xmath14 $ ] , where for each ordered set partition @xmath28 , its set of vertices is given by the corresponding set of nodes @xmath49 ; * in general @xmath50 is a simplex if and only if there exists a maximal simplex @xmath28 such that @xmath51 .
the color of the vertex of @xmath48 is the color of its indexing node , in particular , @xmath52 can be applied to any subset of @xmath53 .
furthermore , when @xmath28 is an @xmath0-simplex of @xmath44 , we set @xmath54 .
note that @xmath48 is a _ pure _ simplicial complex of dimension @xmath0 , meaning that all of its maximal simplices have the same dimension .
the simplicial complex @xmath48 has been introduced by herlihy & shavit , @xcite , see also a recent book @xcite , and is a widely used gadget in theoretical distributed computing .
it has been proved in @xcite that @xmath48 is a subdivision of an @xmath0-simplex .
a wide generalization of this fact has been proved in @xcite .
@xmath5 the construction from definition [ df : chi1 ] can be used to define chromatic subdivision of an arbitrary simplicial complex due to the following simple fact : when restricted to any of its boundary simplices @xmath55 , the standard chromatic subdivision @xmath44 is naturally isomorphic to @xmath56 , where @xmath57 is the dimension of @xmath55 .
indeed , given any simplicial complex @xmath58 , we can simply replace each of its simplices with its standard chromatic subdivision , these will fit together nicely , and we can call the result the _ standard chromatic subdivision of @xmath58_. in particular , this means that we can define _ iterated chromatic subdivisions _ @xmath59 . the simplicial complex @xmath60 is shown on figure [ fig:2 ] . for any finite set @xmath21 we let @xmath61 denote the standard simplex whose vertices are indexed by the elements of @xmath21 .
in this setting , our standard simplex @xmath15 would be called @xmath62}$ ] .
the boundary simplices are called @xmath63 , for all subsets @xmath64 $ ] .
given two equicardinal subsets @xmath65 $ ] , the corresponding boundary simplices @xmath63 and @xmath66 have the same dimension , and any bijection @xmath67 induces a simplicial isomorphism from @xmath63 to @xmath66 .
clearly , the construction of the iterated standard chromatic subdivision will also induce a simplicial isomorphism from @xmath68 to @xmath69 .
we let @xmath70 denote the simplicial isomorphism from @xmath68 to @xmath69 induced by the unique order - preserving bijection from @xmath71 to @xmath72 .
[ df : cmp ] a binary labeling @xmath73 is called * compliant * if for all @xmath65 $ ] , such that @xmath74 , and all vertices @xmath75 , we have @xmath76 in simple terms , this means that the restriction of the labeling to @xmath77 will only depend on the cardinality of the ordered set @xmath71 . the following well - known statement ,
see @xcite , is a useful reformulation of the solvability of the weak symmetry breaking in the layered immediate snapshot model in purely mathematical terms .
[ thm : red ] in the layered immediate snapshot computational model , the weak symmetry breaking task for @xmath0 processes is solvable in @xmath78 rounds if and only if there exists a compliant binary labeling @xmath79 , such that for every @xmath2-simplex @xmath80 , the restricted map @xmath81 is surjective .
clearly , the reduction in theorem [ thm : red ] means that theorem [ thm : main - math ] immediately implies theorem [ thm : main - cs ] .
@xmath5 we are now ready to introduce the main function for our study .
assume @xmath0 is an arbitrary natural number .
we let @xmath1 denote the minimal number @xmath78 such that there exists a compliant binary labeling @xmath79 , such that for every @xmath2-simplex @xmath80 , the restricted map @xmath81 is surjective .
if no such @xmath78 exists , we set @xmath82 . note , that once such a labeling @xmath10 exists for some @xmath78 , for any other @xmath83 it is easy to extend it to a labeling @xmath84 satisfying the same conditions . furthermore , note that theorem [ thm : red ] implies the following remark .
the weak symmetry breaking task for @xmath0 processes is solvable in the layered immediate snapshot model if and only if @xmath85 . the actual value @xmath1 is the minimal number of rounds needed for the distributed protocol to solve this task .
the table [ table : sb ] summarizes our current knowledge of the function @xmath1 .
@xmath5 even though definition [ df : chi1 ] provides a well - defined simplicial complex , it is somewhat cumbersome to work with , mainly due to the fact that the lower - dimensional simplices lack a direct combinatorial description .
we shall now mitigate this situation by giving an alternative combinatorial description of the simplicial structure of @xmath48 , which has been first obtained in @xcite .
however , before we can do this , we need some additional terminology .
a * partial ordered set partition * of the set @xmath14 $ ] is a pair of ordered set partitions of nonempty subsets of @xmath14 $ ] , @xmath87 , which have the same number of parts , such that for all @xmath88 , we have @xmath89 . given such a partial ordered set partition @xmath28 , we introduce the following terminology . * the union @xmath90 is called the * carrier set * of @xmath28 , and is denoted by @xmath91 . * the union @xmath92 is called the * color set * of @xmath28 , and is denoted by @xmath93 . *
the * dimension * of @xmath28 is defined to be @xmath94 , and is denoted @xmath95 .
when appropriate , we shall also write @xmath96 which we shall call the _ table form _ of @xmath28 . we note , that both nodes @xmath97 , for @xmath98 $ ] , as well as ordered set partitions of @xmath14 $ ] are special cases of partial ordered set partitions of @xmath14 $ ] .
indeed , a node @xmath97 , such that @xmath98 $ ] corresponds to the somewhat degenerate partial ordered set partition of @xmath14 $ ] @xmath99 whereas an ordered set partition @xmath100 corresponds to the partial ordered set partition of @xmath14 $ ] @xmath101 i.e. , a partial ordered set partition @xmath102 , such that @xmath103 for all @xmath104 , and @xmath105 $ ] . each partial ordered set partition has non - empty color set , which in turn is contained in its carrier set .
the nodes correspond to the partial ordered set partitions with minimal color set , consisting of just one element , and ordered set partitions correspond to the partial ordered set partitions with maximal color set , namely the whole set @xmath14 $ ] .
[ df : dl1 ] assume we are given a partial ordered set partition @xmath106 of the set @xmath14 $ ] , such that @xmath107 , and we are also given an element @xmath108 , say @xmath109 , for some @xmath110 . to define the * deletion of @xmath24 from @xmath28 * we consider three different cases . 1 .
if @xmath111 , then the deletion of @xmath24 from @xmath28 is set to be @xmath112 2 .
if @xmath113 , and @xmath114 , then the deletion of @xmath24 from @xmath28 is set to be @xmath115 3 .
if @xmath113 , and @xmath116 , then the deletion of @xmath24 from @xmath28 is set to be @xmath117 we denote the deletion of @xmath24 from @xmath28 by @xmath118 .
it is easy to see that @xmath118 is again a partial ordered set partition of @xmath14 $ ] .
we have @xmath119 , @xmath120 , and @xmath121 .
definition [ df : dl1 ] can be generalized as follows .
[ df : dl2 ] let @xmath122 be a partial ordered set partition of the set @xmath14 $ ] , and assume we are given a non - empty set @xmath123
. let @xmath124 index all sets @xmath125 , such that @xmath126 , and set @xmath127 . for all @xmath128 , we set @xmath129 the obtained partial ordered set partition @xmath130 is called * deletion of @xmath131 from @xmath28 * , and is denoted @xmath132 .
the properties of the deletion of a part of its color set from a partial ordered set partition are summarized in the following proposition .
[ pr:2.7 ] let @xmath122 be a partial ordered set partition of the set @xmath14 $ ]
assume we have a non - empty subset @xmath123 .
then @xmath132 is again a partial ordered set partition of the set @xmath14 $ ] , such that 1 .
@xmath133 ; 2 .
@xmath134 ; 3 .
2 . assume @xmath131 and @xmath136 are disjoint non - empty subsets of @xmath137 , such that @xmath138 .
then , we have @xmath139 * proof . *
all statements are immediate from definition [ df : dl2 ] .
generalizing definition [ df : osp ] we can define the set of nodes of an arbitrary partial ordered set partition of the set @xmath14 $ ] .
[ df : osp2 ] let @xmath122 be a partial ordered set partition of the set @xmath14 $ ] .
we shall call the set @xmath140 , the * set of nodes * of @xmath28 , where again @xmath33 denotes the unique index such that @xmath141 . note , that @xmath142 , @xmath143 , and @xmath144 .
comparing definitions [ df : dl2 ] and [ df : osp2 ] yields the identity @xmath145{\setminus}\{x\}){\,|\,}x\in c(\sigma)\}.\ ] ] furthermore , when an ordered set partition is viewed as a partial one , its set of nodes does not depend on which one of the definitions [ df : osp ] and [ df : osp2 ] is used .
crucially , the set of nodes completely determines any partial ordered set partition of the set @xmath14 $ ] .
[ pr : unique ] assume @xmath28 and @xmath55 are both partial ordered set partitions of the set @xmath14 $ ] , such that @xmath146 , then @xmath147 . * proof .
* assume @xmath148 .
without loss of generality , we can write @xmath122 and @xmath149 , such that @xmath150 . to start with , the sets of sets @xmath151 and @xmath152 must be equal .
this immediately implies that @xmath153 , and that @xmath154 , for all @xmath155 . since @xmath156
, we have @xmath157 .
let @xmath158 denote the minimal number such that @xmath159 .
without loss of generality we can assume that we can find an element @xmath24 , such that @xmath109 , @xmath160 , and @xmath161 , for some @xmath162 .
then , @xmath28 has a node @xmath163 , whereas @xmath55 has a node @xmath164 . since @xmath165 we arrive at a contradiction .
we are now ready to prove that partial ordered set partitions provide the right combinatorial language to describe the simplicial structure of @xmath48 .
[ pr : chi2 ] the nonempty simplices of @xmath48 can be indexed by all partial ordered set partitions of @xmath14 $ ] .
this indexing satisfies the following properties : 1 .
the dimension of the simplex indexed by @xmath166 is equal to @xmath95 .
the vertices of the simplex @xmath28 indexed by @xmath167 are indexed by @xmath49 .
3 . in general , the set of subsimplices of the simplex @xmath28 indexed by @xmath168 is precisely the set of simplices indexed by partial ordered set partitions from the set @xmath169 .
* proof . *
assume @xmath170 , such that @xmath26 forms a simplex in @xmath44 .
by definition [ df : chi1 ] there exists an @xmath0-simplex @xmath28 of @xmath48 , such that @xmath171 .
furthermore , by we get @xmath172{\setminus}x){\,|\,}x\in c(w)\}$ ] .
we now set @xmath173{\setminus}c(w))\ ] ] to be the partial ordered set partition of the set @xmath14 $ ] which indexes the simplex @xmath26 .
first , we note that by proposition [ pr:2.7](1)(a ) we have @xmath174{\setminus}([n]{\setminus}c(w))=c(w).\ ] ] furthermore , we derive @xmath175{\setminus}c(w)),c(w){\setminus}x ) { \,|\,}x\in c(w)\}=\{{\text{\rm dl}\,}(\sigma,[n]{\setminus}x ) { \,|\,}x\in c(w)\}=w,\end{gathered}\ ] ] where the penultimate equality is due to .
assume @xmath55 is another @xmath0-simplex of @xmath48 , such that @xmath176 .
set @xmath177{\setminus}c(w))$ ] .
we have shown in that @xmath178 .
it follows from proposition [ pr : unique ] that @xmath179 , and hence the partial ordered set partition of the set @xmath14 $ ] , which indexes @xmath26 does not depend on the choice of @xmath28 .
we can now show that the indexing given by satisfies the @xmath4 properties in the formulation of the proposition . to see property @xmath180 ,
let @xmath170 be the simplex of @xmath48 indexed by @xmath181 . on one hand , @xmath182 . on the other hand
, picking an @xmath0-simplex @xmath28 such that @xmath171 , we get @xmath183{\setminus}c(w)))=\dim\sigma-(n+1-|c(w)|)= n-(n+1-|w|)=|w|-1,\ ] ] hence @xmath184 .
finally note , that property @xmath185 has been proved by , whereas property @xmath186 follows directly from .
@xmath5 the following combinatorial concept is the key to describing the iterated standard chromatic subdivisions . [ df : linked ] assume we are given a tuple @xmath187 of partial ordered set partitions of the set @xmath21 .
we say that the tuple @xmath28 is * linked * if for all @xmath188 , we have @xmath189 .
assume we have a partial ordered set partition @xmath87 , and @xmath123 .
let @xmath104 be the minimal index , such that @xmath190 .
we define @xmath191 to be @xmath192 , if such an index @xmath104 exists , and we let @xmath191 be empty otherwise
. we can now generalize proposition [ pr : chi2 ] .
[ pr : chi3 ] the simplices of @xmath193 can be indexed by linked @xmath78-tuples of partial ordered set partitions of the set @xmath14 $ ] .
this indexing satisfies following properties : 1 .
the dimension of the simplex indexed by @xmath194 is equal to @xmath195 .
the subsimplices of @xmath194 are all tuples @xmath196 , where @xmath197 , and @xmath198 , for all @xmath199
. * proof . *
clearly , each simplex of @xmath193 can be obtained by first choosing a simplex of @xmath48 , then viewing this simplex as @xmath200 , for some @xmath201 , then picking the next simplex in @xmath200 , and so on , repeating @xmath78 times in total . by proposition
[ pr : chi2 ] each next simplex can be indexed by a partial ordered set partition , and each time the color set of the previous simplex is the carrier of the next one .
this is exactly the same as requiring for this tuple of simplices to be linked in the sense of definition [ df : linked ] .
both ( 1 ) and ( 2 ) now follow immediately from proposition [ pr : chi2 ] and the definition of the deletion operation .
generalizing , the index of a simplex in @xmath193 can be visualized as an array of subsets : @xmath202 satisfying @xmath203 , for all @xmath204 .
in particular , the top - dimensional simplices of @xmath193 are indexed by tuples @xmath205 .
we find it practical to use the following shorthand notation : @xmath206 .
@xmath5 one of the reasons , why the layered immediate snapshot computational model is amenable to detailed analysis is because the corresponding protocol complexes have a useful structure of a pseudomanifold .
a pure @xmath0-dimensional simplicial complex @xmath58 is called a * pseudomanifold * of dimension @xmath0 if the following two conditions are satisfied : 1 .
any @xmath2-simplex belongs to at most two @xmath0-simplices ; 2 .
any two @xmath0-simplices @xmath28 and @xmath55 can be connected by a path of @xmath0-simplices @xmath207 , such that for any @xmath208 , the @xmath0-simplices @xmath209 and @xmath210 share an @xmath2-simplex .
it is well known , see e.g. , @xcite , that the iterated chromatic subdivision of an @xmath0-simplex is an @xmath0-pseudomanifold .
we now provide the combinatorial language for describing how to move between @xmath0-simplices of this pseudomanifold . to start with , consider an @xmath0-simplex @xmath28 of @xmath48 , say @xmath31 .
we set @xmath211{\setminus}a_t , & \text { if } |a_t|=1 ; \\ [ n ] , & \text { otherwise . } \end{cases}\ ] ] [ df : flip ] given an @xmath0-simplex @xmath28 of the simplicial complex @xmath48 , and @xmath212 , we let @xmath213 denote the @xmath0-simplex of @xmath48 obtained in the following way .
let us say @xmath31 , and @xmath214 . 1 .
assume @xmath215 , then we set @xmath216 2 .
assume @xmath217 ( that is @xmath218 ) and @xmath219 .
then we set @xmath220 note that since @xmath212 , we can not have the case @xmath221 , hence @xmath213 is well - defined .
we say that @xmath213 is obtained from @xmath28 by a _ flip _ of @xmath28 with respect to @xmath24 . in this sense
, @xmath222 is the set of all colors which can be flipped . given any simplex @xmath28 we can always flip with respect to all colors except for at most one color .
we can also flip back , so we have @xmath223 for all @xmath28 , @xmath24 .
furthermore , we remark that for any @xmath212 , the @xmath0-simplices @xmath28 and @xmath213 share an @xmath2-simplex @xmath55 given by @xmath224 . definition [ df : flip ] can be generalized to iterated chromatic subdivisions . for an @xmath0-simplex @xmath225 of @xmath193 we set @xmath226 . effectively this means that @xmath227 $ ] , unless @xmath228{\setminus}p$ ] , in which case we have @xmath227{\setminus}p$ ] .
[ df : gflip ] assume we are given an @xmath0-simplex @xmath225 of the simplicial complex @xmath193 , and @xmath212 .
let @xmath158 be the maximal index such that @xmath229 , by the definition of @xmath222 , such @xmath158 must exist .
we define @xmath213 to be the following @xmath78-tuple of ordered partitions : @xmath230 again , it is easy to see that for any @xmath212 , the @xmath0-simplices @xmath28 and @xmath213 will share an @xmath2-simplex , and that for all @xmath28 , @xmath24 , we have the identity @xmath231
@xmath5 let us review some basic terminology of the graph theory , more specifically the matching theory . to start with , recall that a graph @xmath232 is called _ bipartite _ if its set of vertices @xmath233 can be represented as a disjoint union @xmath234 , such that there are only edges of the type @xmath235 , with @xmath236 , @xmath237 .
we shall call @xmath238 a _ bipartite decomposition_. note , that such a decomposition need not be unique .
classically , a _ matching _ in a graph @xmath232 is a subset @xmath239 of its set of edges @xmath240 , such that two different edges from @xmath239 do not share vertices .
we call edges which belong to @xmath239 the _ matching edges _ and all other edges the _ non - matching edges_. a vertex @xmath241 is called _ matched _ if there exists an edge in @xmath239 having @xmath241 as an endpoint .
we call the unmatched vertices _ critical _ , which is consistent with the terminology we used in the previous sections .
a matching is called _ perfect _ if there are no critical vertices , otherwise we may call the matching _
partial_. clearly , an existence of a perfect matching implies that the sets @xmath21 and @xmath242 have the same cardinality . a matching is called _ near - perfect _ if there is exactly one critical vertex . in this paper , the fundamental instance of a graph on which matchings are constructed
is provided by @xmath243 .
this is the graph whose vertices are all @xmath0-simplices of @xmath48 , and two vertices are connected by an edge if the corresponding @xmath0-simplices share an @xmath2-simplex .
sometimes , we shall abuse our language and call the vertices of @xmath243 simplices .
we color the edges of @xmath243 as follows : the edge connecting @xmath28 with @xmath55 gets the color of the vertex of @xmath28 which does not belong to @xmath244 .
this graph has also been studied in @xcite ; an example is shown on figure [ fig:1 ] .
the graph @xmath243 is bipartite and the bipartite decomposition is unique . recall that @xmath0-simplices of @xmath48 are indexed by ordered set partitions @xmath100 of the set
@xmath14 $ ] . the bipartite decomposition of @xmath243 is then provided by sorting the @xmath0-simplices of @xmath48 into two groups according to the parity of the number @xmath245 . for convenience of notations
we shall define a function @xmath246 , which we call _ orientation _ of the simplex , as follows : @xmath247 the matching itself in this context will mean to group @xmath0-simplices in pairs , so that in each pair the @xmath0-simplices share a boundary @xmath2-simplex .
accordingly , we can talk about critical @xmath0-simplices etc . in general , for any @xmath248 ,
we let @xmath249 denote the graph , whose vertices are the @xmath0-simplices of @xmath193 , and two vertices are connected by an edge if the corresponding @xmath0-simplices share an @xmath2-simplex .
the edges of @xmath249 are colored in the same way as those of @xmath243 .
@xmath5 in this paper we will primarily need the combinatorial description of the simplicial structure of the second chromatic subdivision of an @xmath0-simplex . in this case , the proposition [ pr : chi3 ] says that simplices of @xmath250 are indexed by pairs @xmath251 of partial ordered partitions of @xmath14 $ ] , such that @xmath252 .
the @xmath0-simplices of @xmath250 are simply pairs @xmath251 of @xmath0-simplices of @xmath48 .
the vertices of @xmath250 are indexed by pairs of partial ordered set partitions @xmath253 such that @xmath254 .
assume we are given a simplex @xmath255 of @xmath250 , say @xmath122 , and @xmath256 .
assume furthermore , we are given some subset @xmath257 .
let @xmath158 be the minimal index such that @xmath258 , then @xmath259 .
finally , we remark that for an @xmath0-simplex @xmath260 , and @xmath261 $ ] , the flip @xmath262 is always defined unless @xmath263 , and is explicitly given by .
@xmath5 for an arbitrary simplex @xmath264 of @xmath193 we set @xmath265 .
[ df : int ] let @xmath241 be a vertex of @xmath193 .
we say that the vertex @xmath241 is an * internal * vertex if @xmath266 $ ] , otherwise we say that @xmath241 is a * boundary vertex*. in the case @xmath267 , there is a handy criterion for deciding whether all vertices of an @xmath0-simplex are internal .
[ prop : int ] let @xmath268 be an @xmath0-simplex of @xmath250
. then all vertices of @xmath269 are internal if and only if @xmath270
. in general , assume that @xmath271 , let @xmath272 be the largest index such that @xmath273 for all @xmath274 .
the boundary vertices of @xmath269 are precisely the vertices with colors from @xmath275 .
* recall , that since @xmath269 is an @xmath0-simplex , we have @xmath276 $ ] . pick @xmath261 $ ] , say @xmath277 .
the vertex of @xmath269 , which is colored by @xmath24 has the index @xmath278 , where @xmath279 , and @xmath280 .
this vertex is internal if and only if @xmath281 $ ] . by the definition of the deletion operation ,
this is the case if and only if @xmath282 , i.e. , @xmath283 clearly , the fact that is true for all @xmath284 is equivalent to the condition @xmath270 , and in general yields the description of all boundary vertices of @xmath269 .
we are now in a position to describe the initial labeling of the vertices of @xmath285 @xmath286 before we do this , we would like to designate certain boundary vertices of @xmath287 as exceptional .
[ df : exv ] the set @xmath288 consists of all vertices of @xmath287 listed in table [ table : exv2 ] . the vertices in @xmath288 are called * exceptional vertices * , while all other boundary vertices are called * regular vertices*. @xmath289 \begin{array}{|c||c| } \hline a , b & a \\ \hline a & a \\
\hline \end{array } & \begin{array}{|c||c| } \hline a , b & a , b \\ \hline a , b & b \\
\hline \end{array } & & \text { for all } 0\leq a < b\leq 5;\\[15pt ] \begin{array}{|c||c| } \hline a , b , c & a \\ \hline a & a \\
\hline \end{array } & \begin{array}{|c||c| } \hline a , b , c & a , b \\
\hline a , b & a \\
\hline \end{array } & \begin{array}{|c||c| } \hline a , b , c & a , b \\ \hline a , b & b \\
\hline \end{array } & \\[15pt ] \begin{array}{|c||c| } \hline a , b , c & a , b , c \\
\hline a , b , c & b \\
\hline \end{array } & \begin{array}{|c||c| } \hline a , b , c & a , b , c \\
\hline a , b , c & c \\
\hline \end{array } & & \text { for all } 0\leq a < b < c\leq 5 . \end{array}\ ] ] clearly , there are 136 exceptional vertices .
note , that @xmath290 whenever @xmath241 is an exceptional boundary vertex .
the general rule for the labeling @xmath71 is now as follows : @xmath291 clearly , this labeling is compliant .
we remark that since any @xmath12-simplex contains an internal vertex , there are no @xmath7-monochromatic @xmath12-simplices in the labeling @xmath71 .
of course there are quite many @xmath8-monochromatic @xmath12-simplices .
we will eliminate them by passing to the third chromatic subdivision .
@xmath5 our next goal is to produce a perfect matching on the set of @xmath8-monochromatic @xmath12-simplices , so that each pair of matched @xmath12-simplices shares a @xmath292-simplex
. we will start with producing a partial matching . as a warm - up ,
let us as above consider the graph @xmath243 , whose set of vertices is given by @xmath0-simplices of @xmath48 .
for @xmath293 , let @xmath294 denote the number of vertices of @xmath295 .
we have @xmath296 , @xmath297 etc .
there is an easy recursion @xmath298 see ( * ? ? ?
* subsection 4.4 ) for a more general formula .
one can see by induction that @xmath294 is always odd . to do this , simply evaluate the right hand side of modulo 2 and see that @xmath299 since @xmath294 is odd we can not hope for a perfect matching . however , it is easy to produce a number of near - perfect matchings , each one matching all but one simplex .
we need the following piece of terminology .
[ df : level ] let @xmath300 be a fixed order on a nonempty subset of @xmath14 $ ] , and let @xmath31 be an @xmath0-simplex of @xmath48 .
we shall define the number @xmath301 $ ] , called the * level of @xmath28 with respect to @xmath302 * as follows . if @xmath303 , then @xmath304 is not defined ; else @xmath305 , where @xmath57 is the highest number @xmath306 , such that @xmath307 .
definition [ df : level ] can be rephrased as follows .
if the last set in the ordered set partition @xmath28 is not @xmath308 , the @xmath309 , else we proceed with the next set in @xmath28 . if the last set in @xmath28 is @xmath308 , but the penultimate one is not @xmath310 , then @xmath311 , else we proceed with the next set in @xmath28 .
one then repeats this step , moving down in the ordered set partition @xmath28 .
the level is then not defined , if we went through whole @xmath302 and did not stop .
this will happen precisely when @xmath28 has the form @xmath312 .
we are now ready to define a special kind of matching , which we shall call the _
standard matching_. [ df : sm ] let @xmath313 be any fixed order on the set @xmath14 $ ] .
recall that @xmath314 is an ordered set partition of @xmath14 $ ] associated to @xmath302 .
the * standard matching * @xmath315 associated to @xmath302 is a bijection @xmath316 defined by @xmath317 there is a simple verbal formulation for the standard matching . in order to match a simplex @xmath28 , simply flip with respect to the highest possible color , such that the flip preserves the level function , see also the proof of proposition [ prop : sm ] .
the next proposition says that in the standard matching on the set of all @xmath0-simplices of @xmath48 with respect to any permutation @xmath302 , there is exactly one critical simplex , namely @xmath314 .
this is illustrated by figure [ fig:3 ] .
[ prop : sm ] the standard matching @xmath315 described in definition [ df : sm ] is a well - defined near - perfect matching on @xmath243 , with the only critical @xmath0-simplex being indexed by @xmath314 .
* let @xmath318 . the crucial fact which we need is that @xmath319 to see , assume @xmath320 , such that @xmath321 , i.e. , @xmath322 .
the ordered set partition @xmath323 is obtained from @xmath28 by either merging two sets , or by splitting one of the sets . if in the process the sets @xmath324 , @xmath325 , @xmath326 , @xmath327 are untouched , then obviously @xmath328 as well
. on the other hand , the only way the sets @xmath325 , @xmath326 , @xmath327 could be modified would be by merging them with a neighbor . however , this is impossible , since merges must always involve the set @xmath329 , and we have assumed that @xmath321 .
it remains to consider the case that the set @xmath324 was altered .
if it was split into @xmath329 and @xmath330 , then again is satisfied , since @xmath331 . finally ,
if the set @xmath324 was merged with the set @xmath332 , then again @xmath333 , and is valid again .
now it follows from together with that @xmath334 this means that we have a well - defined matching , and that all simplices @xmath28 for which @xmath304 is defined are matched .
we conclude that we have a near - perfect matching with the critical @xmath0-simplex being indexed by @xmath314 .
note , that even if we have @xmath335 , for @xmath336 , the matching @xmath315 is still defined for those simplices @xmath28 , for which the number @xmath304 is defined . in this case
, we will have a number of critical simplices , namely all those indexed by the tuples @xmath337 , and the proof is simply the same as that of proposition [ prop : sm ] . for a proper nonempty subset @xmath338
$ ] we let @xmath339 denote the subgraph of @xmath243 induced by set of all @xmath0-simplices @xmath31 , such that @xmath340 .
[ prop : sm2 ] let @xmath300 be the set @xmath14{\setminus}v$ ] taken in an arbitrary order .
then @xmath315 provides a perfect matching of the simplices in @xmath339 .
* the part of the proof of proposition [ prop : sm ] showing did not use that fact that @xmath302 has cardinality @xmath16 , so we still have that identity .
this means that @xmath315 is still a well - defined matching on @xmath243 , only now with many more critical simplices .
first , we note that @xmath304 is defined for all @xmath341 .
indeed , of @xmath304 is not defined , then @xmath342 , but then @xmath343{\setminus}\{x_0,\dots , x_m\}=v$ ] yields a contradiction .
furthermore , we have the implication @xmath344 indeed , assume @xmath345 , and @xmath31 , with @xmath340
. then , by definition of @xmath315 , either we have @xmath346 , or @xmath347 , where @xmath348{\setminus}v$ ] .
either way , we get @xmath349 , so @xmath350 .
we now conclude that @xmath315 is a perfect matching on @xmath339 .
[ df : st ] given an @xmath0-simplex @xmath31 , with @xmath351 , such that @xmath352 , we set @xmath353 . given a subset @xmath338 $ ] , we shall call a tuple @xmath354 of nonempty subsets of @xmath355 a * prefix * in @xmath355 , if @xmath356 are disjoint .
we shall say that a prefix is * full * if @xmath357 .
we also allow an empty prefix .
furthermore , given a prefix @xmath354 , we let @xmath358 denote the subgraph of @xmath243 induced by the set of all @xmath0-simplices @xmath359 of @xmath48 , such that @xmath360 .
in particular , the set @xmath361 coincides with the previously defined set @xmath339 .
we note , that for any @xmath338 $ ] , the set of vertices of @xmath243 , which is the set of all @xmath0-simplices of @xmath48 , is a disjoint union of the sets of vertices of @xmath358 , where @xmath362 runs over all prefixes in @xmath355 .
the following lemma , whose statement is illustrated by figure [ fig:4 ] , provides a useful generalization of proposition [ prop : sm2 ] .
[ lm : sml ] assume @xmath338 $ ] , and @xmath335 is an arbitrary permutation of the set @xmath14{\setminus}v$ ] .
let furthermore @xmath354 be an arbitrary prefix of @xmath355 . 1 .
if the prefix @xmath362 is not full , then the standard matching @xmath315 gives a perfect matching of simplices in @xmath358 .
if the prefix @xmath362 is full then the standard matching @xmath315 of @xmath358 has precisely one critical @xmath0-simplex , namely @xmath342 . *
* our argument is very close to the proof of proposition [ prop : sm2 ] .
take @xmath363 .
if @xmath304 is not defined , then @xmath28 has the form @xmath364 . since @xmath365
, we see that @xmath342 , which is only possible if the prefix @xmath362 is full .
so , if @xmath362 is full , we do have one critical simplex , otherwise we do not .
assume now @xmath359 , such that @xmath366 , and let @xmath367 .
we have @xmath368 .
the ordered set partition @xmath323 is obtained from @xmath28 by merging two sets or splitting a set .
if the sets @xmath369 are untouched , then clearly @xmath370 . in the other hand , the sets @xmath356
can not be involved in any modification , so the only remaining case is that @xmath371 has been modified .
either @xmath371 has been replaced by @xmath372 followed by @xmath373 , or @xmath374 and we replace @xmath371 , @xmath375 with @xmath376 . in any case , @xmath377 , where @xmath378 , hence @xmath379 .
thus , we derive the implication @xmath380 for non - critical @xmath28 . hence @xmath315 gives a perfect matching on the set of all @xmath0-simplices in @xmath358 , with the only possible exception being @xmath342 .
@xmath5 before we proceed with describing the partial matching on the set of all @xmath8-monochromatic @xmath12-simplices of the labeling @xmath71 , let us give combinatorial description of all the elements in this set . to start with , all @xmath12-simplices with only internal vertices are @xmath8-monochromatic .
these are described by a simple criterion in proposition [ prop : int ] .
the rest of @xmath8-monochromatic @xmath12-simplices are those which contain internal vertices and some exceptional boundary vertices . in the table [ table : exc - vert ]
we list all exceptional boundary vertices and the @xmath12-simplices which contain them .
@xmath381 next we need to understand which of the @xmath12-simplices from the right column of the table [ table : exc - vert ] contain _ only _ exceptional boundary vertices .
recall that a @xmath12-simplex of @xmath287 is indexed by a pair of ordered set partitions of subsets of @xmath382 $ ] .
we shall describe the set of @xmath12-simplices containing only exceptional boundary vertices by grouping them together according to the first ordered set partition .
for a @xmath12-simplex @xmath28 of @xmath383 , let @xmath384 denote the subgraph of @xmath295 induced by all @xmath12-simplices @xmath55 of @xmath383 such that the @xmath12-simplex @xmath251 of @xmath287 is @xmath8-monochromatic .
@xmath385 the explicit description of the graph @xmath384 is given in proposition [ prop:3.11 ] , using table [ table : exc - simp ] .
we remark that the types of the @xmath12-simplices listed in the left column of that table are meant to be mutually exclusive .
[ prop:3.11 ] let @xmath386 be a @xmath12-simplex of @xmath383 , and set @xmath387{\setminus}s_p$ ] .
then we have @xmath388 , where the subset @xmath389 is given as follows . if @xmath28 is of the type listed on the left column of table [ table : exc - simp ] , then @xmath390 where the union is taken over all prefixes @xmath362 which are listed in the corresponding row of table [ table : exc - simp ] .
we set @xmath389 to be an empty set otherwise
. * proof . * to start with , we see that @xmath338 $ ] .
if @xmath355 is empty , then we are looking at the central @xmath12-simplex @xmath391)$ ] , in which case all of the @xmath12-simplices of @xmath384 are internal .
assume now @xmath355 is non - empty .
let @xmath392 be a @xmath12-simplex of @xmath383 . by proposition [ prop : int ] , all the vertices of the @xmath12-simplex @xmath251 of @xmath285 are internal if and only if @xmath393 , or equivalently @xmath394 . in other words ,
the 5-simplex @xmath251 is internal if and only if @xmath395 .
all internal @xmath12-simplices are 1-monochromatic , and all further @xmath8-monochromatic @xmath12-simplices appear due to the exceptional vertices .
so if did not have any , we would simply have @xmath396 . to take into account the influence of the exceptional vertices , we obtain the correction set @xmath389 by a direct case - by - case analysis .
table [ table : exc - vert ] shows the cases which need to be considered .
the table [ table : exc - simp ] describes in each one of these cases , which new @xmath8-monochromatic simplices will arise .
@xmath5 in order to look at matchings in depth , we need certain standard tools from the matching theory . given a partial matching ,
there is a classical way of modifying it , possibly enlarging , or even making it perfect .
following definition provides the key concepts . given a matching @xmath239 on a graph @xmath232 , assume @xmath397 is an edge path in @xmath232 with endpoints @xmath241 and @xmath398 .
we call the path @xmath397 * * alternating * if the edges of @xmath397 are alternatively matching and non - matching ; * * properly alternating * if is alternating and those endpoints of the path which are not matched on the path are critical ; * * augmenting * if it is properly alternating , starting and ending with non - matching edges ; * * semi - augmenting * if it is properly alternating , starting with a matching edge and ending with a non - matching one ; * * weakly semi - augmenting * if it is alternating , starting with a matching edge and ending with a non - matching one ; * * non - augmenting * if it is alternating , starting and ending with matching edges .
properly alternating non - self - intersecting edge paths allow us to modify matchings .
[ df : mdef ] assume we are given a matching @xmath239 on a graph @xmath232 , and a properly alternating non - self - intersecting edge path @xmath397 .
we define @xmath399 be a new matching on @xmath232 consisting of all edges from @xmath239 which do not belong to @xmath397 together with all edges from @xmath397 which do not belong to @xmath239 .
the next proposition shows that having a properly alternating edge path is enough to find a larger matching , and eventually to turn partial matchings into perfect ones .
[ prop : ap ] assume @xmath239 is a matching on a bipartite graph @xmath232 , and @xmath400 are critical vertices of @xmath232 with respect to @xmath239 .
assume furthermore , that there exists a properly alternating edge path @xmath397 from @xmath241 to @xmath398 .
then , there exists a matching @xmath401 on @xmath232 , such that the set of critical vertices with respect to @xmath401 is obtained from the set of critical vertices with respect to @xmath239 by removing the vertices @xmath241 and @xmath398 .
* if the path @xmath397 is non - self - intersecting , then the new matching @xmath401 can be taken to be @xmath399 .
assume @xmath397 is a self - intersecting path , and its vertices , listed in the path order , are @xmath402 , @xmath403 , @xmath326 , @xmath404 .
since @xmath397 is self - intersecting , there exist @xmath405 , such that @xmath406 , and there is no vertex duplication in the sequence @xmath407 , @xmath408 , @xmath326 , @xmath409 .
let @xmath410 be obtained from @xmath397 by removing the loop consisting of the edges @xmath411 , @xmath412 , @xmath326 , @xmath413 .
since the graph @xmath232 is bipartite , that loop must have an even length , hence the new path @xmath410 is again properly alternating . repeating this procedure we will eventually arrive at a properly alternating non - self - intersecting edge path connnecting @xmath241 with @xmath398 .
we note , that if the edge path @xmath397 in proposition [ prop : ap ] is semi - augmenting , then the number of critical vertices does not change , whereas , if @xmath397 is non - augmenting , then the number of critical vertices increases by @xmath414 . even though modifying a matching along a semi - augmenting path does not change the number of critical vertices , this modification is still useful as it could be thought of as transporting the critical vertex from one endpoint of the semi - augmenting path to the other .
[ df : strm ] let @xmath232 be a bipartite graph , with the corresponding bipartite decomposition @xmath238 .
we say that @xmath232 is * conducting * if one of the following situations occur : 1 .
we have @xmath415 , and for any vertex @xmath236 , the graph @xmath232 has a near - perfect matching , with the critical vertex @xmath241 .
we have @xmath416 , and for any vertices @xmath236 , @xmath237 , the graph @xmath232 has a matching , with precisely two critical vertices @xmath241 and @xmath398 . more specifically , if a conducting graph @xmath232 satisfies condition ( 1 ) we shall call it * conducting graph of the first type * , else we shall call it * conducting graph of the second type*. one handy way of showing that a graph is conducting is by presenting a certain collection of paths , as the next proposition details . [ prop : bg ] let @xmath232 be a bipartite graph , with the corresponding bipartite decomposition @xmath238 . 1 .
assume @xmath239 is a near - perfect matching of @xmath232 with critical vertex @xmath236 , such that for any other vertex @xmath417 there exists a semi - augmenting path from @xmath241 to @xmath398 , then the graph @xmath232 is conducting of the first type .
assume @xmath239 is a perfect matching of @xmath232 , such that for any vertices @xmath236 , @xmath237 there exists a non - augmenting path from @xmath241 to @xmath398 , then the graph @xmath232 is conducting of the second type . * proof .
* this is an immediate consequence of proposition [ prop : ap ] .
[ df : closed ] let @xmath418 be a family of prefixes of a set @xmath338 $ ] containing the empty prefix .
we say that @xmath418 is * closed * if the following two conditions are satisfied : 1 . if @xmath419 , where @xmath420 , @xmath421 , then @xmath422 ; 2 . if @xmath422 , and @xmath423 , for some @xmath424
, then there exists @xmath425 , such that @xmath426 furthermore , we let @xmath427 denote the subgraph of @xmath295 induced by the vertices of @xmath428 .
[ rem : sn ] by lemma [ lm : sml ] , the standard matching will induce a matching on @xmath427 .
however , this matching does not have to be perfect or near - perfect . as a matter of fact
, the number of critical vertices will be equal to the number of full prefixes in @xmath418 .
the next observation can be shown by direct inspection .
[ rem : cl ] all families of prefixes @xmath418 listed in table [ table : exc - simp ] are closed .
we shall now proceed with analysing conductibility of graphs @xmath427 .
let us fix a natural number @xmath429 .
let @xmath430 denote the central simplex of @xmath48 , indexed by @xmath431)$ ] , and let @xmath432 denote the neighboring @xmath0-simplex indexed by @xmath433)$ ] .
consider the standard matching @xmath315 , associated to @xmath434 .
the only critical simplex of @xmath435 is @xmath436 , and clearly , @xmath435 matches @xmath430 with @xmath432 .
our central tool will be the following techical lemma .
[ lm:1 ] let @xmath28 be a non - critical @xmath0-simplex in @xmath48 , for @xmath429 .
there exists a weakly semi - augmenting ( with respect to @xmath315 ) path @xmath437 starting at @xmath28 and ending either at @xmath430 or at @xmath432 .
* since the endpoints of a weakly semi - augmenting path have the same orientation , we know that if such a path @xmath437 exists , then its endpoint from the set @xmath438 is uniquely determined . for brevity , we shall say that @xmath437 _ goes from @xmath28 to @xmath438 . _
taking the empty path we see that the claim is true for @xmath439 and for @xmath440 .
assume that @xmath441 , @xmath442 .
we shall use the notation @xmath31 throughout the proof .
when describing edges in a path , we shall usually not list entire combinatorial indices of the simplices , but rather the part of the ordered set partition where the actual change occurs .
the path from @xmath28 to @xmath438 will be constructed by concatenating different pieces .
let us briefly introduce some notations in which we encode our paths .
assume @xmath31 is an @xmath0-simplex of @xmath48 , and set @xmath443 .
* we let @xmath444 denote the matching edge between @xmath28 and @xmath445 . *
assume @xmath446 , for some @xmath199 , and @xmath447 .
we let @xmath448 denote the edge between @xmath28 and @xmath449 . *
assume instead that @xmath450 , for some @xmath88 , @xmath451 , and @xmath447 .
we let @xmath452 denote the edge between @xmath28 and @xmath453 . *
if @xmath450 , for some @xmath88 , and @xmath454 , we use the notation @xmath455 to denote the set of edges @xmath452 , for all @xmath451 . we shall call the move @xmath456
the _ generic split_. when a generic split is used in our notation of the path , it means that it does not matter at this point which of the elements of @xmath21 we split off , and any will do .
formally , this means that we output a set of paths , rather than one path .
later , we will need to choose paths satisfying certain additional conditions .
this can be done from this set of paths , by substituting specific splits instead of generic ones .
for example , condition ( 2 ) of definition @xmath457 only guarantess the existance of _ some _ element @xmath24 , which we are allowed to split .
if at this point we used generic split of @xmath324 in our path , then we will be able to pick a suitable path from our set .
we break our argument into considering three different cases
. * case 1 . *
assume that @xmath458 .
pick @xmath104 such that @xmath459 .
our assumption means that either @xmath460 , and @xmath461 , or @xmath450 .
our first goal , is to reduce this to the case @xmath462 .
assume that @xmath463 , and @xmath464 , say @xmath465 .
we can use the moves and to lower the index @xmath104 by @xmath8 .
@xmath466 @xmath467 assume now @xmath463 , and @xmath468 .
we can use the moves and to lower the cardinality @xmath469 by @xmath8 , eventually reducing this to the case above .
@xmath470 @xmath471 in either case , we are able to lower the index @xmath104 by @xmath8 , and in the end to reach the case @xmath472 .
next , we want to achieve @xmath473 , or @xmath474 . assume first @xmath475 , then use the move to achieve @xmath474 .
@xmath476 assume now @xmath477 , but @xmath478 .
since @xmath442 , we must have @xmath479 .
if @xmath480 , then use the move to reduce to the case @xmath475 , which we just dealt with .
@xmath481 if @xmath482 and @xmath483 , then use the move to reduce to the case @xmath475 again .
@xmath484 finally , assume @xmath482 , @xmath485 .
then , use the move , picking @xmath486 , to reduce to the case @xmath473 .
so at this point we have @xmath472 , and @xmath473 , or @xmath474 . using one of the moves and , whenever @xmath215
, we will eventually arrive at the simplex which is either encoded by @xmath488 or by @xmath489 .
@xmath490 @xmath491 we now continue with moves and reducing our simplex either to @xmath492 or to @xmath493 .
@xmath494 @xmath495 if @xmath496 , we use the move to arrive at @xmath432 .
@xmath497)\ ] ] so we can assume that @xmath498 . if @xmath499 , we use the move to arrive at @xmath430 .
@xmath500)\ ] ] else write @xmath501 and use the move to arrive at the considered case @xmath499 . @xmath502 * case 2 . *
assume @xmath503 .
in this case we have @xmath504 and @xmath505 .
we use the move to reduce it to case 1 , where @xmath506 .
* assume @xmath508 .
if @xmath509 , iteratively use the moves and to reduce it to case 1 .
@xmath510 @xmath511 if @xmath512 , iteratively use the moves and to reduce it to case 1 .
@xmath513 @xmath514 assume now @xmath515 .
use the moves and to reduce it to the previously considered cases .
@xmath517 finally assume @xmath518 .
we consider three cases .
if @xmath519 , we use the move ; if @xmath520 , we use the move ; if @xmath521 use the move , which is possible since @xmath429 .
@xmath522 @xmath523 @xmath524{\,|\,}n)\stackrel{n-1}{\longrightarrow}([n-1]{\,|\,}n ) \stackrel{n-2{\,|\,}}{\rightsquigarrow}(n-2{\,|\,}[n-3]\cup n-1{\,|\,}n)\ ] ] in each case we reduce to the already considered case @xmath525 .
curiously , lemma [ lm:1 ] is not true when @xmath526 for @xmath527 .
it still remains true for all other simplices @xmath28 , and it is also trivially true for @xmath528 .
all of this can be seen by direct inspection .
[ thm : cond1 ] for any natural number @xmath0 , the graph @xmath243 is conducting of the first type . * proof . * the cases @xmath528 and @xmath526 can be verified directly , so we assume @xmath429 .
recall now , that by proposition [ prop : sm ] the standard matching @xmath315 is a near - perfect matching with a critical @xmath0-simplex indexed by @xmath302 .
let @xmath55 be any other @xmath0-simplex such that @xmath529 . by proposition [ prop : bg](1 )
we need to show that there exists a weakly semi - augmenting path between @xmath55 and @xmath302 .
let @xmath28 be any of the neighboring @xmath0-simplices of @xmath302 .
since both @xmath28 and @xmath55 are non - critical , lemma [ lm:1 ] implies that there exist a weakly semi - augmenting path @xmath530 between @xmath55 and @xmath438 , and a weakly semi - augmenting path @xmath531 between @xmath28 and @xmath438 . since @xmath532 , the paths @xmath530 and @xmath531 will link them to the different vertices in @xmath438 . let @xmath533 be the concatenation of @xmath530 , followed by the edge @xmath534 , then by @xmath531 , and finally by the edge @xmath535
clearly , @xmath533 is a semi - augmenting path from @xmath55 to @xmath302 .
[ thm : cond2 ] assume @xmath14\supset v\neq\emptyset$ ] , such that @xmath536 , then @xmath339 is conducting of the second type . *
* note , that the assumption @xmath429 is automatic here , since @xmath537 . without loss of generality
, we can assume that @xmath538 $ ] , with @xmath539 , and set @xmath540 . by proposition [ prop : sm2 ] , the standard matching @xmath315 is a perfect matching on the graph @xmath339 .
let @xmath430 and @xmath432 be as above , we still have @xmath541 and @xmath542 .
pick an arbitrary @xmath543 , which is non - critical with respect to @xmath315 . by lemma [ lm:1 ] there
exists a weakly semi - augmenting path @xmath533 from @xmath28 to @xmath438 .
all we need to do is to see that this path lies entirely within @xmath339 .
recall , that there are 4 types of edges used in the construction of our path : @xmath444 , @xmath448 , @xmath452 , and @xmath455 . clearly ,
if @xmath544 , then any of the edges @xmath444 , @xmath448 , @xmath455 will lead to a vertex in @xmath545 as well .
the edges @xmath452 occur only in paths , , , and . in all these cases we have @xmath546 , in particular @xmath547 ; hence also here the edge @xmath452 leads to a vertex in @xmath545 .
we conclude that the entire path stays within @xmath339 .
now , we have seen in the proof of ( 1 ) that the existence of such paths means that any two vertices @xmath548 of @xmath339 , such that @xmath549 , are connected by an augmenting path . by proposition [ prop : bg](2 )
this implies that @xmath339 is conducting of the second type .
[ thm : cond3 ] the graphs @xmath550 are conducting for all families of prefixes @xmath418 listed in table [ table : exc - simp ] , under the assumption @xmath551
. * proof . *
again without loss of generality we can assume that @xmath552 , all other cases can be reduced to this one by appropriate renaming .
choose @xmath302 to be any order on @xmath14{\setminus}v$ ] with @xmath553 coming last , i.e. , @xmath554 .
by remark [ rem : sn ] the standard matching @xmath315 is a well - defined matching on the graph @xmath427 with one critical simplex for each full prefix in @xmath418 .
as above , we still have @xmath541 and @xmath542 .
let @xmath28 be a non - critical simplex in @xmath427 .
by lemma [ lm:1 ] there exists a weakly semi - augmenting path @xmath533 from @xmath28 to @xmath438 .
this time , what we need to do is to see that this path lies entirely within @xmath427 .
again we need to perform a detailed analysis of moves through .
take any edge @xmath555 on our path .
assume @xmath556 , say @xmath557 , where @xmath420 , and @xmath360 .
of course we have @xmath422 .
we need to show that @xmath558 .
we consider separately each of the 4 possible types of edge @xmath559 . * case 1 . *
if @xmath559 is of the type @xmath444 , then the @xmath55 starts with the same prefix as @xmath28 , so the claim follows .
* case 2 . *
assume @xmath559 is of the type @xmath455 .
if @xmath560 , then , since @xmath418 is a closed prefix , by definition [ df : closed](2 ) there exists @xmath451 , such that the edge @xmath452 will lead to @xmath558 . if @xmath561 , then @xmath55 starts with the same prefix as @xmath28 , so again @xmath558 . finally , the case @xmath562 never occurs on our path .
* case 3 . * assume @xmath559 is of the type @xmath452 .
as mentioned in the proof of theorem [ thm : cond2 ] , in all of these cases we have @xmath546 .
this means that @xmath547 , and @xmath55 starts with the same prefix as @xmath28 , so @xmath558 . *
* assume finally @xmath559 is of the type @xmath448 .
we must have @xmath563 , for some @xmath564 . if @xmath565 , then @xmath55 starts with the same prefix as @xmath28 , so @xmath558 . if @xmath566 , then @xmath55 starts with the prefix @xmath567 .
since @xmath418 is a closed prefix , definition [ df : closed](1 ) implies that @xmath568 , hence @xmath558 .
finally , the case @xmath569 never occurs on our path .
thus we have shown that @xmath558 in all 4 cases .
if @xmath418 has no full prefixes , then by lemma [ lm : sml](1 ) @xmath570 is a perfect matching and @xmath427 is conducting .
if @xmath418 has a single full prefix , then by lemma [ lm : sml](2 ) it has a near - perfect matching .
let @xmath397 denote the critical @xmath0-simplex . by an argument identical to the one used in the proof of lemma [ lm:1
] , we see that there is a weakly semi - augmenting path from @xmath397 to any @xmath0-simplex @xmath28 of @xmath427 , such that @xmath571 . using proposition [ prop : bg](1 )
, we then conclude that @xmath427 is conducting .
finally , when considering the row @xmath572 , with @xmath573 , @xmath574 , we get a family of prefixes with 3 full prefixes , hence we get 3 critical simplices .
these simplices are @xmath575 , @xmath576 , @xmath577 .
we now extend our matching @xmath315 by matching @xmath578 with @xmath579 , obviously these share an @xmath2-simplex , so the new matching is well - defined .
all that remains is to find semi - augmenting paths from @xmath578 and @xmath579 to @xmath438 . to find such a path starting from @xmath578 we use the move ,
while to find such a path starting from @xmath579 we use the move .
@xmath581 in each case , we arrive at a non - critical simplex , from which we have already found a weakly semi - augmenting path to @xmath438 .
@xmath5 let us now return to the labeling @xmath71 , where we want to find a perfect matching on the set of all @xmath8-monochromatic @xmath12-simplices . using the notation of subsection [ ss:3.4 ] , this set is a disjoint union of the sets @xmath384 , where @xmath28 ranges over all @xmath12-simplices of @xmath383 . the complete description of @xmath384 has been obtained in proposition [ prop:3.11 ] .
at this point we would like to distinguish some of the @xmath12-simplices of @xmath582 .
we set @xmath583{\setminus}a){\,|\,}a\in[5]\}\\ { { \mathcal w}}_2&:=&\{(\{a , b\}{\,|\,}[5]{\setminus}\{a , b\}){\,|\,}a , b\in[5],\ , a < b\}\\ { { \mathcal w}}_3&:=&\{(\{a , b , c\}{\,|\,}[5]{\setminus}\{a , b , c\}){\,|\,}a , b , c\in[5],\,a < b < c\}\\ { { \mathcal w}}&:=&{{\mathcal w}}_1\cup{{\mathcal w}}_2\cup{{\mathcal w}}_3\cup([5 ] ) . \end{array}\ ] ] clearly , @xmath584 , and we shall call the @xmath12-simplices in @xmath585 _ exceptional simplices_. for every @xmath586 , @xmath31 , we now fix a matching on @xmath384 . specifically , we set @xmath587 , see definition [ df : st ] , and let @xmath588 to be the standard matching @xmath589 , for all @xmath590
. for @xmath591 , say @xmath592 , with @xmath573 , @xmath574 , we obtain @xmath588 from the standard matching @xmath589 , by extending it by one more edge , matching @xmath593 with @xmath594 , as in the proof of theorem [ thm : cond3 ] .
we summarize what we know about these matchings in the following proposition .
[ prop:42 ] the matching @xmath588 is perfect for all @xmath595 , and near - perfect for all @xmath596 .
* proof . *
if @xmath391)$ ] , proposition [ prop : sm ] implies that @xmath588 is near - perfect .
assume now @xmath597)$ ] .
by proposition [ prop : sm2 ] the matching @xmath588 is perfect for all @xmath28 which do not appear in table [ table : exc - simp ] .
assume finally that @xmath28 does appear in table [ table : exc - simp ] .
a line - by - line analysis shows that the corresponding prefix family @xmath418 has full prefixes if and only if @xmath596 .
it has one full prefix if @xmath598 , and it has three full prefixes if @xmath591 .
the result now follows from lemma [ lm : sml ] , and the fact that two out of three critical ( with respect to the standard matching ) simplices for @xmath591 have been matched in the extended matching @xmath588 .
we let @xmath599 denote the matching on @xmath582 obtained as a union of all matchings @xmath588 . by proposition [ prop:42 ]
this matching has @xmath600 critical simplices .
we connect these @xmath600 critical simplices by @xmath601 non - intersecting augmenting paths ; the table [ table:21paths ] contains the explicit description of the @xmath601 paths .
this will allow to use the construction from definition [ df : mdef ] and to deform the matching @xmath599 to a perfect matching @xmath602 .
@xmath603 @xmath604 we define @xmath605 , by setting @xmath606 , where @xmath607 is the unique critical @xmath12-simplex in @xmath384 .
note that @xmath608)=-1 $ ] , @xmath609 for @xmath598 , and @xmath610 for @xmath591 .
we now need the following result .
[ lm : path ] assume that we have a path @xmath533 in @xmath582 connecting two exceptional @xmath12-simplices @xmath28 and @xmath55 , such that 1 . @xmath611 ; 2 .
any @xmath12-simplex @xmath397 on the path @xmath533 other than @xmath28 and @xmath55 is non - exceptional , and the corresponding graph @xmath612 is conducting of the second type .
then there exists an augmenting path @xmath136 in @xmath613 such that 1 .
@xmath136 connects the critical simplex in @xmath384 with the critical simplex in @xmath614 ; 2 .
@xmath136 lies entirely in @xmath615
. * proof . *
assume @xmath616 , with @xmath617 , @xmath618 .
for all @xmath199 , choose @xmath12-simplices @xmath619 , @xmath620 such that 1 .
@xmath621 and @xmath622 are adjacent in @xmath613 ; 2 . @xmath623 . to see that such simplices can be chosen let us pick @xmath269 and @xmath624 to be two adjacent @xmath12-simplices in @xmath582 , say @xmath625 , @xmath626 , and @xmath627 . without loss of generality
we can assume that @xmath628 ; in fact , we could always assume that either @xmath629 , or @xmath630 , but we do not need such detail . pick any @xmath631 .
then any @xmath12-simplex @xmath632 in @xmath613 , such that @xmath633 belongs to @xmath634 .
furthermore , it is adjacent to the @xmath12-simplex @xmath635 , which in turn belongs to @xmath636 .
there are @xmath637 such @xmath12-simplices , exactly half of which have orientation @xmath8 , so there is plenty of choice for the simplices @xmath621 and @xmath622 .
note that @xmath638 follows automatically from @xmath639 and @xmath611 .
set @xmath640 to be the critical @xmath12-simplex in @xmath384 , and set @xmath641 to be the critical @xmath12-simplex in @xmath614 . by theorem [ thm : cond3 ] there exist semi - augmenting paths @xmath642 , for @xmath88 , such that each @xmath642 connects @xmath643 with @xmath621 . concatenating these paths
will yield the desired path @xmath136 .
since the paths presented in table [ table:21paths ] are non - intersecting , it follows from lemma [ lm : path ] that the corresponding @xmath601 augmenting paths in @xmath613 are non - intersecting as well . as we said above
, this means that we can deform the matching @xmath599 to a perfect matching @xmath602 , which leads to the proof of our main theorem .
before we proceed with the proof , we need one last piece of terminology .
assume @xmath644 is a vertex of @xmath193 .
the * predecessor * of @xmath241 is the vertex of the simplex @xmath645 of @xmath646 , which has the same color as @xmath241 ; we shall denote it by @xmath647 .
* proof of theorem [ thm : main - math]*. let us describe an assignment @xmath648 of values @xmath7 and @xmath8 to vertices of @xmath11 .
let @xmath649 be such a vertex , and set @xmath650 , @xmath55 is a simplex of @xmath285 . if @xmath651 , then we set @xmath652 .
it is a compliant labeling , since the initial labeling @xmath71 was chosen to be compliant .
we shall call such the assignment @xmath653 the _ default value_. assume now @xmath654 .
if there is only one @xmath12-simplex of @xmath285 containing @xmath55 , or if the two @xmath12-simplices containing @xmath55 are not matched by @xmath602 , then we let @xmath655 be the default value . otherwise , we set @xmath656
. finally , consider the case @xmath657 .
if @xmath55 is not a @xmath8-monochromatic @xmath12-simplex in @xmath71 , then we let @xmath655 be the default value . otherwise , let @xmath658 be the @xmath12-simplex which is matched to @xmath55 by @xmath602 .
we set @xmath659 we shall now see that the obtained assignment @xmath648 has no monochromatic @xmath12-simplices . let @xmath660 be a @xmath12-simplex of @xmath11 , and let @xmath661 be the corresponding @xmath12-simplex of @xmath285 containing @xmath28 .
if @xmath55 is not @xmath8-monochromatic in @xmath71 , then by our construction , all values of @xmath648 on vertices of @xmath55 are the default ones , which means that the vertices of @xmath28 have the same values under @xmath648 as vertices of @xmath55 of the same color under @xmath71 . in particular , @xmath28 is not monochromatic .
assume now @xmath55 is @xmath8-monochromatic under @xmath71 , and assume @xmath55 is matched to @xmath658 under @xmath602 . by our construction the vertex of @xmath28 which has color @xmath662 is assigned the value @xmath8 , hence @xmath28 can not be @xmath7-monochromatic .
let @xmath663 $ ] , such that @xmath664 , and @xmath78 is almost maximal with respect to @xmath665 ; such an element exists due to remark [ rem : amax ] .
let @xmath398 be the color of @xmath28 with the color @xmath78 . by our construction @xmath666 ,
hence @xmath28 is not @xmath8-monochromatic either .
in this section we describe the application of our main theorem in theoretical distributed computing and present an explicite distributed protocol solving weak symmetry breaking for @xmath3 processes in @xmath4 rounds . to start with ,
we record the information obtained in section [ sect:4 ] in a table , which we call * exsimp*. namely , for each path from table [ table:21paths ] use lemma [ lm : path ] to produce an augmenting path @xmath533 in @xmath613 . for every simplex @xmath251 from @xmath533 we add an entry @xmath667 to * exsimp * , where @xmath662 is the color of the edge from @xmath533 which is adjacent to @xmath251 , and which does not belong to original matching @xmath599 ( it does belong to the final matching @xmath602 ) . .... protocol wsb6 ( input : ` $ p$ ` ) ` { \bf write } ` ; ` { \bf read } ` ; ` { \bf write } ` ; ` { \bf read } ` ; assume the current view is ` $ ( \sigma'\ds\tau')$ ` if ` $ \car(\sigma')\neq[n]$ ` and ` $ ( \sigma'\ds\tau')$ ` is not in table ` \ref{table : exv2 } ` then ` { \bf decide 0 } ` ` { \bf write } ` ; ` { \bf read } ` ; assume the current view is ` $ ( \sigma\ds\tau\ds\gamma)$ ` if ` $ |\car(\gamma)|\leq n-1 $ ` then ` { \bf decide~1 } ` if ` $ \car(\gamma)=[n]$ ` then ` $ \xi:=\tau$ ` else assume ` $ \tau=(a_1\s\dots\s a_t\ds b_1\s\dots\s b_t)$ ` , set ` $ q:=[n]\setminus(b_1\cup\dots\cup b_t)$ ` if ` $ q\notin a_1\cup\dots\cup a_t$ ` then ` { \bf decide~1 } ` else pick ` $ 1\leq k\leq t$ ` such that ` $ q\in a_k$ ` set ` $ \xi:=(a_1\s\dots\s a_{k-1}\s q\s a_k\sm q\s a_{k+1}\s\dots\s a_t)$ ` if ` $ p=\match(\sigma,\xi)$ ` then ` { \bf decide~1 } ` else ` { \bf decide~0 } ` function ` \match ` ( input : ` $ \xi,\sigma$ ` ) if ` $ ( \xi,\sigma , c)$ ` is in ` \exsimp ` then return ` $ c$ ` else return ` $ \lev(\xi,\st(\sigma))$ ` ....
the formal protocol is given in figure [ figure : two : protocol ] .
the verbal description is as follows .
let @xmath668 be the i d of the process running the protocol
. * step 1 .
* execute two rounds of write - read sequence .
assume that the view of the process @xmath668
after @xmath414 rounds is @xmath669 .
if @xmath670 $ ] , then proceed to step 2 .
else we have @xmath671 $ ] , in which case we check if @xmath241 is one of the exceptional second round views from definition [ df : exv ] , see table [ table : exv2 ] .
if it is an exceptional second round view from that table then proceed to step 2 , else * decide 0 * and stop
. * step 2 . * execute one more round of write - read .
assume that the view of the process @xmath668
after @xmath4 rounds is @xmath672 . *
if @xmath673 , then * decide 1 * and stop .
* if @xmath674 $ ] , set @xmath675 and proceed to step 3 .
* if @xmath676 , assume @xmath677 .
we have @xmath678 , so there exists @xmath679 $ ] , @xmath680 , such that @xmath681{\setminus}q$ ] . if @xmath105{\setminus}q$ ] , then * decide 1 * and stop .
else , there exists @xmath110 , such that @xmath682 .
now , set @xmath683 * step 3 . *
if @xmath684 , then * decide 1 * , else * decide 0 * and stop . here is the procedure @xmath685 , which has two ordered set partitions @xmath28 , @xmath686 of @xmath14 $ ] as input and returns a value from @xmath14 $ ] .
procedure @xmath685 .
if @xmath687 is in * exsimp * , set @xmath688 , else set @xmath689 we note that the views from table [ table : exv2 ] can also be described verbally .
for example , the view @xmath690 means that process @xmath691 has only seen itself both in the first and in the second round . the view @xmath692 has a more lengthy interpretation .
it means , that the process @xmath691 has seen one other process @xmath693 in the second round , whose i d was larger than that of @xmath691 .
furthermore , @xmath691 has seen 3 processes in the first round , including itself and @xmath693 .
the third process it has seen has i d larger than that of both @xmath691 and @xmath693 . finally , the views of @xmath693 and @xmath691 in the first round were the same ( @xmath691 knows that after the second round ) .
we leave it to the reader to provide similar interpretations for other views listed in the table [ table : exv2 ] .
currently , there are no lower bounds for the number of rounds needed to solve the weak symmetry breaking task . here
we give the first such lower bound , stating that if weak symmetry breaking can be solved at all , one would need at least two rounds to do that .
* let @xmath695 denote the set of pairs of integers @xmath696 .
let us assume that we have a distributed protocol running exactly one round and having output in the set @xmath697 .
after the execution of this protocol each process @xmath430 has seen a certain number of other processes and has determined the relative position of its own i d among the i d s of the processes it has seen .
such an information can be encoded by a pair of integers @xmath698 , such that @xmath699 , this is the number of processes , including itself , which @xmath430 has seen , and @xmath700 , which is the relative position of the i d of @xmath430 .
hence , the decision function for such a protocol is precisely a function @xmath701 .
assume now we are given such a decision function @xmath702 solving the weak symmetry breaking .
let us break up @xmath703 , where @xmath704 .
let @xmath158 be the maximum index such that @xmath705 is not surjective .
since @xmath706 is not surjective , the number @xmath158 is well - defined . without loss of generality
we can assume that @xmath707 .
for all @xmath708 we let @xmath709 be any number such that @xmath710 .
such a number exists for all @xmath711 , since by our construction @xmath712 is surjective for these values of @xmath104 .
we shall now construct an execution @xmath713 of the protocol after which the decision function will assign the value @xmath7 to all processes , which of course contradicts to the claim that the protocol solves weak symmetry breaking . the execution @xmath713 will start with a simultaneous activation of @xmath158 processes , after which the rest of the processes will activate one at a time . in total
the execution @xmath713 has @xmath714 rounds .
these numbers can be determined constructively as follows . to start with
, the numbers @xmath716 , @xmath326 , @xmath721 are initialized to be @xmath7 .
we then proceed in @xmath714 steps . in the first step , we set @xmath722 for all @xmath560
. next , for @xmath104 running from @xmath723 to @xmath0 , repeat the following : set @xmath724 , and increase by @xmath8 each @xmath725 , such that @xmath726 and @xmath727 .
the result after @xmath714 steps is then taken as the final output .
it easy to see that each process activated in round @xmath719 , for @xmath728 , will see @xmath729 processes , including itself , and that the relative position of his i d in what he sees will be @xmath730 .
h. attiya , a. castaeda , _ a non - topological proof for the impossibility of k - set agreement _ ,
stabilization , safety , and security of distributed systems , lecture notes in computer science * 6976 * , ( 2011 ) , 108119 .
h. attiya , a. paz , _ counting - based impossibility proofs for renaming and set agreement _ , 26th international symposium , disc 2012 , salvador , brazil , lecture notes in computer science * 7611 * , ( 2012 ) , 356370 .
a. castaeda , s. rajsbaum , _ new combinatorial topology upper and lower bounds for renaming _ ,
proceeding of the 27th annual acm symposium on principles of distrib .
acm , new york , ( 2008 ) , 295304 .
kozlov , _ topology of immediate snapshot complexes _ , topology appl . * 178 * ( 2014 ) , 160184 .
kozlov , _ structure theory of flip graphs with applications to weak symmetry breaking _ , preprint 42 pages , 2015 . | in this paper we study a family of discrete configuration spaces , the so - called _ protocol complexes _ , which are of utmost importance in theoretical distributed computing . specifically , we consider questions of the existance of compliant binary labelings on the vertices of iterated standard chromatic subdivisions of an @xmath0-simplex .
the existance of such labelings is equivalent to the existance of distributed protocols solving weak symmetry breaking task in the standard computational model . as a part of our formal model ,
we introduce function @xmath1 , defined for natural numbers @xmath0 , called the _ symmetry breaking _ function . from the geometric point of view @xmath1
denotes the minimal number of iterations of the standard chromatic subdivision of an @xmath2-simplex , which is needed for the compliant binary labeling to exist . from the point of view of distributed computing , the function @xmath1 measures the minimal number of rounds in a protocol solving the weak symmetry breaking task .
in addition to the development of combinatorial topology , which is applicable in a broader context , our main contribution is the proof of new bounds for the function @xmath1 .
accordingly , the bulk of the paper is taken up by in - depth analysis of the structure of adjacency graph on the set of @xmath0-simplices in iterated standard chromatic subdivision of an @xmath0-simplex .
on the algorithmic side , we provide the first distributed protocol solving weak symmetry breaking task in the layered immediate snapshot computational model for some number of processes .
it is well known , that the smallest number of processes for which weak symmetry breaking task is solvable is @xmath3 . based on our analysis ,
we are able to find a very fast explicit protocol , solving the weak symmetry breaking for @xmath3 processes using only @xmath4 rounds .
furthermore , we show that no protocol can solve weak symmetry breaking in fewer than 2 rounds . |
the theoretical description of electronic correlations in solids is still a challenge .
even one of the simplest models for correlated electrons , the hubbard model,@xcite exhibits enormous mathematical complexity .
its analysis is far from being complete .
recently , a new class of models for tight - binding electrons with infinitely strong on - site repulsion has been discovered . for particle numbers of two or more particles per unit cell , one can construct the ground states of these models.@xcite these ground states are spin - singlets and have the structure of resonating - valence - bond states@xcite for certain cases .
the class of models which are solvable by this method has been generalized by several authors.@xcite one - dimensional models of this class@xcite allowed for the calculation of equal - time correlation functions@xcite using a transfer - matrix technique .
all equilibrium correlations studied so far decay exponentially .
dynamical correlations obtained numerically for the one - dimensional hubbard chain indicate dispersionless excitation spectra for an electron density of two per cell.@xcite the propagation of one hole , however , shows dispersive delocalized features .
the regime of particle numbers below two per unit cell has not been accessible by analytical methods so far .
the main result of this paper is an analytical upper bound on the ground state energy of systems containing @xmath4 particles ( @xmath0 denotes the number of cells ) .
the trial state and the corresponding upper bound on the ground - state energy are discussed for two examples .
the first example is a hypercubic hubbard model as introduced in ref .. it is shown that the trial state becomes an asymptotically exact eigenstate of this particular @xmath2 -particle system in the thermodynamic limit . for the second system , a linear hubbard chain , the upper bound on the ground - state energy
may be calculated analytically as well . a comparison to exact numerical results@xcite
is presented .
in this section we recall the most general description of the class of solvable models.@xcite this class contains hubbard , anderson and emery models in the limit of infinitely strong interaction ( @xmath5 ) .
the solution presented below requires certain lattice structures .
( in the following , the term _ graph _ is used , because _ lattice _ implies translational invariance which needs not be assumed . )
the models are defined on graphs , where each of the @xmath0 vertices contains a _ cell _ of sites .
( the cells need not be identical either . )
the cells are labeled by an index @xmath6 .
let the set of all sites within the cell @xmath7 be called @xmath8 .
( it is mentioned in passing that @xmath8 is _ not _ a unit cell if the graph is translationally invariant , because neighboring cells may share some sites . ) the sites ( or electronic orbitals ) within a cell @xmath7 are labeled by an index @xmath9 .
some of those orbitals ( not necessarily all ) may carry an infinitely strong on - site repulsion , e.g. a cell may contain `` @xmath10-sites '' which may be empty or occupied by one particle only ( with spin up or down ) and `` @xmath11-sites '' which can be occupied by up to two particles .
the set of @xmath10-sites is called @xmath12 .
the repulsion is incorporated into the model by a projection operator @xmath13 which strictly excludes double occupancy on all sites @xmath14 , where @xmath15 .
the fermi operators @xmath16 ( @xmath17 ) denote creation ( annihilation ) operators and are defined as usual .
the connection between cells is established by common sites .
we denote the set of common sites by the intersection @xmath18 .
examples for some models are given in ref .. figure [ fig:1 ] illustrates the two cases discussed below . in order to construct the ground state of the models defined above
consider the linear combination of fermi operators @xmath19 they obey the following algebra : @xmath20}_+ = { [ { \psi}_{i,\sigma}^{},{\psi}_{j,\sigma ' } ^{}]}_+ = 0 & \\ \nonumber & { [ { \psi}_{i,\sigma}^{},{\psi}_{j,\sigma ' } ^{\dagger}]}_+ = \,\delta_{\sigma,\sigma ' } ^ { } & \\ \nonumber & \times \left\ { \begin{array}{ll } \sum\limits_{\alpha\in{\cal n}_i}\ , { |\lambda_{i,\alpha,\sigma}^{}|}_{}^{2 } & \quad { \rm for}\quad i = j\\ \sum\limits_{\alpha,\beta\in{\cal c}_{ij } } \delta_{(i\alpha),(j\beta)}^ { } \lambda_{i,\alpha,\sigma}^{\star } \lambda_{j,\beta,\sigma}^ { } & \quad { \rm otherwise}\\ \end{array } \right . \ , . & \end{aligned}\ ] ] ( the kronecker symbol @xmath21 assumes the value one if the indices @xmath22 and @xmath23 label the same site and equals zero otherwise . ) the coefficients @xmath24 define the parameters of the hamiltonian which will be constructed below .
the @xmath25-operators allow to write down a positive semidefinite@xcite hamiltonian in a very concise way : @xmath26 however , the physical meaning of this hamiltonian becomes more transparent if it is rewritten in terms of the fermion @xmath27-operators by using the relation @xmath28 and the definition of the @xmath25 operators as defined in eq .
( [ eq:2 ] ) .
the hamiltonian ( [ eq:4 ] ) is transformed into the following form : @xmath29 the first term in @xmath30 describes particle hopping within a cell where each site ( orbital ) is connected to all others .
( particle transfer between interacting orbitals and orbitals without interaction is usally called a hybridization . in that case
, the model would describe e.g. an anderson model . )
the remaining terms contain trivial constants and sums over occupation numbers which add up to the total particle - number operator for the case of the @xmath3-dimensional hubbard model as discussed below .
if the sets @xmath31 and @xmath8 are not identical or if not every site within a cell is connected to a neighbor cell , the sums over occupation number operators may add up to an additional field which shifts the energy of certain sites .
such a model then contains more than one parameter and the exact ground state as derived below is only valid on a parameter surface defined by the quantities @xmath32 . in the following ,
all @xmath32-coefficients are assumed to be real@xcite .
the models as defined above allow to write down an exact ground state if the system contains two or more particles per cell .
the ground state is a gutzwiller projected slater determinant : @xmath33 . because of the equation @xmath34 , the state eq .
( [ eq:7 ] ) is an eigenstate of @xmath35 ( [ eq:4 ] ) , the corresponding eigenvalue equals zero .
further , @xmath35 is positive semi - definite@xcite .
therefore the state ( [ eq:7 ] ) belongs to the ground - state manifold . in ref
tasaki provides a proof of the uniqueness of the ground state if the state @xmath36 is equal to the vacuum state @xmath37 ( @xmath1 particles ) .
larger fillings can be discussed if @xmath36 contains additional particles , for that case it immediately follows by construction that the state ( [ eq:7 ] ) is degenerate .
if all sites carry an infinitely strong on - site repulsion and , e.g. the @xmath32-coefficients are set to unity , the state ( [ eq:7 ] ) exhibits a structure often called resonating - valence - bond state.@xcite in that case the ground state consists of a linear combination of local singlet bonds:@xcite @xmath38 .
the projector ensures that no two singlet bonds have a site in common .
although the ground state is known for two or more particles per cell , the only analytical result that has been obtained for particle numbers @xmath39 below the `` magic '' number @xmath40 so far is the trivial lower bound on the ground state energy @xmath41 which is valid for _ any _ number of particles .
[ the hamiltonian ( [ eq:4 ] ) is non - negative . ]
the present section provides an upper bound on the ground state energy of @xmath35 for one additional hole , e.g. a particle number of @xmath42 using a variational state .
the quality of variational methods is not very well controlled , there exists no small parameter , only an inequality for the energies .
therefore the method requires some intuition in order to `` hit '' the correct physics by guessing a good state .
the first guess of a variational state , the state latexmath:[\[\label{eq:8 } : = p { \psi}_{i,\uparrow}^{\dagger } \prod_{j,\,\ , j\neq i } { \psi}_{j,\uparrow}^{\dagger } { \psi}_{j,\downarrow}^{\dagger } @xmath1 -particle physics unchanged as possible .
the state is constructed in a way that most terms of the hamiltonian generate zero - states due to the relation @xmath44 . from @xmath45 and @xmath46
( @xmath47 denotes the ground state of the system with @xmath1 particles of the previous section ) one obtains the diagonal matrix @xmath48 if the norm of the states @xmath49 and @xmath47 were known , the expectation value of @xmath35 could be evaluated .
unfortunately , this is not possible in general .
however , the following orthogonal and normalized [ see eq .
( [ eq:9 ] ) ] trial states latexmath:[\[\label{eq:10 } : = \frac { \sqrt{h } } { \sqrt{\langle\phi_0|\phi_0\rangle } } value .
( the positive square root of @xmath35 is well defined because @xmath51 . )
surprisingly , the hamiltonian matrix elements @xmath52 of the @xmath2 -particle system can be expressed in terms of occupation - number expectation - values of the @xmath1-particle system in the following way : @xmath53 where @xmath54 in the following section , the upper bound eq .
( [ eq:11 ] ) on the ground - state energy is discussed for two examples which allow us to calculate the expectation value @xmath55 .
the first example , where the upper bound on the ground - state energy of the @xmath2 -particle system may be evaluated is the hubbard model as introduced in ref .
, where all @xmath32 s are set to unity and the spatial dimension is larger than one . for the three - dimensional case ,
the cells have the topology of octahedra with @xmath56 sites .
all sites carry interaction , therefore the sets @xmath31 and @xmath8 are identical .
further , every site in @xmath8 is shared by exactly one neighboring cell , forming a @xmath3 dimensional ( hyper)cubic lattice of cells .
periodic boundary conditions are assumed .
the corresponding lattice is visualized in fig .
[ fig:1]a .
the hamiltonian of this system reads @xmath57 the last term in ( [ eq:13 ] ) contains a trivial constant @xmath58 and the total electron - number operator @xmath59 ( with eigenvalue @xmath39 ) which is a conserved quantity . with the fourier
transform @xmath60 ( the occupation probabilities for all sites are equal for symmetry reasons ) one obtains an upper bound@xcite on the @xmath2 -particle ground state energy @xmath61 the minimum @xmath62 with respect to the @xmath63-vector is obvious : for @xmath64 the right hand side of ( [ eq:15 ] ) becomes zero .
however , this @xmath63-vector has to be _ excluded _ for this particular model , because it can be shown@xcite ( c.f .
[ app:1 ] ) that the state @xmath47 ( [ eq:7 ] ) vanishes if @xmath65 is allowed . that particular @xmath63-vector can be excluded by the restriction to systems where one of the @xmath3 system dimensions contains an odd number of cells .
then the @xmath63-vector component corresponding to this particular direction can not equal @xmath66 .
this result indicates that the state @xmath67 becomes an asymptotically exact @xmath2 -particle ground state of @xmath35 for the thermodynamic limit .
the upper bound on the ground state energy vanishes as @xmath68 and asymptotically approaches the lower bound @xmath69 .
this result is even relevant for @xmath70 , where the asymptotically exact state describes one hole in a half filled rvb - background .
the absence of ferromagnetism does not contradict nagaoka s theorem@xcite because the present lattice is not bipartite .
( for a bipartite lattice , the state would be a ferromagnet , i.e. the total spin would assume its maximum value . )
unfortunately , no rigorous statements on the uniqueness of this ground state can be made .
further , it can be seen that for @xmath71 the chemical potential of the system for particle numbers of @xmath72 remains constant .
the non - interacting tight - binding analogue exhibits quite similar behavior : the band structure of this particular lattice consists of one cosine band with @xmath73 and @xmath74 flat bands at the upper band edge of the extended band .
however , the `` bandwidth '' of the interacting system is reduced by a factor @xmath75 , which may be interpreted as an increased mass .
this result sheds new light on the conductivity in the system .
it might be that the many - particle state describes a dispersive delocalized hole which may contribute to conduction .
unfortunately , no results on hole _ densities _ have been obtained and the dispersive state is not exact except in the vicinity of @xmath65 . as a second example
, we study a linear chain with @xmath0 cells and @xmath76 sites per cell @xmath8 ( @xmath77 sites per _ unit cell _ ) .
the topology of this chain is illustrated in fig .
[ fig:1]b ; it may be interpreted as a chain of connected tetrahedra .
again , periodic boundary conditions and the case of infinite on - site repulsion on every site are assumed , and thus the sets @xmath8 and @xmath31 are identical .
the cells @xmath7 and @xmath78 are connected by one site ( backbone site ) : the sites with indices @xmath79 and @xmath80 are identical .
all @xmath32 s are set to unity .
the hamiltonian of this particular chain reads @xmath81 in addition to the hopping terms , one obtains a local field which shifts the energy of the backbone sites ( @xmath82 ) . for this model ,
the upper bound eq .
( [ eq:11 ] ) reduces to the simple expression@xcite @xmath83 the right hand side of eq .
( [ eq:18 ] ) assumes a minimum for @xmath84 ( which is allowed for this model ) .
this leads to the following upper bound on the ground state energy : @xmath85 the occupation probability on the backbone sites may be calculated by a transfer - matrix technique for the thermodynamic limit .
this calculation is briefly outlined in appendix [ app:2 ] and yields @xmath86 in contrast to the previous example this result indicates a gap in the @xmath2 -particle spectrum .
equation ( [ eq:20 ] ) estimates this gap from above .
therefore the upper bound is compared to numerically obtained exact results as described in ref .. the numerical results have been obtained by the conjugate - gradient algorithm.@xcite the energies are listed in tab .
[ tab:1 ] . for chain lengths
up to four unit cells , the ground state is calculated without applying symmetry restrictions . in order to check for ground - state degeneracy
, the determination of the lowest eigenstate is repeated in the subspace orthogonal to the ground state .
for an even number of unit cells , the energies of these orthogonal eigenstates do not coincide with the ground - state energy , therefore the ground states for even chain length are non - degenerate .
( chains with an odd number of unit cells are degenerate because of reflection symmetry which induces degeneracy with respect to @xmath87 . ) as a second step , the same calculations have been performed assuming site - interchange symmetry as decribed in ref . and translational invariance with @xmath84 .
the energies of the symmetrized states coincide with the previously calculated non - degenerate ground - state energies , therefore the symmetry of the ground state is uniquely determined .
the chain with six cells is numerically treated by application of site interchange symmetry only and the result for @xmath88 is obtained assuming translational invariance in addition to site - interchange symmetry .
the numerically obtained exact ground - state energies @xmath89 are shown in tab .
[ tab:1 ] .
the upper bound eq .
( [ eq:20 ] ) of the infinite system deviates from the numerical result for @xmath88 by @xmath90 .
this agreement is not very close .
however , the symmetries of the variational state and the numerically exact state coincide .
better quantitative agreement may be obtained by the expectation value of @xmath35 with respect to the state @xmath91 ) . ] , which may be evaluated numerically . for a chain with @xmath92 one
obtains @xmath93 .
this result deviates from the numerically exact result by @xmath94 only .
the state @xmath67 is proportional to @xmath95 , therefore a difference in the expectation values @xmath96 and @xmath97 simply reflects the fact that @xmath98 ( and hence @xmath99 ) is not an eigenstate .
the operator @xmath100 `` amplifies '' the deviation from the exact ground state .
the single - particle band - structure of this particular chain is illustrated in fig .
[ fig:2 ] and consists of two extended bands , one broad band separated from a narrower one by a gap of the size of one energy unit .
the highest energy band collapses to one dispersionless flat band which is again separated by a small gap . for @xmath1 particles the non - interacting system would have a completely filled lower valence band describing an insulator .
it is obvious that the interacting system shows different behavior here : in the framework of the one - particle language , one would expect a gap if the particle number is changed from @xmath40 to @xmath101 , however , for that case the chemical potential of the interacting system remains constant .
in contrast to the one - particle picture , the chemical potential of the interacting system changes if the particle number is changed from @xmath42 to @xmath40 . this contradiction illustrates the breakdown of one - particle physics in this particular model .
for a class of models describing strongly interacting fermions it is possible to write down the exact ground state if the particle number equals 2 per unit cell or more .
the particle number regime below this threshhold has not been accessible by rigorous methods so far .
we present a tractable variational state for systems with one additional hole .
the corresponding expectation value of the hamiltonian approaches a lower bound for the case of a particular @xmath3-dimensional hubbard model in the thermodynamic limit .
we conclude that the presented variational state approaches the ground state or an eigenstate which is degenerate to the ground state if the ground state is not unique . for a second model , a special hubbard chain ,
the analytical expectation value deviates significantly from the numerically calculated ground state energy . although it is shown that the numerical ground state and the variational state exhibit the same symmetries , it remains unclear , how well the ground - state physics of the chain is described by our variational ansatz .
this appendix provides a proof that the state @xmath47 [ eq . ( [ eq:7 ] ) ] of the @xmath1- particle @xmath3- dimensional system vanishes , if the system extends an even number of unit cells in each direction .
the state and the hamiltonian are expressed in terms of creation ( annihilation ) operators @xmath102 ( @xmath103 ) referring to non - orthogonal one - particle states .
however , they must be linear independent , otherwise the corresponding slater determinant vanishes
. there is no problem for models with non - equivalent sites , however , the @xmath3-dimensional hubbard model requires some care .
consider the slater determinant @xmath104 the matrix @xmath105 has the elements @xmath106 where @xmath107 equals one if @xmath7 and @xmath108 are nearest neighbors and zero otherwise .
the structure of @xmath105 is similar to a @xmath3-dimensional hypercubic tight - binding problem with periodic boundary conditions .
the determinant is equal to the product over all eigenvalues of the above tridiagonal matrix .
the eigenvalues are given by the familiar cosine bands @xmath109 .
if all system length contain an even number of unit cells , one of the eigenvalues vanishes , because one of all possible @xmath63- vectors equals @xmath110 . therefore , the state @xmath111 and the projected state @xmath112 vanishes as well .
the condition that at least one system length must be odd is a _ necessary _ condition for this model .
we briefly recall the transfer matrix calculation of the occupation probability for certain sites of the hubbard chain .
this calculation requires open boundary conditions .
however , the influence of the boundary conditions decreases with increasing system size .
the ground state of this chain can be decomposed into a sum over all possible configurations of @xmath0 singlet bonds ( dimers ) .
every component of the ground state has exactly one dimer in each cell .
let the singlets ( dimers ) be written as @xmath113
. excludes any double occupancy .
therefore the indices @xmath9 and @xmath115 have to be different within a cell .
this condition is not sufficient , because double occupancy on the backbone sites could occur due to certain configurations of neighbor cells : if e.g. cell @xmath7 contains the singlet @xmath116 and cell @xmath78 the singlets @xmath117 , @xmath118 , or @xmath119 the connection site would be occupied by two particles .
these neighbor configurations have to be excluded .
the restriction to allowed neighbor configurations is coded by the transfer matrix : @xmath120 with these definitions , the norm of the ground state @xmath121 for a chain with @xmath0 cells may be written as @xmath122 the occupation of one @xmath123-site well within the chain ( the cell index @xmath124 should be @xmath125 ) is given by the expression @xmath126 which may be simplified by using a spectral representation of the matrix @xmath127 .
the eigenvalues of @xmath127 are @xmath128 , @xmath129 and a fourfold degenerate eigenvalue @xmath130 .
the right and left eigenstates may be calculated by @xmath131 and @xmath132 .
they are mutually orthogonal and have to be normalized according to @xmath133
. then the matrix @xmath134 reads @xmath135 as usual , the contribution of the largest eigenvalue dominates the results with growing system size . in the thermodynamic limit
the total occupation of the @xmath123-sites becomes @xmath136 the backbone occupation may be calculated analogously or simply by applying @xmath137 .
it is possible to construct a model with periodic boundary conditions , where not all phases of the @xmath32-coefficients can be removed by a gauge transformation.@xcite the implications of this fact are disregarded here . | we study a class of exactly solvable models for strongly correlated electrons , defined on a set of @xmath0 cells , and with infinite on - site repulsion on part of the sites of each cell . for @xmath1 or more electrons
the exact ground state is known .
we construct a tractable @xmath2-particle state which becomes asymptotically degenerate to the ground state in the thermodynamic limit for one special @xmath3-dimensional model . for other models ,
that state may be used to calculate variational upper bounds on the ground - state energy . for a hubbard chain with three sites per unit cell ,
the analytical bound is compared to numerically exact results . |
active galactic nuclei ( agn ) are envisioned to be accreting supermassive black holes at the centers of galaxies .
monitoring variability in these sources is an important way to extract information about the compact object within , as the timescales over which this variability is seen can provide information about the size and structure of the region from which the emission originates in order to place physical constraints on the central engine within .
agn are known to show variability on a variety of timescales : the optical line emission and continuum flux from quasars ( qsos ) has been observed to vary on timescales of months to years , while x - ray flux from these sources varies on shorter timescales of hours to days ( peterson 2001 @xmath1 references therein ) .
the mechanism behind this variability is still uncertain , though many theories have been formulated to explain this phenomenon .
the leading explanation for the observed variability is that changes in magnitude are related to accretion events or disk instabilities @xcite .
other explanations for variability include changes in magnitude due to supernovae , starbursts , or microlensing events ( e.g. , aretxaga @xmath1 terlevich 1994 ) .
it has been shown that 80 - 100% of agn candidates vary in the optical over the course of several years ( e.g. , koo , kron , @xmath1 cudworth 1986 ) .
webb & malkan ( 2000 ) found that 60% of low - luminosity agns ( llagns , those with luminosities m@xmath2@xmath3 - 23 ) varied when studied over a period of months .
thus , optical variability is an important and successful method for identifying agn in imaging surveys .
indeed , there is some evidence that llagn exhibit greater amplitude flux changes than do their brighter counterparts @xcite , making variability a particularly effective means for their identification .
llagn represent a significant contribution to the x - ray , ir , and uv backgrounds , so knowledge of their space and luminosity distribution at a range of redshifts is important .
they may also represent an important phase of galaxy evolution , which connects normal and active galaxies @xcite .
this paper presents the initial results from a variability survey of the great observatories origins deep survey ( goods ) south and north fields to identify optically varying agn , particularly llagn , at z@xmath41 using small aperture photometry to isolate galactic nuclei .
this technique has been shown to successfully identify faint agn in high - resolution hst multi - epoch imaging surveys @xcite .
the goods images were obtained with the hst advanced camera for surveys ( acs ) over five epochs separated by 45-day intervals , allowing for optical variability to be monitored over a six - month period .
our initial study focuses on a pre - selected sample of galaxies in the goods - south field based on x - ray and/or ir properties . in this paper
, we investigate the optical variability of agn candidates having a range of obscuring properties , from soft x - ray sources to highly obscured mid - ir sources lacking x - ray emission .
the aim of our study is twofold : 1 ) to confirm via optical variability the agn nature of candidates identified in multiwavelength surveys , and 2 ) to quantify the use of optical variability in identifying various types of agn for use in interpreting the results of our larger forthcoming survey of all optical sources in the goods south and north fields .
the galaxies in this study were chosen from two catalogs .
the first is a sample of mid - infrared - selected agn candidates from alonso - herrero et al .
( 2006 ) with optical counterparts in the goods - south field .
these galaxies were selected using spitzer / mips 24 @xmath5 m observations and are well - fit with a power law spectral energy distribution ( sed ) through the spitzer irac bands at 3.6 , 4.5 , 5.8 , and 8 @xmath5 m .
the second sample consists of x - ray sources compiled by giacconi et al .
( 2002 ) and alexander et al .
( 2003 ) using the chandra x - ray observatory .
the chandra deep field south ( cdfs ) , which encompasses the goods - south acs field , is a combination of 11 individual pointings taken with the advanced camera imaging spectrometer ( acis - i ) covering a range of 0.2 - 10 kev with a total exposure time of 942 ks . in the next sections we discuss how
the sources in these catalogs were cross - matched with optical sources in the acs fields for inclusion in our variability survey .
it was first necessary to estimate the optical magnitude to which photometry could be accurately and robustly obtained . to do this , we obtained photometry of optical sources in four epoch 1 frames ( sections 4 , 8 , 10 , and 12 ) to cover a large spatial range in the goods field .
we used the iraf tool daofind to locate all sources in order to construct a histogram of number vs. nuclear ( r = 0.075 " or 2.5 pixels ) v magnitude .
we measured the magnitude of the galaxy nuclei as we would be doing to identify nuclear variability in the selected samples .
this histogram revealed a sudden drop in number counts beginning at a nuclear v magnitude of 27 .
we therefore chose a magnitude cutoff of 27 and any objects with average magnitudes fainter than v@xmath6 = 27 were not considered in our analysis .
this limit is also consistent with the limiting magnitude to which nuclear variability may expect to be detected using fixed aperture photometry for galactic nuclei extending to z@xmath41 . in the current dataset , we expect to become less sensitive to a varying nucleus within a host galaxy at nuclear v magnitudes of @xmath427.5 .
kasm ( 2003 ) , section 3 , for further discussion of this effect . ]
alonso - herrero et al .
( 2006 ) identified 92 galaxies within the cdfs which were detected at 24 @xmath5 m and display power law - like emission from 3.6 to 8 @xmath5 m .
these galaxies were selected at 24 @xmath5 m because at z@xmath71 a large fraction of these objects should be ultra - luminous infrared galaxies ( ulirgs , with l@xmath8@xmath710@xmath9 l@xmath10 ) .
ulirgs are known to have a steep power law - like sed in the infrared @xcite and many are classified as qsos .
alexander et al .
( 2005 ) showed that at least 40% of ir - luminous high - redshift galaxies contain agn , and similar results at z=1 - 2 have been found with recent spitzer observations @xcite .
all of the sample galaxies are luminous in the ir : 30% are hyper luminous ( l@xmath8@xmath710@xmath11 l@xmath10 ) , 41% are ulirgs ( l@xmath8@xmath410@xmath9 - 10@xmath11 l@xmath10 ) , and all but one are luminous ir galaxies ( lirgs , l@xmath8@xmath410@xmath12 - 10@xmath9 l@xmath10 ) .
alonso - herrero et al .
( 2006 ) have further quantified the multiwavelength seds to classify their sources as either broad line agns ( blagns ) or narrow line agns ( nlagns ) .
we matched sources identified in the acs goods field from the version 1.1 source catalog derived from the stacked v - band image with the ir catalog and found 37 matches within a 3.5 `` radius , 30 of which fell within 1 '' .
of these matches , 11 were too faint for optical photometry ( fainter than 27th magnitude in a single v - band epoch ) and 5 were not visible in all five goods epochs .
this left a total of 22 targets .
all final optical sources matched the ir coordinates within 1.1 " .
agn as a class are known to be x - ray sources and are commonly selected via x - ray surveys @xcite .
we matched the coordinates of objects in the alexander et al .
( 2003 ) chandra x - ray catalog with the goods version 1.1 source catalog and found 200 matches within a 1.9 " radius . of these optical counterparts , 80 were too faint for optical photometry , 16 were not visible in all five epochs , one was contaminated by a cosmic ray , and one optical source was added due to a double nucleus .
finally , one additional source was found by matching the goods catalog coordinates with the giacconi et al .
( 2002 ) x - ray catalog , which also contains the 102 x - ray sources from alexander et al .
thus , a total of 104 x - ray - selected objects were photometered to look for optical variability .
fourteen of these objects overlap the ir - selected sources discussed in section 2.1 .
aperture photometry was performed on all pre - selected sources in each epoch of the version 1.0 goods - south v - band images using an aperture radius of 2.5 pixels ( 0.075 " ) .
all images were visually inspected to ensure consistent centering across all five epochs .
the photometry yielded a light curve for each source . in order to quantify variation and pick out galaxies varying significantly above the photometric noise
, we calculated the average magnitude , the standard deviation around the average ( @xmath13 ) , and the typical photometric error which was calculated using the following formula : @xmath14 here , error@xmath15 is the formal magnitude error for a galaxy nucleus in each epoch and n is the total number of measurements ( in this case n=5 , for 5 epochs ) .
error@xmath16 is then the rms of the magnitude errors ( error@xmath15 ) , essentially giving an error bar on the standard deviation ( @xmath13 ) .
we then defined the `` significance parameter '' of each object s variability in the following way : @xmath17 by dividing the standard deviation by the rms of its errors , we compare the amount of an object s variation around its average magnitude by its typical photometric error .
thus , the significance parameter of each object s variability is simply a measure of its standard deviation normalized by the rms of the photometric error .
this significance parameter should not be interpreted in terms of statistical probabilities , since the number of points ( 5 epochs of data ) is not enough to determine a gaussian distribution .
nonetheless , this quantity does provide a valid measurement of the level of variability found for each galaxy in our sample .
we compare the significance values against optical , ir , and x - ray properties to search for correlations between optical variability and properties at other wavelengths .
table 1 lists all of our objects and their optical , ir , and x - ray properties .
figure 1 shows the standard deviation for each galaxy nucleus ( @xmath13 ) vs. average nuclear v magnitude .
the error bars are error@xmath16 for each source , which is the rms of the photometric errors in each epoch , and the solid line is 3 @xmath18 error@xmath16 .
objects which have large standard deviations and relatively small values of error@xmath16 clearly stand out above the solid line .
we have chosen a significance value of 3 as our variability threshold , although throughout the paper and figures we carry along the significance value for each source .
a significance of 3 or greater means that the standard deviation is at least three times the amount of change in the photometric errors over the five epochs .
figure 2 shows variability significance vs. average nuclear v magnitude for all objects .
the solid line at a significance value of 3 is the same as the 3 @xmath18 error@xmath16 line in figure 1 .
x - ray sources are shown as black triangles , ir sources are shown as gray squares , and objects that appear in both catalogs are shown as asterisks . we find a total of 29 variables out of the 112 galaxies selected via x - ray and mid - ir emission , resulting in a 26@xmath0 variability rate for our candidates .
eight of the 14 objects ( 57@xmath0 ) which are both mid - ir power law - selected agn candidates and x - ray sources show significant optical variability .
many of these are also among the brightest optical sources .
detection in both x - rays and the mir indicates the presence of an agn with some dust present to reprocess a portion of the light and re - emit it in the infrared . among the other variable objects , those detected only in x - rays are likely
agn enshrouded by little dust , resulting in no significant reprocessed light and were thus not identified in the mid - ir - selected sample .
two of the objects detected as optical variables were identified via mid - ir power - law seds only and do not show x - ray emission ( squares ) .
these sources are likely obscured agns that do not show up in x - rays because much of the x - ray , uv , and optical light is reprocessed into the mid - ir .
thus , while these objects would not be selected as agn via x - ray emission , they do show optical and ir evidence of their agn nature . in total , we find that 10 out of 22 mid - ir - selected agn candidates ( 45@xmath0 ) and 27 out of 104 x - ray - selected agn candidates ( 26@xmath0 ) show optical variability over the course of 6 months .
we further explore the properties of the mid - ir - selected agn in figure 3 .
alonso - herrero et al .
( 2006 ) separate blagn - like seds from nlagn - like seds at @xmath19 @xmath20 -0.9 , where steeper ( i.e. , more negative ) values represent nlagns and shallower power law seds are classified as blagns .
this figure reveals that 7 out of 11 galaxies classified as blagn are optical variables ( 64@xmath0 ) , 2 out of 4 borderline sources are variable ( 50@xmath0 ) , and only one of 7 nlagns shows optical variability ( 14@xmath0 ) .
traditionally we expect that optical variability should be most observable among type 1 agn , since the leading theory for the cause of variability is based on instabilities within the accretion disk @xcite . these instabilities would be most apparent in type 1 agn , which give us a clear view of this part of the system in the unified model .
therefore , our findings are consistent with this expectation , but also show that many borderline and a small fraction of nlagns can also be identified as optical variables . of the x - ray sources in our study , only 26@xmath0 are found to be optical variables . to interpret these statistics , we first calculated the x - ray - to - optical flux ratios for the x - ray sources to determine that they are indeed agn .
we use the r - band magnitudes for the x - ray sources from giacconi et al .
( 2002 ) , taken with the fors1 camera at the vlt and the wide field imager ( wfi ) on the eso - mpg 2.2 meter telescope at la silla .
there are 93 objects with published r magnitudes in our sample of 104 x - ray objects ( 89@xmath0 ) .
we find that the x - ray sources in our survey cluster around log(f@xmath21/f@xmath22)=0 and cover the range between log(f@xmath21/f@xmath22)=1 to -2 , which is the general range of values for agn @xcite .
we also find four sources with log(f@xmath21/f@xmath22)@xmath23 - 2 ( aid 182 , 189 , 192 , and 207 ) .
these sources have optical variability signficance values of 0.83 , 4.88 , 2.09 , and 4.64 , respectively .
the two with high significance values ( @xmath73 ) may be agn which simply have low f@xmath21/f@xmath22 values , consistent with some optical variables identified in the hdf - n @xcite .
the other two , having both low variability significance and low f@xmath21/f@xmath22 , may in fact not be agn . thus , 98@xmath0 of the x - ray sources have f@xmath21/f@xmath22 values consistent with agn . to investigate trends among variability and x - ray band ratio , we used the published values of alexander et al .
( 2003 ) , where the band ratio ( br ) is the ratio of the hard x - ray band ( 2 - 8 kev ; hb ) source counts to the soft x - ray band ( 0.5 - 2 kev ; sb ) counts .
we arbitrarily separate hard " sources from soft " sources at a band ratio of br=0.5 , with soft sources having [email protected] . of the 100 x - ray - selected agn candidates for which a band ratio is measured , 37@xmath0 are soft sources and 63@xmath0 are hard sources .
figure 4 shows variability significance vs. x - ray band ratio .
we find that 19 out of the 37 x - ray sources with soft band ratios are variable ( 51@xmath0 ) .
there are very few variable objects with band ratios greater than 0.5 and there are no variables among those sources with band ratios harder than 2 .
we find only 8 variables out of 62 sources with a band ratio greater than 0.5 .
these findings are consistent with the expectation that harder band ratios indicate a more obscured source , which will likely show less optical variability as dust may be obscuring and reprocessing the optical agn light . objects which exhibit soft band ratios are less obscured and therefore much easier to detect as variables . in summary ,
51@xmath0 of soft x - ray sources ( [email protected] ) are optical variables , 16@xmath0 of sources having band ratios in the range 0.5@xmath23br@xmath232 are variables ( 8 of 48 ) , and no objects with br@xmath72 are detected as optical variables .
we clearly observe a decreasing level of optical variability with increasing x - ray hardness .
webb @xmath1 malkan ( 2000 ) find that the amplitude of optical variability in agn increases on timescales of 1 - 100 days , with 60@xmath0 of agns varying on month - to - month timescales .
the objects in their study are estimated to have magnitudes similar to the objects in our sample .
the results from figure 4 are consistent with this expectation , as we see 51@xmath0 of objects with soft band ratios showing variability .
the 49@xmath0 of soft sources that do not show variability are thus likely to be agn that do not have variability timescales to which our survey is sensitive - our baseline of months is not sufficient to pick up variability in these sources .
webb @xmath1 malkan also found that the observed variability does not appear to depend on other agn properties ; a similar study over the same timescale of the same galaxies would reveal the same percentage of variability , though the specific galaxies found to show the greatest variability would change .
thus , among the soft x - ray sources , it is plausible that all would show optical variability if observed over a longer time baseline ( i.e. , several years ) .
this interpretation is also valid for the mid - ir - selected sources .
we found that 50 to 63@xmath0 of the blagn / borderline seds are optical variables .
again , all may be variable when observed over longer time baselines .
the drop in the fraction of optical variables among harder x - ray sources and the nlagn seds among ir - selected sources , however , is likely due to increased levels of obscuring material in these types of agn . for the many non - variable x - ray sources ,
a coexisting possibility is that some fall into the class of x - ray bright optically normal galaxies ( xbongs ) , objects which appear as agns via x - ray flux but lack optical evidence for accretion ( comastri 2002 ) .
such objects tend to have hard x - ray flux ratios indicative of obscured accretion .
some of these objects can be explained as agn which are dominated by their host galaxy light , which would dilute the agn light and mask optical evidence for accretion
. additionally , rigby et al . ( 2006 ) studied hard x - ray - selected agns in host galaxies having a wide range of inclination angles .
they concluded that the optical dullness in some agn within host galaxies that are not face - on or spheroidal may be a result of obscuring material aligned with the host galaxy and far from the ionizing nuclear activity .
this is consistent with our finding of fewer optical variables among harder x - ray sources .
finally , we have investigated whether there is any relationship among the published x - ray band ratios and the multiwavelength sed classifications of alonso - herrero et al .
( 2006 ) , and how that might relate to the detection of optical variability .
interestingly , of those sources selected via both x - ray and mid - ir criteria , most ( 10 out of 14 ) have soft band ratios .
somewhat surprisingly , the hardest source from this subset of galaxies ( aid 179 ) is classified with a blagn sed .
all of the optical variables detected in both x - rays and via mid - ir selection are soft x - ray sources with [email protected] .
thus , we do not observe a clear correlation between the x - ray and multiwavelength sed agn classifications .
similarly , barmby et al .
( 2006 ) finds only marginal agreement between agn classification determined separately based on mid - ir seds and x - ray band ratios , although the amount of obscuration present should in principle affect both wavelength regimes .
they list variations in the gas - to - dust ratio or a range of intrinsic agn properties as possible reasons that the ir and x - ray classifications do not tend to agree .
such reasons may also explain the many non - x - ray - detected galaxies with ir properties similar to those that are detected in x - ray surveys .
we have investigated the presence of optical variability in agn candidates selected via x - ray emission and mid - ir power law seds .
we find a range of variability amplitudes among these candidates , with 26@xmath0 having optical flux variations around the mean that are at least 3 times the typical photometric error .
the most significant variables are both mid - ir- and x - ray - selected agn and are bright optical sources .
most of the variables , however , are x - ray - only selected sources with soft band ratios , indicating a relatively unobscured agn with negligible amounts of dust near the ionizing source
. these findings are consistent with the expectation that optical variability selects primarily type 1 , unobscured agn .
we find several indications , however , that optical variability is also observable among more obscured agn , albeit with a much lower detection rate .
eight of the optically variable x - ray sources in our survey have band ratios greater than 0.5 .
a higher band ratio indicates higher column densities of obscuring material along the line of sight .
we find that the optical variability significance decreases with increasing amounts of obscuration .
using the multiwavelength sed classifications of alonso - herrero et al .
( 2006 ) , we find that while most ( 70@xmath0 ) of the optical variables are classified as blagns , 20@xmath0 are borderline nlagn / blagn seds and 10@xmath0 ( 1 of 10 ) have a nlagn sed . finally , we detected optical variability for two mid - ir power law - selected agn that are not detected in x - rays .
such sources may be heavily obscured agn , where x - rays and much of the optical / uv light is blocked by dust which re - emits the light in the mid - ir .
these sources are among the faintest optical sources in our survey and also lie close to the variability threshold . a large fraction of the agn candidates ( 74@xmath0 ) ,
however , do not show optical variability significant enough to produce flux changes greater than 3 times the typical photometric error . among the relatively unobscured sources ( i.e. , those detected in the x - ray with band ratios less than 0.5 ) , about 50@xmath0 are not variable .
this is consistent with the expectation that only @xmath460@xmath0 of agn show optical variability on month - to - month timescales @xcite .
thus , the majority of unobscured agn that do not show optical variability may be optical variables when observed on longer time intervals of years . the inclusion of additional epochs of imaging data for goods - s will help to answer this question .
another possibility is that some of the sources are in the class of optically dull " galaxies which exhibit agn x - ray luminosities but show no optical evidence for accretion .
some may be dominated by light from the non - varying host galaxy , while others may reside in galaxies with large amounts of obsuring material in the host .
this paper is based on observations with the nasa / esa hubble space telescope , obtained at the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 and observations with the keck telescope , made possible by the w. m. keck foundation and nasa .
funding was provided by stsci grant ar-09948.01-a and nsf career grant 0346691 .
alexander , d.m . ,
bauer , f.e . ,
brandt , w.n . ,
schneider , d.p . ,
hornschemeir , a.e . ,
vignali , c. , barger , a.j . ,
broos , p.s . ,
cowie , l.l . ,
garmire , g.p . ,
townsley , l.k . ,
bautz , m.w . ,
chartas , g. , sargent , w.l .
2003 , aj , 126 , 539 alonso - herrero , a. , prez - gonzlez , p. g. , alexander , d. m. , rieke , g. h. , rigopoulou , d. , le floch , e. , barmby , p. , papovich , c. , rigby , j. r. , bauer , f. e. , brandt , w. n. , egami , e. , willner , s. p. , dole , h. , huang , j .- s .
2006 , , 640 , 167 barmby , p. , alonso - herrero , a. , donley , j.l . ,
egami , e. , fazio , g.g . ,
georgakakis , a. , huang , j .- s . , laird , e.s . ,
miyazaki , s. , nandra , k. , park , s.q .
, prez - gonzlez , p. g.,rieke , g.h . ,
rigby , j.r .
, willner , s.p .
2006 , , 642 , 126 croton , d.j . ,
farrar , g.r . ,
norberg , p. , colless , m. , peacock , j.a . ,
baldry , i.k . ,
baugh , c.m .
, bland - hawthorn , j. , bridges , t. , cannon , r. , cole , s. , collins , c. , couch , w. , dalton , g. , de propris , r. , driver , s.p . , efstathiou , g. , ellis , r.s .
, frenk , c.s .
, glazebrook , k. , jackson , c. , lahav , o. , lewis , i. , lumsden , s. , maddox , s. , madgwick , d. , peterson , b.a . ,
sutherland , w. , taylor , k. 2005 , mnras , 356 , 1155 giacconi , r. , zirm , a. , wang , j. , rosati , p. , nonino , m. , tozzi , p. , gilli , r. , mainieri , v. , hasinger , g. , kewley , l. , bergeron , j. , borgani , s. , gilmozzi , r. , grogin , n. , koekemoer , a. , schreier , e. , zheng , w. , norman , c. 2002 , , 139 , 369 szokoly , g.p . , bergeron , j. , hasinger , g. , lehmann , i. , kewley , l. , mainieri , v. , nonino , m. , rosati , p. , giacconi , r. , gilli , r. , gilmozzi , r. , norman , c. , romaniello , m. , schreier , e. , tozzi , p. , wang , j.x .
, zheng , w. , zirm , a. 2004 , , 155 , 271 113 & - & j033217.14 - 275402.5 & 53.0714194 & -27.9007025 & 25.83 & 0.03 & 0.03 & 0.95 & - & - & 1.22 + - & mips003133 & j033246.84 - 275121.2 & 53.1951520 & -27.8558940 & 25.25 & 0.06 & 0.02 & 2.76 & -0.6 & 0.2 & - + - & mips003149 & j033214.55 - 275256.6 & 53.0606141 & -27.8823767 & 25.71 & 0.02 & 0.03 & 0.61 & -0.9 & 0.2 & - + 6a & - & j033216.16 - 274941.7 & 53.0673266 & -27.8282610 & 24.63 & 0.03 & 0.02 & 1.55 & - & - & @xmath23 0.37 + - & mips003485 & j033225.24 - 275226.6 & 53.1051776 & -27.8740506 & 25.81 & 0.06 & 0.03 & 1.73 & -1.4 & 0.2 & - + - & mips003528 & j033234.46 - 275005.0 & 53.1436200 & -27.8347720 & 26.48 & 0.05 & 0.05 & 0.93 & -0.9 & 0.2 & - + - & mips003618 & j033235.71 - 274916.0 & 53.1487963 & -27.8211171 & 26.66 & 0.06 & 0.06 & 1.09 & -1.2 & 0.2 & - + 9a & - & j033218.45 - 274555.9 & 53.0768751 & -27.7655258 & 24.94 & 0.01 & 0.02 & 0.48 & - & - & @xmath23 1.27 + - & mips003537 & j033240.75 - 274926.5 & 53.1699410 & -27.8240740 & 26.31 & 0.38 & 0.05 & 7.31 & -0.6 & 0.2 & - + - & mips003871 & j033233.02 - 274200.4 & 53.1375810 & -27.7001130 & 25.81 & 0.06 & 0.03 & 2.02 & -0.5 & 0.2 & - + - & mips010818 & j033227.19 - 274051.4 & 53.1133032 & -27.6809373 & 26.98 & 0.31 & 0.07 & 4.24 & -0.9 & 0.2 & - + 39 & - & j033201.58 - 274327.0 & 53.0065900 & -27.7241626 & 24.43 & 0.02 & 0.01 & 1.31 & - & - & 0.42 + 43 & - & j033203.04 - 274450.1 & 53.0126503 & -27.7472400 & 26.33 & 0.13 & 0.04 & 2.83 & - & - & 0.30 + 44 & - & j033203.65 - 274603.7 & 53.0152257 & -27.7676828 & 23.60 & 0.02 & 0.01 & 1.96 & - & - & 3.56 + 53 & - & j033206.27 - 274536.7 & 53.0261253 & -27.7601830 & 24.59 & 0.03 & 0.02 & 2.10 & - & - & @xmath23 0.67 + 60 & mips004281 & j033207.98 - 274239.5 & 53.0332562 & -27.7109593 & 25.32 & 0.02 & 0.02 & 0.91 & -1.1 & 0.2 & 0.89 + 62 & - & j033208.00 - 274657.3 & 53.0333350 & -27.7825706 & 26.09 & 0.01 & 0.04 & 0.33 & - & - & 0.51 + 63 & - & j033208.27 - 274153.5 & 53.0344441 & -27.6982065 & 25.84 & 0.33 & 0.03 & 9.74 & - & - & @xmath23 0.37 + 65 & - & j033208.53 - 274648.3 & 53.0355495 & -27.7800797 & 23.28 & 0.00 & 0.01 & 0.35 & - & - & @xmath7 1.85 + 66 & mips013611 & j033208.66 - 274734.4 & 53.0360946 & -27.7928810 & 19.82 & 0.10 & 0.00 & 50.11 & -0.8 & 0.2 & 0.34 + 68 & - & j033209.45 - 274806.8 & 53.0393617 & -27.8018870 & 21.04 & 0.08 & 0.00 & 25.78 & - & - & 0.93 + 73 & - & j033210.76 - 274234.6 & 53.0448370 & -27.7096055 & 22.99 & 0.01 & 0.01 & 1.22 & - & - & @xmath23 1.59 + 73 & - & j033210.76 - 274234.6 & 53.0448370 & -27.7096055 & 22.92 & 0.03 & 0.01 & 2.97 & - & - & @xmath23 1.59 + 75 & - & j033210.91 - 274343.1 & 53.0454675 & -27.7286273 & 25.65 & 0.05 & 0.03 & 1.64 & - & - & @xmath23 1.39 + 76 & mips004258 & j033210.91 - 274414.9 & 53.0454704 & -27.7374846 & 23.39 & 0.11 & 0.01 & 12.86 & -0.9 & 0.2 & 0.35 + 80 & - & j033211.41 - 274650.0 & 53.0475311 & -27.7805481 & 24.36 & 0.02 & 0.01 & 1.17 & - & - & 0.98 + 83 & - & j033212.20 - 274530.1 & 53.0508213 & -27.7583569 & 24.18 & 0.02 & 0.01 & 1.24 & - & - & @xmath23 1.51 + 84 & - & j033212.22 - 274620.6 & 53.0509337 & -27.7724017 & 25.51 & 0.01 & 0.03 & 0.43 & - & - & 1.37 + 88 & - & j033213.25 - 274240.9 & 53.0551915 & -27.7113507 & 22.43 & 0.01 & 0.01 & 1.63 & - & - & 1.22 + 91 & - & j033213.85 - 274248.9 & 53.0577260 & -27.7135792 & 25.72 & 0.05 & 0.04 & 1.25 & - & - & @xmath7 2.33 + 94 & - & j033214.00 - 275100.7 & 53.0583525 & -27.8502044 & 25.13 & 0.04 & 0.02 & 1.73 & - & - & 1.84 + 95 & - & j033214.08 - 274230.4 & 53.0586652 & -27.7084382 & 25.46 & 0.23 & 0.03 & 8.40 & - & - & @xmath23 1.04 + 96 & - & j033214.43 - 275110.8 & 53.0601151 & -27.8530055 & 24.82 & 0.04 & 0.02 & 2.25 & - & - & 2.61 + 97 & - & j033214.45 - 274456.6 & 53.0601915 & -27.7490538 & 26.42 & 0.06 & 0.05 & 1.19 & - & - & @xmath23 2.16 + 102 & - & j033214.98 - 274224.9 & 53.0624129 & -27.7069101 & 25.02 & 0.02 & 0.02 & 1.24 & - & - & 0.62 + 106 & - & j033215.80 - 275324.7 & 53.0658257 & -27.8902004 & 25.80 & 0.06 & 0.03 & 1.89 & - & - & @xmath23 0.27 + 107 & - & j033215.79 - 274629.7 & 53.0657934 & -27.7749284 & 24.62 & 0.02 & 0.02 & 0.94 & - & - & - + 113 & - & j033216.74 - 274327.5 & 53.0697345 & -27.7243116 & 24.11 & 0.01 & 0.01 & 0.94 & - & - & @xmath23 1.20 + 114 & - & j033216.85 - 275007.5 & 53.0702116 & -27.8354160 & 26.14 & 0.07 & 0.04 & 1.74 & - & - & @xmath23 1.71 + 115 & - & j033217.06 - 274921.9 & 53.0710682 & -27.8227401 & 22.78 & 0.02 & 0.01 & 2.78 & - & - & 1.46 + 117 & - & j033217.14 - 274303.3 & 53.0714326 & -27.7175864 & 23.12 & 0.14 & 0.01 & 19.5 & - & - & 0.32 + 121 & - & j033218.01 - 274718.5 & 53.0750617 & -27.7884859 & 24.12 & 0.01 & 0.01 & 1.00 & - & - & @xmath23 0.27 + 122 & - & j033218.24 - 275241.4 & 53.0760024 & -27.8781606 & 24.62 & 0.14 & 0.02 & 8.94 & - & - & 0.38 + 123 & - & j033218.34 - 275055.2 & 53.0764119 & -27.8486599 & 25.79 & 0.03 & 0.03 & 1.03 & - & - & 5.28 + 125 & - & j033218.83 - 275135.5 & 53.0784670 & -27.8598545 & 26.60 & 0.05 & 0.06 & 0.90 & - & - & 2.57 + 126 & - & j033219.00 - 274755.5 & 53.0791461 & -27.7987408 & 25.77 & 0.03 & 0.03 & 1.12 & - & - & 1.17 + 129 & - & j033219.81 - 274122.7 & 53.0825338 & -27.6896504 & 22.39 & 0.02 & 0.01 & 1.95 & - & - & @xmath23 0.76 + 131 & - & j033220.05 - 274447.2 & 53.0835448 & -27.7464484 & 26.17 & 0.07 & 0.04 & 1.69 & - & - & 4.35 + 132 & - & j033220.31 - 274554.7 & 53.0846122 & -27.7651813 & 25.19 & 0.02 & 0.04 & 0.69 & - & - & - + 134 & - & j033220.48 - 274732.3 & 53.0853209 & -27.7923093 & 25.20 & 0.03 & 0.02 & 1.16 & - & - & 1.80 + 137 & - & j033221.42 - 274231.2 & 53.0892636 & -27.7086599 & 25.14 & 0.03 & 0.02 & 1.15 & - & - & @xmath7 2.89 + 144 & - & j033222.54 - 274804.3 & 53.0939086 & -27.8011896 & 26.40 & 0.06 & 0.05 & 1.35 & - & - & @xmath23 0.39 + 145 & - & j033222.54 - 274603.8 & 53.0939227 & -27.7677328 & 25.76 & 0.07 & 0.03 & 2.35 & - & - & 1.96 + 146 & - & j033222.55 - 274949.8 & 53.0939514 & -27.8305093 & 26.11 & 0.05 & 0.04 & 1.30 & - & - & 1.18 + 148 & - & j033222.58 - 274425.8 & 53.0941038 & -27.7405110 & 24.87 & 0.02 & 0.02 & 0.75 & - & - & - + 149 & - & j033222.76 - 275224.0 & 53.0948392 & -27.8733220 & 22.10 & 0.01 & 0.01 & 2.02 & - & - & 0.62 + 155 & - & j033224.26 - 274126.4 & 53.1010655 & -27.6906716 & 24.19 & 0.01 & 0.01 & 0.83 & - & - & 0.37 + 158 & - & j033224.86 - 274706.4 & 53.1035834 & -27.7851064 & 23.71 & 0.01 & 0.01 & 0.96 & - & - & @xmath23 0.73 + 160 & - & j033224.96 - 275008.0 & 53.1039901 & -27.8355688 & 26.26 & 0.04 & 0.04 & 1.04 & - & - & 3.23 + 161 & - & j033224.98 - 274101.5 & 53.1040892 & -27.6837528 & 23.67 & 0.02 & 0.01 & 1.64 & - & - & 2.46 + 162 & - & j033225.11 - 275043.3 & 53.1046078 & -27.8453489 & 25.09 & 0.03 & 0.02 & 1.47 & - & - & 2.44 + 163 & mips010787 & j033225.17 - 274218.8 & 53.1048578 & -27.7052197 & 24.07 & 0.02 & 0.01 & 1.49 & -0.7 & 0.2 & 0.28 + 164 & - & j033225.16 - 275450.1 & 53.1048319 & -27.9139259 & 25.29 & 0.02 & 0.02 & 0.73 & - & - & @xmath7 5.53 + 167 & - & j033225.74 - 274936.4 & 53.1072675 & -27.8267670 & 23.94 & 0.01 & 0.01 & 0.50 & - & - & @xmath23 1.37 + 173 & mips013087 & j033226.49 - 274035.5 & 53.1103938 & -27.6765399 & 20.27 & 0.07 & 0.00 & 34.21 & -1.3 & 0.2 & 0.23 + 174 & mips010823 & j033226.67 - 274013.4 & 53.1111149 & -27.6703838 & 25.80 & 0.03 & 0.03 & 0.97 & -1.1 & 0.2 & 0.53 + 177 & mips000309 & j033227.01 - 274105.0 & 53.1125287 & -27.6847238 & 19.75 & 0.11 & 0.00 & 67.25 & -0.6 & 0.2 & 0.31 + 179 & mips003888 & j033227.62 - 274144.9 & 53.1150981 & -27.6958053 & 23.76 & 0.03 & 0.01 & 2.63 & -0.7 & 0.2 & 3.11 + 181 & - & j033228.74 - 274620.4 & 53.1197554 & -27.7723316 & 24.68 & 0.03 & 0.02 & 1.60 & - & - & 0.20 + 182 & - & j033228.81 - 274355.6 & 53.1200615 & -27.7321250 & 21.88 & 0.00 & 0.01 & 0.83 & - & - & 0.29 + 188 & - & j033229.85 - 275105.9 & 53.1243696 & -27.8516331 & 25.62 & 0.05 & 0.03 & 1.66 & - & - & 1.58 + 189 & - & j033229.88 - 274424.4 & 53.1244949 & -27.7401248 & 21.20 & 0.03 & 0.01 & 4.88 & - & - & @xmath23 0.12 + 191 & mips014635 & j033229.98 - 274529.9 & 53.1249148 & -27.7583013 & 21.50 & 0.05 & 0.00 & 14.81 & -0.7 & 0.2 & 0.36 + 192 & - & j033229.99 - 274404.8 & 53.1249588 & -27.7346753 & 20.87 & 0.01 & 0.01 & 2.09 & - & - & 0.29 + 193 & - & j033230.06 - 274523.5 & 53.1252547 & -27.7565350 & 23.63 & 0.09 & 0.01 & 9.63 & - & - & 0.28 + 195 & - & j033230.22 - 274504.6 & 53.1258995 & -27.7512749 & 22.45 & 0.14 & 0.01 & 26.77 & - & - & 0.28 + 196 & - & j033231.36 - 274725.0 & 53.1306522 & -27.7902710 & 25.71 & 0.03 & 0.04 & 0.77 & - & - & @xmath23 1.05 + 197 & mips003938 & j033231.46 - 274623.2 & 53.1310696 & -27.7731060 & 24.77 & 0.02 & 0.02 & 1.13 & -1.2 & 0.2 & @xmath23 1.04 + 199 & - & j033232.04 - 274451.7 & 53.1335033 & -27.7477042 & 25.14 & 0.06 & 0.02 & 2.65 & - & - & - + 203 & - & j033232.99 - 274117.0 & 53.1374396 & -27.6880584 & 26.41 & 0.06 & 0.05 & 1.21 & - & - & @xmath23 3.86 + 207 & - & j033233.46 - 274312.8 & 53.1394133 & -27.7202143 & 20.94 & 0.02 & 0.00 & 4.64 & - & - & 0.90 + 212 & - & j033234.34 - 274350.1 & 53.1430925 & -27.7305813 & 24.64 & 0.03 & 0.02 & 2.16 & - & - & 0.72 + 214 & - & j033234.95 - 275511.2 & 53.1456346 & -27.9197732 & 25.14 & 0.01 & 0.02 & 0.51 & - & - & 0.49 + 215 & - & j033235.04 - 274932.6 & 53.1459896 & -27.8257333 & 26.72 & 0.06 & 0.06 & 0.91 & - & - & @xmath23 0.76 + 216 & - & j033235.10 - 274410.6 & 53.1462691 & -27.7362726 & 25.43 & 0.03 & 0.03 & 1.33 & - & - & @xmath23 0.70 + 227 & - & j033236.72 - 274406.4 & 53.1529803 & -27.7351237 & 25.03 & 0.04 & 0.02 & 1.98 & - & - & 0.45 + 229 & - & j033237.46 - 274000.1 & 53.1560803 & -27.6666919 & 23.63 & 0.09 & 0.01 & 8.82 & - & - & 0.30 + 230 & mips010697 & j033237.76 - 275212.3 & 53.1573435 & -27.8700853 & 25.45 & 0.07 & 0.03 & 2.61 & -1.9 & 0.2 & 0.33 + 234 & mips003915 & j033238.12 - 273944.8 & 53.1588307 & -27.6624444 & 20.79 & 0.08 & 0.00 & 35.21 & -0.8 & 0.2 & 0.23 + 236 & - & j033238.87 - 274733.2 & 53.1619461 & -27.7925487 & 25.87 & 0.10 & 0.04 & 2.67 & - & - & @xmath23 0.79 + 237 & - & j033238.76 - 275121.6 & 53.1615043 & -27.8560057 & 26.88 & 0.11 & 0.07 & 1.66 & - & - & 1.27 + 242 & - & j033239.09 - 274601.8 & 53.1628593 & -27.7671602 & 21.25 & 0.09 & 0.00 & 30.07 & - & - & 0.70 + 244 & - & j033239.46 - 275031.8 & 53.1644211 & -27.8421693 & 24.20 & 0.01 & 0.01 & 1.15 & - & - & 0.35 + 248 & - & j033241.40 - 274717.1 & 53.1725205 & -27.7880947 & 24.37 & 0.01 & 0.01 & 0.80 & - & - & @xmath23 1.66 + 250 & - & j033241.87 - 274359.9 & 53.1744526 & -27.7332996 & 25.71 & 0.23 & 0.03 & 7.30 & - & - & 0.36 + 251 & mips003108 & j033241.85 - 275202.5 & 53.1743886 & -27.8673533 & 23.10 & 0.05 & 0.01 & 6.80 & -0.5 & 0.2 & 0.47 + 254 & - & j033242.86 - 274702.7 & 53.1785932 & -27.7840820 & 25.27 & 0.09 & 0.02 & 3.93 & - & - & 0.52 + 256 & mips009708 & j033243.24 - 274914.2 & 53.1801493 & -27.8206046 & 23.24 & 0.37 & 0.01 & 45.95 & -0.6 & 0.2 & 0.36 + 258 & - & j033244.01 - 274635.0 & 53.1833644 & -27.7763817 & 25.38 & 0.03 & 0.03 & 1.24 & - & - & 0.23 + 260 & - & j033244.27 - 275141.1 & 53.1844471 & -27.8614196 & 22.17 & 0.02 & 0.01 & 3.74 & - & - & @xmath23 0.17 + 261 & - & j033244.31 - 275251.3 & 53.1846401 & -27.8809173 & 24.25 & 0.21 & 0.01 & 15.99 & - & - & 0.29 + 262 & - & j033244.44 - 274819.0 & 53.1851517 & -27.8052754 & 23.81 & 0.02 & 0.01 & 1.56 & - & - & 0.36 + 263 & - & j033244.45 - 274940.2 & 53.1852272 & -27.8278365 & 25.29 & 0.16 & 0.02 & 7.06 & - & - & 1.47 + 264 & - & j033244.60 - 274835.9 & 53.1858335 & -27.8099684 & 25.48 & 0.03 & 0.03 & 1.23 & - & - & 0.55 + 265 & - & j033245.02 - 275439.6 & 53.1875770 & -27.9109950 & 23.15 & 0.02 & 0.01 & 1.25 & - & - & @xmath23 0.85 + 266 & - & j033245.11 - 274724.0 & 53.1879518 & -27.7899986 & 23.28 & 0.01 & 0.01 & 1.45 & - & - & @xmath23 0.66 + 267 & - & j033245.68 - 275534.4 & 53.1903431 & -27.9262231 & 22.79 & 0.01 & 0.01 & 1.34 & - & - & 0.41 + 269 & - & j033246.41 - 275414.0 & 53.1933716 & -27.9038774 & 23.88 & 0.02 & 0.01 & 1.49 & - & - & 1.04 + 273 & - & j033247.18 - 275147.5 & 53.1965772 & -27.8632067 & 25.95 & 0.16 & 0.04 & 4.35 & - & - & @xmath23 1.08 + 276 & - & j033248.18 - 275256.6 & 53.2007350 & -27.8823907 & 23.58 & 0.03 & 0.01 & 3.74 & - & - & 1.88 + 286 & - & j033252.88 - 275119.8 & 53.2203537 & -27.8555099 & 24.13 & 0.19 & 0.01 & 15.25 & - & - & 0.40 + 292 & - & j033256.71 - 275319.1 & 53.2362800 & -27.8886316 & 23.51 & 0.01 & 0.01 & 1.40 & - & - & @xmath23 0.62 + | we investigate the use of optical variability to identify and study active galactic nuclei ( agn ) in the goods - south field . a sample of 22 mid - infrared power law sources and 102 x - ray sources with optical counterparts in the hst acs images were selected .
each object is classified with a variability significance value related to the standard deviation of its magnitude in five epochs separated by 45-day intervals .
the variability significance is compared to the optical , mid - ir , and x - ray properties of the sources .
we find that 26@xmath0 of all agn candidates ( either x - ray- or mid - ir - selected ) are optical variables .
the fraction of optical variables increases to 51@xmath0 when considering sources with soft x - ray band ratios . for the mid - ir agn candidates which have multiwavelength seds , we find optical variability for 64@xmath0 of those classified with seds like broad line agns .
while mostly unobscured agn appear to have the most significant optical variability , some of the more obscured agns are also observed as variables . in particular , we find two mid - ir power law - selected agn candidates without x - ray emission that display optical variability , confirming their agn nature . |
there is a good consensus from many types of observations that dark matter makes up @xmath1 of the mass - energy density of the universe @xcite .
however , the nature of dark matter is unknown .
the most popular class of candidates is the weakly interacting massive particle ( wimp ) , since such particles may be produced thermally in the early universe at the right abundance and with behavior consistent with a large set of cosmological observations @xcite .
the canonical wimp is electrically neutral , and it is stable . once produced in the early universe , it interacts with itself and with ordinary matter only gravitationally .
if the wimp is the dark matter , then the dark halos of galaxies and galaxy clusters are described by a collisionless gas of self - gravitating wimps .
however , there is both theoretical and observational room for other types of dark - matter candidates .
while observations are _ consistent _ with wimps , the observations do not _ require _ the dark matter to be wimps . moreover , some observations
, of the inner mass distribution in galaxies and/or the abundance subhalos in dark - matter halos @xcite have inspired theoretical searches for dark - matter candidates with properties beyond those of the canonical wimp .
there is a large amount of literature on dark - matter particles with additional physical properties beyond those of the canonical stable collisionless wimp .
examples include ( but are by no means limited to ) particles with small electric charges @xcite or dipoles @xcite , self - interacting particles @xcite , or particles with long - range forces @xcite ( see refs .
@xcite for recent reviews ) . in the spirit of these lines of investigation
, we consider in this work a neutral dark - matter particle @xmath2 of mass @xmath3 that decays with lifetime @xmath4 to a slightly less massive neutral particle @xmath5 of mass @xmath6 with @xmath7 and an effectively massless particle @xmath8 which is itself assumed to be noninteracting .
we imagine that such a scenario may arise in some implementations of models of inelastic dark matter @xcite .
when the @xmath2 particle decays , the daughter @xmath5 particle receives a nonrelativistic velocity kick @xmath9 .
now suppose that these particles make up a self - gravitating halo of virial velocity @xmath10 .
if @xmath11 , then the halo may be heated slightly , and its mass distribution thus rearranged slightly , by the velocity kicks imparted to the @xmath5 particles once most of the @xmath2 particles have decayed .
if , on the other hand , @xmath12 , then these halos will be completely disrupted after most of the @xmath2 particles have decayed .
there should thus be no halos with @xmath13 if these particles decay with lifetimes @xmath4 small compared with the age @xmath14 of the universe , and this has been postulated as a possible explanation for the low number of dwarf galaxies relative to that expected from dissipationless dark - matter simulations @xcite .
alternatively , if @xmath12 but @xmath15 , then only a small fraction ( @xmath16 ) of the halo particles will be kicked out of the halo .
the resulting halo mass and mass distribution may then be affected slightly without being completely disrupted .
the canonical - wimp halo is , of course , recovered in the limits @xmath17 and/or @xmath18 . in this paper
, we report on simulations of self - gravitating halos with decaying particles . while the results of these decays can be understood in the limits @xmath19 and @xmath12 @xcite using the adiabatic - expansion model @xcite , detailed evolution of the halo over the full range of the @xmath4-@xmath20 parameter space requires numerical simulation .
we use the simulations and observations of the galaxy - cluster mass function and the halo mass - concentration relation to constrain the decay parameter space .
our central results are presented in fig .
[ fig : summary ] , which shows that ( roughly ) the combination of decay times less than a few times the age of the universe and kick velocities greater than 100 km s@xmath0 are ruled out .
the outline of the paper is as follows . in sec .
[ sec : sims ] , we describe our simulations . in sec .
[ sec : results ] , we characterize the evolution of dark - matter halos as a function of the decay parameters using the simulations , and show how the simulations , in conjunction with observational and theoretical ( in the context of wimps ) determinations of the cluster mass function and the mass - concentration relation , allow us to constrain the decay parameter space . in sec .
[ sec : discussion ] , we discuss various aspects of our findings , including resolution effects ( with an eye towards the requirements for cosmological simulations ) and observational biases . we summarize our work in sec .
[ sec : conclusion ] .
we simulate isolated equilibrium dark - matter halos using the parallel @xmath21-body code gadget-2 in its newtonian , noncosmological mode @xcite .
we have modified gadget-2 in the following way to handle decays .
the maximum time step @xmath22 allowed in the simulations is chosen such that the probability for a decay within that maximum time step is well - described by @xmath23 . at each time step , there is a monte - carlo simulation of decays , with the decay probability being @xmath24 for each particle .
if a particle is designated for decay , it receives a kick speed @xmath25 in a random direction , and is flagged to ensure that the particle can not decay again . because we assume the mass difference is negligible throughout the simulations ( @xmath26 ) , we maintain the mass of the particles ; the change in the kinetic energy of the particle due to the kick is much greater than the change to the potential energy due to the decrease in particle mass .
the massless @xmath8 particle escapes from the halo and is not of interest in this calculation .
our halos are spherically symmetric , and the density @xmath27 as a function of galactocentric radius @xmath28 is taken to be the navarro - frenk - white ( nfw ) form @xcite , @xmath29 where @xmath30 is the scale density , and @xmath31 is the scale radius , such that the halo concentration is @xmath32 where @xmath33 is the virial radius , or size , of the halo .
we define virial quantities with respect to the spherical top - hat overdensity , such that the virial mass @xmath34 where @xmath35 is the mean matter density of the universe , and @xmath36 in a @xmath37cdm universe , where @xmath38 and @xmath39 is the fraction of the critical density @xmath40 in the form of matter ( @xmath41 ) @xcite .
the initial velocity distribution must be taken to be consistent with the mass distribution so that the effects of decay are not confused with relaxation of an initially nonequilibrium halo to equilibrium .
we take the velocity ellipsoid to be isotropic , a reasonable first approximation to dark - matter halos @xcite .
this allows us to use eddington s formula , @xmath42 \end{aligned}\ ] ] to relate the particle distribution function ( df ) @xmath43 as a function of the negative energy @xmath44 ( @xmath45 for all particles bound to the halo ) to known functions , the density and gravitational potential profile @xmath46 .
we can not use eq .
( [ eq : rho ] ) in eq .
( [ eq : eddington ] ) for arbitrary @xmath28 ; the mass of a halo corresponding to a density profile given by eq .
( [ eq : eddington ] ) is formally infinite .
thus , we must truncate the mass profile of the halo , but with care such that the distribution function is everywhere positive - definite .
we use the truncation scheme advocated by kazantzidis et al .
( 2004 ) @xcite , @xmath47 the density profile has been exponentially truncated at the cut off radius @xmath48 and with the parameter @xmath49 controlling the sharpness of the transition .
continuity of the function and its logarithmic slope demand that @xmath50 to set the initial positions and velocities of the particles , we sample the df using an acceptance - rejection method to find the initial radii and speeds @xcite .
the radius and speed are then isotropically mapped to three spatial coordinates .
we begin every simulation with identical initial conditions , varying only the concentration @xmath51 of the nfw profile , particle lifetimes , and kick velocities .
we simulate @xmath52 halos only , but we can extrapolate our results to other halo masses by considering the structural changes to the halos as a function of the decay parameters with respect to virial parameters , as discussed in sec . [
sec : results : theo ] . to determine the virial spherical top - hat overdensity @xmath53 , and hence the virial radius , we use a cosmology with @xmath54 , @xmath55 , and @xmath56 .
we perform a set of 50 simulations , with @xmath57 50 , 100 , 200 , and 500 km s@xmath0 and @xmath58 1 , 10 , 50 , and 100 gyr , and with @xmath59 and @xmath60 . since the virial speed of the halo is @xmath61 using a spherical top - hat overdensity definition of virial parameters , the kick speeds were chosen to bracket the virial value .
the typical crossing time of a particle in the halo is @xmath62 myr , and the virial time @xmath63 is a few gyr , meaning that our choices for @xmath4 bracket the typical dynamical time scales of particles in the halo .
we choose @xmath59 and @xmath64 , since these are typical @xmath37cdm concentrations for galaxies , groups , and clusters .
we use observations of such systems to set constraints on the decay parameter space in sec .
[ sec : results : obs ] .
the fiducial simulations used for our analysis in sec .
[ sec : results ] contain @xmath65 particles total , or @xmath66 particles within the virial radius at the start of the simulations . to test the convergence of the simulations
, we run the same simulations over again but with an order of magnitude fewer particles .
the convergence tests are further discussed in sec .
[ sec : discussion : res ] .
we set gadget-2 parameters , such as the softening , maximum time step , and the force accuracy criterion , according to the recommendations of power et al .
( 2003 ) @xcite . for the fiducial simulations ,
the softening length is set to 1.3 kpc , and it is set to 3.3 kpc in the low - resolution simulations .
for most of the simulations , the maximum allowed time step is set to @xmath67 myr , in accordance with power et al .
. however , for simulations with @xmath68 myr and @xmath69 gyr , @xmath70 myr such that @xmath71 .
we run the simulations for 10 gyr .
we also performed simulations without decay to check that our equilibrium initial conditions were indeed in equilibrium , and to determine the extent of numerical relaxation at the halo centers .
we found that the halos were stable over the 10 gyr duration of the simulations , and that numerical relaxation ceased to be important for @xmath72 kpc for the fiducial simulations and @xmath73 kpc for the low - resolution simulations .
in sec . [ sec : results : theo ] , we present and summarize the results of our simulations .
we show how the halo mass , nfw concentration , and nfw goodness of fit depend on the decay parameters @xmath20 and @xmath4 , initial concentration , and time . in sec .
[ sec : results : obs ] , we show how the results of the simulation can be used in combination with the observed galaxy - cluster mass function and the halo mass - concentration relation to constrain the decay parameter space .
we present the halo density profiles , multiplied by the radius to highlight differences , as a function of @xmath25 , @xmath4 , and time in fig .
[ fig : rrho ] .
we show the density profiles for halos with initial @xmath59 since the qualitative features of the @xmath59 and @xmath64 halos are similar , and omit figures for @xmath75 and 1 gyr with @xmath76 since the halos are effectively destroyed within a few gyr .
the higher lines in each subplot correspond to earlier times , and the lines go ( top to bottom ) @xmath77 , 2.5 , 5 , 7.5 , and 10 gyr after the start of the simulation .
there are other interesting features of note in fig .
[ fig : rrho ] .
first , we see that for small @xmath25 relative to the virial speed , there is little change to the dark - matter profile .
this is expected , as such small @xmath25 leads to neither significant mass loss nor kinetic energy injection in the halo .
however , for @xmath78 ( @xmath79 ) , changes to halo structure begin to become significant for @xmath80 gyr . in general , changes to halo mass and halo structure are most extreme for large @xmath25 and small @xmath4 .
changes are abrupt when @xmath4 is small , although it takes @xmath81 gyr for the halos to reach equilibrium if @xmath82 gyr .
the halo dark - matter density profile times the radius , as a function of radius and time .
the lines correspond to different time slices in the simulation .
the solid line represents the density at the beginning of the simulations ; the dashed line is the density 2.5 gyr after the simulation start ; the long - dashed line is that after 5 gyr ; the dot - dashed line is the density profile after 7.5 gyr ; and the long - dash - dotted line represents the density profile after 10 gyr .
density profiles for @xmath83 and @xmath82 gyr are not shown because the halos are destroyed within a few gyr.,scaledwidth=45.0% ] we see similar trends in the velocity data , which we present in figs . [
fig : sigma ] and [ fig : beta ] . in fig .
[ fig : sigma ] , we show the velocity dispersion profiles @xmath84^{1/2}\end{aligned}\ ] ] as a function of decay parameters and time , and in fig .
[ fig : beta ] , we show the velocity anisotropy @xmath85 where @xmath86 , @xmath87 , and @xmath88 are the velocity dispersions in spherical coordinates .
we set up the initial conditions such that @xmath89 ( such that @xmath90 ) and given the equilibrium initial conditions , the velocity dispersion is initially equivalent to the rms velocities . even at late times in the cases of small @xmath4 or large @xmath25 ,
the rms velocities are only slightly different from the velocity dispersions .
as expected , the velocity dispersion profiles @xmath91 decrease in normalization as more halo particles decay , since the negative heat capacity of a self - gravitating system ensures that an injection of kinetic energy due to decays results in an overall decrease in the equilibrium kinetic energy of the halo ( cf .
the most extreme drops in the velocity dispersion occur for small @xmath4 or large @xmath25 for which the kinetic energy injected into the halos due to decays is large .
the velocity dispersion profile quickly settles to a new equilibrium for @xmath92 gyr .
when both @xmath25 is large and @xmath4 is relatively small , the velocity dispersion increases as a function of time at large @xmath28 .
these regions are no longer within the virial radius and contain a number of unbound particles , the latter of which in particular drives up the velocity dispersion . the three - dimensional velocity dispersion as a function of radius and time , tiled according to @xmath25 and @xmath4 .
the lines correspond to the same time slices as in fig .
[ fig : rrho].,scaledwidth=45.0% ] the velocity anisotropy @xmath93 hovers around zero for most times and radii of most of the decay parameters we consider .
this is unsurprising , since the initial conditions and the decays have a high degree of spherical symmetry in both configuration and velocity space .
the exceptions in @xmath94 occur , unsurprisingly , for small @xmath4 and large @xmath25 . in the case of @xmath82 gyr for @xmath95 or @xmath96 , @xmath97 in the outskirts of the halos at late times , indicating that orbits in the outskirts of those halos become tangentially biased .
radial orbits in the outskirts of galaxies are easiest to strip since their binding energy is lower at fixed radius than more circular orbits . for large @xmath25and @xmath4 ,
though , the orbits clearly become radially biased in the outskirts of the halo .
these halos have not settled into a true dynamical equilibrium , and there is a significant net outflow of particles .
this net outflow drives the radial bias in the orbits , and is more significant in the outskirts because any particle which decays onto an unbound orbit in the interior of the halo must pass through the exterior regions .
moreover , the particles originating in the outskirts are more vulnerable to ejection , so the daughter particles originating in the outskirts will also have a significant radial bias .
the interior regions of the halo are less vulnerable to decay ( unless @xmath12 ) and the dynamical time is short relative to the decay time , both of which serve to preserve the initial orbital structure deep within the halos .
however , since the dark - matter mass profile may be inferred observationally but the dark - matter velocities are not , we focus our analysis on the mass distribution in halos as a function of decay parameters and time .
perhaps the most striking feature of fig .
[ fig : rrho ] is that the nfw halo profile , eq .
[ eq : rho ] , is clearly not a good fit to many of the profiles . in the way in which we have chosen to display the density profile information , @xmath98 as a function of @xmath28 ,
the profile should be flat for @xmath99 and be monotonically decreasing for larger @xmath28 .
however , there is a clear turnover in @xmath98 in some cases , most notably in the @xmath82 gyr cases , but also at late times for @xmath100 or 50 gyr and @xmath101 .
the velocity anisotropy @xmath102 as a function of radius and time , tiled according to @xmath25 and @xmath4 .
the lines correspond to the same time slices as in fig .
[ fig : rrho].,scaledwidth=45.0% ] in order to find the goodness of fit of the nfw profile quantitatively , as well as the best - fit nfw parameters , we bin the particles radially with respect to the center of mass of the system , and use two different likelihood estimation routines : a simple @xmath103 model , which should work in the limit of a large number of particles in each bin ; and a poisson - based likelihood .
we treat the uncertainty in the number of particles in each bin as @xmath104 , where @xmath21 is the observed number of particles in a bin . instead of using @xmath105 and @xmath106 as free parameters in the nfw fit in eq .
( [ eq : rho ] ) , we use @xmath107 and @xmath51 .
we use 100 bins between @xmath108 ( @xmath109 kpc , the minimum radius at which numerical relaxation ceases to matter ) and @xmath110 ( the approximate virial radius at @xmath77 ) .
we bin the particles either linearly or logarithmically in @xmath28 , and obtain identical fits , as expected .
the minimum @xmath103 parameters are slightly biased relative to those obtained with the poisson likelihood function , but not at a significant level .
thus , we can use statistics associated with @xmath103 to describe the fits , and define @xmath103 per degree of freedom ( @xmath103/d.o.f . ) as our metric for the fit quality .
we summarize the best - fit nfw concentrations and the goodness of fit ( parametrized by @xmath111d.o.f . ) in fig .
[ fig : c ] .
the concentrations are extrapolated to a hubble time @xmath112 gyr from our simulations , and their values correspond to the shading for each grid point representing a different ( @xmath20,@xmath4 ) pair used for a simulation .
the size of the grid points corresponds to the quality of the fit .
if the @xmath103/d.o.f . exceeds 1.5 at any time during the simulation , the grid point is small ; if @xmath103/d.o.f.@xmath113 for the whole simulation ( sampled every 0.5 gyr ) , then the point is large .
in some cases , the fit is somewhat marginal @xmath114 much of the time , in the dividing regions of @xmath115 parameter space between good and really bad nfw fits .
the general features between the simulations with initial @xmath116 and @xmath117 are similar , though ; for @xmath118 gyr and @xmath119 ( @xmath120 ) , the nfw profile is generally not a good fit to the halo profile at some ( especially late ) times . for lower @xmath25 or high @xmath4
, the nfw density profile provides a reasonable fit throughout the lifetime of a halo , even if the halo mass and concentration change .
the concentrations drop rapidly as @xmath20 increases , and the final concentration is less sensitive , in general , into @xmath4 than to @xmath20 .
the concentration always decreases as a result of decay , which is expected if kinetic energy is injected to the halo or in the case of adiabatic expansion . at large @xmath4 ( specifically , @xmath121 and 100 gyr ) and @xmath122 , the changes to the halo structure with time are well - described by the adiabatic - expansion scheme explored in ref
. @xcite .
if @xmath4 is much greater than the halo dynamical time ( @xmath123 gyr for particles in an @xmath52 halo ) and the kick speed is high enough that daughter particles are cleanly ejected from the halo , then changes to the gravitational potential are adiabatic and easy to model .
if all particles are on circular orbits , the density profile remains constant , but with the scale radius rescaled as @xmath124 and the nfw scale density rescaled as @xmath125 , where @xmath126 is the fraction of the initial dark - matter particles that have decayed .
we find that nfw profiles are a good fit for @xmath127 ( @xmath128 ) and @xmath129 gyr with the adiabatic - expansion parameters , but this does not work so well for @xmath130 gyr or for @xmath131 ( @xmath132 ) .
thus , it appears that @xmath133 and @xmath134 are required for the analytic adiabatic - expansion description to describe halo evolution .
we reported the goodness of fit and concentration in terms of @xmath135 instead of @xmath20 because we would like to extrapolate the findings of our simulations to other halo masses .
it is natural to parametrize the kick speeds @xmath20 in terms of the virial speed ; the virial speed is a measure of the binding energy of the halo , and @xmath20 is a measure of the kinetic energy injected into the halo .
following ref .
@xcite , we link the decay time @xmath4 with the crossing time @xmath136 at the half - mass radius , where @xmath137^{1/2}$ ] , such that the typical dynamical time of a particle in a halo is @xmath138 thus , we can extend our simulation results to halos of other masses .
we summarize the mass remaining after a hubble time @xmath14 in fig .
[ fig : mass ] . in this plot
, we show regions in the decay parameter space in terms of @xmath135 and @xmath139 on the left , and @xmath135 and @xmath4 on the right , in order to highlight both the absolute and relative time scales .
the hubble time is a fixed time scale , and one can consider @xmath4 in reference to that timescale , or think of @xmath14 in terms of dynamical times , and then consider the relative time scale @xmath139 .
the long - dashed line shows @xmath140 for the @xmath116 simulations , which has been found via interpolation of our grid of @xmath141 as a function of the decay parameters .
this line indicates the transition from decays having little effect on the virial mass of the halo to having a nontrivial effect .
the short - dashed line shows @xmath142 , which is around the transition between decays causing total destruction of the halo and moderate mass loss in the halo .
the dotted and dot - dashed lines show @xmath140 and @xmath142 for the halos which originally had @xmath117 .
when the decays are parametrized in terms of @xmath139 and @xmath135 , the @xmath140 and the @xmath142 lines are largely similar between the two different initial concentrations .
the largest discrepancy occurs near @xmath143 and @xmath144 gyr .
this effect is real , and is due to the fact that more concentrated halos have more particles that are more tightly bound to the halos , which makes the particles harder to eject .
however , for @xmath145 or for the @xmath146 line , the differences in mass retention between the two sets of simulations are small .
the differences are greatest at small @xmath139 and @xmath135 because the main particles affected and likely to be kicked out of the halo due to such decays are in the outer halo . if @xmath147 , particles even relatively tightly bound to the halo have a high chance of being ejected , which ameliorates some differences due to differences in concentration . in general , though , even though the dynamical times of the @xmath116 and @xmath117 halos are different by a factor of 1.7 , and even though there are different numbers of dynamical times in @xmath14 for the two sets of halos , the remaining halo mass at @xmath14 is not a strong function of concentration in @xmath148 parameter space .
the differences are more pronounced in the right - hand side plot in fig .
[ fig : mass ] , in which the remaining mass is plotted as a function of @xmath4 and @xmath135 .
the @xmath117 halos tend to retain more mass at fixed @xmath10 and @xmath4 because the typical particle has a higher binding energy than in the @xmath59 case .
since the dynamical time depends on @xmath51 , which is related to the distribution of particle energies within the halo , some of the dependence of halo mass loss on concentration is divided out by parameterizing mass loss in terms of @xmath149 instead of @xmath4 .
a more direct comparison would be to examine mass loss and concentration changes as a function of @xmath135 and @xmath139 for a fixed number of dynamical times .
when we do this , we find that the @xmath150 lines lie practically on top of each other for @xmath116 and @xmath117 , but the @xmath151 line is offset by approximately a factor of 2 in @xmath139 between @xmath116 and @xmath117 .
however , what matters in comparisons to observations is the effect of decay at a fixed time . instead of mapping results in terms of @xmath135 and @xmath139
, we interpolate in terms of @xmath135 , @xmath4 , and @xmath51 .
our simulations span the typical @xmath37cdm concentrations expected for the sizes of halos relevant to the observations , and the virial speeds span a factor of @xmath152 , a relatively small range . in summary , we find that for halos that initially have @xmath116 or 10 and have the equilibrium nfw halo profiles expected in @xmath37cdm simulations , the effects of decay in the regime @xmath153 and @xmath154 are to drastically lower the halo concentration in a hubble time , drive substantial mass loss , and drive changes to the halo profile such that the nfw profile is not a good fit . for @xmath155 gyr or @xmath156 ,
the nfw profile remains a good fit to the density profile , mass loss is not too significant , and the concentration does not drop too much . mass loss and concentration reduction
are more severe for fixed @xmath139 for large @xmath20 until @xmath128 , at which point the decay can be described analytically in the adiabatic - expansion approximation , and the effects of decay no longer depend on @xmath20 as long as @xmath157 . at such high @xmath20 , any daughter particle in a decay
is ejected from the halo , so there is no kinetic energy injection into the halo as a result of decays . in the intermediate regime
[ roughly corresponding to the space between the @xmath146 and @xmath158 lines ] , mass loss is moderate although the concentration may drop quite a lot , and the inner part of the density profile is shallower than for nfw .
there are some differences between the @xmath116 and @xmath117 simulations beyond that which can be factored out by parameterizing decay in terms of the virial time scales @xmath135 and @xmath139 .
this is due to the fact that the gravitational potential , and hence binding energy , does depend on the concentration in a way that is not completely factored out by the virial scaling .
even though we parametrize halos in terms of `` typical '' dynamical times and typical speeds , there is a diversity of particle speeds and time scales within a halo .
nonetheless , since our simulations span the range of @xmath159 and @xmath51 of the halo observations below , we are able to interpolate our simulations and map @xmath37cdm dark - matter halos to halos in a decaying - dark - matter cosmology . under the hierarchical picture of structure formation in the universe
, it is expected that low - mass dark - matter halos form earlier than high - mass halos . as a consequence of this early formation time ,
the inner parts of low - mass halos are far more dense relative to the outskirts of the halo relative to high - mass halos @xcite . in other words , low - mass halos are more concentrated than high - mass halos .
this trend has been observed and quantified in simulations of @xmath37cdm cosmologies @xcite .
it is found that cosmologies with higher matter fractions , @xmath39 , and a higher amplitude of matter fluctuations averaged in @xmath160 mpc spheres , @xmath161 , produce higher concentrations for fixed halo mass than cosmologies in which @xmath39 and @xmath161 are low .
this is due to the fact that structures collapse earlier when the universe is more dense if @xmath39 and @xmath161 are higher .
decays will make halos appear less concentrated ; hence , @xmath39 and @xmath161 inferred from halos that have experienced decay _ if _
one begins with a wimp dark - matter ansatz would be lower than that inferred from observations of earlier epochs .
we can set conservative constraints on the decay parameter space by comparing observationally inferred mass - concentration relations with the mass - concentration relation derived from decays in a high-@xmath39 , high-@xmath161 cosmology .
the cmb last - scattering surface is the earliest window into cosmological parameters .
the six - parameter cosmological fit to the _ wilkinson microwave anisotropy probe _ ( wmap ) seven - year data set suggests @xmath162 ( 1-@xmath163 ) and @xmath164 ( 1-@xmath163 ) @xcite .
the 2-@xmath163 upper limits of @xmath39 and @xmath161 correspond approximately to the mean wmap first - year mean parameters @xcite . the mean mass - concentration relation found in simulations of a @xmath37cdm cosmology using the wmap-1 parameters is illustrated in fig .
[ fig : mc ] , and is denoted by the upper solid line @xcite .
we use our grid of mass loss and concentration as a function of @xmath4 and @xmath165 , found using the simulations and described in sec .
[ sec : results : theo ] , to map @xmath37cdm halo masses and concentrations in a wmap-1 cosmology to those of halos at @xmath166 .
we show the mass - concentration relation as a function of decay parameters in fig .
[ fig : mc ] , along with the mean theoretical and observed relations .
it should be noted that we use the best - fit nfw concentrations [ eq .
( [ eq : c ] ) ] in fig .
[ fig : mc ] , even if the nfw profile is not a good fit to the density profile .
nonetheless , the nfw concentration provides a reasonable measure of the density contrast within the halo . in fig .
[ fig : mc ] , the solid lines represent the mean mass - concentration relations found in dissipationless , baryon - free cosmological simulations using wmap one - year ( upper line ) @xcite and wmap three - year ( bottom line ) @xcite cosmologies @xcite .
the long - dashed lines represent decaying - dark - matter models with ( top to bottom ) @xmath68 , 40 , 20 , and 5 gyr under the assumption that the wmap one - year cosmology is the underlying cosmology .
the short - dashed and dotted lines are observationally inferred mean mass - concentration relations using a sample of clusters observed in a variety of ways @xcite and x - ray observations @xcite , respectively .
the error bar in the upper left corner of the plots shows the scatter in the mass - concentration relation for simulations , and the typical uncertainty in the mean mass - concentration relation from the observations .
the shaded region corresponds to the 1-@xmath163 region of the mean mass - concentration relation inferred from weak - lensing observations of sloan digital sky survey ( sdss ) galaxies @xcite .
we use only the weak - lensing mass - concentration relation to constrain the decay parameters .
this is because the cluster and x - ray samples have serious selection effects which have not been fully quantified .
reference @xcite finds that the concentration of strong - lens clusters is even higher than that predicted by ref .
@xcite , who identified strong - lens systems in @xmath21-body simulations .
dark - matter profiles can only be inferred from x - ray observations if the halos are relaxed ; relaxed halos have , on average , higher concentration than the population of halos as a whole @xcite .
thus , until the selection effects are better quantified for the x - ray and strong - lens systems , one should not use the mass concentration inferred from observations of those systems to constrain dark - matter models . even though there are also systematics that have not been fully quantified in the weak - lensing measurement ( see sec .
iiid of ref .
@xcite for a sampling of possible effects ) , the measurement provides conservative constraints because it lies slightly below simulation values , on which we base our decay predictions .
we consider a ( @xmath167 ) pair of parameters to be excluded if its mass - concentration relation lies below the 1-@xmath163 lower limit of the mean weak - lensing mass - concentration relation . using this criterion , we see from fig . [ fig : mc ] that current weak - lensing measurements are unable to constrain @xmath4 for @xmath168 . however , @xmath169 gyr is ruled out for @xmath170 , and @xmath171 for larger values of @xmath25 .
interestingly , the constraints from the simulations for large @xmath25 are nearly identical to those found analytically using a simple model of adiabatic expansion ( valid only if @xmath172 and @xmath173 ) in ref . @xcite .
we have neglected the effects of baryons on the halo concentration .
this is largely because it is not clear how baryons affect the dark - matter halo . in cosmological simulations with two different prescriptions for baryon physics , rudd et al .
( 2008 ) @xcite found either little difference relative to dark - matter - only simulations , or an increase in concentration of @xmath174 above the dark - matter - only values .
however , even in the latter case , the minimum allowed @xmath4 for fixed @xmath25 would only decrease @xmath175 .
better constraints may be obtained in the future .
upcoming deep wide - field surveys will increase the statistics of weak and strong lensing and x - ray cluster profiles , and probe down to smaller halo masses @xcite .
however , improvements in using the ( redshift - dependent ) mass - concentration relation to constrain dark - matter decays ( or cosmological parameters ) will require a better understanding of the systematics and selection effects for the various methods of measuring the dark - matter halo masses of galaxies and clusters .
this is likely to be a subject of significant future work , since the selection effects are relevant to determining halo mass functions .
the evolution of halo mass functions depends on the growth function .
since the growth function is sensitive to the equation of state of dark energy , there will be significant efforts to understand any systematic errors that may obscure inferences about dark energy from observations @xcite .
in addition , there needs to be a better theoretical understanding of which processes of galaxy evolution drive changes to dark - matter halo profiles .
this subject has generated recent interest @xcite , and with continuing improvements in the spatial resolution and star - formation and feedback recipes in simulations , it should be possible to determine which physical processes influence the distribution of dark matter within galaxies .
the mass function of galaxy clusters @xmath176 is sensitive to @xmath39 and @xmath161 for similar reasons as for the mass - concentration relation ; if the amplitude of matter fluctuations is higher , structure forms earlier and is more abundant in the local universe @xcite
. since galaxy clusters are the largest and rarest of virialized structures , corresponding to small fluctuations in the linear matter density field smoothed on @xmath177 mpc scales , small changes in @xmath161 and @xmath39 will result in large changes to the cluster mass function .
if dark matter were to decay into relativistic particles , the cluster mass function would look different than the @xmath37cdm mass function because the continuous pumping of relativistic particles into the universe would alter the background evolution of the universe and hence the density threshold for halo collapse .
this effect was investigated by oguri et al .
( 2003 ) @xcite .
however , the main effect of decays in which the mass splitting between dark - matter species is small is to reduce the mass of halos due to ejection resulting from decays .
since @xmath161 and @xmath39 should be the same regardless of the redshift of the matter tracer in a @xmath37cdm universe , the smoking gun for decays would be if @xmath39 and @xmath161 inferred from an early epoch ( the cosmic microwave background or lyman-@xmath178 forest ) were systematically higher than that inferred from the @xmath179 cluster mass function .
we use parameter estimates from the wmap seven - year data set and low - redshift observations of clusters to constrain decay parameter space . as with the analysis using the mass - concentration relation ,
we find conservative bounds on the decay parameter space by considering the parameters for which the cluster mass function with @xmath39 and @xmath161 set to 2-@xmath163 above the wmap-7 parameters maps to a cluster mass function that is 1-@xmath163 below @xmath39 and @xmath161 inferred from observations of low - redshift clusters .
both optical and x - ray observations of clusters indicate that the 1-@xmath163 lower limits on @xmath39 and @xmath161 are approximately the same as the 2-@xmath163 lower limits on those parameters from wmap-7 , @xmath180 and @xmath181 @xcite .
we use the relation between the @xmath37cdm halo masses and the halo masses after @xmath182 with decay to find the cluster mass function , @xmath183 here , @xmath184 is the comoving number density of halos with masses larger than @xmath185 after a hubble time of decay , @xmath186 is the tinker et al .
( 2008 ) @xcite @xmath37cdm differential mass function , @xmath187 is the initial @xmath37cdm mass of the halos , and @xmath185 is the final halo mass after a hubble time of decay . since the relation between @xmath187 and @xmath185 masses depends on concentration ( sec . [
sec : results : theo ] ) , we use the wmap-1 mass - concentration relation from ref .
@xcite to set the initial concentrations of the initial @xmath37cdm halos and interpolate among our simulations to find @xmath185 from @xmath187 .
we show the galaxy - cluster mass function as a function of decay parameters in fig .
[ fig : dn ] .
the dotted line is the mass function for the high-@xmath39 , high-@xmath161 cosmology , the solid line is for the mean wmap-7 cosmology , and the long - dashed line represents the mass function for a cosmology set to the 2-@xmath163 lower limits of @xmath39 and @xmath161 parameters from wmap-7 from the six - parameter cosmological fit @xcite .
all other cosmological parameters have been set to the wmap-7 mean values , as those parameters do not affect the normalization or shape of the mass function nearly as much as @xmath39 and @xmath161 .
the short - dashed lines show the mass functions assuming ( top to bottom ) @xmath68 , 40 , 20 , and 5 gyr .
each panel shows a different @xmath25 .
it is clear that the cluster mass function ( for which we consider @xmath188 to be classified as a cluster ) is relatively insensitive to small kick speeds , @xmath189 .
this is because the typical virial speed of a cluster is @xmath190 . from fig .
[ fig : mass ] , it is apparent that severe mass loss in dark - matter halos requires @xmath134 and @xmath191 gyr .
however , the shape of the mass function is somewhat different than those predicted in @xmath37cdm cosmologies , since higher mass halos are less affected by lower @xmath20 than low - mass halos . for @xmath192 ,
the number density of @xmath193 is approximately the same for @xmath4 as the lowest number density allowed by low - redshift observations of clusters .
however , even for higher @xmath4 , the cluster mass function is far less steep than predicted by @xmath37cdm . at @xmath194 ,
the cluster number density is too low for @xmath195 gyr . at even higher @xmath20
, the cluster mass functions converge to the analytic estimates of ref .
@xcite in the case of @xmath173 , and constrain @xmath196 gyr .
the shape of the cluster mass function as well as the number density @xmath197 can be used to constrain decays , but the statistics are generally lower for higher cluster masses .
constraints on dark - matter decays from the cluster mass function are more robust to baryon physics than the mass - concentration relation . using @xmath21-body hydrodynamic simulations with two different models for baryon physics , rudd et al .
( 2008 ) @xcite found that the normalization of the halo mass function deviates by @xmath198 with the inclusion of baryons relative to the halo mass function found in @xmath21-body cold - dark - matter - only simulations .
the redshift evolution of the cluster mass function is also a useful probe of decays @xcite .
the signature of decays will be that , for a fixed @xmath179 mass function , the high - redshift mass functions will have higher normalization and a different shape than in a @xmath37cdm universe . if one were to estimate @xmath39 and @xmath161 for different redshift slices with the ansatz of a @xmath37cdm universe , one would estimate systematically higher @xmath161 and @xmath39 at higher redshifts . currently , @xmath161 and @xmath39 inferred from small samples of x - ray clusters up to @xmath199 are consistent at different redshift slices , although the error bars are still fairly large @xcite . upcoming deep optical , x - ray , and sunyaev - zeldovich surveys @xcite should allow for better constraints on decays , since they will find large samples of clusters at a variety of redshifts .
summary of the mass - concentration and @xmath179 cluster mass function limits on the decay parameter space.,scaledwidth=45.0% ] we summarize the constraints on the decaying - dark - matter parameter space in fig .
[ fig : summary ] .
the exclusion region , marked `` ruled out '' on the plot , comes from three types of observations . for relativistic @xmath25
, the wmap-1 data set the strongest constraint on @xmath4 , with @xmath200 gyr at 68% c.l .
this restriction is essentially a restriction on @xmath201 , the radiation density of the universe as a function of time . for @xmath202 ,
both the galaxy - cluster mass function and the mass - concentration relation require @xmath203 gyr . at lower @xmath25 , the mass - concentration relation sets best constraints on @xmath4 , at approximately the level @xmath204 gyr for @xmath205 and @xmath206 gyr for @xmath207 .
we note that this exclusion region looks similar to that of ref .
however , the exclusion region in this paper is more robust ; the exclusion region for @xmath208 in ref .
@xcite was based on noting that the clustering of galaxies in the sdss appeared to be consistent with the clustering of @xmath37cdm halos with @xmath209 @xcite .
the constraint was set based on not letting @xmath37cdm halos with @xmath210 lose more than half their mass due to decays .
this is a fairly hand - waving constraint .
it will likely take cosmological simulations of decaying dark matter in order to map galaxy clustering to constraints on dark - matter decays .
the regions labeled `` cdm - like '' ( cold - dark - matter - like " , such as wimps ) and `` allowed '' are still plausible regions of parameter space .
however , it will be challenging to probe the region marked `` cdm - like '' .
for @xmath211 gyr , there is hardly any change to halo properties ; mass loss relative to @xmath37cdm halos is on the @xmath212 level , and concentrations go down by at most @xmath213 .
the nfw profile is still a reasonable fit to the halo density profiles .
it is not clear if the systematics in even the largest and deepest all - sky surveys will be small enough to probe such subtle deviations from @xmath37cdm halo properties .
decay models with @xmath214 will also be challenging to probe , as such speeds are far less than the typical virial speeds of galaxies or larger systems .
the smallest halos that have been found observationally have @xmath215 , and are subhalos at that @xcite .
such halos have @xmath216 . as we saw in sec .
[ sec : results : theo ] , the structural properties of dark - matter halos hardly change if @xmath217 . moreover , there are poor statistics of such small halos , and it is not clear how many halos of that size will host a luminous component .
the statistics of such small objects may be probed using gravitational milli - lensing , but we are years away from building any sort of statistical sample of such halos @xcite . there is some hope of further constraining the region marked `` allowed '' in the near future .
sunyaev - zeldovich , x - ray , and deep optical surveys should uncover @xmath218 galaxy clusters up to quite high redshift @xcite ; then , both the @xmath179 cluster mass function and the evolution of the cluster mass function could be used to more accurately constrain the decay parameter space .
the deep optical surveys should also allow for a more precise measurement of the mean mass - concentration relation , and down to lower halo masses .
on the theoretical side , a better understanding of selection effects will allow for a better comparison of the decaying - dark - matter predictions with observations .
in addition , if cosmological simulations of decaying dark matter are performed , one can hope to characterize the halo occupation distribution ( hod ) @xcite , which will open up galaxy clustering as an avenue to better explore the decay parameter space .
so far , we have not explored using dwarf galaxies to constrain the allowed region of parameter space , as suggested by ref .
this is because there are no robust measurements of the abundance of @xmath219 halos , nor are there strong constraints on the inner profiles of dwarf galaxies @xcite .. ]
one could imagine that the @xmath215 dwarf galaxies could be the result of decays of halos born with much higher virial masses , although the density profile would be significantly different than expected in @xmath37cdm . in the future
, large optical surveys such as lsst will look farther down the luminosity function , both in the local group and in the local universe @xcite , which will allow for better constraints using the abundance of low - mass halos .
astrometric observations of stars in local group dwarf galaxies will be critical getting the three - dimensional stellar kinematic velocity data for better constraints on the dark - matter density profiles in those galaxies @xcite
. this will be possible with next - generation astrometric satellites , such as _ gaia _ and sim @xcite .
in the previous section , we presented the results of our simulations and showed how those results , along with the mass - concentration relation and cluster mass functions inferred from existing observations , could be used to constrain the decaying - dark - matter parameter space . in this section , we address two issues that will affect future attempts to further constrain the decay parameter space : numerical convergence in simulations , and more accurate mapping of simulations to observations .
we based all our analysis in sec .
[ sec : results ] on simulations that initially had @xmath220 particles within the virial radius .
we would like to know which of the results we described above are fully converged , and which results may be spurious due to resolution effects . moreover , we would like to understand resolution effects before embarking on cosmological simulations , which are far more costly than the simulations described in sec . [
sec : sims ] . if @xmath221 particles are required within a virial radius in order to determine halo properties , it becomes prohibitive to get good statistics on dark - matter halos ( and subhalos ) on a variety of mass scales in cosmological simulations . in order to understand the robustness of our results , and to determine if one can get away with fewer particles in the virial radius for cosmological simulations , we perform simulations with the same @xmath51 , @xmath25 , and @xmath4 values as those used in sec . [
sec : results ] , but with @xmath222 particles within the virial radius .
the low - resolution simulations are briefly described in sec .
[ sec : sims ] .
our goal is to determine how the virial mass , concentration , and goodness of fit to the nfw density profile compare between the two sets of simulations .
density profile for the @xmath64 , @xmath223 , @xmath130 gyr simulations at 9 gyr divided by the best - fit nfw density profile .
solid line : @xmath66 particles in the virial radius .
dashed line : @xmath222 particles within the virial radius.,scaledwidth=45.0% ] we find that the virial mass derived from fits to the nfw profile are consistent between the two sets of simulations , and are consistent with the virial mass as determined by finding the average density of dark - matter particles within a sphere centered on the halo center of mass .
similarly , the nfw concentration parameter from the fits is also consistent between the two sets of simulations . therefore , if one is concerned with accurately determining halo masses or nfw concentration parameters in a simulation with a decaying - dark - matter cosmology , one requires no more than @xmath222 particles within the virial radius .
by contrast , the quality of the nfw fit can be sensitive to the number of particles within the virial radius .
for example , @xmath224 for fits to all particles within the virial radius for @xmath223 and @xmath100 gyr for @xmath222 particles within the virial radius , regardless of the initial concentration . however , the @xmath103 is bad for @xmath225 for both @xmath59 and @xmath64 if there are initially @xmath66 particles in the virial radius , even though the nfw fit parameters of the two different - resolution simulations are identical ( see fig . [
fig : fit ] for @xmath103 in the @xmath64 case ) .
this is illustrated in fig .
[ fig : res ] , in which we show the density profile from the simulation ( @xmath226 ) divided by the best - fit nfw profile for the two simulations at @xmath227 gyr with @xmath64 , @xmath228 , and @xmath130 gyr ( @xmath229 ) .
the innermost points ( @xmath230 kpc ) show evidence of numerical relaxation . for the rest of the halo
, one can tell that there is significant curvature in @xmath231 for the @xmath66-particle simulation , but not for the @xmath222-particle simulations .
this difference is likely in large part due to the increased poisson noise in the @xmath222-particle simulation . in general
, the discrepancy in the quality of the fit appears to be exacerbated in cases in which @xmath118 gyr and @xmath232 .
the consequences for our results are as follows .
our constraints on the decay parameter space from cluster mass functions depended on the mapping between the @xmath37cdm halo mass and the halo mass after a time @xmath14 .
the mass estimated in the high - resolution simulations used in sec .
[ sec : results ] is virtually identical to that found in the lower - resolution simulations , so we believe the halo mass ( as a function of decay parameters and time ) to be converged and reliable .
similarly , the use of the mass - concentration relation depends on the mapping of @xmath37cdm halo masses and nfw concentrations to those of halos that have experienced decays .
again , we believe the virial masses and nfw concentrations to be converged in our simulations .
the shape of the dark - matter density profile is not used for mapping either of the observations to constraints on decay parameter space .
thus , we believe the mapping of our simulations to observational constraints to be robust .
the implications for future cosmological @xmath21-body simulations are as follows .
we believe it is possible to get accurate estimates of halo masses and nfw concentrations using no more than @xmath218 particles in the virial radius .
it may be possible to get good statistics using fewer particles , but it will likely take convergence tests in fully cosmological @xmath21-body simulations to determine the minimum particle number required for convergence , especially for subhalos .
this means that it may be possible to explore decaying - dark - matter cosmologies with reasonable statistics using modest computational resources .
the only situation in which higher particle numbers may be required is a detailed comparison of the shape of decaying - dark - matter halo profiles with respect to @xmath37cdm halos . in this case
, we showed that @xmath220 particles are required within the virial radius to determine the goodness of the nfw fit , at least down to @xmath233 for @xmath135 and @xmath191 gyr .
in other regions of the decay parameter space , the @xmath222- and @xmath66-particle simulations converge on the goodness of the nfw fit .
if one is mainly concerned with radical differences between @xmath37cdm and decaying - dark - matter halo density profiles , then @xmath218 particles within the virial ( or tidal truncation ) radius is likely sufficient ; more subtle differences require higher particle numbers . in sec .
[ sec : results ] , we used nfw fits to the entire halos to relate halo properties to observables . however , astronomical probes of the gravitational potential of galaxies and their halos are rarely sensitive to the entire halo .
rotation curves rarely extend beyond the inner @xmath234 of the virial radius of the halo @xcite , strong lensing is generally also most sensitive to the mass within @xmath235 of the virial radius @xcite , and weak lensing is generally most sensitive to the outer parts of halos ( excluding the inner @xmath213 of the virial radius ) @xcite .
x - ray gas profiles in galaxies , groups , and clusters can extend up to several tens of percents from the halo centers @xcite .
if dark - matter halos were well described by the nfw profile and there were no systematics in the observations that would bias the determination of the dark - matter profile , then all probes should yield the same profile fits .
different probes sometimes yield different profile fits , though .
for example , it has been claimed that nfw fits to dwarf - galaxy rotation curves trend lower in concentrations than predicted in dissipationless cold - dark - matter simulations , and that nfw profiles are not always good fits to the data @xcite .
strong - lens systems and x - ray clusters appear to have higher concentrations than the average concentrations predicted by simulations , although a large part of that may be due to selection effects @xcite . in some cases ,
multiple probes are used on the same system and show different fits .
for example , ref .
@xcite found moderate tension between the weak - lensing and strong - lensing nfw fits to abell 611 ( although the mass contained within 100 kpc , which is approximately the transition point between the domains of the probes , is consistent ) , and strong tension with nfw fits derived from stellar kinematic data , so much so that density profiles significantly shallower than nfw are preferred if the three data sets are combined @xcite .
we would like to explore the possibility that these discrepancies could be explained in the context of decaying dark matter . in sec .
[ sec : results : obs ] , we showed that the nfw profile is not always a good fit to dark - matter halos after some of the @xmath2 particles have decayed , especially if @xmath236 .
thus , profile fits to subsections of the halo may deviate from fits to the whole halo .
nfw fits to a simulation of a @xmath237 , @xmath64 halo with @xmath238 and @xmath239 gyr .
the solid line represents a fit using all data down to the inner cut - off ( 4 kpc ) , the short - dashed line represents fits using particles within half of the initial virial radius , the long - dashed line shows fits within the inner tenth of the initial virial radius , and the dot - dashed line shows fits to the outer half of the halo .
the error bars are the 90% limits from bootstrap resampling particles at each timestep , although the error bars lose meaning as the @xmath103 becomes large .
top panel : nfw fits to the virial mass .
the true virial mass follows the solid line .
middle panel : the nfw concentration .
bottom panel : @xmath103/d.o.f . for the nfw fit.,scaledwidth=45.0% ] to demonstrate trends in the fit parameters as a function of where in the halo one performs the fit , we show in fig .
[ fig : fit ] nfw fit parameters and the goodness of fit ( @xmath111d.o.f . ) as a function of time for a halo initially in equilibrium with @xmath240 and @xmath64 , with @xmath223 and @xmath130 gyr .
the solid ( blue ) line shows the nfw fit parameters @xmath107 and @xmath51 using all particles within the virial radius .
the error bars indicate the 90% range of parameter fits based on bootstrap resampling particles at each time slice .
the short - dashed line shows the best - fit parameters using only those particles in the inner half of the virial radius , and the long - dashed line shows the fits to particles within the inner 10% of the virial radius . in these three cases , the inner cut - off to the fits
is @xmath241 kpc , corresponding to the minimum radius at which numerical effects on the density profile can be ignored .
the dot - dashed line shows the fit resulting from considering only those particles with radii @xmath242 . in many cases ,
the nfw profile does not provide a good fit to the density profiles , but it is also not always a good fit to observational data .
the trends we discuss below are there regardless of the goodness of the nfw fit .
we find that the concentration derived from the inner parts of the halo is systematically lower than the concentration derived from the halo as a whole at late times , although at early times , the inner concentration is sometimes higher for @xmath243 or @xmath15 .
however , if the maximum radius used in the fit is @xmath244 , the concentration from the nfw fit is always lower than the concentration from the nfw fit to the halo as a whole . in general , the discrepancy among concentrations becomes less severe as a larger fraction of the halo particles are used for the fit .
the nfw fit becomes better and concentrations tend to be slightly higher if the minimum radius is raised to 10 kpc from 4 kpc , and in general , nfw fits become better as the maximum radius considered in the fit decreases
. fits to the outer part of the halo show a trend that the halo concentration is initially higher than that estimated from the halo as a whole , then dips low , and then becomes relatively high again .
this phenomenon is due to the finite time it takes for particles ejected from the inner halo to reach the outer halo , and the longer dynamical ( and hence , equilibration ) time in the outer halo .
the nfw profile is almost always a reasonable fit to the data with @xmath245 . are these trends consistent with observations ?
the relatively low concentration measured in the innermost part of the halo relative to the halo as a whole is qualitatively consistent with the low concentrations inferred from low - surface - brightness and dwarf - galaxy rotation curves .
however , in the absence of selection biases , one would expect the weak - lensing - derived concentrations to be slightly higher than strong - lensing - derived concentrations for the same halos , and also slightly higher than concentrations inferred from x - ray gas profiles , which is the opposite of what is observed .
x - ray and strong - lensing halo samples are likely highly biased towards higher concentrations , and so a better understanding of selection effects is necessary in order to understand the differences in inferred concentration among the various lensing , x - ray , and dynamical probes of dark - matter density profiles .
nonetheless , the prediction from decaying - dark - matter models is that it is possible that different probes of dark - matter halo profiles will indicate different concentrations for the same halo , even if the nfw profile is shown to be a good fit to data from probes sensitive to the innermost or outermost parts of the halo , such as weak lensing or rotation curves .
in this work , we have used simulations of two - body dark - matter decay in isolated halos initially in equilibrium to characterize the response of halos as a function of the center - of - mass kick speed @xmath25 , the decay time @xmath4 , and time .
we find that for the union of the parameter space @xmath246 and @xmath247 , dark - matter halos are essentially unchanged ; the mass loss is @xmath198 over a hubble time , and the shape of the density profile does not change , although the concentration declines somewhat for @xmath248 .
for the union of @xmath134 and @xmath249 , there is severe mass loss and a radical change to the shape of the dark - matter density profile .
for the rest of the decaying - dark - matter parameter space , there is moderate mass loss and significant deviations of the shape of the dark - matter density profile relative to the initial halo properties .
we used our simulations in conjunction with the observed mass - concentration relation and the galaxy - cluster mass function to constrain the two - parameter decay parameter space .
we find that the mass - concentration relation yields the stronger constraint on @xmath4 for a broad range of @xmath25 , constraining @xmath250 gyr for @xmath251 .
constraints on the decay parameters should improve greatly in the next 5 - 10 years with blossoming wide - field sunyaev - zeldovich and optical surveys , and with cosmological simulations of decaying dark matter . | we consider models in which a dark - matter particle decays to a slightly less massive daughter particle and a noninteracting massless particle .
the decay gives the daughter particle a small velocity kick .
self - gravitating dark - matter halos that have a virial velocity smaller than this velocity kick may be disrupted by these particle decays , while those with larger virial velocities will be heated .
we use numerical simulations to follow the detailed evolution of the total mass and density profile of self - gravitating systems composed of particles that undergo such velocity kicks as a function of the kick speed ( relative to the virial velocity ) and the decay time ( relative to the dynamical time ) .
we show how these decays will affect the halo mass - concentration relation and mass function . using measurements of the halo mass - concentration relation and galaxy - cluster mass function to constrain the lifetime
kick - velocity parameter space for decaying dark matter , we find roughly that the observations rule out the combination of kick velocities greater than 100 km s@xmath0 and decay times less than a few times the age of the universe . |
the standard model of cosmology , the so - called @xmath4cdm with the minimal six parameters , can explain well a wide range of cosmological observations , such as primordial abundance of light elements , cosmic microwave background(cmb ) and large scale structures(lss ) . meanwhile , there are still some hints that indicate that new physics beyond the minimal @xmath4cdm model maybe is needed in order to explain cdm sector better
. there are three controversies for cdm paradigm on small cosmological scales , _
cusp vs. core _ , _ too - big - to - fail _ and _ missing satellites _ ( see ref .
@xcite for a review ) , which have triggered both astrophysical @xcite and dm - related investigations @xcite .
a solution that resolves simultaneously these controversies has been proposed in ref .
@xcite , where both dm and active neutrino interact with a new gauge boson with mass round @xmath5 .
then the dm s velocity - dependent self - scattering cross section can be around @xmath6@xmath7 g at the dwarf satellites scale , and evades the constraints from milky way galaxy and galaxy cluster .
thus one can resolve the first two controversies .
meanwhile , the dm s scattering off cosmic neutrino background leads to its late kinetic decoupling at temperature @xmath8 , which is translated into a cut - off of the smallest protohalo mass @xmath9 , resolving the 3rd puzzle , namely missing satellites problem .
however , since active neutrino couples to a mev particle , such scenario is restrictively constrained @xcite .
the cmb data indicates that a small amount of relativistic species or hot dark matter(hdm ) could exist at cmb time @xcite , in addition to the standard three generations of active neutrinos .
this is often parametrized as the effective number of additional neutrino species @xmath10 .
it has been shown in ref .
@xcite that the best fit to all available data is given by @xmath11 where @xmath12 is the effective hdm mass .
also , it was shown very recently that a similar amount of hdm can help to relieve the tension of tensor - to - scalar ratio ( @xmath13 ) between planck data @xcite and the recently announced measurement of b - mode polarization by bicep2 @xcite , without a running spectral index @xcite .
it is well known that sterile neutrino can serve as a hdm component of the universe .
sterile neutrino is also well motivated in order to solve accelerator @xcite , reactor @xcite and gallium anomalies @xcite in neutrino oscillation experiments .
both reactor and gallium anomalies prefer a new mass - squared difference , @xmath14 @xcite , while accelerator experiments @xcite prefer @xmath15 . in all three cases
the favored mixing angles are around @xmath16 .
such a large mixing angle would in general lead to fully thermalized sterile neutrinos by oscillation and thus an increase of @xmath17 for each sterile neutrino .
this is in some tension with the above cosmological data eq .
( 1 ) , as shown in global fit @xcite including bbn , cmb and lss data
. the above tension can be relieved by introducing new interaction for sterile neutrino .
the new interaction can generate a matter potential @xmath18 that results in a tiny effective mixing angle @xmath19 in matter @xcite for @xmath20 , @xmath21 where @xmath22 is the mixing angle in vacuum . as a result
, the thermalization of sterile neutrino by oscillation can be suppressed and @xmath23 is easily obtained @xcite .
recently it has been shown that the tension in the data can be reconciled at @xmath24 level within an _ effective _ theory @xcite where a dim-5 operator is responsible for the active - sterile neutrino mixing . in this paper , we propose a _ ultraviolet _
complete theory for dm and sterile neutrino that can accommodate the aforementioned cosmological data and neutrino oscillation experiments within @xmath0 level .
the model includes both cdm and hdm , and we call it the @xmath25mdm(the first m stands for mixed ) .
we consider the standard seesaw model with two right - handed ( rh ) neutrinos ( gauge singlet ) @xmath26 in the lagrangian [ eq : lag ] for leptogenesis @xcite , which will not affect our discussions in the following . ] , and add a dark sector with @xmath1 gauge symmetry and coupling @xmath27 , and dark photon field @xmath28 , and dark higgs field @xmath29 and two different dirac fermion @xmath30 and @xmath31 in the dark sector .
all the new fields are sm gauge singlets .
we assign equal @xmath32 charges to @xmath33 and @xmath30 , which is normalized to @xmath6
. then the most general gauge invariant renormalizable lagrangian is given by @xmath34^{2 } - \lambda_{\phi h}\left[\phi_{x}^{\dagger}\phi_{x } - \frac{v_{\phi}^{2}}{2}\right]\left[h^{\dagger}h-\frac{v_{h}^{2}}{2}\right],\end{aligned}\ ] ] [ eq : lag ] where @xmath35 are the sm left - handed lepton doublets , @xmath36 is the sm higgs doublet , and @xmath37 is the field strength for sm @xmath38 . the covariant derivative on a field @xmath39 is defined as @xmath40 we have chosen the @xmath1 charge for @xmath31 in such a way that the @xmath41 term is forbidden by @xmath1 gauge symmetry ( otherwise @xmath31 may decay if kinamatically allowed ) .
thus @xmath31 would be stable and the dm candidate .
the local gauge symmetry is broken by the following vacuum configurations : @xmath42 where @xmath43gev and @xmath44(mev ) for our interest .
there will be mixings among various fields after the spontaneous gauge symmetry breaking .
the gauge kinetic mixing term results in tiny mixings among the physical gauge fields , @xmath45 and @xmath46 .
also there is a mixing between higgs fields @xmath47 and @xmath48 with @xmath49 two scalar excitations @xmath47 and @xmath48 can be expressed in terms of mass eigenstates , @xmath50 and @xmath51 , as @xmath52 with a mixing angle @xmath53 . because of the higgs portal interaction ( @xmath54 term ) and the additional scalar @xmath48 , the electroweak vacuum could be stable up to planck scale without additional new physics beyond the particle contents presented in eq .
( 2 ) ( see refs .
@xcite for example ) .
a novel feature of this model is that there can be mixing among three active neutrinos @xmath55 , sterile neutrinos @xmath56 and dark fermion @xmath30 due to @xmath57 , @xmath58 and @xmath59 after spontaneous gauge symmetry breaking . in order to correctly explain the active neutrino oscillation data ,
at least two @xmath60 s are needed , in which case two of @xmath61 are massive and the other one is massless .
then after diagonalization of @xmath62 mass matrix for @xmath63 and @xmath30 , mass eigenstates are composed of 7 majorana neutrinos , @xmath64 and @xmath65 , or collectively @xmath66 : @xmath67 where @xmath68 is the unitary mixing matrix that diagonalizes the mass matrix @xmath69 , @xmath70_{3\times 2 } & 0_{3\times 2 } \\ \dfrac{v}{\sqrt{2}}[y_{\alpha i}]^t_{2\times 3 } & \left[m^{r}_{ij}\right]_{2\times 2 } & \dfrac{v_\phi}{\sqrt{2}}(f_i \ ; g_i)_{2\times 2}\\ 0_{2\times 3 } & \dfrac{v_\phi}{\sqrt{2}}(f_i \ ; g_i)^t_{2\times
2 } & \left ( \begin{array}{cc } 0 & m_\phi \\ m_\phi & 0 \end{array } \right ) \\ \end{array } \right).\ ] ] in the following discussion , if not specified , we shall use @xmath61 and @xmath71 to collectively denote three active neutrinos and four sterile neutrinos , respectively .
the mixing also distributes the new @xmath1 gauge / yukawa interaction to all neutrinos with actual couplings depending on the exact mixing angles .
we assume that the mixing angles between @xmath55 and @xmath30 are negligible , compared to the mixing between @xmath56 and @xmath30
. this can be easily achieved by adjusting @xmath72 s , @xmath73 s and @xmath74 s .
a more straightforward way is to work in the flavor basis , in which only @xmath56 and @xmath48 have dark yukawa and gauge interactions , respectively . because of the new dark interactions for @xmath71 , all sterile neutrinos @xmath71 s are not thermalized by oscillation from active neutrinos and thus can contribute to the number of effective neutrino by a proper amount , @xmath75 after bbn @xcite .
the exact mass spectrum and mixing angles for @xmath71 are free , subject to conditions for fitting the data .
we shall take at least one @xmath71 is around 1 ev and others as free , lighter or heavier , and the mixing angles among @xmath71 are large enough for suppressing their production by oscillation from active neutrino .
based on a different setup , our model improves a similar attempt presented in a recent paper @xcite in two aspects .
first , our model is renormalizable and thus ultraviolet complete , while the model in ref .
@xcite assumed a dim-5 operator for generating the active - sterile neutrino mixing and therefore depends on the uv completion .
second , we shall show below that the model presented in the present paper can reconcile the current cosmological data with neutrino oscillation experiments within @xmath0 rather than only within @xmath24 as discussed in @xcite .
and ( b)@xmath76 scattering where @xmath77 for @xmath78 s majorana nature , @xmath79.,title="fig:",scaledwidth=23.0% ] and ( b)@xmath76 scattering where @xmath77 for @xmath78 s majorana nature , @xmath79.,title="fig:",scaledwidth=23.0% ] communication between dark sector and sm particles or thermal history before bbn time is determined mostly by two mixing parameters , @xmath80 and @xmath54 .
@xmath80 is contrained by dm direct searches around @xmath81 for @xmath82(tev ) @xmath31 and @xmath82(mev ) @xmath46 @xcite . and @xmath54 as small as @xmath83 would be enough to thermalize the dark sector at @xmath84tev @xcite . after the cross sections of dark particles scattering off sm particle drop below the expansion rate of the universe
, the dark sector decouples from the thermal bath of the visible sector and entropy density would be conserved separately in each sector .
the decoupling temperature of the dark sector , @xmath85 , would determine how much @xmath86 is left at a later time .
the exact value for @xmath86 will be given in the following .
chemical decoupling of dm from the heat bath sets its relic density today .
after the temperature drops below @xmath87 , @xmath31 starts to leave the chemical equilibrium and would finally freeze out at @xmath88 . to account for the correct thermal relic density , the thermal cross section for @xmath89 annihilation @xmath90 should be around @xmath91@xmath92s .
the dominant annihilation channel in this model is @xmath93 , and the relic density requires the gauge coupling @xmath27 to be @xcite @xmath94 where @xmath95 is the @xmath1 charge of @xmath31 and shall be taken @xmath3 for definiteness in later discussion .
we shall focus on the cdm @xmath31 with mass @xmath96 tev , which is preferred region as shown in ref .
@xcite .
kinetic decoupling of @xmath31 from @xmath71 happens at much later time when the elastic scattering rate for @xmath97 drops below some value determined by hubble parameter @xmath36 .
the feynman diagram is shown in fig .
[ fig : dmnu](b ) . for a thermal distribution of sterile neutrino ,
the decoupling temperature is given by @xmath98 where @xmath99 and @xmath100 are the temperatures of cmb and sterile neutrinos , respectively .
except that dm is dominantly scattering off sterile neutrinos in our model rather than active ones , the above formula is similar to the one in ref .
@xcite and gives the approximate order - of - magnitude estimation , although the precise formula may depend on the neutrino mixing angles from the couplings @xmath101 .
the kinetic decoupling of dm from the relativistic particles imprints on the matter power spectrum , for which there are two relevant scales @xcite : the comoving horizon @xmath102 and free - streaming length @xmath103 . for our interested regime ,
@xmath104 is much larger and relevant .
thus @xmath105 can be translated into a cutoff in the power spectrum of matter density perturbation with @xmath106 where @xmath107 is the sum of matter densities today , @xmath108 .
then @xmath9 can be easily obtained for explanation of _ missing satellites problem _ for @xmath82(tev ) @xmath31 and @xmath82(mev ) @xmath46 .
because of the light mediator @xmath46 , the dm self - scattering @xmath109 can have a large cross section , @xmath110@xmath7 at small scales , while relative small values at milky way and larger scales .
this can flatten the dark halo , decrease the total mass of halo centre and resolve both _
cusp vs. core _ and _ too - big - to - fail _ controversies .
the quantity that is usually used to describe the efficiency for the dm - dm self - scattering is the transfer cross section @xmath111 @xmath112 can be easily calculated from fig .
[ fig : dmnu](a ) in the perturbative regime @xmath113 as , @xmath114 \ , \\ & & \alpha_x=\dfrac{g_{x}^{2}}{4\pi } , \beta = \frac{2\alpha_{x}m_x}{m_\chi v_{\textrm{rel}}^{2 } } \ , r=\dfrac{m_\chi v_{\textrm{rel}}}{m_x},\end{aligned}\ ] ] where @xmath115 is the relative velocity of @xmath31 and @xmath116 .
@xmath115 is around @xmath117 km / s for dwarf galaxies , milky way and the galaxy clusters , respectively . in the non - perturbative regime @xmath118
, we have @xcite @xmath119 as an illustration , in fig .
[ fig : sigmat ] , we show the case with @xmath120 and @xmath121 , in which @xmath122 can be achieved properly for dwarf galaxies with @xmath123km / s . as function of relative velocity for @xmath120 and @xmath121.,scaledwidth=45.0% ]
and @xmath124 . the blue(solid ) and purple(dashed ) contours @xcite correspend to the @xmath0 and @xmath24 for the cosmological data with the best fit point @xmath125 .
the region between two red vertical lines can be achieved in our model .
and the horizontal dotted line marks the centre value for @xmath124 from the global fit for neutrino oscillation data in @xmath126 scenario @xcite .
we use @xmath127gev and @xmath128 is the confinement - deconfinement transition between quarks and hadrons . see the text for detail.[fig : fit],scaledwidth=45.0% ] after the decoupling of dark sector from the visible thermal bath , relativistic particles can still contribute to the radiation density . for 4 light sterile neutrinos
, their contributions to @xmath86 can be parametrized as @xmath129^{\frac{4}{3}}\nonumber \\ & = & 4\times\left[\frac{g_{\ast s}\left(t\right)}{g_{\ast s}^{x}\left(t\right)}\times \frac{g_{\ast s}^{x}\left(t_{x}^{\textrm{dec}}\right ) } { g_{\ast s}\left(t_{x}^{\textrm{dec}}\right)}\right]^{\frac{4}{3}},\end{aligned}\ ] ] where @xmath130 is the photon temperature , and @xmath131 counts the total number of relativistic degrees of freedom for entropy ( @xmath132 for dark sector ) .
conservation of entropy density has been used in the last step of the above equations .
when only sterile neutrinos are relativistic at the time just before bbn epoch , we have @xmath133 the parameter @xmath134 is well - known in sm @xcite and depends on the decoupling temperature .
for example , @xmath135 for @xmath136 . together with @xmath137 we can get @xmath138^{\frac{4}{3 } } \simeq 0.579.\ ] ] increasing(decreasing ) @xmath139 gives smaller(larger ) @xmath86 due to the changes in @xmath134 .
for instance , if @xmath140 , we would have @xmath141 and @xmath142 . if @xmath143 , we would have @xmath144 and @xmath145 . here
@xmath128 is the temperature for confinement - deconfinement transition between quarks and hadrons in qcd .
decoupling temperature lower than @xmath128 would give too large @xmath146 and therefore is excluded at high confidence level .
the available range for @xmath86 is the region between two red vertical lines in fig .
[ fig : fit ] .
if @xmath46 and @xmath51 are also relativistic around bbn time , we have @xmath147 in eq .
( [ eq : dneff ] ) and additional contributions from the bosonic part @xmath148 where the factor @xmath149 accounts for bosonic degrees of freedom normalized to fermonic one , @xmath150 . the ratio of @xmath86 for two cases is about @xmath151 so the difference is small and we shall not distinguish two cases in the later discussion
. these extra sterile neutrinos can also be relativistic even at cmb time with @xmath152 ev and play the role of hdm .
their effects on cosmology can be parametrized by the effective mass defined as @xmath153 where only relativistic sterile neutrinos are summed over .
sterile neutrino masses can be chosen to fit the neutrino oscillation data .
we take the face values from the global fit @xcite : for instance , with @xmath126 scenario @xcite gives @xmath154 and @xmath155 .
since @xmath156 is massless in our model , we have @xmath157 ev and @xmath158 ev . then using eq .
( [ eq : masseff ] ) , we depict the central value of @xmath124 as the horizontal dotted line in fig .
[ fig : fit ] .
we can see that cosmological data can be reconciled with neutrino oscillation experiments within @xmath0 in our model , which is quite remarkable .
the crucial difference between our model and ref .
@xcite is due to eqs .
( [ eq : dneff ] ) and ( [ eq : masseff ] ) , because of `` 4 '' sterile neutrinos in our model .
usually , only one sterile neutrino is responsible for @xmath86 and the relation among @xmath159 , for which one would have @xmath160 then this is consistent with neutrino oscillation data only at @xmath24 level as shown in ref .
@xcite . in the above discussion
we have assumed that @xmath161 for illustration .
this assumption may not be necessarily true when either oscillation brings all neutrinos into equilibrium or some sterile neutrinos are heavy enough such that they become non - relativistic at the time before cmb and heat other neutrinos . in both cases we have @xmath162 , and our model predictions are still consistent with neutrino oscillation data within @xmath0 level .
there are a few different ways to test our model
. direct detection of cdm @xmath31 will be possible for no vanishing @xmath80 . also , @xmath163 will annilhilate to two @xmath46s , which in turn decay into sterile neutrinos immediately .
these high energy sterile neutrinos can oscillate to active neutrinos which can be detected by neutrino telescropes , such as icecube , whose current limit on @xmath90 is around @xmath164@xmath92s @xcite .
taking into account boost factors due to light mediators in our model , future detection of these neutrino flux will be possible .
since we have more sterile neutrino species than other models , oscillation experiments could also be used to test the model even though this depends on the exact mixing angles and mass spectrum .
in this paper , we have proposed a _ ultraviolet _ complete renormalizable model for self - interacting cdm and sterile neutrinos that can accommodate the cosmological data and neutrino oscillation experiments simultaneously within @xmath0 level .
the model is based on a dark sector with local @xmath1 dark gauge symmetry that is spontaneously broken at @xmath82(mev ) scale .
the resulting @xmath82(mev ) gauge boson ( dark photon ) can mediate a dm self - scattering cross section around @xmath165@xmath7 g which is of right order to resolve two issues for cdm at small cosmological scales , _
cusp vs. core _ and _ too - big - to - fail_. in our model , two light rh gauge singlet neutrinos ( @xmath166 ) can mix with a dark fermion @xmath30 and therefore can interact with dm through the new dark gauge boson .
the relics of these sterile neutrinos serve as the hot dark matter with a right amount of @xmath86 ( see fig . 1 ) , which relieves the tension between planck and bicep2 .
the masses of these sterile neutrinos are consistent with neutrino oscillation experiments within @xmath0 .
meanwhile , the interaction between dm and sterile neutrino delays the dm s kinetic decouple to sub - kev temperature and induces a lower cut - off in the primordial matter power spectrum , resolving the _ missing satellites problem_. the model could be tested further by ( in)direct detection of cdm @xmath31 , and also through neutrino oscillation experiments if favorable parameters are realized in nature .
we are grateful to wan - il park for useful discussions .
this work is supported in part by national research foundation of korea ( nrf ) research grant 2012r1a2a1a01006053 , and by the nrf grant funded by the korea government ( msip ) ( no .
2009 - 0083526 ) through korea neutrino research center at seoul national university ( pk ) .
999 d. h. weinberg , j. s. bullock , f. governato , r. k. de naray and a. h. g. peter , arxiv:1306.0913 [ astro-ph.co ] . g. efstathiou , j. bond , and s. d. white , mon.not.roy.astron.soc .
* 258 * ( 1992 ) 16 . c. pfrommer , p. chang , and a. e. broderick , astrophys.j .
* 752 * ( 2012 ) 24 , arxiv:1106.5505 [ astro-ph.co ] .
j. silk and a. nusser , http://dx.doi.org/10.1088/0004-637x/725/1/556 [ astrophys.j .
* 725 * ( 2010 ) 556560 ] , arxiv:1004.0857 [ astro-ph.co ] .
p. r. shapiro , i. t. iliev , and a. c. raga , http://dx.doi.org/10.1111/j.1365-2966.2004.07364.x [ mon . not
* 348 * ( 2004 ) 753 ] , arxiv : astro - ph/0307266 .
mac low and a. ferrara , http://dx.doi.org/10.1086/306832 [ astrophys .
j. * 513 * ( 1999 ) 142 ] , arxiv : astro - ph/9801237 .
f. hopkins , e. quataert , and n. murray , mon.not.roy.astron.soc .
* 421 * ( 2012 ) 35223537 , arxiv:1110.4638 [ astro-ph.co ] . a. pontzen and f. governato , http://dx.doi.org/10.1111/j.1365-2966.2012.20571.x [ mon . not .
* 421 * ( 2012 ) 3464 ] , arxiv:1106.0499 [ astro-ph.co ] . c. a. vera - ciro , a. helmi , e. starkenburg , and m. a. breddels , arxiv:1202.6061 [ astro-ph.co ] .
j. wang , c. s. frenk , j. f. navarro , and l. gao , tt arxiv:1203.4097 [ astro-ph.ga ] . d. n. spergel and p. j. steinhardt , http://dx.doi.org/10.1103/physrevlett.84.3760 [ phys .
* 84 * ( 2000 ) 37603763 ] , arxiv : astro - ph/9909386 .
p. bode , j. p. ostriker , and n. turok , http://dx.doi.org/10.1086/321541 [ astrophys .
j. * 556 * ( 2001 ) 93107 ] , arxiv : astro - ph/0010389 .
j. j. dalcanton and c. j. hogan , http://dx.doi.org/10.1086/323207 [ astrophys .
j. * 561 * ( 2001 ) 3545 ] , arxiv : astro - ph/0004381 .
a. r. zentner and j. s. bullock , http://dx.doi.org/10.1086/378797 [ astrophys.j .
* 598 * ( 2003 ) 49 ] arxiv : astro - ph/0304292 .
k. sigurdson and m. kamionkowski , http://dx.doi.org/10.1103/physrevlett.92.171302 [ phys .
* 92 * ( 2004 ) 171302 ] , arxiv : astro - ph/0311486 . m. kaplinghat , http://dx.doi.org/10.1103/physrevd.72.063510 [ phys . rev . *
d72 * ( 2005 ) 063510 ] , arxiv : astro - ph/0507300 .
f. borzumati , t. bringmann , and p. ullio , http://dx.doi.org/10.1103/physrevd.77.063514 [ phys . rev .
* d77 * ( 2008 ) 063514 ] , arxiv : hep - ph/0701007 . c. boehm , p. fayet , and r. schaeffer , http://dx.doi.org/10.1016/s0370-2693(01)01060-7 [ phys
b518 * ( 2001 ) 814 ] , arxiv : astro - ph/0012504 .
m. kaplinghat , l. knox , and m. s. turner , http://dx.doi.org/10.1103/physrevlett.85.3335 [ phys .
* 85 * ( 2000 ) 3335 ] , arxiv : astro - ph/0005210 . w. hu , r. barkana , and a. gruzinov , http://dx.doi.org/10.1103/physrevlett.85.1158 [ phys .
* 85 * ( 2000 ) 11581161 ] , arxiv : astro - ph/0003365 . m. r. lovell , v. eke , c. s. frenk , l. gao , a. jenkins , et al .
, http://dx.doi.org/10.1111/j.1365-2966.2011.20200.x [ mon.not.roy.astron.soc . * 420 * ( 2012 ) 23182324 ] , arxiv:1104.2929 [ astro-ph.co ] . j. l. feng , m. kaplinghat , and h .- b .
yu , http://dx.doi.org/10.1103/physrevlett.104.151301 [ phys .
* 104 * ( 2010 ) 151301 ] , arxiv:0911.0422 [ hep - ph ] .
m. r. buckley and p. j. fox , http://dx.doi.org/10.1103/physrevd.81.083522 [ phys.rev .
* d81 * ( 2010 ) 083522 ] , arxiv:0911.3898 [ hep - ph ] .
a. loeb and n. weiner , http://dx.doi.org/10.1103/physrevlett.106.171302 [ phys.rev.lett .
* 106 * ( 2011 ) 171302 ] , arxiv:1011.6374 [ astro-ph.co ] .
m. vogelsberger , j. zavala , and a. loeb , mon.not.roy.astron.soc .
* 423 * ( 2012 ) 3740 , arxiv:1201.5892 [ astro-ph.co ] . s. tulin , h .- b .
yu , and k. m. zurek , http://arxiv.org/abs/1302.3898 [ arxiv:1302.3898 [ hep - ph ] ] . p.
ko and y. tang , arxiv:1402.6449 [ hep - ph ] .
l. g. van den aarssen , t. bringmann and c. pfrommer , phys .
lett . * 109 * , 231301 ( 2012 ) [ arxiv:1205.5809 [ astro-ph.co ] ] .
b. r .
ahlgren , t. ohlsson and s. zhou , phys .
rev . lett .
* 111 * , 199001 ( 2013 ) [ arxiv:1309.0991 [ hep - ph ] ] .
i. m. shoemaker , phys .
dark univ .
* 2 * , no . 3 , 157 ( 2013 ) [ arxiv:1305.1936 [ hep - ph ] ] .
r. j. wilkinson , c. boehm and j. lesgourgues , jcap * 1405 * , 011 ( 2014 ) [ arxiv:1401.7597 [ astro-ph.co ] ] .
r. a. burenin , astron .
lett . * 39 * , 357 ( 2013 ) [ arxiv:1301.4791 [ astro-ph.co ] ] .
m. wyman , d. h. rudd , r. a. vanderveld and w. hu , phys .
lett . * 112 * , 051302 ( 2014 ) [ arxiv:1307.7715 [ astro-ph.co ] ] .
j. hamann and j. hasenkamp , jcap * 1310 * , 044 ( 2013 ) [ arxiv:1308.3255 [ astro-ph.co ] ] .
r. a. battye and a. moss , phys .
* 112 * , 051303 ( 2014 ) [ arxiv:1308.5870 [ astro-ph.co ] ] .
s. gariazzo , c. giunti and m. laveder , jhep * 1311 * , 211 ( 2013 ) [ arxiv:1309.3192 [ hep - ph ] ] .
p. a. r. ade _ et al .
_ [ planck collaboration ] , arxiv:1303.5076 [ astro-ph.co ] .
a. r. ade _ et al .
_ [ bicep2 collaboration ] , arxiv:1403.3985 [ astro-ph.co ] .
e. giusarma , e. di valentino , m. lattanzi , a. melchiorri and o. mena , arxiv:1403.4852 [ astro-ph.co ] . j. -f .
zhang , y. -h .
li and x. zhang , arxiv:1403.7028 [ astro-ph.co ] .
cora dvorkin , mark wyman , douglas h. rudd , wayne hu , arxiv:1403.8049 [ astro-ph.co ] . c. athanassopoulos _ et al . _
[ lsnd collaboration ] , phys . rev .
lett . * 77 * , 3082 ( 1996 ) [ nucl - ex/9605003 ] .
a. aguilar - arevalo _ et al . _
[ lsnd collaboration ] , phys .
d * 64 * , 112007 ( 2001 ) [ hep - ex/0104049 ] .
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 ] ] .
j. n. abdurashitov , v. n. gavrin , s. v. girin , v. v. gorbachev , p. p.
gurkina , t. v. ibragimova , a. v. kalikhov and n. g. khairnasov _ et al .
_ , phys .
c * 73 * , 045805 ( 2006 ) [ nucl - ex/0512041 ] . c. giunti , m. laveder , y. f. li , q. y. liu and h. w. long , phys .
d * 86 * ( 2012 ) 113014 [ arxiv:1210.5715 [ hep - ph ] ] .
k. n. abazajian , m. a. acero , s. k. agarwalla , a. a. aguilar - arevalo , c. h. albright , s. antusch , c. a. arguelles and a. b. balantekin _ et al .
_ , arxiv:1204.5379 [ hep - ph ]
. b. armbruster _ et al .
_ [ karmen collaboration ] , phys .
d * 65 * ( 2002 ) 112001 [ hep - ex/0203021 ] .
a. a. aguilar - arevalo _ et al . _ [ miniboone collaboration ] , phys .
* 110 * ( 2013 ) 161801 [ arxiv:1207.4809 [ hep - ex ] , arxiv:1303.2588 [ hep - ex ] ] .
m. antonello , b. baibussinov , p. benetti , e. calligarich , n. canci , s. centro , a. cesana and k. cieslik _ et al .
_ , eur .
j. c * 73 * ( 2013 ) 2345 [ arxiv:1209.0122 [ hep - ex ] ] .
j. hamann , s. hannestad , g. g. raffelt and y. y. y. wong , jcap * 1109 * , 034 ( 2011 ) [ arxiv:1108.4136 [ astro-ph.co ] ] .
e. k. akhmedov , hep - ph/0001264 .
s. hannestad , r. s. hansen and t. tram , phys .
* 112 * ( 2014 ) 031802 [ arxiv:1310.5926 [ astro-ph.co ] ] . b. dasgupta and j. kopp , phys .
* 112 * ( 2014 ) 031803 [ arxiv:1310.6337 [ hep - ph ] ] .
t. bringmann , j. hasenkamp and j. kersten , arxiv:1312.4947 [ hep - ph ] .
m. fukugita and t. yanagida , phys .
b * 174 * , 45 ( 1986 )
. s. baek , p. ko and w. -i .
park , jhep * 1307 * , 013 ( 2013 ) [ arxiv:1303.4280 [ hep - ph ] ]
. b. batell , m. pospelov and a. ritz , phys .
d * 79 * ( 2009 ) 115019 [ arxiv:0903.3396 [ hep - ph ] ] .
s. baek , p. ko , w. -i .
park and y. tang , arxiv:1402.2115 [ hep - ph ] .
j. l. feng , m. kaplinghat and h. -b .
yu , phys .
d * 82 * , 083525 ( 2010 ) [ arxiv:1005.4678 [ hep - ph ] ] .
a. loeb and m. zaldarriaga , phys .
d * 71 * ( 2005 ) 103520 [ astro - ph/0504112 ] .
e. bertschinger , phys .
d * 74 * ( 2006 ) 063509 [ astro - ph/0607319 ] .
j. beringer _
[ particle data group collaboration ] , phys . rev .
d * 86 * , 010001 ( 2012 ) .
j. kopp , m. maltoni and t. schwetz , phys .
lett . * 107 * , 091801 ( 2011 ) [ arxiv:1103.4570 [ hep - ph ] ] .
m. archidiacono , n. fornengo , c. giunti and a. melchiorri , phys .
d * 86 * , 065028 ( 2012 ) [ arxiv:1207.6515 [ astro-ph.co ] ] .
j. kopp , p. a. n. machado , m. maltoni and t. schwetz , jhep * 1305 * , 050 ( 2013 ) [ arxiv:1303.3011 [ hep - ph ] ] .
m. g. aartsen _ et al . _
[ icecube collaboration ] , arxiv:1309.7007 [ astro-ph.he ] . | we propose a _ ultraviolet _ complete theory for cold dark matter(cdm ) and sterile neutrinos that can accommodate both cosmological data and neutrino oscillation experiments within @xmath0 level .
we assume a new @xmath1 dark gauge symmetry which is broken at @xmath2(mev ) scale resulting light dark photon .
such a light mediator for dm s self - scattering and scattering - off sterile neutrinos can resolve three controversies for cold dm on small cosmological scales : _
cusp vs. core _ , _ too - big - to - fail _ and _ missing satellites_. we can also accommodate @xmath3 ev scale sterile neutrinos as the hot dark matter(hdm ) and can fit some neutrino anomalies from neutrino oscillation experiments within @xmath0 .
finally the right amount of hdm can make a sizable contribution to dark radiation , and also helps to reconcile the tension between the data on the tensor - to - scalar ratio reported by planck and bicep2 collaborations . |
@xmath0 aurigae ( 7 aur , hd 31964 ; hr 1605 , @xmath0 aur hereafter ) is an eclipsing binary with the longest known orbital period of 27.1 years and showing an eclipse depth of 0.75 mag in the optical and almost a year long duration of totality phase . the eclipse depth is independent of wavelength from @xmath6 1400 to below the far - ir wavelength , but the depth , the duration of the eclipse and comparable masses of the components imply that the components should be almost equally bright .
however , no secondary eclipse was observed .
eclipse characteristics indicate that the occulting object is very elongated parallel to the binary orbit and it must be huge to block half of the primary s light .
it was suggested that the primary is occulted by a disk shaped secondary causing the two year long eclipse ( huang 1965 , kopal 1971 , gyldenkarne 1970 , wright 1970 , sahade & wood 1978 ) .
ir observations long - ward of 5@xmath7 region made by backman et al .
( 1984 ) revealed an excess emission in the infrared which can be attributed to the thermal radiation from the dust in the disk with a temperature of @xmath6 550 k ( also see hoard et al . 2010 ; muthumariappan & parthasarathy 2012 ) .
the k i 7699 line variation and hence the presence of neutral gas in and around the disk shaped secondary of @xmath0 aur was for the first time discovered by parthasarathy ( 1982 ) ( see also stencel 1982 , parthasarathy & lambert ( 1983a , b , c ) from the systematic increase in the strength of k i 7699 line during the 1982 - 1984 eclipse .
hack & selvelli ( 1979 ) first observed uv excess in the @xmath8 uv spectrum of @xmath0 aur .
the @xmath8 uv spectra obtained during the 1982 - 1984 eclipse also show uv excess and indicates a b5 v star at the centre of the disk ( parthasarathy & lambert 1983d ; boehm , ferluga & hack 1985 ; chapman , kondo & stencel 1983 ; altner et al .
1986 , ake 1984 , 2006 ) .
light and colour variations during the eclipse of @xmath0 aur was discussed by parthasarathy & frueh ( 1986 ) .
@xmath0 aur has recently finished an eclipse cycle which started in 2009 august and ended in 2011 august .
numerous photometric , spectroscopic and interferometric observations were carried out from uv to ir during 1982 - 1984 and 2009 - 2011 eclipses ; stencel ( 2012 ) presented a comprehensive summary of the recent eclipse results ( see table . 1 for photometric contact details ) .
the interferometric images published by kloppenborg et al .
( 2010 ) show that the eclipsing object is clearly disk shaped .
kloppenborg et al .
( 2012 ) presented the light - curve variation of @xmath0 aur during the 2009 - 2011 eclipse .
recently , stefanik et al . (
2010 ) reported an updated single - lined spectroscopic binary solution for the orbit of the f0ia primary based on 20 years of monitoring at the cfa , combined with historical velocity observations dating back to 1897 .
they presented two solutions .
one uses the velocities outside the eclipse phases together with mid - times of previous eclipses , from photometry dating back to 1842 , which provide strongest constraint on the ephemeris . from this
they find an orbital period of 9882 days ( 27.0938 years ) , an orbital eccentricity of 0.29 , system radial velocity of 2.26 @xmath9 0.15 km s@xmath10 and a mass function of 2.51 m@xmath4 for the binary using a combined fit to the data ( where the eclipse times were incorporated with the radial velocities into the least - square fit ) . by using only radial velocities
they find that the predicted middle of the current eclipse is nine months earlier than estimated from photometric study , implying that the gravitating companion in the binary is not the same as the eclipsing object .
they conclude that , the purely spectroscopic solution may be biased by perturbations in the velocities due to the short - period oscillations of the f0ia primary star .
other notable recent results are the interferometric studies by stencel et al .
( 2008 ) , photospheric abundance analysis of sadakane et al .
( 2010 ) , spectral and photometric analysis presented by chadima et al .
( 2011 ) and the detailed infrared studies of @xmath0 aur during the 2009 - 2011 eclipse ( stencel et al .
more recently , griffin & stencel ( 2013 ) conclude that the structure of the disk does not alter appreciably on a time scale of a century and they discovered a mass transfer stream from the f0ia star on to the disk .
they suggested that the f0ia may be a horizontal branch star .
the exact nature of the disk and the masses and the evolutionary nature of the binary components are not yet known well .
orbital characteristics and spectral properties of the primary indicate two different models for the @xmath0 aur system .
one is the high - mass star model with the f0ia primary having a mass of 15m@xmath4 ( carroll et al .
1990 ) in which the disk is a proto - planetary or proto - stellar disk and it is most unlikely to have resulted from the primary overflowing its roche lobe and transferring mass to the secondary .
the other one is the low - mass star model ( eggleton & pringle 1984 ) in which the f0ia primary has a mass of 4m@xmath4 or less , which supports an accretion disk for @xmath0 aur which is a remnant of post - main sequence mass transfer from the primary to the secondary .
distance measurement made by hipparcos ( 650 pc , perryman et al .
1997 ) and in turn an estimate of intrinsic luminosity of the primary does not favour a high - mass star model . however , hipparcos distance estimate has significant error and it was argued based on the interstellar absorption and reddening that the distance to @xmath0 aur is 1500 pc ( guinan et al .
2012 ) , which will set a large intrinsic luminosity for the system and indicates a high - mass star system .
though this issue is not yet resolved , it is now clear that @xmath0 aur is a binary system for which the brighter f0ia component can be less massive than its companion ( parthasarathy & pottasch 1986 ) .
disks are among most common astrophysical systems .
the @xmath0 aur disk has been described variously as thick ( huang 1965 ) or thin , flat or twisted , opaque or semi transparent , fully solid or possessing a central hole .
hoard et al . ( 2012 ) showed that their photometric data obtained with the @xmath11 @xmath12 @xmath13 are consistent with a cool disk model .
similar results were found from their 2-d radiative transfer modelling of the dusty disk by muthumariappan & parthasarathy ( 2012 ) .
they also discussed that the disk is not continuous , but has a void of radius 2 au at its centre .
if the low - mass star model suits for @xmath0 aur system , then it offers an unique opportunity to study in detail a new class of disks which are associated with post - agb stars .
@xmath0 aur shows a shell spectrum during the eclipse which manifests itself by the presence of additional absorption lines arising from the optically thin outer atmosphere of the disk .
numerous absorption lines are seen in the optical spectrum of @xmath0 aur which showed significant variation during the recent eclipse ( leadbeater et al .
the vast majority of the primary s lines are accompanied by one or two velocity shifted absorption components formed as the light of f0ia star passes through the neutral gas in and around the disk . these spectral lines which show absorption due to the disk and display spectral variation during the eclipse can be very useful tools to trace the disk structure and to derive its parameters .
the low - excitation k i line at 7699 is one of the key spectral lines to understand the disk as , unlike other lines mentioned above , this line was absent ( except the interstellar component ) in the out - of - eclipse spectrum of the f0ia star and shows systematic variation in strength during the eclipse phase ( parthasarathy 1982 , parthasarathy & lambert 1983a , b , c , lambert & sawyer 1986 , leadbeater et al . 2012 ) .
this line has been used to study the disk structures ( leadbeater & stencel 2010 ; potravnov & grinin 2013 ) .
gas associated with the secondary is most clearly traced by the k i resonance lines at 7664 and 7699 originated from in and around the disk shaped secondary . the stronger line at 7664 is generally blended with strong telluric lines of o@xmath3 .
it was also discussed that several low - excitation lines of fe peak elements like cr i , sc i , ti i are most likely shell lines which can be used to trace the disk structure ( sadakane et al .
however , unlike the k i at 7699 , these shell lines are not present in the spectra throughout the eclipse .
fe i lines appear from the mid - eclipse to the third contact and cr i lines are present from third contact to fourth contact .
all other lines appear only for a very short duration near the third contact ( see also fig . 4 of sadakane et al .
2013 ) .
in this paper we study the variations shown by the shell lines k i at 7699 ( excitation potential 1.6 ev ; k i line hereafter ) , fe i at 5110.435 and 5434.53 ( excitation potential 1.01 ev ; fe i lines hereafter ) and cr i at 5345.807 and 5348.326 ( excitation potential 1.00 ev ; cr i lines hereafter ) in their strength , line profile shape and in their radial velocity using high- and medium - resolution spectroscopic observations obtained during the 2009 - 2011 eclipse .
we also present the variation shown by the na d and h@xmath1 lines . using these results
we construct a model for the eclipse of @xmath0 aur and derive the geometrical parameters of the disk .
we further constrain the masses of the stellar components using our model and discuss the evolutionary status of @xmath0 aur and the origin of its disk .
the high - resolution spectroscopic observations used in this study were obtained from a fibre fed echelle spectrograph , attached to the prime focus of the vainu bappu telescope ( vbt ) located at the vainu bappu observatory , kavalur , india .
the echelle spectra in the 4000 to 9000 wavelength region were recorded on a 4k @xmath14 4k ccd detector with a pixel size of 12 @xmath7 m .
these observations were taken in stable and clear sky conditions with exposure times varying from 3 to 15 minutes for each spectrum .
we have recorded 52 echelle spectra during december 2010 to april 2012 .
our spectra hence were taken before the third contact ( 27 february 2011 ) to the end of the eclipse ( 26 august 2011 ) .
in addition , some spectra were recorded during eight months period after the end of the eclipse .
most of the spectra ( 33 of 52 ) were taken with the slit mode giving a spectral resolution of @xmath6 75 000 .
some of the spectra ( 19 of 52 ) were obtained in the slit - less mode which gives a resolution of 30 000 .
75 000 is the highest spectral resolution ever obtained on this object during its 2009 - 2011 eclipse .
these data were reduced and analysed with a pipelined data analysis steps using @xmath15 software .
we have taken the following data reduction sequence : bias subtraction , flat fielding , scattered light subtraction , spectral extraction and wavelength calibration .
the wavelength calibration was done using th - ar comparison spectra which were taken along with @xmath0 aur observations .
the observed velocities are then converted to the helio - centric values .
the telluric lines present at different spectral locations were used to find the error in our radial velocity calibration . the maximum error in radial velocities in the slit mode derived from un - blended lines
is found to be 78 m sec@xmath10 and for the blended lines we have employed multiple gaussian fitting which leads to a maximum error of 100 m sec@xmath10 in this observing mode .
spectra taken in the slit - less mode has a maximum error of 320 m sec@xmath10 for the un - blended lines and for the blended lines the maximum error is 400 m sec@xmath10 .
the signal - to - noise ratio of our observations varies from 200 to 400 .
table 2 lists the measurements of k i line equivalent widths and radial velocities obtained from vbt data .
we have combined these observations with 247 measurements of the k i line equivalent widths and radial velocities made by leadbeater at three hills observatory ( tho ) using a lhires iii spectrograph ( with a spectral resolution of @xmath16 = 22 000 ) with 200 mm and 280 mm telescopes .
the measurements cover the full eclipse period from 29 may 2009 to 21 march 2012 which are listed in table 3 . in this line ,
a constant interstellar component is blended with the system component for most of the eclipse .
this was removed before measuring the line parameters ( based on measurements made outside the eclipse ) .
further details of the data reduction can be found in leadbeater et al .
in addition , we have included the values of equivalent widths and radial velocities of k i lines measured from 47 high - resolution ( @xmath16 = 45 000 ) spectra ( after removing the interstellar component ) obtained by potravnov & grinin ( 2013 ) using maestro echelle spectrograph attached to the 2 m telescope at the terskol observatory ( to ) .
table 4 lists the measurements of k i line equivalent widths and radial velocities obtained from to data .
our data includes a total of 346 values of equivalent widths and of radial velocities of the k i 7699 line covering entire eclipse during 2009 to 2011 .
table 5 . shows the equivalent width and radial velocity values for the fe i 5110 line extracted from spectra submitted as part of the international @xmath0 aurigae campaign 2009 - 2011 and available on - line at @xmath17 .
the spectra were recorded using eshel echelle spectrographs at a resolving power @xmath16 = 10 000 ( @xmath18 ) .
telescope aperture was 280 mm ( buil , thizy ) and 355 mm ( garrel ) .
for the na d@xmath3 line , the equivalent width and radial velocity values from gorodenski ( 2012 ; observations made with a spectral resolution of @xmath16 = 18 000 ) were included in our study .
all the spectral lines presented in this study from three different observatories show significant variation in line profile shape and in radial velocity during the eclipse .
a time series of a few selected spectra taken near k i , fe i 5110 , fe i 5440 cr i , h@xmath1 and na d lines are shown respectively in fig . 1 , fig . 2 , fig . 3 , fig .
4 . fig . 5 , and in fig . 6 .
in addition , also shown in the figures are the spectra taken at these lines after the end of the eclipse ( on 7 april 2012 ) for comparison .
all the spectra shown in the figures were taken using vbt echelle spectrograph with a spectral resolution of 75 000 .
interstellar absorption components were well resolved and hence were not removed from these spectra .
the shell lines were seen in the spectrum only during or near the eclipse .
the k i line was present throughout the eclipse and was seen even after the eclipse for a few months .
this line shows a blue asymmetry until 18 february 2011 , and this asymmetry breaks into two blended components thereafter .
the line profiles are increasingly blue shifted until 18 march 2011 ( near the third contact ) , then onwards the trend becomes opposite ( see fig .
1 ) . as shown in fig . 2 , and fig .
3 , the fe i lines are blue shifted until 24 march 2011 and shows an opposite trend thereafter .
the absorption strength of the fe i lines also increases towards this turning point and the line strength is maximum on 2 april 2011 , nearly a month after the third contact .
however , after this date the fe i line is very weak .
cr i lines show very similar behaviour with the fe i lines both in radial velocity pattern and in absorption strength variation ( see fig .
the above findings on the fe i and cr i lines are in good agreement with the results of sadakane et al .
( 2013 ) . unlike the k i line
, the other three shell lines do not show any noticeable line profile asymmetry .
these lines were seen in the spectra only for a short duration near the third contact in our data and also in the archive .
this was also noted by sadakane et al .
( 2013 ) .
the h@xmath1 absorption line profiles are broad and display several sub - structures during the eclipse ( see fig . 5 ) .
spectra which were obtained after the eclipse show relatively narrower line profiles with no noticeable sub - structures .
line profiles are mostly asymmetrical ; the violet wing of the profile shows a steep fall and the red wing gradually increases to the continuum level .
the variation shown by these profiles is quite complex during the eclipse .
the h@xmath1 absorption line profile consists of both photospheric and shell components .
weak emission components are often seen at the violet and at the red edges of the absorption profile .
the red edge emission component is relatively brighter than the violet edge component .
the h@xmath1 absorption component also shows variability outside the eclipse phase ( guangwei et al .
the interpretation of h@xmath1 line profile variation during the eclipse hence needs to take the variations outside the eclipse into account .
for example , the complex behaviour of the h@xmath1 line during the eclipse was examined by chadima et al .
( 2011 ) after subtracting the mean out - of - eclipse profile .
as it can be seen from the spectra shown in fig 6 , the na doublet lines are quite broad ( and some are possibly saturated ) during the eclipse and do not have sharp dip . unlike the h@xmath1 line , these absorption lines have a more symmetrical structure and do not show any sub - structures .
they also do not have emission components at the edges of the absorption lines .
to understand the disk structure , we have analysed the variations shown by the shell absorption lines in their radial velocities ( rvs ) , equivalent widths ( ews ) and in full - width at half - maxima ( fwhms ) during the eclipse .
the values of these parameters were measured after fitting a continuum to the spectrum using a polynomial regression model .
the beginning and the end points of a line are defined where the line is judged to come in contact with , or cross through the estimated continuum .
the split seen in the k i line near the third contact was not taken into account for our analysis and we have considered this as a single line with mean velocity of its components . the wavelength , ew and fwhm of a symmetrical line profile were obtained by making a gaussian fit . wherever an asymmetry is seen in the line profile
we have employed multi - gaussian fitting method and the mean position of the velocity components was taken to estimate the rv .
the ew of individual components were quadratically added to get the total value .
the width of a line profile at its half maximum is taken as the fwhm , for both symmetrical the asymmetrical profiles . in the following sections we discuss the variations shown by these parameters for different spectral lines during the eclipse .
the k i line is strong enough throughout the eclipse so that we could measure these parameters in all the spectra .
however , the other shell lines show absorption features with measurable strengths only during a part of the eclipse phase ( near the third contact ) .
the values of ew and rv values for the k i line listed in table 4 , are re - measured from the spectra obtained by potravnov & grinin ( 2013 ) after the subtraction of the interstellar component .
the ew at a particular phase of eclipse measures the quantity of absorbing matter in the disk along the line - of - sight to the f0ia star .
the variation of ew against the julian date during the eclipse for three shell lines and na d@xmath3 line considered in this study are shown in fig .
the interstellar absorption component near the k i line is well resolved in all the vbt spectra , hence the ew of the disk component can be measured without any confusion .
the interstellar component of the k i line has a mean ew of 0.1036 @xmath9 0.0007
, as measured from the vbt spectra .
the ew of the interstellar k i line measured from the tho observations is 0.111 and potravnov & grinin ( 2013 ) reported a value of 0.119 .
the estimated uncertainty in ew measurements from vbt spectra , as arrived from repeated measurements , is less than 4@xmath19 for the k i and na d lines .
the error could be larger ( upto 7@xmath19 ) for the fe i , cr i lines as their absorption strengths are weak .
the k i line is nearly , but not quite optically thin during the eclipse .
the ew curve of the k i line shows a dip before 25 december 2010 ( see table 2 ) , which was not covered by vbt observations .
the vbt data do not cover the second contact when the ew reached its first maximum and the mid - eclipse phase where it reached a dip .
the second contact as seen from the k i line observations of tho was on jd 2455250 and the ew on this date is 0.476 .
the closest point to the second contact in the to data is on jd 2455275 on which the measured ew is 0.4 .
the mid - eclipse seen in the k i line had occurred on 27 july 2010 ( jd 2455405 ) as it is seen from the spectroscopic data of the tho ( see table 3 ) when the ew has a value of 0.332 .
the ew steeply rises to 0.860 on 20 feb 2011 ( jd 2455613 ) , which is the nearest point we have to the third contact in the vbt data ( see table 2 ) . the third contact , as seen from the spectroscopic data of tho , was on 26 feb 2010 ( jd 2455619 ) on which the measured ew was 0.796 ( see table 3 ) .
this disk - origin ew decreases gradually then after to a lowest value of 0.100 on 14 december 2011 ( jd 2455909 ) in the vbt data , and a value of 0.114 on jd 2455888 in the to data . in the tho data , the lowest ew value originated from the disk is 0.021 which was obtained on jd 2456005 .
the ew curve of na d@xmath3 line steeply rises to 1.83 on 29 march 2011 ( jd 2455650 ) in the vbt data then it decreases gradually to its lowest value of 1.15 which occurred on 5 december 2011 ( jd 2455900 ) .
the end of the optical photometric eclipse is on 26 august 2011 ( stencel 2012 ) and the k i line in principle should disappear completely , except for the interstellar part , after this date . however , the decline in ew continues well after the fourth contact , for about three months after the end of the eclipse .
the spectral line is seen in tho data with measurable strength until jd 2456005 , when the primary f0ia star should be far clear from the disk .
the fe i line and the cr i lines also seem to show a similar trend within the eclipse phase where they were observed .
the maximum ew values for the fe i line and cr i lines obtained from vbt data were near to the third contact ( jd 2455612 ) and their respective values are 0.071 and 0.052 .
this is very close to the date on which k i line also showed its maximum ew .
as noted earlier , these low - excitation lines are seen only for a short duration near the third contact and so there are no measurements available during the earlier eclipse phases to compare them with the k i line ew curve .
the ew curves of all shell lines do not show a smooth pattern and they display jumps at various eclipse phases .
this indicates a non - uniform neutral gas density distribution and presence of sub - structures in and around the disk ( see also leadbeater et al .
however vbt observations show that the positions of these jumps in ew curves are not same for k i , fe i , cr i and na d@xmath3 lines .
the rv at a particular phase of the eclipse measures the mean velocity of the absorbing material of the disk in front of the f0ia star , projected along the line - of - sight .
this includes the orbital motion of the disk around the centre of mass of the binary system and the rotation motion of the disk around the unseen secondary star .
the radial velocity curve can be used to deduce the rotation curve of the disk , as near the mid - eclipse the line - of - sight projected orbital velocity will be negligible but the observed rv of the disk is still larger than the rv of the f0ia star .
the variation of rv against the julian date during the eclipse for the three shell lines and na d@xmath3 line considered in this study are shown in fig .
the rv curves of k i and na d@xmath3 lines show similar trend : both show their lowest values around jd 2455613 ( near the third contact ) in vbt data and it increases there after . however , the lowest rv seen in these lines are different : 41.42 km @xmath20 for k i line ( see table 2 ) and 24.12 km s@xmath10 for na d@xmath3 line . the lowest rv obtained from the tho data at k i line was on jd 2455649 which is 40.5 , and on the third contact ( jd 2545619 )
its value is 38.6 ( see table 3 ) .
as the vbt data do not cover the second contact , the maximum value of rv could not be obtained .
the second contact as seen from the k i line observations of tho was on jd 2455250 with a rv of 17.5 km s@xmath10 .
however , the largest rv measured from k i line from tho data is 20.3 km s@xmath10 which was on jd 2455205 .
the largest rv measured from the na d@xmath3 line is 16.53 km s@xmath10 ( gorodenski 2012 ) .
the data obtained from to at k i line do not have a rv point close to the third contact but the rv value obtained close to the second contact is 21.8 km s@xmath10 which occurred on jd 2455275 .
radial velocities measured from the fe i line and cr i line do not show their minimum values on jd 2455613 .
the minimum rv observed for the fe i line is 45.5 km s@xmath10 which occurred on jd 2455649 ( near the third contact ) .
the minimum rv measured for the cr i line was on the same date .
its value is 46.8 km s@xmath10 , which is a little larger than the value shown by the fe i line .
both the fe i and cr i lines show minimum velocities which are lower than that observed for the k i line , however , the minimum velocities of these two shell lines were seen 30 days after the occurrence of the minimum velocity of the k i line .
these findings are in good agreement with sadakane et al .
the rv curve derived from the k i line ( fig .
8) shows maximum and minimum values which are not symmetric with respect to the radial velocity of the binary system ( 2.26 km s@xmath10 ) as also can be seen in fig .
6 of sadakane et al .
this radial velocity asymmetry is also seen in the rv curve of the na d@xmath3 line ( fig .
11 ) , however , not as large as it is seen for the k i line .
a possible explanation for this asymmetry is discussed in section 4.2 .
the fwhm traces the broadening of the spectral lines at a particular phase of the eclipse .
the variations of the fwhm against the julian date are shown in fig . 9 for the three shell lines and the na d@xmath3 line considered in our study .
the fwhm measured from the k i line and the na d@xmath3 line decrease to their minimum values which had occurred near jd 2455900 ( about three months after the end of optical eclipse ) and they increase then after . the minimum value of fwhm seen for the k i line is 0.286 and for the na d@xmath3 line it is 1.02 .
k i line fwhm could be measured in our data until jd 2456005 .
this steady increase of fwhm curve may show the presence of matter radially outside the main disk , possibly in the form of a ring with large velocity dispersion .
as seen earlier , evidence for this extended disk was also seen in the ew curves of the shell lines .
though a trend of decreasing fwhm was seen for the fe i and cr i lines , similar to the k i and na d@xmath3 lines , there are no reliable value of fwhm available beyond jd 2455660 due to the weakness of the spectral line after the third contact .
the minimum fwhm for the fe i
line and cr i lines which were observed near the third contact have their respective values of 0.25 ( 14.6 km s@xmath10 ) and 0.26 ( 14.5 km s@xmath10 ) .
we attempt here to understand the trends shown by the ew and the rv curves of the shell lines of @xmath0 aur during the eclipse using a geometrical model and analytically solving the line parameters in the optically thin limit . using this model
we further try to constrain the disk geometrical parameters and the stellar masses of the binary system . in this model
we assume that the disk is cylindrical in shape in which gas is distributed uniformly .
we do not address here the disk sub - structures shown by the spectral lines , however , to understand the trends shown by the ew and rv curves of the shell lines this can be sufficient .
the gas in the disk is in keplerian rotation around the star located at the centre of the disk .
this assumption establishes a relation between the rv due to the disk rotation , the size of the disk and the mass of the secondary .
the optically visible f0ia star and the central star of the disk rotate around their centre of mass in an orbit which has an inclination of 90@xmath21 .
the inner disk is opaque and seen edge - on ( as suggested by earlier studies ; hoard et al .
2010 ; muthumariappan & parthasarathy 2012 ) and the shell absorption lines arise from the peripheral and transparent atmosphere of the disk located around the opaque region .
we consider that the disk has an outer diameter ( the diameter of the atmosphere ) of @xmath22 and has an inner diameter ( the diameter of the opaque disk ) of @xmath23 .
the shell absorption lines are formed between @xmath22 and @xmath23 .
the disk has a central void of diameter 4 au ( muthumariappan & parthasarathy 2012 ) which is surrounded by the opaque disk and the gaseous atmosphere surrounds the opaque disk , hence , @xmath24 4 au .
we further assume here that the disk diameter is significantly larger than the photospheric size of the f0ia star .
kloppenborg et al .
( 2010 ) showed that the angular diameter of the opaque disk is about six times the angular diameter of the star .
hence , @xmath22 is larger than six times the size of the photosphere and this assumption is appropriate .
as the disk passes in front of the f0ia star , neutral gas which is present in the disk column overlapping the line - of - sight cone absorbs the continuum flux from the f0ia star ( primary ) producing the observed shell lines .
the thickness of the disk is taken to be constant .
the total light absorbed by an infinitesimal cell element at a position @xmath25 and @xmath26 from the central star ( secondary ) of the disk ( @xmath25 is normal and @xmath26 is parallel to the line of sight ) , is hence @xmath27 where , @xmath28 is a constant and @xmath29 is the density of the cell .
integrating this value over the line - of - sight column will give the total ew .
the integration limits of @xmath26 are from @xmath30 to @xmath31 .
the integration limits of @xmath25 depends on the line - of - sight cone size of the primary , which in turn depend on the distance .
but the computed value is normalised with respect to the peak value and the distance does not constrain on the shape of ew curve .
the variation of ew with the phase of the eclipse as the disk moves in front of the f0ia star produces the ew curve .
the radial velocity of an infinitesimal cell element is contributed by its line - of - sight projected keplerian velocity around the secondary ( @xmath32 ) , the line - of - sight projected orbital velocity of the secondary around the centre of mass of the binary system ( @xmath33 ) and the system radial velocity ( @xmath34 ) .
we included a constant value @xmath35 to fit the rv curve ( see also section 4.2 ) .
if @xmath36 and @xmath37 are respectively the masses of the primary and the secondary , the rv of a cell element at a distance @xmath38 from the centre of the disk which makes an angle @xmath39 with the line - of - sight is @xmath40 where , @xmath41 the line - of - sight projected orbital velocity of the secondary around the centre of mass can be found as follows .
the disk comes along the line - of - sight to the primary during the mid - eclipse and the model requires that the secondary is in its apastron point of its orbit around the centre of mass of the binary system during the mid - eclipse ( fitting ew and rv curves could not be achieved otherwise using the allowed variable parameters ) .
the distance from the focus to the apastron of the secondary s elliptical orbit with a period @xmath42 and an eccentricity @xmath43 is given by @xmath44 \times [ p^{2}g(m_{1}+m_{2})/4\pi^2]^{1/3}\ ] ] a similar relation for the primary can be obtained by interchanging @xmath36 and @xmath37 . the radial distance ( @xmath45 ) and the velocity ( @xmath46 ) of the secondary at an angular location @xmath47 ( measured from @xmath48 to the radial vector to the centre of mass ;
the angular location of the primary is 180@xmath21 + @xmath47 ) is @xmath49 @xmath50\times[1+e^{2}-2ecos\theta]}\ ] ] the angle between this velocity vector and the line perpendicular to the radial vector @xmath51 is given as @xmath52 the line - of - sight component of the secondary s orbital velocity is hence @xmath53 the density - weighted values of @xmath32 of the cell elements along the line - of - sight cone has a peak value , which depends on @xmath23 , @xmath22 and the density function .
this is added with @xmath33 , @xmath34 and @xmath35 to get the observed radial velocity at any instant .
the time variation of rv as the disk eclipses the primary gives the observed rv curve .
the primary moves in the opposite direction with the same angular velocity of the secondary with respect the the centre of mass .
the ingress and egress times of the eclipse take this into account .
10 shows the variation of line - of - sight integrated radial velocity components of @xmath32 , @xmath33 and their addition against the eclipse phase .
* as it can be seen from equations 1 to 8 , only the disk geometrical parameters , the radial density function and the stellar masses are used to fit the rv curve and the shape of the ew curve .
if the angular diameter of the optical disk is known accurately , then the distance is given from the disk linear diameter derived from the model .
further , the masses of the stellar components depend on the distance ( higher mass for a larger distance ) due to the observed luminosity of the primary and the binary mass function . hence
, the distance is implied by the outcome of the model , though it is not an input variable . * @xmath23 and @xmath22 constrain both the ew and rv curves . as it can be seen from equation 2 .
, the rv curve is also determined by the secondary mass .
the diameter of the disk is related to @xmath45 and also to @xmath48 as the eclipse phases are fixed and hence the ew is indirectly connected to the primary to the total stellar mass ratio .
the rv curve is strongly stellar mass dependent .
if the geometrical and density parameters are constrained from the ew curve then fitting the rv curve will constrain the secondary mass . the variable parameters of our model
are @xmath23 , d@xmath54 and the secondary mass and we worked for a model which can fit both the ew and the rv curves simultaneously .
the radial density function in the disk is taken as the density function in the disk mid - plane of the accretion disk given by lynden - bell @xmath55 pringle ( 1974 ) . the mass function of the binary system , its orbital period and the eccentricity of the orbit were taken from stefanik et al .
( 2010 ) corresponding to their combined fit .
these values are 2.51 m@xmath4 , 9896 days and 0.227 respectively .
the masses of the component stars allowed by the mass function are from 5.0 to 14.3 m@xmath4 for the secondary and correspondingly from 2.0 to 19.9 m@xmath4 for the f0ia primary ( taking the orbital inclination as 90@xmath21 ) . for a set of input values of the stellar masses allowed by the mass function
, the semi - major axis of the secondary s orbit was calculated using the orbital period . from this , the apastron radius and in turn the radial velocity were estimated as described above .
this was exercised for different sets of allowed stellar masses to reach a best fit to the observed rv curve using the disk inner and outer diameter derived from the ew fit .
different models were computed to fit the ew and the rv curves of k i , fe i and na d@xmath3 lines during the eclipse . for the fe
i line , significant absorption was seen from the mid - eclipse to the end of the eclipse .
we could not fit these curves for the cr i lines , as discussed earlier , these low - excitation lines appear only for a very short duration near the third contact and there are no sufficient data points available tracing the full eclipse phase to perform a model fit . the best - fit model ew and rv curves are plotted against their observations in fig .
11 and their model parameters are listed in table 6 . as it can be seen in fig 11 ,
while the rv curve of the k i line fits well by the model ( by making a velocity offset , see below ) , its observed ew curve shows an asymmetry which could not be reproduced by the model .
the ingress region of the disk deviates significantly from the model fit and the egress region makes a more reasonable fit by an uniform disk model .
the ew points after jd 2455800 are quite far from the model estimation .
the ew and rv curves of na d@xmath3 line deviate quite significantly from the model .
in addition to the observed asymmetry between the egress and the ingress seen in na d@xmath3 ew curve which is not reproduced in the model , the model curve is narrower than the observed one in the mid - eclipse to the end of the eclipse region .
the rv curves of the k i and the na d@xmath3 lines also show asymmetry : to fit the rv curve of the k i and fe i lines , offset values of 4.0 km s@xmath10 and 3.0 km s@xmath10 were required to be added to their respective model curves .
the rv curve of the na d@xmath3 line looks relatively more symmetrical , which needed an offset value of @xmath6 1.0 km s@xmath10 .
this may indicate that the disk is not circular rather it is elliptical in shape and the central star is located in the focus located away from the observer .
the major axis of the elliptical disk should also have an inclination with respect to the line - of - sight .
such a system can make a larger radial velocity towards the observer and a smaller radial velocity away from the observer when compared to a circular disk .
the combined rv curve can be fitted reasonably well by just making an offset to the centre of mass velocity .
the offset value is significantly smaller for the rv curve of the na d@xmath3 line than for the k i and fe i lines .
though the rv points of na d@xmath3 line before jd 2455300 and after jd 2455700 could not be fit by the model and the overall fit also is not as good as that for the k i and fe i lines , the difference seen in the offset is significant .
a possible explanation could be that the region of the disk traced by the na d@xmath3 line is relatively more circular .
the k i line absorption occurs in the disk from @xmath23 = 5.8 au to @xmath22 = 8.9 au , whereas , for the fe i line the absorption occurs between @xmath23 = 5.4 au and @xmath22 = 8.7 au . for the na d@xmath3 line , the absorption occurs in a region of the disk with @xmath23 = 4.1 au to @xmath22 @xmath6 8.1 au .
these model parameters of the na d@xmath3 line rv curve should be taken with caution , as we could not reach a good fit to the observations .
the best possible fit shows that the outer and inner diameters of the disk traced by the na d@xmath3 line is smaller than the disk diameters modelled from the other two lines .
the mass of the b5v star at the centre of the disk derived from our model is 5.4 m@xmath4 and the corresponding mass of the f0ia primary , given by the mass function , is 2.5 m@xmath4 . the model for a higher primary mass of 12 m@xmath4
was also calculated and are presented in fig 11 . while the ew curves are the same for these two models ( as the geometrical parameters are the same ) , the rv curve of the high - mass star model shows a large deviation from the observations ( see fig .
11 ) suggesting a low - mass binary system . it should be noted that the observed ingress and egress times of the eclipse and the relative position of the primary put stringent conditions on the value of @xmath48 and in turn the size of the disk . large @xmath48 will allow a large disk but can also change the beginning of the eclipse earlier and the end of the eclipse later than observed . for the high - mass stars
, our model ew curve constrains that the disk can not be larger than 1.1 times the disk size required for the low - mass star model .
but the rv curve requires this factor should be about 2 .
hence our model fit to the ew and the rv curves of the shell lines during eclipse , within its assumptions , favours a low - mass star binary for @xmath0 aur .
taking into account of the enhancement of @xmath5-process elements reported by sadakane et al .
( 2010 ) , the f0ia primary is most likely in post - agb stage of evolution .
as mentioned earlier , the ew curves of the k i and other shell lines show jumps at different locations .
this indicates that the disk is not homogeneous and fine structures with density jumps are present in the disk .
presence of rings and gaps in the disk plane were suggested earlier by leadbeater & stencel ( 2010 ) .
leadbeater et al .
( 2012 ) have discussed in detail on the jumps observed in the k i line ew curve in terms of different structures present in the disk .
the ingress region of the disk deviates significantly from an uniform - disk model , showing an asymmetry in the disk column density between the ingress and egress ( see also ake 2006 , parthasarathy & frueh 1986 ) .
the gas column density at the trailing edge found from our model is roughly twice that of the leading edge .
further , the spectral lines are seen in our data ( both from vbt and tho ) with measurable strength of ew until jd 2456010 , when the primary f0ia star should be far clear from the disk .
this indicates that the disk material does not come to an end abruptly after the fourth contact and the disk may have a halo which is significantly larger than its diameter .
as noted earlier , the fwhm curve also shows the presence of gaseous matter around the disk after the end of the photometric eclipse .
it was noted by leadbeater et al .
( 2012 ) that the ew of the k i line gradually increased before the eclipse , for nearly the same duration as seen after the eclipse .
this shows that with the k i line we are seeing through the full thickness of the disk atmosphere and the line profile has contributions from all depths within the absorbing region at a given phase of the eclipse .
the difference in the observed minimum velocity and a delay of 36 days in the time of occurrence in our measurements between the k i line and the fe i and cr i lines indicates that the k i line originates relatively in the outer regions when compared to the other two lines in the rotating disk .
this is also brought out from our models ( see table 6 ) .
the time delay was also noticed earlier by sadakane et al .
( 2013 ) , though the value found by them is 40 days .
the occurrence of fe i and cr i lines for only a short duration may imply that these elements are present only in a limited region .
ake ( 2006 ) obtained @xmath56 far - uv spectrum of @xmath0 aur which show rich emission line spectrum due to low - lying lines and he discussed one possibility of the emission was due to resonance scattering of photons in the disk from an occulting hot source .
griffin & stencel ( 2013 ) studied the blue and near - uv spectra of @xmath0 aur obtained during the recent eclipse .
they have also studied 130 digitized historic spectra from mount wilson ( dating back to 1930 ) and from the dao ( dating from 1971 ) .
they found precise repetition of disk - related spectral line changes during the three successive eclipses of @xmath0 aur .
they conclude that the structure of the disk does not alter appreciably on a time scale of a century .
they find that the disk is having an extensive and optically thin outer layer and a flat structure that is tilted near to edge on .
interferometric ( chara + mirc ) imaging decisively identified the eclipse - causing object to be an opaque disk .
the he i 10830 line near the mid - eclipses phases clearly indicates that there is hot object at the centre of the disk which was earlier suggested by parthasarathy & lambert ( 1983d ) based on the analysis of uv ( iue ) spectra during the 1982 - 84 eclipse .
one of the long standing and key unresolved question of the @xmath0 aur system is whether the f0ia star a low - mass star or a high - mass star .
the choice between these two not only shows the evolutionary status of the binary components , but also constrains the nature of the disk .
following arguments were put forward to support these two cases .
the hipparcos distance to @xmath0 aur of 650 pc is quite uncertain and a reliable distance estimate is very essential to address the evolutionary nature of the system .
distance estimates using other methods like the spectroscopic method , pulsation theory , membership in an ob association , etc .
, as discussed by carroll et al .
( 1991 ) , were also made and they indicate the distance to be upto 1.3 kpc .
guinan et al . ( 2012 ) derive a distance of 1.5 kpc based on the interstellar absorption and reddening .
the large distance and hence the large absolute luminosity of the f supergiant supports the high - mass star model ( lissauer & backman 1984 , guinan et al .
in addition , the semi - regular light variation of the f0ia star resembles to those found in luminous a - f supergiants ( carroll et al .
the high - mass star advocates that @xmath0 aur has a proto - planetary disk .
accurate determination of the mass function indicates that a primary of 12m@xmath4 should have a corresponding secondary of mass 11 m@xmath4 .
the major problem with the high - mass star model is that the secondary is highly under - luminous for its mass of 12m@xmath4 . to explain this luminosity problem , lissauer & backman ( 1984 ) proposed that the central object in the disk is a close binary where the total mass is equally shared by the binary components .
such a configuration can be stable if the close binary had a separation of @xmath57 5 au ( pendleton & black 1983 ) .
however , as discussed by hoard et al .
( 2010 ) , the disk should be 95@xmath19 transparent in this case for the minimum possible circumstellar and interstellar reddening of @xmath58 = 0.45 for @xmath0 aur and thus the disk can not eclipse the f0ia star .
our present study of the k i line ew and rv curves during recent eclipse shows that the ew curve does not allow the disk size to be very large as required by the rv curve , and the models we worked out do not favour high - mass for @xmath0 aur .
eggleton & pringle ( 1985 ) proposed the f0ia star to be a low - mass post - agb star to account for the luminosity problem of the secondary .
the low - mass star advocates that @xmath0 aur has an accretion disk . earlier discussion on the low - mass star model given by takeuti ( 1986 ) .
analysis of the radial velocities by lambert & sawyer ( 1986 ) showed a mass for the secondary between 3m@xmath4 to 6m@xmath4 , supporting a low - mass primary .
similar conclusion was also arrived by hinkle & simon ( 1987 ) from their accurate two micron co radial velocity measurements .
the interferometric images of @xmath0 aur during ingress were published by kloppenborg et al .
( 2010 ) which showed that the eclipsing body is an opaque disk which was modelled as a cylinder .
the mass ratio of the primary to the secondary obtained by them from binary solution is 0.62 which constrains a lower - mass for the f0ia primary of @xmath0 aur .
stencel ( 2013 ) summarized the results from the extensive multi - wavelength observations of @xmath0 aur during the recent eclipse .
hinkle & simon ( 1987 ) showed from their co infrared molecular line observations of f0ia star that the isotopic abundance ratio of @xmath59c/@xmath60c = 10 @xmath9 3 indicating a late evolutionary stage for the primary .
the material in the disk seems to be over abundant in @xmath60c and rare - earth elements which may be the result of accretion and/or mass transfer during the agb and post - agb stage of the primary ( see griffin & stencel 2012 ; stencel 2013 ) .
mourard et al .
( 2012 ) from the vega / chara visibility measurements of @xmath0 aur from 2009 to 2011 eclipse find that the formation of emission wings of h@xmath1 in an expanding zone almost twice the photospheric size of the f0ia star .
sadakane et al .
( 2010 ) obtained from their high resolution spectra taken outside the eclipse the abundances of mg , si , s , ca , sc , ti , cr and fe and these abundances are close to solar abundances .
they found , however , a slight , but definite over - abundances of @xmath5-process elements y , zr and ba and enhancement of n , na in the photosphere of the f0ia primary .
the over - abundances of n , na and @xmath5-process elements clearly indicates that f0ia star has experienced third dredge - up and @xmath5-process nucleosynthesis .
these results support the idea that the f0ia primary is a post - agb star .
recently ishigaki et al . (
2012 ) found similar @xmath5-process elements over abundance in the photosphere of post - agb f supergiant crl 2688 .
the presence of post - agb supergiant binaries with dusty disks was first suggested by parthasarathy & pottasch ( 1986 ) .
high - mass model of f0ia star can not explain the over abundance of @xmath5-process elements . for high - mass f0ia star , by the time it evolves for third dredge up and @xmath5-process overabundance to occur
, carbon burning will takes place and it will become a supernova ( iben & renzini 1983 ) . during the stable evolution of massive stars third
dredgeup and @xmath5-process nucleosynthesis does not occur ( maeder 1993 , martins 2014 ) .
muthumariappan & parthasarathy ( 2012 ) suggested from their 2-d radiative transfer modelling of @xmath0 aur disk that the grains in the disk of @xmath0 aur are most likely amorphous carbon , and it is quite unlikely that silicate dust is present in the disk .
this indicates that the dust in the disk was processed and does not show interstellar composition and hence it is not a proto - planetary disk .
they concluded that the disk was formed from mass transfer and/or accretion from the c rich post - agb star ( during slow - wind and/or super - wind mass - loss phase on the agb , the mass - loss rate estimated by castelli ( 1978 ) from the p - cygni profiles is @xmath6 10@xmath61 m@xmath4/yr ) to its main sequence companion .
they have also suggested very extended gaseous envelope around the f0ia star .
griffin & stencel ( 2013 ) have discovered that the disk is receiving from the f0ia primary a very confined stream of material that is enriched in rare - earth elements .
this mass - transfer stream is visible between egress phases of the eclipse .
griffin & stencel ( 2013 ) conclude from their study that the f0ia primary may be a horizontal branch star .
the above mentioned recent results on @xmath0 aur system indicate that the f0ia star is a post - agb star .
modelling of the ew and the rv curves during the recent eclipse carried out in our study supports that it is a low - mass binary system .
we argue that the disk of @xmath0 aur was formed very recently when the post - agb primary filled its roche lobe and transferred mass to the secondary , first suggested by eggleton & pringle ( 1985 ) .
we present high - resolution ( 30 000 & 70 000 from vbt and r = 45 000 from to ) and medium resolution ( 22 000 from tho and 10 000 from archive ) spectroscopic observations of the eclipsing binary @xmath0 aur taken during 2009 - 2011 eclipse .
strong variability in the shell spectral lines were seen both in their line profile shape and in their radial velocity .
low - excitation lines of fe i and cr i appear only for a short duration near the third contact .
non - uniformity in the disk structure is clearly seen from the ew and fwhm variations of the k i line .
it is also observed that the shell lines are present in the disk much after the end of the optical eclipse indicating an extended halo around the disk . using a geometrical model
, we fit the ew and rv curves of the k i derived from 346 data points , fe i lines and the na d@xmath3 line obtained from our data and from the archival data . the model for the k
i line shows that the shell absorption occurs in a transparent atmosphere of diameter 8.9 au around an opaque disk of diameter about 5.8 au .
it constrains that the @xmath0 aur binary system has low - mass constituents with the f0ia primary having a mass of 2.5 m@xmath4 and the mass of the hot star at the centre of the disk is 5.4 m@xmath4 .
our results and the overabundance of @xmath5- process elements found by sadakane et al .
( 2010 ) constrain that the f0ia primary is a post - agb star .
we thank the vbt observing staff for their co - ordination during our observations .
we thank c. buil , t. garrel and o. thizy of the international epsilon aurigae campaign 2009 - 2011 for allowing access to their spectra .
part of the work on this paper was carried during the mp s visiting professor position in iucaa .
mp is thankful to prof .
ajit kembhavi , prof .
kandaswamy subramanian and prof .
t. padmanabhan for their kind encouragement , support and hospitality .
we thank the anonymous referee for his / her useful comments .
ake t.b . , 2006 , in sonneborn , moos & anderson , eds , astrophysics in the far - uv , five years of discovery with @xmath56 , asp conf .
348 , p. 156
, simon t. , 1984 , in j.m .
mead et al .
eds . , future of uv astronomy based on six years of @xmath8 research , p. 361 altner b. , chapman r. d. , kondo y. , stencel r. e. , 1986 , a&a suppl .
65 , 199 backman d.e . , becklin e. e. , cruikshank d. , p. , joyce r. r. , simon t. , tokunaga a. , 1984 , apj , 284 , 799 boehm c. , ferluga s. , hack m. , 1984 , a&a 130 , 419 carrol s.m .
et al . , 1991 ,
apj 367 , 278 castelli f. , 1978 , a&a , 69 , 23 chadima p. et al . , 2011 , a&a , 530 , 146 chapman r.d . ,
kondo y. , stencel r.e .
, 1983 , apj 269 , l17 eggleton p.p . , pringle j.e . , 1985 , apj , 288 , 275 ferluga s. , hack m. , 1985 , a&a 144 , 395 gorodenski s.a . , 2012 , jaavso , 40 , 743 griffin r.e .
, stencel r. e. , 2013 , pasp , 125 , 775 guangwei c. , huisong t. , jun x. , youngsheng l. , 1994 , a&a , 284 , 874 guinan e.f .
, dewarf l.e . , 2002 , in tout c.a . ,
hamme w.v .
, eds , asp conf . ser .
279 , exotic stars as challenges to evolution , san francisco , p. 121
guinan e. f. et al . , 2012 , a&a , 546 , 123 gyldenkarne k. , 1970 , vistas astron .
, 12 , 199 hack m. , selvell p.l . , 1979 , a&a , 75 , 316 hinkle k.h . , simon t. , 1987 , apj , 315 , 296 hoard d.w . , howell s.b .
, stencel r.e . , 2010 , apj , 714 , 549 hoard d.w .
et al . , 2012 , apjl , 748 , l28 huang s. , 1965 , apj , 141 , 976 hinkle k.h . , simon t. , 1987 , apj , 315 , 296 ishigaki m. , parthasarathy m. , reddy b.e .
, garcia - lario p. , takada y. , aoki w. , garcia - hernandez d.a .
manchado a. , 2012 , mnras , 425 , 997 iben i. , renzini a. , 1983 , ara&a , 21 , 271 kloppenborg b. et al . , 2010 , nature , 464 , 870 kopal z. , 1971 , ap&ss , 10 , 332 lambert d. , sawyer s. , 1986 , pasp , 98 , 389 leadbeater r. et . al . , 2012 , jaavso , 40 , 729 leadbeater r. , stencel r.e . , 2010 , arxiv:1003.3617v2 lissauer j.j . ,
backman d.e . , 1984 , apj , 286 , l39 lynden - bell d. , pringle j.e . , 1974 , mnras , 168 , 603 maeder , a . , 1993 , ssrv 66 , 349 martins f. , 2014 , arxiv:1404.0166 mourard d. et al . , 2012 , a&a , 544 , 91 muthumariappan c. , parthasarathy p. 2012 ,
mnras , 423 , 2075 muthumariappan c. , parthasarathy m. , jayakumar k. , appakutty m. , 2012 , in manchado a. , stanghellini l. & schonberner d. eds , proc .
283 , planetary nebulae : an eye to the future .
cambridge univ . press ( cambridge ) , p. 454
- 455 454 parthasarathy m. 1982 , in @xmath0 aurigae newsletter , no . 5 , p.5 , december 1982 , edited by r.e .
parthyasarathy m. , frueh m.l .
, 1986 , ap & ss , 123 , 31 parthasarathy m. , lambert d.l . , 1983a , iau circular 3766 , 2 parthasarathy m. , lambert d.l .
, 1983b , iau circular , 3848,2 parthasarathy m. , lambert d.l .
, 1983c , iau circular , 3857 , 1 parthasarathy m. , lambert d.l .
, 1983d , pasp 95 , 1012 parthasarathy m. , pottasch s.r .
1986 , a&a 154 , l 16 pendleton y.j . ,
black d.c . , 1983 , aj , 88 , 1415 perryman m.a.c .
et al . , 1997 , a&a 323 , l49 potravnov i.s . , grinin v.p . , 2013 , a.rep , 57 , 991 sadakane k. , kambe e. , sato b. honda s. , hashimoto o. , 2010 , pasj , 62 , 1381 sadakane k. , kambe , e. , hashimoto o. , honda s. , sato b. , 2013 , pasj , 65 , l1 sahade j. , wood f.b . , 1978 , interacting binary stars , pergamon press , new york , p.152 stefanik r.p .
, torres g. , lovegrove j. , pera v.e . , latham d.w . , zajac j. , mazeh t. , 2010 , aj . , 139 , 1254 stencel r.e . , 1982 , in @xmath0 aurigae newsletter , no . 5 , p. 5 . ,
december 1982 , edited by r.e .
stencel stencel r.e . , 2012 ,
jaavso , 40 , 618 stencel r.e . , 2013 , giec conf . , 2010 stencel r. e , creech - eakman m. , hart a. , hopkins j.l . ,
kloppenborg b.k .
, mais d.e . , 2008 ,
apj , 689 , l137 stencel r.e . ,
kloppenborg b.k . ,
wall r.e . , 2011 ,
aj , 142 , 174 takeuti m. , 1986 , ap&ss , 120 , 1 wright k.o . , 1970 , vistas astron . , 12 , 147 | we report high - resolution ( r = 30 000 , 45 000 and 75 000 ) echelle and medium - resolution ( r = 22 000 and 10 000 ) spectroscopic observations of the long - period , eclipsing binary @xmath0 aurigae during the 2009 - 2011 eclipse .
low - excitation shell lines , viz , the k i line at 7699 ( with 346 data points ) , cr i lines at 5345.807 and 5348.326 and fe i line at 5110.435 which originated from the disk shaped secondary , h@xmath1 and the shell components of the na d@xmath2 and d@xmath3 lines show significant variation in their shapes and radial velocities during the eclipse .
the equivalent width curve shown by the k i line around the ingress and egress phases indicates that the gas density in the trailing edge is about a factor of two higher than the density in the leading edge . using a geometrical model ,
in which a homogeneous , cylindrical keplarian disk eclipses the f0ia primary star and the shell absorption lines originate from the gaseous atmosphere around an opaque disk , we fit the equivalent width and the radial velocity curves of the k i line covering the full eclipse .
a reasonably good fit can be achieved by a low - mass binary model where the mass of the central star of the disk is 5.4 m@xmath4 and the mass of the primary is 2.5 m@xmath4 and a disk size of 8.9 au .
the low - mass of the primary , with enhanced @xmath5-process elements found by sadakane et al . ( 2010 ) , supports that it is a post - agb f supergiant . for the high - mass binary model ,
the modelled radial velocity curve deviates significantly from the observations .
[ firstpage ] @xmath0 aurigae - stars : individual : eclipsing binaries - spectroscopy- stars : circumstellar matter - stars : agb and post - agb - stars : evolution |
borexino is a unique detector able to perform measurement of solar neutrinos fluxes in the energy region around 1 mev or below because of its low level of radioactive background .
after several years of efforts and tests with the prototype ctf detector the design goals have been reached and for some of the radioactive isotopes ( internal @xmath28u and @xmath29th ) largely exceeded .
the low background is an essential condition to perform the measurement : in fact solar neutrinos induced scintillations can not be distinguished on an event by event analysis from the ones due to background .
the energy shape of the solar neutrino is the main signature that has to be recognized in the experimental energy spectrum by a suitable fit procedure that includes the expected signal and the background .
the basic signature for the mono - energetic 0.862 mev @xmath0be neutrinos is the compton - like edge of the recoil electrons at 665 kev .
the detector is located deep underground ( approximately 3800 m of water equivalent , mwe ) in the hall c of the laboratori nazionali del gran sasso ( italy ) , where the muon flux is suppressed by a factor of 10@xmath30 .
the main goal of the experiment was the detection of the monochromatic neutrinos that are emitted in the electron capture of @xmath0be in the sun with 5% precision .
sketch of the borexino detector .
the base of the dome - like structure is 18 m in diameter . ]
the complete up to date technical description of the borexino detector has been reported in @xcite and @xcite .
the detector is schematically depicted in fig.[detector ] .
the inner part is an unsegmented stainless steel sphere ( sss ) that is both the container of the scintillator and the mechanical support of the photomultipliers . within this sphere ,
two nylon vessels separate the scintillator volume in three shells of radii 4.25 m , 5.50 m and 6.85 m , the latter being the radius of the sss itself .
the inner nylon vessel ( iv ) contains the liquid scintillator solution , namely pc ( pseudocumene , 1,2,4-trimethylbenzene c@xmath31h@xmath32(ch@xmath32)@xmath32 ) as a solvent and the fluor ppo ( 2,5- diphenyloxazole , c@xmath33h@xmath34no ) as a solute at a concentration of 1.5
g / l ( 0.17% by weight ) .
the second and the third shell contain pc with a small amount ( 5
g / l ) of dmp ( dimethylphthalate ) that is added as a light quencher in order to further reduce the scintillation yield of pure pc . the pc / ppo solution adopted as liquid scintillator satisfies specific requirements : high scintillation yield ( @xmath35 photons / mev ) , high light transparency ( the mean free path is typically 8 m ) and fast decay time ( @xmath363 ns ) , all essential for good energy resolution , precise spatial reconstruction , and good discrimination between @xmath37-like events and events due @xmath38 particles . furthermore ,
several conventional petrochemical techniques are feasible to purify the hundred of tons of fluids needed by borexino .
the scintillation light is collected by 2212 photomutipliers ( pmts ) that are uniformly attached to the inner surface of the sss .
all but 384 photomultipliers are equipped with light concentrators that are designed to reject photons not coming from the active scintillator volume , thus reducing the background due to radioactive decays originating in the buffer liquid or @xmath39 s from the pmts .
the tank has a cylindrical base with a diameter of 18 m and a hemispherical top with a maximum height of 16.9 m. the water tank ( wt ) is a powerful shielding against external background ( @xmath39rays and neutrons from the rock ) and is also used as a cherenkov muon counter and muon tracker . the muon flux , although reduced by a factor of @xmath40 by the 3800 m.w.e . depth of the gran sasso laboratory , is of the order of 1 m@xmath41 h@xmath42 , corresponding to about 4000 muons per day crossing the detector .
this flux is well above borexino requirements and a strong additional reduction factor ( about @xmath43 ) is necessary .
therefore the wt is equipped with 208 photomultipliers that collect the cherenkov light emitted by muons in water . in order to maximize the light collection efficiency the sss and the interior of the wt surface
are covered with a layer of tyvek , a white paper - like material made of polyethylene fibers .
the borexino has an excellent energy resolution for its size , this is the result of the high light yield of @xmath44 p.e./mev/2000 pmts .
the energy resolution ( 1@xmath45 ) at the @xmath0be compton edge energy ( 662 kev ) is as low as 44 kev ( or 6.6% ) .
analysis of the borexino data showed that the main goals concerning the natural radioactivity have been achieved .
the contamination of the liquid scintillator with respect to the u / th is at the level of @xmath46 g / g ; the contamination with @xmath47k is at the level of @xmath48g / g ( @xmath49 in natural potassium ) ; the @xmath50c content is @xmath51 g / g with respect to the @xmath52c . among the other contamination sources only @xmath53kr , @xmath54bi and @xmath54po
have been identified . the @xmath53kr counts @xmath360.3 ev / day / tone , it is @xmath37- emitter with 687 kev end - point .
the @xmath54po is the most intense contamination ( with initial activity of 60 counts / day / tone ) , it decays emitting monoenergetic @xmath38 with 5.3 mev energy , the half - life time of the isotope is 134 days .
the residual contaminations do not obscure the expected neutrino signal , the presence of the 862 kev monoenergetic @xmath0be solar neutrino is clearly seen in the experimental spectrum . in such a way , the collaboration succeeded to purify the liquid scintillator from residual natural radioactive isotopes down to the levels much lower than was initially envisaged for the @xmath0be neutrino measurement , which resulted in broadening of the initial scientific scope of the experiment .
the main goal of the experiment was the detection of the monochromatic neutrinos that are emitted in the electron capture of @xmath0be in the sun with 5% precision .
this goal has been achieved during the first stage of the experiment @xcite .
the borexino reported the first measurement of neutrino @xmath55b neutrinos with liquid scintillator detector with a 3 mev threshold on electrons recoil @xcite .
the stability of the detector allowed also to study the day - night effect of the @xmath0be solar neutrino signal , thereby allowing to completely exclude the low solution of the neutrino oscillation based on solar data alone @xcite .
finally , the low background of the detector , the refined analysis on threefold coincidences @xcite and the positronium discrimination method based on the positronium formation study made it possible to explore the 1 - 2 mev region with unprecedented sensitivity .
this led to the first observation of solar neutrinos from the basic pep reaction @xcite .
in addition , the best limit for the cno production in a star has been established . in this way
, borexino has completed direct detection of be-7 , pep and b-8 solar neutrino components thereby providing complete evidence of the transition from msw and vacuum oscillation of the lma solution of the solar neutrino problem ( fig .
[ pee ] ) .
electron neutrino survival probability , three points in this plot are derived from the borexino data . it should be noted that the precision of the be-7 neutrino flux itself is much better than the pee error at the corresponding point , this is related to the solar model uncertainties . ,
one of the goal of the borexino experiment is the measurement of all the solar neutrino fluxes , with the exception of the hep flux , too faint for detection in borexino .
the @xmath0be , pep and @xmath55b ( this last with the lowest threshold to date ) have been already measured , but the experimental uncertainties can be reduced .
in addition borexino will try to measure the pp and cno fluxes .
.[ssmvsdata]ssm predictions and current experimental results [ cols="^,^,^,^,^",options="header " , ] in table [ ssmvsdata ] the solar fluxes measured by borexino so far are compared with the ssm prediction , for low and high metallicity .
the experimental results agree , within the errors , with the ssm predictions , but can not distinguish between the two metallicities , due to the uncertainties of the model and the experimental errors .
it would be useful , at this moment , to recall what the metallicity puzzle is .
the solar surface heavy element abundance has been calculated about ten years ago with a 1d model , which uses data from spectroscopic observations of the elements present in the photosphere ( gs98 @xcite ) .
this model agrees with the helioseismology observations , namely the measurement of the speed of the mechanical waves in the sun .
more recently a 3d hydro - dynamical model ( agss09 @xcite ) of the near - surface solar convection , with improved energy transfer , has changed the z / x ratio with respect to the previous 1d treatment : 0.0178 ( low metallicity ) to be compared with the previous 0.0229 ( high metallicity ) .
the 3d model results perfectly reproduce the observed solar atmospheric line ( atomic and molecular ) profiles and asymmetries , but are in clear disagreement with the helioseismology data . at present
there is no satisfactory solution to this controversy @xcite .
the 1d and the 3d models predict different neutrino fluxes from the various nuclear reactions , as shown in table 1 , where they are compared with the experimental results obtained until now .
as stated above , it is not possible , at present , to discriminate the solutions due to model uncertainties and experimental errors .
a measurement of the cno flux , with reasonable errors , could distinguish between the two models which predict substantially different fluxes .
the pp solar neutrino flux has been never measured directly .
gallex and sage have measured the integrated solar flux from 233 kev , which , together with the borexino @xmath0be neutrino flux measurement and the experimental data on the @xmath55b neutrino flux , can be used to infer the pp neutrino flux with a relatively small uncertainty , once the luminosity constraint is applied ( fig .
[ pee ] ) .
nevertheless a direct experimental observation , which can be compared with the solar luminosity and the ssm prediction , would be an important achievement .
the pp flux measurement is part of the borexino phase 2 program .
improving the @xmath0be flux measurement is one of the goals for borexino phase ii .
the physics goals of this study can be summarized in three main points .
1 . reduction of the total error ( statistical+systematic ) down to hopefully @xmath363% .
even with such a precision this measurement can not solve unambiguously the metallicity puzzle , because of the uncertainties of the ssm .
the @xmath0be flux measured in phase i falls in between the two predicted flux values , for high and low metallicities , and a smaller uncertainty would not help if this will be the case also for phase ii .
a very precise experimental determination of the @xmath0be flux remains nonetheless an important tool for testing the solar model as well as a remarkable technical achievement .
2 . in the context of the neutrino physics , the non standard
neutrino interactions are currently debated .
one way to study them is to analyze the shape of the oscillation vacuum - matter transition region . while @xmath0be can not have a conclusive role in this matter ,
it can nevertheless help in restricting the range of the nsi flavor diagonal terms .
it is possible to constraint the nsi parameters studying the @xmath56 elastic scattering .
bounds are imposed by various other experiments on solar , atmospheric and reactor ( anti)neutrinos . but @xmath0be neutrinos have the strong advantage of being mono - energetic ( @xmath55b neutrino detected by the other solar experiments in real time have a continuous energy spectrum ) .
in borexino , the limitation to this analysis comes from the residual background , especially @xmath53kr , and , to lesser extend , @xmath54bi , which can mimic non - zero values of the nsi parameters @xmath57 and @xmath58 .
an increase of statistics does not help much if not accompanied by a reduction of such background .
this is the most important target of opportunity for phase 2 .
the very low @xmath53kr and reasonably low @xmath54bi achieved , make a direct pp measurement a reality . a careful understanding of the spectrum response in the @xmath50c end - point region is crucial , its study is possible through a dedicated effort .
the main problem in the pp - neutrino study is the disentanglement of the very tail of the @xmath50c spectrum ( with possible pile - up ) from the pp- neutrino spectrum .
the feasibility of the measurement is under study .
a direct detection of pp neutrinos would be a spectacular result and would justify the phase ii alone .
the analysis group performed a study of sensitivity to pp solar neutrinos with the current background levels achieved , the expected statistical precision of the pp - neutrino flux measurement is below 10% , the systematics error will be mainly connected with uncertainty of the @xmath59 pile - up spectrum and intensity determination .
the first indication for pep solar neutrinos has been reported by the collaboration @xcite .
the value for the pep interaction rate obtained in phase i ( 590 live - days ) was @xmath60 cpd / day/100 tones , the absence of a pep signal was rejected at 98% c.l .
the current measurement , in conjunction with the ssm ( the uncertainty in the pep flux is as low as 1.2% ) , yields a survival probability of pee = @xmath61 though the uncertainties are far from gaussian .
the precision is dominated by the statistical uncertainty ( about 20% ) , though with more ( background - free ) data , systematic uncertainties ( 10% ) will start to become important .
this is an extremely important result , but shy of the first measurement of pep solar neutrinos .
the addition of a modest batch of data with @xmath54bi reduced at or below 30 counts/(100 ton @xmath62 day ) will result in the first measurement ( 3@xmath45 ) of pep solar neutrinos .
a much prolonged data taking could also result in a 5@xmath45 precision measurement .
the measurement will allow to gauge the survival probability in the immediate proximity of the transition between two different oscillation regimes .
the borexino detector is the first large volume liquid scintillator detector sensitive to the low - energy solar neutrinos .
it possesses a very good energy resolution in comparison to the water cherenkov detectors , what allows to search for the solar @xmath55b neutrinos starting practically from the energies of the so called thallium limit ( maximum energy of @xmath39 rays from the chains of radioactive decay of @xmath29th and @xmath28u ; gamma - quantum with maximum energy e=2.6 mev is emitted in the decay of @xmath63tl ) .
the measurements of the @xmath55b above 2.8 mev has been performed using one year statistics ( 246 days of live- time ) of the borexino data @xcite .
the threshold of 2.8 mev is the lowest achieved so far in the @xmath55b neutrino real - time measurements .
the interest in the neutrino flux measurement with low threshold comes from the peculiar properties of the survival probability in this energy region .
the electron neutrino oscillations at e<2 mev are expected to be driven by the so called vacuum oscillation , and at energies e>5 mev - by resonant matter - enhanced mechanism . the energy region in between has never been investigated in spectrometric regime , and is of particular interest because of the expected smooth transition between the two types of oscillations .
the rate of @xmath55b solar neutrino interaction as measured through their scattering on the target electrons is [email protected](stat)@xmath640.01(sys ) cpd/100 tons .
this corresponds to an equivalent electron neutrino flux of @xmath65=([email protected]@xmath640.1)@xmath6210@xmath30 cm@xmath41s@xmath42 , as derived from the elastic scattering only , in good agreement with existing measurements and predictions .
the corresponding mean electron neutrino survival probability , assuming the bs07(gs98 ) standard solar model ( high z model ) , is [email protected] at the effective energy of 8.6 mev .
the ratio between the measured survival probabilities for @xmath0be and @xmath55b is 1.9@xmath45 apart from unity ( see fig.[pee ] ) .
for the first time the presence of a transition between the low energy vacuum - driven and the high - energy matter - enhanced solar neutrino oscillations is confirmed using the data from a single detector , the result is in agreement with the prediction of the msw - lma solution for solar neutrinos . acquiring more statistics ( of up to 5 years of the calendar time ) the borexino will provide the competitive measurement of the @xmath55b neutrino flux .
the borexino large size and possibility to reconstruct an interaction point ( with a precision of 14 cm at 1 mev energy deposit ) makes it an appropriate tools for searching of sterile neutrinos .
if the oscillation baseline is about 1 m ( which corresponds to @xmath671 ev@xmath68 ) , exposure of the detector to a compact powerful neutrino source should give rise to a typical oscillation picture with dips and rises in the spatial distribution of events density with respect to the source .
right beneath the borexino detector , there is a cubical pit ( side 105 cm ) accessible through a small squared tunnel ( side 95 cm ) that was built at the time of construction with the purpose of housing neutrino sources for calibration purposes . using this tunnel ,
the experiment with neutrino source can be done with no changes to the borexino layout .
the center of the pit is at 8.25 m from the detector center , requiring a relatively high activity of the neutrino source in order to provide detectable effect .
the experiment sox ( for short distance neutrino oscillations with borexino ) @xcite will be carried in three stages with gradually increasing sensitivity : phase a : : a @xmath69cr neutrino source of 200 - 400 pbq activity deployed at 8.25 m from the detector center ( external with respect to the detector ) ; phase b : : deploying a @xmath70ce-@xmath70pr antineutrino source with 2 - 4 pbq activity at 7.15 m from the detector center ( placed in the detector s water buffer ) ; phase c : : a similar @xmath70ce-@xmath70pr source placed right in the center of the detector .
layout of the borexino detector and the approximate location of the neutrino and anti - neutrino sources in the three phases.,scaledwidth=80.0% ] figure 1 shows a schematic layout of the borexino detector and the approximate location of the neutrino and anti - neutrino sources in the three phases .
two types of neutrino sources are considered : the @xmath69cr source and the @xmath70pr based source .
@xmath69cr decays via electron capture into @xmath69v , emitting two neutrino lines of 750 kev ( 90% ) and 430 kev ( 10% ) , while @xmath70pr decays @xmath37 into @xmath70nd with an end
point of 3 mev ( parent @xmath70ce decays too , but end - point of its @xmath37-decay is below the ibd threshold ) .
the portion of the @xmath70pr spectrum above the 1.8 mev detection threshold is the only of importance for the experiment .
elastic scattering of @xmath71 on electrons induce negligible background .
the source activity of 200 - 400 pbq is challenging , but only a factor 2 - 4 higher than what already done by gallex and sage in the 90 s . the @xmath70ce@xmath70pr experiment in phases
b and c does nt require high source activity .
the phase c is the most sensitive but it can be done only after the shutdown of the solar neutrino program , because it needs modification of the detector .
the phases a and b will not disturb the solar neutrino program of the experiment , which is supposed to continue until the end of 2015 , and do not require any change to borexino hardware .
the challenge for the phase c is constituted by the large background induced by the source in direct contact with the scintillator , that can be in principle tackled thanks to the correlated nature of the @xmath71 signal detection . in phase
b this background , though still present , is mitigated by the shielding of the buffer liquid .
borexino can study short distance neutrino oscillations in two ways : by comparing the detected number of events with expected value ( disappearance technique , or total counts method ) , or by observing the oscillation pattern in the events density over the detector volume ( waves method ) . in the last case the typical oscillations length is of the order
the variations in the survival probability @xmath72 could be seen on the spatial distribution of the detected events as the waves superimposed on the uniformly distributed background .
oscillation parameters can be directly extracted from the analysis of the waves .
the result may be obtained only if the size of the source is small compared to the oscillation length .
the @xmath69cr source will be made by about 10 - 35 kg of highly enriched cr metal chips which have a total volume of about 4 - 10 l. the source linear size will be about 15 - 23 cm , comparable to the spatial resolution of the detector .
the @xmath70ce@xmath70pr source is even more compact .
all simulations shown below takes into account the source size . in phase a the total counts method sensitivity
is enhanced by exploiting the fact that the life - time of the @xmath69cr is relatively short . in phases
b and c this time - dependent method is not effective because the source life - time is longer ( 411 days ) , but this is compensated by the very low background and by the larger cross - section .
the total counts and waves methods combined together yield a very good sensitivity for both experiments . besides , the wave method is independent on the intensity of the source , on detector efficiency , and is potentially a nice probe for un - expected new physics in the short distance behavior of neutrinos or anti - neutrinos .
the sensitivity of sox with respect to oscillation into sterile neutrino was evaluated with a toy monte carlo .
expected statistical samples ( 2000 events ) were generated for each pair of oscillation parameters .
we assume a period of 15 weeks of stable data taking before the source insertion in order to accurately constrain the background .
the background model includes all known components , identified and accurately measured during the first phase of borexino .
we built the confidence intervals from the mean @xmath73 for each couple of parameters with respect to the non - oscillation scenario .
the result is shown in fig .
[ fig : sensitivity ] , as one can see fro the figure the reactor anomaly region of interest is mostly covered .
sensitivity of the phase a ( @xmath69cr external , blue ) , of phase b ( @xmath70ce@xmath70pr external , red ) and phase c ( @xmath70ce@xmath70pr center , green ) .
the gray area is the one indicated by the reactor anomaly , if interpreted as oscillations to sterile neutrinos .
both 95% and 99% c.l .
are shown for all cases .
the yellow line indicates the region already excluded in @xcite.,scaledwidth=95.0% ] the simulation for the phase a are shown for a single irradiation of the @xmath69cr source up to the initial intensity of 370 pbq ( 10 mci ) at the site .
a similar result can be obtained with two irradiations of about 200 pbq if higher intensity turns out to be beyond the technical possibilities .
the single irradiation option is preferable and yields a slightly better signal to noise ratio .
the physics reach for the @xmath70ce@xmath70pr external ( phase b ) and internal ( phase c ) experiments , assuming 2.3 pbq ( 75 kci ) source strength and one and a half year of data taking ) is shown in the same figure ( fig .
[ fig : sensitivity ] ) .
the @xmath73 based sensitivity plots are computed assuming significantly bigger volume of liquid scintillator ( spherical vessel of 5.5 m radius ) , compared to the actual volume of liquid scintillator ( limited by a sphere with 4.25 m radius ) used for the solar phase .
such an increase will be made possible by the addition of the scintillating fluor ( ppo ) in the inner buffer region ( presently inert ) of the detector .
we have also conservatively considered exclusion of the innermost sphere of 1.5 m radius from the analysis in order to reject the gamma and bremsstrahlung backgrounds from the source assembly . under all these realistic assumptions ,
it can be noted from fig .
[ fig : sensitivity ] that the intrinsic @xmath70ce@xmath70pr sensitivity is very good : for example the 95% c.l .
exclusion plot predicted for the external test covers adequately the corresponding reactor anomaly zone , thus ensuring a very conclusive experimental result even without deploying the source in the central core of the detector .
borexino achieved its main goal , but is still a challenging detector . due to achieved level of purification ( much higher than it was needed for @xmath0be neutrino detection )
some measurements of solar neutrino fluxes beyond the original program were performed .
borexino - ii is operating with repurified ls , at new levels of radiopurity , two years of data are collected and are being analyzed aiming the pp - neutrino flux and cno neutrino flux measurement .
borexino is also an ideal detector to test the sterile neutrinos through the disappearance / wave effects .
the proposed staged approach ( @xmath69cr source at the first stage and two @xmath70ce@xmath70pr experiments ) is a comprehensive sterile neutrino search which will either confirm the effect or reject it in a clear and unambiguous way .
in particular , in case of one sterile neutrino with parameters corresponding to the central value of the reactor anomaly , sox will surely discover the effect , prove the existence of oscillations and measure the parameters through the `` oscillometry '' analysis .
borexino was made possible by funding from infn ( italy ) , nsf ( usa ) , bmbf , dfg , and mpg ( germany ) , nrc kurchatov institute ( russia ) , mnisw ( poland , polish national science center ( grant dec-2012/06/m / st2/00426 ) ) , russian foundation for basic research ( grant 13 - 02 - 92440 aspera , the nsfc - rfbr joint research program ) and rscf research program ( russia ) .
we acknowledge the generous support of the gran sasso national laboratories ( lngs ) .
sox is funded by the european research council . | borexino is a unique detector able to perform measurement of solar neutrinos fluxes in the energy region around 1 mev or below due to its low level of radioactive background .
it was constructed at the lngs underground laboratory with a goal of solar @xmath0be neutrino flux measurement with 5% precision .
the goal has been successfully achieved marking the end of the first stage of the experiment .
a number of other important measurements of solar neutrino fluxes have been performed during the first stage .
recently the collaboration conducted successful liquid scintillator repurification campaign aiming to reduce main contaminants in the sub - mev energy range . with the new levels of radiopurity borexino can improve existing and challenge a number of new measurements including : improvement of the results on the solar and terrestrial neutrino fluxes measurements ; measurement of pp and cno solar neutrino fluxes ; search for non - standard interactions of neutrino ; study of the neutrino oscillations on the short baseline with an artificial neutrino source ( search for sterile neutrino ) in context of sox project .
= 100000 * solar neutrino with borexino : results and perspectives * o. smirnov@xmath1 , g. bellini@xmath2 , j. benziger@xmath3 , d. bick@xmath4 , g. bonfini@xmath5 , d. bravo@xmath6 , b. caccianiga@xmath2 , f. calaprice@xmath7 , a. caminata@xmath8 , p. cavalcante@xmath5 , a. chavarria@xmath7 , a. chepurnov@xmath9 , d. dangelo@xmath2 , s. davini@xmath10 , a. derbin@xmath11 , a. empl@xmath10 , a. etenko@xmath12 , k. fomenko@xmath13 , d. franco@xmath14 , g. fiorentini@xmath15 , c. galbiati@xmath7 , s. gazzana@xmath5 , c. ghiano@xmath8 , m. giammarchi@xmath2 , m. gger - neff@xmath16 , a. goretti@xmath7 , c. hagner@xmath4 , e. hungerford@xmath10 , aldo ianni@xmath5 , andrea ianni@xmath7 , v. kobychev@xmath17 , d. korablev@xmath1 , g. korga@xmath10 , d. kryn@xmath18 , m. laubenstein@xmath5 , b. lehnert@xmath19 , t. lewke@xmath16 , e. litvinovich@xmath12 , f. lombardi@xmath5 , p. lombardi@xmath2 , l. ludhova@xmath2 , g. lukyanchenko@xmath12 , i. machulin@xmath12 , s. manecki@xmath6 , w. maneschg@xmath20 , f. mantovani@xmath15 , s. marcocci@xmath8 , q. meindl@xmath16 , e. meroni@xmath2 , m. meyer@xmath4 , l. miramonti@xmath2 , m. misiaszek@xmath21 , p. mosteiro@xmath7 , v. muratova@xmath11 , l. oberauer@xmath16 , m. obolensky@xmath18 , f. ortica@xmath22 , k. otis@xmath23 , m. pallavicini@xmath8 , l. papp@xmath24 , l. perasso@xmath8 , a. pocar@xmath23 , g. ranucci@xmath2 , a. razeto@xmath5 , a. re@xmath2 , b. ricci@xmath15 , a. romani@xmath22 , n. rossi@xmath5 , r. saldanha@xmath7 , c. salvo@xmath8 , s. schnert@xmath25 , h. simgen@xmath20 , m. skorokhvatov@xmath12 , a. sotnikov@xmath1 , s. sukhotin@xmath12 , y. suvorov@xmath26 , r. tartaglia@xmath5 , g. testera@xmath8 , d. vignaud@xmath18 , r.b .
vogelaar@xmath6 , f. von feilitzsch@xmath16 , h. wang@xmath27 , j. winter@xmath16 , m. wojcik@xmath21 , a. wright@xmath7 , m. wurm@xmath4 , o. zaimidoroga@xmath1 , s. zavatarelli@xmath8 , k. zuber@xmath19 , and g. zuzel@xmath21 .
( borexino collaboration ) |
the main goal of this paper is to provide a classification algorithm for a large class of compact hyperbolic coxeter polytopes .
a coxeter polytope in hyperbolic space is a convex fundamental domain for reflection group , i.e. a polytope whose all dihedral angles are integer parts of @xmath2 . for a compact hyperbolic coxeter polytope @xmath3
we denote by @xmath4 , @xmath5 and @xmath6 the dimension , the number of codimension one faces ( _ facets _ in the sequel ) , ans the number of pairs of disjoint facets respectively .
let @xmath7 be the set of polytopes satisfying the following condition : @xmath8 the set @xmath7 is finite .
all polytopes of this set can be listed by a finite algorithm provided in section [ algorithm ] . in general ,
the number of compact hyperbolic coxeter polytopes is infinite . in @xcite ,
makarov constructed infinite series of polytopes in dimensions @xmath9 and @xmath10 . in @xcite
, allcock used the @xmath0-dimensional polytope constructed by bugaenko @xcite to present an infinite series of @xmath0-polytopes .
another source of examples is given by right - angled compact coxeter polytopes , they exist in hyperbolic spaces of dimension at most @xmath9 ( see @xcite ) .
recently nikulin @xcite and agol , belolipetsky , storm , whyte @xcite proved that the number of maximal arithmetic hyperbolic reflection groups is finite .
a similar statement concerning general reflection groups does not hold : examples by makarov provide fundamental domains of infinite number of non - commensurable reflection groups .
however , all known examples of infinite series of polytopes ( in dimension at least @xmath9 ) can be obtained from a finite number of polytopes by a composition of two operations : taking a fundamental domain of a finite index reflection subgroup of the corresponding reflection group , or gluing two coxeter polytopes along congruent facets .
this gives rise to a notion of _ essential _ hyperbolic coxeter polytope which is minimal with respect to operations above ( see section [ essential ] for precise definitions ) .
a classification problem of compact hyperbolic coxeter polytopes can be easily reduced to a classification of essential polytopes . in dimensions
@xmath11 and @xmath12 compact hyperbolic coxeter polytopes are completely classified by poincar @xcite and andreev @xcite , so in this paper we restrict our attention to polytopes in the spaces of dimension at least @xmath9 . as we have mentioned above
, the number of known essential compact hyperbolic coxeter polytopes is finite ( all known polytopes can be obtained by gluing several copies of finite number of polytopes ) .
moreover , all polytopes of dimension greater than five that are known to be essential belong to @xmath7 . in section [ essential ] we discuss properties of essential polytopes and present a list of them .
while proving that a polytope is essential , the main tool is the result of @xcite which , roughly speaking , states that gluing operations can not decrease the number of facets .
a possible counterpart of the result by nikulin and agol , belolipetsky , storm , whyte in non - arithmetic case whould be a finiteness of the number of essential polytopes , which we conjecture in section [ essential ] .
all the remaining sections are devoted to the proof of the main theorem . here
the plan of the proof is .
first , we split the set of polytopes @xmath7 in the following way : @xmath13 where @xmath14 consists of polytopes from @xmath7 of dimension @xmath4 with @xmath5 facets .
further , each @xmath14 can be presented as @xmath15 where @xmath16 consists of polytopes from @xmath14 with dihedral angles not less than @xmath17 . for given @xmath18 and @xmath19 there are finitely many combinatorial types of polytopes and finitely many ways to assign angles between facets , which implies that @xmath16 is finite due to the following fact proved by andreev @xcite : an acute - angled polytope @xmath3 is completely determined by its dihedral angles .
our aim is to prove that for given @xmath4 and @xmath5 the set @xmath16 is empty for large @xmath19 ( which will imply that @xmath14 is finite ) , we do it in section [ multiplicity ] .
the key tool here is the trick allowing us to reduce dimension .
it is based on the following result due to borcherds @xcite : if a face @xmath20 of codimension @xmath11 is an intersection of two facets composing a dihedral angle less than @xmath21 , then @xmath20 itself is a coxeter polytope .
then we show ( section [ fin ] ) that for given @xmath4 the set @xmath14 is empty for large @xmath5 ( which will imply that @xmath22 is finite ) . in @xcite , vinberg proved that the dimension of a compact hyperbolic coxeter polytope does not exceed @xmath23 .
therefore , we need to consider sets @xmath22 for a finite number of dimensions only , so @xmath7 is finite . in section [ cox ]
we recall basic facts about coxeter polytopes and their facets , and reproduce the notation introduced in @xcite ; in section [ tools ] we describe the technique of local determinants and prove several technical lemmas . in section [ algorithm ]
we construct an algorithm which allows us to list all the polytopes of @xmath7 .
section [ dim4 ] is devoted to investigation of @xmath9-dimensional case .
we would like to thank the university of fribourg , where the most part of the work was carried out , for hospitality and nice working atmosphere , we thank r. kellerhals for invitation , numerous stimulating discussions , and for partial support by snf projects 200020 - 113199 and 200020 - 121506/1 .
we are grateful to v. emery for communicating unpublished results of @xcite and to a. postnikov for the idea of the proof of lemma [ esm ] .
we also thank ihes ( where the current version of the paper was prepared ) for hospitality and excellent working conditions .
in this section we list the essential facts about coxeter polytopes and their faces .
we mainly follow @xcite and @xcite , see also @xcite . in what follows
we write _ @xmath4-polytope _ instead of `` @xmath4-dimensional polytope '' , _ @xmath19-face _ instead of `` @xmath19-dimensional face '' and _ facet _ instead of `` face of codimension one '' .
an abstract _ coxeter diagram _
@xmath24 is a finite @xmath25-dimensional simplicial complex with weighted edges , where weights @xmath26 are positive , and if @xmath27 then @xmath28 for some integer @xmath29 .
a _ subdiagram _ of @xmath24 is a subcomplex with the same weights as in @xmath24 .
the _ order _ @xmath30 is the number of nodes of the diagram @xmath24 . if @xmath31 and @xmath32 are subdiagrams of an abstract coxeter diagram @xmath24 , we denote by @xmath33 a subdiagram of @xmath24 spanned by the nodes of @xmath31 and @xmath32 . given an abstract coxeter diagram @xmath24 with nodes @xmath34 and weights @xmath26 , we construct a symmetric @xmath35 matrix @xmath36 , where @xmath37 , @xmath38 if @xmath39 and @xmath40 are adjacent , and @xmath41 otherwise . by determinant , rank and signature of @xmath24 we mean the determinant , the rank and the signature of @xmath42 .
we can suppress the weights but indicate the same information by labeling the edges of a coxeter diagram in the following way : if the weight @xmath26 equals @xmath43 , @xmath39 and @xmath40 are joined by an @xmath44-fold edge or a simple edge labeled by @xmath45 ; if @xmath46 , @xmath39 and @xmath40 are joined by a bold edge ; if @xmath47 , @xmath39 and @xmath40 are joined by a dotted edge labeled by @xmath26 ( or without any label ) . by a _
multiple _ edge we mean an edge of weight @xmath43 for @xmath48 . by a _
multi - multiple _ edge we mean an edge of weight @xmath43 for @xmath49 .
an abstract coxeter diagram @xmath24 is _ elliptic _ if @xmath42 is positive definite .
a diagram @xmath24 is _ parabolic _ if any indecomposable component of @xmath42 is degenerate and positive semidefinite ; a connected diagram @xmath24 is a _
diagram if @xmath24 is indefinite but any proper subdiagram of @xmath24 is elliptic ; @xmath24 is _ hyperbolic _ if its negative inertia index is equal to @xmath25 .
the list of connected elliptic and parabolic diagrams with their standard notation is contained in ( * ? ? ?
* tables 1,2 ) .
see also ( * ? ? ?
* table 3 ) for the list of lannr diagrams .
we need the following property of these lists : any lannr diagram of order greater than @xmath11 contains a multiple edge .
it is convenient to describe coxeter polytopes by their coxeter diagrams .
let @xmath3 be a coxeter polytope with facets @xmath50 .
the coxeter diagram @xmath51 of the polytope @xmath3 is a diagram with nodes @xmath52 ; two edges @xmath39 and @xmath40 are not joined if the hyperplanes spanned by @xmath53 and @xmath54 are orthogonal ; @xmath39 and @xmath40 are joined by an edge with weight @xmath55 if @xmath56 , then @xmath42 coincides with the gram matrix of outer unit normals to the facets of @xmath3 ( referring to the standard model of hyperbolic @xmath4-space in @xmath57 ) .
it is shown in @xcite that a coxeter diagram @xmath51 of a compact @xmath4-dimensional hyperbolic polytope @xmath3 is a connected diagram of signature @xmath58 without parabolic subdiagrams . in particular , @xmath51 contains no bold edge , and any indefinite subdiagram is hyperbolic and contains a lannr diagram .
moreover , it is shown there that any compact hyperbolic coxeter @xmath4-polytope @xmath3 is simple , and elliptic subdiagrams of @xmath51 are in one - to - one correspondence with faces of @xmath3 : a @xmath19-face @xmath59 corresponds to an elliptic subdiagram of order @xmath60 whose nodes correspond to the facets of @xmath3 containing @xmath59 .
[ same ] let @xmath51 be a coxeter diagram of a hyperbolic coxeter @xmath4-polytope @xmath3 of finite volume . then no proper subdiagram of @xmath51 is a diagram of a hyperbolic coxeter @xmath4-polytope of finite volume .
we denote by @xmath61 the set of all coxeter diagrams of compact hyperbolic coxeter @xmath4-polytopes , and by @xmath62 the set of coxeter diagrams of compact hyperbolic coxeter @xmath4-polytopes containing no edges of multiplicity greater than @xmath63 .
let @xmath3 be a compact hyperbolic coxeter @xmath4-polytope , and denote by @xmath51 its coxeter diagram .
let @xmath64 be an elliptic subdiagram of @xmath51 .
by ( * ? ? ?
* theorem 3.1 ) , @xmath64 corresponds to a face of @xmath3 of dimension @xmath65 .
denote this face by @xmath66 .
@xmath66 itself is an acute - angled polytope , but it might not be a coxeter polytope .
borcherds proved the following sufficient condition for @xmath66 to be a coxeter polytope .
[ bor ] suppose @xmath3 is a coxeter polytope with diagram @xmath51 , and @xmath67 is an elliptic subdiagram that has no @xmath68 or @xmath69 component .
then @xmath66 itself is a coxeter polytope .
facets of @xmath66 correspond to those nodes that together with @xmath64 comprise an elliptic subdiagram of @xmath51 .
let @xmath70 be a _ neighbor _ of @xmath64 , so that @xmath71 attaches to @xmath64 by some edges .
we call @xmath71 a _ good neighbor _ if @xmath72 is an elliptic diagram , and _ bad _ otherwise .
we denote by @xmath73 the subdiagram of @xmath51 consisting of nodes corresponding to facets of @xmath66 .
the diagram @xmath73 is spanned by good neighbors of @xmath64 and by all vertices not joined with @xmath64 ( in other words , @xmath74 is spanned by all nodes of @xmath75 except bad neighbors of @xmath64 ) .
if @xmath66 is a coxeter polytope , denote its coxeter diagram by @xmath76 . in ( * ?
* theorem 2.2 ) , allcock gives a receipt how to compute the dihedral angles of @xmath66 .
we need the following partial case of this result : [ cor_all ] suppose that @xmath66 is a coxeter polytope .
if @xmath64 has no good neighbors then @xmath77 . in particular , this always holds for @xmath78 where @xmath49 .
proposition [ cor_all ] allows us to identify diagrams @xmath73 and @xmath76 for @xmath64 being multi - multiple edge .
in this section we recall the technique of local determinants derived in @xcite and prove several properties of determinants of hyperbolic coxeter diagrams .
let @xmath24 be a coxeter diagram , and let @xmath79 be a subdiagram of @xmath24 such that @xmath80 .
a _ local determinant _ of @xmath24 on a subdiagram @xmath79 is @xmath81 the following result is the main tool in the consideration of multiplicities of edges .
[ loc_sum ] if a coxeter diagram @xmath24 consists of subdiagrams @xmath82 having a unique vertex @xmath83 in common , and no vertex of @xmath84 attaches to @xmath85 for @xmath86 , then @xmath87 now we prove several properties of determinants of hyperbolic diagrams . the first one concerns monotonicity .
[ monoton ] let @xmath88 be an elliptic diagram and @xmath89 be any node such that @xmath90 is hyperbolic .
then @xmath91 is a strictly decreasing function on label of each edge joining @xmath92 with @xmath88 .
moreover , @xmath91 is a strictly decreasing function on label of each edge @xmath93 such that @xmath94 is an elliptic diagram .
denote @xmath95 , @xmath96 .
let @xmath97 be the negative of the label of the edge @xmath98 , @xmath99 , so that @xmath100 .
we will prove that @xmath91 is a strictly decreasing function on label @xmath101 of the edge @xmath93 .
consider a standard expanding of @xmath91 into @xmath102 summands @xmath103 ( each obtained as a product of @xmath4 matrix elements ) .
then @xmath104 it is easy to see that @xmath105 ( since @xmath88 is an elliptic diagram ) .
recall that @xmath106 as a label of an edge of a coxeter diagram , so it is sufficient to prove that @xmath107 .
each of @xmath102 summands of @xmath91 can be written as @xmath108 where @xmath109 are products of all the weights of edges of pairwise disjoint cycles in @xmath90 ( following @xcite , we call @xmath110 _ cyclic products _ ) , @xmath111 , and @xmath112 notice that @xmath113 for @xmath86 implies @xmath114 .
if a summand @xmath103 contributes to a linear part of @xmath115 then @xmath103 contains exactly one of @xmath116 and @xmath117 .
we will prove that the sum of the terms containing @xmath116 or @xmath117 is non - positive .
let @xmath118 be a cyclic product containing @xmath116 , denote by @xmath119 the sum of all terms @xmath103 containing @xmath120 .
it is easy to see that @xmath121 where @xmath122 .
since @xmath120 contains @xmath116 or @xmath117 , @xmath123 contains both @xmath124 and @xmath125 , which imply that the diagram @xmath126 is elliptic , so @xmath127 .
hence , @xmath128 .
since at most one of @xmath110 in the decomposition ( @xmath129 ) contains @xmath116 or @xmath117 , each of @xmath103 contributes to at most one of @xmath130 , so , @xmath131 which is negative . in view of the inequality
@xmath106 we get @xmath107 , and the first statement of the lemma is proved .
now one can notice that we used only the fact that @xmath132 is an elliptic diagram , so the second statement holds by the same argument .
the next two lemmas provide upper bounds for determinants of some hyperbolic diagrams .
[ separate1 ] there exists a constant @xmath133 which depends only on the dimension @xmath4 such that for any diagram @xmath90 satisfying * @xmath88 is an elliptic diagram of order @xmath4 , @xmath90 is hyperbolic ; * either @xmath134 , or @xmath88 contains no multi - multiple edges ; * @xmath135 contains no dotted edges ; holds @xmath136 denote by @xmath137 the diagram obtained from @xmath90 by replacing all the edges joining @xmath92 with @xmath88 with labels higher than @xmath138 by edges labeled by @xmath138 .
it is easy to see that each hyperbolic subdiagram of @xmath90 remains hyperbolic after this change , and each elliptic subdiagram remains elliptic , so @xmath139 . by lemma [ monoton ]
, @xmath91 is a decreasing function on a multiplicity of each edge joining @xmath92 with @xmath88 .
hence , @xmath140 .
if @xmath88 contains no multi - multiple edges , then given a dimension @xmath4 there are finitely many possibilities for @xmath137 ( since @xmath141 ) , so we may take for @xmath133 any positive number smaller than minimum of @xmath142 .
if @xmath88 is a diagram of order @xmath11 or @xmath12 containing multi - multiple edges , then the second statement of lemma [ monoton ] implies that @xmath91 is a decreasing function on a multiplicity of each edge of @xmath90 ( here we use the third assumption of the lemma ) .
we replace each multi - multiple edge of @xmath90 with label higher than @xmath138 by an edge labeled by @xmath138 , denote the obtained diagram by @xmath143 and obtain finitely many possibilities for @xmath143 , which finishes the proof .
denote by @xmath144 a lannr diagram of order @xmath12 containing subdiagrams of types @xmath145 , @xmath146 and @xmath147 ( see fig .
[ lanner ] ) .
let @xmath83 be the vertex of @xmath144 that does not belong to @xmath147 .
denote by @xmath148 the local determinant @xmath149 .
it is easy to compute ( see e.g. @xcite ) that @xmath150 the explicit expression for @xmath148 shows that @xmath151 is an increasing function on each of @xmath152 tending to infinity while @xmath153 tends to infinity . in particular
, one can check that @xmath154 .
the next lemma shows that any @xmath155 contains a certain hyperbolic subdiagram with determinant of relatively large absolute value .
[ separate ] for each dimension @xmath156 and positive integer @xmath157 there exists a constant @xmath158 depending on @xmath4 and @xmath157 such that for any coxeter diagram @xmath159 there exists an elliptic subdiagram @xmath160 of order @xmath4 and a node @xmath92 joined with @xmath88 by at most one dotted edge satisfying @xmath161 moreover , if @xmath162 or @xmath163 , then @xmath164 does not depend on @xmath157 , and the subdiagram @xmath90 may be chosen in such a way that for some @xmath165 the subdiagram @xmath166 contains no dotted edges and @xmath167 contains no edges at all .
@xmath168we prove the first statement of the lemma separately for @xmath11-dimensional case , @xmath12-dimensional case , and the case of bounded multiplicity .
let @xmath3 be a polytope with @xmath56 .
* case 1 : * @xmath169 .
we consider 3 cases : @xmath3 is either a triangle , or a right - angled polygon , or a polygon with more than @xmath12 sides and at least one strictly acute angle .
if @xmath3 is a coxeter triangle , then @xmath170 is a decreasing function on the labels of all edges of @xmath24 ( see lemma [ monoton ] ) , and it is easy to check that @xmath171 suppose now that @xmath3 has at least @xmath9 sides .
suppose that @xmath3 is not right - angled , and @xmath172 and @xmath173 are its consecutive sides such that @xmath174 .
denote by @xmath175 the nodes of @xmath24 corresponding to @xmath176 respectively .
we will prove that either @xmath177 or @xmath178 .
since @xmath3 has at least @xmath9 sides , the line containing @xmath179 does not intersect the line containing @xmath180 .
let @xmath181 be a unique common perpendicular to @xmath179 and @xmath180 ( it is easy to see that @xmath182 intersects the sides @xmath179 and @xmath180 themselves but not the prolongations ) . since @xmath174 , @xmath181 does not coincide with @xmath183 .
let @xmath184 be a common perpendicular to @xmath181 and @xmath183 ( it may coincide with @xmath180 ) .
then the lines @xmath185 and @xmath181 bound a 3 quadrilateral ( i.e. a quadrilateral with @xmath12 right angles ) , see fig .
[ lambert].a .
denote by @xmath186 and @xmath187 the lengths of its sides contained in @xmath181 and @xmath184 respectively .
then ( see e.g. ( * ? ? ? * theorem 7.17.1 ) ) @xmath188 therefore , either @xmath189 or @xmath190 .
notice that the distance @xmath191 between @xmath179 and @xmath180 is not less than @xmath186 , and the distance @xmath192 between @xmath183 and @xmath173 is not less than @xmath187 .
so , either @xmath193 or @xmath194 . in the former case
@xmath195 similarly , in the latter case , @xmath196 suppose now that @xmath3 is a right - angled polygon .
let @xmath197 be its consecutive sides and let @xmath198 ( see fig .
[ lambert].b ) .
let @xmath181 be a line through @xmath199 orthogonal to @xmath180 and @xmath184 be a line through @xmath199 orthogonal to @xmath200 ( notice that since @xmath3 is right - angled @xmath201 , so @xmath202 ) .
then the lines @xmath172 and @xmath181 bound a lambert quadrilateral , as well as the lines @xmath203 and @xmath184 do . furthermore , at least one of the angles @xmath204 and @xmath205 does not exceed @xmath206 ( since @xmath207 ) . without loss of generality
we may assume that @xmath208 .
then @xmath209 as it is shown above for the case of non- right - angled polygon , this implies that either @xmath210 or @xmath211 .
* case 2 : * @xmath212 .
first suppose that @xmath3 is a right - angled @xmath12-polytope .
then all @xmath11-faces of @xmath3 are right - angled polygons , and as it is shown in case 1 , each of them has at least one edge with length @xmath213 satisfying @xmath214 .
let @xmath215 be such a long edge , @xmath216 and @xmath217 be the facets of @xmath3 such that @xmath218 , and @xmath219 and @xmath220 be the remaining facets containing the vertices of @xmath215 .
clearly , we have @xmath221 for the distance between @xmath216 and @xmath217 .
denote by @xmath175 the nodes of @xmath24 corresponding to @xmath222 and @xmath220 respectively .
then @xmath223 suppose now that @xmath3 is not right - angled .
let @xmath216 and @xmath217 be the facets composing an acute dihedral angle , denote @xmath224 the corresponding edge , and let @xmath225 be any vertex of @xmath226
. denote by @xmath219 a unique facets of @xmath3 such that @xmath227 .
then at least one of the angles formed by @xmath219 with @xmath216 and @xmath217 is right .
assume that @xmath228 , and denote @xmath229 .
let @xmath230 be the other vertex of @xmath215 , and @xmath220 be the facet such that @xmath231 .
if @xmath232 then @xmath3 is one of the nine @xmath12-dimensional hyperbolic coxeter tetrahedra , so @xmath233 for some constant @xmath234 .
suppose that @xmath217 does not intersect @xmath220 .
then we have @xmath235 , @xmath236 , and @xmath116 is not equal to @xmath237 , i.e. is not greater than @xmath238 . expanding the determinant @xmath239 and taking into account that @xmath235 and @xmath113 for @xmath86 , we get @xmath240 if @xmath241 , then @xmath242 and @xmath243 .
if @xmath244 , then @xmath245 and @xmath246 .
* case 3 : * @xmath247 .
first suppose that @xmath3 is not right - angled , so @xmath24 contains at least one non - dotted edge , say @xmath248 .
then there exists an elliptic subdiagram @xmath249 containing @xmath248 .
a subdiagram @xmath250 defines an edge ( @xmath25-face ) of @xmath3 , so it has one more good neighbor @xmath92 in @xmath24 besides @xmath124 . by lemma [ monoton ] , @xmath170 is a decreasing function on label of each edge joining @xmath92 with @xmath88 .
notice that @xmath251 is the only pair of nodes in @xmath135 which can be joined by a dotted edge .
if it is , replace @xmath252 by the edge labeled by @xmath138 and denote the obtained diagram by @xmath137 , otherwise let @xmath253 . since @xmath93 is a non - empty edge , the diagram @xmath137 is also hyperbolic .
so , @xmath139 and @xmath254 .
since @xmath255 , there are finitely many possibilities for @xmath137 , so we may take as @xmath158 any positive number smaller than minimum of @xmath142 . now suppose that @xmath3 is right - angled .
any @xmath11-face of @xmath3 is right - angled polygon , so , by case 1 , it has an edge @xmath256 of length @xmath257 .
denote by @xmath258 the facets containing @xmath225 , we may assume that @xmath230 is the intersection of facets @xmath259 .
let @xmath260 be the corresponding nodes of @xmath24 .
then the diagram @xmath261 contains a unique edge @xmath262 , and @xmath263 .
so , the first statement of the lemma is proved .
the second statement follows immediately from the choice of @xmath90 and the constant @xmath264 in low - dimensional cases .
the following lemma provides an upper bound for a local determinant of hyperbolic diagram of special type .
later on , we find such a subdiagram in any coxeter diagram of a polytope from @xmath7 , and use lemma [ local ] to get a bound for a multiplicity of an edge .
[ local ] given @xmath265 , integer @xmath266 and dimension @xmath156 there exists a constant @xmath267 such that for any diagram @xmath268 satisfying the following five conditions * @xmath88 is an elliptic subdiagram of order @xmath4 , while @xmath90 and @xmath268 are hyperbolic diagrams ; * @xmath92 is joined with @xmath88 by at most one dotted edge ; * if @xmath269 and @xmath270 is a dotted edge , then labels of edges of @xmath167 do not exceed @xmath157 ; * @xmath271 is not joined with @xmath90 by any dotted edge ; * @xmath272 ; holds @xmath273 the inequality @xmath274 follows from the fact that both @xmath275 and @xmath135 are hyperbolic , so their determinants are non - positive .
suppose that the diagram @xmath268 contains no dotted edges .
then we have @xmath276 since each of the @xmath277 summands in the standard expansion can not exceed @xmath25 .
so , @xmath278 .
now , suppose that @xmath268 contains a unique dotted edge @xmath279 and denote by @xmath213 the label of this edge .
then @xmath280 where @xmath281 moreover , @xmath282 , at the same time @xmath283 is an elliptic diagram containing no edges labeled by @xmath284 , so we obtain @xmath285 for some constant @xmath286 .
we may assume @xmath286 to be less than @xmath25 .
therefore , for @xmath287 we have @xmath288 for @xmath289
we have @xmath290
so @xmath291 does not exceed maximum of these two numbers for any @xmath213 . at the end of the section ,
we prove several elementary facts about coxeter diagrams of polytopes belonging to @xmath7 .
[ dplus2 ] let @xmath292 , and let @xmath24 be a coxeter diagram of @xmath3 .
then for any elliptic subdiagram @xmath293 of order @xmath4 there exist nodes @xmath294 such that the diagram @xmath268 contains no dotted edges .
denote by @xmath295 the set of all nodes of @xmath296 which are not joined with @xmath88 by a dotted edge .
denote by @xmath297 the number of nodes in @xmath295 , and by @xmath298 the number of the remaining nodes in @xmath296 .
clearly , @xmath299 now suppose that any two elements of @xmath295 are joined by a dotted edge .
then , using that @xmath300 , we may write down the following inequality for @xmath6 : @xmath301 the contradiction shows that there are at least two nodes @xmath302 which are not joined by a dotted edge .
therefore , @xmath268 contains no dotted edges .
the same reasoning as above gives rise to the following statement .
[ dplus11 ] let @xmath292 , and let @xmath24 be a coxeter diagram of @xmath3 .
then for any elliptic subdiagram @xmath293 of order @xmath4 and any @xmath303 such that @xmath90 contains no dotted edges there exists a node @xmath304 such that the diagram @xmath268 contains no dotted edges .
[ dots ] suppose that @xmath292 and @xmath67 is an elliptic diagram of the type @xmath305 , @xmath306 . if each neighbor of @xmath64 is joined with @xmath307 by at least one dotted edge then * @xmath308 ( where @xmath309 and @xmath310 are the dimension of @xmath66 and the number of nodes in @xmath76 respectively ) ; * for any elliptic subdiagram @xmath311 of order @xmath312 there exists a node @xmath313 which is not joined with @xmath88 by a dotted edge .
moreover , there exists a neighbor @xmath271 of @xmath64 such that the diagram @xmath268 contains no dotted edges .
denote by @xmath314 the number of neighbors of @xmath64 .
denote by @xmath315 the number of dotted edges in @xmath76 . by assumption , each neighbor of @xmath64 is joined with @xmath307 by at least one dotted edge , so there are at least @xmath314 dotted edges `` outside '' of @xmath307 .
taking into account that @xmath316 , we have @xmath317 on the other hand , @xmath318 , so @xmath319 .
thus , we get @xmath320 and the first statement is proved .
the proof of the second statement is similar to the proof of lemma [ dplus2 ] .
denote by @xmath295 the set of all nodes of @xmath321 which are not joined with @xmath88 by a dotted edge .
denote @xmath322 , and denote by @xmath323 the number of the remaining nodes in @xmath321 . by lemma [ dplus2 ] , @xmath300 .
now suppose that each neighbor of @xmath64 is joined with each node of @xmath295 by a dotted edge
. then there are @xmath324 dotted edges joining the neighbors of @xmath64 with the nodes of @xmath295 .
therefore , @xmath325 in contradiction to the assumption .
thus , there exist a neighbor @xmath271 of @xmath64 and @xmath326 which are not joined by a dotted edge , so the lemma is proved .
in this section we prove that for each @xmath4 there is a lower bound for dihedral angles of polytopes from @xmath327 .
we start from low - dimensional case .
[ no_multi ] let @xmath3 be a coxeter @xmath4-polytope , and let @xmath67 be a subdiagram of the type @xmath305 , @xmath306 .
suppose that either @xmath328 , or @xmath329 , or @xmath307 contains no multi - multiple edges .
there exists a constant @xmath330 depending on @xmath4 only such that @xmath331 ( i.e. , @xmath332 ) .
consider the diagram @xmath307 . by assumptions
, we may apply lemma [ separate ] with @xmath333 to @xmath66 .
thus , there exists an elliptic subdiagram @xmath311 of order @xmath312 and a node @xmath334 such that @xmath335 for some constant @xmath234 depending only on @xmath312 , and @xmath90 contains at most one dotted edge .
suppose that there exists a neighbor @xmath271 of @xmath64 such that @xmath271 is not joined with @xmath90 by a dotted edge .
then the diagram @xmath268 fits into all the conditions of lemma [ local ] ( with @xmath333 and @xmath336 ) , so there exists a constant @xmath337 ( depending on @xmath4 only ) such that @xmath338 . by proposition [ loc_sum ] , @xmath339
since the diagram @xmath340 contains @xmath341 nodes , @xmath342 and hence , @xmath343 and @xmath344 since @xmath345 tends to ( negative ) infinity while @xmath19 tends to infinity , we obtain some constant @xmath330 such that @xmath332 .
suppose now that each neighbor of @xmath64 is joined with @xmath90 by a dotted edge .
then we are in assumptions of lemma [ dots ] , so there exists a node @xmath346 and a neighbor @xmath347 of @xmath64 such that @xmath348 contains no dotted edges .
now we can apply lemma [ separate1 ] to @xmath349 , which implies that there exists a constant @xmath350 depending on @xmath4 only such that @xmath351 .
therefore , @xmath348 fits into assumptions of lemma [ local ] with @xmath333 and @xmath352 , so we can proceed with the diagram @xmath348 in the same way as we did in the preceding paragraph with @xmath268 .
this finishes the proof .
now we are able to prove the main result of the section .
the following theorem provides a uniform upper bound for multiplicity of edges of coxeter diagrams @xmath51 for @xmath353 .
[ multi ] if @xmath354 , then there exists a constant @xmath355 such that for any @xmath353 the coxeter diagram @xmath51 belongs to @xmath356 .
the proof is by induction on @xmath4 .
the base consists of the cases @xmath328 and @xmath329 , for which we refer to lemma [ no_multi ] .
suppose that @xmath357 and for each @xmath358 the theorem is already proved , i.e. for any @xmath359 there is a constant @xmath360 such that for any @xmath361 and for any @xmath362 of type @xmath305 we have @xmath363 .
now take any subdiagram @xmath67 of type @xmath305 and consider the diagram @xmath307 .
we may assume that @xmath306 , in particular @xmath364 .
if @xmath307 contains no multi - multiple edges , we refer to lemma [ no_multi ] .
suppose that @xmath307 contains at least one multi - multiple edge
. then choose any elliptic subdiagram @xmath311 of order @xmath312 containing a multi - multiple edge .
we can write @xmath88 as @xmath365 , @xmath366 , where each of @xmath367 is a subdiagram of order @xmath11 consisting of a multi - multiple edge , and @xmath368 is either a subdiagram containing no multi - multipe edges or a subdiagram of order @xmath11 or @xmath12 ( such a decomposition of @xmath88 may not be unique ) .
denote @xmath369 , and consider @xmath370 ( so , @xmath79 consists of all nodes of @xmath307 not joined with @xmath371 ) .
this is a coxeter diagram of a @xmath372-polytope , where @xmath373 .
we may assume that either @xmath374 or @xmath79 contains no multi - multiple edges , otherwise we may add a multi - multiple edge ( denote it by @xmath375 ) to @xmath376 to get @xmath377 , and consider @xmath378 . by lemma [ separate ] , there exists a constant @xmath234 ( depending on @xmath4 only ) , an elliptic subdiagram @xmath379 of order @xmath372 and a node @xmath380 such that @xmath381 .
first , suppose that there exists a neighbor @xmath271 of @xmath64 not joined with @xmath382 by a dotted edge .
then , applying proposition [ loc_sum ] , we get @xmath383 notice that @xmath384 while each of @xmath385 either equals to 1 ( if @xmath386 is not joined with @xmath271 ) or is negative ( otherwise ) . therefore , @xmath387 for @xmath388 , and we obtain @xmath389 at the same time , by lemma [ local ] applied to @xmath390 with @xmath333 and @xmath336 , the local determinant @xmath391 is bounded by some constant @xmath337 ( depending on @xmath4 only ) .
now consider the diagram @xmath340 . by construction , @xmath392 , so @xmath393 is not connected , and we use proposition [ loc_sum ] to obtain @xmath394 since the diagram @xmath340 consists of @xmath341 nodes , we get @xmath395 which implies @xmath344 since @xmath345 tends to ( negative ) infinity while @xmath19 tends to infinity , we obtain some constant @xmath396 such that @xmath397 . we are left to consider the case when each neighbor of @xmath64 is joined with @xmath398 by a dotted edge , suppose now that this holds .
let @xmath88 be any elliptic subdiagram of @xmath307 . according to the second statement of lemma [ dots ] ,
there exists a node @xmath399 and a neighbor @xmath347 of @xmath400 such that the diagram @xmath348 contains no dotted edges . by the first statement of lemma [ dots ]
, we have @xmath401 .
we may apply the induction assumption to get @xmath402 .
in particular , the subdiagram @xmath349 contains no edges labeled by @xmath403 .
so , by lemma [ separate ] , @xmath404 for some @xmath350 depending on @xmath312 only , and by lemma [ local ] there exists a constant @xmath405 such that @xmath406 .
this implies that @xmath407 and hence , we obtain some constant @xmath408 such that @xmath409 . now , taking as @xmath355 the maximum of @xmath396 and @xmath408 , we complete the proof .
in the previous section we proved that multiplicity of edges of coxeter diagrams of polytopes from @xmath327 is bounded .
the current goal is to prove that the number of facets of polytopes from @xmath327 is also bounded by some constant depending on @xmath4 only .
this will imply finiteness of @xmath327 .
first , we estimate the order of subdiagrams without dotted edges .
[ n_0 ] there exists a constant @xmath410 such that for any diagram @xmath411 satisfying * @xmath412 for some diagram @xmath413 ; * @xmath411 contains an elliptic subdiagram @xmath88 of order @xmath4 ; * @xmath411 contains no dotted edges ; holds @xmath414 .
assume that @xmath415 .
then there exists @xmath416 .
since @xmath417 , multiplicity of each edge of @xmath90 is uniformly bounded , so the number of possible configurations @xmath90 is finite , denote the number of them by @xmath418 .
recall that the signature of @xmath90 is @xmath58 .
this implies that the corresponding vectors in @xmath419 ( denote them @xmath260 ) form a basis , so for any @xmath420 the corresponding vector @xmath421 is completely determined by the weights of the edges joining @xmath271 with @xmath90 .
again , since @xmath417 , for each @xmath90 there is at most @xmath422 ways to join @xmath271 with @xmath90 .
thus , for each of finite number of configurations of vectors ( @xmath260 ) there are finitely many vectors @xmath83 which may correspond to nodes of @xmath423 , more precisely the number is bounded by @xmath424 . adding to this number the order of @xmath90 , we get the bound @xmath410 . to illustrate application of lemma [ n_0 ] we provide a very rough bound for @xmath9-dimensional case .
let @xmath425 .
we will estimate the number of facets of @xmath3 .
first , compute the number of distinct elliptic diagrams @xmath88 of order @xmath9 .
there are @xmath10 connected ones ( @xmath426 , @xmath427 , @xmath428 , @xmath429 , and @xmath430 ) , @xmath12 non - connected ones with connected component of order @xmath12 ( direct sum of @xmath431 and one of @xmath432 , @xmath433 , @xmath434 ) , and @xmath435 diagrams of type @xmath436 , @xmath437 .
so , the number of elliptic diagrams is @xmath438 .
next , compute @xmath439 .
a node @xmath440 may be attached to an elliptic diagram @xmath88 in @xmath441 ways ( this bound is very rough : we forget about huge number of symmetries ; moreover , a large part of these configurations gives rise to elliptic or parabolic diagrams ) .
so , @xmath442 .
now , any node @xmath443 may be attached to @xmath444 in at most @xmath445 ways ( again , this bound is very rough ) .
hence , we may estimate that @xmath446 , where @xmath447 [ n_1 ] there exists a constant @xmath448 such that for any diagram @xmath411 satisfying * @xmath412 for some diagram @xmath413 ; * @xmath411 contains an elliptic subdiagram @xmath88 of order @xmath4 and a node @xmath89 such that * * @xmath90 contains no dotted edges ; * * at most one node of @xmath449 is joined with @xmath90 by a dotted edge ; holds @xmath450 .
[ lin ] before proving lemma [ n_1 ] we recall some elementary fact from linear algebra .
let @xmath451 be a non - degenerate quadratic space , and let @xmath452 be a basis .
then a system of equations @xmath453 has at most two solutions . in other words ,
any vector is determined ( up to some symmetry ) by its scalar products with all but one elements of basis of @xmath225 .
we may assume that there exist @xmath420 and @xmath454 joined by a dotted edge , otherwise we take @xmath455 ( see lemma [ n_0 ] ) .
now the proof is similar to the proof of lemma [ n_0 ] .
there are finitely many diagrams @xmath90 , and finitely many ways to join @xmath271 with @xmath456 . in view of remark [ lin ] ,
the weight of the edge @xmath457 is determined by the multiplicities of the edges joining @xmath271 with @xmath456 .
thus , we get some bound for the number of nodes in @xmath411 .
more precisely , we may take a rough bound @xmath458 where the multiple @xmath459@xmath460 is the number of ways to choose @xmath461 in @xmath90 and @xmath418 is the estimate for the number of possible configurations @xmath90 ( see proof of lemma [ n_0 ] ) .
[ n ] for any @xmath354 there exists a constant @xmath462 such that any polytope from @xmath463 has at most @xmath462 facets .
let @xmath464 and @xmath56 , i.e. @xmath465 , where @xmath466 and @xmath6 is the number of dotted edges in @xmath24 .
by lemma [ multi ] , @xmath467 for @xmath468 in notations of lemma [ multi ] .
let @xmath88 be any elliptic subdiagram of @xmath24 of order @xmath4 .
by lemma [ dplus2 ] , there exist @xmath469 such that the diagram @xmath268 contains no dotted edges .
let @xmath470 be the set of nodes of @xmath471 joined with @xmath268 by at most one dotted edge . by lemma [ n_1 ] , @xmath472 .
denote by @xmath473 the set of nodes of @xmath24 joined with @xmath90 by two or more dotted edges . clearly , @xmath474 .
furthermore , @xmath475 since @xmath465 , we have @xmath476 , which implies @xmath477 [ finite ] the number of polytopes in @xmath463 is finite .
let @xmath464 and @xmath56 .
by lemma [ n ] , the number of nodes in @xmath24 is bounded by some constant @xmath462 depending on @xmath4 only . according to lemma [ multi ] , @xmath478 .
clearly , the number of coxeter diagrams with bounded number of nodes and bounded multiplicities of edges is finite ( we do not write weights of dotted edges ) . by andreev s theorem @xcite
each of the possible diagrams corresponds to at most one coxeter polytope .
so , there are finitely many coxeter diagrams that may occur to be diagrams of polytopes from @xmath463 , which implies that @xmath463 is finite .
combining corollary [ finite ] with the result of vinberg @xcite stating that the dimension of compact hyperbolic coxeter polytope does not exceed @xmath23 , we obtain the following theorem .
[ main ] the set @xmath7 is finite .
in section [ fin ] we proved that @xmath463 is finite .
since we have a bound on the number of facets of any polytope @xmath479 , we are limited to finitely many combinatorial types of polytopes . given a combinatorial type , we have finitely many possibilities to assign weights to non - dotted edges of @xmath56 ( due to lemma [ multi ] ) .
however , the possible labels of dotted edges remain undefined . in particular , given a combinatorial type of @xmath4-polytope ( satisfying the condition @xmath465 ) we even can not still check whether there exists a compact coxeter polytope of this combinatorial type . in this section
we describe an algorithm which allows us to determine the labels of the dotted edges one after another . using the algorithm
one can check if the given combinatorial type is realizable in @xmath463 , as well as to find all possible realizations in this class .
the main idea of the algorithm is the following .
first , for any polytope @xmath479 and its coxeter diagram @xmath51 ( with unknown labels of dotted edges ) we show the way to find out the weights of dotted edges . for this
we point out a sequence of subdiagrams @xmath480 together with a way to determine weights of dotted edges in @xmath481 by the weights of dotted edges of @xmath482 .
next , we use results of the previous two sections to show that for any @xmath483 there are finitely many ways only to assign weights to @xmath481 .
then , starting from all possible elliptic diagrams @xmath484 and adding vertices one by one , we get a ( huge ) list of diagrams which are candidates to be subdiagrams of @xmath51 for @xmath479 . at each step
we check the signature of every diagram and preserve only ones with correct signature ( which is @xmath58 ) . for diagrams with correct one we also check whether it is already a diagram of a polytope .
if at some step @xmath485 the set of diagrams @xmath481 is empty , the procedure is done .
this can not happen later than at step number @xmath486 . in this way we get coxeter diagrams of all the polytopes from @xmath463 .
[ alg_p ] given a polytope @xmath479 and its coxeter diagram @xmath56 with unknown weights of dotted edges , there exists a finite algorithm which provides the weights of dotted edges of @xmath24
. let @xmath487 be a node incident to the maximal number of dotted edges ( in particular , this means that any node of @xmath24 other than @xmath488 is incident to at most @xmath489 dotted edges ) .
let @xmath490 be any elliptic subdiagram of order @xmath4 containing @xmath488 ( i.e. , @xmath64 is a diagram of some vertex @xmath491 of @xmath3 , and @xmath491 belongs to facet @xmath492 corresponding to @xmath488 ) .
define @xmath493 .
as we have explained above , for each positive integer @xmath494 we will construct a subdiagram @xmath495 of order @xmath496 such that in each @xmath481 all the weights are determined .
* step 1 : attaching nodes . * by lemma [ dplus2 ]
, there exist @xmath497 such that @xmath498 contains no dotted edges .
define @xmath499 , @xmath500 .
denote by @xmath501 the set of nodes of @xmath502 joined with @xmath32 by at most one dotted edge , and let @xmath503 . according to remark [ lin ]
, we can find the label of the dotted edge ( if any ) in @xmath504 , so we can find the vector in @xmath419 corresponding to @xmath461 as a linear combination of the vectors corresponding to @xmath505 , which allows us to find all the labels of the edges joining @xmath461 with @xmath32 .
hence , we can take @xmath506 . in the same way
, we can attach to @xmath507 any other element of @xmath501 to get @xmath508 .
indeed , we find the corresponding vector in @xmath419 as a linear combination of the vectors corresponding to @xmath505 , and then compute scalar products with all the vectors corresponding to the remaining nodes @xmath509 of @xmath507 . applying this procedure for each node from @xmath501
, we get a subdiagram @xmath510 , where @xmath511 , @xmath512 .
each of the remaining nodes is joined with @xmath499 by at least two dotted edges .
since @xmath465 , this implies that there are at most @xmath513 nodes in @xmath514 . in other words , to this moment @xmath515 consists of @xmath341 nodes of @xmath498 and at least @xmath516 attached nodes , so @xmath517 . if @xmath518 then the lemma is proved , otherwise we go to the next step .
* step 2 : walking along edges of @xmath3 . *
let @xmath519 , and suppose that we have already constructed a subdiagram @xmath520 .
let @xmath521 be any diagram of a vertex ( i.e. , an elliptic diagram of order @xmath4 ) , and let @xmath522 be a node such that @xmath523 is a complete diagram of an edge . * claim : * it is possible to attach @xmath71 to @xmath520 to get @xmath524 .
let @xmath525 be any node of @xmath526 .
since @xmath527 is a complete diagram of an edge , @xmath71 can be joined with @xmath88 by at most one dotted edge .
suppose that @xmath71 is not joined with @xmath88 by a dotted edge .
take any @xmath528 and notice that @xmath349 has signature @xmath58 , so it corresponds to a basis of @xmath419 .
therefore , we are able to find the corresponding vector in @xmath419 as a linear combination of the vectors corresponding to @xmath349 , and then compute scalar products with all the vectors corresponding to the remaining nodes of @xmath520 , so @xmath529 .
now suppose that @xmath71 is joined with @xmath88 by a dotted edge .
if @xmath71 is also joined by a dotted edge with each of at least @xmath530 nodes of @xmath526 , then there are at least @xmath531 dotted edges incident to @xmath71 , which contradicts the choice of @xmath488 .
thus , @xmath71 is not joined by a dotted edge with some node @xmath528 .
therefore , @xmath71 is joined with @xmath349 by exactly one dotted edge , so we can attach @xmath71 to @xmath349 and then to @xmath520 to get @xmath524 . according to the claim above , for any diagram @xmath532 of a vertex @xmath533
contained in @xmath520 , @xmath519 , we can attach to @xmath520 any complete diagram of edge emanating from @xmath533 . walking along the edges of @xmath3 we can pass from @xmath534 to any other vertex of @xmath3 .
therefore , for each elliptic diagram @xmath293 of order @xmath4 there exists some number @xmath535 such that @xmath536 . collecting diagrams of all the vertices of @xmath3 one by one ,
we obtain @xmath537 .
[ alg_fix ] given abstract coxeter diagram @xmath24 with unknown weights of dotted edges , and an integer @xmath354 satisfying @xmath465 , where @xmath6 is the number of dotted edges and @xmath466 , there exists a finite algorithm which * verifies if there exists @xmath479 such that @xmath56 ; * provides the weights of dotted edges if @xmath56 .
let @xmath24 be an abstract coxeter diagram with unknown weights of dotted edges . according to ( * ? ? ?
* theorem 2.1 and proposition 4.2 ) , @xmath538 if and only if its signature equals @xmath58 and the poset of elliptic subdiagrams of @xmath24 coincides with a poset of faces of some compact euclidean @xmath4-polytope .
so , the following conditions on @xmath24 are essential : * the order of maximal elliptic subdiagram of @xmath24 equals @xmath4 ; * each elliptic diagram of order @xmath539 is contained in exactly two elliptic diagrams of order @xmath4 ( in terms of faces of @xmath3 , each edge has exactly two vertices ) .
both conditions can be easily verified without knowing weights of dotted edges .
if any of them does not hold then @xmath24 is not a diagram of a compact polytope .
otherwise , we need to check , whether it is possible to assign weights to the dotted edges of @xmath24 to get signature @xmath58 .
now we proceed as in the algorithm constructed in lemma [ alg_p ] .
the only difference is in some additional computations .
suppose that we have already constructed a subdiagram @xmath540 , @xmath541 .
this means that we know a configuration of @xmath542 vectors in @xmath419 with gram matrix corresponding to @xmath543 .
while attaching new node @xmath71 by expressing the corresponding vector as a linear combination of vectors corresponding to some subdiagram @xmath544 of order @xmath545 , we need to compute scalar products of this vector with all the preceding @xmath542 vectors , including ones corresponding to nodes joined with @xmath71 by non - dotted edges .
if all the scalar products coincide with ones prescribed by @xmath24 , then we add the node to obtain @xmath546 and a configuration of @xmath547 vectors in @xmath419 , otherwise we stop the process and deduce that @xmath24 is not a diagram of a polytope .
[ alg ] there exists a finite algorithm listing coxeter diagrams of all the polytopes from @xmath327 .
general algorithm . * in fact
, lemma [ alg_fix ] provides a way to classify all the polytopes @xmath353 . indeed , according to theorem [ multi ]
, @xmath51 does not contain edges of multiplicity greater than @xmath548 , and by lemma [ n ] , @xmath549 .
the number of coxeter diagrams ( without labels of dotted edges ) is finite , and for each of them we can find out whether there exists a corresponding polytope @xmath353 .
however , the number of diagrams we need to check in this way is extremely large even for small @xmath355 and @xmath550 .
below we provide a shorter algorithm which becomes realizable for reasonable @xmath355 .
an example of application to @xmath328 ( with some further refinements ) is presented in section [ dim4 ] .
* 2 . reduced algorithm .
* we proceed as in the algorithm constructed in lemma [ alg_p ] by listing all possible diagrams @xmath481 on each step .
first , we list all possible diagrams @xmath493 . according to theorem [ multi ]
, the multiplicity of edges of @xmath484 does not exceed @xmath548 , so we can list all elliptic diagrams of order @xmath4 . denote the list of all obtained diagrams by @xmath551 .
then we construct the list @xmath552 of diagrams @xmath553 without dotted edges . after constructing any diagram @xmath481 we immediately check whether @xmath481 is already a diagram of a polytope . by proposition [ same ] ,
a coxeter diagram of @xmath4-polytope can not contain a diagram of another @xmath4-polytope as a proper subdiagram .
therefore , if we get a diagram @xmath481 of @xmath4-polytope , we put it in the resulting list of polytopes and exclude from further considerations . by lemma [ dplus2 ] , there exists @xmath554 without dotted edges .
in particular , the diagram @xmath555 should appear in @xmath552 .
therefore , we do the following : for each pair @xmath556 of ( possibly isomorphic ) diagrams from @xmath552 constructed by the same @xmath484 ( i.e. @xmath553 , @xmath557 ) we compose a diagram @xmath558 and compute the weight of the edge @xmath559 from the equation @xmath560 .
we put @xmath32 in the list @xmath561 if and only if the weight equals @xmath562 for some positive integer @xmath563 .
any @xmath51 contains some diagram from @xmath561 as a subdiagram . *
2.1 attaching @xmath564 * define the set @xmath501 as the set of nodes of future @xmath51 joined with @xmath32 by at most one dotted edge . as we have already proved ( see step 1 of the proof of lemma [ alg_p ] ) , after attaching of all the nodes of @xmath501 to @xmath32 we get at least @xmath565 nodes , so we obtain an inequality @xmath566 which is equivalent to @xmath567 further procedure depends on @xmath568 in @xmath51 we are looking for .
if @xmath569 , then @xmath570 and we are done . now suppose that @xmath571 .
the procedure to attach the nodes is the following .
for each node @xmath572 we join a new node @xmath71 with all nodes of @xmath573 by non - dotted edges in all possible ways , and for each of them compute the weights of the edges @xmath574 and @xmath575 . if one of these weights is equal to @xmath562 for some positive integer @xmath563 , and another is either equal to @xmath576 or is greater than one , then we get @xmath577 and put it in the list @xmath578 .
if we assume that @xmath579 , then we need to attach other nodes from @xmath501 .
for this we take all the pairs @xmath580 of ( possibly isomorphic ) diagrams from @xmath578 with the same @xmath32 ( i.e. @xmath577 , @xmath581 ) , compose a diagram @xmath582 and compute the weight of the edge @xmath583 from the equation @xmath584 ( since @xmath585 ) .
we put @xmath508 in the list @xmath586 if and only if the weight either is equal to @xmath562 for some positive integer @xmath563 or is greater than one . in this case
we call the pair @xmath580 _ compatible_. so , the list @xmath586 consists of unions of pairs of compatible diagrams from @xmath578 .
in the same way we can construct a list @xmath587 as the set of triples of mutually compatible diagrams from @xmath578 , and so on .
in other words , if we are looking for @xmath51 with @xmath588 , then @xmath51 should contain a subdiagram from the list @xmath589 composed of @xmath590-tuple of mutually compatible diagrams from @xmath578 .
notice that @xmath591 , so the number of lists @xmath592 is finite .
* 2.2 walking along edges * denote by @xmath45 the maximal index of @xmath592 obtained at the previous step , i.e. @xmath593
but @xmath594 . we fix @xmath5 and look for polytopes @xmath595 .
notice that we have a bound on @xmath5 : @xmath596 define the list @xmath597 .
let @xmath598 be the subset of @xmath599 consisting of all diagrams which are not diagrams of @xmath4-polytopes and which contain no more than @xmath600 dotted edges .
consider any diagram @xmath601 .
@xmath481 is not a diagram of a polytope , so it contains at least one elliptic subdiagram @xmath88 of order @xmath4 which belongs to less than @xmath4 complete diagrams of edges
. we will attach to diagram @xmath481 a node @xmath83 to create a new complete diagram of an edge in the following way : for each @xmath602 we attach @xmath83 to @xmath603 with at most one dotted edge , and then compute the weights of all the other edges joining @xmath83 with @xmath481 .
we need to attach nodes with at least @xmath11 dotted edges only , otherwise we get a diagram @xmath604 . here
we assume that the diagram of the initial vertex @xmath64 contains a node incident to a maximal number of dotted edges , which implies that @xmath83 is not joined by a dotted edge with some node @xmath602 ( the proof is identical to the one provided in the claim in step 2 of lemma [ alg_p ] ) .
let @xmath605 be the set of all such diagrams @xmath606 for all @xmath601 and @xmath607 that contain no more than @xmath600 dotted edges . creating in this way complete diagrams of edges and attaching them to all @xmath608 , we obtain the set @xmath609 .
in the same way we get the sets @xmath610 , @xmath611 and so on ( we attach a new node to some @xmath481 only if @xmath481 is not a diagram of a @xmath4-polytope ) .
the union @xmath612 contains diagrams @xmath51 of all the polytopes @xmath613 ( where @xmath614 is the list of all @xmath4-polytopes obtained at the step 2.1 ) .
[ fast ] even the reduced algorithm is not too fast .
below we list some method to obtain better estimates for @xmath355 and to improve the algorithm .
* 1 . * better estimate for @xmath355 .
+ let @xmath24 be a diagram of a polytope @xmath615 and let @xmath616 be the maximal multiplicity of an edge in @xmath24 .
let @xmath160 be an elliptic subdiagram of order @xmath4 containing an edge of multiplicity @xmath19 .
by lemma [ dplus2 ] , there exist two points @xmath617 such that @xmath618 contains no dotted edges and all edges in @xmath275 have multiplicity less or equal to @xmath19 .
this gives us a way to obtain a better estimate for the maximal multiplicity of an edge .
we consider all elliptic diagrams @xmath88 of order @xmath4 containing no edges of multiplicity greater than @xmath548 and let @xmath619 be the maximal multiplicity of an edge in @xmath88 .
for each of these diagrams @xmath88 we consider all indefinite diagrams @xmath275 with zero determinant containing neither dotted edges nor edges of multiplicity greater than @xmath619
. the number of these diagrams may not be too large .
then the maximal multiplicity amongst ones appearing in all these diagrams can be taken into account instead of @xmath548 .
* 2 . * changing starting vertex .
+ in the algorithm provided in the proof of lemma [ alg ] we start from diagram @xmath64 of a vertex satisfying some requirements .
more precisely , @xmath64 should contain a node @xmath488 incident to maximal number of dotted edges . however , sometimes this condition can be avoided .
suppose that we have already attached all the nodes from @xmath501 to get @xmath620 .
now we want to attach to @xmath620 a new node @xmath83 to get a complete diagram of some edge @xmath621 for diagram of some vertex @xmath88 .
as we have seen above , @xmath622 so we have an inequality @xmath623 if we assume that @xmath620 already contains at least @xmath624 dotted edges , then any new node @xmath83 is incident to at most @xmath485 dotted edges , so there exists @xmath625 which is not joined with @xmath83 by a dotted edge .
therefore , we can apply our algorithm .
the requirement for @xmath620 to contain at least @xmath624 dotted edges is not too restrictive .
since @xmath626 , this condition holds , for example , if there are no triples @xmath627 of mutually compatible diagrams such that @xmath628 contains no dotted edges . but while avoiding the requirement on @xmath64 , we hugely reduce the computations .
indeed , we can now start from diagram @xmath64 of any vertex .
in particular , we may assume that @xmath64 contains an edge of the maximal multiplicity in @xmath24 .
this allows us in each case to make calculations up to some multiplicity depending on the starting elliptic diagram , but not up to @xmath548 .
in this section , we apply algorithm provided in lemma [ alg ] ( taking into account remark [ fast ] ) to classify polytopes from @xmath629 .
we list the main steps and further essential refinements .
suppose that @xmath56 is a coxeter diagram of a polytope , and consider an edge @xmath64 of maximal multiplicity @xmath63 in @xmath24 .
according to the algorithm , we want to find a subdiagram @xmath392 and @xmath630 satisfying the assumptions of lemma [ local ] and to estimate its local determinant .
@xmath66 is a polygon , so it is either a triangle , or has a pair of non - intersecting sides .
we consider these cases separately .
* case 1 . * @xmath66 has more than @xmath12 sides , @xmath631 .
+ in this case its diagram @xmath76 contains a subdiagram described in case 1 of lemma [ separate ] with a dotted edge of weight at least @xmath632 .
we introduce the notation on fig .
[ poly ] : @xmath213 is the weight of the dotted edge , and the remaining variables are integers .
writing down the explicit formula of @xmath291 , we can see that the local determinant is a decreasing function of @xmath213 ( in the appropriate domain of all the variables ) .
therefore , while looking for the estimate we may assume that @xmath633 .
we get the following expression : @xmath634 where @xmath635 now we will use the fact that every unknown weight is less than one . a rough estimate ( all the entries in the denominator and positive ones in the numerator are zeros , and negative ones in the numerator are equal to @xmath25 )
gives an inequality @xmath636 .
applying proposition [ loc_sum ] and taking into account that @xmath637 , we see that @xmath638 .
since the local determinant @xmath345 is a decreasing function on all the weights , this implies that @xmath639 , which , in its turn , implies that @xmath640 .
the bound above is very rough .
we can improve it in the following way .
suppose that @xmath641 .
then , repeating the considerations above for the simplified expression , we obtain a bound @xmath642 . by symmetry
, we get the same if @xmath643 . applying the same procedure to the case @xmath644 ( or @xmath645 )
, we obtain an upper bound @xmath646 .
thus , if we assume now that @xmath647 , then each of @xmath648 should exceed @xmath11 . differentiating the expression @xmath649 with respect to @xmath6
we see that the local determinant decreases .
this implies that maximum is attended for @xmath650 , and , similarly , for @xmath651 . repeating the procedure for simplified expression we get an estimate @xmath652 ,
so the contradiction shows that @xmath642 , and if @xmath653 then at least one of @xmath648 is equal to @xmath11 .
let us now improve the estimate another one time .
assume that @xmath643 .
then we have two possibilities : either @xmath641 or @xmath654 .
applying the same procedure in the both cases we get an estimate @xmath655 . in other words , if we assume that @xmath656 , then @xmath657 .
suppose that @xmath656 .
suppose also that @xmath658 .
then the derivative of the local determinant with respect to @xmath6 occurs to be negative , which implies @xmath652 as we have seen above .
thus , if @xmath659 , then @xmath657 , and @xmath644 .
similarly , @xmath645 . computing the local determinant in these settings , we get a bound @xmath660 .
the contradiction shows that if @xmath66 is not a triangle then @xmath655
. now we make the following computations . for each @xmath661-tuple of integers @xmath662 ( where @xmath663 is the maximal one ) we compute @xmath213 to satisfy the equation @xmath637 , and list all of them satisfying @xmath664 .
if @xmath665 and @xmath66 is not a triangle , then @xmath51 contains one of these subdiagrams .
the list ( denote it by @xmath666 ) is large : it contains around @xmath667 diagrams .
the number is approximate since while running the program we allow @xmath213 to be a bit smaller to prevent errors .
now it is easy to see that there exists a node @xmath668 which is not joined with @xmath669 by dotted edges ( the proof is similar to lemma [ dplus2 ] ) .
thus , the diagram @xmath670 should be contained in the list @xmath666 .
more precisely , we are looking for a pair of diagrams from @xmath666 parametrized by @xmath671 and @xmath672 , such that the diagram @xmath673 has signature @xmath674 , and the nodes @xmath271 and @xmath675 are not joined by a dotted edge , see fig .
[ dot4 ] .
so , we look for pairs of diagrams from @xmath666 with the same @xmath676 , and @xmath213 , and for every such pair compute the weight of the edge @xmath677 .
the result is empty list .
therefore , for each polytope from @xmath629 with @xmath64 being the edge of maximal multiplicity @xmath678 the polygon @xmath66 is a triangle .
* @xmath66 is a triangle , @xmath631 .
+ in this case @xmath679 , and @xmath271 is any vertex of @xmath24 which is not joined with @xmath680 by a dotted edge ( it does exist by lemma [ dplus11 ] ) .
we get the subdiagram shown on fig [ triangle ] .
first , we are looking for reasonable @xmath681 .
the considerations are very similar to the previous case , but are much more detailed , so we show the plan .
we may assume that @xmath682 . the first attempt to get a bound is to consider two cases : either @xmath641 or @xmath654 .
if @xmath654 then a very rough estimate gives @xmath683 , where @xmath684 is the number of negative summands in the denominator of the expansion of the expression for local determinant @xmath291 , and @xmath685 is a diagram of triangle with minimum possible determinant ( while @xmath654 ) .
this implies that @xmath686 . doing the same for @xmath641 we get a bound @xmath687 , which implies @xmath688 . at this point
we could stop and say that @xmath689 .
however , this estimate does not look reasonable .
making several similar refinements ( considering more detailed cases for triples @xmath690 ) we can improve the bound .
for example , subdividing the case @xmath641 into @xmath651 and @xmath691 we get an estimate @xmath692 for the first case and @xmath693 for the latter . therefore , we see that @xmath694 , and if @xmath695 then @xmath641 and @xmath651 . according to our algorithm ( see remark [ fast ] ) we run the following computation for @xmath696 in these settings : we are looking for the diagram @xmath697 whose determinant vanishes , and all the entries do not exceed @xmath19 . actually , we find nothing , which implies that @xmath698 . repeating the procedure for more detailed cases of triples @xmath690 and running short computations , we come to a bound @xmath699 .
in fact , the estimate @xmath700 is sharp : there exists a unique diagram @xmath697 with zero determinant and @xmath701 , it is shown on fig .
[ k30 ] . here
@xmath702 .
so , we found @xmath355 .
now we go to the first step of the reduced algorithm ( see lemma [ alg ] ) : we want to list all the diagrams @xmath703 .
we find one diagram with @xmath701 ( mentioned above ) , one with @xmath704 , and @xmath705 with @xmath706
. then we proceed exactly as it is prescribed by the reduced algorithm . on our way
we get all the esselmann polytopes and @xmath9-polytopes with @xmath138 facets with multi - multiple edges .
however , we do not get anything new .
* case 3 . * @xmath707 .
+ in this case we follow the reduced algorithm _ verbatim_. again , we meet all the remaining esselmann polytopes and @xmath9-polytopes with @xmath138 facets , but we do not find any new polytope .
this leads to the following theorem completely describing @xmath629 : [ d4 ] all the polytopes belonging to @xmath629 have at most @xmath138 facets .
the way to deal with the diagrams such that @xmath66 contains dotted edge can be easily generalized to higher dimensions .
if the estimate for maximal multiplicity in this case is much better than in general one ( as we have for @xmath328 ) , this may speed up the process .
in section [ ess_def ] we define essential polytopes , present the list of known ( to the authors ) compact coxeter polytopes and formulate the main result of this section , theorem [ ess - t ] . in section [ ess_prop ] we discuss the methods allowing to differ essential polytopes from non - essential ones . in section [ ess_proof ]
we use these methods to show that all the polytopes claimed to be essential in theorem [ ess - t ] are essential , and that all other known polytopes ( with possible exclusion of two polytopes listed in question [ qop ] ) are not essential .
denote by @xmath708 the reflection group generated by reflections in the facets of @xmath3 .
suppose that there exists a finite index subgroup @xmath709 generated by reflections .
then the fundamental polytope @xmath710 of @xmath711 consists of several copies of @xmath3 .
as it is proved in @xcite , this procedure gives rise to infinitely many coxeter polytopes of finite volume in all dimensions up to @xmath712 , and infinitely many compact coxeter polytopes in all dimensions up to @xmath0 .
conversely , taking a polytope @xmath20 such that @xmath713 , we obtain a tessellation of @xmath3 by copies of @xmath20 ( in this case we say that _
@xmath3 is decomposed with fundamental polytope @xmath20 _ ) .
clearly , a polytope minimal with respect to this procedure is a fundamental domain of maximal reflection group .
another operation is the following .
suppose that for some facet @xmath59 of @xmath3 there exists a coxeter polytope @xmath710 with a facet @xmath714 congruent to @xmath59 , such that sums of corresponding dihedral angles composing by @xmath59 and @xmath714 respectively are integral divisors of @xmath2 .
then , gluing @xmath3 and @xmath710 along @xmath59 and @xmath714 , we get a new coxeter polytope .
makarov @xcite used this procedure to construct infinite series of polytopes in hyperbolic spaces of dimension @xmath9 and @xmath10 .
in particular , he produced fundamental domains of infinite number of non - commensurable reflection groups .
it is easy to describe the converse operation , i.e. _ dissecting _ a coxeter polytope into two smaller ones .
a compact coxeter polytope @xmath3 can be divided into two smaller coxeter polytopes @xmath715 and @xmath716 if there exists a hyperplane @xmath717 dissecting @xmath3 such that for every facet @xmath59 of @xmath3 \2 ) if @xmath718 contains no interior points of @xmath59 and @xmath719 then @xmath720 for some facet @xmath714 of @xmath3 , and the dihedral angles composed by @xmath717 with @xmath59 and @xmath714 are integral divisors of @xmath2 .
we say that a coxeter polytope @xmath3 is _ essential _ if it is minimal with respect to both operations above , i.e. the corresponding reflection group @xmath708 is maximal , and @xmath3 can not be dissected by a hyperplane into two smaller coxeter polytopes .
clearly , the classification of compact coxeter hyperbolic polytopes can be reduced to classification of essential ones .
we show that all known compact coxeter polytopes of dimension greater than @xmath12 can be glued from a finite number of essential ones , and list all known essential compact coxeter polytopes .
the main question we want to ask is * simplices @xcite ( in other words , @xmath4-polytopes with @xmath545 facets ) ; * @xmath4-polytopes with @xmath341 facets : * * simplicial prisms @xcite ; * * esselmann @xmath9-polytopes , i.e. products of two triangles @xcite ; * products of simplices @xcite ( which occurs to be a sublist of the previous list ) ; * @xmath4-polytopes with @xmath721 facets @xcite , @xcite ; * napier cycles @xcite ( which is a sublist of the previous list ) ; * truncated simplices @xcite . * @xmath661-dimensional polytope with @xmath722 facets ( the only polytope in dimension @xmath723 known so far ) ; * @xmath138-dimensional polytope @xmath724 facets ( besides its double , the only polytope known in dimension @xmath138 ) ; * @xmath0-dimensional polytope with @xmath725 facets ( see ( * ? ? ?
* table 2.2 ) where the gram matrix of this polytope is reproduced ) . to the authors
best knowledge , all the other examples of compact coxeter hyperbolic polytopes are defined by gluing smaller ones ( see e.g. @xcite , @xcite , @xcite ) , but we do not care about them now since they are already glued .
[ ess - t ] all the polytopes listed in table [ ess - list ] are essential .
all the remaining known compact hyperbolic coxeter polytopes of dimension at least @xmath9 except ones listed in question [ qop ] can be glued from these ones and the polytopes listed in question [ qop ] .
a _ nerve _ of a coxeter group is a simplicial complex with vertices indexed by standard generators ; a set of vertices span a simplex if the corresponding reflections generate a finite group . in case of equal numbers of facets we have the following result .
[ eq ] let @xmath726 be an infinite indecomposable coxeter group , and let @xmath727 be a finite index reflection subgroup of the same rank as @xmath726
. then the nerve of @xmath728 can be obtained from the nerve of @xmath726 by deleting some simplices .
[ eqp ] let @xmath3 and @xmath20 be compact hyperbolic coxeter polytopes of the same dimension .
suppose that @xmath713 and the numbers of facets of @xmath20 and @xmath3 are equal .
then @xmath20 and @xmath3 are combinatorially equivalent .
we can reformulate the corollary in terms of simplicial polytopes . recall that a fundamental polytope of cocompact reflection group is simple . therefore , for a cocompact reflection group @xmath729 its nerve coincides with face - lattice of simplicial polytope @xmath730 dual to @xmath20 .
hence , in view of theorem [ eq ] , corollary [ eqp ] is equivalent to the following statement .
[ esm ] let @xmath730 be a simplicial @xmath4-polytope , and let @xmath731 be its face - lattice .
then any proper simplicial subcomplex of @xmath731 with the same number of vertices is not a face - lattice of a simplicial @xmath4-polytope . indeed , suppose that @xmath732 is a proper subcomplex of @xmath731 , and @xmath732 is a face - lattice of a simplicial polytope @xmath733 .
then both @xmath732 and @xmath731 can be thought as triangulations of a @xmath734-sphere .
but since @xmath731 contains @xmath732 and they do not coincide , the number of vertices of @xmath731 should exceed the number of vertices of @xmath732 , so we come a contradiction .
now we are able to list some properties of decompositions of polytopes into smaller ones .
lemma [ pr_s ] concerns the case of a finite index subgroup ( i.e. , decomposition ) , and lemma [ pr_d ] concerns the case of dissection . for any polytope @xmath3 we denote by @xmath735 and by @xmath736 its volume and the number of its facets respectively . by a _
standard subgroup _ of @xmath708 we mean a subgroup generated by reflection with respect to some of the facets of @xmath3 .
to prove @xmath744 , consider the group @xmath753 . if @xmath750 has more facets than @xmath3 does , then @xmath717 intersects interior of all the facets of @xmath3 , so @xmath753 contains @xmath754 .
since @xmath750 is a coxeter polytope , this contradicts corollary [ subp ] , so @xmath744 is proved .
statements @xmath749 and @xmath751 are evident . to prove @xmath747 ,
notice first that @xmath716 has a unique facet @xmath755 which is not a facet of @xmath715 .
let @xmath756 be all the facets of @xmath3 and @xmath757 be a hyperplane containing @xmath53 .
it is sufficient to prove that for any subset @xmath758 an intersection @xmath759 intersects @xmath717 if and only if it intersects @xmath755 . without loss of generality we may consider maximal intersections only , namely the intersections of @xmath539 facets , or , in other words , the edges .
notice also , that any edge of @xmath3 which does not belong to @xmath755 is an edge of @xmath715 ( or an edge lying on a line that contains an edge of @xmath715 ) .
let @xmath760 be any edge of @xmath716 such that @xmath761 and @xmath762 .
then @xmath763 ( otherwise @xmath764 is an intersection of @xmath4 faces of @xmath715 , so @xmath765 , by andreev s theorem @xcite ) .
hence , @xmath766 , which implies @xmath767 in contradiction to the assumption .
this implies that if some edge of @xmath3 intersects @xmath717 then it intersects @xmath755 . on the other hand , if some edge @xmath768 of @xmath3 intersects @xmath769 ( @xmath770 ) and @xmath771 , then @xmath772 where @xmath773 is a line containing some edge of @xmath715 , so @xmath773 intersects @xmath717 and the condition @xmath747 is proved . to check if @xmath3 can be dissected in `` orthogonal way '' ( i.e. , by a hyperplane orthogonal to all facets of @xmath3 which it does meet )
we do the following .
for every @xmath4-tuple of facets @xmath774 of @xmath3 ( except those composing vertices of @xmath3 ) we draw a hyperplane orthogonal to all the facets @xmath258 ( it is unique ) , and look at the angles of @xmath717 with other facets .
if all of them are right for some @xmath4-tuple of facets , we get a dissection . to check if @xmath3 can be dissected in `` non - orthogonal way '' , for every @xmath775-face of @xmath3 we draw a hyperplane @xmath717 dissecting the corresponding dihedral angle into two , each of those is integral divisor of @xmath2 , and compute the angles of @xmath717 with the remaining facets of @xmath3 .
clearly , the number of such hyperplanes is finite since one of the new dihedral angles is not less than half of the initial one .
if all the new angles are integral divisors of @xmath2 , we obtain a dissection .
however , conditions @xmath776 of lemma [ pr_d ] usually give a much shorter way to understand if @xmath3 can be dissected or not .
suppose that @xmath3 is a union of @xmath715 and @xmath716 .
due to @xmath744 , we usually know the complete list of candidates to be @xmath715 and @xmath716 ( the only exceptions are the simplex truncated @xmath10 times , and @xmath0-polytope with @xmath725 facets we discuss later ) . condition @xmath747 further reduces the number of candidates
. then we glue pairs of candidates along congruent facets with appropriate dihedral angles and look if we get @xmath3 .
condition @xmath751 also allows to reduce the number of pairs in consideration when we know volumes of polytopes ( e.g. in even - dimensional case due to poincar formula ( * ? ? ?
* part i , chapter 7 ) or for simplices @xcite ) . similarly , lemma [ pr_s ] helps to check if @xmath3 is a fundamental domain of a subgroup of some group @xmath729 .
however , here we do not have an algorithm providing the answer .
for example , we can not decide if @xmath0-polytope with @xmath725 facets found by bugaenko @xcite ( the corresponding set of normal vectors is reproduced in ( * ? ? ?
* table 2.1 ) ) is essential or not : we do not know the lists of @xmath0-polytopes with @xmath777 or more facets . for any abstract coxeter diagram @xmath24 denote by @xmath778 the diagram obtained from @xmath24 by removing all nodes which are not incident to any edge labeled by an odd number @xmath19 ( or to a simple or triple edge ) ( see fig . [ odd ] ) .
[ dis_subd ] suppose that @xmath3 can be dissected into two coxeter polytopes @xmath715 and @xmath716 , both in the list of currently known polytopes .
then for any subdiagram @xmath779 such that @xmath780 holds @xmath781 .
the dihedral angles of all currently known polytopes are in the set @xmath782 , @xmath21 , @xmath206 , @xmath783 , @xmath784 and @xmath785 .
let @xmath717 be the dissecting hyperplane and suppose that @xmath717 dissects some dihedral angle of @xmath3 formed by facets @xmath786 and @xmath787 . clearly , @xmath788 ( where ( @xmath789 stays for the angle formed by @xmath717 and @xmath755 ) .
this is possible ( for the values listed above ) in the following three cases only : @xmath790 this implies that if @xmath216 and @xmath217 are facets of @xmath715 forming an angle @xmath791 for some @xmath792 then neither @xmath216 nor @xmath217 coincides with @xmath717 .
so , both @xmath216 and @xmath217 are facets of @xmath3 .
suppose that @xmath3 is dissected into some polytopes @xmath715 and @xmath716 .
then by property @xmath744 of lemma [ pr_d ] each of @xmath715 and @xmath716 has at most 8 facets .
hence , both @xmath715 and @xmath716 are in the list of currently known polytopes , and we may apply lemma [ dis_subd ] .
notice that each diagram of a @xmath10-polytope with at most 8 facets contains a subdiagram @xmath411 such that @xmath793 is connected and @xmath411 contains at least 3 simple edges . on the other hand , @xmath51 contains only two simple edges .
the contradiction to lemma [ dis_subd ] shows that the dissection is impossible . * in dimensions
68 * there is a unique known polytope not listed above ( see ( * ? ? ?
* table 4.9 ) ) , it is the @xmath0-polytope shown in fig .
[ rest6 ] , left .
this polytope is doubled polytope from fig .
[ rest6 ] , right , where the latter should be reflected in the facet marked white .
* in dimension 5 * all the remaining known polytopes are once truncated simplices ( i.e. , simplicial prisms ) or twice truncated simplices ( see @xcite and ( * ? ? ?
* table 4.10 ) ) .
it is shown in @xcite that if a simplex is not truncated in orthogonal way then it is not essential .
however , one of the simplices truncated in orthogonal way ( i.e. , from the list ( * ? ? ?
* table a.5 ) ) is not essential either : the polytope shown in fig .
[ rest5 ] , left , consists of two copies of the prism to the right , the copy to glue may be obtained by reflecting in the facet marked white . * in dimension 4 * the situation is more involved due to larger number of polytopes and combinatorial types .
the simplex not listed in table [ ess - list ] can be glued according to ( * ? ? ?
* section 4 ) ( it is a double of the simplex with both triple and double edges in the diagram ) .
all the prisms with one `` right '' base are contained in table [ ess - list ] .
the @xmath10 missed esselmann polytopes ( * ? ? ?
* theorem 1.1 ) are fundamental chambers for subgroups of index @xmath11 or @xmath9 of the groups generated by reflections in the facets of two esselmann polytopes listed in the table .
we are left to consider @xmath9-polytopes with at least @xmath138 facets .
the twice truncated simplices can be processed as in dimension @xmath10 . the only twice truncated simplex truncated in orthogonal way (
i.e. , from the list @xcite ) that is not essential is shown on fig .
[ rest4 t ] to the left , it can be tessellated by @xmath12 copies of the prism shown on fig .
[ rest4 t ] to the right ( the tessellation can be obtained by reflecting the polytope in two facets marked white and gluing all the three copies together ) .
the polytopes with @xmath138 facets whose diagrams have two multi - multiple edges and one dotted edge clearly tessellate @xmath11 or @xmath25 polytopes drawn in the same row ( see ( * ? ? ? * table 4.11 ) ) : one should reflect these polytopes in the facets having exactly one non - right angle @xmath784 or @xmath785 .
the polytopes shown on fig .
[ rest4s],left and fig . [ rest4s ] , middle ones , consist of two copies of one shown on fig .
[ rest4s ] to the right .
one shown in the middle of fig .
[ rest4s ] can be obtained by reflecting the right one .
however , the polytope shown on fig
. [ rest4s ] , left , does not define an index two subgroup .
the facet of the initial polytope marked white is a @xmath12-simplex whose coxeter diagram has a symmetry . to get the left polytope
we should glue a copy along this facet , not reflected one but symmetric to it . amongst the three times
truncated simplices ( they all have 8 facets ) only two are not listed in table [ ess - list ] : these are the polytopes shown on fig .
[ rest4s3 ] to the left .
each of these two polytopes can be obtained by gluing of two copies of the corresponding polytope shown on fig .
[ rest4s3 ] , right , where the latter should be glued by the facet marked white .
* 2 . * to prove that all the polytopes listed in the table [ ess - list ] are essential we use lemmas [ pr_s ] , [ pr_d ] , and [ dis_subd ] . as
above , we go through all combinatorial types by increasing complexity and decreasing dimension . notice that we should also consider decompositions of polytopes into non - essential polytopes .
the main consideration is that for any polytope in the table [ ess - list ] we know the complete list of polytopes of the same dimension with smaller number of facets or with the same combinatorics .
indeed , we know all @xmath4-polytopes with at most @xmath721 facets ; it is shown in @xcite that there exists a unique @xmath138-polytope with @xmath724 facets , and all the other polytopes from table [ ess - list ] are three times truncated simplices , the complete list of which is also known due to @xcite . * in dimension 5 * , using lemmas [ pr_s ] and [ pr_d ] together with lemma [ dis_subd ] we reduce the problem to the following : is the group generated by the polytope shown on fig .
[ 5left ] maximal or not ?
denote this polytope by @xmath794 . to see that @xmath795 is maximal
, we compare the distances between the disjoint facets in @xmath794 and in all possible candidates to generate a larger group .
namely , each of the polytopes satisfying the conditions of lemma [ pr_s ] has a pair of disjoint facets at larger distance one from another than the maximal of the three distances in @xmath794 , so , none of the candidates can be embedded in @xmath794 . * in dimension 4 * , we have many possibilities to consider , but we always can compute volumes of polytopes applying poincar formula ( see e.g. ( * ? ? ?
* part i , chapter 7 ) ) and then we can use properties ( 4 ) and @xmath751 of lemmas [ pr_s ] and [ pr_d ] . applying lemmas [ pr_s ] and [ pr_d ] together with lemma [ dis_subd ] ,
we reduce the problem to the following two questions : to solve the second question notice , that if @xmath798 is a subgroup of @xmath799 , then the neighborhood of the vertex of @xmath802 corresponding to the subdiagram @xmath804 is tiled by the neighborhood of the vertex of @xmath803 corresponding to @xmath805 , which implies that the facets of @xmath803 corresponding to the nodes @xmath124 , @xmath125 , @xmath806 , and @xmath807 are the facets of @xmath802 , while the remaining two facets are not .
so , two copies of @xmath803 should be attached by the facets @xmath808 and @xmath809 , however this does not lead to a coxeter polytope ( the polytope obtained has an angle @xmath810 ) .
v. o. bugaenko , _ groups of automorphisms of unimodular hyperbolic quadratic forms over the ring @xmath812 $ ] _ , moscow .
math . bull .
39 ( 1984 ) , 614 .
v. o. bugaenko , _ on reflective unimodular hyperbolic quadratic forms _ , selecta math .
( 1990 ) , 263271 .
v. o. bugaenko , _ arithmetic crystallographic groups generated by reflections , and reflective hyperbolic lattices , _ from : `` lie groups , their discrete subgroups , and invariant theory '' , adv .
soviet math . 8 , amer .
soc . , providence , ri ( 1992 ) 3355 . | we introduce a notion of essential hyperbolic coxeter polytope as a polytope which fits some minimality conditions .
the problem of classification of hyperbolic reflection groups can be easily reduced to classification of essential coxeter polytopes .
we determine a potentially large combinatorial class of polytopes containing , in particular , all the compact hyperbolic coxeter polytopes of dimension at least @xmath0 which are known to be essential , and prove that this class contains finitely many polytopes only .
we also construct an effective algorithm of classifying polytopes from this class , realize it in four - dimensional case , and formulate a conjecture on finiteness of the number of essential polytopes
. [ section ] [ theorem]conjecture [ theorem]lemma [ theorem]corollary [ theorem]property [ theorem]proposition [ theorem]definition [ theorem]question [ theorem]remark [ theorem]example h s @xmath1 ] u |
the unification of standard model and general relativity is one of the most intriguing and desirous goals of modern physics , and it has stimulated many theoretical ideas towards quantum gravity ( qg ) .
some of them predict lorentz violation ( lv ) , where the lorentz symmetry of space - time breaks down at high energies , and it introduces some tiny `` lv relics '' at lower energies .
this probability has arisen in the space - time foam @xcite , loop gravity @xcite , vacuum condensate of tensor fields @xcite , and the so - called doubly special relativity ( dsr ; also known as the deformed special relativity ) @xcite .
lv is an important issue in qg theories , mainly because it provides some available scenarios to validate or falsify theories through `` windows on qg '' @xcite .
generally , we know that the quantum - gravitational effects are relevant only when the energy approaches the planck energy , @xmath2 gev .
this huge number is obviously out of reach in the foreseeable future .
however , at lower energies , the `` relic '' effects can demonstrate themselves and produce a modified energy - momentum dispersion relation with extra terms suppressed by @xmath3 . here
the power @xmath4 and the coefficient @xmath5 both depend on the specific theoretical framework under consideration @xcite . to see the manifestations of the lorentz - violating physics , novel insights are needed , and fortunately , some have already been proposed practically to probe tiny modifications to conventional physics .
accurate measurements in the laboratories on the earth @xcite , violating processes in astrophysics with energetic particles @xcite , and accumulated effects through cosmological amplification mechanisms @xcite , are three basic considerations till now to reveal new physics arisen in the lv scenarios . here
we try to review and discuss some candidate theories on lv , and then utilize the modified energy - momentum dispersion relation to search observational clues in the photon sector ( sometimes jointly , the electron and positron sectors ) .
high energy photons originated from astrophysics are extremely promising to probe or falsify lv theories in various aspects @xcite .
this paper is organized as follows . in sec .
[ sec2 ] , several selected underlying models are briefly reviewed and discussed with the most recent observational constraints . in sec .
[ sec3 ] , we focus on photons from astrophysical sources
. these photons can reveal lv effects through their time - of - flight with energy - dependent velocities , and their interactions with low energy background photons , namely the extra - galactic background light ( ebl ) and/or the @xmath6 k cosmic microwave background ( cmb ) radiations .
multi - tev @xmath7-rays may also be involved in self - decays into @xmath8 pairs with proper lv parameters . in sec .
[ sec4 ] , the most updating lv constraints from the spectra of ultra - high energy cosmic rays ( uhecrs ) are reviewed and discussed .
[ sec5 ] summaries the paper with discussions on several current and upcoming instruments related to lv physics .
in a rough manner , lv theories can mainly be classified into two categories .
one is effective field theory ( eft ) , which provides an excellent framework where tiny lv effects are introduced into the lagrangian through lv operators renormalizable ones with mass dimension 3 and/or 4 , see , _ e.g. _ , the standard model extension ( sme ) @xcite , and the further extended non - renormalizable ones with mass dimension 5 and/or 6 @xcite .
the other category includes those qg theories that can not be embedded into the eft framework , such as the quantum space - time foam model @xcite and dsr @xcite . in the eft framework ,
all renormalizable operators , which guarantee the gauge symmetry preserved , are introduced by colladay and kosteleck @xcite perturbatively and elegantly .
the lv terms can be characterized by dimensionless tiny coefficients of order @xmath9 or even smaller .
lately , myers and pospelov @xcite proposed dimension-5 non - renormalizable lv terms into the eft lagrangian .
the dimension-5 cpt - odd terms in quantum electrodynamics ( qed ) read as @xcite , @xmath10 while the dimension-5 cpt - even terms appear as @xcite @xmath11 in above lagrangians , @xmath12 is a fixed timelike four - vector indicating the preferred frame , which is usually chosen as the frame where the cmb radiation is isotropic ; @xmath13 is the dual of @xmath14 , defined as @xmath15 ; @xmath16 are projection operators , and @xmath17 is the qed covariant derivative ; @xmath18 , @xmath19 , and @xmath20 are dimensionless lv coefficients naturally expected to be of order @xmath21 . lately , the dimension-6 cpt - even terms are also introduced , by mattingly @xcite , @xmath22 we can see that in the non - renormalizable eft , the lv terms are spontaneously suppressed by @xmath23 or @xmath24 , compared to the renormalizable lv terms with tiny coefficients of order @xmath9 put in by hand @xcite .
however , there remains a `` naturalness problem '' @xcite in the non - renormalizable eft radiative corrections may generate `` baby '' renormalizable terms which subsequently become even more important than the original `` parent '' non - renormalizable operators .
luckily , supersymmetry can play a role to prevent or alleviate this problem @xcite . from the above lagrangians
, we finally arrive at the modified energy - momentum dispersion relations for photons and fermions , respectively , @xmath25 @xmath26 where @xmath27 is the rest mass of fermions , and @xmath28 are combinations of @xmath19 and/or @xmath29 @xcite , where their subscribes `` @xmath30 '' denote two opposite helicities . in the eft , @xmath31 , and @xmath32 @xcite .
however , @xmath33 is not necessary to equal to @xmath34 in general . for more details on non - renormalizable eft calculations and related issues ,
see _ e.g. _ , refs .
@xcite . from aspects of observational limits ,
unfortunately , the dimension-5 operators seem unnatural , even with supersymmetry included and softly broken @xcite .
the linear energy dependence in eft is largely unfavored @xcite .
the most strict observations come from the vacuum birefringence and synchrotron radiations @xcite .
so we would like to fix @xmath35 in the following analysis , and impose @xmath36 for simplicity ( denoted as @xmath37 hereafter ; liberati and maccione also treated eft in the same way in ref .
this choice preserves cpt symmetry and helicity symmetry , though they are not the essential ingredients generally .
there are also several models where the local description in terms of effective lagrangian breaks down . for instance , ellis _
proposed a liouville - inspired stringy analogue of space - time foam model @xcite . in this model
, a gas of d - particles roams in the bulk space - time of a high - dimension cosmology , wherein our universe is represented as a d3-brane .
photons are represented by open strings .
their interaction with d - particles excites and stretches the strings , and then decays and emits photons .
thus `` gravitational medium effects '' modify the canonical dispersion relation of photons to include an lv term which depends linearly on the energies of photons , @xmath38 $ ] , while the dispersion relation for charged particles remains untouched and no helicity dependence is predicted @xcite .
hence , it avoids many tight constraints from the vacuum birefringence of photons , the electron / positron sector , and the uhecrs sector @xcite .
most recently , ref .
@xcite claimed that the space - time foam model is unable to explain simultaneously the time lags of tev @xmath7-rays , and the non - observation of high energy photons above eev .
however , ellis , mavromatos , and nanopoulos argued that there are ways to avoid the constraints @xcite , hence the issue is still in dispute . on the other hand , the time - of - flight of photons from distant @xmath7-ray bursts ( grbs ) and active galactic nuclei ( agns )
strikingly favors the linearly energy dependent lv corrections @xcite .
therefore , it still has strong motivations and advantages to study this linearly dependent lv theory .
the above discussed lv theories are far from complete .
especially , dsr @xcite is a well - motivated candidate where two invariant scales are included , namely the velocity of low energy photons , _
i.e. _ , the conventional light speed @xmath39 , and a length scale @xmath40 ( or equivalently , an energy scale @xmath41 ) . in the dsr
, the energy - momentum conservation laws may also be modified @xcite , thus brings further complications to the analysis of lv phenomena . in this paper , the canonical energy - momentum conservation is assumed . for a phenomenologically uniform description ,
we use the following dispersion relations , for photons and fermions , respectively , @xmath42 \,,\ ] ] @xmath43\,,\ ] ] where only @xmath44 ( the linear modification ) and @xmath45 ( the quadratic modification ) are currently observationally relevant .
more specifically , we should keep in mind that the case with @xmath46 , @xmath47 corresponds to the space - time foam model , while the case with @xmath48 , @xmath49 , @xmath50 represents the dimension-6 eft with cpt - even lv operators . in addition , the case with @xmath51 , @xmath52 is also of great interests , for the reason that it can be a consequence of the geometrically small scale of space - time and the equivalence principle to all species of particles .
as the dispersion relations are modified , we can suspect extra lv impacts beyond the standard model .
the tiny corrections can be magnified when the energy @xmath53 is large , and also maybe through cosmological amplification mechanisms .
hence when two magnification mechanisms are combined , the astrophysical photons with very high energies provide plentiful motivations to be regarded as an important source to investigate lv physics .
the dynamics of lv is poorly understood , hence we focus on the kinematics of photon propagation from distant galaxies towards us .
the effect of lv on high energy photon propagation is threefold .
firstly , the velocity of photons , defined as @xmath54 , is no longer a constant , according to eq .
( [ photon ] ) .
consequently , photons with different energies spend slightly different time - of - flight to arrive at the earth .
it was firstly recognized in refs .
@xcite , and is discussed extensively by the scientific society with great passion recently on the fermi , magic , and hess observations @xcite . secondly , the modified dispersion relation may allow the strictly forbidden reactions in the lorentz - symmetric theories to occur , such as photon decay , @xmath55 .
it results in a cutoff in the spectra of @xmath7-rays above the decay threshold .
thirdly , high energy photons interact with low energy background photons ( cmb and/or ebl ) on the way of propagation , through the reaction @xmath56 ; lv effects can influence the threshold of this process , and lead to interesting phenomena @xcite .
we discuss the above three scenarios in the following .
worthy to stress that , since the lorentz symmetry is broken in the lv theory , the transformation laws between two relatively moving frames are unknown , hence the laboratory system is the only proper frame to discuss relevant physics before we get access to proper transformation laws .
amelino - camelia _ et al . _
@xcite first suggested using grbs to test lv physics . due to the large cosmological distance and the fine time structure of grbs ,
tiny lv effects can be amplified into observable quantities . by taking into account cosmological expansion of the universe , the time lag induced by the lv modified dispersion relation between photons with high energies @xmath57 , and
those with low energies @xmath58 , is @xcite , @xmath59 where @xmath44 and @xmath45 stand for linear and quadratic energy dependence ; @xmath60 km s@xmath61 mpc@xmath61 is the hubble expansion constant ; @xmath62 is the vacuum energy density , and @xmath63 is the matter energy density in the current universe ; @xmath64 is the redshift of the source .
fermi lat discovered that high energy photons have a tendency to arrive later relative to low energy ones @xcite , which might present potential evidence for lv with @xmath65 for the space - time foam model or @xmath66 for the dimension-6 cpt - even eft .
however , the determination of time lag from observational data is highly nontrivial and affected by many facets , both artificially and instrumentally @xcite .
the primary uncertainty comes from the unknown effects from source activities , mainly due to our imperfect knowledge of the radiation mechanism of grbs .
however , we can separate the source effects if we can achieve a survey of grbs at different redshifts @xcite .
the time lag induced by lv accumulates with propagation distance , as it is a gravitational medium effect . on the contrary ,
the intrinsic source induced lag is likely to be a distance - independent quantity , which can be approximated as a constant .
the relation between the observed delay @xmath67 and the intrinsic time lag @xmath68 is @xmath69 in ref .
@xcite , we systematically studied this scenario of fermi lat grbs , and got the quantum gravity scale @xmath70 gev for the linear energy dependence , which corresponds to @xmath71 .
worthy to mention that , @xmath72 represents superluminal photon propagation , while @xmath73 corresponds to subluminal propagation .
the time - of - flight of high energy photons is the clearest test to search lv effects of the photon sector , hence provides the most convincing results .
agns and pulsars are also utilized to study lv effects through time - of - flight analysis @xcite .
the observational results that high energy photons arrive relatively later than low energy ones suggest a negative @xmath5 , and current data largely favor linearly energy dependent lv effects @xcite
. the alteration of the dispersion relations of photons and fermions might also lead to some peculiar reactions to occur , like the basic qed vertex @xmath1 , called photon decay , which is generally forbidden by the canonical energy - momentum conservation in the standard model .
photon decay might be realistic in the framework of lv theories with proper lv parameters .
we consider a very high energy photon of momentum @xmath74 , which decays into an electron of momentum @xmath75 , @xmath76 $ ] , and a positron of momentum @xmath77 .
utilizing the `` threshold theorem '' @xcite , energy conservation leads to @xcite @xmath78 = xk \left [ 1 + \frac{m^2}{2(xk)^2}+\frac{\eta_{n}}{2 } \left(\frac{xk}{e_{\rm pl}}\right)^n \right ] + \{x \leftrightarrow 1-x\ } \,,\ ] ] where we only keep the leading corrections of planck - scale terms , and the leading term in the series of @xmath79 .
the consistence of the truncation can be checked .
after a few steps , the equation turns into @xcite @xmath80 .\ ] ] to get the threshold to valid photon decay , is equivalent to minimize @xmath74 on the left - hand side , hence to maximize the right - hand side of eq .
( [ pd ] ) @xcite .
we consider three theoretical scenarios in the following .
\(i ) @xmath81 this is the lv parameter configuration for the space - time foam model @xcite . under these circumstances , eq .
( [ pd ] ) reduces to @xmath82 \xi_1 $ ] .
we can see clearly that @xmath83 leads to @xmath84 , which is the very case in the standard model where no photon decay is kinematically allowed . for @xmath85
, there is no meaningful solution for the photon decay process . on the contrary , for @xmath86 , any high energy photon with its momentum larger than @xmath87 would decay rapidly into an @xmath8 pair @xcite , thus we are not supposed to observe photons with momentum larger than @xmath88 from astrophysical sources . on the other hand , the observation of multi - tev photons can cast a constraint on the lv parameter @xmath89 from the @xmath86 side @xcite .
if @xmath89 is of order @xmath90 , then the relevant energy lays within the currently observational capability .
actually , the observations of the @xmath91 and @xmath92 photons from the crab nebula @xcite have already constrained @xmath89 to be less than @xmath93 in this scenario .
it should be noticed that , here the power dependence of @xmath94 on @xmath89 , @xmath95 , differs from that of ref .
@xcite , @xmath96 , because there the energy dependence of lv in terms of the maximal attainable velocity @xcite is different from that of the space - time foam model .
however , in the relevant energy range , the results are roughly the same .
\(ii ) @xmath97 if the modification of the dispersion relation is due to pure planck - scale geometry of space - time , and the equivalence principle applies to all species of particles , then the fermions ( electrons and positrons ) share the same lv parameters as the photons . and if the dispersion relation is modified to the first order , _
i.e. _ , linearly energy dependent , then from eq .
( [ pd ] ) , the energy - momentum conservation leads to @xmath98 \xi_1 $ ] .
thus again , for @xmath99 , no photon decay is allowed , and for @xmath100 , there exists a threshold for photons with momentum larger than @xmath101 to decay .
this is also within the energy range of observational practicability if the lv parameters are of order @xmath90 .
\(iii ) @xmath102 in the above two cases , the threshold happens when @xmath103 , _
i.e. _ , the electron and the positron in the final state have the same amount of momentum . however , this is not a general property in the lv photon decay phenomenon , as first pointed out by jacobson , liberati and mattingly @xcite . in the dimension-6 cpt
- even eft framework , lv terms of photons and fermions both depend quadratically on the energy .
now eq . ( [ pd ] )
is reduced to @xmath104 \times \left [ \xi_2 - \eta_2 \left((1-x)^{3}+x^{3}\right ) \right]$ ] , where the right - hand side is quartic of @xmath105 if @xmath106 , and the discussions to derive the threshold energy become more subtle .
if @xmath107 , then it is a similar case as the space - time foam model , where @xmath108 corresponds to no photon decay , while @xmath109 leads to a threshold at @xmath110 . and
@xmath111 in arbitrary unit .
[ pdiagram],width=264 ] for the @xmath106 case , after introducing @xmath112 , @xmath113 $ ] , the constraint becomes quadratic of @xmath64 @xcite , @xmath114 the maximum of the right - hand side can happen at @xmath115 ( no asymmetry ) , @xmath116 ( the maximum asymmetry ) , or @xmath117 ( a mediate asymmetry ) , depending on the realistic values of lv parameters @xcite . after a detailed calculation
, we can get the `` phase diagram '' for photon decay in the dimension-6 cpt - even eft , illustrated in fig .
[ pdiagram ] @xcite , where @xmath118 @xmath119^{1/4 } \simeq 0.15 \left [ \frac{-\eta_2}{(\xi_2-\eta_2)^2 } \right]^{1/4}~{\rm eev } \,.\ ] ] we can see that the physical process @xmath120 heavily depends on the phase space of lv parameters . in the region where @xmath121 ( @xmath103 thereof ) , it appears the naturally expected configuration of final state , where the electron and the positron share the same amount of the initial momentum .
however , in some other phase space , the asymmetric momentum configuration of final state may appear , namely that the electron and the positron gain different proportions of the initial momentum , satisfying @xmath122 ( @xmath123/2 \neq 1/2 $ ] thereof ) @xcite .
besides , there is also a regime where no photon decays can ever happen . from the above discussions
, we can see that within different theories , there can be a cutoff in the @xmath7-ray spectra above a threshold energy with lv parameters in some certain regions .
the cutoff is predominant , instead of a suppression , because the time scale of the photon decay is quite short and the process takes place in no time compared to their time - of - flight @xcite .
thus , the photon decay feature is obvious in the photon spectra if it exists , and hence should be observed explicitly .
the @xmath45 case has the cutoff energy four orders higher than that of the @xmath44 case , thus it is more challenging to practical observations .
however , the observations concerning the quadratic lv modification , as well as the linear lv modification , are already emerging .
another interesting phenomenon comes from the interaction of a high energy photon @xmath7 with a low energy photon @xmath124 through the pair - production reaction @xmath125 .
cmb is the relic radiation from the big bang which evolutes in the expanding universe after the lights decouple with the matters .
and ebl is the faint diffuse light of the sky , consisting of the historically accumulated flux of all extragalactic sources .
it is relevant for the formation of stars and galaxies , and also the large - scale structure of the universe .
its wavelength varies from the radio band to the ultraviolet band .
the observed spectra of high energy photons from distant galaxies are modified on the way of propagation , due to the attenuation by these low energy background `` absorbers '' .
the presence of lv will change the reaction threshold and modify the reaction patterns in an intriguing and physically meaningful manner @xcite .
it can be extended to the ultra - high energy photons , as well as the uhecrs , without any further difficulties @xcite .
this might be the most extensively investigated scenario among the lv - induced `` windows on qg ''
the lv effects depend on the energy of photons involved , and the magnitude and also the sign of the lv parameters , as well as the theoretical framework under considerations .
as we can see in the photon decay section , different parameter configurations lead to different lv physical phenomena .
since four particles are involved in the pair - production absorption , we naturally expect more subtle and complicated dependence on the phase space of lv parameters than the photon decay case .
_ provided the most detailed discussions on this issue @xcite . here
we look into two possible scenarios which are the most relevant and well constrained from observational aspects : ( i ) the propagation of multi - tev photons @xcite , ( ii ) the photo - meson production through the gzk mechanism and the subsequently pion - produced photons with energies above eev @xcite .
we would work in the simple framework provided by jacob and piran @xcite , who studied the linearly lv suppressed scenarios in a clear manner .
assuming that the lv parameters of fermions vanish , then by utilizing the `` threshold theorem '' @xcite , we can arrive at the threshold of the low energy background lights to interact with an incidental photon with energy @xmath126
@xcite , @xmath127 @xmath128 leads to the conventional case @xmath129 , of course .
if the lv parameters of electrons and positrons are also introduced , the lv modified patterns are hardly further modified .
it only introduces a numerical rescaling of order @xmath130 for the second term in eq .
( [ abeq ] ) , by assuming the likely magnitude of the lv parameters of fermions compared to that of photons @xcite .
thus we here ignore them for simplicity . in the superluminal propagation case with @xmath131 , the photon decay process dominates , where it results in a cutoff , and no photons are observable above energy larger than the decay threshold .
thus we focus on the subluminal case with @xmath132 here .
it is easily seen that there is a minimum on the right hand side of eq .
( [ abeq ] ) .
the minimum appears as the most crucial feature from the pair - production absorption within lv theories @xcite .
firstly , it means that the background photons with energies lower than the minimum never interact with the high energy photons .
secondly , most importantly , lv predicts a `` reemergence '' of the high energy photon flux when the incidental photons have energies above the critical energy @xcite , @xmath133^{1/(n+2 ) } \,.\ ] ] for the @xmath44 case , @xmath134 , while for the @xmath45 case , @xmath135 .
the critical energies are of the same order as the threshold energies of photon decays , because the photon decay can be viewed as a high energy @xmath7-ray photon interacting with a `` low '' energy one with infinite wavelength ( therefore zero energy actually ) .
thus photon decay is a special case of the reaction @xmath125 with a vanishing @xmath124 . above the critical energy , with the energies of high energy photons increasing , the threshold of low energy photons to interact also increases . in the relevant energy range ,
the photon density of the background lights decreases with energy .
hence , there are less low energy photons to annihilate with high energy photons above the critical energy , therefore , a recovery of the high energy photon flux is expected in the lv theories above the critical point @xcite . as for the photons with energies above eev , galaverni and sigl
considered the ultra - high energy photons from gzk - cutoff - induced pion decays @xcite . in their work ,
the lv parameters for the electron and the positron are also fully taken under considerations .
for the non - lv case , these ultra - high energy photons would interact with the cmb photons and lead to a cutoff in the spectra ( actually , a suppression here ) above the pair - production threshold . and
if lv - corrected terms exist in the dispersion relations , as shown above , the threshold will shift and there is a reemergence above some critical energy if @xmath136 .
hence , it is supposed to observe the `` reemergent flux '' through auger , agasa , and yakutsk observations , according to numerical calculations @xcite .
however , these observations only find upper limits for photon flux in the relevant energy range .
the non - observation situation allows very tight constraints on lv parameters
@xcite reported that , because of the non - observation of photons above @xmath137 , @xmath138 is limited to be below @xmath139 , and @xmath140 . a recent work utilized the same scenario , intending to rule out the space - time foam model mentioned above , after taking into the energy loss of photons in the stochastic interactions with the d - branes @xcite .
due to the most violating processes and long baseline propagations in astrophysics , there has been accumulating significant constraints for lv parameters from the astrophysical society @xcite . aside from the above mentioned scenarios ,
other processes are also taken under considerations by many authors and have achieved good progresses on the lv issues , _
e.g. _ , vacuum birefringence , vacuum @xmath141erenkov radiation , helicity decay , photon splitting @xcite .
the lv modifications are usually suppressed by the ratio of the particle energy to the planck scale , hence the high energy particles have a merit to amplify these tiny effects .
we here review some most recent limits from uhecrs , because they possess the highest energies we have ever observed , and can cast tight constraints on lv theories .
uhecrs are among the most studied phenomena of lv for historical and practical reasons . in a certain period of time
, people doubted whether a `` standard '' gzk cutoff exists , and many models are proposed to account for the `` shift '' or `` disappearance '' of the gzk cutoff . among them , the lv explanation was a promising candidate .
nowadays , since the gzk cutoff is found around the predicted point , it is used to constrain the lv parameters @xcite .
_ got the difference of the maximal attainable velocity @xcite of pions to photons , @xmath142 and @xmath143 , by utilizing the hires monocular spectra and the auger combined spectra , respectively , under the assumptions that the lv modifications for nucleons are negligible and that the dominant component of uhecrs is the proton @xcite .
stecker and scully derived a best fit to the lv parameter of @xmath144 from studies on the auger spectra @xcite .
the above two studies both refer to the conception of the center of mass system , which might not be a well - defined conception in presence of lv .
et al . _ worked in the framework of eft with dimension-5 and dimension-6 operators , and utilized the auger data to establish two - sided bounds on the coefficients for the proton and the pion , whose lv scales are found to be well above the planck energy @xcite .
though various uhecrs spectra studies favor the standard gzk explanation , the lv scenarios are not ruled out thoroughly .
small derivations from lorentz symmetry are still under serious considerations .
however , the fact that the cutoff point together with the spectrum of uhecrs are consistent with the standard theoretical expectation favors the gzk explanation without lv . some authors argued that , the compound of uhecrs is not well understood yet . even for general knowledge
, we believe that they are protons or heavy ions
. then uhecrs are composite entities instead of fundamental particles , and their underlying degrees of freedom are actually quarks and gluons .
this situation may limit the constraints derived from uhecrs @xcite .
moreover , the gzk cutoff may also be due to the accelerating capability of the uhecrs host galaxies .
therefore , more works on uhecrs from observational and theoretical aspects are needed to address these related issues
. on the other hand , as mentioned , refs .
@xcite realized that uhecrs lead to photo - pion production when undergoing inelastic collisions with cmb photons , and subsequently pions decay into the ultra - high energy photons , _
i.e. _ , the process @xmath145 always accompanies with the gzk cutoff process @xmath146 . after a research into the influence of lv terms in this scenario , they reached the conclusion that @xmath138 is limited to be below @xmath139 , and @xmath140 , in order to explain the non - observation of photons in the uhecrs flux above @xmath137 @xcite . ending this section ,
let us compare and discuss some lv scenarios .
the lag caused by the time - of - flight @xcite is the cleanest test ever , and it introduces no interaction into the phenomenological analysis , and some primary studies are emerging with attempt to disentangle the unknown ( thus controversial ) source effects @xcite .
hence , it provides the most convincing evidence for lv effects on the photon sector .
time - of - flight analysis favors linear energy dependence within current observations .
however , the results are contradictory with other constraints from kinematics of reactions , where the linearly corrected lv terms are constrained to be very unlikely @xcite .
even the stringy space - time foam model , where no corrections for charged particles are predicted thus avoiding many constraints from charged species , is debatably unfavored by the non - observation of the ultra - high energy photon flux @xcite .
as for species other than photons , _
e.g. _ , electrons , positrons , quarks , and gluons , they can enjoy their own individual lv parameters differing from those of photons in principle . in some theories , particles with different helicities
can also possess unrelated lv parameters .
thus lots of lv parameters are involved totally , and detailed analysis is needed and fortunately is performing to constrain all these tiny lv terms @xcite . since most reactions contain more than one species , combined analysis is severely in demand .
as we can see from the photon decays in the dimension-6 cpt - even eft , the lv physics generally depends highly , sometimes even nonlinearly , on the phase space of lv parameters .
many efforts are still focusing to combine constraints from different types of analysis to derive a robust result @xcite . as for composite particles , the jump from fundamental degrees of freedom to the compounded entities , _
e.g. _ , the protons and the heavy ions in the uhecrs , requires more careful arguments and treatments .
lorentz violation ( lv ) may arise from many quantum gravity ( qg ) models . and apart from its origin
, it is apparently important for its own sake .
theoretically , lv physics can originate from stringy considerations , effective field theories ( efts ) , doubly special relativity , loop gravity theories , and so on .
correspondingly , fruitful and intriguing phenomena emerge from lv corrections , and fortunately they are testable on several operating platforms from various aspects nowadays , _
e.g. _ , the fermi satellite and the pierre auger collaboration .
this issue becomes more and more attractive to the scientific society nowadays , because evidence and constraints from experiments and observations are appearing continuously and steadily reachable and reliable .
the most promising standpoint of the testability comes from the modified energy - momentum dispersion relation , where extra lv terms are introduced through different theoretical scenarios , compared to the canonical energy - momentum relation , @xmath0 . in this paper , we extensively review the studies of the lv influence on the astrophysical photons with very high energies .
they are classified into three categories : ( i ) the velocity of photons @xmath147 now depends on the wavelength of the light , hence introduces a time lag between photons with different energies , which can be observed from @xmath7-ray bursts ( grbs ) and the flares of active galactic nuclei ( agns ) ; ( ii ) newfangled reactions that are strictly forbidden in the standard model due to the non - conservation of energy - momentum might happen within lv theories , and a cutoff of photon flux above the threshold energy of photon decay , @xmath148 , is expected with proper lv parameters ; ( iii ) lv also alters the threshold energy of normal reactions , hence the spectra of the very high energy photons can be renewed to provide physically intriguing phenomena , due to the modified patterns of the reaction @xmath149 , for multi - tev photons and eev photons through interacting with low energy background lights . to search for the lv physics in above scenarios ,
there are several indispensable steps .
most importantly , more sensitive detectors should play a decisive role , since generally , the higher the energy , the more diluted the flux .
currently , the auger collaboration and the hess collaboration are playing significant importance in relevant fields .
an `` auger north '' array intending to behave better than the present southern hemisphere auger array is also under consideration .
an improved knowledge of the astrophysically intrinsic mechanisms is also essential in searching lv physics from astrophysical side , _
e.g. _ , the details of the background lights , the un - attenuated spectra of sources , the acceleration and radiation mechanisms of uhecrs .
space - based satellites , fermi and the incoming `` extreme universe space observatory onboard japanese experiment module '' ( jem - euso ) are promising missions to pindown astrophysical issues . as we can see , lv physics reveals itself at high energy bands more obviously , thus the techniques to detect higher energy cosmic rays , _
e.g. _ , @xmath150 ev , are also of great significance . in summary ,
theoretically , we can not predict the values of lv parameters in lv theories from the first principle most of the time though , fortunately , experimental and observational data are serving tight constraints from various aspects to falsify theories and/or to constrain the lv parameters .
these observations have developed and are developing rapidly within new platforms nowadays . until now ,
no explicit lv effects deviating from the special relativity are observed determinately , and many theoretical scenarios are ruling out .
meanwhile , new clues of lv physics are emerging rapidly from astrophysical observations and ground - based laboratory experiments . in the following dozens of years
, the lv issue is expected to meet a more clear situation then .
we thank helpful conversations with zhi xiao .
this work is partially supported by national natural science foundation of china ( nos . 11005018 , 10721063 , 10975003 , 11035003 ) .
it is also supported by hui - chun chin and tsung - dao lee chinese undergraduate research endowment ( chun - tsung endowment ) at peking university , and by national fund for fostering talents of basic science ( nos .
j0630311 , j0730316 ) .
l. maccione , s. liberati and g. sigl , _ phys .
lett . _ * 105 * , 021101 ( 2010 ) [ arxiv:1003.5468 [ astro-ph.he ] ] .
j. alfaro , _ phys .
* 94 * , 221302 ( 2005 ) [ arxiv : hep - th/0412295 ] .
j. alfaro , _ phys .
d _ * 72 * , 024027 ( 2005 ) [ arxiv : hep - th/0505228 ] .
z. xiao and b. q. ma , _ int .
j. mod .
phys . a _ * 24 * , 1359 ( 2009 ) [ arxiv:0805.2012 [ hep - ph ] ] .
k. murase , _ phys .
lett . _ * 103 * , 081102 ( 2009 ) [ arxiv:0904.2087 [ astro-ph.he ] ] .
w. bietenholz , arxiv:0806.3713 [ hep - ph ] .
d. mattingly , arxiv:0802.1561 [ gr - qc ] . v. a. kostelecky and m. mewes , _ phys .
lett . _ * 99 * , 011601 ( 2007 ) [ arxiv : astro - ph/0702379 ] . v. a. kostelecky and m. mewes , _ astrophys .
j. lett . _
* 689 * , l1 ( 2008 ) [ arxiv:0809.2846 [ astro - ph ] ] .
v. a. kostelecky and m. mewes , _ phys .
d _ * 80 * , 015020 ( 2009 ) [ arxiv:0905.0031 [ hep - ph ] ] .
j. alfaro and l. f. urrutia , _ phys .
d _ * 81 * , 025007 ( 2010 ) [ arxiv:0912.3053 [ hep - ph ] ] .
j. collins , a. perez , d. sudarsky , l. urrutia and h. vucetich , _ phys .
lett . _ * 93 * , 191301 ( 2004 ) [ arxiv : gr - qc/0403053 ] .
j. alfaro , a. a. andrianov , m. cambiaso , p. giacconi and r. soldati , _ int .
j. mod .
phys . a _ * 25 * , 3271 ( 2010 ) [ arxiv:0904.3557 [ hep - th ] ] . | lorentz violation ( lv ) is predicted by some quantum gravity ( qg ) candidates , wherein the canonical energy - momentum dispersion relation , @xmath0 , is modified .
consequently , new phenomena beyond the standard model are predicted . especially , the presence of lv highly affects the propagation of astrophysical photons with very high energies from distant galaxies . in this paper , we review the updating theoretical and experimental results on this topic .
we classify the effects into three categories : ( i ) time lags between photons with different energies ; ( ii ) a cutoff of photon flux above the threshold energy of photon decay , @xmath1 ; ( iii ) new patterns in the spectra of multi - tev photons and eev photons , due to the absorption of background lights .
as we can see , the details of lv effects on astrophysical photons depend heavily on the `` phase space '' of lv parameters . from observational aspects ,
available and upcoming instruments can study these phenomena hopefully , and shed light onto lv issues and qg theories .
the most recent progresses and constraints on the ultra - high energy cosmic rays ( uhecrs ) are also discussed . |
recent type ia supernova data @xcite confirm , at high significance , that the cosmological expansion is currently accelerating .
if we assume that general relativity is valid on cosmological length scales , this cosmological acceleration requires that the universe s current energy density budget is dominated by far by an approximately spatially uniform component dark energy with negative pressure @xmath0 , where @xmath1 is the dark energy density .
the most economic and the oldest form of dark energy is einstein s cosmological constant @xmath2 @xcite which is time independent and has an equation of state @xmath3 .
the time - independent cosmological constant model ( @xmath2cdm ) provides a reasonably good fit to most current cosmological data ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ) , but despite this success a lot of alternative models of dark energy have been proposed over the years .
one reason for this is that cosmological data can not yet tightly constrain the various options currently under debate , although it is thought that in next decade a large amount of more precise new data will tightly constrain or measure departures from the now standard @xmath2cdm model ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) .cdm model is based on the cold dark matter ( cdm ) structure formation model which might have trouble accounting for some observations ( for a summary of some of the problems see @xcite ) . ]
the second reason is that a cosmological constant is not straightforward to understand on a more fundamental level ; in particular , it is difficult to accept as fundamental the needed new energy scale of a few @xmath4 . in some scalar field
dark energy models a nonlinear attractor solution ensures that this scale of a few @xmath4 is not fundamental , rather it follows from a much higher energy scale as the scalar field energy density decreases during the cosmological expansion ( @xcite ) . in this paper we also consider this @xmath5cdm model in which the dark energy is modeled as a slowly rolling scalar field @xmath5 with self - interaction potential energy density @xmath6 , where @xmath7 .
a number of other models with time - varying dark energy have been proposed .
we also consider the xcdm parametrization which is often used to describe time - varying dark energy models.cdm dark energy at late times ( see , e.g. , @xcite ) . ] in xcdm dark energy is assumed to be a perfect fluid with effective equation of state @xmath8 , where @xmath9 is a number less than @xmath10 . given a model it is possible to compute quantities such as the hubble parameter @xmath11 and the angular diameter distance @xmath12 as a function of redshift @xmath13 .
since these quantities are observable ( or observational results depend on them ) , given perfect data it is possible in principle to compare observational results with theoretical predictions and determine which dark energy model provides a better fit to the data .
there are a few difficulties however .
theoretical predictions not only depend on dark energy model parameters , they also can depend on a number of other cosmological parameters , such as the energy density of non - relativistic matter or baryonic matter or radiation , the energy density perturbation spectral index , the total mass of neutrinos , the hubble constant , etc . while some of these parameters are directly measurable , some of them have to be simultaneously determined from the data .
this leads to degeneracies in cosmological parameter space . the other issue is possible systematic errors in cosmological observations , which are very difficult to trace .
it is therefore reassuring that the supernova data indication of dark energy is confirmed by cosmic microwave background ( cmb ) anisotropy data .
more precisely , assuming the cdm model of structure formation and assuming that dark energy does not evolve in time , cmb anisotropy data are consistent with negligible space curvature ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , so in conjunction with low measured non - relativistic matter density ( see , e.g. , * ? ? ?
* ) cmb anisotropy data also demand dark energy . to convincingly remove degeneracies and understand and cancel the effect of unwanted systematic errors it s necessary to have many independent cosmological tests .
other recently discussed cosmological probes include the angular size of radio sources and quasars as a function of redshift ( see , e.g. , @xcite ) , strong gravitational lensing ( see , e.g. , @xcite ) , weak gravitational lensing ( see , e.g. , @xcite ) , measurements of the hubble parameter as a function of redshift ( see , e.g. , @xcite ) , galaxy cluster gas mass fraction versus redshift data ( see , e.g. , @xcite ) , and the growth of large - scale structure ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
a measurement of the large - scale structure baryon acoustic oscillation ( bao ) peak length scale provides another possible cosmological test .
this was first measured a few years ago in the two - point correlation function of luminous red galaxies in the sdss catalog at @xmath14 ( @xcite , also see @xcite ) and from 2dfgrs data @xcite , and then later in the joint sdss and 2df catalogs at @xmath15 and @xmath16 @xcite .
one attractive feature of the bao measurements is that they do not explicitly depend on the value of hubble constant .
these bao measurements were used in the original papers to constrain parameters of the @xmath2cdm and xcdm models . see @xcite , @xcite , @xcite , and @xcite , and references therein , for recent discussions of bao data constraints on these and other models .
the @xcite measurement was made at only two redshifts and by itself does not provide a robust and reliable test of dark energy models , but in combination with other data it does provide useful constraints on cosmological parameters .
however , a number of surveys are planned in next few years that will measure the bao scale accurately and at a variety of redshifts up to @xmath17 .
this upcoming bao data , especially when combined with other data , will prove very useful in tightly constraining dark energy parameters ( see , e.g. , * ? ? ?
recently @xcite argued that these measurements of the bao scale were essentially measurements orthogonal to the line of sight and so statistically independent from a line of sight measurement of the bao scale , even if the same galaxy catalog is used for both measurements .
@xcite used the sdss data to compute the line of sight or radial bao scale in redshift space for two ranges of redshift and @xcite showed the resulting constraints on the spatially - flat xcdm model .
these constrains from the radial bao scale data are quite similar to the constraints derived from earlier `` non - radial '' bao scale measurements . in this paper
we extend the analysis to the @xmath2cdm and @xmath5cdm models of dark energy.cdm parameters but do not show constraint contours for @xmath2cdm .
] the data are consistent with spatially - flat @xmath2cdm . however , these current radial bao measurements ( like current non - radial bao measurements ) can not tightly constrain time - varying dark energy , although the situation is anticipated to improve in the next few years .
we also derive constraints on these models from a combined analysis of the radial bao data and the @xcite union type ia supernova data .
since the radial bao and type ia supernova data constraints are almost orthogonal , the constraints from the combined data are significantly tighter than those from either individual data sets .
the combined constraints favor a close to spatially - flat @xmath2cdm model ( more so than the type ia supernova data ) , but do not yet completely rule out time - varying dark energy .
our paper is organized as follows . in sec . 2
we describe the radial bao measurements we use . in sec .
3 we explain how we derive constraints on different dark energy models from radial bao and type ia supernova data .
we present and discuss our results in sec .
in a spherically symmetric universe the two - point correlation function is a function of two variables , @xmath18 , where @xmath19 is the separation along the line of sight and @xmath20 is the separation on the sky .
it can also be expressed as a function of absolute separation @xmath21 and the cosine of the angle between the line of sight and the direction of separation , @xmath22 .
the correlation function can then be decomposed into multipole moments @xmath23 where @xmath24 is the @xmath25 order legendre polynomial .
multipole moments of different orders can be related to each other if one has a complete theory of linear and nonlinear evolution .
although high multipoles that describe the `` shape '' of baryon acoustic oscillation imprints on the matter distribution are very difficult to measure in practice , theoretically they are independent of the monopole and could provide additional structure formation tests .
initial work considered only the averaged over direction monopole part of the correlation function , @xmath26 and found a bao peak signal at a comoving distance of @xmath27 ( @xmath28 is the hubble constant @xmath29 in units of 100 @xmath30 ) .
this measurement was however mostly transverse to the line of sight direction @xmath20 ; the weight of separation along the line of sight contributes less then @xmath31 @xcite .
consequently , it is fair to assume that the radial baryon acoustic peak scale measurement in the line of sight direction from @xmath32 is statistically independent from that measured from @xmath33 , even if the same galaxy sample is used for both measurements .
@xcite recently used sdss data to measure the radial baryon acoustic scale in two redshift ranges @xmath34 with radial bao peak scale @xmath35 and @xmath36 with @xmath37 ( both one standard deviation errors ) .
theoretically the radial bao peak scale is given by @xmath38 where @xmath11 is the hubble parameter at redshift @xmath13 , @xmath39 is the sound horizon size at the drag redshift @xmath40 , at which baryons were released from photons , and @xmath41 is the speed of light .
@xmath11 can be easily computed in a given cosmological model and depends on model parameters such as the non - relativistic matter density and the time dependence of dark energy . to compute @xmath42 @xcite use two different methods .
one is to use the ratio @xmath43 between the distance to the last - scattering surface and @xmath42 measured by cmb anisotropy experiments and compute the sound horizon at photon decoupling from @xmath44 here @xmath45 is the redshift at the photon decoupling and @xmath46 is the angular diameter distance .
alternatively , one can use priors on the fractional energy density parameters of baryonic matter , @xmath47 , nonrelativistic matter , @xmath48 , and relativistic matter , @xmath49 , from , e.g. , cmb anisotropy measurements , and compute the sound horizon at the drag redshift from @xmath50 both options have similar drawbacks .
one has to assume priors on `` nuisance '' parameters like @xmath43 or various energy densities .
cmb anisotropy measurements themselves have measurement errors that must be accounted for , otherwise the errors on the estimates of dark energy model parameters of interest will be underestimated .
also , the best fit values for nuisance parameters given by cmb anisotropy data are in general different for every cosmological model and also depend on model parameter values .
to be fully consistent when using priors one would have to reanalyze cmb experiments for each cosmological model ( and model parameter value ) instead of using a single set of values for @xmath43 , @xmath47 , @xmath48 , and @xmath49 . at present , however , the bao scale is measured only in two redshift ranges and does not provide very tight parameter constraints compared to other observational tests . hence , as long as we are interested in preliminary constraints on dark energy from bao scale measurements we may use the simplified approach of @xcite , keeping in mind that when more and better quality bao scale measurements become available a more complete , careful , and time - consuming analysis will be warranted .
we compare the two measured radial bao peak scales @xcite with the predictions of three dark energy models .
the models we consider are standard @xmath2cdm , the xcdm parametrization of the dark energy s equation of state , and the @xmath5cdm model with an inverse power law potential energy density @xmath6 . in these dark energy models , at late times , we can compute the redshift - dependent hubble parameter from @xmath51 in the @xmath52 model the energy density of the scalar field has to be computed as a function of the redshift and @xmath53 parameter by numerically solving the equations of motion . in all three models
the background evolution is described by two parameters .
one is the nonrelativistic matter fractional energy density parameter @xmath48 and the other one is a parameter @xmath54 that characterizes the dark energy . for the @xmath2cdm model @xmath54
is the cosmological constant fractional energy density parameter @xmath55 , in the xcdm parametrization it is the equation of state parameter @xmath9 , and in the @xmath56 model it is @xmath53 which describes the steepness of the scalar field self - interaction potential . in this paper
we consider only spatially - flat xcdm and @xmath5cdm models , while in the @xmath2cdm case spatial curvature is allowed to vary , with the space curvature fractional energy density parameter @xmath57 .
the angular diameter distance is determined in terms of the hubble parameter through @xmath58}\quad & ( \omega_{\rm k}>0),\\ & d_a(z)=\frac{1}{(1+z)}\int_0^z{\frac{dz'}{h(z')}}\quad & ( \omega_{\rm k}=0),\\ & d_a(z)=\frac{1}{\sqrt{-\omega_{\rm k}}h_0(1+z)}\sinh{\left[\sqrt{-\omega_{\rm k}}h_0\int_0^z{\frac{dz'}{h(z')}}\right]}\quad & ( \omega_{\rm k}<0).\end{aligned}\ ] ] we use both methods proposed in @xcite to estimate the sound horizon at recombination .
we compute theoretically predicted model values of the radial baryon acoustic scale in redshift space , @xmath59 , at redshift @xmath60 and @xmath61 , where by @xmath62 we denote the `` nuisance '' parameters .
we assume that the measurements at @xmath15 and @xmath16 are independent and that the errors are gaussianly distributed . to constrain model parameters
we compute @xmath63 where the two observed values , @xmath64 , are listed above eq .
( 3 ) and define a likelihood function @xmath65 we integrate over nuisance parameters to get a two dimensional likelihood for the cosmological parameters of interest @xmath66 . for the prior likelihood of nuisance parameters we use gaussian distribution functions with wmap 5-year recommended means and variances @xmath67 , @xmath68 ( with @xmath69 positive correlation between two measurements ) , @xmath70 , @xmath71 , and @xmath72 @xcite .
for each dark energy model we define the best - fit values of parameters as a pair ( @xmath48 , @xmath54 ) that maximizes the two - dimensional likelihood function @xmath73 .
the 1 , 2 , and 3@xmath19 confidence level contours are defined as the sets of parameters ( @xmath48 , @xmath54 ) for which @xmath73 is less than its maximum value by multiplicative factors of @xmath74 , @xmath75 , and @xmath76 , respectively . in figs .
13 we show two - dimensional constraints on cosmological parameters from radial bao data for each of three dark energy models and two sets of priors . since radial bao measurements alone , at the moment , can not tightly constrain cosmological parameters we combined them with the @xcite union type ia supernova data . in our derivation of the type ia supernova constraints we closely follow @xcite .
we define a likelihood function for cosmological parameters , @xmath77 , for each dark energy model .
we marginalize over @xmath29 with a flat noninformative prior in the range @xmath78 to get a two - dimensional likelihood @xmath79 .
the constraints on the @xmath5cdm model from the supernova data are shown in fig .
4.cdm and xcdm models from the supernova `` union '' data set see fig . 12 in @xcite .
] since radial bao and supernova measurements are independent , we multiply the individual likelihoods to get a joint likelihood function @xmath80 .
we show constraints derived from the joint likelihood function in figs .
57 . from the joint two - dimensional likelihood function
we define the marginal one - dimensional likelihood functions for parameters @xmath48 and @xmath54 ( assuming uniform priors on both parameters in the ranges @xmath81 , @xmath82 , @xmath83 , and @xmath84 ) through @xmath85 in fig .
8 we show one - dimensional likelihood functions for @xmath48 and @xmath54 for the three dark energy models . from these one - dimensional likelihood functions , in all three models , we define the best - fit value of parameters as the values that maximize the likelihood , with highest posterior density 1@xmath19 confidence level intervals as the value of parameters for which @xmath86 and with @xmath87 higher everywhere inside the interval than outside .
the values of one - dimensional best - fit parameters and corresponding 1@xmath19 confidence level intervals are listed in table 1 .
figure 1 shows the @xcite radial bao scale constraints on the @xmath2cdm model .
when we use the wmap measured value of the ratio @xmath43 ( thick lines ) the model is constrained to be very close to the spatially - flat case .
when we use the wmap measured value of the sound horizon at recombination @xmath42 ( thin lines ) the constraints are weaker , the spatial curvature is not well constrained , and the nonrelativistic matter @xmath48 has to be less than @xmath88 at about 3@xmath19 .
the contours in fig .
1 are in reasonable accord with those shown in fig . 2 of @xcite which were derived using the `` non - radial '' bao peak scales measured by @xcite and @xcite .
when the radial bao data is combined with supernova data the resulting constraints are significantly stronger , see fig .
the nonrelativistic matter density parameter is now constrained to be @xmath89 at about 3@xmath19 confidence level while the cosmological constant density parameter lies in the @xmath90 range .
the best - fit values are close to the spatially - flat model .
the constraints computed from the joint one - dimensional likelihoods shown in figs .
8a and 8b are listed in table 1 .
they are much more restrictive than those derived from the individual data sets , because those constrain combinations of dark energy parameters that are in some sense `` orthogonal '' in parameter space . for the xcdm parametrization ( for which we consider only spatially - flat models ) the confidence level contours derived from radial bao measurements are broad and a range of @xmath48 and @xmath9 values are acceptable .
the constraints are shown in fig .
they are similar to the results shown in ( * ? ? ?
* fig . 1 ) .
when we use the wmap measured value of the ratio @xmath43 ( thick lines ) the nonrelativistic matter @xmath48 has to be less than @xmath91 at about 3@xmath19 .
these confidence level contours are in reasonable accord with the ones shown in ( * ? ? ?
3 ) which were derived using the non - radial bao data of @xcite and @xcite .
however , compared to the non - radial bao scale measurements the radial bao scale measurements better constrain @xmath9 from below and tend to favor higher values of it .
the joint constraints from radial bao and supernova data are shown in fig .
the joint likelihood constrains the equation of state parameter to be @xmath92 at about 3@xmath19 confidence , while @xmath93 . for xcdm the joint constraints
do not depend much on the method used for the analysis of the radial bao data , unlike the @xmath2cdm and @xmath5cdm cases .
one - dimensional joint likelihood functions for @xmath48 and @xmath9 are shown in figs .
8c and 8d and the corresponding constraints are given in table 1 .
the spatially - flat @xmath5cdm model confidence level contours are shown in fig .
3 . here , the radial bao measurements constrain @xmath48 to be between @xmath94 and @xmath91 at about 3@xmath19 , but the @xmath53 parameter is not constrained well and large values of @xmath53 ( relatively rapidly evolving dark energy ) are not ruled out , although the likelihood peaks at @xmath95 .
these results are similar to the ones derived in using the non - radial bao peak scale measurements of @xcite and @xcite .
figure 6 shows type ia supernova constraints on @xmath5cdm . from the union data set alone , the nonrelativistic matter energy density parameter @xmath96 , and if @xmath97 , @xmath98 at about 3@xmath19 confidence .
these contours are similar to and a little more constraining than the ones derived using the @xcite type ia supernova gold data ( * ? ? ?
* fig . 1 ) .
when we combine the radial bao measurements with the union supernova data , fig . 7 ,
the @xmath53 parameter is constrained to be less than 1.5 at 3@xmath19 while @xmath93 .
one - dimensional joint likelihood functions for @xmath48 and @xmath53 are shown in figs .
8e and 8f and the corresponding constraints are listed in table 1 . for both radial bao analysis methods
@xmath48 is relatively well constrained and the likelihood for @xmath53 peaks at low values of @xmath53 , close to the corresponding spatially - flat @xmath2cdm values . from the analysis presented here and in @xcite
, it is clear that neither the current radial or non - radial bao peak scale measurements by themselves are able to tightly constrain cosmological parameters .
however , given the shape of the resulting confidence contours , typically an elongated ellipse - like shape , current bao data can provide very useful constraints when combined with other data like type ia supernova measurements . as another example , the constraints derived from radial and non - radial bao peak measurements complement the constraints derived from galaxy cluster gas mass fraction data @xcite and when used jointly they can tightly constrain parameters of dark energy models @xcite . it
is interesting to note that for the @xmath2cdm and xcdm models the best fit values from the bao and galaxy cluster gas mass fraction data sets are about 3@xmath19 apart ( see * ? ? ?
23 ) , but for the two data sets together the minimum @xmath99 is acceptable .
this is , most probably , due to unknown systematic errors in one ( or both ) of the data sets and deserves further investigation .
it is also of interest to compare the constraints shown in figs .
13 to those derived using other data . the radial bao constraints shown in figs .
13 are more restrictive than the ones derived from hubble parameter versus redshift data @xcite and from gravitational lensing data @xcite .
however , they are comparable to the ones derived from earlier type ia supernova data @xcite by @xcite . current bao peak scale data is sparse and can not tightly constrain dark energy parameters or differentiate between different dark energy models .
in next few years a number of planned surveys like pau @xcite , boss , and wigglez @xcite should measure the bao peak scale with @xmath31 accuracy possibly up to a redshift of @xmath17 .
these future bao measurements should prove to be of great significance in discriminating between different currently viable dark energy models and constraining cosmological parameters .
c c c c c @xmath100 & @xmath101 & @xmath102 & @xmath103 & @xmath104 + @xmath105&@xmath106 & @xmath107 & @xmath108 & @xmath109 + @xmath110 & @xmath111 & @xmath112 & @xmath113&@xmath114 + @xmath115 & @xmath106 & @xmath116 & @xmath117&@xmath118 + @xmath119 & @xmath111 & @xmath120 & @xmath121&@xmath122 + @xmath123 & @xmath106 & @xmath124 & @xmath125&@xmath126 + confidence level contours for the @xmath2cdm model from radial bao measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath127 and @xmath128 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath129 and @xmath128 , derived from the two - dimensional likelihood function .
the dotted line demarcates spatially - flat @xmath2cdm models.,width=680,height=680 ] confidence level contours for the xcdm parametrization from radial bao measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath130 and @xmath131 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath130 and @xmath132 , derived from the two - dimensional likelihood function .
the dotted horizontal line demarcates spatially - flat @xmath2cdm models.,width=680,height=680 ] confidence level contours for the @xmath52 model from radial bao measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath129 and @xmath133 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath134 and @xmath133 , derived from the two - dimensional likelihood function .
the horizontal axis with @xmath95 corresponds to spatially - flat @xmath2cdm models.,width=680,height=680 ] confidence level contours for the @xmath5cdm model from the supernova data .
the cross shows best - fit model parameters , @xmath135 and @xmath133 , derived from the two - dimensional likelihood function .
the horizontal axis with @xmath95 corresponds to spatially - flat @xmath2cdm models.,width=680,height=680 ] confidence level contours for the @xmath2cdm model derived using the joint likelihood from radial bao and type ia supernova measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath136 and @xmath103 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath137 and @xmath108 , derived from the two - dimensional likelihood function .
the dotted line demarcates spatially - flat @xmath2cdm models .
the @xmath48 and @xmath55 ranges shown here are smaller than those shown in fig .
1.,width=680,height=680 ] confidence level contours for the xcdm parametrization derived using the joint likelihood from radial bao and type ia supernova measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath134 and @xmath113 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath137 and @xmath138 , derived from the two - dimensional likelihood function .
the dotted horizontal line demarcates spatially - flat @xmath2cdm models .
the @xmath48 and @xmath9 ranges shown here are smaller than those shown in fig .
2.,width=680,height=680 ] confidence level contours for the @xmath52 model derived using the joint likelihood from radial bao and type ia supernova measurements .
thick lines are derived using the wmap measured ratio @xmath43 and the cross shows best - fit model parameters , @xmath134 and @xmath121 , derived from the two - dimensional likelihood function .
the thin lines are derived using the wmap value for the sound horizon at recombination @xmath42 and the circle shows best - fit model parameters , @xmath137 and @xmath125 , derived from the two - dimensional likelihood function .
the horizontal axis with @xmath95 corresponds to spatially - flat @xmath2cdm models .
the @xmath48 and @xmath53 ranges shown here are smaller than those shown in figs . 3 and 4.,width=680,height=680 ] | we use the radial baryon acoustic oscillation ( bao ) measurements of @xcite to constrain parameters of dark energy models .
these constraints are comparable with constraints from other non - radial " bao data . the radial bao data are consistent with the time - independent cosmological constant model but do not rule out time - varying dark energy .
when we combine radial bao and the @xcite union type ia supernova data we get very tight constraints on dark energy . |
gauge theories at finite temperatures have been explored intensively by means of monte carlo simulations on a lattice . for yang - mills theories without fermions
many calculations have been done for different gauge groups , see for example @xcite .
a phase transition has been found , separating a low - temperature phase with confinement of static quarks from a high - temperature deconfined phase . in full qcd , including also up , down , and strange quarks , a crossover separates the confined nuclear matter phase at low temperatures from the quark - gluon plasma in the high temperature regime @xcite .
the realisation of chiral symmetry in qcd is another temperature dependent phenomenon . at low temperatures
chiral symmetry is broken , while it is restored at high temperatures .
this provides another ( pseudo-)critical temperature .
recent numerical investigations have shown that the restoration of chiral symmetry takes place near the deconfinement transition @xcite .
the physical relation between the two critical temperatures remains , however , unclear due to the lack of an exact order parameter @xcite . for supersymmetric yang - mills theory ( sym )
there are only few non - perturbative results about its behaviour at finite temperatures .
a great interest in the subject comes from the application of the ads / cft conjecture @xcite to the description of the deconfinement transition of qcd .
the ads / cft conjecture is a duality between low - energy string theory in ten dimensions and strong coupling @xmath1 sym in four dimensions @xcite .
@xmath1 sym is a conformal theory and therefore reductions are needed in order to relate the results to a theory like qcd with a mass - gap .
finite temperature is a possibility to break both supersymmetry and conformal invariance of @xmath1 sym @xcite and therefore it could be possible that many fundamental properties are shared between the weakly interacting quark - gluon plasma and supersymmetric models at finite temperature .
supersymmetric models at finite temperatures have a different behaviour than other models , due to the difference between the thermal statistic of bose and fermi particles . in the euclidean time direction
periodic and anti - periodic boundary conditions must be imposed on fermionic and bosonic fields , respectively . at zero temperature , in the infinite volume limit , this difference can be neglected and exact susy can be formulated consistently . at finite temperatures ,
the temporal direction is compactified and boundary conditions will break the supersymmetry between fermions and bosons @xcite . therefore there is no high temperature limit in which a possible spontaneously or explicitly broken supersymmetry can be effectively restored @xcite .
this intriguing property was subject of many studies in the past , in particular for understanding the nature and the pattern of this temperature induced susy breaking , see @xcite for a review .
supersymmetry opens the possibility to study the relation between the deconfinement transition and chiral symmetry restoration .
dual gravity calculations proved that confinement implies chiral symmetry breaking for a class of supersymmetric yang - mills theories @xcite .
the object of our investigations is @xmath2 sym at finite temperatures .
this theory describes the strong interactions between gluons and their superpartners , the gluinos , which are majorana fermions in the adjoint representation of the gauge group . at zero temperature
the theory is in a confined phase and chiral symmetry is spontaneously broken by a non - vanishing expectation value of the gluino condensate .
this theory has been subject of intensive theoretical investigations .
relations between sym and qcd have been found in terms of the orientifold planar equivalence @xcite .
they have lead to conjectures about sym relics in qcd @xcite .
@xmath0 sym has also a crucial role in the context of the gauge / gravity duality of the @xmath3 theory @xcite .
numerical simulations of @xmath2 sym are possible with the monte carlo methods @xcite and they provide an important non - perturbative tool for exploring the phase diagram at finite temperatures .
a mass term for the gluinos is added in our numerical simulations and the results are extrapolated to the chiral limit .
the gauge group chosen is su(2 ) .
the results obtained show clearly that deconfinement occurs at a temperature which decreases with decreasing gluino mass .
the distribution of the order parameter and the finite size scaling support the possibility that the order of the associated phase transition is the same for a pure gauge theory and its supersymmetric extension , at least for the range of masses considered .
possible scenarios for the relation between chiral symmetry breaking and deconfinement are also discussed .
the chiral symmetry is found to be restored near the temperature of the deconfinement phase transition , even if it requires a more careful extrapolation to the chiral limit .
a chiral phase transition of the same order of the deconfinement transition is argued considering the general symmetries of the model .
the @xmath2 sym theory is the supersymmetric extension of pure gauge theory .
the model is constructed imposing su(@xmath4 ) gauge invariance and a single conserved supercharge , obeying the algebra @xmath5 where the generators of the supersymmetry @xmath6 are majorana spinors , @xmath7 is the charge conjugation matrix and @xmath8 the momentum operator .
the theory contains gluons as bosonic particles , and gluinos as their fermionic superpartners .
the gluino is a spin- majorana fermion in the adjoint representation of the gauge group .
a majorana fermion obeys the `` reality '' condition @xmath9 supersymmetry relates the gauge fields @xmath10 and gluino fields @xmath11 : @xmath12 where @xmath13 is a global majorana fermion , parametrising the transformation .
the euclidean on - shell action for @xmath2 sym theory in the continuum is @xmath14 the @xmath15-term can be added to the action as in qcd without violating the underlying symmetries of the model .
the operator @xmath16 is topologically invariant and the theory is periodic in the parameter @xmath15 , i. e.@xmath15 and @xmath17 are equivalent . in the following @xmath18
will be assumed .
the additional parameter @xmath19 introduces a bare mass for the gluino .
this mass in the fermionic sector breaks supersymmetry softly , i. e. this kind of breaking guarantees that the main features of the supersymmetric theory , concerning the ultraviolet renormalisability , remain intact . at zero temperature
, gluons and gluinos can be found only in colourless bound states .
those bound states are expected to form supermultiplets of equal masses if exact supersymmetry is realised .
a low - energy effective lagrangian has been formulated @xcite , predicting a bound spectrum of mesons , glueballs and gluino - glueballs , which has been subject of many numerical lattice investigations @xcite .
on the lattice , the gauge fields @xmath20 are associated with the links of the lattice using the exponential map @xmath21 where @xmath22 are the lie group generators in the representation @xmath23 . in the following
, @xmath24 and @xmath25 will denote the link variables in the fundamental and in the adjoint representation , respectively .
the adjoint links @xmath25 are related to the fundamental links @xmath24 through the well - known formula @xmath26 in our investigations the gauge group is su(2 ) , therefore @xmath27 su(2 ) and @xmath28 so(3 ) .
the generators in the fundamental representation are normalised such that : @xmath29 in our simulations , the gauge part of @xmath30 in eq .
( [ continuum_action ] ) is discretised with a tree - level symanzik improved action : @xmath31 where @xmath32 is the standard plaquette term formed out of four links , and @xmath33 represents a rectangle with lower left corner on the point @xmath34 .
the gluino part of @xmath30 in eq .
( [ continuum_action ] ) is represented on the lattice using the discretised version of the dirac operator , depending on the links in the adjoint representation @xmath25 : @xmath35(y , x ) \lambda(x).\ ] ] the action of the wilson - dirac operator @xmath36 on the gluino field @xmath37 is given by ( dirac and colour indices suppressed ) @xmath38 where @xmath39 is the hopping parameter .
supersymmetry and chiral symmetry are explicitly broken using this discretisation scheme .
euclidean invariance is explicitly broken on the lattice and therefore it is impossible to construct a local action invariant under supersymmetry transformations for finite lattice spacing @xmath40 @xcite . a fine tuning is needed to recover the broken supersymmetry and chiral symmetry in the continuum limit . in supersymmetric yang - mills theory the tuning of a single parameter , namely the bare gluino mass @xmath19 , is enough to recover both symmetries @xcite .
the tuning to the chiral limit can be defined by the vanishing of the adjoint pion mass , which is defined in a partially quenched setup @xcite .
we use it here to define different lines of constant physics for theories with a softly broken supersymmetry .
the @xmath0 sym is an asymptotically free theory , expected to behave at high temperatures as a conformal gas of free gluons and gluinos @xcite . at zero temperature confinement and gluino
condensation take place .
the possible phases are characterised by the expectation value of their related order parameters considered as a function of the temperature .
a useful order parameter for the deconfinement transition is the polyakov loop @xmath41 the expectation value of the polyakov loop has the physical meaning of the exponential of the negative free energy of a single static dirac quark in the fundamental representation @xmath42 therefore a non - vanishing value of @xmath43 means that a state with a single isolated quark exists , i. e. deconfinement .
deconfinement is associated with the spontaneous breaking of the center symmetry , defined by the following transformation of the gauge fields in a fixed time - slice at @xmath44 : @xmath45 in contrast to qcd with fermions in the fundamental representation , this transformation leaves invariant both the gauge and the fermionic part of the action of @xmath2 sym .
the wilson - dirac operator is written in terms of links in the adjoint representation that are unaffected by the complex rotation .
on the other hand , the polyakov loop transforms non - trivially under the center transformations : @xmath46 it is thus an exact order parameter for the deconfinement transition at any value of the gluino mass @xmath19 .
the pattern for the center symmetry breaking is hence the same as in pure su(@xmath4 ) gauge theories , and it is possible that the svetitsky - yaffe conjecture @xcite is valid for @xmath2 sym .
this conjecture implies a deconfinement transition of second order for the gauge group su(2 ) , corresponding to the universality class of the @xmath47 ising model in three dimensions .
sym with gauge group su(2 ) .
the chiral condensate @xmath48 is the order parameter of the chiral phase transition .
each non - zero value of the renormalised gluino mass @xmath49 introduces a source of chiral symmetry breaking . at low temperatures , moving from positive to negative gluino mass @xmath49 , the chiral condensate jumps from a positive to a negative expectation value .
the phase transition is therefore of first order with a coexistence of two phases .
the line of first order phase transitions is expected to terminate in a second order endpoint .
[ chiral_phase_diagram],scaledwidth=53.0% ] the @xmath0 sym has a classical u(1)@xmath50 axial symmetry , meaning that the transformation @xmath51 leaves the action invariant when the gluino mass is exactly zero .
this symmetry is known as the r - symmetry u(1)@xmath52 and it corresponds to the relative rotation of the left- and the right - handed weyl components of the gluino field @xmath37 .
beyond the classical level , quantum fluctuations break chiral symmetry by a term proportional to the gauge coupling and to the number of colours : @xmath53 the dependence on @xmath4 is absent in qcd , and it is typical of gauge models with fermions in the adjoint representation . the anomalous contribution to the axial transformations can be absorbed in the periodicity of the parameter @xmath15 : @xmath54 if the angle @xmath55 assumes one of the values @xmath56 , @xmath57 .
the remaining chiral symmetries thus from the group @xmath58 .
numerical investigations @xcite have confirmed the conjecture @xcite that this invariance is spontaneously broken at zero temperature by a non - vanishing expectation value of the gluino condensate @xmath59 to a remaining @xmath47 symmetry corresponding to the sign flip @xmath60 .
the complete pattern of chiral symmetry breaking is thus @xmath61 the phase transition associated with the spontaneous breaking @xmath62 is of first order at zero temperature , related to the jump of the expectation value of the chiral condensate .
the system is in this respect similar to a @xmath63 ising model ( @xmath64 ) , with the gluino condensate corresponding to the spontaneous magnetisation and the renormalised gluino mass to the external magnetic field .
this similarity suggests that for the gauge group su(2 ) there is a critical temperature @xmath65 of a second order phase transition and a phase with restored @xmath66 symmetry at high temperatures , see fig .
[ chiral_phase_diagram ] .
there are three possible scenarios for the relation of deconfinement and chiral symmetry restoration
. @xmath65 might coincide with the deconfinement transition temperature @xmath67 , but there is no restriction to this scenario from first principles .
if they do not coincide either a mixed deconfined phase with broken chiral symmetry or a mixed confined phase with restored chiral symmetry exists , see fig .
[ scenarios ] . in addition , the remaining part of the u(1)@xmath52 symmetry broken by the anomaly could be effectively restored in high temperature limit .
in order to perform monte carlo simulations of @xmath0 sym , the gluino field is integrated out in the path integral . for majorana fermions
the result is the pfaffian of the wilson - dirac operator @xmath68 the pfaffian of an antisymmetric matrix is related to the square root of the determinant by @xmath69 the additional factor leads to the notorious sign problem of this theory . at a fixed lattice spacing ,
configurations with a negative pfaffian sign can appear .
this happens in particular at small residual gluino masses close to the supersymmetric limit .
these contributions are reduced moving to smaller lattice spacings and the pfaffian is strictly positive in the continuum limit .
it is hence possible to stay in the region where the sign problem is irrelevant . on the other hand , a reliable extrapolation to the supersymmetric limit
requires small gluino masses , and negative pfaffian signs can not be excluded .
we have monitored the pfaffian signs for the runs with the most critical parameters using the method introduced in @xcite to keep this effect under control .
our simulations have been performed using the hybrid monte carlo algorithm ( hmc ) .
we have applied two different approaches for the approximation of the square root of the determinant : an exploratory study was done with a code based on the polynomial ( phmc ) approximation ; the second , and main part of the work , was performed instead using a new code , based on the rational ( rhmc ) approximation .
the phmc algorithm is used with one - level stout links in the wilson - dirac operator , while the rhmc is used with unsmeared links . when the renormalised gluino mass is sent to zero ,
the wilson - dirac operator becomes ill - conditioned and the computational demand for the numerical integration of the classical trajectory in the hmc increases drastically .
the most reliable approach is therefore to perform the simulations for several non - zero values of the gluino mass and obtain the supersymmetric limit by extrapolation of the results .
the phase diagram of @xmath0 sym theory is investigated on lattices of finite size @xmath70 for different values of the bare couplings @xmath71 and @xmath72 . the boundary conditions are anti - periodic in the euclidean time direction for fermions and periodic in all other cases . in order to convert the bare parameters into physical units the size of the lattice spacing in physical units is needed .
this is done by means of the sommer parameter @xmath73 @xcite and the @xmath74 parameter @xcite .
the calculation of the scale is based on results of simulations at zero temperature . at fixed @xmath72 and @xmath71
the temperature is proportional to the inverse of the number of lattice sites in the temporal direction @xmath75 , @xmath76 the continuum limit is obtained when @xmath77 at fixed @xmath78 .
the phase transitions are determined as a function of the renormalised value of the residual gluino mass @xmath49 that breaks supersymmetry softly .
the extrapolation of the transition temperatures to the supersymmetric limit @xmath79 is the final result of our calculation .
it has been shown in a partially quenched setup that @xmath49 is proportional to the square of the adjoint pion mass @xmath80 @xcite . for several critical couplings @xmath81 and
@xmath82 of the deconfinement transition the adjoint pion mass is measured in zero temperature simulations on lattices of size @xmath83 .
these results are summarised in table [ tablemeaszero ] . at small values of @xmath80
the dependence of @xmath49 on @xmath84 for fixed @xmath72 is approximately linear . in cases , where zero temperature results at @xmath81 and different values of @xmath71 were present from previous investigations
we apply a linear fit to interpolate @xmath49 at @xmath82 . in the continuum
the scaling function is given by the novikov - shifman - vainshtein - zakharov beta - function @xcite to all orders in perturbation theory . from this beta - function
the one - loop perturbative scaling of the lattice spacing as a function of the coupling @xmath72 in the supersymmetric limit is given by @xmath85 however , at finite lattice spacings it is more feasible to consider a non - perturbative scale setting based on measurable scale parameters .
we consider two different observables for the scale setting .
the first one is the sommer parameter @xmath86 obtained from the static quark - antiquark potential @xcite . since in sym there is no string breaking for static quarks in the fundamental representation , this scale can be measured in the same way as for pure yang - mills theory .
the second observable is the recently proposed alternative @xmath87 @xcite .
this observable is obtained from the gradient flow of the gauge action density @xmath88 with @xmath89 represented by clover plaquettes , @xmath90 the gradient flow is defined as continuous smearing procedure using , in our case , the functional derivative of the wilson plaquette action .
the scale parameter @xmath87 is defined by the flow time @xmath91 , where @xmath92\right|_{t = w_0 ^ 2}=0.3\ ; .\ ] ] the dependence of the observable on the flow time is shown in fig .
[ wilson_flow_integrated ] .
as expected , the dependence of @xmath93 on @xmath91 is approximately linear for large @xmath91 .
an advantage of @xmath86 and @xmath87 is their weak dependence on the residual gluino mass . in both cases
a mild , but not negligible , linear dependence is observed , see fig .
[ w0_mass_dependence ] . as a consequence
two different approaches can be applied to set the scale : in a _ mass dependent _ renormalisation scheme the scale is set separately at each value of @xmath49 and consequently the lattice spacing depends on the gluino mass , @xmath94 in the second approach the lattice spacing is taken to be independent of the gluino mass @xmath49 , @xmath95 therefore , in this case the linear behaviour of @xmath87 and @xmath86 is interpreted as a physical dependence of their value on the fermion mass @xcite .
this approach is called _ mass independent _ renormalisation scheme and it requires an extrapolation of @xmath86 and @xmath87 to a fixed reference value of @xmath49 , as shown in fig .
[ w0_mass_dependence ]
. extrapolated to the chiral limit ( @xmath96 ) .
[ fig : scalesetting1lstout],scaledwidth=70.0% ] in our exploratory study , where the phmc approximation with one - level of stout smearing has been used , the mass independent approach has been applied . from previous investigations and
@xmath97 will be presented in a paper which is in preparation.[note1 ] ] we have already some zero temperature results at @xmath98 , @xmath99 , @xmath100 , and @xmath101 , see table [ betar0values ] . in these studies
the sommer parameter @xmath86 has been extrapolated to the value @xmath102 at the supersymmetric limit , corresponding to a reference scale of @xmath103 .
we have completed the data with additional simulations at @xmath104 , see table [ tablemeasstout2 ] .
the dependence of the scale @xmath102 on @xmath72 is fitted with a similar parametrisation as used in @xcite , @xmath105 due to the limited amount of data the error of the coefficients @xmath106 , @xmath107 , @xmath108 , and @xmath109 is not reliably obtained from a single fit . to get a better estimate we have assembled several samples of data points by taking values within the given error bounds of each point . with the fits of these samples
one obtains a set of curves that determines the error bound of the interpolation , see fig .
[ fig : scalesetting1lstout ] .
this result determines the values of @xmath102 for the @xmath72 values without enough data from zero temperature simulations . in the second part of the work , where the simulations have been performed with the rhmc algorithm and without stout smearing ,
the scale has been set by the parameter @xmath74 and both the mass dependent and independent schemes . in the mass independent scheme @xmath74
is extrapolated to the chosen reference point @xmath110 , see fig .
[ w0_mass_dependence ] .
this reference point can be accurately extrapolated already from a small number of zero temperature simulations .
the obtained value @xmath111 is used to fix the scale for all the simulations with the same value of @xmath72 . in order to exclude possible systematic errors with this approach
we have also applied a mass dependent scale setting prescription , see table [ tablemeaszero ] . in that case
the scale at @xmath72 and @xmath71 is set with the value of @xmath74 measured at the same combination of the parameters in a zero temperature simulation .
we have done the first scan of the phase diagram of supersymmetric yang - mills theory with the same parameters and settings as in our zero temperature studies of the particle spectrum @xcite .
one level of stout smearing has been applied in these simulations .
they have been performed at fixed lattice size of @xmath112 and @xmath113 .
a few runs , to check the finite volume effects , have been done using @xmath114 .
we have collected for each value of @xmath72 the critical value @xmath82 determined by the peak in the polyakov loop susceptibility .
the results can be found in table [ tablemeasstout ] and are represented by the red symbols in fig .
[ fig : roughps - a ] .
the value of @xmath115 is calculated with the value of @xmath102 , for each value of @xmath72 , obtained from an interpolation based on eq .
( [ interpolatingeq ] ) .
the red line indicates the transition between confined and deconfined phase for a residual gluino mass @xmath49 different from zero . in the limit of @xmath116 , i. e. infinite @xmath49 , the result of pure su(2 )
yang - mills theory is obtained .
lowering the mass , i.e. increasing @xmath71 , towards the chiral limit the phase transition temperature decreases .
since @xmath75 is fixed these lower temperatures correspond to a smaller value of @xmath72 and a larger lattice spacing . the blue symbols in fig .
[ fig : roughps - a ] indicate the lines of constant @xmath117 in zero temperature simulations .
they are based on the results of simulations at @xmath104 , @xmath118 , @xmath99 , @xmath100 , and @xmath101 , where theses five values correspond to the five points along each blue line .
a linear interpolation of @xmath80 as a function of @xmath84 has been used .
each of the intersections between the ( blue ) lines of constant @xmath49 and the phase transition ( red ) line corresponds to the phase transition at temperature @xmath119 of a theory with softly broken supersymmetry .
the phase transition of the supersymmetric yang - mills theory would correspond to the intersection between the red line and the @xmath103 line . from these results
it can only be estimated to be around @xmath120 .
a systematic extrapolation can be done as a function of the physical parameter @xmath49 instead of the bare parameter @xmath71 .
we take the four largest values of @xmath71 , corresponding to @xmath121 .
they are converted to @xmath122 using the values of @xmath123 of table [ tablemeasstout2 ] .
we can perform a linear fit to determine the critical temperature , see fig .
[ fig : roughps - b ] . already with these rough data the phase transition point can thus be estimated to be around @xmath124 in these first investigations it turned out that much larger statistics and a more precise scale estimation is necessary .
compared to these uncertainties the improvement by stout smearing of the links is not important .
we have therefore developed a new more flexible update program and performed a careful investigation in the region with small @xmath49 at the phase transition . in that way we have obtained a more reliable extrapolation of the supersymmetric limit without stout smearing .
as explained above , the parameter @xmath74 is used as a more recent alternative scale setting .
different spatial and temporal lattice extents are considered with lattice sizes @xmath125 = @xmath126 , @xmath127 , @xmath128 , and @xmath129 to estimate the influence of finite size effects and lattice artifacts . for each lattice size
, simulations are done with different bare gauge couplings @xmath72 and gluino masses ( @xmath71 ) .
the details of the simulations , done at the critical value @xmath82 , are summarised in table [ tablemeasunstout ] .
the autocorrelations between consecutive configurations generated by the hmc algorithm increase drastically near the critical point of the deconfinement transition .
we have increased the statistics near the phase transition in order to compensate this effect , and we have investigated accurately the finite volume effects and the scaling behaviour . hence these points require a huge amount of computer time . the bare gluino mass ( @xmath71 ) is varied for each @xmath75 and for each @xmath72 to locate the point of the deconfinement phase transition @xmath82 .
[ polyakov4_12 ] demonstrates this approach for a lattice size @xmath127 and @xmath130 , where the polyakov loop starts to rise at @xmath131 .
the point of the phase transition can be determined more clearly by the maximum of the polyakov loop susceptibility @xmath132 here the susceptibility is defined in terms of the modulus of the polyakov loop . while this choice does not alter the position of the peak , it introduces a non - zero value of @xmath133 below the critical point . as can be seen in fig .
[ susc4_12 ] , the location of the transition is found at @xmath134 , where the susceptibility shows a clear peak .
we have found that @xmath82 defined in this way has only a mild finite volume dependence , which is impossible to distinguish with our current precision . at @xmath135 and @xmath136 .
the red line is the expected scaling for a second order phase transition in the universality class of the @xmath47 ising model , the purple and the green lines represent instead a first order phase transition and a cross - over , respectively .
the coloured shadows indicate the errors from the extrapolation of eq .
( [ fssformula ] ) .
[ fss],scaledwidth=67.0% ] the finite size scaling of the susceptibility @xmath133 contains information about the nature of the phase transition in the infinite volume limit .
the susceptibility has a scaling dependence on the volume , @xmath137 which is linear for a first order phase transition ( @xmath138 ) , flat for a crossover behaviour ( @xmath139 ) , and non - linear , with @xmath140 , for a second order phase transition in the universality class of three - dimensional @xmath47 ising model @xcite .
we have obtained the ratios of the polyakov loop susceptibility for @xmath141 and @xmath142 with @xmath113 .
the resulting volume dependence is shown in fig .
the svetitsky - yaffe conjecture is in good agreement with the data and a possible change from the second order phase transition of pure gauge theory to first order induced by gluinos seems to be excluded . as a further evidence for this statement the distributions of the polyakov loop at different values of @xmath71
demonstrate the slow continuous emergence of a new peak in addition to the central distribution of the absolute value , see fig .
[ distribution ] .
this is in accordance with the divergence of the correlation length at a second order phase transition .
the peak of the polyakov loop susceptibility , eq .
( [ eq : sus ] ) , defines a critical combination of bare parameters @xmath143 , see table [ tablemeasunstout ] . at these values new simulations
have been performed at zero temperature , see table [ tablemeaszero ] , to determine the value of the adjoint pion mass in lattice units @xmath144 and to set the scale .
note that in table [ tablemeaszero ] , for each @xmath81 , in addition to the result obtained at the corresponding @xmath82 , other values determined at different @xmath71 are present : these are necessary to extrapolate the mass independent scale @xmath145 as explained in sec .
[ sec : scale ] . with the previous determinations ,
the critical temperature and the adjoint pion mass are obtained in dimensionless units , i. e. @xmath146 and @xmath147 . in the mass independent scheme the points @xmath148
are linearly interpolated , and the deconfinement temperature is extrapolated to @xmath149 the linear fit shown in fig .
[ extrapolation - mass - independent ] clearly indicates that the deconfinement transition occurs at lower temperatures when the gluino mass is decreased , @xmath150 the final extrapolation to the supersymmetric limit @xmath151 leads to @xmath152 where the quoted error is only statistical . as an alternative we employ a mass - dependent renormalisation scheme .
the points @xmath153 with @xmath74 determined at the same value of the bare parameters @xmath71 and @xmath72 at the phase transition
are linearly interpolated , @xmath154 where @xmath155 . the interpolation is shown in fig .
[ extrapolation - mass - dependent ] .
the slope is different due to the change of renormalisation scheme .
however , the final extrapolation to the supersymmetric limit leads to the compatible result @xmath156 but with a smaller error due to the more precise determination of the scale for points with heavier gluino masses . for a comparison of @xmath2 sym and pure gauge theory , we computed the scale @xmath87 at infinite gluino mass on a lattice @xmath157 and @xmath158 .
the chosen @xmath72 is the critical value for the deconfinement transition of pure su(2 ) yang - mills theory with a symanzik improved gauge action at @xmath159 @xcite .
the measured value of @xmath160 leads to @xmath161 for pure su(2 ) yang - mills theory .
the ratio of the deconfinement temperatures for pure and supersymmetric yang - mills theory is thus @xmath162 introducing physical units by setting @xmath163 mev for the critical temperature in pure gauge theory , we obtain a physical value of the deconfinement phase transition temperature for @xmath0 su(2 ) supersymmetric yang - mills theory , @xmath164 this value is consistent with the one level stout data , see eq .
( [ eq : roughstout1 ] ) , where we obtain the rough estimate @xmath165 from our data , assuming @xmath166 as in qcd . in the analysis of the deconfinement transition
we have found no evidence for contributions from negative pfaffians even at the largest values of @xmath71 .
the order parameter of the chiral phase transition is the gluino condensate . a non - zero expectation value of this parameter signals the breaking of the @xmath47 remnant of the u(1)@xmath52 symmetry .
the bare gluino condensate is defined as the derivative of the logarithm of the partition function with respect to the bare gluino mass parameter , @xmath167 chiral symmetry is broken by our lattice action with the wilson - dirac operator for the fermions , and the bare gluino condensate @xmath168 acquires an additive and multiplicative renormalisation : @xmath169 at zero temperature a first order transition is expected when the bare gluino mass is changed , crossing a critical value corresponding to @xmath103 .
close to such a transition the histogram of @xmath170 shows a two peak structure in a finite volume .
the transition can be identified with the point where the symmetry of the two peaks changes , as done in @xcite .
such an analysis is independent of the renormalisation described in eq .
( [ condensate - ren ] ) . at finite temperatures the first order chiral phase transition extends to a phase transition line at @xmath103 , terminating in a second order end - point . beyond that point
the transition changes from first order to a cross over , see fig .
[ scenarios ] . for this reason ,
considerations on the renormalisation procedure become important for the precise localisation of the phase transition .
the additive renormalisation is removed by a subtraction of the zero temperature result : @xmath171 the calculation of the renormalisation constant @xmath172 can be avoided in a fixed scale approach , where the bare coupling @xmath72 and @xmath71 are fixed and the temperature is changed by a variation of @xmath75
. the bare gluino condensate is obtained from the trace of the inverse wilson - dirac operator , @xmath173 here and in the following @xmath174 denotes the functional integral with respect to the gauge part of the action , i. e. @xmath175 , where a positive pfaffian is assumed .
the trace of the inverse wilson - dirac operator is evaluated with 20 random noise vectors using the stochastic estimator technique .
the simulations are done on a lattice @xmath176 , with @xmath177 , @xmath178 , and @xmath179 .
a simulation at zero temperature , i. e. @xmath180 , has been performed to determine the adjoint pion mass @xmath181 and the value of the scale @xmath182 .
the subtracted chiral condensate starts to rise at @xmath183 , but its behaviour is quite smooth due the crossover nature of the transition away from the supersymmetric limit , see fig .
[ chiralcond12c ] . for a better identification of the pseudo - critical transition point
, we determine the peak of the chiral susceptibility @xmath184 .
this observable is proportional to the derivative of the gluino condensate and it has connected and disconnected contributions : @xmath185 the connected contribution is expected to vanish in the supersymmetric limit . in the range of parameters that we have considered in this investigation the disconnected contribution is already dominant and the connected contribution can be neglected for a localisation of the peak , see fig .
[ connecteddisconnecedchiraldistribution ] . in that respect the relevant dynamics of the phase transition is already similar to the one at a vanishing residual gluino mass . even though the connected contribution is negligible
, we consider the complete observable in the following . in that way we ensure that absence of additional systematic uncertainties in our extrapolations .
the results of our simulation are shown in fig .
[ chiralcondsusc12c ] .
even though there is quite a broad central region , a visible peak can be identified corresponding to the value at @xmath186 . for comparison ,
the polyakov loop is shown in fig .
[ polyakov12c ] .
it acquires a non - vanishing expectation value at @xmath187 .
this small deviation is still consistent with the scenario of a chiral symmetry restoration and a deconfinement phase transition at the same temperature . in order to provide an upper limit for the chiral symmetry restoration temperature ,
we have done new simulations approximately at the supersymmetric limit , i. e. around the value of @xmath71 where the adjoint pion mass is expected to vanish , following the approach of @xcite .
this corresponds to @xmath188 at @xmath189 .
the lattice sizes were chosen to be @xmath190 with @xmath191 .
note that these simulations can not be done at large values of @xmath75 , due to the long time needed for a convergence of the conjugate gradient algorithm in that limit .
the distributions are displayed in fig .
[ chiraldistribution ] . at high temperatures , like @xmath192 ,
the distribution is close to a gaussian without any indications for a double peak in the gluino condensate . at @xmath193 ,
on the other hand , we observe a small second peak emerging from the distribution .
therefore , at these low temperatures the transition , close to the supersymmetric limit , becomes consistent with a first order chiral phase transition .
moreover , this suggests that the transition happens in the region between @xmath194 and @xmath193 .
it is another indication that the chiral phase transition and the deconfinement transition are close to each other : assuming that the value of @xmath74 does not change so much going from @xmath179 to @xmath195 we are able to estimate an upper limit for the supersymmetric chiral critical temperature : @xmath196 .
we have checked our assumption of a positive pfaffian by a measurement of its sign on 200 configurations . at @xmath179
we have found no contribution , whereas at @xmath195 around @xmath197 of the configurations have a negative sign on the @xmath198 lattice .
hence the simulations at the supersymmetric limit provide only an estimate of the transition point .
in further studies the contributions with a negative sign have to be taken into account more carefully ; alternatively , with a larger amount of computing time , the supersymmetric limit can extrapolated from a region without a relevant sign problem .
we have investigated the deconfinement and the chiral phase transitions in @xmath0 supersymmetric yang - mills theory .
different from the case of qcd , the polyakov loop is a well - defined order parameter at all values of the gluino mass , and the deconfinement transition can be identified in an unambiguous way .
the u(1)@xmath52 chiral symmetry is only partially broken by the anomaly and a remnant @xmath58 symmetry survives .
the expectation value of the gluino condensate is the order parameter for the breaking of this remnant symmetry down to @xmath47 .
we have investigated the dependence of both order parameters on the temperature and on the gluino mass .
we have determined the temperature , where the deconfinement phase transition takes place , with a good accuracy , and extrapolated the transition temperature in the supersymmetric limit .
different scale setting prescriptions lead to consistent results , indicating the reliability of the result .
the transition happens at a temperature , which is around 80% of the transition temperature in pure yang - mills theory .
the identification of the chiral phase transition point , on the other hand , needs more effort , since the transition becomes a crossover at finite gluino masses . in a fixed scale approach
we have identified the transition region taking also the renormalisation into account .
we were able to narrow the range for the chiral transition down to a region close to the deconfinement transition .
this situation can be compared with @xmath199 adjoint qcd ( aqcd ) , a theory similar to sym . in the case of aqcd
, there exists a mixed phase with deconfinement but a broken chiral symmetry .
the deconfinement temperature is eight times smaller than the point of chiral symmetry restoration @xcite . from this perspective
, sym appears to be more similar to qcd , where the deconfinement and chiral symmetry restoration seem to occur at the same temperature . in order to confirm scenario ( a ) of fig .
[ scenarios ] with coincident phase transitions , it will be necessary to perform simulations in more parameter points , with higher statistics , and on larger lattices .
a study of the finite size scaling is of great importance to test the existence of a second order endpoint for the chiral phase transition in the supersymmetric limit .
presently we have been able to study the phase transitions only at rather low values of @xmath72 , i. e. at relatively large lattice spacings . at these parameters
there is still a considerable deviation from the degeneracy of the particle masses in supermultiplets .
hence studies at larger values of @xmath75 are needed for reliable extrapolations to the supersymmetric limit .
this is the largest source of a systematic uncertainty for our current determination of the deconfinement transition point . in the future
we also plan to study different numbers of colours @xmath4 , and different boundary conditions , since the phase transitions seem to be sensitive to these parameters .
we thank i. montvay for helpful instructions and comments .
the authors gratefully acknowledge the computing time granted by the john von neumann institute for computing ( nic ) and provided on the supercomputer juropa at jlich supercomputing centre ( jsc ) , and by the leibniz - rechenzentrum ( lrz ) in mnchen provided on the supercomputer supermuc .
further computing time has been provided by the compute cluster palma of the university of mnster .
10 j. engels , j. fingberg and m. weber , nucl .
b * 332 * ( 1990 ) 737 .
b. lucini , m. teper and u. wenger , phys .
b * 545 * ( 2002 ) 197 , [ arxiv : http://arxiv.org / abs / hep - lat/0206029[hep - lat/0206029 ] ] .
y. aoki , g. endrodi , z. fodor , s. d. katz and k. k. szabo , nature * 443 * ( 2006 ) 675 , [ arxiv : http://arxiv.org / abs / hep - lat/0611014[hep - lat/0611014 ] ] .
y. aoki , z. fodor , s. d. katz and k. k. szabo , jhep * 0601 * ( 2006 ) 089 , [ arxiv : http://arxiv.org / abs / hep - lat/0510084[hep - lat/0510084 ] ] .
a. bazavov _ et al .
_ , phys .
d * 80 * ( 2009 ) 014504 , [ arxiv : http://arxiv.org / abs/0903.4379[0903.4379 ] [ hep - lat ] ] .
y. aoki , z. fodor , s. d. katz and k. k. szabo , phys .
b * 643 * ( 2006 ) 46 , [ arxiv : http://arxiv.org / abs / hep - lat/0609068[hep - lat/0609068 ] ] .
s. borsanyi _ et al . _
[ wuppertal - budapest collaboration ] , jhep * 1009 * ( 2010 ) 073 , [ arxiv : http://arxiv.org / abs/1005.3508[1005.3508 ] [ hep - lat ] ] .
j. m. maldacena , adv .
* 2 * ( 1998 ) 231 , [ arxiv : http://arxiv.org / abs / hep - th/9711200[hep - th/9711200 ] ] .
s. s. gubser and a. karch , ann .
nucl . part .
* 59 * ( 2009 ) 145 , [ arxiv : http://arxiv.org / abs/0901.0935[0901.0935 ] [ hep - th ] ] .
l. girardello , m. t. grisaru and p. salomonson , nucl .
b * 178 * ( 1981 ) 331 .
t. e. clark and s. t. love , nucl .
b * 217 * ( 1983 ) 349 .
a. k. das , physica a * 158 * ( 1989 ) 1 .
o. aharony , j. sonnenschein and s. yankielowicz , annals phys .
* 322 * ( 2007 ) 1420 , [ arxiv : http://arxiv.org / abs / hep - th/0604161[hep - th/0604161 ] ] .
a. armoni , m. shifman and g. veneziano , in : _
shifman , m. ( ed . )
et al . : from fields to strings , vol . 1 _ , 353 - 444 , [ arxiv : http://arxiv.org / abs / hep - th/0403071[hep - th/0403071 ] ] . a. armoni , m. shifman and g. veneziano , phys . rev
. lett .
* 91 * ( 2003 ) 191601 , [ arxiv : http://arxiv.org / abs / hep - th/0307097[hep - th/0307097 ] ] .
l. girardello , m. petrini , m. porrati and a. zaffaroni , nucl .
b * 569 * ( 2000 ) 451 , [ arxiv : http://arxiv.org / abs / hep - th/9909047[hep - th/9909047 ] ] .
i. montvay , nucl . phys .
b * 466 * ( 1996 ) 259 , [ arxiv : http://arxiv.org / abs / hep - lat/9510042[hep - lat/9510042 ] ] .
g. veneziano and s. yankielowicz , phys .
* b 113 * ( 1982 ) 231 .
g. r. farrar , g. gabadadze and m. schwetz , phys .
* d 58 * ( 1998 ) 015009 , [ arxiv : http://arxiv.org / abs / hep - th/9711166[hep - th/9711166 ] ] .
g. bergner , t. berheide , i. montvay , g. mnster , u. d. zugurel and d. sandbrink , jhep * 1209 * ( 2012 ) 108 , [ arxiv : http://arxiv.org / abs/1206.2341[1206.2341 ] [ hep - lat ] ] .
g. bergner , i. montvay , g. mnster , u. d. zugurel and d. sandbrink , jhep * 1311 * ( 2013 ) 061 , [ arxiv : http://arxiv.org / abs/1304.2168[1304.2168 ] [ hep - lat ] ] .
g. bergner , jhep * 1001 * ( 2010 ) 024 , [ arxiv : http://arxiv.org / abs/0909.4791[0909.4791 ] [ hep - lat ] ] .
m. kato , m. sakamoto and h. so , jhep * 0805 * ( 2008 ) 057 , [ arxiv : http://arxiv.org / abs/0803.3121[0803.3121 ] [ hep - lat ] ] .
g. curci and g. veneziano , nucl .
* b 292 * ( 1987 ) 555 .
h. suzuki , nucl .
* b 861 * ( 2012 ) 290 , [ arxiv : http://arxiv.org / abs/1202.2598[1202.2598 ] [ hep - lat ] ] . g. mnster and h. stwe , jhep ( 2014 ) , in press , [ arxiv : http://arxiv.org / abs/1402.6616[1402.6616 ] [ hep - th ] ] . d. amati , k. konishi , y. meurice , g. c. rossi and g. veneziano , phys .
162 * ( 1988 ) 169 .
b. svetitsky and l. g. yaffe , nucl .
b * 210 * ( 1982 ) 423 .
r. kirchner , i. montvay , j. westphalen , s. luckmann and k. spanderen , phys .
b * 446 * ( 1999 ) 209 , [ arxiv : http://arxiv.org / abs / hep - lat/9810062[hep - lat/9810062 ] ] .
n. seiberg , phys .
d * 49 * ( 1994 ) 6857 , [ arxiv : http://arxiv.org / abs / hep - th/9402044[hep - th/9402044 ] ] .
n. seiberg , nucl . phys .
b * 435 * ( 1995 ) 129 , [ arxiv : http://arxiv.org / abs / hep - th/9411149[hep - th/9411149 ] ] .
g. bergner and j. wuilloud , comput .
* 183 * ( 2012 ) 299 , [ arxiv : http://arxiv.org / abs/1104.1363[1104.1363 ] [ hep - lat ] ] .
r. sommer , nucl .
b * 411 * ( 1994 ) 839 , [ arxiv : http://arxiv.org / abs / hep - lat/9310022[hep - lat/9310022 ] ] .
s. borsanyi _ et al .
_ , jhep * 1209 * ( 2012 ) 010 , [ arxiv : http://arxiv.org / abs/1203.4469[1203.4469 ] [ hep - lat ] ] .
v. a. novikov , m. a. shifman , a. i. vainshtein and v. i. zakharov , nucl .
b * 229 * ( 1983 ) 381 .
a. k. de , a. harindranath and j. maiti , [ arxiv : http://arxiv.org / abs/0803.1281[0803.1281 ] [ hep - lat ] ] .
s. necco and r. sommer , nucl .
b * 622 * ( 2002 ) 328 , [ arxiv : http://arxiv.org / abs / hep - lat/0108008[hep - lat/0108008 ] ] .
a. m. ferrenberg and d. p. landau .
b * 44 * ( 1991 ) 5081 .
g. cella , g. curci , r. tripiccione and a. vicere , phys .
d * 49 * ( 1994 ) 511 , [ arxiv : http://arxiv.org / abs / hep - lat/9306011[hep - lat/9306011 ] ] .
f. karsch and m. ltgemeier , nucl .
b * 550 * ( 1999 ) 449 , [ arxiv : http://arxiv.org / abs / hep - lat/9812023[hep - lat/9812023 ] ] .
.parameters for the additional zero temperature simulations with one level of stout smearing . at @xmath104 the value of @xmath86
is extrapolated to the supersymmetric limit . [ cols="^,^,<,<,^,<,>",options="header " , ] | supersymmetry ( susy ) has been proposed to be a central concept for the physics beyond the standard model and for a description of the strong interactions in the context of the ads / cft correspondence
. a deeper understanding of these developments requires the knowledge of the properties of supersymmetric models at finite temperatures .
we present a monte carlo investigation of the finite temperature phase diagram of the @xmath0 supersymmetric yang - mills theory ( sym ) regularised on a space - time lattice .
the model is in many aspects similar to qcd : quark confinement and fermion condensation occur in the low temperature regime of both theories .
a comparison to qcd is therefore possible .
the simulations show that for @xmath0 sym the deconfinement temperature has a mild dependence on the fermion mass .
the analysis of the chiral condensate susceptibility supports the possibility that chiral symmetry is restored near the deconfinement phase transition . |
genomic dna is packaged into chromatin in eukaryotic cells .
the building block of chromatin is the nucleosome , @xcite a 147 base pair ( bp ) dna segment wrapped in @xmath0 superhelical coils on the surface of a histone octamer .
@xcite the unstructured tails of the histones are the targets of numerous covalent modifications @xcite and may influence how the ordered nucleosome array folds into higher order chromatin structures .
chromatin can both block access to dna @xcite and juxtapose sites far apart on the linear sequence .
@xcite the regulation of gene expression by dna - bound factors can be partially understood from their binding energies and thermodynamics .
nucleosomes formation energies vary by about 5 kcal / mol _ in vitro _ , @xcite comparable to the variation of dna - binding energies in other factors .
it is then natural to ask whether the locations of nucleosomes _ in vivo _ can be predicted thermodynamically , either in isolation or in competition with other dna - binding factors . under this scenario ,
the role of chromatin and its remodeling enzymes would be catalytic , modifying the rate of assembly but not the final disposition of factors on dna .
the alternative is that chromatin remodeling within a regulatory unit actively positions nucleosomes to control access to the dna , in analogy with motor proteins . it has not been possible to quantify by genetics where living cells fall between these extremes .
recent computational approaches @xcite have used collections of dna positioning sequences isolated _ in vivo _ to train pattern matching tools that were then applied genome - wide
. however , the training data may not be representative of direct histone - dna binding if other factors reposition nucleosomes _ in vivo _ by exclusion .
furthermore , models based on alignments of nucleosome positioning sequences @xcite require a choice of background or reference sequence and it is known that nucleotide composition varies among functional categories of dna .
* biophysical model of a nucleosome core particle . * for these reasons we developed a biophysical model for nucleosome formation that resolves the energy into the sum of two potentials modeling the core histone - dna attraction and the intra - dna bending energy . the first potential is assumed to be independent of the dna sequence since there are few direct contacts between the histone side chains and dna bases .
@xcite for the dna bending , we constructed an empirical quadratic potential @xcite using a database of 101 non - homologous , non - histone protein - dna crystal structures .
more specifically , we model dna base stacking energies by defining three displacements ( rise , shift , and slide ) and three rotation angles ( twist , roll and tilt ) for each dinucleotide ( fig . 1a ) . @xcite together the six degrees of freedom completely specify the spatial position of bp @xmath1 in the coordinate frame of bp @xmath2 ( supporting information ( si ) fig .
geometric transformations based on these degrees of freedom can be applied recursively to reconstruct an arbitrary dna conformation in cartesian coordinates .
conversely , the atomic coordinates in a crystal structure allow the inference of the displacements and rotations for each dinucleotide ( si methods ) .
the elastic force constants in the quadratic dna bending energy are inferred from the protein - dna structural data by inverting a covariance matrix of deviations of dinucleotide geometric parameters from their average values .
@xcite strongly bent dinucleotides ( with one or more geometric parameters further than 3 standard deviations from the mean ) are iteratively excluded from the data set ( si methods ) .
our model does not use any higher order moments of empirical geometry distributions , which would lead to a non - quadratic elastic potential ; nor are there sufficient data to model more than successive base pairs .
we assume that the histone - dna potential is at minimum along an ideal superhelix whose pitch and radius are taken from the nucleosome crystal structure , @xcite and varies quadratically when the dna deviates from the superhelix .
this sequence - independent term represents the balance between the average attractive electrostatic ( and other ) interactions between the histones and the dna phosphate backbone , @xcite and steric exclusion between the proteins and dna .
the sum of the bending and histone - dna potentials is minimized to yield the elastic energy and the dinucleotide geometries for each nucleosomal sequence ( see methods ) . because the total energy is quadratic ,
energy minimization is equivalent to solving a system of linear equations and thus the algorithm can be applied on a genome - wide scale .
there is no background in this model , but the results may depend on the choice of the protein - dna structural data set .
since our bending energy is empirical and inferred from co - crystal structures , it lacks a physical energy scale . by comparison with the worm - like chain model our units can be converted to kcal / mol through multiplication by 0.26 ( si methods ) .
we position multiple nucleosomes and proteins bound to dna using standard thermodynamics and enforce steric exclusion between bound entities in any given configuration ( si methods ) .
our program dnabend can compute the sequence - specific energy of dna restrained to follow an arbitrary spatial curve . * analysis of dna geometries from the nucleosome crystal structure .
* one way to validate dnabend is to predict the dna conformation in the high - resolution ( 1.9 ) nucleosome crystal structure @xcite ( pdb code 1kx5 ) using only the dna sequence .
the dnabend predictions are significantly correlated with the experimental geometries for twist , roll , tilt and slide ( @xmath3 ) , but are less successful for shift and rise ( si fig .
10 ) , although the peak positions are generally correct .
dnabend does not reproduce rapid shift oscillations in the region between bps 35 and 105 , and underestimates the magnitude of observed deviations in rise and slide .
we under - predict slide since for certain key base pairs our training structures imply a much smaller mean value of slide ( _ e.g. _ 0.18 for ca steps ) than that observed in the nucleosome structures ( 0.91 for ca steps ) . changing just this one mean value
gives more reasonable magnitudes of slide ( si fig .
10 , black curve ) .
et al . _ also observed that slide is large and positive in the nucleosome crystal structure , and makes a significant contribution to the superhelical trajectory .
@xcite because nucleosomal dna is highly bent , different degrees of freedom are strongly coupled ( si fig .
11a ) : for example , base pairs tend to tilt and shift simultaneously to avoid a steric clash .
these couplings are much less pronounced in the non - histone protein - dna complexes used to derive the elastic energy model ( si fig .
11b ) , but nonetheless appear prominently when the 1kx5 dna is positioned by dnabend ( si fig .
analysis of available nucleosome crystal structures shows that the tight dna wrapping is facilitated by sharp dna kinks if flexible dinucleotides ( _ e.g. _ 5-ca / tg-3 or 5-ta-3 ) are introduced into the region where the minor groove faces the histone surface . for other sequences and other structural regions the bending
is distributed over several dinucleotides .
@xcite we substituted all possible dinucleotides into the 1kx5 atomic structure ( keeping dna conformation fixed ) , and computed the elastic energy for each sequence variant .
the most sequence specific regions are those where the minor groove faces the histone octamer ( si fig .
the specificity is especially dramatic if dna is kinked ( _ e.g. _ at positions 109 , 121 and 131 ; si fig .
@xcite although these positions are occupied by ca / tg dinucleotides in the crystal structure , the model assigns the lowest energy to ta dinucleotides , consistent with the periodic ta signal previously observed in good nucleosome positioning sequences @xcite ( si fig .
. * comparison with _ in vitro _ measurements .
* dnabend accurately predicts experimental free energies of nucleosome formation @xcite ( fig .
2a , si fig .
in contrast , the alignment model of segal _ et al . _
@xcite trained on yeast nucleosomal sequences has little predictive power for the sequences in this set , most of which were chemically synthesized .
dnabend also correctly ranks sequences selected _ in vitro _ for their ability to form stable nucleosomes , or to be occupied by nucleosomes _ in vivo _
@xcite because the alignment model is constructed using the latter set it assigns anomalously low ( more favorable ) energies to some of the sequences from it , and higher ( less favorable ) energies to the artificial sequences known from experiment to have excellent binding affinities .
@xcite finally , dnabend correctly ranks mouse genome sequences selected _ in vitro _ on the basis of their high or low binding affinity @xcite ( @xmath4 ) , whereas the alignment model has less resolution ( @xmath5 ) , though on average it does assign better scores to the high affinity set ( si fig .
12b ) . a further test of how dnabend actually positions nucleosomes on dna can be provided by a collection of sequences where _ in vitro _
positions are known with 1 - 2 bp accuracy .
we have determined nucleosome positions on synthetic high - affinity sequences 601 , 603 , and 605 @xcite using hydroxyl radical footprinting ( si figs .
13 and 14 ) , and obtained 3 more sequences from the literature .
@xcite the measured position is always within 1 - 2 bp of a local minimum in our energy , and that energy minimum in 5 out of 6 cases is within 0.5 - 1.0 kcal / mol of the global energy minimum ( si fig .
15 ; note that the total range of sequence - dependent binding energies is @xmath6 kcal / mol ) .
we also asked if dnabend could be used to design dna sequences with high and low binding affinities for the histone octamer .
free energies of computationally designed sequences were measured _ in vitro _ using salt gradient dialysis @xcite ( si table 1 , si methods ) .
the free energy of the designed best sequence was lower than the free energy of the designed worst sequence , although only by 1.6 kcal / mol , which is less than the experimentally known range of free energies ( si table 1 , si fig .
these results underscore both the ranking power and the limitations of our current dna mechanics model . * periodic dinucleotide distributions in high and low energy sequences .
* dnabend - selected nucleosome sequences exhibit periodic dinucleotide patterns that are consistent with those determined experimentally : @xcite for example , with lowest energy sequences , 5-aa / tt-3 and 5-ta-3 dinucleotide frequencies are highest in the negative roll regions ( where the minor groove faces inward ) , while 5-gc-3 frequencies are shifted by @xmath6 bp ( si fig .
surprisingly , the distributions of at and aa / tt , ta dinucleotides are in phase , despite a very low flexibility of the former ( si fig . 8b ) .
it is possible that at steps are used to flank a more flexible kinked dinucleotide .
we estimate the energy difference between the best and the worst 147 bp nucleosome forming sequences to be 15.2 kcal / mol , with the energies of 95% of genomic sequences separated by
less than 6.4 kcal / mol ( si methods ) .
this is larger than the experimental range ( si fig .
12a ) because nucleosomes can not be forced in experiments to occupy the worst - possible location on a dna , but instead tend to find the most favorable locations with respect to the 10 bp helical twist .
* prediction of nucleosome positions in the yeast genome .
* we use nucleosome energies ( computed using the 147 bp superhelix ) and the binding energies of other regulatory factors to construct a thermodynamic model in which nucleosomes form while competing with other proteins for access to genomic sequence .
a typical configuration thus contains multiple dna - bound molecules of several types and explicitly respects steric exclusion .
we take all such configurations into account using dynamic programming methods @xcite that enable us to compute a boltzmann - weighted statistical sum and thus the probability @xmath7 for each factor to bind dna starting at every possible position along the genome ( fig .
1b , si methods ) .
@xcite we also compute the occupancy of each genomic bp , defined as its probability to be covered by any protein of a given type ( fig .
there are two adjustable parameters for each dna - binding factor : the temperature , which determines how many factors are bound with high probability , and the free protein concentration , which determines the average occupancy . with our choice of parameters ( which have not been optimized , rather , were fixed to allow for comparison with previous studies - si methods ) , the average nucleosome occupancy is 0.797 , and stable , non - overlapping nucleosomes ( with @xmath8 ) cover 16.3% of the yeast genome .
* bioinformatic models based on alignments of nucleosome positioning sequences .
* several existing analyses of nucleosome positions are based on alignment models , and these in turn explicitly @xcite or implicitly @xcite include a background model .
we examined the sensitivity of one alignment - based approach to the choice of background by implementing it in two different ways : alignment model i is identical to segal et al .
@xcite and uses genomic background for one strand and uniform background for the other , while alignment model ii employs genomic background for both strands ( si methods ) . comparing two models allows us to separate more robust predictions from those that depend strongly on the implementation details . although the alignment models correlate well with each other ( @xmath9 ) , we find a small _ negative _ correlation between dnabend - predicted occupancies and energies and those from the alignment model i ( si fig .
dnabend energies are strongly correlated if two nucleosomes are separated by a multiple of 10 bp , and anti - correlated if the nucleosomes are separated by a multiple of 5 bp , which puts the helical twist out of phase ( si fig .
log scores predicted by the alignment model exhibit weaker oscillations ( si fig .
17d ) , probably due to ambiguities in aligning the training sequences . because all three models make somewhat different predictions of average nucleosome occupancies in broad genomic regions ( si fig .
19 ) , additional experimental data are required to establish which model is the most accurate .
in contrast , all models predict that the distribution of center - to - center distances between proximal stable nucleosomes ( with @xmath10 ) exhibits strong periodic oscillations ( si fig .
20 ) . however , similar oscillations occur in a non - specific model where we first create a regular nucleosomal array and then randomly label a given fraction of nucleosomes as stable ( data not shown ) . because steric exclusion is respected by every model , nucleosomes form regular arrays ( si figs .
17a , b ) which help induce the observed periodicity in the positions of stable nucleosomes . *
nucleosome depletion upstream of the orfs . * microarray - based maps of _ in vivo _ nucleosome positions show striking depletion of nucleosomes from the promoter regions ( fig .
@xcite we find that this depletion is difficult to explain using nucleosome models alone .
however , the nucleosome - free region upstream of the open reading frames ( orfs ) becomes much more pronounced when tata box - binding proteins ( tbps ) and other factors bind their cognate sites ( fig .
3a , si fig .
displaced nucleosomes are re - arranged in regular arrays on both sides of factor - occluded sites , @xcite creating the characteristic oscillatory structure around the nucleosome - free region which includes the occupancy peak over the translation start observed in both regular @xcite and h2a.z @xcite nucleosomes ( fig .
3a and si fig .
similar oscillations occur when non - specific nucleosomes compete with tbp ( si fig .
21 ) , making intrinsic nucleosome sequence preferences hard to disentangle from the larger phasing effects in this data set .
* nucleosome occupancy of tata boxes . * nonetheless , nucleosome stabilities can play a crucial role in gene activation : for example , dnabend predicts that both tata boxes in the promoter of the yeast _ mel1 _ ( @xmath11-galactosidase ) gene are occupied by a stable nucleosome , in agreement with the extremely low level of background gene expression observed in _ mel1 _
promoter - based reporter plasmids @xcite ( fig .
in contrast , the tata elements of the _ cyc1 _ promoter were shown to be intrinsically accessible _ in vivo _ , @xcite resulting in high background expression levels .
consistent with these findings , we predict that one of the _ cyc1 _ tata boxes has intrinsically low nucleosome occupancy , and moreover that the nucleosome is easily displaced in competition with tbp ( fig .
it has been suggested that tata box - containing genes may be repressed through steric exclusion of tbp by a nucleosome placed over the tata box .
@xcite in agreement with this hypothesis , in the absence of other factors dnabend predicts slightly higher nucleosome occupancy over tata boxes ( fig .
stress - induced genes are tata - rich and may be nucleosome - repressed under non - stress conditions .
@xcite thus , this prediction of dnabend can be tested experimentally by measuring nucleosome occupancy over tata boxes of stress - induced genes in non - inducing conditions , genome - wide ( to ensure adequate statistics which were not provided by the data sets available to us , @xcite although it was found in a very recent genome - wide study of nucleosome occupancy in yeast that promoters of stress - induced genes tend to be covered by nucleosomes when cells are grown in ypd media , whereas the transcription start site for the typical gene was nucleosome - free @xcite ) . * nucleosome - induced tf cooperativity and nucleosome depletion over tf binding sites . *
it has been known from chromatin immunoprecipitation experiments that some near - consensus dna sequences are not occupied by their cognate transcription factors ( tfs ) , while poorer sites may be occupied .
@xcite thus it is natural to ask whether histone binding preferences can help distinguish functional and non - functional tf binding sites .
_ found that nucleosome occupancy was intrinsically smaller at functional sites @xcite ( si fig .
24a ) , while dnabend predicts the opposite ( fig . 4a , upper panel ) ( but
see contrary low resolution data from liu _
however , if tfs are allowed to compete with nucleosomes at all sites , the nucleosomes become preferentially depleted ( and the tf occupancy becomes higher ) at the functional sites ( fig .
4a , lower panel ) .
this depletion is not due to the slightly more favorable binding energies of the functional sites ( si fig .
25a ) , because it is reversed when their positions are randomized within intergenic regions ( si fig .
furthermore , dnabend - predicted nucleosomes are not depleted over functional sites simply because they are less stable - fig .
4a showed the opposite .
in fact , functional sites tend to occur just upstream of the orfs ( si fig .
25b ) , where dnabend - predicted nucleosomes exhibit enhanced stability ( si fig . 23 ) .
the only remaining possibility is that the depletion of dnabend - predicted nucleosomes is caused by the spatial clustering of functional binding sites ( fig .
4b , si fig .
25c ) . in this scenario several dna - binding proteins
cooperate to evict the nucleosome and thus enhance their own binding , in a phenomenon known as the nucleosome - induced cooperativity ( fig .
4c ) . @xcite the cooperativity does not depend on direct interactions between tfs and requires only that two or more tf sites occur within a 147 bp nucleosomal footprint ( fig .
4d ) . a model in which tfs compete with nucleosomes for their cognate sites provides the best explanation for the strong nucleosome depletion over functional sites observed in microarray experiments @xcite ( fig .
although it is not yet conventional to do so , @xcite it would be trivial to reward clustering within 147 bp in codes that predict regulatory sites . * predictions of experimentally mapped nucleosome positions .
* we used a set of 110 nucleosome positions reported in the literature for 13 genomic loci to see how well dnabend predictions match _ in vivo _ chromatin structure ( fig .
5 , si figs .
26a - n ) .
the predictive power of nucleosome models improves considerably when sequence - specific factors are included ( fig .
5 ) , though not in genome - wide sets ( si fig . 27 ) ; this result may be due to the lack of accurate genome - wide knowledge of tf binding sites and energies .
for example , at @xmath12 , @xmath13 , and recombination enhancer loci , nucleosomes are positioned in regular arrays whose boundaries are determined by the origin recognition complex , abf1 , rap1 , and mcm1/mat@xmath14 bound to their cognate sites ( si figs .
26f - h ) . both intrinsic sequence preferences and boundaries created by other factors contribute to positioning nucleosomes : on one hand , at the _ gal1 - 10 _ locus a nucleosome covering a cluster of gal4 sites is evicted , making @xmath15 bp of promoter sequence accessible to tbp and other factors ( si fig .
26a ) , as observed _ in vivo_. @xcite on the other hand , gcn4 and abf1 sites at the _
his3-pet56-ded1 _ locus are intrinsically nucleosome - depleted , because histones have lower affinity for _
his3-pet56 _ and _ ded1 _ promoters ( fig .
6 , si fig .
26k ) . @xcite
dnabend correctly predicts chromatin re - organization caused by sequence deletions in the _
promoter region : @xcite sequence - dependent nucleosome positions are refined through the action of gcn4 and tbp to improve the agreement with experiment ( si figs .
26l - n ) .
we have developed a dna mechanics model capable of accurately predicting _ in vitro _
free energies of nucleosome formation , optimal base pairs in the minor and major grooves of nucleosomal dna , and dna geometries specific to each base step .
we only get an agreement with the available genome scale data sets @xcite for nucleosome depletion from the tata box and functional tf binding sites when we include competition with the relevant factors . in the absence of dna - binding factors dnabend predicts a weak enhancement of nucleosome occupancy over their sites and thus agrees with the conjecture , not yet demonstrated on a genome - wide scale , that nucleosomes provide a default repression .
@xcite the two alignment models we examined do not support this conjecture ; it is conceivable that they have implicitly captured a hybrid signal from both nucleosomes and dna - bound factors .
the highest quality data on generic nucleosome occupancy come from the specific loci summarized in fig .
5 . there is a relatively weak signal from all models in the absence of other factors , originating from a few correct predictions at short distances .
because dnabend predicts correct binding energies , the weak signal suggests that nucleosomes positioned by thermodynamics on bare dna have relatively weak correlation with _ in vivo _
positions , while inclusion of other factors substantially improves the picture .
dnabend presents a useful biophysical framework for analysis of _ in vivo _ nucleosome locations and tf - nucleosome competition
. it will be interesting to examine its predictions for metazoan genomes , and to modulate gene expression levels in model systems through computational design of nucleosome occupancy profiles .
for full details , see si methods . dnabend software and additional supporting data ( including a _ s.cerevisiae
_ nucleosome viewer ) are available on the nucleosome explorer website : _
the total energy of a nucleosomal dna is given by a weighted sum of two quadratic potentials : @xmath16 where @xmath17 is the sequence - specific dna elastic energy @xcite and @xmath18 is the histone - dna interaction energy ( see si methods ) .
the potentials are quadratic with respect to deviations of global displacements @xmath19 and local angles @xmath20 from their ideal superhelical values .
the weight @xmath21 is fit to maximize the average correlation coefficient between the distributions of geometric parameters observed in 1kx5 @xcite and the dnabend predictions ( si fig .
the conformation adopted by the dna molecule is the one that minimizes its total energy @xmath22 , yielding a system of linear equations : @xmath23 where @xmath24 and @xmath25 is the number of dinucleotides .
the nucleosome energies ( assumed to be given by @xmath17 ) and the energies of other dna - binding factors at each genomic position are then used as input to a dynamic programming algorithm @xcite which outputs binding probabilities and bp occupancies for each dna element .
a.v.m . was supported by the lehman brothers foundation through a leukemia and lymphoma society fellowship .
e.d.s . was funded by the nsf grant dmr-0129848 . j.w .
acknowledges support from nigms grants r01 gm054692 and r01 gm058617 , and the use of instruments in the keck biophysics facility at northwestern university .
v.m.s . was supported by nsf ( 0549593 ) and nih ( r01 gm58650 ) grants .
we thank eran segal , michael y. tolstorukov , wilma k. olson , and victor b. zhurkin for useful discussions and for sharing nucleosome data .
99 khorasanizadeh s ( 2004 ) the nucleosome : from genomic organization to genomic regulation .
_ cell _ 116:259272 .
richmond tj , davey ca ( 2003 ) the structure of dna in the nucleosome core .
_ nature _ 423:145150 .
boeger h , griesenbeck j , strattan js , kornberg rd ( 2003 ) nucleosomes unfold completely at a transcriptionally active promoter .
_ mol cell _ 11:15871598 .
wallrath ll , lu q , granok h , elgin scr ( 1994 ) architectural variations of inducible eukaryotic promoters : preset and remodeling chromatin structures . _ bioessays _ 16:165170 . thastrom a , lowary pt , widlund hr , cao h , kubista m , widom j ( 1999 ) sequence motifs and free energies of selected natural and non - natural nucleosome positioning dna sequences . _
j mol biol _ 288:213229 .
segal e , fondufe - mittendorf y , chen l , thastrom a , field y , moore ik , wang jz , widom j ( 2006 ) a genomic code for nucleosome positioning .
_ nature _ 442:772778 .
ioshikhes ip , albert i , zanton sj , pugh bf ( 2006 ) nucleosome positions predicted through comparative genomics .
_ nat genet _ 38:12101215 .
peckham he , thurman re , fu y , stamatoyannopoulos ja , noble ws , struhl k , weng z ( 2007 ) nucleosome positioning signals in genomic dna .
_ genome res _ 17:11701177 .
luger k , mder aw , richmond rk , sargent df , richmond tj ( 1997 ) crystal structure of the nucleosome core particle at 2.8 resolution .
_ nature _ 389:251260 .
olson wk , gorin aa , lu x , hock lm , zhurkin vb ( 1998 ) dna sequence - dependent deformability deduced from protein - dna crystal complexes .
_ proc nat acad sci _ 95:1116311168 .
morozov av , havranek jj , baker d , siggia ed ( 2005 ) protein - dna binding specificity predictions with structural models .
_ nucl acids res _ 33:57815798 .
arents g , moudrianakis en ( 1993 ) topography of the histone octamer surface : repeating structural motifs utilized in the docking of nucleosomal dna .
_ proc nat acad sci _ 90:1048910493 .
tolstorukov my , colasanti av , mccandlish dm , olson wk , zhurkin vb ( 2007 ) a novel roll - and - slide mechanism of dna folding in chromatin : implications for nucleosome positioning .
_ j mol biol _ 371:725738 .
widom j ( 2001 ) role of dna sequence in nucleosome stability and dynamics .
_ q rev biophys _ 34:269324 .
shrader te , crothers dm ( 1989 ) artificial nucleosome positioning sequences .
_ proc nat acad sci _ 86:74187422 .
shrader te , crothers dm ( 1990 ) effects of dna sequence and histone - histone interactions on nucleosome placement .
_ j mol biol _ 216:6984 .
lowary pt , widom j ( 1998 ) new dna sequence rules for high affinity binding to histone octamer and sequence - directed nucleosome positioning .
_ j mol biol _ 276:1942 .
widlund hr , cao h , simonsson s , magnusson e , simonsson t , nielsen pe , kahn jd , crothers dm , kubista m ( 1998 ) identification and characterization of genomic nucleosome - positioning sequences . _
j mol biol _ 267:807817 .
cao h , widlund hr , simonsson t , kubista m ( 1998 ) tgga repeats impair nucleosome formation .
_ j mol biol _ 281:253260 .
thastrom a , bingham lm , widom j ( 2004 ) nucleosomal locations of dominant dna sequence motifs for histone - dna interactions and nucleosome positioning . _
j mol biol _ 338:695709 .
durbin r , eddy sr , krogh a , mitchison g ( 1998 ) _ biological sequence analysis : probabilistic models of proteins and nucleic acids _
( cambridge university press , cambridge , ma ) .
yuan g , liu y , dion mf , slack md , wu lf , altschuler sj , rando oj ( 2005 ) genome - scale identification of nucleosome positions in
_ s.cerevisiae_. _ science _ 309:626630 .
ozsolak f , song js , liu xs , fisher de ( 2007 ) high - throughput mapping of the chromatin structure of human promoters .
_ nat biotech _ 25:244248 .
kornberg rd , stryer l ( 1988 ) statistical distributions of nucleosomes : nonrandom locations by a stochastic mechanism .
_ nucl acids res _ 16:66776690 .
fedor mj , lue nf , kornberg rd ( 1988 ) statistical positioning of nucleosomes by specific protein - binding to an upstream activating sequence in yeast .
_ j mol biol _ 204:109127 .
albert i , mavrich tn , tomsho lp , qi j , zanton sj , schuster sc , pugh bf ( 2007 ) translational and rotational settings of h2a.z nucleosomes across the _ saccharomyces cerevisiae _ genome .
_ nature _ 446:572576 .
ligr m , siddharthan r , cross fr , siggia ed ( 2006 ) gene expression from random libraries of yeast promoters .
_ genetics _ 172:21132122 .
melcher k , sharma b , ding wv , nolden m ( 2000 ) zero background yeast reporter plasmids .
_ gene _ 247:5361 .
chen j , ding m , pederson ds ( 1994 ) binding of tfiid to the cyc1 tata boxes in yeast occurs independently of upstream activating sequences .
_ proc nat acad sci _ 91:1190911913 .
kuras l , struhl k ( 1999 ) binding of tbp to promoters _ in vivo _ is stimulated by activators and requires pol ii holoenzyme .
_ nature _ 399:609613 .
struhl k ( 1999 ) fundamentally different logic of gene regulation in eukaryotes and prokaryotes .
_ cell _ 98:14 .
lee w , tillo d , bray n , morse rh , davis rw , hughes tr , nislow c ( 2007 ) a high - resolution atlas of nucleosome occupancy in yeast .
_ nat genet _
39:12351244 .
macisaac kd , wang t , gordon db , gifford dk , stormo gd , fraenkel e ( 2006 ) an improved map of conserved regulatory sites for _ saccharomyces cerevisiae_. _ bmc bioinformatics _ 7:113 .
adams cc , workman jl ( 1995 ) binding of disparate transcriptional activators to nucleosomal dna is inherently cooperative .
_ mol cell biol _ 15:14051421 .
liu x , lee c - k , granek ja , clarke nd , lieb jd ( 2006 ) whole - genome comparison of leu3 binding in vitro and in vivo reveals the importance of nucleosome occupancy in target site selection . _
genome res _
16:15171528 .
miller ja , widom j ( 2003 ) collaborative competition mechanism for gene activation _ in vivo_.
_ mol cell biol _ 23:16231632 .
li s , smerdon mj ( 2002 ) nucleosome structure and repair of n - methylpurines in the gal1 - 10 genes of _
saccharomyces cerevisiae_. _ j biol chem _ 277:4465144659 .
sekinger ea , moqtaderi z , struhl k ( 2005 ) intrinsic histone - dna interactions and low nucleosome density are important for preferential accessibility of promoter regions in yeast .
_ mol cell _ 18:735748 .
basehoar ad , zanton sj , pugh bf ( 2004 ) identification and distinct regulation of yeast tata box - containing genes .
_ cell _ 116:699709 .
morozov av , siggia ed ( 2007 ) connecting protein structure with predictions of regulatory sites .
_ proc nat acad sci _ 104:70687073 . only for the _
pho5 _ , @xmath13 , and _
deletion loci with dnabend ( si fig .
26 ) , and for the _ gal1 - 10 _ , _ ste6 _ loci with alignment model i , @xcite for distances @xmath26 bp .
see si fig .
27 for details of implementation , including the definition of the null ( random ) model with steric exclusion .
results from the alignment model ii are similar ( data not shown ) . ]
we model each dna basepair as a rigid body and specify its position in space using a local coordinate frame attached to each basepair and defined using @xmath27 $ ] and @xmath28 ( @xmath29 ) . here
@xmath30 are the orthonormal basis vectors of the local frame , and @xmath28 is its origin in the fixed global coordinate frame .
both the basis vectors and the vector to the origin can be expressed in terms of 6 independent parameters that are sufficient for reconstructing the spatial position of any rigid body .
@xcite thus apart from a single global translation and rotation an arbitrary dna conformation with @xmath31 basepairs is uniquely specified if @xmath25 sets of 6 geometric parameters are known . by convention , @xcite the geometric parameters
are chosen to be the three angular degrees of freedom which define unit vectors of the local frame attached to the basepair @xmath2 in the local frame attached to the basepair @xmath32 , and the displacement vector which gives the origin of frame @xmath2 with respect to origin of frame @xmath32 : @xmath33 . here
@xmath34 are the helical twist , roll and tilt angles , and @xmath35 is the displacement vector with the x , y , z components called slide , shift and rise ( fig
@xcite since the geometric parameters which specify the spatial position of the basepair @xmath2 are defined with respect to the frame rigidly attached to the previous basepair , their values capture local deviations in the dna conformation .
however , they can also be used to recursively construct the rotation matrix for the basepair @xmath2 in the global frame : @xmath36 where each @xmath37 matrix is the product of three rotations : @xmath38 here , @xmath39 @xmath40 @xmath41 note that @xmath42 and @xmath43 are the rotation matrices around the y and z axes , respectively . the middle term on the right - hand side of equation ( [ ti ] ) introduces both roll and tilt with a single rotation through angle @xmath44 around a roll - tilt axis .
the roll - tilt axis lies in the x - y plane of the mid - step coordinate frame ( mid - step triad , or mst : supplementary fig .
7 ) , and is in general inclined at an angle @xmath45 to the y - axis . @xcite in this scheme , it is the mid - step triad that is used to transform the displacement vector @xmath35 into the fixed global frame : @xmath46 . introducing a single roll - tilt rotation axis
is preferable to separate roll and tilt rotations because it eliminates the non - commutativity problem associated with the roll and tilt rotation operations .
@xcite similarly to the basepair rotation matrices , the mid - step triads are also constructed recursively : @xmath47 where @xmath48 thus having a complete set of local geometric parameters @xmath49 is equivalent to knowing the global dna conformation : first , the recursive relation ( [ ri ] ) is employed to determine the orientations of all basepair coordinate frames ( except for the first one which has to be fixed in space independently and thus provides the overall orientation and position of the dna molecule ) .
second , equation ( [ ri : mst ] ) is used to construct the mst frames , which are then employed to transform all local displacements @xmath35 into the global frame . finally , all displacement vectors are added up vectorially to determine the origins of the basepair coordinate frames . if necessary , the basepair frames can then be used to reconstruct the positions of all dna basepair atoms ( using an idealized representation which neglects flexibility of bases in a basepair ) . note that knowing a basepair coordinate frame is in general insufficient for predicting the positions of the phosphate backbone and sugar ring atoms because of their additional degrees of freedom
. finally we remark without showing the details of the calculations that the inverse problem is also well - defined : a full set of basepair and mst rotation matrices in the global frame is sufficient to reconstruct all local degrees of freedom @xmath50 .
we have implemented the solution to the inverse problem following a previously published description .
@xcite while the helical twist serves to rotate the consecutive dna basepairs with respect to one another , introducing non - zero roll and tilt angles imposes curvature onto the dna conformation . indeed ,
if both roll and tilt are negligible @xmath37 is simply a rotation through @xmath51 around the z - axis ( which coincides with the helical axis in this case ) , and @xmath52 is a rotation through @xmath53 around the same axis . the action of roll and tilt can be seen more clearly by considering the curvature vector , defined as the difference between the tangent vectors for two consecutive basepairs : @xmath54 .
the expression for curvature is simplified in the limit of small roll and tilt : since the magnitude of roll ( @xmath55 ) and tilt ( @xmath56 ) are typically much smaller than the helical twist per basepair ( @xmath57 ) , @xmath58 can be expanded to @xmath59 : @xmath60 because we are primarily interested in the magnitude of the curvature vector , we can compute it in the local frame of basepair @xmath2 : @xmath61 where @xmath62 is the z - axis unit vector . using equation ( [ ti_linear ] )
we obtain : @xmath63 thus in this limit the magnitude of the curvature vector contains equal contributions from both roll and tilt : @xmath64 .
furthermore , the roll and tilt contributions to curvature are shifted by @xmath65 with respect to one another .
transformations between local and global descriptions of the dna molecule are important because the sequence - specific dna elastic energy is best described in terms of the local degrees of freedom @xmath50 , whereas it is the global coordinates that can be most naturally restrained to follow an arbitrary spatial curve . in nucleosome core particles the spatial curve used to restrain
dna conformation is an ideal left - handed superhelix with pitch @xmath7 and radius @xmath66 , described by the following parametric equation : @xmath67 using the arc length @xmath68 to parameterize the curve we obtain : @xmath69 where @xmath70 .
a local frame at position @xmath71 is given by a set of three orthonormal frenet vectors ( tangent , normal , and binormal ) : @xmath72 we position @xmath31 sets of frenet basis vectors equidistantly along the superhelical curve : @xmath73 ( @xmath29 ) , where @xmath11 is the number of nucleosomal superhelical turns .
fitting an ideal superhelix to the high - resolution crystal structure of the nucleosome core particle @xcite reveals that @xmath74 in order to transform the slowly rotating frenet basis vectors into the helically twisted local frames associated with dna basepairs , we impose an additional helical rotation around the tangent vector @xmath75 : @xmath76 here @xmath77 is the cumulative helical twist , @xmath78 is the average helical twist from the structure , and the rotation specified by equations ( [ nb : rot ] ) is performed in the @xmath79 basis .
similarly , the mst frames are located at @xmath80 ( @xmath81 ) , and are rotated through @xmath82 to bring them into register with the helical twist .
a superhelix described by equation ( [ ideal : t ] ) has constant curvature which is manifested by the constant difference between the consecutive tangent vectors : @xmath83 .
this allows us to determine the length of the roll - tilt vector : @xmath84 ( which also gives the maximum value of roll and tilt ) .
in this idealized picture , roll and tilt can be assumed to make equal contributions to the curvature . indeed , reconstruction of the local geometric parameters @xmath85 from the full set of helically twisted frenet frames results in @xmath86 and @xmath87 , where @xmath88 and @xmath89 is the initial phase determined by the location of the first basepair .
thus twist and rise are constant for every basepair in the ideal superhelix , slide and shift are zero , whereas roll and tilt exhibit oscillations resulting from the superhelical curvature , and shifted by @xmath65 with respect to one another ( supplementary fig .
10 ) .
the total dna conformation energy is a weighted sum of two terms : the sequence - specific dna elastic energy designed to penalize deviations of the local geometric parameters from their average values observed in the protein - dna structural database , and the restraint energy which penalizes deviations of the dna molecule from the nucleosomal superhelix .
thus all interactions ( primarily of electrostatic nature ) between the dna molecule and the histone octamer are modeled implicitly by imposing the superhelical restraint .
* dna elastic energy . * using the local degrees of freedom @xmath49 , we represent the sequence - specific dna elastic energy by a quadratic potential : @xcite @xmath90^t f^{n(s ) } [ \alpha^{s } - \langle \alpha^{n(s ) } \rangle],\ ] ] where the sum runs over all consecutive dinucleotides ( basesteps ) @xmath71 and @xmath91 are the average values computed for the basestep type @xmath92 using a collection of oligonucleotides extracted from a set of 101 non - homologous protein - dna structures .
@xcite the matrix of force constants @xmath93 is evaluated by inverting the covariance matrix @xmath94 of deviations of local geometric parameters from their average values ( @xmath95 ) : @xcite @xmath96 note that our elastic energy model utilizes only and first and second moments of the empirical geometry distributions , and thus disregards all non - gaussian effects such as skewness and bimodality . including higher order moments amounts to introducing non - quadratic corrections to the elastic energy model , which makes the computation much less tractable .
this problem is partially alleviated however by removing dinucleotides from the data set if one or more of their geometric parameters are further than 3 standard deviations away from the mean .
the mean is then recomputed and the procedure is repeated until convergence .
@xcite dna elastic energy is a function of 10 basestep type - dependent average values for each of the 6 local degrees of freedom ( all averages and covariances for the dinucleotides related by reverse complementarity are equal by construction ) , and of 15 independent force constants in each symmetric 6d matrix @xmath97 . in order to be able to combine the dna elastic term with the superhelical restraint term defined in the global frame , we apply the following coordinate transformation to equation ( [ e_wo ] ) : @xmath98 ( @xmath99 and @xmath100 denote 3d unit and zero matrices , respectively ) .
applying @xmath101 to the local degrees of freedom leaves the angles ( twist , roll , and tilt ) invariant , but transforms the local displacements ( shift , slide , and rise ) into the global frame . to keep the elastic energy invariant , the force constants have to be transformed as well : @xmath102 .
equation ( [ e_wo ] ) then becomes : @xmath103^t \mathbf{f}^{n(s ) } [ { \mathbf r}_s ( \alpha^{s } - \langle \alpha^{n(s ) } \rangle ) ] , \ ] ] finally , it is convenient to change variables so that all degrees of freedom are expressed in terms of their deviations from the ideal superhelix : @xmath104 note that all the transformations described in this section involve no additional approximations to the original dna elastic energy ( equation ( [ e_wo ] ) ) .
thus we are free to choose the most convenient rotation matrix in equation ( [ rs ] ) , and use @xmath105 from the ideal superhelix in equation ( [ e_wo_global ] ) .
* superhelical restraint energy . *
superhelical restraint energy is used to bend nucleosomal dna into the superhelical shape .
we define the restraint energy as the quadratic potential : @xmath106 where @xmath107 and @xmath108 are the nucleosomal and the ideal superhelix radius - vectors in the global frame ( @xmath109 ) : @xmath110 then to the lowest order the difference between the radius vectors is given by : @xmath111,\ ] ] where @xmath112 label the vector components , and @xmath113 are the connectivity coefficients constructed as a product of the first derivative of the rotation matrix with respect to the frame angles @xmath114 and the ideal displacements @xmath115 in the local frame . the first term in equation ( [ d : exp ] )
represents the net change in the global radius vector @xmath116 caused by the changes in the displacements @xmath117 up to and including the basestep @xmath71 , while the second term reflects the change in the global radius vector resulting from modifying one of the rotation angles @xmath118 .
note that changing a single rotation angle at position @xmath119 affects every downstream basepair position linearly by introducing a kink into the dna chain , whereas making a displacement change at that position simply shifts all downstream coordinates by a constant amount .
the first derivative of the rotation matrix is evaluated as : @xmath120 upon substitution of the expansion ( [ d : exp ] ) into the restraint energy we obtain an effective quadratic potential : @xmath121 where @xmath122 , and the 6x6 matrix of force constants is given by three distinct 3x3 submatrices : @xmath123 here , @xmath124 where @xmath125 is the theta function , and @xmath126 is the kronecker delta .
@xmath127 couples displacements with displacements , @xmath128 couples displacements with angles ( and thus has one connectivity coefficient ) , and @xmath129 couples angles with angles through two connectivity coefficients . summing over some of the theta functions we obtain : @xmath130 where @xmath131 .
* total energy and dna conformation minimization . * the total energy of nucleosomal dna is given by : @xmath132 where @xmath21 is the fitting weight introduced to capture the balance between favorable histone - dna interactions and the unfavorable energy of bending dna into the nucleosomal superhelix .
we fit @xmath21 to maximize the average correlation coefficient between the distributions of geometric parameters observed in the high - resolution crystal structure of the nucleosome core particle ( 1kx5 ) , @xcite and the corresponding dnabend predictions ( supplementary fig .
this procedure yields @xmath133 for the 147 bp superhelix and @xmath134 for the 71 bp superhelix bound by the @xmath135 tetramer ( bps 39 through 109 in the 147 bp superhelix ) .
dnabend is not very sensitive to the exact value of @xmath21 : we found a correlation of 0.99 between the free energies computed using @xmath133 and @xmath134 and the 71 bp superhelix ( data not shown ) .
note that the total energy is a sum of two quadratic potentials written out in terms of the deviations of the global displacements @xmath19 and the local frame angles @xmath20 from their ideal superhelical values .
given the value of the fitting weight , the conformation adopted by the dna molecule is the one that minimizes its total energy @xmath22 .
since the energy is quadratic , finding the energy minimum is equivalent to solving a system of linear equations : @xmath136 the numerical solution of the system of equations ( [ ederiv ] ) provides a set of geometric parameters corresponding to the minimum dna energy : @xmath137 .
these can be used to find the original geometric parameters in the local frame : @xmath138 .
* worm - like chain . * according to the worm - like chain model , the energy required to bend a dna molecule is sequence - independent and given by : @xcite @xmath139 where @xmath140 is the contour length of the molecule , @xmath141 is the persistence length ( estimated to be @xmath142 ) , @xcite @xmath143 kcal / mol , and @xmath75 is the unit tangent vector .
the contour length of the ideal 147 bp superhelix is given by @xmath144 . from equations ( [ ideal :
s ] ) and ( [ tnb ] ) we obtain @xmath145 , and thus @xmath146 in fig
. 2b , the mean and the standard deviation @xmath147 of dnabend energies computed for chromosome iii are @xmath148 . equating the worm - like chain model estimate with the mean dnabend energy ,
we obtain a scaling coefficient of 0.26 which can be used to express dnabend energies in kcal / mol .
this gives the difference of 15.2 kcal / mol between the best and the worst chromosome iii sequences , and @xmath149 kcal / mol .
most sequences differ by @xmath150 or less in fig .
2b and are thus separated by @xmath151 kcal / mol .
a similar value of the scaling coefficient ( 0.21 ) arises from a linear model fit between experimental and dnabend - predicted free energies in supplementary fig .
12a ( red circles ) .
chromatin structure plays an important role in regulating eukaryotic gene expression .
@xcite eukaryotic dna is packaged by histones into chains of nucleosomes that are subsequently folded into @xmath152 nm chromatin fibers . because arrays of nucleosomes form on chromosomal dna under physiological conditions , in order to predict genomic nucleosomal occupancies we need to describe dna packaged into multiple non - overlapping nucleosomes .
furthermore , nucleosomes may compete with other dna binding proteins for genomic sequence .
for example , several closely spaced tf binding sites may serve to displace the nucleosomes from the promoter region upon tf binding . to take the competition between nucleosomes and other factors into account
we need to consider configurations with multiple regularly spaced nucleosomes ( accomodating a 20 - 25 bp linker between the end of one nucleosome and the beginning of the next ) as well as other dna - binding proteins .
the probability of any such configuration can be computed if we can construct a statistical sum ( partition function ) over all possible configurations .
though the sum has exponentially many terms and is thus impossible to evaluate by brute force for all but very short dna sequences , it can nonetheless be efficiently computed with a dynamic programming algorithm .
@xcite here we develop the dynamic programming approach for a general case of @xmath153 objects of length @xmath154 that are placed on genomic dna ( _ i.e. _ occupy @xmath155 dna basepairs ) .
the objects could represent nucleosomes , tfs , or any other dna - binding proteins .
the binding energy of object @xmath119 with dna at each allowed position @xmath2 is assumed to be known : @xmath156 , where @xmath25 is the number of basepairs . for nucleosomes we will carry out a genome - wide computation of dna elastic energies as described above , while for other factors
the energy landscapes will be constructed based on their binding preferences inferred from footprinting experiments , selex assays , bioinformatics predictions , etc .
we assign index @xmath157 to the background which can be formally considered to be an object of length 1 : @xmath158 . in the simplest case which we consider here the background energy is zero everywhere , but more sophisticated models could incorporate a global bias by making positions near dna ends less favorable , etc .
we wish to compute a statistical sum over all possible configurations in which object overlap is not allowed ( including the background `` object '' ) : @xmath159 where @xmath160 is the total energy of an arbitrary configuration of non - overlapping objects , @xmath161 is the number of objects of type @xmath119 , and @xmath162 is the pre - computed energy of the object of type @xmath119 which occupies positions @xmath163 through @xmath164 .
it is possible to evaluate @xmath165 ( or the free energy @xmath166 ) efficiently by recursively computing the partial statistical sums : @xmath167 with the initial condition @xmath168 .
the theta function is defined as : @xmath169 it is computationally more efficient to transform equation ( [ z : forward ] ) into log space by defining the partial free energies @xmath170 : @xmath171 with the initial condition @xmath172 . for numerical stability ,
we rewrite equation ( [ f : forward ] ) as an explicit update of the partial free energy computed at the previous step : @xmath173 the free energy computed in the final step is the full free energy where all possible configurations are taken into account : @xmath174 . since the algorithm proceeds by computing partial sums from 1 to n it is often called the forward pass .
similar equations can be constructed for the backward pass which proceeds from n to 1 : @xmath175 with the initial condition @xmath176 .
we transform equation ( [ z : backward ] ) into log space by defining the backward partial free energies @xmath177 : @xmath178 with the initial condition @xmath179 , or equivalently : @xmath180 note that @xmath181 by construction . with the full set of forward and backward partial free energies we can evaluate any statistical quantity of interest . for example
, the probability of finding an object of type @xmath119 at positions @xmath182 is given by : @xmath183 another quantity of interest is the occupancy of the basepair @xmath2 by object @xmath119 , defined as the probability that basepair @xmath2 is covered by any object of type @xmath119 : @xcite @xmath184 ( note that @xmath185 for @xmath186 ) .
if the object is composite ( _ i.e. _ consists of @xmath155 basepairs extended symmetrically on both sides by @xmath187 to take the nonzero linker lengths into account : @xmath188 ) , we may be interested in the partial occupancy by @xmath155 basepairs at the center of the object . equation ( [ occ : def ] ) then becomes : @xmath189 recursively , @xmath190 here we assume that all quantities on the right - hand side are set to zero if their indices are outside their definition domains ( such as @xmath191 ) .
finally , we need to take into account the fact that the objects can bind dna in both directions , and thus there are two binding energies for each position : @xmath192 ( object @xmath119 starts at @xmath2 and extends in the 5 to 3 direction ) and @xmath193 ( object @xmath119 starts at @xmath194 and extends in the 3 to 5 direction ) .
it is easy to show that the formalism developed above applies without change if the binding energies @xmath192 are replaced by the free energies @xmath195 which take both binding orientations into account : @xmath196.\ ] ] in the case of a single type of dna - binding object ( such as the nucleosome energies predicted with dnabend ) there are two free parameters : the mean energy of nucleosomes @xmath197 over a chromosome or a given genomic region which affects overall nucleosome occupancy , and the standard deviation @xmath198 which plays the role of inverse temperature . note that dnabend nucleosomes are 157 bp long because a 147 bp histone binding site is flanked on both sides by 5 bp long linkers which do not contribute to the energy . because the background energy is set to zero , making @xmath197 more negative results in fewer bases being devoid of nucleosomes ( this corresponds to increasing free concentration of histone octamers in the histone - dna solution ) , and vice versa .
the inverse temperature is related to the nucleosome stability : making @xmath198 smaller results in fewer stable nucleosomes ( _ e.g. _ those present in half or more of all the configurations in equation ( [ z : part ] ) ) , while allowing for a bigger difference between favorable and unfavorable nucleosomal energies results in more configurations with frozen , stable nucleosomes .
for all calculations in this paper we have chosen to set @xmath199 , @xmath200 for every chromosome .
the nucleosome energies with zero mean result in the average nucleosome occupancy of 0.797 over all chromosomes , and 12538 stable , non - overlapping nucleosomes ( with @xmath8 ) covering 16.3% of the yeast genome .
in addition , this value of @xmath198 is close to the scale of dnabend elastic energies .
the free parameters described here are also present in the alignment model ( albeit in a more implicit manner , see section [ aln : methods ] ) , and are required in general for each type of dna - binding object .
all tf binding energies were computed using position - specific weight matrices ( pwms ) from morozov _ et al
. _ @xcite and macisaac _ et al . _
: @xcite @xmath201 where @xmath202 is the length of the binding site , @xmath203 is the nucleotide at position @xmath2 , @xmath204 is the frequency of base @xmath49 at position @xmath2 in the pwm , and @xmath205 is the background frequency of base @xmath49 : @xmath206 ( the same background frequencies as used in the alignment - based nucleosome model , see section [ aln : methods ] ) .
@xmath207 is an energy scale , set to 1 for all tfs .
the distribution of energies is bounded from below by the energy of the consensus sequence @xmath208 which is a free parameter of the model since it depends on the tf concentration in solution .
thus the energy of a tf bound to site @xmath209 is given by : @xmath210 , where @xmath211 is typically set to -6.0 ( relative to the 0.0 average energy of dnabend nucleosomes ) , but can be also varied through a range of energies ( cf .
finally , the two energies of binding the same site in both directions are combined using : @xmath212,\ ] ] where @xmath213 is the strand bias .
a set of tf binding energies computed in this way ( together with the nucleosome binding energy profile ) is used as input to the recursive algorithm from section [ dynapro ] . in practice , we restrict a set of sites to which tfs are allowed to bind by assigning a highly unfavorable energy to all tf positions that are not listed as predicted sites by macisaac _
@xcite tbp binding energies were obtained in the same way , except that the tbp pwm was derived using an alignment of tata box sites from basehoar _ et al . _
@xcite tbps were only allowed to bind the sites listed in basehoar _
_ as tata boxes .
yuan _ et al .
_ have developed a hidden markov model - based approach for inferring nucleosome positions from high resolution tiled microarray data .
@xcite the data was collected for most of _
chromosome iii , plus additional regulatory regions .
we have used yuan _
`` hmm calls '' listed in the supplementary data to identify probes occupied by either a well - positioned or a fuzzy nucleosome ( manually added `` hand calls '' were not considered ) .
we then extended nucleosomal regions from the mid - point of the first occupied probe to the mid - point of the last occupied probe , plus additional 10 bp on each side ( making it consistent with the definition of nucleosome coverage of tf binding sites employed in fig . 3 of yuan _ et al . _ ) .
@xcite thus the nucleosomal length is given by @xmath214 , where @xmath25 is the number of contiguous probes occupied by a single nucleosome .
we find that most nucleosomes occupy 6 contiguous probes and thus cover 120 bp , somewhat less than 147 bp expected from the structural point of view .
the total number of nucleosomes placed on chromosome iii is 1045 ; every basepair covered by either well - positioned or fuzzy nucleosome was assigned an occupancy of 1.0 , resulting in the overall average occupancy of 0.488 ( that is , almost half of the genome is covered by the non - overlapping nucleosomes ) .
the resulting occupancy profile and the corresponding nucleosome positions were used to compile all yuan _ et al .
_ results in this work .
albert _ et al .
_ have used high throughput dna sequencing to identify genomic positions of h2a.z nucleosomes in _ s.cerevisiae _ with a median 4 bp resolution .
@xcite we converted their nucleosome positioning data into the occupancy profile by using the `` fine - grain '' nucleosome center coordinates ( reported for both strands in albert _ et al . _
supplementary materials ) to place 147 bp nucleosomes on the genome . since many `` fine - grain '' nucleosomes overlap
, we end up with longer nucleosome - occupied regions containing multiple hits and surrounded by regions with zero nucleosome coverage . to each base pair in these occupied regions
we then assign an integer equal to the number of nucleosomes that cover this particular base pair , thus producing an unnormalized nucleosome occupancy profile .
finally , the occupancy profile is normalized by rescaling the integer counts separately in each occupied region so that the highest occupancy is assigned a value of 1.0 .
the total number of overlapping h2a.z nucleosome reads is 34796 ; with the conventions described above the average occupancy is 0.118 over all chromosomes . the resulting occupancy profile and the corresponding nucleosome positions were used to produce albert _ et al .
_ results for supplementary figs . 22 and 27b .
segal _ et al .
_ have recently developed a nucleosome positioning model based on the alignment of 199 sequences occupied by nucleosomes _ in vivo _ in the yeast genome .
@xcite the main assumption of the model is that it is possible to infer nucleosome formation energies using dinucleotide distributions observed in aligned nucleosomal sequences . to this end
, @xmath215 experimentally determined mononucleosome sequences of length @xmath216 bp and their reverse complements were aligned around their centers as described in segal _
_ @xcite the rules for choosing the center of the alignment should be clear from the following example : .... x aagcgttaaacgc seq1 gcgtttaacgctt seq1_rc atgcgacg seq2 cgtcgcat seq2_rc .... here x labels the center of the alignment , which is chosen in an obvious way for sequences with an odd number of nucleotides ( seq1 ) , and by making the number of nucleotides to the left of the central nucleotide one less than the number of nucleotides to the right of the central nucleotide for sequences with an even number of nucleotides ( seq2 ) .
a local average over neighboring dinucleotide positions was carried out by shifting the original alignment of @xmath217 by 1 bp to the right as well as 1 bp to the left , and adding the shifted sequences to the original alignment .
thus the final alignment has @xmath218 sequences .
next , the conditional probability of a nucleotide of type @xmath11 at position @xmath2 ( @xmath219 ) given a nucleotide of type @xmath220 at position @xmath32 ( @xmath221 ) is computed at every position within 67 bp of the center of the alignment : @xmath222 where @xmath223 , @xmath224 is the number of times the @xmath225 dinucleotide occurs in the augmented alignment , and @xmath157 marks the alignment center .
keeping conditional probabilities only for 135 central positions disregards dinucleotide counts from the outer edges of the alignment to which fewer sequences contribute .
the final model is 157 bp long because 135 conditional probabilities are flanked on each side by 11 background nucleotide frequencies , to approximate steric replusion between nucleosomes : @xmath226 where @xmath227 are the background frequencies from the yeast genome .
finally , the model assigns a negative log score ( interpreted as the energy of nucleosome formation ) to a 157 bp long nucleotide sequence @xmath209 : @xmath228 where @xmath49 denotes the nucleotide type at position @xmath2 in the sequence @xmath209 , and @xmath229 are either the genomic background frequencies defined above or uniform background frequencies : @xmath230 ( _ i.e. _ @xmath231 ) .
note that by construction the flanking nucleotides do not contribute to equation ( [ logscore ] ) ( apart from a possible difference in background frequencies ) , and thus play the role of embedding linker regions . finally , to reproduce results from segal _ et al .
_ @xcite the energies of both strands are combined using an empirical formula which implicitly sets the temperature , resulting in 13978 stable , non - overlapping nucleosomes ( with probability @xmath10 ) that cover 18.2% of the yeast genome , and the average occupancy of 0.844 over all chromosomes : @xmath232 where @xmath233 is a 157 bp sequence starting at position @xmath234 in the genome , and @xmath235 is its reverse complement .
note that genomic background frequencies are used for one strand while uniform background frequencies are used for the other .
we refer to the log scores computed in this way , and the corresponding nucleosome probabilities and occupancies as alignment model i. alternatively , in alignment model ii we use genomic background frequencies for both dna strands and rescale the temperature explicitly to match dnabend and alignment model i : @xmath236 + \langle l^{old}(s_k ) \rangle,\ ] ] where @xmath237 and @xmath238 signifies an average over the chromosome .
note that while the log scores are rescaled in equation ( [ logscore : rescaled ] ) , their mean stays the same ( and close to zero because the background scores have been subtracted , cf .
equation ( [ logscore ] ) ) .
we treat the log scores from equations ( [ logscore : combined ] ) and ( [ logscore : rescaled ] ) as nucleosome energies , and use them to compute nucleosome probabilities and occupancies as described in section [ dynapro ] .
we find that alignment model ii predictions do not strongly depend on the exact value of the scaling coefficient .
alignment model ii results in the average nucleosome occupancy of 0.810 over all chromosomes ; 26670 stable , non - overlapping nucleosomes ( with @xmath10 ) are placed , covering 34.7% of the yeast genome .
the overall correlation between alignment models i and ii is 0.66 for log scores and 0.60 for occupancies .
we used a standard competitive nucleosome reconstitution procedure to measure the relative affinity of different dna sequences for binding to histones in nucleosomes .
@xcite in this method , differing tracer dna molecules compete with an excess of unlabeled competitor dna for binding to a limiting pool of histone octamer .
the competition is established in elevated [ nacl ] , such that histone - dna interaction affinities are suppressed and the system equilibrates freely .
the [ nacl ] is then slowly reduced by dialysis , allowing nucleosomes to form ; further reduction in [ nacl ] to physiological concentrations or below then `` freezes - in '' the resulting equilibrium , allowing subsequent analysis , by native gel electrophoresis , of the partitioning of each tracer between free dna and nucleosomes .
the distribution of a given tracer between free dna and nucleosomes defines an equilibrium constant and a corresponding free energy , valid for that competitive environment .
comparison of the results for a given pair of tracer dnas ( in the identical competitive environment ) eliminates the dependence on the details of the competitive environment , yielding the free energy difference ( @xmath239 ) of histone - interaction between the two tracer dnas . to allow for comparison with other work , we include additional tracer dnas as reference molecules : a derivative of the 5s rdna natural nucleosome positioning sequence @xcite and the 147 bp nucleosome - wrapped region of the selected high affinity non - natural dna sequence 601 .
@xcite the 5s and 601 reference sequences were prepared by pcr using plasmid clones as template .
146 and 147 bp long dnas analyzed in x - ray crystallographic studies of nucleosomes ( pdb entries 1aoi and 1kx5 , respectively ) were prepared as described @xcite using clones supplied by professors k. luger and t.j .
richmond , respectively .
new 147 bp long dna sequences designed in the present study were prepared in a two - step pcr - based procedure using chemically synthesized oligonucleotide primers .
all synthetic oligonucleotides were gel purified prior to use .
the central 71 bp were prepared by annealing the two strands .
the resulting duplex was gel purified and used as template in a second stage pcr reaction to extend the length on each end creating the final desired 147 bp long dna .
the resulting dna was again purified by gel electrophoresis .
dna sequences to be analyzed were 5 end - labeled with @xmath240p , and added in tracer quantities to competitive nucleosome reconstitution reactions .
reconstitution reactions were carried out as described @xcite except that each reaction included 10 @xmath241 g purified histone octamer and 30 @xmath241 g unlabeled competitor dna ( from chicken erythrocyte nucleosome core particles ) in the 50 @xmath241l microdialysis button .
* dna templates . *
plasmids pgem-3z/601 , pgem-3z/603 and pgem-3z/605 containing nucleosome positioning sequences 601 , 603 and 605 , respectively , were described previously .
@xcite to obtain templates for hydroxyl radical footprinting experiments the desired @xmath15-bp dna fragments were pcr - amplified using various pairs of primers and taq dna polymerase ( new england biolabs ) .
the sequences of the primers will be provided on request . to selectively label either upper or lower dna strands one of the primers in each pcr reaction was 5-end radioactively labeled with polynucleotide kinase and @xmath242p - atp .
@xcite the single - end - labeled dna templates were gel - purified and single nucleosomes were assembled on the templates by dialysis from 2 m nacl .
@xcite nucleosome positioning was unique on at least 95% of the templates .
* hydroxyl radical footprinting . *
hydroxyl radicals introduce non - sequence - specific single - nucleotide gaps in dna , unless dna is protected by dna - bound proteins .
@xcite hydroxyl radical footprinting was conducted using single - end - labeled histone - free dna or nucleosomal templates as previously described .
@xcite in short , 20 - 100 ng of single - end - labeled dna or nucleosomal templates were incubated in 10 mm hepes buffer ( ph 8.0 ) in the presence of hydroxyl radical - generating reagents present at the following final concentrations ( 2 mm fe(ii)-edta , 0.6% h2o2 , 20 mm na - ascorbate ) for 2 min at @xmath243 c. reaction was stopped by adding thiourea to 10 mm final concentration .
dna was extracted with phe : chl ( 1:1 ) , precipitated with ethanol , dissolved in a loading buffer and analyzed by 8% denaturing page .
* data analysis and sequence alignment . * the denaturing gels were dried on whatman 3 mm paper , exposed to a cyclone screen , scanned using a cyclone and quantified using optiquant software ( perklin elmer ) .
positions of nucleotides that are sensitive to or protected from hydroxyl radicals were identified by comparison with the sequence - specific dna markers ( supplementary fig .
the dyad was localized by comparison of the obtained footprints with the footprints of the nucleosome assembled on human @xmath11-satellite dna .
the latter footprints were modeled based on the available 2.8 resolution x - ray nucleosome structure .
* table 1 .
* predicted ( @xmath244 ) and measured ( @xmath245 ) free energies of computationally designed sequences .
experimental free energies are shown relative to the reference sequence from the _ l.variegatus _
5s rrna gene .
histone binding is dominated by the contribution from the @xmath135 tetramer with the 71 bp binding site .
@xcite the best binder is created by using simulated annealing to introduce mutations and thus minimize the energy of a 71 bp dna molecule . b71s1 and b71s2 have different sequences flanking the 71 bp designed site ( whose contribution dominates the total free energy ) . 601s1 and 601s2 consist of the 71 bp site from the center of the 601 sequence @xcite and flanking sequences from b71s1 and b71s2 , respectively .
w147s is a 147 bp sequence whose free energy ( with contributions from multiple @xmath135 binding sites ) was maximized by simulated annealing . x146 and x147 are 146 bp and 147 bp dna sequences from nucleosome crystal structures 1aoi @xcite and 1kx5.@xcite + twist & 0.130 & 0.041 & @xmath2460.046 & @xmath2460.159 & -0.305 & 0.739 + roll & 0.041 & 0.069 & @xmath2460.026 & @xmath2460.053 & -0.213 & -0.135 + tilt & @xmath2460.046 & @xmath2460.026 & 0.406 & -0.408 & @xmath2470.073 & @xmath2470.403 + shift & @xmath2460.159 & @xmath2460.053 & -0.408 & 11.948 & @xmath2462.162 & @xmath2472.511 + slide & -0.305 & -0.213 & @xmath2470.073 & @xmath2462.162 & 8.241 & -5.353 + rise & 0.739 & -0.135 & @xmath2470.403 & @xmath2472.511 & -5.353 & 35.715 + + twist & 0.179 & 0.053 & @xmath2460.015 & @xmath2470.083 & -0.118 & 1.244 + roll & 0.053 & 0.085 & @xmath2460.003 & @xmath2460.048 & 0.049 & 0.089 + tilt & @xmath2460.015 & @xmath2460.003 & 0.323 & -0.312 & @xmath2470.298 & @xmath2470.219 + shift & @xmath2470.083 & @xmath2460.048 & -0.312 & 4.912 & @xmath2461.672 & @xmath2470.872 + slide & -0.118 & 0.049 & @xmath2470.298 & @xmath2461.672 & 10.089 & -3.277 + rise & 1.244 & 0.089 & @xmath2470.219 & @xmath2470.872 & -3.277 & 35.709 + + twist & 0.113 & 0.038 & @xmath2470.001 & @xmath2460.168 & 0.040 & 0.850 + roll & 0.038 & 0.077 & @xmath2460.022 & @xmath2460.006 & -0.023 & 0.067 + tilt & @xmath2470.001 & @xmath2460.022 & 0.280 & -0.365 & @xmath2460.252 & @xmath2470.995 + shift & @xmath2460.168 & @xmath2460.006 & -0.365 & 4.954 & @xmath2460.527 & @xmath2460.093 + slide & 0.040 & -0.023 & @xmath2460.252 & @xmath2460.527 & 4.516 & -2.966 + rise & 0.850 & 0.067 & @xmath2470.995 & @xmath2460.093 & -2.966 & 29.330 + + twist & 0.098 & 0.027 & @xmath2470.034 & @xmath2460.061 & -0.254 & 0.799 + roll & 0.027 & 0.059 & @xmath2470.046 & @xmath2460.165 & -0.016 & 0.202 + tilt & @xmath2470.034 & @xmath2470.046 & 0.393 & -0.965 & @xmath2460.174 & @xmath2470.593 + shift & @xmath2460.061 & @xmath2460.165 & -0.965 & 5.740 & @xmath2470.117 & @xmath2461.963 + slide & -0.254 & -0.016 & @xmath2460.174 & @xmath2470.117 & 2.772 & -4.449 + rise & 0.799 & 0.202 & @xmath2470.593 & @xmath2461.963 & -4.449 & 23.870 + + twist & 0.114 & 0.045 & @xmath2460.008 & @xmath2470.139 & -0.328 & 1.587 + roll & 0.045 & 0.075 & @xmath2470.014 & @xmath2460.005 & -0.083 & 0.797 + tilt & @xmath2460.008 & @xmath2470.014 & 0.218 & -0.537 & @xmath2470.215 & @xmath2460.066 + shift & @xmath2470.139 & @xmath2460.005 & -0.537 & 3.917 & @xmath2460.085 & @xmath2470.401 + slide & -0.328 & -0.083 & @xmath2470.215 & @xmath2460.085 & 3.795 & -7.972 + rise & 1.587 & 0.797 & @xmath2460.066 & @xmath2470.401 & -7.972 & 41.678 + + twist & 0.133 & 0.055 & @xmath2460.027 & @xmath2460.073 & -0.350 & 1.301 + roll & 0.055 & 0.097 & @xmath2460.042 & @xmath2470.099 & -0.158 & 0.336 + tilt & @xmath2460.027 & @xmath2460.042 & 0.408 & -1.012 & @xmath2460.016 & @xmath2460.139 + shift & @xmath2460.073 & @xmath2470.099 & -1.012 & 10.434 & @xmath2463.814 & @xmath2471.105 + slide & -0.350 & -0.158 & @xmath2460.016 & @xmath2463.814 & 6.278 & -7.676 + rise & 1.301 & 0.336 & @xmath2460.139 & @xmath2471.105 & -7.676 & 41.988 + + twist & 0.101 & 0.020 & 0.000 & 0.000 & -0.278 & 0.806 + roll & 0.020 & 0.040 & 0.000 & 0.000 & 0.001 & -0.046 + tilt & 0.000 & 0.000 & 0.255 & -0.510 & 0.000 & 0.000 + shift & 0.000 & 0.000 & -0.510 & 3.104 & 0.000 & 0.000 + slide & -0.278 & 0.001 & 0.000 & 0.000 & 3.991 & -2.936 + rise & 0.806 & -0.046 & 0.000 & 0.000 & -2.936 & 20.174 + + twist & 0.069 & 0.006 & 0.000 & 0.000 & -0.217 & 0.646 + roll & 0.006 & 0.057 & 0.000 & 0.000 & 0.094 & 0.168 + tilt & 0.000 & 0.000 & 0.256 & -0.542 & 0.000 & 0.000 + shift & 0.000 & 0.000 & -0.542 & 3.473 & 0.000 & 0.000 + slide & -0.217 & 0.094 & 0.000 & 0.000 & 4.030 & 1.322 + rise & 0.646 & 0.168 & 0.000 & 0.000 & 1.322 & 34.392 + + twist & 0.189 & 0.068 & 0.000 & 0.000 & 0.111 & 1.195 + roll & 0.068 & 0.124 & 0.000 & 0.000 & 0.438 & -0.397 + tilt & 0.000 & 0.000 & 0.641 & -0.043 & 0.000 & 0.000 + shift & 0.000 & 0.000 & -0.043 & 6.670 & 0.000 & 0.000 + slide & 0.111 & 0.438 & 0.000 & 0.000 & 15.942 & -2.611 + rise & 1.195 & -0.397 & 0.000 & 0.000 & -2.611 & 47.789 + and @xmath1 are shown in blue , and the mst coordinate frame
is shown in green .
for illustrative purposes , only rise @xmath248 and twist @xmath249 are set to non - zero values .
the origin of the mst frame is at the midpoint of the line connecting the origins of two base pair frames ( which are separated by @xmath248 along the z - axis ) ; the mst frame is rotated through @xmath250 with respect to the frame @xmath2 . ] ( @xmath251 ) .
black curve in the slide panel is a dnabend prediction based on a modified elastic potential in which the mean slide for the ca dinucleotide was increased from 0.18 ( inferred from a set of non - nucleosome protein - dna complexes and used in this work ) to 0.91 ( inferred from a limited set of available nucleosome structures ) .
note that the ca dinucleotides occur at positions ( 3,10,16,26,50,53,60,66,77,85,102,109,115,133 ) in the 1kx5 crystal structure .
@xcite whereas the slide correlation coefficient between the native and minimized structures remained essentially the same at 0.533 , the absolute magnitude of predicted slide peaks is in better correspondence with the observed values .
thus an elastic potential trained on non - nucleosome complexes may not reproduce all structural aspects of highly bent nucleosomal dna . ] .
@xcite ( b ) the 183 bp sequence from the pgub plasmid [ bps 11,31 ] .
@xcite ( c ) the 215 bp fragment from the sequence of the chicken @xmath252 gene [ bp 52 ] .
@xcite ( d , e , f ) synthetic high - affinity sequences @xcite 601 [ bp 61 ] , 603 [ bp 81 ] , and 605 [ bp 59 ] .
experimentally known nucleosome starting positions are listed in a one - based coordinate system with consecutively numbered base pairs .
nucleosomes on sequences 601 , 603 , and 605 were mapped by hydroxyl radical footprinting ( supplementary figs .
13 and 14 ) .
all dna sequences used in this calculation are available on the nucleosome explorer website : _
http://nucleosome.rockefeller.edu_. ] ) predicted using dnabend ( a ) and alignment model i ( b ) .
red : null model in which the same number of stable nucleosomes is randomly positioned on the genome without overlap .
solid red lines are mean values for 100 random placements , dashed red lines are one standard deviation away .
alignment model ii exhibits similar oscillations ( data not shown ) . ]
( with pwm constructed using a combination of a transfac mcm1 pwm and a structure - based 1mnm pwm @xcite ) .
note that the experimentally mapped nucleosomes are phased off the region occupied by @xmath253 ; however , in the dnabend simulation nucleosomes move to the left instead , perhaps because we overestimate the stability of the nucleosome starting at bp 322200 . ]
( same pwm as in supplementary fig .
note that in contrast to the _ bar1 _ locus ( supplementary fig .
26d ) , nucleosomes are shifted in the right direction when @xmath253 binds , phasing off the tf - occupied region . ]
locus , with experimental nucleosome positions mapped by weiss _
@xcite dark blue : origin recognition complex ( pwm based on the consensus sequence from ref .
@xcite ) , dark orange : abf1 , pink : rap1 .
note that two prominent gaps in the otherwise regularly spaced array of experimentally mapped nucleosomes coincide with the regions of high tf occupancy . ]
locus , with experimental nucleosome positions mapped by ravindra _
@xcite dark blue : origin recognition complex ( pwm based on the consensus sequence from ref .
@xcite ) , dark orange : abf1 , pink : rap1 .
similar to the @xmath12 locus ( supplementary fig .
26f ) , two prominent gaps in the otherwise regularly spaced array of experimentally mapped nucleosomes coincide with the regions of high tf occupancy . ]
( same pwm as in supplementary fig .
similar to the @xmath12 and @xmath13 loci ( supplementary figs .
26f , g ) , two prominent gaps in the otherwise regularly spaced array of experimentally mapped nucleosomes coincide with the regions of high tf occupancy . ]
. distances between neighboring nucleosomes are sampled from a gaussian distribution with the mean chosen to reproduce the average occupancy predicted by the actual model , and the standard deviation set to 0.5 of the mean ( negative linker lengths are not allowed ) .
thus in the null models nucleosomes are positioned non - specifically but on average form a regular array .
the only parameter borrowed from the actual model is the predicted average occupancy .
the center - to - center nucleosome distance is determined by locating in the periodic occupancy profile the nearest region with occupancy of at least 0.35 over the nucleosomal length .
error bars that correspond to 100 random realizations of the null models are too small to be shown .
note that in both ( a ) and ( b ) red and cyan curves nearly overlap ; adding tf and tbp binding to the alignment models does not improve their performance ( data not shown ) . ] bouchiat , c. , wang , m.d . ,
allemand , j .- f . , strick , t. , block , s.m . and
croquette , v. ( 1999 ) estimating the persistence length of a worm - like chain molecule from force - extension measurements .
_ biophys.j .
_ , * 76 * , 409413 .
macisaac , k.d . , wang , t. , gordon , d.b . , gifford , d.k . , stormo , g.d . and
fraenkel , e. ( 2006 ) an improved map of conserved regulatory sites for _ saccharomyces cerevisiae_. _ bmc bioinformatics _ , * 7 * , 113 .
albert , i. , mavrich , t.n . ,
tomsho , l.p . , qi , j. , zanton , s.j . , schuster , s.c . and
pugh , b.f .
( 2007 ) translational and rotational settings of h2a.z nucleosomes across the _ saccharomyces cerevisiae _ genome .
_ nature _ , * 446 * , 572576 .
dyer , p.n . ,
edayathumangalam , r.s . , white , c.l . , bao , y. , chakravarthy , s. , muthurajan , u.m . and luger , k. ( 2004 ) reconstitution of nucleosome core particles from recombinant histones and dna .
_ methods enzymol .
_ , * 375 * , 23-44 .
thastrom , a. , lowary , p.t .
, widlund , h.r . ,
cao , h. , kubista , m. and widom , j. ( 1999 ) sequence motifs and free energies of selected natural and non - natural nucleosome positioning dna sequences .
_ j.mol.biol._ , * 288 * , 213229 .
widlund , h.r . ,
cao , h. , simonsson , s. , magnusson , e. , simonsson , t. , nielsen , p.e . ,
kahn , j.d . , crothers , d.m . and kubista , m. ( 1998 ) identification and characterization of genomic nucleosome - positioning sequences .
_ j.mol.biol._ , * 267 * , 807817 .
kassabov , s.r . ,
henry , n.m . ,
zofall , m. , tsukiyama , t. and bartholomew , b. ( 2002 ) high - resolution mapping of changes in histone - dna contacts of nucleosomes remodeled by isw2 . _ mol.cell.biol._ , * 22 * , 75247534 .
moreira , j.m . and
holmberg , s. ( 1998 ) nucleosome structure of the yeast cha1 promoter : analysis of activation - dependent chromatin remodeling of an rna - polymerase - ii - transcribed gene in tbp and rna pol ii mutants defective in vivo in response to acidic activators . _
embo j. _ , * 17 * , 60286038
. almer , a. , rudolph , h. , hinnen , a. and hrz , w. ( 1986 ) removal of positioned nucleosomes from the yeast pho5 promoter upon pho5 induction releases additional upstream activating dna elements . _ embo j. _ , * 5 * , 26892696 .
venter , u. , svaren , j. , schmitz , j. , schmid , a. and hrz , w. ( 1994 ) a nucleosome precludes binding of the transcription factor pho4 _ in vivo _ to a critical target site in the pho5 promoter .
_ embo j. _ , * 13 * , 48484852 .
shimizu , m. , roth , s.y .
, szent - gyorgyi , c. and simpson , r.t .
( 1991 ) nucleosomes are positioned with base pair precision adjacent to the @xmath14 operator in _
saccharomyces cerevisiae_. _ embo j. _ , * 10 * , 30333041 .
ravindra , a. , weiss , k. and simpson , r.t .
( 1999 ) high - resolution structural analysis of chromatin at specific loci : _ saccharomyces cerevisiae _ silent mating - type locus @xmath13 .
_ mol.cell.biol._ , * 19 * , 79447950 .
sekinger , e.a .
, moqtaderi , z. and struhl , k. ( 2005 ) intrinsic histone - dna interactions and low nucleosome density are important for preferential accessibility of promoter regions in yeast .
_ mol.cell_ , * 18 * , 735748 . | in eukaryotic genomes , nucleosomes function to compact dna and to regulate access to it both by simple physical occlusion and by providing the substrate for numerous covalent epigenetic tags . while nucleosome positions _ in vitro _ are determined by sequence alone , _ in vivo _ competition with other dna - binding factors and action of chromatin remodeling enzymes play a role that needs to be quantified .
we developed a biophysical model for the sequence dependence of dna bending energies , and validated it against a collection of _ in vitro _
free energies of nucleosome formation and a nucleosome crystal structure ; we also successfully designed both strong and poor histone binding sequences _
ab initio_. for _ in vivo _ data from _ s.cerevisiae_ , the strongest positioning signal came from the competition with other factors .
based on sequence alone , our model predicts that functional transcription factor binding sites have a tendency to be covered by nucleosomes , but are uncovered _ in vivo _ because functional sites cluster within a single nucleosome footprint , making transcription factors bind cooperatively .
similarly a weak enhancement of nucleosome binding in the tata region for naked dna becomes a strong depletion when the tata - binding protein is included , in quantitative agreement with experiment .
predictions at specific loci were also greatly enhanced by including competing factors .
our physically grounded model distinguishes multiple ways in which genomic sequence can influence nucleosome positions and thus provides an alternative explanation for several important experimental findings .
= 1 |
a long sought - after goal is the all - optical processing of information , replacing state - of - the - art electronic handling .
it is well known , however , that major obstacles arise in particular at spatial scales small compared to the involved wavelengths .
light can not be focused to arbitrarily small areas , and it can not directly be used to image or write structures below a scale given by the involved wavelength @xcite . a key origin for these restrictions is diffraction , leading , e.g. , to the rayleigh limit for the case of imaging @xcite . due to the enormous significance for applications ,
e.g. , in nano- or life - sciences , a number of schemes have been invented to circumvent or even surpass the standard limits in a various setups .
examples are sub - wavelength microscopy @xcite , and optical sub - wavelength lithography @xcite .
but more fundamentally , diffraction also already impacts the propagation of a light beam in free space .
for example , a gaussian probe beam with a certain width at the focal plane spreads as it propagates away from the focal plane , on a length scale given by the rayleigh length @xcite .
this paraxial diffraction spreading occurs in free space as well as inside a medium , and affects applications such as the transmission of small images , the transfer of small mask structures to the target surface in lithography , or the steering or optical manipulation of light beams .
it was recognized that this spreading due to paraxial diffraction can be decreased , increased or completely eliminated by propagating the light beam through a medium with spatially varying optical properties .
this can be achieved via the use of external coherent control fields , inducing a spatially varying susceptibility and inducing phase shifts via different physical processes @xcite .
the spatial profile of a strong control beam renders the medium inhomogeneous along the transverse direction . since absorption and dispersion of the medium are crucially dependent on the intensity and the shape of the strong control beam
, the control beam will strongly affect the propagation of the probe beam . the ability to control the spatially varying susceptibility of the medium via external fields holds the key to eventually being able to control the spreading of optical beams . these and related techniques are of significance also because they have the potential of complementing optical information transmission with all - optical information processing , thereby replacing the currently ubiquitous electronic information processing .
( color online ) diffractionless beam propagation through a vapor cell .
( a ) shows a schematic setup to study the propagation of a probe laser field through an atomic vapor cell .
the propagation dynamics is controlled using an additional control laser field with a laguerre - gaussian transversal profile .
( b ) three - level @xmath0 configuration of the vapor atoms.,width=226 ] thus it is not surprising that a great number of schemes both in theory and in experiment demonstrated the ability to optically control the propagation of light .
most of these works can be understood in analogy to a waveguide , confining the light in specific modes transversally .
this could enable one to implement applications originally proposed for physical waveguides @xcite with all - optically induced waveguides .
for example , it was shown that spatial variations in a medium induced by a control laser field can lead to focusing or de - focusing of a probe beam @xcite .
an optically written waveguide realized by applying a laguerre - gaussian control field near - resonantly to an atomic medium to propagate a probe beam was demonstrated in @xcite and discussed in @xcite .
waveguiding in a warm rubidium vapor based on an off - resonant raman configuration was reported in @xcite .
a signal beam was shown to propagate with small spot size over several diffraction lengths . in a variant of such setups ,
the spatial profile of a control beam can be cast onto a probe beam .
recently , this concept was implemented in rubidium atomic vapor , based on coherent population trapping @xcite . in this experiment ,
a spatial profile was transferred from the coupling beam onto the probe beam , which was found to exhibit feature sized five times smaller than the diffraction - limited size .
however , all of these schemes suffer from the fact that the modes other than the specific supported mode can not propagate without diffraction , or even not at all .
this prohibits the desired transmission of arbitrary images . in order to overcome this limitation , very recently
, a scheme for the direct propagation of arbitrary probe beam images without diffraction was put forward theoretically @xcite and verified experimentally @xcite .
this setup is based on an atomic three - level system reminiscent of electromagnetically induced transparency , but assumes an explicit breaking of the usual two - photon resonance condition such that strong absorption occurs .
the underlying physics in this setup is fundamentally different from previous setups , as it relies on the movement of the atoms in order to compensate the diffraction , rather than on a stationary waveguide .
a method to propagate optical images with reduced diffraction based on the saturated absorption mechanism in a two - level atom was proposed in @xcite .
future ideas along this line therefore can be expected to advance key applications in optical information transmission and processing , as well in reading and writing at the sub - wavelength scale . in this paper
, we discuss the light propagation through waveguides written by a spatially structured control laser field into a warm atomic vapor . the laguerre - gaussian control field and the probe laser field couple with the atoms to form a three - level system in @xmath0 configuration as shown in fig .
we start by deriving analytical results for the steady state susceptibility experienced by the probe beam .
we then consider a thermal medium by averaging the susceptibility over a maxwellian velocity distribution of the vapor atoms .
the probe beam propagation dynamics is analyzed by numerically solving the maxwell - schrdinger equation including diffraction and dispersion in two spatial dimensions transversal to the propagation direction , the spatial profile of the control field , and the thermal motion of the vapor atoms .
we find that for suitable parameter choices , a single gaussian probe field mode can be propagated diffraction - less or even with sub - diffraction through the warm vapor .
we then proceed to analyze more complex mode structures of the probe beam , and show that a gaussian double pulse can be propagated without distortion or diffraction through the optically written waveguide .
finally , we demonstrate that a beam formed out of multiple secant - hyperbolic shaped pulses can be propagated diffraction - less through the warm atomic vapor .
the results on non - gaussian beam propagation open perspectives for the propagation of arbitrary images in warm atomic vapors .
the paper is organized as follows . in sec .
[ theory ] , we introduce our model , discuss the susceptibility of a thermal vapor , and describe the maxwell - schrdinger equation used to analyze the propagation dynamics of the probe beam through the optically written waveguide . in sec .
[ tailor ] , we discuss the static properties of the medium susceptibility under the action of the spatially structured control laser beam . in sec . [ prop ] , we then discuss our results on the propagation of different probe beams through the optically written waveguide . finally , sec .
[ summary ] discusses and summarizes our results .
we consider the geometry as shown in fig . [ fig1 ] .
a weak probe laser field with frequency @xmath1 is focused at the entry of an atomic vapor , and co - propagating with an additional control laser field with frequency @xmath2 focused at the back of the vapor cell .
the cell is filled with a gas of three level atoms in @xmath0-configuration as shown in fig . [ fig1](b ) .
the electric fields of the two beams can be defined as [ probe ] @xmath3 where @xmath4 is the slowly varying envelope and @xmath5 is unit polarization vector of the field .
the index @xmath6 denotes the probe @xmath7 or control @xmath8 field , respectively .
the polarization vectors of the beams are chosen in such a way that the control beam @xmath9 is tuned to the atomic transition @xmath10 while the probe beam @xmath11 is coupled to the atomic transition @xmath12 .
the hamiltonian of the @xmath0-system under electric - dipole and rotating - wave approximation can then be written as [ hschroed ] @xmath13 where @xmath14 denotes the resonance frequency on the transition @xmath15 , @xmath16 is the matrix element of the electric dipole moment operator @xmath17 , and we have set the energy of state @xmath18 to zero .
we now make use of the unitary transformation @xmath19 to rewrite the hamiltonian as @xmath20 here , @xmath21 for @xmath6 are the detunings of the fields defined as @xmath22 the rabi frequencies of the probe and control fields @xmath23 and @xmath24 are related to the slowly varying amplitudes of @xmath25 and @xmath26 as @xmath27 as shown in fig .
[ fig1](b ) , the upper atomic state @xmath28 decays to @xmath18 with rate @xmath29 , and to @xmath30 with rate @xmath31 .
furthermore , @xmath32 is the decay rate of the ground state atomic coherence . in order to incorporate these incoherent processes
, we use the master equation approach to describe the dynamics of the atomic populations and coherences of the three level atoms . from eq .
( [ h ] ) , the equations of motion for the elements of the density matrix @xmath33 can be written as [ full_density ] @xmath34\rho_{12 } + { \imath}g \rho_{22 } \nonumber\\ & + { \imath}g \rho_{32 } - { \imath}g\rho_{11 } \,,\\ \dot{\rho}_{13}=&-\left[\frac{1}{2}(\gamma_1 + \gamma_2 ) + { \imath}{\delta_c}\right]\rho_{13 } + { \imath}g\rho_{23 } \nonumber\\ & + { \imath}g\rho_{33 } - { \imath}g\rho_{11}\,,\\ \dot{\rho}_{23}= & -[\gamma - { \imath}(\delta_p-\delta_c)]\rho_{23 } + { \imath}g^*\rho_{13 } - { \imath}g\rho_{21}\ , . \end{aligned}\ ] ] the third population follows from the conservation law @xmath35 . in the following , we for simplicity restrict the analysis to the case @xmath36 . in the steady state
, the density matrix equations ( [ full_density ] ) can be solved analytically .
we are in particular interested in the steady state value of the atomic coherence @xmath37 , which determines the probe field susceptibility @xmath38 of the medium at frequency @xmath1 .
the analytical expression for @xmath38 to all orders in the control and probe rabi frequencies is given in appendix [ susceptibility ] .
the susceptibility crucially depends on the two photon detuning of the raman system defined as @xmath39 for a warm atomic medium , the thermal motion of the atoms causes inhomogeneous broadening of the atomic spectra
. this broadening plays a important role when considering the optical properties of the medium to determine the probe beam propagation dynamics through the medium @xcite .
we incorporate the motion in the susceptibility @xmath38 in eq .
( [ mollow_chi ] ) by defining velocity - dependent detuning of a given atom [ detuning_dop ] @xmath40 where @xmath41 is the velocity of the atom . for simplicity ,
we have assumed assume @xmath42 .
the velocities of all atoms obey the maxwellian distribution @xmath43 with @xmath44 is the doppler width defined by @xmath45 at temperature @xmath46 . for a warm atomic system ,
the averaged susceptibility is given by @xmath47 in general , the medium susceptibility is a function of the position in the vapor cell due to the spatial structure of the applied laser fields . in order to account for the spatial variations , we study the propagation of the light beams through the medium using maxwell s equations . for this
, we use the slowly varying envelope approximation , and assume that the changes along the longitudinal direction @xmath48 of the probe envelope is very small compared to the changes which occur along the transverse directions @xmath49 .
the beam propagation equation for the rabi frequency @xmath50 of the probe field under the paraxial wave approximation then follows as @xmath51 where @xmath52 and @xmath53 is the speed of light in free space .
the thermal motion of the atoms is included in the the beam propagation by considering the average response of the susceptibility @xmath54 .
the terms in the parentheses in eq .
( [ max1 ] ) are responsible for the diffraction of the probe beam . in order to study the effect of diffraction and spatial dispersion on the probe beam propagation
, we solve eq .
( [ max1 ] ) numerically using a higher order split step operator method @xcite .
( color online ) normalized spatial intensity profile of the control laser field .
the solid ( dashed ) line shows the profile at the entry ( exit ) plane of the vapor cell .
@xmath55 is the radial distance from the beam center .
the beam waist and the rayleigh length of the laguerre - gaussian beam are chosen as 120 @xmath56 m and 5.7 cm , respectively.,width=6 ] in order to illustrate how the spatial structure of the control field enables us to guide the weak probe beam through warm atomic vapor , we first study the behavior of the susceptibility of the medium as a function of the spatial variation of the control field intensity .
we chose a doughnut shaped laguerre - gaussian ( lg ) control field to induce a waveguide - like refractive index pattern inside the atomic vapor .
the rabi frequency of the lg control field can be written as [ control_field ] @xmath57\ , , \\
r & = \sqrt { x^2 + y^2 } \,,\label{beam_waist}\\ w_{z}&=w_{c } \sqrt { 1 + \left ( { \frac{z - z_{0}}{{z_{r } } } } \right)^2}\,\\ q&=\imath z_{r } -z + z_0\,,\\ \theta&=\tan ^ { - 1 } \left [ { \frac{y}{x } } \right]\,,\\ z_{r } & = \frac{{\pi w_{c}^2 } } { \lambda } \,,\end{aligned}\ ] ] where @xmath58 and @xmath59 are the location of the beam waist and the rayleigh length , respectively .
figure [ fig2 ] shows the intensity variation of the control field against radial position @xmath55 at the entry and exit faces of the medium . for this figure ,
we have considered a vapor cell of length 5 cm and filled with rb atoms .
the intensity of the control field approaches zero for @xmath60 and maxima occur at @xmath61 cm at the entry face of the medium .
it should be noted that there is no significant change in the structure of the control field envelope throughout the propagation through the medium .
this is consistent with the results in the experiment by howell et al @xcite .
from the point of view of an additional probe beam , the control field structure renders the susceptibility inhomogeneous along the transverse direction .
( color online ) real part ( a ) and imaginary part ( b ) of the averaged susceptibility @xmath62 in a thermal vapor obtained from eq .
( [ chi_average ] ) in the presence of the lg control beam .
the results are shows as a function of the radial position @xmath55 in the beam and the raman detuning @xmath63 .
the other parameters are chosen as @xmath64 , @xmath65 m , @xmath66 , @xmath67 , @xmath68 , @xmath69 atoms/@xmath70 and @xmath71 . ]
( color online ) normalized intensity of the transmitted probe beam plotted against the radial distance from the beam center @xmath55 .
the solid black line shows the profile at the input of the vapor cell .
the other two curves show the profile at the output after a propagation length of 5 cm .
the red short - dashed curve is obtained without control field , whereas the blue long - dashed line is obtained with a lg control laser field .
the inset shows the variation of the unnormalized output intensity of the probe beam .
the initial width and raman detuning of the off - resonant probe field are chosen as @xmath72=48@xmath56 m and @xmath73 , respectively .
all other parameters are as in fig.[fig : fig3].,width=226 ] next , we have solved eq .
( [ chi_average ] ) numerically in the presence of a continuous wave probe field to analyze the spatial inhomogeneity of the susceptibility .
figure [ fig : fig3 ] shows surface plots of the real and imaginary parts of @xmath62 as a function of the radial position @xmath74 and the raman detuning @xmath75 .
for the calculation , we have fixed the single photon probe detuning to @xmath67 , and use typical parameters for the @xmath76rb transition : @xmath77 rad / sec , @xmath78 rad / sec , and @xmath69 atoms/@xmath79 .
it can be seen from fig .
[ fig : fig3](a ) that the refractive index is high at radial positions symmetrically located around @xmath80 at which the control intensity is high . towards the beam center and for larger radial distances
@xmath81 , it decreases .
the difference between the maximum and the minimum of the refractive index is of order 10@xmath82 , which is in agreement with the experimental results of howell et .
al @xcite .
this shows that a lg control beam allows one to mimic a fibre like refractive index profile inside the atomic medium . from fig .
[ fig : fig3](b ) we find that the absorption of the medium has a minimum at @xmath83 , whereas an increase or decrease of the raman detuning from this value results in increased absorption by the medium
. the analytical expressions eqs .
( [ control_field ] ) for the control field rabi frequency allow us to study the effect of the spatial structure of the control beam on the probe beam propagation dynamics . for this
, we have included the control field profile eqs .
( [ control_field ] ) in the propagation equation ( [ max1 ] ) .
first , we study the propagation dynamics of a gaussian probe beam .
the shape of the probe beam is defined as @xmath84 where @xmath85 is the beam width .
numerical results are shown in fig .
[ fig4 ] , for a rayleigh length of the probe beam chosen as @xmath86 cm .
the figure compares the propagation of the probe beam through the medium in the presence and the absence of the control field . in free space or in the absence of the control beam , we observe from fig .
[ fig4 ] that the probe beam spreads to nearly 5.64@xmath72 during the propagation over a distance of 5 cm .
the most intuitive way to understand the spreading of the beam is by decomposing it into a set of plane waves . in the course of the propagation
, each plane wave acquires its own unique phase shift .
the superposition of the phase shifted plane waves at the output gives rise to spreading .
in contrast , for bessel , mathieu and airy beams each constituent plane wave acquires exactly the same phase shift . in these cases ,
diffractionless propagation is possible through free space .
analogously , the manipulation of diffraction - induced spreading is possible by changing the spatial dispersion of the medium .
this spatial dispersion can be substantially modified by employing a suitable structure of the control field .
in particular , a suitable structure of the control field can cancel or even reverse the effect of diffraction .
our numerical results in fig .
[ fig4 ] obtained with the lg control field show that the width of the probe beam is 37@xmath56 m after a 5 cm propagation distance .
thus , the width of the output probe beam in the presence of the lg control beam is narrower than the width of the input probe beam , demonstrating that the lg control field can be used to reverse and even over - compensate the effect of diffraction .
these results are in good agreement with the experiment of howell et al @xcite .
we also notice from the inset of fig .
[ fig4 ] that the integrated transmission intensity of the output probe beam is 44@xmath87 of the integrated intensity of the input probe beam .
this indicates that the probe beam can be guided efficiently in an atomic vapor .
next , we study the propagation of a beam with two gaussian modes , using the same method as for the single gaussian probe beam . figure [ fig : fig5a ] shows the normalized intensity of both transmitted gaussian beams in the @xmath88 plane along the @xmath48 axis for a propagation length up to @xmath89 cm in the presence of the lg control field . in comparison , without the control beam or in free space the incident probe beam undergoes spreading due to its geometrical shape as shown in fig .
[ fig : fig5b ] .
it can be seen from fig [ fig : fig5b ] that again , the spreading due to diffraction can be eliminated using the lg control beam .
the inset of fig .
[ fig : fig5b ] shows the unnormalized transmission of the two gaussian beams in the @xmath88 plane .
it is clear from the inset of the fig .
[ fig : fig5b ] that the transmitted energy of probe beam gets reduced by @xmath90 while the width of the lg control beam is changed from @xmath91 m to @xmath92 m .
we also notice that the spacing between two peaks of the output probe beam can be increased by decreasing the width of the lg control beam .
this shift in the probe beam peak positions arises if the probe beam spectrum is located on the sides of the control field intensity pattern , rather than symmetrically around the control field intensity maximum .
as an application , the finesse of the output probe beam can be controlled by changing the width of the lg control beam .
figure [ fig6 ] shows the radial distribution of the output probe beams at @xmath93 cm , for a medium without and with the lg control beam , respectively . as in fig .
[ fig : fig5b ] it can be seen that in free space or without the control beam , the beam profiles are severely distorted .
in contrast , with the lg control beam , the shape of the two gaussian beams is preserved .
this shows that the spatial structure of the control field can be used to transmit more complicated structures over a large distance without much diffraction and attenuation .
we proceed with a probe beam formed out of multiple secant - hyperbolically shaped patterns as an example for the transmission of a non - gaussian field mode .
the probe beam is defined as @xmath94,\end{aligned}\ ] ] where @xmath95 , @xmath96 and @xmath97 are the total number , the center , and the widths of the peaks , respectively . for simplicity , in our calculation , we assume that the individual peaks have equal widths @xmath97 .
figure [ fig : fig7a ] shows the radial variation of probe beam intensity in the presence of the lg control beam at the exit plane of the vapor at @xmath93 cm . a corresponding comparison of the non - gaussian beam propagation through the medium with and without the control beam is shown in fig .
[ fig : fig7b ] .
it is important to note from fig .
[ fig : fig7b ] that the output beam retains sub - peaks of secant - hyperbolic profile as they were present in the input beam .
thus the lg control beam allows to propagate gaussian as well as non - gausssian beams through the medium without diffraction .
we have discussed diffraction - less propagation of gaussian and non - gaussian probe beams through a warm atomic vapor suitably tailored using an additional control beam .
the laguerre - gaussian control beam writes a spatially modulated index of refraction in the vapor , which leads to a modified dispersion experienced by the probe beam , and subsequently to diffraction - less propagation of the probe beam .
we have shown that probe beams of single gaussian , multiple gaussion , and multiple non - gaussian structure can be propagated through the medium without diffraction .
tnd thanks p. k. vudyasetu for correspondence on off - resonant absorption in a @xmath0-type system .
the steady - state susceptibility is given by @xmath98 where the numerator @xmath99 and the denominator @xmath100 evaluate to @xmath101\ , , \\
d=&\gamma |g|^6 + |g|^4\left\{3|g|^2(\gamma+2\gamma)+2\gamma\left(\gamma\gamma+{\delta_{r}}{\delta_c}\right)\right\ } \nonumber\\ & + \gamma |g|^2\left\{(\gamma^2+{\delta_{r}}^2)(\gamma^2+{\delta_p}^2)+2|g|^2(\gamma\gamma+{\delta_{r}}{\delta_p } ) \right .
\nonumber \\ & \left .
+ |g|^4\right\ } + |g|^2 \left\{3 |g|^4 ( \gamma + 2 \gamma ) \right .
\nonumber \\ & \left .
+ \gamma \left(\gamma ^2+{\delta_c}^2\right ) \left(\gamma ^2+{\delta_{r}}^2\right)+\left(4 \gamma ^2 \gamma + \gamma \left(6 \gamma ^2 + 4{\delta_{r}}^2\right ) \right .
\nonumber \\ & \left .
+ \gamma ( { \delta_c}+{\delta_p})^2\right ) |g|^2 \right\}\,.\end{aligned}\ ] ] p. storey , m. collett , and d. f. walls , phys .
lett . * 68 * , 472 ( 1992 ) ; phys .
a * 47 * , 405 ( 1993 ) ; r. quadt , m. collett , and d. f. walls , phys . rev . lett . * 74 * , 351 ( 1995 ) .
s. kunze , k. dieckmann , and g. rempe , phys .
lett . * 78 * , 2038 ( 1997 ) .
a. m. herkommer , w. p. schleich , and m. s. zubairy , j. mod . opt . * 44 * , 2507 ( 1997 ) .
chang , j. evers , m. o. scully , and m. s. zubairy , phys .
rev . a * 73 * , 031803(r ) ( 2006 ) . | diffractionless propagation of optical beams through atomic vapors is investigated .
the atoms in the vapor are operated in a three - level raman configuration . a suitably chosen control
beam couples to one of the transitions , and thereby creates a spatially varying index of refraction modulation in the warm atomic vapor for a probe beam which couples to the other transition in the atoms .
we show that a laguerre - gaussian control beam allows to propagate single gaussian probe field modes as well as multi - gaussian modes and non - gaussian modes over macroscopic distances without diffraction .
this opens perspectives for the propagation of arbitrary images through warm atomic vapors . |
modern high - speed optical communication systems require high - performing implementations that support throughputs of 100 gbit / s or multiples thereof , that have low power consumption , that realize close to the theoretical limits at a target of @xmath0 , and that are preferably adapted to the peculiarities of the optical channel . especially with the advent of coherent transmission schemes and the utilization of high resolution , soft - decision decoding has become an attractive means of reliably increasing the transmission reach of lightwave systems . currently , there are two popular classes of codes for soft - decision decoding that are attractive for implementation in optical receivers at decoding throughputs of 100 gbit / s and above : codes and .
the latter can be decoded with a highly parallelizable , rapidly converging soft - decision decoding algorithm , usually have a large minimum distance , but require large block lengths of more than @xmath1bits to realize codes with small overheads , leading to decoding latencies that can be detrimental in certain applications . with overheads of more than 15% to 20%
, these codes no longer perform well , at least under hard - decision decoding @xcite .
codes are understood and are suited to realize codes with lengths of a few @xmath2bits and overheads above @xmath3% .
recently , the class of codes@xcite has gained widespread interest due to the fact that these codes are asymptotically capacity - achieving , have appealing encoding and decoding complexity and show outstanding practical decoding performance .
codes are an extension of existing coding schemes by a superimposed convolutional structure .
the technique of spatial coupling can be applied to most existing codes , the most popular are however codes @xcite and @xcite , which have found use in optical communications ( _ staircase codes _ ) and show outstanding performance , operating within @xmath4db of the capacity of the hard - decision awgn channel . in this paper , we discuss the use of codes in optical communications and especially focus on - codes .
we summarize some recent advances and design guidelines for - codes and show by means of an -based decoding platform that large gains at low bit error rates can be realized with relatively small codes when compared with state - of - the - art ldpc codes .
the aim of this paper is to show that - codes are mature channel codes that are viable candidates for future optical communication systems with large .
furthermore , their universality makes them attractive for flexible transceivers with adaptive modulation .
an code is defined by the null space of a _ sparse _ parity - check matrix @xmath5 of size @xmath6 where the code contains all binary code words @xmath7 of length @xmath8 such that @xmath9 , i.e. , @xmath10 .
each row of @xmath5 is considered to be a _ check node _ , while each column of @xmath5 is usually termed _
variable node_. we say that the _ variable degree _ ( or _ variable node degree _ ) of a code is _ regular _ with degree @xmath11 if the number of `` 1 ' 's in each column is constant and amounts to @xmath11 .
we say that the _ check degree _ ( or _ check node degree _ ) of a code is _ regular _ with degree @xmath12 if the number of `` 1 ' 's in each row of @xmath5 is constant and amounts to @xmath12 .
the class of _ irregular _ codes has the property that the number of `` 1 ' 's in each column and/or row is not constant .
the _ degree profile _ of an irregular code indicates the fraction of columns / rows of a certain degree .
more precisely , @xmath13 represents the fraction of columns with @xmath14 `` 1 ' 's ( e.g. , if @xmath15 , half the columns of @xmath5 have three `` 1 ' 's ) .
note that @xmath16 has to hold .
similarly , @xmath17 represents the fraction of rows ( i.e. , checks ) with @xmath14 `` 1 ' 's .
codes form an important class of codes in optical communications @xcite .
codes with soft - decision decoding are currently being deployed in systems operating at 100gbit / s and , e.g. , utilizing 16 iterations @xcite .
modern high - performance systems in optical communications are sometimes constructed using a soft - decision inner code which reduces the to a level of @xmath18 to @xmath19 and a hard - decision algebraic outer cleanup code which pushes the system to levels below @xmath20 @xcite .
the outer cleanup code is used to combat the _ error floor _ that is present in most codes . note that the implementation of a coding system with an outer cleanup code requires a thorough understanding of the code and a properly designed interleaver between the and the outer code .
recently , there has been some interest to avoid the use of an outer cleanup code and to use only soft - decision codes with very low error floors , leading to coding schemes with less rate loss and less latency . with increasing computational resources , it is now also feasible to evaluate very low target of codes and optimize the codes to have very low error floors below the system s target @xcite .
although the internal data flow of an ldpc decoder may be larger by more than an order of magnitude @xcite than that of a btc , several techniques can be used to lower the data - flow , e.g. , the use of layered decoding @xcite and min - sum decoding , requiring only two @xmath21-ary , @xmath22 binary and one @xmath23-ary message per check node
. - codes were introduced more than a decade ago@xcite but their outstanding properties have only been fully realized recently , when lentmaier et al .
noticed@xcite that the estimated decoding performance of a certain class of terminated protograph - based - codes with a simple message passing decoder is close to the performance of the underlying code ensemble under decoding as @xmath24 grows , which was subsequently proven rigorously in @xcite , if certain particular conditions on the code structure are fulfilled .
a left - terminated - code is basically an code with a structured ,
infinitely extended parity - check matrix @xmath25 \bm{h}_1(1 ) & \bm{h}_0(1 ) & & & \\[-0.3em ] \vdots & \bm{h}_{1}(2 ) & & & \\[-0.3em ] \bm{h}_\mu(\mu ) & \vdots & \ddots & & \\[-0.7em ] & \bm{h}_{\mu}(\mu+1 ) & \ddots & & \\[-0.5em ] & & & \bm{h}_0(t ) & \\[-0.5em ] & & \ddots & \bm{h}_1(t+1 ) & \ddots \\[-0.3em ] & & & \vdots & \ddots \\[-0.3em ] & & & \bm{h}_\mu(t+\mu ) & \\[-0.5em ] & & & & \ddots \end{array } \right)}\label{eq : timeinvarmatrix}\end{aligned}\ ] ] with @xmath26 being sparse binary parity - check matrices with @xmath27 and @xmath28 denoting the _ syndrome former memory _ of the code .
every code word @xmath7 of the code has to fulfill @xmath29 .
one advantage of - codes is that the infinitely long code words can conveniently be decoded with acceptable latency using a simple windowed decoder @xcite . in practice , in order to construct codes of finite length , e.g. , to adhere to certain framing structures in the communication system at hand , the infinitely extended matrix @xmath30 is _ terminated _ resulting in finite length code .
one example of termination is _ zero - termination _ , where the matrix @xmath30 is cut off after @xmath31 parts , resulting in a code of length @xmath32 and a parity - check matrix @xmath33}$ ] of size @xmath34 } = ( l+\mu)m\times ln$ ] .
note that this termination leads to a _ rate loss _
, which can however be kept small if @xmath35 is chosen large enough . for a discussion of termination schemes ,
we refer the interested reader to @xcite .
codes are now emerging in various applications .
two examples of product codes are the staircase code@xcite and the braided bch codes @xcite , for hard - decision decoding in optical communications . sc - ldpc codes may also be viable for pragmatic coded modulation schemes @xcite . in order to simplify the design of hardware , we first drop the time dependency and only consider the time - independent ( left - terminated ) parity - check matrix with @xmath36 , @xmath37 , which is attractive for implementation as the sub - matrices @xmath38 can be easily reused in the encoder and decoder hardware . in this time - invariant construction with @xmath39
, we can give the following upper bound on the minimum distance of the code ( * ? ? ?
* eq . ( 7 ) )
@xmath40 to construct codes with large enough minimum distances , we maximize the size of the sub - matrices @xmath38 , i.e. , @xmath41 , which has a quadratic influence on .
in order to keep the complexity of the so - constructed code small , we restrict ourselves to small values of the syndrome former memory @xmath28 with either @xmath42 or @xmath43 .
we call such codes _ weakly coupled _ codes @xcite .
in the past , irregular block codes have been used to design codes that perform very well for low snrs , but these schemes do sometimes suffer from relatively high error floors requiring the use of an outer code that leads to inherent rate losses . in the case of - codes , we can use the irregularity to control the propagation speed of the decoding wave of a windowed decoder , i.e. , we can minimize the number of iterations @xmath44 that are necessary until a windowed decoder can advance by one step @xcite . to simplify the code construction and to illustrate the concept
, we only use the most simple form of irregularity and construct slightly irregular - codes with and additionally with either degree-4 or degree-6 variable nodes .
we avoid degree-2 variables nodes due to their potentially detrimental effect on the error floor .
also , in contrast to block ldpc codes , degree-2 nodes are _ not _ of the same importance for - codes .
we vary the fraction of degree-4 or degree-6 nodes between 0 and 1 and select the check nodes such that a rate @xmath45 ( 25% overhead ) code is constructed .
we perform full density evolution using the irregular version of kudekar s @xmath46 ensemble @xcite for random spatial coupling with @xmath47 using an awgn channel and measure the required @xmath48 values to advance the decoding wave by @xmath44 steps .
the density evolution results are shown in fig .
[ fig : speed ] for varying @xmath49 and @xmath50 .
we can see that using additionally degree-4 ( besides degree-3 ) variables does not lead to noteworthy gains , which is why we focus on additional degree-6 nodes in this paper .
the convergence speed improves by selecting a proper value of @xmath50 leading to a smaller required @xmath48 .
the selection depends however on @xmath44 . as we intend to construct low complexity decoders with @xmath51
, we can see that in this case , the optimum is achieved with @xmath52 ( 20% of degree-6 variable nodes , 80% of degree-3 variable nodes ) .
we can see that by proper selection of @xmath50 , we can obtain codes that have an improved decoding convergence , however , we also see that depending on the selection , a worse convergence behavior than for the regular case can result .
we also observe that if we want a code that operates extremely close to capacity , the optimum value of @xmath50 is larger ( around @xmath4 ) than for the more practical case , where the optimum lies at @xmath52 .
note that although we use kudekar s ensemble for density evolution , the codes we construct in the next section are generated from protographs , similar to those in @xcite , as these exhibit better finite length performance . in order to verify the performance of the rapidly converging weakly coupled - codes ,
we use a platform , whose high - level diagram is illustrated in fig .
[ fig : fpga_schematic ] @xcite .
this platform is similar to other platforms reported in the literature @xcite and consists of three parts : a gaussian noise generator , an decoder and an error detecting circuit .
the gaussian noise generator generates gaussian distributed , stemming from bpsk transmission over an awgn channel , using uniform random number generators and the box - muller transform .
these are then fed to the decoder after quantization to 15 levels .
the decoder is based on the layered decoding algorithm @xcite and uses a scaled - minsum check node computation rule with constant scaling factor .
the windowed decoder that is implemented can be sub - divided into three steps . in the first step
, a new sub - block of @xmath24 quantized is received from the random number generator and put into the vacant position of the decoder s memory .
decoding takes place by considering @xmath53 copies of @xmath54 .
the windowed decoded considers an equivalent matrix of size @xmath55 which it processes before shifting in @xmath24 new values . in order to maximize the hardware utilization , within a window , we use two parallel decoders that operate on non - overlapping portions of that matrix . in a first step ,
the first decoding engine operates on the first @xmath41 check nodes of the matrix under consideration while the second engine operates in parallel on the @xmath41 check nodes starting at position @xmath56 . in general , the first engine processes the check nodes at position @xmath57 $ ] while the second engine processes the check node @xmath58 .
note that only a single iteration is carried out to guarantee the required throughput , corresponding effectively to @xmath59 iterations per bit ( due to the use of two engines ) .
@xmath60 ' '' '' the output of the decoder is connected to the evaluation unit , which counts the bit errors and reports the error positions .
we use virtex-7 allowing for a throughput of several gbit / s to evaluate the ber performance of several coding schemes of rate @xmath61 , i.e. , of 25% coding overhead .
we select this particular rate due to its importance in today s systems .
current and future @xmath62gbit / s ( with qpsk ) or @xmath63gbit / s ( with 16-qam ) systems are often operated in @xmath64ghz channels with an exploitable bandwidth of roughly @xmath65ghz due to with non - flat frequency characteristic . with almost rectangular pulse shapes ( root - raised cosine with small roll - off @xmath66 ) and today s generation of , symbol rates of @xmath67gbaud can be realized . with dual - polarization qpsk transmission ,
gross bit rates of 128gbit / s can be realized .
assuming signaling and protocol overheads of 3gbit / s , this leads to a code that adds 25gbit / s parity overhead ( i.e. , of rate @xmath68 ) .
we compare three codes : * as reference , we consider a regular block qc - ldpc code ( marker @xmath69 ) with variable node degree @xmath70 and check node degree @xmath71 .
the code is a quasi - cyclic code of girth 10 and block length @xmath72 , constructed using cyclically shifted identity matrices of size @xmath73 and decoded with @xmath74 row - layered iterations .
* - code a ( @xmath75 ) is the rapidly converging irregular code with syndrome former memory @xmath43 , @xmath52 and @xmath76 and check node degree @xmath77 .
the sub - block size is @xmath78 ( @xmath79 ) .
* - code b ( @xmath80 ) is a regular @xmath81 code with @xmath82 and syndrome former memory @xmath42 .
the size of the sub - matrices is identical to those of - code a , however , we select @xmath42 . both codes are constructed from cyclic permutation matrices of size @xmath83 and are terminated after @xmath84 subblocks .
the simulation results are shown in fig . [
fig : fpga_simres ] .
the block code , which has a matrix that has been optimized for low error floors , is outperformed by both - codes .
- code a offers a coding gain of around 0.3db at a ber of @xmath20 compared to the conventional block ldpc code , but an error floor starts to manifest .
this error floor is not due to any trapping sets , but due to a few uncorrected bits after windowed decoding , which can be recovered with a few - error correcting outer code .
code b has a ber curve that starts to decay at worse channels , but the ber curves cross at @xmath85 .
for the next simulated point , we did not observe any bit errors , and hence we conjecture a lower error error floor than for code a. note that no special measures have been taken to combat an error floor : only a plain scaled min - sum decoder has been used . with the block code
, post - processing @xcite may be necessary to combat the error floor .
another advantage of - codes is that they are future - proof : while the block code does not benefit from further decoding iterations , as its performance is already close to its decoding threshold , the scaling behavior of the sc - ldpc code allows to carry out further iterations and achieve still larger coding gains , as the gap to the decoding threshold is still non - negligible .
this makes these codes attractive for standardization .
as future optical networks tend to become increasingly flexible and _ elastic _ , transceivers that integrate a certain amount of flexibility with respect to coding and modulation formats are required . especially the modulation format is expected to change when transceivers are designed for long - haul or short - haul applications , where the latter require high spectral efficiencies ( e.g. , data center interconnects ) . in this section ,
we show that codes are perfectly suited to be combined with varying modulation formats due to their _ universality _ properties @xcite .
we combine - codes with a modulator and use _ density evolution _ to show how the detector front - end influences the performance of the codes . in conventional ( block ) code design ,
usually the code needs to be `` matched '' to the transfer curve of the detection front - end @xcite . if the code is not well matched to the front - end , a performance loss occurs .
if the detector front - end has highly varying characteristics , due to , e.g. , varying modulation formats or channels , several codes would need to be implemented and selected depending on the conditions , which is not feasible in optical networks , where feedback is usually difficult to realize and where different codes can not be implemented due to hardware constraints .
in contrast to many block ldpc codes , spatially coupled ldpc codes can converge below the pinch - off in the exit chart due to the effect of threshold saturation @xcite .
hence , even if the code is not well matched to the demodulator / detector from a traditional point of view , we can hope to successfully decode .
we can hence use a single code which is _ universally _ good in all scenarios and the code design can stay _ agnostic _ to the channel / detector behavior . in order to illustrate the concept , we model the detector by a linear exit characteristic @xmath86 } = f_d(i_a^{[d ] } ) = a\cdot i_a^{[d ] } + i_c - \frac{a}{2}\ ] ] where @xmath87 controls the slope of the characteristic and @xmath88 describes the mutual information of the communication channel .
the slope @xmath87 models the effect of e.g. , different modulation formats , different bit labelings in higher order modulation and different detectors .
we assume that the output of the detector can be modeled using a
. there therefore also use message passing .
we compare two different code approaches ; first we use the spatially coupled @xmath89 ensemble presented in @xcite with the density evolution equation for iterative detection given by where @xmath90 denotes the node - perspective degree distribution polynomial , @xmath91 the edge - perspective degree distribution , and @xmath92 the edge message erasure probability of spatial position @xmath14 at iteration @xmath93 . additionally , we generate protograph based codes end employ density evolution @xcite including iterative detection .
we consider two code families of rate @xmath94 : the first family is the rapidly converging code from sec .
[ sec : weakly ] with @xmath95 and @xmath77 where we use @xmath47 and @xmath96 in kudekar s ensemble and @xmath97 with @xmath98 in the protograph ensemble .
the second code is a regular code where we use kudekar s @xmath99 ensemble and a protograph ensemble with @xmath100 and @xmath101 with @xmath43 @xcite .
figure [ fig : det_thresholds ] shows the de results where we use solid lines ( ' '' '' ) to show the decoding thresholds for kudekar s ensemble and dashed lines ( ) for the protograph - based ensemble .
all sc codes have decoding thresholds close to the theoretical limit of @xmath102 and the decoding threshold is almost independent of the detector characteristic s slope @xmath87 .
a regular block ldpc code has a highly varying threshold for different slopes @xmath87 .
the flat threshold behavior for codes indicates a _ universal _ , _ channel - agnostic _ behavior .
even an optimized irregular ldpc code will only be good for a single slope parameter @xcite . in order to improve the decoding threshold
, we may deliberately select a precoder that has an exit characteristic with slope @xmath103 , however , as the inset of fig .
[ fig : det_thresholds ] shows , the slope affects the decoding speed ( measured at @xmath104 ) , i.e. , the number of iterations required to advance the decoding wave by one step , so that the complexity will grow alongside . for the case of the rapidly converging code
, @xmath103 further increases the decoding speed .
we have presented an example of such a system with differential detection ( @xmath105 ) that is adapted to a channel with varying phase noise in @xcite .
therein , a single spatially coupled code was able to outperform two different ldpc codes optimized for different channel characteristics .
in this paper , we have highlighted -ldpc codes as potential candidates for future lightwave transmission systems .
we have optimized - codes for convergence speed and shown by means of an fpga - based simulation that very low error rates can be obtained . finally , we have shown that - can be good candidates if they employed in a system with iterative decoding and detection : a single code can be used in various channel conditions .
a. r. iyengar , m. papaleo , p. h. siegel , j. k. wolf , a. vanelli - coralli , and g. e. corazza , `` windowed decoding of protograph - based ldpc convolutional codes over erasure channels , '' _ ieee trans .
inf . theory _
58 , no . 4 ,
pp . 23032320 , april 2012 .
c. hger , a. graell i amat , f. brnnstrm , a. alvarado , and e. agrell , `` comparison of terminated and tailbiting spatially coupled ldpc codes with optimized bit mapping for pm-64-qam , '' in _ proc .
ecoc _ , cannes , france , 2014 , paper th.1.3.1 .
jian , h. d. pfister , k. r. narayanan , r. rao , and r. mazareh , `` iterative hard - decision decoding of braided bch codes for high - speed optical communication , '' in _ proc .
globecom _ ,
atlanta , usa , 2013 .
v. aref , l. schmalen , and s. ten brink , `` on the convergence speed of spatially coupled ldpc ensembles , '' in _ proc .
allerton conference on communications , control , and computing _ , oct .
2013 , arxiv:1307.3780 .
l. schmalen , s. ten brink , and a. leven , `` spatially - coupled ldpc protograph codes for universal phase slip - tolerant differential decoding , '' in _ proc .
ofc _ , los angeles , ca , usa , mar .
2015 , paper th3e.6 . | in this paper , we highlight the class of spatially coupled codes and discuss their applicability to long - haul and submarine optical communication systems .
we first demonstrate how to optimize irregular spatially coupled ldpc codes for their use in optical communications with limited decoding hardware complexity and then present simulation results with an fpga - based decoder where we show that very low error rates can be achieved and that conventional block - based ldpc codes can be outperformed . in the second part of the paper , we focus on the combination of spatially coupled ldpc codes with different demodulators and detectors , important for future systems with adaptive modulation and for varying channel characteristics .
we demonstrate that sc codes can be employed as universal , channel - agnostic coding schemes .
@ps@headingsps .
@headings @ps@ieeetitlepagestyleps .
@ieeetitlepagestyle # 1 ps .
@headings @ps@headings oddhead # 1 evenhead # 1 ps .
@ieeetitlepagestyle @ps@ieeetitlepagestyle oddhead # 1 evenhead # 1 ps .
@headings schmalen _ et .
al _ : next generation error correcting codes for lightwave systems error correction codes , low - density parity - check codes , spatial coupling , optical communications |
js gratefully acknowledges support from the korean science and engineering foundation ( kosef ) through the center for theoretical physics ( ctp ) . the work was supported in part by the kosef - dfg large collaboration project , project no .
96 - 0702 - 01 - 01 - 2 . in the appendix the explicit form of the cp - even term @xmath120 and the cp - odd terms @xmath112 ( @xmath109 to @xmath110 )
is presented : @xmath171 \ ; , \\
\delta_1&=&-8f(q^2)c_km_{\tau}p_k \left(\frac{m_{\tau}}{\sqrt{q^2}}\right ) \cos\vartheta{\rm im}(\xi){\rm im}(f_kf^*_s)\;,\\ \delta_2&=&-8f(q^2)c_km_{\tau}p_k{\rm im}(\xi ) \left[\left\{\cos\theta+ \left(\frac{m_{\tau}}{\sqrt{q^2}}-1\right)\cos\theta\cos\vartheta\right\ } { \rm im}(f_kf^*_s)\right.\nonumber\\ & & \mbox { } \hskip 2 cm + \sin\theta\sin\theta\sin\phi{\rm re}(f_kf^*_s)\bigg]\;,\\ \delta_3&=&-8f(q^2)c_km_{\tau}p_k{\rm im}(\xi ) \left[\left\{\sin\theta\cos\phi+ \left(\frac{m_{\tau}}{\sqrt{q^2}}-1\right)\sin\theta\cos\vartheta\right\ } { \rm im}(f_kf^*_s)\right.\nonumber\\ & & \mbox { } \hskip 2 cm -\cos\theta\sin\theta\sin\phi{\rm re}(f_kf^*_s)\bigg]\;,\\ \delta_4&=&-8f(q^2)c_km_{\tau}p_k{\rm im}(\xi)\bigg[\sin\theta\sin\phi { \rm im}(f_kf^*_s)\nonumber\\ & & \mbox { } \hskip 2 cm + ( \cos\theta\sin\theta\cos\phi -\sin\theta\cos\theta){\rm re}(f_kf^*_s)\bigg]\;,\end{aligned}\ ] ] where the overall function @xmath172 is given by @xmath173 s. narison , n. paver , and d. treleani , nuovo cimento 74 ( 1983 ) 347 ; s. narison , n. paver , e. de rafael and d. treleani , nucl .
b 212 ( 1983 ) 365 ; s. narison , qcd spectral sum rules ( world scientific , singapore , 1989 ) .
barbar collaboration , d. boutigny _ et al .
_ , technical design report , 1995 ; belle collaboration , m.t .
cheng _ et al .
_ , technical design report , kek report no .
95 - 1 , 1995 ; kek report no .
97 - 1 , 1997 ; for relevant physics at b factories , see , e.g. , a.i .
sanda and s. suzuki , in proceedings of international workshop on physics beyond the standard model at the b factory , nagoya , japan , oct .
26 - 28 , 1994 , singapore world scientific ( 1995 ) 476 . | cp violation from physics beyond the standard model ( sm ) is investigated in the cabbibo - suppressed decay @xmath0 with polarized @xmath1 leptons , to which both the @xmath2 resonance @xmath3 and the @xmath4 resonance @xmath5 contribute .
in addition to the cp - odd rate asymmetry , @xmath1 polarization enables us to construct three additional cp - odd polarization asymmetries that can be enhanced due to the interference between the @xmath3 and @xmath5 , and whose magnitudes depend crucially on the @xmath5 decay constant , @xmath6 .
taking a qcd sum rule estimate of @xmath7 mev and the present experimental constraints on the cp - odd parameters into account , we estimate quantitatively the maximally - allowed values for the cp - odd rate and polarization asymmetries in the multi - higgs - doublet ( mhd ) model and the scalar - leptoquark ( slq ) models consistent with the sm gauge symmetry where neutrinos are massless and left - handed as in the sm .
we find that the cp - odd rate and polarization asymmetries are of a similar size for highly - polarized @xmath1 leptons and , for their maximally - allowed values , cp violation in the mhd model and two slq models may be detected with about @xmath8 and @xmath9 @xmath1 s at the @xmath10 level .
0.26 cm 0.26 cm 22.5 cm + + * cp violation in the cabbibo - suppressed decay @xmath11 + with polarized @xmath1 leptons * 1.cm s.y .
choi@xmath12 , jake lee@xmath12 and j. song@xmath13 0.8 cm @xmath12 department of physics , yonsei university 120 - 749 , seoul , korea .
+ 0.2 cm @xmath13 center for theoretical physics , seoul national university , seoul 151 - 742 , korea . the decay of the @xmath1 , the most massive of the known leptons , can serve not only as a useful tool in the investigation of some aspects of the sm but also as a powerful experimental probe of new physics phenomena @xcite .
one phenomenon where new physics can play a crucial role is cp violation . in light of this aspect
the @xmath1 decays into hadrons have been recently studied as probes of cp violation in the scalar sector of physics beyond the sm @xcite . in the present paper
we extend the previous work @xcite by choi , hagiwara and tanabashi to probe cp violation in the cabbibo - suppressed @xmath1 decay @xmath11 with polarized @xmath1 s .
the decay mode is dominated by the contributions of the two lowest vector and scalar resonances , @xmath3 and @xmath5 , with different spins and relatively large width - to - mass ratios @xcite : @xmath14 and the mode is expected to have larger scalar contributions than the @xmath15 or @xmath16 modes due to the @xmath17 quark mass much larger than the @xmath18 quark mass . in the light of these aspects ,
the cabbibo - suppressed mode is worthwhile to be investigated in detail . including possible contributions from new physics with massless left - handed neutrinos
, we can write the matrix element for the decay @xmath19 in the general form @xmath20 , \label{decaym}\end{aligned}\ ] ] where @xmath21 , @xmath22 is the fermi constant , and @xmath23 and @xmath24 are the four momenta of the @xmath1 lepton and the tau neutrino , respectively .
the parameters @xmath25 and @xmath26 , which parametrize contributions from physics beyond the sm , are in general complex . in eq .
( [ decaym ] ) , the helicity of the @xmath27 is denoted by @xmath28 ( @xmath29 ) with its spin quantization direction along its neutrino momentum direction .
the tau neutrino is assumed to be massless and left - handed as in the sm so that the helicity value is @xmath30 as indicated by the negative sign in its spinor @xmath31 .
the vector and scalar hadronic matrix elements @xmath32 with @xmath33 for the cabbibo angle @xmath34 are related through the dirac equations to the @xmath35 and @xmath36 quarks at the quark level and their explicit form can be parametrized in terms of two form factors @xmath37 and @xmath38 : @xmath39 , \nonumber\\ & & j_s=\sqrt{2}\sin\theta_c \left(\frac{m^2_{k^*_0}}{m_s - m_u}\right)c_kf_s(q^2),\end{aligned}\ ] ] where @xmath40 and @xmath41 are the four - momenta of @xmath42 and @xmath43 , respectively , @xmath44 and @xmath45 the @xmath17 and @xmath36 current quark masses , and @xmath46 is the four - momentum of the @xmath47 system ; @xmath48 .
the coupling strength @xmath49 denoting the scalar contributions is determined by the @xmath5 decay constant @xmath6 and the coupling strength @xmath50 of the @xmath5 to @xmath47 under the assumption of br(@xmath51 : @xmath52 the value of @xmath50 is @xmath53 from the measured @xmath54 decay width @xmath55 .
even though the @xmath5 decay constant @xmath6 is not experimentally measured , it has been estimated by several model - dependent methods .
the qcd sum - rule estimate in the @xmath5 narrow - width approximation is @xmath56 mev @xcite and an effective lagrangian estimate of @xmath6 including the width effects leads to a larger value of about 45 mev @xcite , which approaches the pole dominance result of 50 mev @xcite . in light of these present crude estimates of the @xmath5 decay
constant , we adopt for the @xmath5 decay constant the effective lagrangian estimate @xcite @xmath57 anticipating that the decay constant @xmath6 is measured more precisely in future experiments .
0.5 cm to in spite of the several resonance contributions to @xmath58 and @xmath59 , the form factors are approximated to be the propagators of the lowest - level resonance states @xmath3 and @xmath5 : @xmath60 respectively , where @xmath61 and @xmath62 are parametrized in the breit - wigner form with the momentum - dependent widths @xcite : @xmath63 for @xmath64 or @xmath5 .
we adopt for the momentum - dependent widths the parametrizations of the @xmath1-decay program package tauola @xcite @xmath65 ^ 3 & \ \ { \rm for}\ \ q^2>(m_k+m_\pi)^2 \;,\\ 0 & \ \ { \rm for}\ \ q^2\leq ( m_k+m_\pi)^2\ ; , \end{array}\right .
\label{pi1_f}\\ & & { } \nonumber\\ & & d_{k^*_0}(q^2)=\left\{\begin{array}{cl } \frac{m_{k^*_0}}{\sqrt{q^2 } } \left[\frac{p_k(q^2)}{p_k(m^2_{k^*_0})}\right ] & \ \ { \rm for}\ \ q^2>(m_k+m_\pi)^2 \ ; , \\ 0 & \ \ { \rm for}\ \ q^2\leq ( m_k+m_\pi)^2\;. \end{array}\right . \label{pi2_f}\end{aligned}\ ] ] let us now calculate the helicity amplitudes of the cabbibo - suppressed @xmath1 decay .
in general , the @xmath1 three - body decay into @xmath66 is described by five independent kinematic variables . because the hadronic @xmath67 system is solely determined by the lepton momentum transfer , it is convenient to consider two reference frames , namely the @xmath1 rest frame and the @xmath67 rest frame as shown in fig .
[ fig : decay ] .
we define the momentum direction of the virtual @xmath3 or @xmath5 in the @xmath1 rest frame by the polar angle @xmath68 and the azimuthal angle @xmath69 with respect to the @xmath1 momentum direction taken to be along the positive @xmath70-axis and the @xmath43 momentum direction in the @xmath67 rest frame by the angles @xmath71 and @xmath72 .
the rotational invariance of the total system with respect to the @xmath1 momentum direction allows us to take @xmath69 to be zero in the calculation of the helicity amplitudes , while the azimuthal angle @xmath69 is employed to describe @xmath1 polarization , especially the transverse polarization of the @xmath1 . setting @xmath69 to be zero and using the 2-component spinor technique @xcite ,
one can obtain the helicity amplitudes @xmath73 of the cabbibo - suppressed decay @xmath19 in the reference frames defined in fig .
[ fig : decay ] as @xmath74 , \nonumber\\ m_-&=&2g_f\sin\theta_c(1+\chi)\sqrt{m_{\tau}^2 -q^2 } \left [ m_{\tau}\sin\frac{\theta}{2 } \left(\frac{m^2_{k^*_0}}{q^2}+\xi\right)c_k f_s(q^2 ) \right .
\nonumber\\ & & + 2p_k \left .
\left\{\cos\theta\sin\frac{\theta}{2 } -\sin\theta\cos\frac{\theta}{2 } e^{-i\phi }
-\left(\frac{m_{\tau}}{\sqrt{q^2}}-1\right ) \cos\vartheta\sin\frac{\theta}{2}\right\}f_k(q^2 ) \right ] , \label{mtamp } \end{aligned}\ ] ] where @xmath75 is the angle between the @xmath43 momentum in the ( @xmath47 ) rest frame and the ( @xmath47 ) momentum in the @xmath1-rest frame , i.e. @xmath76 , and @xmath77 is the size of the @xmath43 momentum in the virtual @xmath3 and @xmath5 rest frame ; @xmath78 with @xmath79 .
the dimensionless parameter @xmath80 determines the relative size of the scalar contributions to the vector ones : @xmath81 clearly , the initial @xmath27 system is not cp self - conjugate and so any genuine cp - odd observable can be constructed only by considering both the cabbibo - suppressed @xmath27 decay and its charge - conjugated @xmath82 decay , and by identifying the cp relations of their kinematic distributions . before constructing all the possible cp - odd asymmetries explicitly , we calculate the helicity amplitudes for the charge - conjugated process @xmath83 .
the helicity amplitudes @xmath84 of the @xmath82 decay in the same reference frame as in the @xmath27 decay are given by @xmath85 , \nonumber\\ \overline{m}_-&=&- 2 g_f\sin\theta_c(1+\chi^*)\sqrt{m_{\tau}^2 -q^2 } \left [ m_{\tau}\cos\frac{\bar{\theta}}{2 } \left(\frac{m^2_{k^*_0}}{q^2}+\xi^*\right)c_k f_s(q^2 ) \right .
\nonumber\\ & & -2p_k\left.\left\ { \cos\bar{\theta}\cos\frac{\bar{\theta}}{2 } + \sin\bar{\theta}\sin\frac{\bar{\theta}}{2 } e^{-i\bar{\phi } } + \left(\frac{m_{\tau}}{\sqrt{q^2}}-1\right ) \cos\bar{\vartheta}\cos\frac{\bar{\theta}}{2}\right\}f_k(q^2 ) \right ] .
\label{ptamp } \end{aligned}\ ] ] it is easily shown that , if the parameters @xmath26 and @xmath25 are real , the helicity amplitudes ( [ mtamp ] ) for the @xmath27 decay and ( [ ptamp ] ) for the @xmath82 decay satisfy the cp relation : @xmath86 with the results in eqs . ( [ mtamp ] ) and ( [ ptamp ] ) for the @xmath87 decay helicity amplitudes
, one can describe the decay of an arbitrary polarized @xmath1 lepton by superposing the two helicity states .
generally , a pure @xmath1 polarization state , which is polarized in the direction ( @xmath88 ) , is given in terms of the helicity states by @xmath89 where @xmath90 and @xmath91 are the polar and azimuthal angles in the @xmath1 rest frame , respectively , with respect to the @xmath1 momentum direction in the laboratory frame .
in this case the decay amplitude to the final state @xmath92 of the virtual @xmath93 system is expressed in terms of the helicity amplitudes as @xmath94 where the rotational invariance of the total system with respect to the @xmath1 momentum direction has been used to relate the first and second expressions .
it is noteworthy that only the difference @xmath95 between the azimuthal angles @xmath69 and @xmath91 appears . taking into account the polarization degree @xmath96 for a partially - polarized @xmath27 beam , one can decompose the differential decay rate into four independent terms : @xmath97 where the helicity - dependent terms are defined as @xmath98 for the sake of notational convenience , we introduce @xmath99 for the phase space element @xmath100 and denote the four independent terms in eq .
( [ decayrate ] ) as @xmath101 note that ( a ) the @xmath102 term is the unpolarized differential decay rate and the other three terms @xmath103 ( @xmath104 ) are polarization - dependent ; the @xmath105 term is due to longitudinal polarization , the @xmath106 term due to transverse polarization and the @xmath107 due to normal polarization , and ( b ) the transverse and normal components after integrating the differential decay rate over the azimuthal angle @xmath68 or @xmath91 vanish , which is consistent with the so - called null transverse - polarization theorem " @xcite . clearly , to utilize the polarization - dependent terms effectively , the initial @xmath1 beam should be highly polarized , that is to say , @xmath96 should be relatively large . deferring the potential impact of the value of @xmath96 to our later discussion ,
we assume @xmath96=1 for the time being . in this case , each @xmath108 ( @xmath109 to @xmath110 ) term can be decomposed into a cp - even part @xmath111 and a cp - odd part @xmath112 : @xmath113 the four cp - even parts @xmath111 and four cp - odd parts @xmath112 can be easily identified by use of the cp relation ( [ cppro ] ) between the @xmath27 and @xmath82 decay helicity amplitudes and they are expressed in terms of the @xmath114 helicity - dependent terms @xmath108 and @xmath115 as @xmath116 where we have used the same kinematic variables @xmath117 for the @xmath115 except for the replacement of @xmath118 by @xmath119 .
we have numerically estimated the four cp - even terms in the allowed phase - space points and have found that the other three @xmath111 ( @xmath104 ) terms are negligible compared to the @xmath120 term
. therefore we neglect those terms in the following .
the cp - even @xmath120 term and the cp - odd @xmath112 ( @xmath109 to @xmath110 ) can be obtained from the @xmath114 decay helicity amplitudes and their explicit form is listed in appendix a. all the cp - odd terms are proportional to the imaginary part of the parameter @xmath80 in eq .
( [ eq : xi ] ) .
an appropriate real weight function @xmath121 is usually employed to separate the @xmath112 contribution and to enhance its analysis power for the cp - odd parameter @xmath122 through the cp - odd quantity : @xmath123 d\phi_3.\end{aligned}\ ] ] of which the analysis power is determined by the parameter @xmath124 for the analysis power @xmath125 , the number @xmath126 of the @xmath1 leptons needed to observe cp violation at the 1-@xmath28 level is @xmath127 where @xmath128 is the branching fraction of the relevant @xmath1 decay mode .
certainly , it is desirable to find the optimal weight function with the largest analysis power .
it is known @xcite that , when the cp - odd contribution to the total rate is relatively small , the optimal weight function is approximately given by @xmath129 we adopt these optimal weight functions in the following numerical analyses with several concrete models beyond the sm introduced in the following .
although there is no cp violation in the @xmath1 decays within the sm , it is possible to conceive several new sources of cp violation in the @xmath1 decays . among them
we will consider models with new scalar - fermion interactions , which still preserve the sm gauge symmetries and have only the massless and left - handed neutrinos as in the sm . in this case
, only four types of scalar - fermion interactions can contribute to the cabbibo - suppressed decay @xmath130 @xcite ; the mhd model @xcite and three slq models @xcite . in the mhd model
cp violation can arise in the charged higgs sector with more than two higgs doublets @xcite and when not all the charged scalars are degenerate . as in most previous phenomenological analyses
, we also will assume in this mhd model that all but the lightest of the charged scalars effectively decouple from fermions .
the effective lagrangian for the decay @xmath11 in the assumption is then given at energies considerably low compared to @xmath131 by @xmath132+{\rm h.c .
} , \nonumber\\\end{aligned}\ ] ] where @xmath133 , @xmath134 and @xmath135 are complex coupling constants which can be expressed in terms of the charged higgs mixing matrix elements . from the effective lagrangian , one obtain for the mhd cp - violation parameter @xmath136 @xmath137 . \label{mhdp}\end{aligned}\
] ] on the other hand , the effective lagrangians for the three slq models @xcite contributing to the decay @xmath11 are written in the form after a few fierz rearrangements : @xmath138 + { \rm h.c.},\nonumber\\ & & { \cal l}^{ii}_{\rm slq}=-\frac{y_{23}y^{\prime * } _ { 13}}{2m^2_{\phi_2 } } \left[(\bar{s}_lu_r)(\bar{\tau}^c_r\nu^c_{\tau_l } ) + \frac{1}{4}(\bar{s}_l\sigma^{\mu\nu}u_r ) ( \bar{\tau}^c_r\sigma_{\mu\nu}\nu^c_{\tau_l})\right]\nonumber\\ & & \hskip 1.5 cm + \frac{y_{23}y^*_{13}}{2m^2_{\phi_2 } } ( \bar{s}_l\gamma_\mu u_l)(\bar{\tau}^c_l\gamma^\mu\nu^c_{\tau_l } ) + { \rm h.c.},\nonumber\\ & & { \cal l}^{iii}_{\rm slq}=-\frac{z_{23}z^*_{13}}{2m^2_{\phi_3 } } ( \bar{s}_l\gamma_\mu u_l)(\bar{\tau}^c_l\gamma^\mu\nu^c_{\tau_l } ) + { \rm h.c.}\end{aligned}\ ] ] here the coupling constants @xmath139 , @xmath140 and @xmath141 ( @xmath142 ) are in general complex so that cp is violated in the scalar - fermion yukawa interaction terms .
the superscript @xmath143 in the lagrangians @xmath144 and @xmath145 denotes charge conjugation , i.e. @xmath146 in the chiral representation .
although the tensor parts as well as the scalar parts appear in model i and model ii , we do not have the tensor contributions to @xmath130 because we concentrate on the vector and scalar resonance contributions in the present work . in the approximation that all the cp - even contributions from new interactions are neglected ,
the size of the slq cp - violation effects is dictated by the cp - odd parameters @xmath147}{4\sqrt{2}g_f\sin\theta_cm^2_{\phi_1}}\,,\nonumber\\ & & { \rm im}(\xi^{ii}_{\rm slq})=-\frac{m^2_{k^*_0}}{(m_s - m_u)m_\tau } \frac{{\rm im}[y_{23}y^{\prime * } _ { 13}]}{4\sqrt{2 } g_f\sin\theta_cm^2_{\phi_2}}\,,\nonumber\\ & & { \rm im}(\xi^{iii}_{\rm slq})=0\ , .
\label{lqp}\end{aligned}\ ] ] this approximation is justified because the contributions from new physics are expected to be very small compared to those from the sm .
the experimental constraints on the cp - violation parameters in ( [ mhdp ] ) and ( [ lqp ] ) depend on the values for @xmath36 and @xmath17 current quark masses , which are not well determined . here
, we use for the light @xmath36 and @xmath17 quark masses @xmath148 and @xmath149 , which satisfy the mass relation @xmath150 @xcite . inserting the quark mass values into ( [ mhdp ] ) and ( [ lqp ] ) yields @xmath151,\\ { \rm im}(\xi^i_{\rm slq})&\simeq & -39.3 \left(\frac{m_w}{m_{\phi_1}}\right)^2 { \rm im}[x_{23}x^{\prime * } _ { 13}]\;,\nonumber\\ { \rm im}(\xi^{ii}_{\rm slq})&\simeq & -39.3 \left(\frac{m_w}{m_{\phi_2}}\right)^2 { \rm im}[y_{23}y^{\prime * } _ { 13}]\,,\end{aligned}\ ] ] where the @xmath152-boson mass @xmath153 is retained to show the @xmath153 dependence of the parameters explicitly , but 80 gev @xcite will be used for the @xmath152-boson mass in the actual numerical analysis .
the couplings @xmath133 , @xmath134 and @xmath135 in the mhd model can be constrained through the processes such as @xmath154-meson semileptonic decays . because these experimental constraints have been extensively reviewed in ref .
@xcite , we simply follow the analysis from which the combined constraint on @xmath136 is obtained to be @xmath155 when @xmath131 is set to be 45 gev .
although there are at present no direct constraints on the slq cp - odd parameters in ( [ lqp ] ) , a rough constraint to the parameters can be provided on the assumption @xcite that @xmath156 and @xmath157 , that is to say , the leptoquark couplings to quarks and leptons belonging to the same generation are of a similar size ; then the experimental upper bound for the @xmath158 mixing yields @xmath159 which are stronger than the constraint ( [ hdc ] ) on the mhd cp - odd parameter im(@xmath160 ) .
based on the constraints ( [ hdc ] ) and ( [ modeli ] ) to the cp - odd parameters , we quantitatively estimate the number of the cabbibo - suppressed @xmath1 decays to detect cp violation for the maximally - allowed values of the cp - odd parameters : @xmath161 as shown in eq .
( [ eq : number ] ) the branching fraction for the cabbibo - suppressed @xmath1 decay needs to be known beforehand , and experimentally , it is crucial to reconstruct the @xmath1 momentum direction .
so we consider the most easily - identifiable sequential decay channel @xmath162 where an efficient identification of the three charged - pion vertex and the @xmath163 production vertex can be used to determine the momentum of the mother @xmath1 lepton .
the branching fraction for the clean sequential decay mode is approximately @xmath164 @xcite .
table 1 shows the numbers @xmath126 of @xmath1 leptons required to detect cp violation through the cp - odd rate and polarization asymmetries @xmath112 at the 1-@xmath28 level for the maximally - allowed cp - odd mhd and slq parameters ( [ bound ] ) and @xmath165 .
all the cp - odd rate and polarization asymmetries require a similar number of @xmath1 decay events .
but , we note that the results for the cp - odd polarization asymmetries have been obtained for completely - polarized @xmath1 leptons . therefore , in a realistic experiment with partially - polarized @xmath1 leptons
the analysis power of the polarization - dependent observables will be reduced . in his recent works @xcite ,
tsai has claimed that @xmath1 polarization can play a crucial role in probing p , cp and t violation in @xmath1 decays .
however , we see that at least in the cabbibo - suppressed @xmath1 decay @xmath130 it is crucial to highly polarize the @xmath1 leptons to fully utilize the cp - odd polarization - dependent observables .
numerically , we find that cp violation in the mhd model and the two slq models may be detected with about @xmath8 and @xmath9 @xmath1 leptons for the maximally - allowed cp - odd mhd and slq parameters , respectively .
table 1 . _
the number of @xmath1 leptons , @xmath126 , needed for detection with the @xmath125 at the @xmath10 level , are determined for @xmath166 mev with @xmath167 in the mhd model , and @xmath168 and @xmath169 in the two slq models .
_ 0.8 cm [ cols="^,^,^,^,^",options="header " , ] in summary , we have investigated cp violation from the mhd model and slq models in the cabbibo - suppressed decay @xmath0 with polarized @xmath1 leptons to which both the @xmath2 resonance @xmath3 and the @xmath4 resonance @xmath5 contribute .
in addition to the cp - odd rate asymmetry , @xmath1 polarization enables us to construct three additional cp - odd polarization asymmetries whose magnitudes depend crucially on the @xmath5 decay constant , @xmath6 . taking a qcd sum rule estimate of @xmath7 mev and the present experimental constraints on the cp - odd parameters into account ,
we have quantitatively estimated the maximally - allowed values for the cp - odd rate and polarization asymmetries in the mhd and slq models consistent with the sm gauge symmetry with massless left - handed neutrinos .
we have found that the cp - odd rate and polarization asymmetries are of a similar size for highly - polarized @xmath1 leptons and , for their maximally - allowed values , new scalar - fermion interactions may be detected with about @xmath8 or @xmath9 @xmath1 leptons at the @xmath10 level .
consequently , we conclude that since the @xmath1 leptons of the order of @xmath9 or more are expected to be produced yearly at the planned @xmath154 factories @xcite and the proposed @xmath1-charm factories @xcite , it is important to look for cp violation in the cabbibo - suppressed @xmath1 decay @xmath170 even without polarized @xmath1 leptons . |
a common approach to explain the observed hierarchies in fermion masses and mixings is in terms of flavour symmetries .
the matter fields transform under the flavour symmetry , which is spontaneously broken by the vacuum expectation values ( vevs ) of scalar fields that we will call flavons in the following .
most yukawa couplings are forbidden at the renormalisable level and only arise from higher - dimensional operators involving suitable powers of the flavons as determined by the flavour symmetry @xcite .
the flavour hierarchies are then explained by small order parameters given by the ratio of flavon vev and the uv cutoff scale .
this scale itself remains undetermined and can in principle be as large as the planck scale . in case
it is smaller , one can interpret this cutoff scale as the typical mass scale of new degrees of freedom , the so - called `` flavour messengers '' .
the dynamics of this sector may have important impact on low - energy physics even if its characteristic scale is very high . in this work
we want to systematically analyse the structure of the messenger sector and show that it can indeed have important consequences both for yukawa couplings and soft masses in the mssm . in the uv completion that contains the messenger sector , small fermion masses
can be thought to arise from small mixing of light fields either with heavy fermions or with heavy scalars . in the first case , which has been studied extensively in refs .
@xcite , fermion masses correspond to three light eigenvalues of a large mass matrix . in the second case ,
light masses arise from small vevs of the heavy scalars .
this latter case has received less attention despite its structure is much simpler .
such scalar messengers are very suitable to generate texture zeros in the yukawa matrix @xcite , that is , the vanishing of certain entries although allowed by the flavour symmetry .
independently of the kind of messengers , in general one needs many of them .
since these fields carry sm quantum numbers , they contribute to the rg evolution of sm gauge couplings .
the requirement of maintaining perturbativity up to certain scale puts a lower bound on the messenger masses as a function of their number . in this work
we require perturbative physics up to a very high scale of supersymmetry breaking mediation , the planck scale for gravity mediation or the gauge mediation scale . in consequence
the flavour messengers must be very heavy .
direct effects of their exchange are then strongly suppressed , but , when the flavour messenger scale is lower than the susy breaking mediation scale , they are relevant for the flavour dependence of the soft susy breaking terms .
this is because in the supersymmetric theories we are considering the soft susy breaking terms are sensitive to any dynamics that couple to light fields .
one can distinguish tree - level effects on the soft mass terms , generated by integrating out the flavour messengers and determined by the flavour symmetry alone @xcite , and radiative effects summarised by the rg evolution between the susy breaking mediation scale @xmath1 and flavour messenger scale @xmath0 , with the flavour messengers included .
the latter are particularly interesting as they necessarily break flavour universality of soft terms , even if it is assumed ( or naturally present ) as the initial condition at @xmath1 .
we quantitatively estimate these effects and show that they always give in the super - ckm basis flavour off - diagonal terms in soft masses comparable ( in their order of magnitude ) to those one would obtain at tree - level without assuming universality .
thus the evolution of the soft terms according to the mfv hypothesis @xcite from their initial universal values is possible only if @xmath0 is larger than @xmath1 . for gravity mediation , the flavour physics
is then pushed to the planck scale .
the structure of our paper is as follows : in section 2 we first discuss the general structure of the messenger sector , in particular the two possibilities of uv completions with heavy fermions and heavy scalars , respectively .
this allows us to estimate the typical number of messengers that we use in section 3 to calculate the bounds on the messenger scale from perturbativity . in section 4
we discuss the impact of the messenger sector on the flavour structure of soft terms .
in particular we present constraints on light rotation angles valid in large classes of flavour models .
we finally conclude in section 5 . in the appendix we provide explicit examples of uv completions of three popular flavour models in the literature in order to illustrate our general discussion .
we want to consider models with a general flavour symmetry group @xmath10 spontaneously broken by the vevs of the flavon superfields @xmath11 .
the mssm yukawa couplings arise from higher - dimensional @xmath10 invariant operators involving the flavons @xmath12 where the suppression scale @xmath13 is the typical scale of the flavour sector dynamics .
the coefficients of the effective operators are assumed to be @xmath14 , so that yukawa hierarchies arise exclusively from the small order parameters @xmath15 .
the transformation properties of the mssm fields and the flavons under @xmath10 are properly chosen , so that @xmath16 together with their exponents @xmath17 reproduce the observed hierarchies of fermion masses and mixings . in order to uv - complete these models one has to `` integrate in '' heavy fields at the scale @xmath0 .
these messenger fields are vectorlike and charged under @xmath10 .
they couple to the flavons and the light matter . in the fundamental theory
they mix either with mssm matter or higgs fields ( after flavour symmetry breaking ) . in the first case one
has to introduce chiral superfields @xmath18 with the quantum numbers of the mssm matter fields ( see fig .
[ fchain ] ) .
schematic supergraph for fermion uvc.,scaledwidth=90.0% ] in the second case one introduces chiral fields @xmath19 with the quantum numbers of the mssm higgs fields and @xmath20-even gauge singlets , respectively ( see fig . [ hchain ] ) .
schematic supergraph for higgs uvc.,scaledwidth=92.0% ] we are now going to discuss in more detail these two possibilities , to which we refer as fermion uv completion " ( fuvc ) and higgs uv completion " ( huvc ) .
although we restrict to the pure cases , a uv completion involving both kind of fields is also viable .
such a case is a straightforward generalisation of the following discussion . in order to uv -
complete the effective theory above with fermionic messengers , one has to introduce certain numbers of vector - like superfields with light fermion quantum numbers @xmath21 with superpotential interactions of the schematic form @xmath22 where we restricted to the up - sector for simplicity .
moreover we have dropped all couplings that are assumed to be @xmath14 , since all hierarchies should arise from the symmetry breaking alone ( which also implies @xmath23 ) .
the allowed couplings follow directly from the transformation properties of the messengers under @xmath10 .
they have to be chosen appropriately , such that the effective yukawa couplings in eq .
( [ yuk ] ) are generated upon integrating out the messengers .
this choice can be conveniently carried out by drawing tree - level feynman diagrams , see fig .
[ fchain ] , in which a given yukawa entry @xmath24 is induced by the interactions in eq .
( [ flavonint ] ) with the number of flavon insertions given by @xmath25 in eq .
( [ yuk ] ) .
the diagram dictates the required messenger couplings and therefore their @xmath10 quantum numbers . in the following
, we are going to refer to these diagrams as `` chain '' diagrams , for which we introduce the shorthand notation : @xmath26 notice that for generating a yukawa entry with a given number of flavon insertions one can write several chain diagrams using different messengers .
this ambiguity arises from the possible permutations of the higgs and flavon insertions in fig .
[ fchain ] .
for instance , one can choose the position of the higgs insertion , which corresponds to the number of left - handed " ( @xmath27 ) and right - handed " ( @xmath28 ) messengers one wants to use .
in particular it is possible to use only left - handed or right - handed messengers .
all entries of the light yukawas can then be generated using the chain diagrams . to be economic
one can use the same messengers for different chains , but one has to pay attention that the resulting yukawa matrix has full rank , which might not be the case because of possible correlations of different entries . an elegant and simple way to find the minimal number ( meaning minimal number of sm representations ) of messengers needed in each sector
was pointed out in ref .
@xcite . in the full theory
the role of the messenger couplings in eq .
( [ flavonint ] ) is to generate three light eigenvalues of the full mass matrix that involves the chiral fields and the vectorlike messengers .
since the heavy eigenvalues are @xmath29 , the determinant of the full mass matrix must be equal to the determinant of the light mass matrix , up to powers of @xmath0 : @xmath30 because every entry of the full mass matrix involves at most one power of @xmath11 , one needs at least @xmath31 messengers with @xmath32 that is , the minimum number of messengers can be found simply by counting the powers of @xmath16 in the determinant of the yukawa matrix . although this argument gives the minimal number of messengers , it is unclear which messengers have to be included .
we therefore outline a simple , model - independent , procedure to derive minimal sets of fermionic messengers for each sector .
one first identifies the three entries of the effective yukawa matrix that gives the leading order contribution to the determinant .
if the leading order determinant is the sum of several terms one can choose any of the summands . then one constructs the chain diagrams for the three chosen entries using different messengers for different entries , even if they have the same quantum numbers . in this way one
adds a total number of messengers that is precisely @xmath31 .
usually there are several solutions obtained from permuting the higgs and flavon insertions in each diagram . by counting the total number of these permutations for each of the three entries
one can easily find the total number of possible uv completions with minimal number of fermionic messengers . in appendix
a we explicitly construct the fermion messenger sector for three examples in order to illustrate the general procedure .
the above method guarantees that the resulting yukawa matrices are full - rank , i.e. all fermion masses are generated . in general it does not ensure that other yukawa entries besides the three chosen , and therefore the correct mixings , are generated as well .
however , we checked that this is the case for the three example models of appendix a. finally we want to elucidate the necessity of using different messengers for different entries .
let us assume that two yukawa entries @xmath33 , @xmath34 ( @xmath35 ) , contribute at leading order to the determinant and arise from chains that share a messenger .
one can then integrate out all messengers but this one .
the superpotential of eq .
( [ flavonint ] ) reduces to the following form : @xmath36 where @xmath37 and @xmath38 are some effective couplings involving appropriate powers of @xmath16 . by integrating out the last messenger
@xmath39 one clearly obtains a rank 1 matrix ( see the diagram below ) .
one can avoid this by introducing a copy of @xmath39 , i.e. a new messenger @xmath40 with the same quantum numbers as @xmath39 .
this gives rise to a new contribution to the effective superpotential @xmath41 which ensures that the effective yukawa matrix for @xmath42 and @xmath43 has rank 2 .
we now discuss the uv completion of flavour models with heavy higgs fields .
in general one has to introduce certain numbers of vector - like superfields @xmath44 where @xmath45 has the sm quantum number of @xmath46 and @xmath47 is a gauge singlet .
the superpotential interactions are of the schematic form @xmath48 again we dropped all couplings that are assumed to be @xmath14 and take @xmath49 .
the required couplings and therefore the transformation properties of the messengers under @xmath10 can be inferred from chain diagrams like in fig .
[ hchain ] .
note that for generating all yukawa entries , in general corresponding to different charges , one needs different higgs fields coupling to @xmath50 for each @xmath51 . for generating a yukawa entry in general one can write several diagram with different messengers , corresponding to the possible permutations of higgs and flavon insertion in fig .
[ hchain ] .
in particular , it is possible to use only @xmath45 messengers and avoid gauge singlets .
explicit examples with huvc can be found in appendix a. in the fundamental theory , small fermion masses arise from small vevs of the @xmath45 messengers .
these vevs can be calculated by setting the messenger f - terms to zero and using the mssm higgs vev @xmath52 as a background value .
solving the f - term equations @xmath53 is then equivalent to integrating out the messengers by their susy equations of motion .
this is analogous to the supersymmetric type - ii see - saw @xcite , where integrating out the heavy triplet or computing the vev of its scalar component are two equivalent ways to compute the effective neutrino masses .
notice that the interpretation of small fermion masses in the fundamental theory is much easier for the higgs uvc , since the calculation is formally the same as in the effective theory , while in the fermion messenger case one would have to diagonalise huge mass matrices .
this feature of huvcs allows to enforce texture zeros in the yukawa matrices in a very simple way .
although a certain yukawa entry would be allowed by the flavour symmetry , one can set it to zero because of particular dynamics in the messenger sector . from fig .
[ hchain ] it is clear that a specific yukawa entry can only arise if the corresponding coupling to a heavy higgs is present . if such a higgs with the correct transformation properties under @xmath10 is missing , the entry vanishes in the fundamental theory and remains zero in the low - energy effective theory . this elegant possibility to produce
texture zeros has been already outlined in @xcite . in appendix
a we illustrate this for the case of a u(1 ) flavour model .
as we have seen , the strongly hierarchical pattern of fermion masses requires a large number of messengers with sm quantum numbers .
these fields have a strong impact on the running of the sm gauge couplings , so that for a given messenger sector the messenger mass scale can not be too far below the cutoff scale in order to avoid landau poles .
if the messengers are coloured ( as in the case of fuvc ) , the strongest constraint typically arises from the running of @xmath54 which is sizable already at @xmath55 . in the case of huvc , instead the bound is set by @xmath56 and is usually weaker than in the previous case .
moreover some heavy higgs fields can be replaced by heavy singlets with no impact on running of the gauge couplings .
therefore one can expect that models with huvc will be less constrained than fuvc models regarding the bounds from perturbativity of gauge couplings .
similarly , the messenger lagrangian contains many couplings which are @xmath14 and one has to ensure that they do not blow up in the running between the messenger scale and the cutoff .
further constraints on the messenger sector finally arise when one requires that the approximate unification of gauge couplings in the mssm is not spoiled .
this is easily ensured if the messenger fields from complete su(5 ) multiplets .
we are now going to discuss these issues in more detail .
contour plot of the lower bound on the messenger scale @xmath57 for given @xmath58 ( left ) , @xmath59 ( right ) requiring perturbativity up to the cutoff scale @xmath60 ( obtained using eq .
( [ pert - bound])).,scaledwidth=90.0% ] we want to calculate the constraints on the messenger scale for a given uv completion with @xmath61 vector - like su(3)@xmath62 triplet messengers and @xmath63 vectorlike su(2 ) doublet messengers living at a scale @xmath0 .
above @xmath0 the 1-loop @xmath64-function coefficients @xmath65 of the gauge couplings @xmath66 get modified according to @xmath67 where @xmath68 are the mssm coefficients @xmath69 . requiring that @xmath66 remains in the perturbative regime ( @xmath70 ) up to the cutoff scale @xmath60 provides a lower bound on the messenger scale @xmath0 given by @xmath71 } , \label{pert - bound}\end{aligned}\ ] ] where we considered 1-loop running and neglected the susy thresholds .
[ pert ] shows the contours of the lower bound on the messenger scale @xmath57 for @xmath58 ( left ) , @xmath59 ( right ) and the cutoff @xmath60 . comparing the two panels
, one can see that the bound is indeed stronger if the messengers are coloured . in the case of a fuvc
, one can get , for a given effective model , the minimal number of triplets directly from eq .
( [ nmin ] ) . taking into account both up- and
down sector one has @xmath72 , provided that only right - handed messengers are used in both sectors . if also left - handed messengers are used , in general the number of triplets is larger , unless both isospin components of the quark doublet serve as messengers . in turn
, one can use the perturbativity bound as a criterion for defining the `` minimal '' fermionic messenger sector as the solution with the least number of color triplets , i.e. the messenger sector least constrained by perturbativity .
this notion can reduce drastically the degeneracy of minimal messenger sectors since only few solutions efficiently unify up and down sectors .
the uv completions for the example models in the appendix are chosen to fulfil this criterion of minimality , and can therefore be used to illustrate the minimal bounds on the messenger scale for typical flavour models .
we summarise the results in table [ compfh ] .
.[compfh]comparison of huvc and fuvc for the three example models in appendix a. @xmath57 denotes the lower bound on @xmath0 in gev requiring perturbativity up to @xmath73 .
[ cols="^,^,^ " , ] for the leptonic rotation angles we assumed the su(5 ) relations @xmath74 , @xmath75 . comparing to the bounds given in table [ absusytable ] , we see that the u(1 ) model ( as any abelian model ) is seriously challenged by the 1 - 2 sector constraints . even considering an additional suppression @xmath76 from vanishing scalar masses at @xmath1 and small cpv phases , the bounds can be evaded only for a quite heavy susy spectrum , @xmath77 tev ( notice that the bounds in table [ absusytable ] scale like @xmath78 ) . with this setup , the u(1 )
model should still exhibit sizable deviation from the sm in @xmath79 , @xmath80 mixing , as well as a rate for @xmath81 in the reach of the meg experiment . on the other hand ,
the non - abelian example is still perfectly compatible with the bounds .
notice however that large effects in su(3 ) models for @xmath82 cp violation and lfv are still possible , provided that the susy masses are not too heavy .
in this paper , we have discussed the general features of the uv completions of susy flavour models .
we have analysed in detail the structure of the messenger sector of this kind of models , which contains vector - like superfields that mix either with light fermions or with light higgs fields . in the latter case
, it is particularly simple to obtain texture zeros in the yukawa matrices , just by removing certain messengers from the theory , without modifying the transformation properties of the sm fields .
our discussion on the structure of the messenger sector does not rely on susy and can be easily generalised to non - supersymmetric scenarios .
in general , the messenger sector contains many new fields charged under the sm gauge group . as a consequence ,
requiring the theory to remain perturbative up to high - energy scales forces the masses of the messengers to be far above the tev scale , typically @xmath83 gev for the example models we have considered .
this implies that the messengers have no direct impact on low - energy observables .
still , their presence affects the rg running of the sfermion masses .
we have emphasised that the radiative generation of sfermion mass - splittings and misaligned a - terms is unavoidable whenever the flavour symmetry breaking scale is lower than the susy breaking scale , like in gravity mediation .
we have quantitatively estimated this rg effect and found it comparable to the tree - level off - diagonal sfermion masses as expected by the flavour symmetry . therefore the strong flavour constraints can not be evaded even under the assumption of universal soft terms .
these constraints depend only on the diagonal mass splitting and the rotation angles , and therefore apply to large classes of flavour models . in table
[ absusytable ] we have provided bounds on the rotation angles of the light fermions that are valid in any abelian flavour model and can be easily extended to simple non - abelian models .
we find that abelian models are strongly constrained and hence it is difficult to marry them to gravity mediation with susy at the tev scale .
even though it is well - known that abelian models induce large flavour changing effects , still we find it remarkable that this remains true even under the strong assumption of universal soft masses at the susy breaking mediation scale , as a consequence of the presence of flavour messengers affecting the rg running .
not surprisingly , in non - abelian models the sfermion mass - splittings can be suppressed with respect to the abelian case by small flavon vevs and ease the bounds on the rotation angles .
however , large effects for kaon and lfv observables are still possible , provided that the susy spectrum is not too heavy .
we would like to thank c. hagedorn , a. romanino and g. g. ross for useful discussions .
the feynman diagrams have been drawn using jaxodraw @xcite .
we thank the theory division of cern for hospitality during several stages of this work .
l.c . and r.z .
are grateful to the institute of theoretical physics of the university of warsaw for kind hospitality and financial support during their stays in warsaw .
s.p . and r.z .
acknowledge support of the tum - ias funded by the german excellence initiative .
this work has been partially supported by the contract pitn - ga-2009 - 237920 unilhc and the mniszw scientific research grant n202 103838 ( 2010 - 2012 ) .
in this appendix we explicitly construct the uv completions both with fermion and higgs messengers for three example models : u(1 ) , u(1)@xmath84u(1)@xmath85 and su(3 ) .
the first example is based on a u(1)@xmath86 flavour symmetry .
we take for the charges of the mssm superfields @xcite : @xmath87 the higgs fields are neutral and a single flavon is introduced with charge @xmath88 .
this gives rise to the effective yukawa matrices : @xmath89 the correct hierarchy of fermion masses and mixing can be achieved choosing the expansion parameter of the order of the cabibbo angle , @xmath90 .
the determinants of the yukawas matrices in eq .
( [ u1yuk ] ) are @xmath91 where @xmath92 is the charge of the fermion @xmath93 . according to eq .
( [ nmin ] ) , one needs in total @xmath94 messengers in the up - sector and @xmath95 messengers in the down - sector .
all terms contributing to the determinant are of the same order ( which is true in every u(1 ) model ) .
we can choose the term that is the product of the three diagonal entries .
these three entries are then generated from the chain diagrams .
for the chains in the up sector we choose ( from the total number of possibilities that is @xmath96 ) @xmath97 where the messengers are labelled with their u(1 ) charges . note that @xmath98 is present already on the renormalisable level .
for the down sector we choose ( from @xmath99 possibilities ) @xmath100 in total we have used 10 messengers in the up sector and 9 messengers in the down sector , which is exactly the minimal number required .
the charges of the heavy higgs messengers can be easily inferred from the chain diagram of fig .
[ hchain ] , giving 6 messengers in the up sector and 5 in the down sector which are necessarily distinct .
thus one has in total 11 higgs messengers ( plus conjugates ) that we denote by the charge of @xmath46 @xmath101 the relevant part of the superpotential is given by @xmath102 where @xmath46 and @xmath103 are the mssm higgs fields .
taking @xmath104 , the resulting yukawa couplings for the up and the down sector are given by @xmath105 with the shorthand notation @xmath106 notice that in contrast to the fermion messenger case one does not have to add copies of messengers , since in this case every entry comes with a different coupling @xmath107 which implies full rank .
as explained in section [ huvc ] , higgs uv completions easily allow for the presence of texture zeros in the yukawas , simply by removing the corresponding higgs messenger from the theory ( which however sets all yukawas which the same charge to zero ) . but the removal of a certain higgs might interrupt the chain needed for other yukawa entries .
this gap has then to be fixed by using additional singlet messengers . as an example
, we consider the above u(1 ) model where we now remove the heavy down - higgs messenger with charge -4 , that is @xmath108 , together with its conjugate partner @xmath109 .
this results in setting the 1 - 2 , 1 - 3 and 2 - 1 entries in @xmath110 . to zero .
but at the same time this also removes the coupling @xmath111 from the theory , so that also the 1 - 1 entry would be zero as @xmath112 would not take a vev . in order to restore it
, one has to introduce singlet messengers .
specifically , one must add vector - like singlets @xmath113 to the theory . besides mass terms
, the new allowed couplings are @xmath114 the last coupling induces a vev for @xmath115 , this one a vev for @xmath116 , this a vev for @xmath117 and @xmath118 , which finally gives together with the first interaction a vev to @xmath112 .
let us notice that with a similar procedure one can eliminate the 3 - 1 entry ( by dropping @xmath119 ) , whilst the 2 - 3 and 3 - 2 entries can not be set to zero without removing the 2 - 2 and the 3 - 3 entry , respectively . in this u(1 ) example , one can therefore enforce at most the following texture for @xmath110 : @xmath120 that has the interesting feature that the cabibbo angle arises entirely from the up sector .
next we consider the u(1)@xmath84u(1)@xmath121 model in ref .
the charges of the mssm superfields are @xmath122 the higgs are neutral and the three flavons @xmath123 , @xmath124 , @xmath125 have charges @xmath126 , @xmath127 , @xmath128 and take vevs @xmath129 , @xmath130 , @xmath131 .
the effective yukawas are then given by @xmath132 the determinants of the yukawas in eq .
( [ u1xu1yuk ] ) are @xmath133 so that one needs 3 messengers in the up and 5 in the down sector .
we again choose in both sectors the diagonal entries , and build the associated chains with right - handed messengers only .
in the up sector we take ( out of 6 possibilities ) the chain @xmath134 while in the down sector we take ( out of 18 possibilities ) @xmath135 in total we use 3 messengers for the up sector and 5 messengers for the down sector , which is the minimal number needed . by introducing heavy higgs messengers with charges ( plus conjugates ) : @xmath136 one can generate all entries in the yukawa matrices of eq .
( [ u1xu1yuk ] ) .
finally we provide an example with a non - abelian flavour symmetry .
we take the su(3 ) model in ref @xcite .
all the mssm superfields , @xmath138 , @xmath139 and @xmath140 transform as triplets of su(3 ) .
three anti - triplets flavons are introduced with vevs of the form @xmath141 with @xmath142 , @xmath143 and @xmath144 .
the expansion parameter @xmath145 is assumed to be different in the up and down sector , with @xmath146 , @xmath147 . in order to differentiate lepton and down - quark mass matrices , a field @xmath148
is introduced , with @xmath149 .
unwanted operators are forbidden with additional symmetries , under which the flavons are charged while the mssm superfields are neutral . for our purposes
, here we can simply take a single u(1)@xmath86 , with @xmath150 , @xmath151 , @xmath152 , @xmath153 .
the mssm higgs fields have @xmath154 .
the above set - up gives rise to yukawas of the form @xmath155 the determinant of the yukawas in eq .
( [ su3yukawas ] ) is @xmath156 so that 6 messengers in both sectors are required . from eq .
( [ su3yukawas ] ) , we see that the leading contribution to the determinant is given by the yukawa entries @xmath157 , @xmath158 and @xmath159 .
let us construct a minimal set of messengers for the up sector ( the down sector is analogous ) .
we choose to build the three chains with the higgs in the middle ( in total we have 27 possibilities ) @xmath160 where all 6 messengers are su(3 ) singlets and are labelled with the additional @xmath161 charge .
one needs to introduce ( both in the up and in the down sector ) a heavy higgs transforming as @xmath162 under su(3 ) and neutral under u(1)@xmath86 , which couples directly to the sm fermions .
would lead to antisymmetric yukawa matrices . ] in addition one needs three anti - triplets higgs with @xmath163 , @xmath164 , @xmath165 : @xmath166 or alternatively two sexplet singlets with @xmath167 , @xmath168 : @xmath169 plus the corresponding conjugate fields .
we now show an explicit example of the rg effect discussed in section [ pheno ] .
we consider the u(1 ) model described in appendix [ sec : u1 ] in the case of huvc .
the starting point is the superpotential in eq .
( [ higgs - ex ] ) . as an effect of the flavour symmetry breaking ,
higgs messengers mix among themselves .
the rges for the soft masses can be conveniently derived in the messenger mass eigenbasis .
concentrating on the down 1 - 2 sector , we obtain for the mass splittings in leading log approximation : @xmath170 \log\frac{m_s}{m } , \\
( m^2_{\tilde q})_{22 } - ( m^2_{\tilde q})_{11 } & \approx \frac{6}{16 \pi^2 } \tilde{m}_0 ^ 2 \left [ ( \lambda^{u\dag } \lambda^u)_{11 } - ( \lambda^{u\dag } \lambda^u)_{22}+ ( \lambda^{d\dag } \lambda^d)_{11 } - ( \lambda^{d\dag } \lambda^d)_{22}\right]\log\frac{m_s}{m } , \end{aligned}\ ] ] where we have neglected the contribution of the a - terms .
these splittings induce the following contributions to the off - diagonal entries in the super - ckm basis : @xmath171 \theta_{12}^{d r } , \\
( m^2_{\tilde q})^{\rm rot}_{12 } & = \left [ ( m^2_{\tilde q})_{22 } - ( m^2_{\tilde q})_{11 } \right ] \theta_{12}^{d l}. \label{sckm2}\end{aligned}\ ] ] the mixing angles can be easily estimated from eq .
( [ yuk - higgs - ex ] ) : @xmath172 as mentioned above there is also a contribution to the off - diagonal entries directly generated by the running , that at leading log reads : @xmath173 where again the effect of the trilinears is neglected . as we can see , these expressions do not cancel in general against the contributions of eqs .
( [ sckm1 ] , [ sckm2 ] ) , since they involve different @xmath14 coefficients .
they also show that the enhancement factor @xmath17410 we considered for our estimates is a reasonable approximation .
let us now discuss the same effect for the su(3 ) model with huvc .
we start from the following superpotential for the uvc of eq .
( [ su3uvc1 ] ) : @xmath175 where @xmath176 and @xmath177 denote su(3)@xmath137 indices and we omitted the messenger mass terms and other couplings that are not relevant for the present discussion . from the mass eigenbasis for the higgs messengers , we see that @xmath178 acquires small components of the other messengers : @xmath179 plugging this expression in eq .
( [ su3w ] ) , we see that @xmath180 couple with @xmath181 and @xmath182 non - universal in flavour , as @xmath183 for @xmath184 and @xmath185 for @xmath186 . in the running for @xmath187 and @xmath188 ,
this induces a splitting @xmath189 between the first and second generation sfermion masses and a splitting @xmath190 between the third generation mass and the other two .
therefore the induced off - diagonal entries @xmath191 and @xmath192 are just suppressed by an additional @xmath193 factor with respect to the abelian case discussed above and the entries involving the third family have no further suppression .
p. ramond , r. g. roberts and g. g. ross , nucl .
b * 406 * ( 1993 ) 19 [ arxiv : hep - ph/9303320 ] . k. kadota , j. kersten and l. velasco - sevilla , phys .
d * 82 * ( 2010 ) 085022 [ arxiv:1007.1532 [ hep - ph ] ] .
i. de medeiros varzielas and l. merlo , jhep * 1102 * ( 2011 ) 062 [ arxiv:1011.6662 [ hep - ph ] ] .
s. antusch , s. f. king and m. spinrath , phys .
d * 83 * ( 2011 ) 013005 [ arxiv:1005.0708 [ hep - ph ] ] .
l. calibbi , z. lalak , s. pokorski and r. ziegler , arxiv:1204.1275 [ hep - ph ] . for a recent discussion
see z. lalak , s. pokorski , g. g. ross , jhep * 1008 * , 129 ( 2010 ) [ arxiv:1006.2375 [ hep - ph ] ] .
a. masiero , s. k. vempati and o. vives , arxiv:0711.2903 [ hep - ph ] .
m. ciuchini , e. franco , d. guadagnoli , v. lubicz , m. pierini , v. porretti and l. silvestrini , phys .
b * 655 * ( 2007 ) 162 [ hep - ph/0703204 ] .
g. isidori , y. nir and g. perez , ann .
nucl . part .
* 60 * ( 2010 ) 355 [ arxiv:1002.0900 [ hep - ph ] ] .
j. adam _ et al . _
[ meg collaboration ] , phys .
lett . * 107 * ( 2011 ) 171801 [ arxiv:1107.5547 [ hep - ex ] ] .
l. calibbi , j. jones - perez and o. vives , phys .
d * 78 * ( 2008 ) 075007 [ arxiv:0804.4620 [ hep - ph ] ] ; l. calibbi , j. jones - perez , a. masiero , j. -h .
park , w. porod and o. vives , nucl .
b * 831 * ( 2010 ) 26 [ arxiv:0907.4069 [ hep - ph ] ] .
d. binosi , j. collins , c. kaufhold and l. theussl , comput .
commun . * 180 * ( 2009 ) 1709 [ arxiv:0811.4113 [ hep - ph ] ] ; d. binosi and l. theussl , comput . phys .
commun .
* 161 * ( 2004 ) 76 [ hep - ph/0309015 ] . | the flavour messenger sectors and their impact on the soft susy breaking terms are investigated in susy flavour models . in the case when the flavour scale @xmath0 is below the susy breaking mediation scale @xmath1 , the universality of soft terms , even if assumed at @xmath1 , is radiatively broken .
we estimate this effect in a broad class of models . in the ckm basis
that effect gives flavour off - diagonal soft masses comparable to the tree - level estimate based on the flavour symmetry . mpp-2012 - 60 + tum - hep-829/12 * the messenger sector of susy flavour models + and radiative breaking of flavour universality * * lorenzo calibbi@xmath2 , zygmunt lalak@xmath3 , + stefan pokorski@xmath4 , robert ziegler@xmath5 * @xmath2_max - planck - institut fr physik ( werner - heisenberg - institut ) , fhringer ring 6 , + d-80805 mnchen , germany _ @xmath6_institute of theoretical physics , faculty of physics , university of warsaw , hoa 69 , + 00 - 681 , warsaw , poland _ @xmath7_cern physics department , theory division , + ch-1211 geneva 23 , switzerland
_ @xmath8_tum - ias , technische universitt mnchen , lichtenbergstr . 2a , + d-85748 garching , germany _ @xmath9_physik department , technische universitt mnchen , james - franck - strae , + d-85748 garching , germany _ |
what is free - will for a physicist ? this is a very personal question .
most physicists pretend they do nt care , that it is not important to them , at least not in their professional life .
but if pressed during some evening free discussions , after a few beers , surprising answers come out .
everything from `` obviously i enjoy free - will '' to `` obviously i do nt have any free - will '' can be heard .
similarly , questions about time lead to vastly different , though general quite lean discussions : `` time is a mere evolution parameter '' , `` time is geometrical '' are standard claims that illustrate how poorly today s physics understands time .
consequently , a theory of quantum gravity that will have to incorporate time in a much more subtle and rich way will remain a dream as long as we do nt elaborate deeper notions of time .
i like to argue that some relevant aspect of time is not independent of free - will and that free - will is necessary for rational thinking , hence for science .
consequently , this aspect of time , that i ll name creative time - or heraclitus - time - is necessary for science . for different arguments in favor of the passage of time ,
see , e.g. , @xcite .
the identification of time with ( classical ) clocks is likely to be misleading ( sorry einstein ) .
clocks do not describe our internal feeling of the passage of time , nor the objective chance events that characterize disruptive times - the creative time - when something beyond the mere unfolding of a symmetry happens .
indeed , clocks describe only one aspect of time , the geometric , boring , parmenides - time . but let s start from the beginning . before thinking of time and even before physics and philosophy , we need the possibility to decide what we ll consider as correct statements that we trust and believe and which statements we do nt trust and thus do nt buy .
hence : + free - will comes first , in the logical order ; and all the rest follows from this premise .
+ free - will is the possibility to choose between several possible futures , the possibility to choose what to believe and what to do ( and thus what not to believe and not to do ) .
this is in tension with scientific determinism , according to which , all of today s facts were necessary given the past and the laws of nature .
notice that the past could be yesterday or the big - bang billions of years ago .
indeed , according to scientific determinism , nothing truly new ever happens , everything was set and determined at the big - bang .
this is the view today s physics offers and i always found it amazing that many people , including clever people , do really believe in this @xcite .
time would merely be an enormous illusion , nothing but a parameter labeling an extraordinary unraveling of some pre - existing initial ( or final ) conditions , i.e. the unfolding of some symmetry .
what is the explanatory power of such a view ?
what is the explanatory power of the claim that everything was set at the beginning - including our present day feelings about free - will - and that there is nothing more to add because there is no possibility to add anything .
clearly , i am not a compatibilist @xcite , i.e. not among those who believe that free - will is merely the fact that we always happen to `` choose '' what was already pre - determined to occur , hence that nothing goes against our apparently free choices .
i strongly believe that we truly make choices among several possible futures . before elaborating on all this ,
let me summarize my argument .
the following sections do then develop the successive points of my reasoning .
1 . free - will comes first in the logical order .
indeed , without free - will there is no way to make sense of anything , no way to decide which arguments to buy and which to reject .
hence , there would be no rational thinking and no science .
in particular , there would be no understanding .
since free - will is the possibility to choose between several possible futures , point 1 implies that the world is not entirely deterministic .
non - determinism implies that time really exists and really passes : today there are facts that were not necessary yesterday , i.e. the future is open .
4 . in addition to the geometrical time , there is also creative time .
one may like to call the first one parmenides - time , and the second concept of time heraclitus - time @xcite . both exist .
the tension between free - will and creative time on one side and scientific determinism on the other side dissolves once one realizes that the so - called real numbers are not really real : there is no infinite amount of information in any finite space volume , hence initial conditions and parameters defining evolution laws are not ultimately defined , i.e. the real numbers that theories use as inital conditions and parameters are not physically real .
hence , neither newtonian , nor relativity , nor quantum physics are ultimately deterministic . 6 .
consequently , neither philosophy nor science nor any rational argument can ever disprove the existence of free - will , hence of the passage of time .
as already mentioned in the introduction , free - will comes first . indeed ,
free - will is the possibility to choose between several possible futures , like the possibility to choose what to believe and what to do , hence also to choose what not to believe and not to do .
accordingly , without free - will one could not distinguish truth from false , one could not choose between different views .
for example , how could one decide between creationism and darwinism , if we could not use our free - will to choose among these possibilities ? without free - will all supporters of any opinion would be equally determined ( programmed ) to believe in their views . in summary , without free - will there would be no way to make sense of anything , there would be no rational thinking and no science . in particular , there would be no understanding .
furthermore , without free - will one could not decide when and how to test scientific theories .
hence , one could not falsify theories and science , in the sense of popper @xcite , would be impossible .
i was very pleased to learn that my basic intuition , expressed above , was shared and anticipated by a poorly known french philosopher , jules lequyer in the 19th century , who wanted to simultaneously validate science and free - will @xcite . as lequyer emphasized : `` without free - will the certainty of scientific truths would become illusory '' . and ( my addition ) the consistency of rational arguments would equally become illusory .
lequyer continues : `` instead of asking whether free - will is certain , let s realize that certainty requires free - will '' .
lequyer also emphasized that free - will does nt create any new possibilities , it only makes some pre - existing potentialities become actual , a view very reminiscent of heisenberg s interpretation of quantum theory
. however , lequyer continues , free - will is also the rejection of chance . for lequyer - and for me - our acts of
free - will are beginnings of chains of consequences .
hence , the future is open , determinism is wrong ; a point on which i ll elaborate in the next two sections .
lequyer did nt publish anything .
but , fortunately , had an enormous influence on another french philosopher , a close friend , charles renouvier who wrote about lequyer s ideas and published some of lequyer s notes @xcite . in turn
, renouvier had a great influence on the famous american philosopher and psychologist william james who is considered as one of the most influential american psychologists .
william james wrote `` after reading renouvier , my first act of free - will shall be to believe in free - will '' .
this may sound bizarre , but , in fact , is perfectly coherent : once one realizes that everthing rests on free - will , then one acts accordingly .
the existence of genuine free - will , i.e. the possibility to choose among several possible futures , naturally implies that the world is not entirely deterministic . in other worlds , today there are facts that were not necessary , i.e. facts that were not predetermined from yesterday , and even less from the big - bang .
recall that according to scientific determinism everything was set at the beginning , let s say at the big - bang , and since then everything merely unfolds by necessity , without any possible choice .
philosophers include in the initial state not only the physical state of the universe , but possibly also the character of humans - and living beings .
hence , let s recall that according to physical determinism everything is fully determined by the initial state of all the atoms and quanta at any time ( or time - like hypersurface ) and the laws of physics .
for example , given the state of the universe a nanosecond after the big - bang , everything that ever happened and will ever happen - including the characters , desires and reasons of all humans - was entirely determined by this initial condition . in other words , nothing truly new happens , as everything was already necessary a nanosecond after the big - bang . but how can one reconcile ideas about free - will such as summarized in the previous sections with scientific determinism ? or even with quantum randomness ?
this difficulty led many philosophers and scientists to doubt the very existence of free - will .
these so - called compatibilist changed the definition of free - will in order to make it compatible with determinism @xcite .
free - will , they argue , is merely the fact that we are determined to never choose anything that does nt necessary happen .
nevertheless , compatibilists argue , we have the feeling that our `` necessary choices '' are free .
this sounds to me like a game of words , some desperate tentative to save our inner feeling of free - will and scientific determinism .
but , as lequyer anticipated , free - will comes first , hence there is no way to rationally argue against its existence , for rational arguing requires that one can freely buy or not buy the argument : genuine compatibilists must freely decide to buy the compatibilists argument , hence compatibilists must enjoy free - will in lequyer s sense .
moreover , and this is my main point , scientific determinism is wrong , hence there is no need to squeeze free - will in a deterministic world - view .
let me emphasize that since free - will comes first , i.e. the possibility to choose between several possible futures comes first , and since this is incompatible with scientific determinism , the latter is necessarily wrong : the future has to be open , as we show in the next section . before explaining why physics , including classical newtonian physics , is not deterministic ,
let me address first two related questions : when do random ( undetermined ) events happen ? what triggers random events ?
already when i was a high school student , long before thinking seriously about free - will , the concept of randomness and indeterminism puzzled me a lot @xcite .
when can a random event happen ? what triggers its occurrence ?
if randomness is only a characteristic of long sequences , as my teachers told me , then what characterizes individual random events ?
what is the probability of a singular event ?
are nt long sequences merely the accumulation of individual events ?
the only interesting answer to the question `` when do random events happen ? '' i could find was given by yet another 19th century french philosopher ( there is no way to escape from one s cultural environment ) , antoine a. cournot @xcite .
his idea was that chance happens when causal chains meet .
this is a nice idea , illustrated , e.g. , by quantum chance which happens when a quantum system encounters a measuring device .
this idea can be illustrated by everyday chance events .
imagine that two persons , alice and bob meet up by chance in the street ( taken from @xcite ) .
this might happen , for example , because alice was going to the restaurant further down the same street and bob to see a friend who lives in the next street . from the moment they decide to go on foot , by the shortest possible path , to the restaurant for alice and to see his friend for bob , their meeting was predictable .
this is an example of two causal chains of events , the paths followed by alice and bob , which cross one another and thus produce what looks like a chance encounter to each of them .
but that encounter was predictable for someone with a sufficiently global view .
the apparently chance - like nature of the meeting was thus only due to ignorance : bob did not know where alice was going , and conversely .
but what was the situation before alice decided to go to the restaurant ? if we agree that she enjoys the benefits of free - will , then before she made this decision , the meeting was truly unpredictable .
true chance is like this .
true chance does not therefore have a cause in the same sense as events in a deterministic world .
a result subject to true chance is not predetermined in any way .
but we need to qualify this assertion , because a truly chance like event may have a cause .
it is just that this cause does not determine the result , only the probabilities of a range of different possible results are determined . in other words , it is only the propensity of a certain event to be realised that is actually predetermined , not which event obtains @xcite .
let s have a more physicist look at that .
first , consider two colliding classical particles , see fig .
[ figcollidingpart ] .
next , consider a unitary quantum evolution in an arbitrary hilbert space , see fig .
[ figunitaryevol ] .
look for a while at the latter one ; it is especially boring , nothing happens , it is just a symmetry that displays itself .
possibly the symmetry is complex and the hilbert space very large , but frankly , nothing happens as the equivalence between the schrdinger and the heisenberg pictures clearly demonstrates .
likewise , for a bunch of classical harmonic oscillators nothing happens .
somehow , there is no time ( or only the boring geometric time that merely labels the evolution ) .
similarly , as long as the classical particles of fig .
[ figcollidingpart ] merely move straight at a constant speed , nothing happens : in another reference frame they are at rest .
it is only when the classical particles collide , or when the quantum system meets a measuring apparatus , that something happens , as cournot claimed .
but one may object that in phase space the point that represents the 2 particles does nt meet anything . in phase space
, there is no collision , as collisions require at least two objects and in phase space there is only one object , i.e. one point .
moreover , the collision in real space and the consequence of that collision is already entirely determined by the initial conditions : in phase space it s only a symplectic symmetry that displays itself . and even if one assumes that each particle is initially `` independent '' , whatever that could mean , after colliding the 2 particles get correlated
. hence , for cournot s idea to work , one would need a `` correlation sink '' .
this is a bit similar to the collapse postulate of quantum theory which breaks correlations , i.e. resets independence ( separability ) . in summary
, cournot s idea is attractive , but not entirely satisfactory ; it does nt seem to fit with scientific determinism .
it took me a very long time to realize what is wrong with that claim .
consider a finite volume of space , e.g. a one millimeter radius ball containing finitely many particles .
can this finite volume of space hold infinitely many bits of information ?
classical and quantum theories answer is a clear `` yes '' . but why should we buy this assertion ?
the idea that a finite volume of space can hold but a finite amount of information is quite intuitive . however , theoretical physics uses real numbers ( and complex numbers , but let s concentrate on the reals , this suffices for my argument ) .
hence the question : are so - called real numbers really real ? are they physically real ? for sure
, it is not because descartes ( yet another french philosopher , but this time a well - known one ) named the so - called real numbers `` real '' that they are really real .
actually , the idea that real numbers are truly real is absurd : a single real number contains an infinite number of bits and could thus , for example , contain all the answers to all questions one could possibly formulate in any human language @xcite .
indeed , there are only finitely many languages , each with finitely many letters or symbols , hence there are only countably many sequences of letters .
most of them do nt make any sense , but one could enumerate all sequences of letters as successive bits of one real number @xmath0 , first the sequences of length 1 , next of length 2 and so on . the first bit after each sequence tells whether the sequence corresponds to a binary question and , if so , the following bit provides the answer .
such a single real number would contain an infinite amount of information , in particular , as said , it would contain the answer to all possible questions one can formulate in any human language . no doubt , real numbers are true monsters !
moreover , almost all so - called real numbers are uncomputable . indeed , there are only countably many computer programs , hence real numbers are uncomputable with probability one . in other words ,
almost all real numbers are random in the sense that their sequences of digits ( or bits ) are random .
let me emphasize that they are as random as the outcome of measurements on half a singlet maximally entangled with another spin @xmath1 . ] . and
these random numbers ( a better name for `` real '' numbers ) should be at the basis of scientific determinism ?
come on , that s just not serious ! +
imagine that at school you would have learned to name the so - called _ real numbers _ using the more appropriate terminology of _ random numbers_. would you believe that these numbers are at the basis of scientific determinism ? to name `` random numbers ''
`` real numbers '' is the greatest scam and trickery of science ; it is also a great source of confusion in the philosophy of science .
+ note that not all real numbers are random .
some , but only countably many , are computable , like all rational numbers and numbers like @xmath2 and @xmath3 .
actually , all numbers one may explicitly encounter are computable , i.e. are exceptional .
the use of real numbers in physics , and other sciences , is an extremely efficient and useful idealization , e.g. to allow for differential equations .
but one should not make the confusion of believing that this idealization implies that nature is deterministic . a deterministic theoretical model of physics
does nt imply that nature is deterministic . again , real numbers are extremely useful to do theoretical physics and calculations , but they are not physically real .
the fact that so - called real numbers have in fact random digits , after the few first ones , has especially important consequences in chaotic dynamical systems . after a pretty short time
, the future evolution would depend on the thousandth digit of the initial condition .
but that digit does nt really exist .
consequently , the future of classical chaotic systems is open and newtonian dynamics is not deterministic .
actually most classical systems are chaotic , at least the interesting ones , i.e. all those that are not equivalent to a bunch of harmonic oscillators . hence ,
classical mechanics is not deterministic , contrary to standard claims and widely held beliefs .
note that the non - deterministic nature of physics may leave room for emerging phenomena , like e.g. phenomena that could produce top - down causes , in contrast to the usual down - top causes we are used to in physics @xcite .
a well - known example of a set of phenomena that emerges from classical mechanics is thermodynamics which can be deduced in the so - called thermodynamical limit .
but , rather than going to infinite systems , it suffices to merely understand that classical mechanics is not ultimately deterministic , neither in the initial condition , nor in the set of boundary conditions and potentials required to define the evolution equations .
what about quantum theory ?
well , if one accepts that the measurement problem is a real physics problem - as i do , then this theory is also clearly not deterministic @xcite .
if , on the contrary , one believes in some form of a many worlds view , then the details of the enormously entangled wave function of the universe depends again on infinitesimal details , as in classical chaotic systems . note that although quantum dynamics has no hyper - sensitivity to initial conditions , it shares with classical chaotic systems hyper - sensitivity to the parameters that characterize that dynamics , e.g. the hamiltonian .
furthermore , open quantum systems recover classical trajectories also in the case of chaotic systems , see fig .
[ figqchaos ] . hence ,
quantum dynamics is not deterministic .
finally , bohmian quantum mechanics is again hyper - sensitive to the initial condition of the positions of the bohmian particles ; hence , like chaotic classical systems , bohmian mechanics is not deterministic .
admittedly , one may object that now we have an analog of the measurement problem in classical physics , as it is unclear when and how the non - existing digits necessary to define the future of chaotic systems get determined .
this is correct and , in my opinion , inevitable .
first , because free - will comes first , next because mathematical real numbers are physical random numbers . finally , because physics - and science in general - is the human activity aiming at describing and understanding how nature does it . for this purpose one needs to describe also how humans interact with nature , how we question nature @xcite . including the observer inside the description results , at best , in a tautology without any possible understanding : there would result no way to freely decide which description provides explanations , which argument to buy or not to buy . to summarize this section , claiming that classical mechanics is deterministic , or that quantum theory implies a many - world view , is like elevating `` real '' numbers , the determinism of newton s equations and the linearity of the schrdinger equation , to some sort of ultimate religious truth .
it is confusing mathematics with physics .
it is a common but profound epistemological mistake , see fig .
[ figphysreal ] .
so far we saw that free - will comes first in the logical order , hence all its consequences are necessary .
in particular one ca nt argue rationally against free - will and its natural consequence , namely that time really passes .
we also saw that this is not in contradiction with any scientific fact .
actually , quite the opposite , it is in accordance with the natural assumption that no finite region of space can contain more than a finite amount of information .
the widely held faith in scientific determinism is nothing but excessive scientism .
this can be summarized with the simple chain of implications : + _ free - will _ @xmath4 _ non - determinism _
@xmath4 _ time really passes
_ + let us look closer at the implications for time .
there is no doubt that time as an evolution parameter exists . to get convinced it suffices to look at a bunch of classical harmonic oscillators ( like classical clocks ) , or the unitary evolution of a closed quantum system , or at the inertial motion of a classical particle as in fig .
[ figcollidingpart ] .
this time is the boring time , the time when nothing truly new happens , the time when things merely are , time when what matters is _ being _ ,
i.e. parmenides - time .
one could also name this einstein s time . but let s look at the collision between the two particles of fig .
[ figcollidingpart ] .
the detail of the consequences of such a collision depends on non - existing infinitesimal digits , i.e. on mathematically real but physically random numbers . to get convinced just imagine a series of such collisions , this leads to chaos
; hence each collision is the place of some indeterminism , that is of some creative time , time when what matters is _
change_. hence we call this creative time heraclitus - time @xcite .
this creative time is extraordinarily poorly understood by today s science , in particular by today s physics .
this does nt mean that it does nt exist , or that it is not important .
on the contrary , it means that there are huge and fascinating open problems in front of us , scientists , physicists and philosophers .
notice that this is closely related to cournot s idea that random events happen when independent causal chains meet , e.g. when two classical particles meet .
the two particles are independent , at least not fully correlated , because their initial conditions are not fully determined . and their future , after the collision
, is not predetermined , but contains a bit of chance .
similarly , quantum chance happens when a quantum system meets a measurement apparatus , as described by standard textbooks .
admittedly , we do nt know what a measurement apparatus is , i.e. we do nt know which configurations of atoms constitute a measurement apparatus .
this is the so - called quantum measurement problem . according to what we saw ,
there is a similar problem in classical mechanics : despite the indeterminism in the initial conditions and evolution parameters , things get determined as time passes ( as discussed near the end of the previous section ) .
neither philosophy nor science can ever disprove the existence of free - will .
indeed , free - will is a prerequisite for rational thinking and for understanding , as emphasized by jules lequyer .
consequently , neither philosophy nor science can ever disprove that time really passes .
indeed , the fact that time really passes is a necessary consequence of the existence of free - will . the fact that today s science - including classical newtonian mechanics - is not deterministic may come as a huge surprise to many readers ( including the myself of 20 years ago ) .
indeed , the fact that descartes named _ real _ numbers that are actually physically _ random _ had enormous consequences .
this together with the tendency of many scientists to elevate their findings to some sort of quasi - religious ultimate truth - i.e. scientism - lead to great confusion , as illustrated by laplace famous claim about determinism and by believers in some form of the many - world interpretation of quantum mechanics , based respectively on the determinism of newton s equation and on the linearity of schdinger s equation .
once one realizes that science is not entirely deterministic , though it clearly contains deterministic causal chains , one faces formidable opportunities .
this might seem frightening , though i always prefer challenges and open problems to the claim that everything is solved .
non - determinism implies that time really passes , most likely at the junction of causal chains , i.e. when creative time is at work .
this leaves room for emerging phenomena , like thermodynamics of finite systems .
it may also leave room for top - down causality : the initial indeterminism must become determined before indeterminism hits large scale , much in the spirit of quantum measurements . as a side conclusion ,
note that robots based on digital electronics will never leave room for free - will , hence the central thesis of hard artificial intelligence ( the claim that enough sophisticated robots will automatically become conscious and human - like ) is fundamentally wrong .
so , am i a dualist ? possibly , though it depends what is meant by that .
for sure i am not a materialist . note that today s physics already includes systems that are not material in the sense that they have no mass , like electro - magnetic radiation , light and photons .
what about physicalism ?
if this means that everything can be described and understand by today s physics , then physicalism is trivially wrong , as today s theories describe at best 5% of the content of the universe .
more interestingly , if physicalism means that everything can be understood using the tools of physics , then i adhere to this view , though the fact that free - will comes first implies that physics can make endless progress , but without ever reaching a final point .
we will understand much more about time and about free - will , though we ll never get a full rational description and understanding of free - will .
just imagine this debate a century ago .
how naive anyone claiming at that time that physics provides a fairly complete description of nature would appear today .
similarly , for anyone making today similar claims .
let me make a last comment , a bit off - track .
free - will is often analyzed in a context involving human responsibility , `` how could we be responsible for our actions if we do nt enjoy free - will ? '' .
there is another side to this aspect of the free - will question : `` how could we prevent humans from destroying humanity if we claim we are nothing more than sophisticated robots ? '' , and `` how could one argue that human life has some superior value if we pretend we are nothing but sophisticated robots ? '' .
this work profited from numerous discussions , mostly with myself over many decades during long and pleasant walks .
i should also thank my old friend jean - claude zambrini for introducing me to cournot s idea , when we were both students in geneva .
thanks are due to chris fuchs who introduced me to jules lequyer and to many participants to the workshop on _ time in physics
_ organized at the eth - zurich by sandra rankovic , daniela frauchiger and renato renner .
j. norton , www.pitt.edu/ jdnorton / goodies / passage l. smolin , _ time reborn _ , houghton mifflin harcourt , 2013 . j. barbour , _ the end of time _ , oxford univ .
press , 1999 .
r. kane , _ a contemporary introduction to free will _ , oxford univ .
press , 2005 .
ch.h.s temann and ch.h .
stemann , _ heraclitus and parmenides - an ontic perspective _ , grin verlag gmbh , 2013 ; k. jaspers , _ anaximander , heraclitus , parmenides , plotinus , laotzu , nagarjuna _ , mariner books , 1974 .
popper , _ logik der forschung _
, 1934 ; _ the logic of scientific discovery _ , hutchinson , london , 1959 .
j. lequier , _ comment trouver , comment chercher une vrit premire _ , edition de leclat , 1985 ; see also j. grenier , _ la philosophie de jules lequier _ , calligrammes , 1983 ( isbn 2 903258 30 9 ) ; j. wahl , _ jules lequier _ , editions des trois collines , genve - paris , 1948
. w. logue _ charles renouvier , philosopher of liberty _ , louisiana state univ . press , 1993 .
n. gisin , _ propensities in a non - deterministic physics _ , synthese * 98 * , 287 ( 1991 ) .
a. cournot , _ exposition de la thorie des chances et des probabilits _ , librairie hachette , 1843 .
reprinted in part in _
etudes pour le centenaire de la mort de cournot _ , ed .
a. robinet , edition economica , 1978 .
n. gisin , _ quantum chance , nonlocality , teleportation and other quantum marvels _ , springer , 2014 .
g. chaitin , the labyrinth of the continuum , in meta math ! ,
vintage , 2008 .
ellis , in _ downward causation and the neurobiology of free will _ , eds murphy , ellis and oconnor , springer , 2009 .
brun , n. gisin et al .
lett . a 229 , 267 ( 1997 ) .
e. schrdinger , mind and matter , cambridge univ . press , 1958 . | today s science provides quite a lean picture of time as a mere geometric evolution parameter .
i argue that time is much richer .
in particular , i argue that besides the geometric time , there is creative time , when objective chance events happen .
the existence of the latter follows straight from the existence of free - will . following the french philosopher lequyer
, i argue that free - will is a prerequisite for the possibility to have rational argumentations , hence ca nt be denied .
consequently , science ca nt deny the existence of creative time and thus that time really passes . |
the accretion disk models have been developed much in the past several decades .
many disk models have been proposed and some of them are widely adopted in astrophysical studies nowadays , including but not restricted to shakura - sunyaev disk ( ssd ; see shakura & sunyaev 1973 ) , advection dominated accretion flow ( adaf ; see narayan & yi 1994 ; abramowicz et al .
1995 ) , and slim disk ( abramowicz et al . 1988 ) .
however , the outflow structure of accretion disks still remains an open problem .
the equations that describe the hydrodynamic processes are the navier - stokes equations , which are quite difficult to solve in the case of accretion disks which involve viscosity and radiation . therefore , in most works , some kind of simplification , such as one - zone or polytropic distribution and hydrostatic equilibrium , are usually applied in the vertical direction , and the vertical variation ( @xmath4 dependence in cylindrical coordinates ) of the velocity field is usually neglected . in this way the equations are changed to a set of ordinary differential equations ( odes ) in the radial direction , which can be solved numerically . however , by taking these assumptions , one can not get a clear picture of the vertical structure of accretion flows ; the velocity is always radially inward and no mass will cross the disk surface , displaying no outflow structure . among the exceptions is a work done by narayan & yi in 1995(hereafter ny95 ) , which used self - similar assumptions in the radial direction and solved the structure along the @xmath1 direction in spherical coordinates ( @xmath0 ) . however , in their work they assumed @xmath5 and thus can not get a proper velocity field , and their solutions compose of only pure inflow .
they argued that the bernoulli parameter is positive in their solutions so that a bipolar outflow is expected to develop near the vertical axis .
blandford & begelman ( 1999 , hereafter bb99 ) relaxed the mass conservation assumption and assumed that the mass inflow rate varies with radius , and obtained solutions with outflow called adiabatic inflow - outflow solutions ( adios ) .
their solutions are one - dimensional self - similar solutions that are height - averaged and they also applied the bernoulli parameter to argue the presence of outflow .
however , abramowicz et al.(2000 ) pointed out that the positive bernoulli function ( instead of bernoulli parameter , which is defined in convenience of self - similar assumptions ) is not sufficient for outflow ( also see the simulation works done by stone et al .
1999 and yuan & bu 2010 ) .
blandford & begelman ( 2004 ) furthered their work and presented some self - similar two - dimensional solutions of radiatively inefficient accretion flows with outflow .
they assumed hydrostatic equilibrium in the vertical direction and that convection dominates the heat transport , which may only be applicable in certain cases .
xu & chen ( 1997 , hereafter xc97 ) relaxed @xmath5 and obtained two types of solutions with outflow : accretion and ejection solutions .
however their solutions require the net accretion rate to be 0 , which is not realistic .
xue & wang(2005 , hereafter xw05 ) followed ny95 and solved the disk structure along the @xmath1 direction considering @xmath6 .
they arbitrarily set a disk surface , at which @xmath7 , and the sound speed on the surface for their calculation .
their solutions display a field of inflow near the equatorial plane with wind blowing out of the upper boundary , however the boundary is set as an input parameter rather than being calculated , and they only investigated several cases of adafs . in
3.2 we will see that according to our calculation , their assumption does nt hold for accretion flows with large @xmath2 value .
sadowski et al.(2010 ) abandoned the self - similar assumptions and solved the accretion disk structure in the radial and vertical direction simultaneously .
because the navier - stokes equations for accretion disks can not be decoupled intrinsically , they adopted other assumptions , e.g. the disk is not geometrically thick , to derive the equations . as in their work they did nt consider @xmath8 while @xmath9 and @xmath10 were supposed not to vary vertically , they were not able to study outflows . in summary , in the analytical model study , the vertical or @xmath1-direction structure of accretion disks and outflows can still not be dealt with satisfactorily , and many improvements are still needed to be made . on the other hand , observationally there are more and more evidences for outflows in accretion systems , such as sgr a*(marrone et al .
2006 ; xie & yuan 2008 ) , soft x - ray transients(loeb et al . 2001 ) and quasars with blueshifted absorption lines(e.g . ,
pg1115 + 80 ; chartas et al . 2003 ) .
many numerical simulation works have also discovered outflows in their results ( e.g. stone et al . , 1999 ;
igumenshchev & abramowicz , 2000 ; okuda et al . 2005
; ohsuga et al.2005 ; ohsuga & mineshige 2007 ; ohsuga et al .
the common existence of outflows in these works inspire us to investigate the vertical structure of accretion flows explicitly , to find solutions which can deal with @xmath6 and positive @xmath9 to get a clear velocity field , and with more reasonable boundary conditions . as a first step
, we followed the work done by ny95 and xw05 , using self - similar assumptions in the radial direction and solving the odes along the @xmath1 direction in spherical coordinates ( @xmath0 ) .
we used the @xmath2 viscosity prescription and assumed that the @xmath11 component of the viscosity stress tensor is dominant . by neglecting other components of the viscosity stress tensor , the necessary number of boundary conditions is reduced and we will only need the boundary conditions in the equatorial plane , which is obatined by symmetry and is thus quite physical . as we did nt set other arbitrary restrictions to @xmath9 and @xmath6 other than the self - similar assumptions , we can get a velocity field containing positive @xmath9 , to discuss the flow structure with possible outflows .
these assumptions are applicable to different kinds of accretion disk models , ranging through ssd , adaf and slim disk , so we also did many calculations with different sets of parameters ( which is very difficult to do in a numerical simulation work as it would be too time - consuming ) , trying to find the flow structure dependence on the parameters , in order to understand the inflow / outflow mechanism more physically
. these results can also be helpful to future numerical simulation works when they set the input parameters .
it should be noted that there is another branch of researches which investigate accretion flows with an outflow ( usually wind ) plus accretion disk model . in these researches ,
the configuration of the outflow and accretion disk is usually preset , and the calculations are focused on either the outflow or the accretion disk while the influence of the other part is simplified or parameterized ( e.g. fukue , 1989 ; takahara et al . , 1989 ; kusunose , 1991 ; bb99 ; misra & taam , 2001 ; fukue , 2004 ; xie & yuan , 2008 ) .
recently there are also several works done in this way which deal with the outflow and the accretion disk simultaneously ( kawabata & mineshige , 2009 ; dotan & shaviv , 2010 ) . compared with our work , in these studies
the accretion disk is usually height - integrated and the configuration of the accretion flow is assumed rather than being calculated , while in our work we solve the full hydrodynamic equations to get the configuration of the accretion flow .
our work focuses on studying the general structure of disks , outflows and the physical mechanism behind them , and the results can be complementary to one another .
in 2 we present the basic equations and assumptions we used in our calculations . in
3.1 we discuss our numerical methods and present solutions corresponding to typical parameters of ssd , adaf and slim disk . in 3.2 we show the disk structure dependence on different parameters and discuss their physical meanings .
4 we present our summary and discussion .
we consider the hydrodynamic equations of an accretion flow in the spherical coordinates ( @xmath0 ) . the flow is assumed to be steady(@xmath12 ) and axisymmetric(@xmath13 ) .
thus the continuity equation can be written as @xmath14 we assume that for the accretion flow , only the @xmath11-component of the viscous tensor , @xmath15 , is dominant . and
we use the newtonian gravitational potential , @xmath16 .
then the equations of motion can be written as(kato et al .
2008 , also see appendix a of xw05 ) @xmath17 @xmath18 @xmath19 in our calculation , we adopt the @xmath2 prescription of viscosity @xmath20 and @xmath21 is the total pressure .
the shearing box radiation magnetohydrodynamics ( mhd ) simulations done by hirose et al .
( 2009 ) found that the vertically integrated stress is approximately proportional to the vertically averaged total thermal ( gas plus radiation ) pressure . in our work , however , we also apply this relation locally .
following ny95 , we use the advective factor , @xmath22 , i.e. a fraction @xmath3 of the dissipated energy is advected as stored entropy and a fraction @xmath23 is lost due to radiation . in our calculation
we assume that @xmath3 is constant in the accretion flow , so the energy equation can be written as @xmath24 where @xmath25 is the internal energy of the material and can be expressed as(kato et al .
2008 ) @xmath26 where @xmath27 is the heat capacity ratio and is considered to be a constant input parameter in our calculation .
we adopt the self - similar assumptions in the radial direction , therefore we seek a solution of the form @xmath28 this set of self - similar solutions is similar to that of ny95 and is actually identical to that of xw05 . compared with those works , here we use the total pressure @xmath21 in equation ( 11 ) instead of the sound speed @xmath29 , which is proportional to @xmath30 . as stated in ny95 , the only length scale in the problem is @xmath31 and the only frequency is @xmath32 , thus all velocities must scale with radius as @xmath33 . in ny95 , narayan and
yi set @xmath34 , which implies @xmath5 according to the continuity equation , thus they intrinsically set no outflow for the accretion disk .
here we relax this parameter @xmath35 following bb99 and xw05 , to allow outflows from the disk .
if we substitute equations ( 7)-(11 ) into equations ( 1)-(4 ) , the @xmath31 components can be eliminated , leaving only the dimensionless fuctions @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , the variable @xmath1 and some constants which are set as input parameters .
the energy equation ( equation ( 5 ) ) is somewhat more complex . as the internal energy depends differently on gas pressure and radiation pressure ,
we discuss the energy equation more carefully here .
first , one may notice that only the total pressure @xmath21 appears in equations ( 1)-(4 ) .
this is not hard to understand as the dynamical processes do nt recognize whether the pressure is from gas or radiation . @xmath41 and @xmath42 do affect the energy equation in a different way , and to describe this effect , we define the gas pressure ratio @xmath43 , @xmath44 there is one particular case that when @xmath45 , no matter what the value of @xmath43 is , the solution remains the same , as in this case equation ( 6 ) becomes @xmath46 ; if the accretion flow is radiation pressure dominated ( @xmath47 ) , then equation ( 6 ) also becomes @xmath46 , and the flow structure wo nt depend on @xmath27 ; if the accretion flow is gas pressure dominated , then equation ( 6 ) becomes @xmath48 , and the result will depend on the value of @xmath27 . for a general case , equation ( 6 ) can be written as @xmath49p.\ ] ] physically @xmath43 is a value between 0 and 1 , so @xmath50 is always between @xmath51 and @xmath52
. then we can expect that a general solution should be somewhere between two extreme cases : the gas pressure dominated case and the radiation pressure dominated case .
for these two extreme cases , we can calculate the exact solutions of the problem , as equation ( 6 ) will be simplified to forms without @xmath43 .
these two cases also provide the typical solutions we emphasize in 3.1 .
however we also want to know how the solutions change from the gas pressure dominated case to the radiation pressure dominated case . for
the accretion flows in which both gas pressure and radiation pressure are important , we have to assume the gas pressure ratio @xmath43 to be constant to solve the problem . solving
the problem with variant @xmath43 is beyond the capability of this self - similar treatment , and we re planning to do that in our next work .
as we already mentioned , the radiation pressure dominated case has the same result as that of the gas pressure dominated case with @xmath45 , so the question eventually becomes examining how the solution changes according to @xmath27 for the gas pressure dominated case .
we define an equivalent @xmath27 here @xmath53 so that equation ( 6 ) becomes @xmath54 and this @xmath55 represents the equivalent @xmath27 that the accretion flow would have if the flow is treated as gas pressure dominated ( even @xmath56 ) .
we can see that if the flow is radiation pressure dominated ( @xmath57 ) , then no matter what the value of @xmath27 is , @xmath55 is always 4/3 .
here we treat @xmath27 as a constant , and for any specific case , we can get a constant @xmath55 and all the situations can be treated as a gas pressure dominated flow with @xmath55 as an input parameter .
usually @xmath27 is taken as a value between 7/5(the value for diatomic ideal gas ) and 5/3(the value for monatomic ideal gas ) , so that @xmath55 ranges between 4/3 and 5/3 . with equations ( 7)-(11 ) and ( 15 ) , the equations ( 1)-(5 ) can be reduced to a set of five odes : @xmath58=0,\ ] ] @xmath59=0,\ ] ] @xmath60=0,\ ] ] @xmath61=0,\ ] ] @xmath62 - 2\gamma_\mathrm{equ } v_\theta(\theta)\frac{d\rho(\theta)}{d\theta}\right\}=0,\ ] ] with 5 dimensionless fuctions @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , the variable @xmath1 and four input parameters ( @xmath2,@xmath3,@xmath55,@xmath35 ) .
this set of odes can be numerically solved with proper boundary conditions .
we assume the structure of the disk is symmetric to the equatorial plane , and thus we have @xmath63 in which only four conditions are independent .
for the last boundary condition we set @xmath64 , which can be normalized by a scale factor if the effective accretion rate at a certain radius is set ( ny95 , xw05 , etc . ) .
these boundary conditions are enough for our calculations and we did nt introduce other arbitrary boundary conditions .
we obtained numerical solutions of equations ( 16)-(20 ) with different sets of input parameters ( @xmath2,@xmath3,@xmath55,@xmath35 ) .
some typical solutions are shown in figures 1 - 4 .
the calculation starts from the equatorial plane ( @xmath65 ) towards the vertical axis ( @xmath66 ) .
generally the mass density @xmath67 and total pressure @xmath21 will decrease as @xmath1 decreases , and at certain inclination they will get very close to 0 as shown in the figures .
we take this as the upper boundary of the flow structure .
if we continue the calculation through this inclination , it will encounter numerical errors .
we think the reason is that we ca nt describe the flow near the vertical axis with a simple self - similar solution in the radial direction .
this effect is also to be expected .
if we describe the upper boundary ( minimum @xmath1 ) that we reach in our calculation as @xmath68 , then the effective accretion rate @xmath69 across a sphere at radius @xmath31 within the region calculated by us is @xmath70 which is a function of @xmath31 unless @xmath34 . in equation
( 22 ) ( and subsequently in this paper ) negative values of @xmath69 represent inflow while positive values represent outflow .
if we describe the accretion rate in the region between the vertical axis and the inclination @xmath68 as @xmath71 , then according to the steady nature of the flow , we have @xmath72 in which @xmath73 represents the total accretion rate across any sphere at a reasonable radius centered by the central accretor and should be a constant for a steady accretion flow .
if the solution does nt end at an upper boundary , and instead can describe the entire flow structure in the whole space , then @xmath74 and @xmath75 , which should be a constant . according to equation ( 22 ) , this can only happen in the following two cases : ( 1 ) @xmath34 which enforces @xmath69 not to change with radius @xmath31 ; ( 2 ) when @xmath76 , the integration term in equation ( 22 ) must be 0 , in which case @xmath77 and is a constant along radius @xmath31 .
the first case was discussed in ny95 . because @xmath34 , @xmath78 is independent of @xmath31 , and the continuity equation ( 1 ) shows that @xmath5 , resulting in a solution in which the flow is always radial ( with rotation ) .
the second case was discussed in xc97 , and the fact that @xmath79 requires that the outflow rate exactly equals the inflow rate at any radius .
however this is unrealistic for an accretion flow , which is also discussed in xw05 .
the reason is that , when material is accreted in the form of an accretion flow , gravitational energy is released and part of it is changed to internal energy via viscous friction .
the restriction that outflow rate equals inflow rate requires that the internal energy released from gravitational energy should be fully returned to gravitational energy , which violates the second law of thermodynamics .
therefore , the inflow rate must be larger than the outflow rate at a certain radius , in order to compensate for the increase in entropy in the hydrodynamic process .
as we have mentioned before , there are many observations of accretion systems with outflows , and they can be neither of the two cases mentioned above ( case ( 1 ) has no outflow while case ( 2 ) violates the second law of thermodynamics ) .
so the self - similar solutions for an accretion disk with outflow have to be truncated at some inclination .
it should be noted that this effect has already been discussed in xw05 , though they assumed that the self - similar solution is only valid for the inflow part .
our solutions , however , show that at least part of the outflow ( with @xmath80 ) can be described with self - similar assumptions . from the above analysis
, we can also see that @xmath35 is a very important parameter .
when @xmath34 , the entire flow structure can be described by the same set of self - similar solutions , however this kind of solutions have no outflow .
when @xmath81 , self - similar solutions can only describe part of the entire flow structure . according to equation ( 22 ) , when @xmath82 , the effective accretion rate @xmath69 decreases as @xmath31 decreases , implying that material is lost due to outflows .
this kind of solutions contain outflows , which is also consistent with the results of many numerical simulation works ( e.g. okuda et al .
2005 ; ohsuga et al.2005 ;
ohsuga & mineshige 2007 ; ohsuga et al .
2009 ) . on the other hand , when @xmath83 , the solutions have an effective accretion rate @xmath69 that increases towards the central accretor , which implies that there must be matter injection into the accretion flow from high latitudes .
this kind of matter injection is not likely to happen in real cases .
so we would like to investigate solutions with @xmath84 .
solutions with @xmath34 have been examined in ny95 , and they are only for adafs , while solutions with @xmath82 in our work is applicable to a much wider range of accretion disk types .
so in this subsection , we take @xmath85 to calculate the typical solutions . in this way
we can observe how the solutions differ from those of @xmath34 when @xmath35 does nt change much .
we also calculated how the solutions change with different @xmath35 , which will be discussed in detail in 3.2 .
the result shows that , when other parameters are the same , for a large range of @xmath35 with @xmath82 the solutions display similar structures qualitatively . the value of the viscosity parameter @xmath2 can be inferred both from observations and from mhd simulations of magneto - rotational instability ( mri ) . as reviewed by king et al .
( 2007 ) , the value of @xmath2 for ionized disks given by numerical simulations is @xmath86 , while that obtained through observations is @xmath87 . however , the observational determinations of @xmath2 are strongly model - dependent and it still remains an open question what the value of @xmath2 should be for hot , ionized disks . in this subsection
we take @xmath88 , which is a typical value used in theoretical models of accretion disks .
we also discussed how the solutions would change with different @xmath2 in 3.2 . among the most popular analytical disk models
are the ssd , adaf and slim disk models , so here we show the solutions with input parameter corresponding to them respectively . for accretion flows composed mostly of fully ionized hydrogen , the material can be regarded as monatomic ideal gas , of which @xmath89 , and that is the value we take here ( accretion flows composed of unionized @xmath90 would have @xmath91 .
we did nt discuss this situation here , but some relative results can be found in 3.2 ) .
figures 1 & 2 are both for the ssd model which has little advection ( @xmath92 ) , while figure 1 is for the gas pressure dominated region and figure 2 for the radiation pressure dominated region .
the adaf model is treated as advection dominated(@xmath93 ) , and gas pressure dominated ( @xmath94 , so that @xmath95 ) , as shown in figure 3 .
the slim disk model is treated as advection dominated(@xmath93 ) , and radiation pressure dominated ( @xmath57 , so that @xmath96 ) , as shown in figure 4 .
the other two parameters are all set as @xmath88 and @xmath85 . in each figure , we show the function curves of @xmath37 , @xmath38 , @xmath39 , @xmath40 , @xmath36 and the mach numbers of @xmath9 and @xmath6 , in which the adiabatic sound speed is calculated as @xmath97 as shown in equations ( 7 - 11 ) , all the velocities scale with @xmath98 . the density @xmath36 is scaled to 1 in the equatorial plane and it can be determined in a real case once the accretion rate at a certain radius is given .
the pressure @xmath21 is also scaled , and it satisfies the relation that @xmath99 .
so once we get the actual density @xmath67 at a certain radius , we can calculate the corresponding pressure @xmath21 easily .
the mach number curves for @xmath9 first decreases with @xmath1 , reaching 0 at some inclination , and then increases . that is because it changes sign along @xmath1 and we use absolute values in calculating the mach numbers .
there are several common features among figures 1 - 4 .
the value of @xmath6 is always non - positive .
the value of @xmath9 is always negative close to the equatorial plane , and becomes positive at smaller inclinations . in each of these figures
, there exists a certain inclination @xmath100 at which @xmath101 .
this @xmath100 has similar meaning with the @xmath100 in xw05 , except that it is obtained through calculation rather than being preset as an input parameter in xw05 . in the region
@xmath102 , @xmath37 is negative , meaning that the accretion flow is moving towards the central accretor as inflow .
so this region corresponds to the normal accretion disk part in basic models and we call it the inflow region in this paper . in the region
@xmath103 , @xmath37 is positive and the accretion flow is moving away from the central accretor , and we call this region the outflow region .
the region between the upper boundary and the vertical axis contains the outflow which blows out of the upper boundary in the form of wind as shown in figure 5 , and we call this region the wind region , to distinguish it from the outflow region .
it should be noted that our solutions actually can not describe the structure of the wind region .
however , it is natural to suppose that the flow structure in the wind region with @xmath104 is also outflow , and the outflow region ( @xmath103 ) can either be regarded as a transition region between inflow and outflow , or as part of the outflow which can still be described with the self - similar assumptions .
the physical properties shown in figures 1 & 2 agree quite well with the ssd model .
the rotation is very close to being keplerian ; the radial velocity is much smaller than the azimuthal velocity ; and the disk is geometrically thin . however , as mentioned above , our calculation shows that even in the ssd case , there exists outflow in the accretion disk .
as shown in figures 5(a ) & 5(b ) , this outflow is in the form of wind blowing out from the disk surface .
both @xmath9 and @xmath6 are subsonic in the ssd cases .
the advection dominated solutions , as shown in figures 3 & 4 , are geometrically much thicker than the solutions with low advection , and are able to describe the structure of the majority of the space , except only a small region near the vertical axis .
the rotation is sub - keplerian and the radial velocity can be comparable with the azimuthal velocity .
both the adaf and the slim disk case show strong outflows compared with the ssd cases , which can be seen more clearly in figure 5 .
@xmath6 keeps to be subsonic for both cases .
however , for the slim disk case ( figure 4 ) , @xmath9 becomes supersonic near the vertical axis , displaying a stronger outflow than the other three typical solutions .
outflow is the result of the competition among the pressure gradient , the centrifugal force and the gravitational force . as we have mentioned before , the key feature of the outflow in our work is positive @xmath9 , so we would like to investigate the properties of these forces along the radial direction .
the radial components of the pressure gradient , the centrifugal force and the gravitational force can be written respectively as @xmath105 in which @xmath106 we plot the values of @xmath107 , @xmath108 and @xmath109 for the four typical solutions in figure 6 .
the dotted lines correspond to the gravitational force which is scaled as 1 ; the dashed lines correspond to the radial component of the centrifugal force ; the dot - dashed lines correspond to the radial component of the pressure gradient ; and the solid lines correspond to the sum of the radial components of the centrifugal force and the pressure gradient , which drives the outflow .
figure 6(a ) shows the properties of gas pressure dominated ssds , while figure 6(b ) shows the properties of radiation pressure dominated ssds .
it can be seen that for both of the ssd cases , the influence of pressure gradient is very small , and almost in the entire range of the solution , the gravitational force is balanced with the sum of the centrifugal force and the pressure gradient in the radial direction .
this is consistent with the ssd model in which both the advection and the pressure gradient are small enough to be neglected .
therefore these two solutions have similar accretion flow structure and only some quantitative differences exist between each other . on the other hand , for the advection dominated cases , such as the adaf case ( figure 6(c ) ) and the slim disk case ( figure 6(d ) ) , the pressure gradient plays a more important role than the centrifugal force and drives a significant outflow as @xmath1 decreases .
the slim disk case has qualitative differences from the other three cases . in both ssd cases and the adaf case ,
the radial component of pressure gradient starts decreasing at smaller inclination angle , and to balance this effect , the disk rotates faster as @xmath1 decreases to increase the centrifugal force , as shown in figures 1 - 3 .
however in the slim disk case , the radial component of the pressure gradient keeps increasing as @xmath1 decreases .
it not only drives a much stronger outflow than the other three cases , but also reduces the required centripetal force to keep the disk rotating .
thus the disk rotates slower as @xmath1 decreases , as shown in figure 4 . as a comparison with ny95
, we also present here a solution with parameters @xmath88 , @xmath34 , @xmath110 and @xmath93 in figure 7 .
this set of parameters corresponds to @xmath111 in ny95 . in this case
, the accretion flow does nt contain outflow and the three components of velocity and the sound speed remain constant in the structure .
the curves of @xmath36 and @xmath37 do nt agree well with ny95 , however this is because ny95 also sets the boundary conditions on the vertical axis and the values of @xmath36 and @xmath37 are limited by these boundary conditions .
our solution only uses the boundary conditions in the equatorial plane , and can still reach the vertical axis in this case and has a reasonable distribution of physical properties , which further proves the applicability of our method . in this section
we show how the solutions change according to different input parameters respectively , and discuss their physical meanings .
as shown in 3.1 , the whole space is generally divided into three regions : an inflow region , an outflow region and a wind region . as the accretion flow is symmetrical both axially and under reflection in the equatorial plane
, we only need to investigate the flow structure in a quadrant of the @xmath112 plane . in this case , the inflow region starts from the equatorial plane ( @xmath65 ) to the inclination @xmath113 where @xmath101 ; the outflow region starts from the inclination @xmath113 to the upper boundary of the flow structure at @xmath114 , and beyond that is the wind region ( @xmath115 ) in which the motion of the matter does nt satisfy the self - similar assumptions .
here we use @xmath100 and @xmath68 as indicators of the disk size in @xmath1-direction , which directly reflects the size of the inflow region and the whole inflow / outflow region , respectively . in the figures showing how the properties of solutions change according to input parameters ,
each line represents a set of solutions , and we assign a unique label for each solution set for clarity .
table 1 lists all the solution set labels , their input parameters and corresponding lines at the end of 3.2 . `
1 ) solution dependence on \alpha ` the dependence of @xmath100 and @xmath68 on @xmath2 is shown in figure 8 .
first we can clearly see that over a large range of @xmath2 , the advection dominated solutions have both larger inflow region and larger inflow / outflow region . that s because in these solutions most of the viscous heating is stored in the accretion flow as internal energy , resulting in larger pressure at given density and thus larger size of the whole inflow / outflow region . for solutions with the same advective factor @xmath3
, gas pressure dominated solutions generally have larger inflow region .
the inflow region lasts from the equatorial plane to the inclination @xmath100 where @xmath9 reaches 0 , so its size depends on the behavior of @xmath9 . in the equatorial plane , the reflection symmetry gives equation ( 21 ) , and when it is substituted into equations ( 17 ) , ( 19 ) & ( 20 ) , we can get @xmath116 equation ( 30 ) describes the hydrodynamical balance in the radial direction among the pressure gradient , the centrifugal force and the gravitational force , which determines the acceleration of @xmath9 ( the term @xmath117 represents the acceleration of @xmath9 under the self - similar assumptions ) .
equation ( 31 ) describes how the viscosity affects the angular momentum transfer rate in the equatorial plane .
equation ( 32 ) describes the energy mechanism of the accretion flow that a fraction @xmath3 of the dissipated energy is advected as stored entropy .
it should be noted that @xmath9 is generally negative in the equatorial plane , so the absolute value of @xmath9 can be written as ( according to equation ( 32 ) ) @xmath118.\ ] ] if @xmath119 remains the same , then in the range of parameters we study in this paper , @xmath120 increases with @xmath55 .
however @xmath119 also changes with input parameters , and the exact value of @xmath121 should be calculated via equations ( 30)-(32 ) .
figure 10 shows the profile of @xmath121 vs @xmath55 when other parameters are taken as @xmath85 and @xmath88 .
it can be seen that for both the advection dominated solutions and the solutions with little advection , @xmath120 increases with @xmath55 . with a larger value of @xmath9 in the equatorial plane ,
the accretion flow is inclined to undergo a larger range of @xmath1 to change @xmath9 to 0 , resulting in a larger inflow region .
for the two ssd cases , as their flow structures are quite similar , larger values of @xmath9 in the equatorial plane give the gas pressure dominated solution set @xmath122 larger inflow regions than @xmath123 .
for the advection - dominated cases , not only the equatorial value of @xmath9 is smaller in the slim disk case than that in the adaf case , but its radial direction pressure gradient also increases as @xmath1 decreases , giving a larger outward acceleration to @xmath9 , thus the gas pressure dominated solutions @xmath124 have larger inflow regions than the radiation pressure dominated solutions @xmath125 .
the solution set @xmath125 , which corresponds to the slim disk model , has a large outflow size over a broad range of @xmath2 , displaying a large outflow area .
the solutions corresponding to the ssd model , @xmath122 and @xmath123 , have very small outflow regions , and their outflows are mainly in the form of wind blowing out of the disk surface instead of escaping in the outflow region .
the solution set @xmath124 , which corresponds to the adaf model , has a moderate outflow region with small @xmath2 .
however as @xmath2 increases , this outflow region quickly disappears and the outflow takes the form of wind blowing out of the upper surface .
the velocity field displayed in figure 5 is an example for @xmath88 .
@xmath2 is connected with both the viscous heating and the angular momentum transport of the accretion flow .
larger @xmath2 will increase the viscous heating for a fixed surface density .
but at the same time it also increases the angular momentum transfer rate , thus increasing @xmath9 and decreasing surface density , which decreases the total viscous heating rate . therefore the dependence of the viscous heating rate on @xmath2 , is somewhat complex .
viscous heating represents how much gravitational energy is converted to internal energy . but some of it is lost due to radiation and eventually there is only an @xmath3 fraction of the viscous heating which is converted to internal energy .
the increase in the internal energy will raise the temperature and subsequently the pressure , thus increasing the size of the accretion flow .
however , for the ssd cases shown in @xmath122 & @xmath123 , @xmath3 is very small , thus the energy converted to internal energy is negligible and the disk structure is dominated dynamically by the gravity and the rotation of the material , rather than the pressure gradient . as a result , the sizes of accretion flows change very little with different @xmath2 for @xmath122 & @xmath123 . on the other hand ,
for the advection dominated solutions @xmath124 & @xmath125 , a large fraction of the gravitational energy is eventually stored as internal energy , so that the pressure gradient plays a more important role in determining the flow structure as mentioned in 3.1 . therefore the sizes of accretion flows are closely connected with @xmath2 . in figure 8
we can see that for @xmath124 & @xmath125 , when @xmath2 is very small , the viscous process is not quite effective , therefore both the inflow region size and the outflow region size are small ( it s easy to understand if we consider the situation @xmath126 , in which case no accretion flow will be formed ) . from this small value , along with the increasing of @xmath2 , the viscous process starts to take effect and the whole inflow / outflow region size starts to increase , until it reaches a maximum size at @xmath127 for line @xmath128 and @xmath129 for line @xmath130(the exact value of @xmath2 for maximum size also depends on other parameters ) . from then on , the larger values of @xmath9 starts to dominate the total effect and the corresponding accretion flow size starts to decrease .
this property can also be seen in the top left panel of figure 17 , which displays the ratio of the outflow rate to the inflow rate .
it should be noted that , in figure 8 , the size change based on @xmath2 is different between the whole inflow / outflow region and the pure inflow region .
the size of the whole inflow / outflow region decreases after a peak at a certain @xmath2 while the size of the pure inflow region keeps increasing . for solution sets @xmath122 , @xmath124 and @xmath125 ,
this difference is enough to generate a critical @xmath131 respectively , at which the outflow region totally disappears .
so for @xmath132 , the accretion flow that we can resolve with self - similar assumptions is composed of pure inflow and we did nt find a surface at which @xmath133 .
that s why we say that the assumption used in xw05 may not be true for large @xmath2 .
kluzniak and kita ( 2000 ) also found this critical @xmath2 which was calculated to be @xmath134 by them . in our work , however , @xmath131 depends on other parameters and can change significantly , as shown in figure 8 ( @xmath135 for @xmath122 , @xmath136 for @xmath124 and @xmath137 for @xmath125 ) . for @xmath123 ,
this @xmath131 is above 1 and it s not displayed .
figures 9 displays the @xmath9 distribution along the @xmath1-direction for solutions with different @xmath2 . figures
( a ) , ( b ) , ( c ) and ( d ) correspond to the four sets of solutions in figure 8 respectively , and the solid , dashed and dotted lines correspond to different @xmath2 as indicated in the legends . the range of the axes are adjusted to show more details in comparison .
we can see that in the solutions corresponding to the ssd model , i.e. figures 9(a ) & 9(b ) , @xmath9 is almost proportional to @xmath2 .
that is because when alpha increases , so does the angular momentum transfer rate . for the ssd case , the accretion flow structure
is mostly determined by gravity and rotation as discussed in 3.1 , and the rotation is quasi - keplerian , so the increase in angular momentum transfer rate increases @xmath9 almost proportionally .
a special case in the equatorial plane is shown by equation ( 31 ) , in which when @xmath10 is keplerian , the value of @xmath9 is proportional to @xmath2 . for the adaf and the slim disk cases , because pressure gradient plays an important role in the accretion flow structure , @xmath9 is generally not proportional to @xmath2 , but still positively correlated with @xmath2 as shown in figure 9(c ) & 9(d ) .
we can also see that some lines in figure 9 do nt contain positive @xmath9 . that s because the @xmath2 values for these solutions are greater than the critical value @xmath131 mentioned above .
` 2 ) solution dependence on n ` the parameter @xmath35 describes how the density changes along the radius in our self - similar assumptions . as discussed in 3.1 , when @xmath34 , the entire flow structure can be described by the same set of self - similar solutions , but does nt contain outflow ; when @xmath82 , the solution contains outflow but the self - similar assumptions can not be applied to the region near the vertical axis ; solutions with @xmath83 are unlikely to happen in real cases .
the dependence of @xmath100 and @xmath68 on @xmath35 is shown in figure 11 .
figures 12 & 13 display the @xmath9 and @xmath6 distributions along the @xmath1-direction for solutions with different @xmath35 .
we can see that solutions all achieve a maximum size at @xmath34 for both the whole inflow / outflow region and the inflow only region .
this is because when @xmath34 the entire space can be described by the same set of self - similar equations as discussed in 3.1 . in this case ,
@xmath5 and mass is conserved on the streamlines pointing exactly at the central accretor , thus no outflow is required . before reaching the maximum value at @xmath34 , the flow size is positively correlated with @xmath35 as a general trend .
as @xmath35 decreases , according to equation ( 22 ) @xmath69 decreases faster inwards , which means a larger mass loss rate into the area beyond @xmath68 , resulting in the increase in the size of the wind region and the decrease in the size of the whole inflow / outflow region in figure 11 , and the increase in @xmath6 in figure 13 . in figure 11 line @xmath138 has 2 peaks . according to equation ( 25 ) ,
the radial pressure gradient is proportional to @xmath139 , so when @xmath140 does nt change much , the radial pressure gradient increases with @xmath35 . in advection - dominated solutions @xmath141 & @xmath142 ,
the radial pressure gradient is more important than centrifugal force in accelerating @xmath9 , so for larger @xmath35 which is inclined to have larger radial pressure gradient , one would expect to see faster increase in @xmath9 and stronger outflow in the outflow region .
these features of @xmath9 can be seen in figure 12 .
the radial pressure gradient is dominant in the slim disk case as mentioned in 3.1 , so the acceleration of @xmath9 increases significantly with increasing @xmath35 in the slim disk case , which can be seen in figure 12(d ) with a much steeper profile of @xmath9 when @xmath85 .
this increased outflow velocity in the outflow region can cause a steeper density drop along @xmath1 direction towards the upper boundary and therefore make the outflow region become smaller , creating a valley in line @xmath138 around @xmath143 .
the top right panel in figure 17 shows how the ratio of the outflow rate to the inflow rate changes depending on @xmath35 .
we can see that when @xmath35 becomes small enough , the ratio changes very little with @xmath35 . in this range of @xmath35 , the size of different regions also does nt change much in figure 10 , and the flow structure is similar to each other . for @xmath95 , we can not get a reasonable set of equatorial values for physical properties when @xmath35 is larger than 3/2 . for @xmath96 , i.e. solutions @xmath144 and @xmath142 , we can still get reasonable values of equatorial physical properties when @xmath35 is larger than 3/2 , so that we can calculate solutions with @xmath83 in this case .
solutions with @xmath83 have an effective accretion rate @xmath69 that increases towards the central accretor , which implies that there must be matter injection into the accretion flow from high latitudes . this kind of matter injection is not likely to happen in real cases , so we wo nt pay more attention to the solutions with @xmath83 here . solutions with @xmath34 and @xmath82
both exist and have quite different structure .
however , as in ny95 , solutions with @xmath34 are only for adafs , while solutions with @xmath82 in our work is applicable to a much wider range of accretion disks . in this sense
, we believe that outflows should be common in various accretion disks . `
3 ) solution dependence on f ` the dependence of @xmath100 and @xmath68 on @xmath3 is shown in figure 14 .
it has already been discussed that as the advective factor @xmath3 increases , the size of the whole inflow / outflow region also increases , which is shown more clearly here in figure 14 .
the gas pressure dominated solutions have larger inflow region in agreement with figures 8 & 11 , and the reason has been discussed in 3.2.1 . as @xmath3 increases , the size of the outflow region increases faster for radiation pressure dominated flows , displaying a larger outflow size for solutions corresponding to the slim disk model .
the reason is that , as @xmath3 increases , a larger fraction of viscous heating is converted to internal energy instead of being lost via radiation , raising the temperature of the accretion flow .
the gas pressure is proportional to @xmath145 while the radiation pressure is proportional to @xmath146 , so in the radiation pressure dominated solutions , @xmath21 increases faster as @xmath3 increases .
the increased pressure inflates the outflow region and causes a stronger outflow . on the other hand ,
as @xmath3 gets close to 1 , eventually the outflow velocity at small inclination becomes so large , that it will instead reduce the size of the outflow region because the material lost in the outflow region is limited .
as shown in the bottom left panel of figure 17 , at a certain @xmath147 , the fraction of material lost in the outflow region reaches an upper limit , and with larger @xmath3 , this fraction almost remains unchanged . in this range of @xmath3 ,
the larger outflow velocity will instead shrink the size of the outflow region .
this causes line @xmath148 in figure 14 to reach a peak of @xmath149 at @xmath147 ( corresponding to figure 17(c ) ) , and to start decreasing as @xmath3 becomes larger . `
4 ) solution dependence on \gamma_\mathrm{equ } ( and consequently \beta ) ` the dependence of @xmath100 and @xmath68 on @xmath55 is shown in figure 15 .
the top axis displays the corresponding values of @xmath43 when @xmath27 takes the value of 5/3 .
if the intrinsic heat capacity ratio is not 5/3 ( e.g. when dealing with proto - stellar accretion disks with unionized gas ) , the corresponding @xmath43 in figure 15 may not be correct , but one can still calculate the correct @xmath43 with @xmath55 and @xmath27 according to equation ( 14 ) .
figure 16 displays the @xmath9 and @xmath6 distributions along the @xmath1-direction for solutions with different @xmath55 .
the bottom right panel of figure 17 shows how the ratio of the outflow rate to the inflow rate changes depending on @xmath55 ( and consequently @xmath43 ) .
in figure 15 , the size of the pure inflow region generally increases with @xmath55 ( and consequently @xmath43 ) .
as shown in figure 10 & 16 and discussed in 3.2.1 , solutions with larger @xmath55 ( and consequently larger gas pressure ratio @xmath43 ) have larger values of @xmath9 in the equatorial plane . this effect combined with the acceleration properties of @xmath9 ,
causes solutions with larger @xmath55 ( and consequently larger @xmath43 ) to have larger inflow region . for solutions with low advection , as the outflow region is very small , the size of the whole inflow / outflow region also displays similar profile with increasing @xmath55 ( and consequently @xmath43 ) .
that s why lines @xmath150 , @xmath151 and @xmath152 all have positive slope in figure 15 . in figure 17(d ) , there exists a transition point for the advection dominated solutions , and for solutions with @xmath153 ( corresponding to @xmath154 ) , the ratio of the outflow rate to the inflow rate keeps being large . as discussed before , for the advection dominated cases , radiation pressure dominated solutions have larger pressure gradient , shown as steeper velocity profiles in figure 16 , which drives stronger outflow . as shown in figure 17(d ) , the outflow to inflow ratio increases as @xmath55 decreases from 5/3 ( which indicates the decreasing of gas pressure ratio @xmath43 from 1 ) .
however , the outflow rate can not exceed the inflow rate and eventually the increase will cease , reaching a transition point at @xmath155 ( corresponding to @xmath156 ) . for @xmath153 ( corresponding to @xmath157 ) , the outflow to inflow ratio remains almost unchanged .
further decrease in @xmath55 will continue to increase the outflow velocity , while the fraction of the inflow lost in the outflow region is almost unchanged , so the size of the outflow region will shrink .
this feature can be also seen in figure 15 , which shows that the size of the outflow region for advection dominated solutions first increases with decreasing @xmath55 , reaching an upper limit at @xmath155 ( corresponding to @xmath156 ) , and decreases with decreasing @xmath55 after that .
it should be noted that , this transition value of @xmath55 depends on other input parameters , and this value of @xmath158 corresponds to @xmath88 , @xmath85 and @xmath93 . for the solutions with low advection , the ratio of the outflow rate to the inflow rate do nt change much and are always very small as shown in figure 17(d ) . `
5 ) bends in solutions ` we have shown the solution dependence on different input parameters
. the solutions sometimes show bends which we would like to discuss more specifically here .
the bend about @xmath2 is closely connected with the energy mechanism . as discussed in
3.2.1 , the dependence of viscous heating on @xmath2 is somewhat complex .
it has a peak at a certain value of @xmath2 ( which depends on other input parameters ) , and decreases on both sides . for the advection
dominated solutions @xmath124 & @xmath125 , most of the viscous heating is stored as internal energy in the accretion flow , which raises the temperature and subsequently the pressure , thus increasing the size of the accretion flow . on the other hand , the increase in @xmath2 also increases the absolute value of @xmath9 of the inflow near the equatorial plane , in which case more work needs to be done to drive an outflow , decreasing the outflow region as @xmath2 increases .
so naturally the size of the whole inflow / outflow region also has a peak at some value of @xmath2 .
as @xmath2 increases from a very small value , at first the increase in viscous heating dominates the total effect , and the size of the accretion flow increases , until it reaches a maximum value ; from then on , the increase in @xmath9 starts to dominate the total effect , which makes the outflow more difficult to produce and also reduces viscous heating when @xmath2 becomes large enough , thus shrinking the size of the whole accretion flow . to show the energy profile for solutions with different @xmath2 , here we calculate the @xmath1-averaged bernoulli function @xmath159 .
the bernoulli function represents the specific total energy in the accretion flow .
a positive value of the bernoulli function is only a necessary , not a sufficient , condition for outflow formation ( abramowicz et al . 2000 ) .
however , as our solutions have already showed outflow structure in the velocity fields , a larger value of the bernoulli function indicates that the accretion flow is more likely to contain stronger outflow .
locally the bernoulli function can be written as @xmath160,\ ] ] in which @xmath161 is the specific enthalpy , @xmath162 is the velocity ( all three components included ) and @xmath163 is the gravitational potential .
so the @xmath1-averaged bernoulli function can be written as @xmath164 \sin\theta d\theta } { \int_{\theta_\mathrm{b}}^{90\textordmasculine } \rho(\theta)\cdot \sin\theta d\theta}.\end{aligned}\ ] ] figure 18 displays how the @xmath1-averaged bernoulli function @xmath159 changes with variant @xmath2 for the advection dominated cases .
it can be seen that the positions of the peaks agree quite well with the bends of line @xmath128 and line @xmath130 in figure 8 .
figure 19 displays the velocity fields of solutions around the bend for variant @xmath2 .
the values of @xmath2 are chosen so that the solutions on both sides of the peaks have similar upper boundaries ( @xmath165 ) .
it can be seen that the velocities in the outflow region are more aligned to the @xmath1 direction in the right panel of each row than those in the left panel . that s because larger @xmath2 increases the angular momentum transfer rate , thus increases the absolute value of @xmath9 near the equatorial plane where most of the material is accreted .
the value of @xmath9 is negative in the equatorial plane and changes gradually to positive values in the outflow region under the effect of pressure gradient and centrifugal force .
the absolute value of @xmath9 in the equatorial plane in the right panel of each row is significantly larger than that in the left panel , so when pressure gradient does nt differ much , the positive values of @xmath9 in the outflow region will be smaller for larger @xmath2 and the velocities will be more aligned to the @xmath1 direction . the solutions corresponding to slim disks ( the second row ) show velocities more aligned to the radial direction in the outflow region than those corresponding to adafs ( the first row ) , due to larger pressure gradient and thus larger values of @xmath9 in the outflow region .
the bends in solution dependence on @xmath35 , @xmath3 and @xmath55 have different physical mechanism from that on @xmath2 .
it has been mentioned that , as @xmath35 increases , @xmath3 increases or @xmath55 decreases , the pressure gradient increases and drives stronger outflow .
however , the outflow rate can not exceed the inflow rate . at seen in figure 17 , the ratio of the outflow rate to the inflow rate for lines @xmath142 , @xmath166 and @xmath152 will eventually reach an upper limit at certain values of corresponding parameters . for
solution sets @xmath166 and @xmath152 , further increase in @xmath3 or further decrease in @xmath55 will continue to increase the radial outflow velocity , while the outflow rate in the outflow region is almost unchanged , so the size of the outflow region will shrink . for the solution dependence on @xmath35 ,
the bend near @xmath167 has similar physical reason as the bends for @xmath3 and @xmath55 .
however , the size of the accretion flow has another peak at @xmath34 , where the flow is always radial ( with rotation ) and no outflow is needed .
so line @xmath142 in figure 17 starts to drop as @xmath35 gets close to 3/2 and line @xmath138 in figure 11 gets to another peak at @xmath34 .
figure 20 displays the velocity fields of solutions around the bend for variant @xmath35 , @xmath3 and @xmath55 .
it can be seen that , as @xmath35 increases ( the first row ) , @xmath3 increases ( the second row ) or @xmath55 decreases ( the third row ) , the flow patterns display similar evolvement for these parameters : at first both the accretion flow size and the radial outflow velocity increase ; after the corresponding parameter gets past the respective value at the transition point ( the middle panels in all three rows of figure 20 ) , the radial outflow velocity continues to increase while the size of the accretion flow decreases .
the velocities in the outflow regions become more aligned to the radial direction from left to right in all three rows , due to increased radial outflow velocities .
the labels for solution sets and lines which appear in the text and figures are summarized in table 1 along with their corresponding input parameters .
0.2 cm [ cols="<,^,^,^,^,^,<",options="header " , ]
with the self - similar assumptions along the radial direction and boundary conditions obtained from the reflection symmetry in the equatorial plane , we are able to solve the navier - stokes equations along the @xmath1 direction explicitly , and the velocity field is obtained .
the result shows that outflows are common in all kinds of accretion disk models , which is consistent with some numerical simulation works ( e.g. ohsuga et al.2005 ; ohsuga & mineshige 2007 ; ohsuga et al .
generally the accretion flow consists of three different regions : an inflow region near the equatorial plane which contains the largest portion of mass , an outflow region above the inflow region in which the matter starts escaping the central accretor in the @xmath31-direction , and a wind region which contains the material blowing out from the boundary of the outflow region .
the structure of the inflow and the outflow region can be resolved in our solutions , while the wind region does nt obey the self - similar assumptions and can not be resolved by our calculation .
the inflow region and the wind region are essential to form a steady accretion flow structure .
the outflow region is missing in the cases with large viscosity parameter @xmath132 .
the critical value @xmath131 depends on other parameters and is @xmath136 for adafs , @xmath137 for slim disks and @xmath135 for gas pressure dominated ssds when @xmath85 . to compare with the popular analytical models
, we calculated solutions with parameters corresponding to the ssd , the adaf and the slim disk models .
the solutions corresponding to the ssd model have relatively thin , quasi - keplerian disks with very small outflow regions , and the outflows mostly take the form of wind from the disk surface .
the solutions corresponding to the adaf and slim disk models have thick , sub - keplerian disks .
the solution corresponding to the adaf model has a moderate outflow region which shrinks quickly with the increase of @xmath2 , and the outflow mainly takes the form of wind .
however as the upper boundary of the outflow region is close to the vertical axis , eventually the outflow may take the form of bipolar jets , though we can not actually resolve the explicit structure around the vertical axis .
the solutions corresponding to the slim disk model has a large outflow region in which most of the outflow takes place , and near the upper boundary of the outflow region , the material escapes with supersonic velocity .
our calculation displays very small outflow regions for input parameters corresponding to the ssd model , and for these solutions the inflow region has similar structure with the one - dimensional ssd model .
however the accretion rate inside the disk may not be conserved along the radius due to mass lost via wind . for the adaf and the slim disk cases
, our result displays an outflow region too large to be neglected , and we think it s necessary to consider the explicit structure in the @xmath1 or @xmath4 direction to get more realistic results than the traditional analytical models .
it was proposed in several works ( gu & lu , 2007 ; jiao et al .
2009 ) that the slim disk model has an upper limit of accretion rates , above which outflows seem to be inevitable . later in another paper ( jiao & lu 2009 ) , one - dimensional steady transonic global solutions for slim disks with presumed amount of outflows were calculated , and in this case the proposed upper limit of accretion rates for slim disks can be exceeded . another paper written by takeuchi et al .
( 2009 ) used the outflow data from two - dimensional radiation - hydrodynamic ( rhd ) simulations to construct an one - dimensional steady solution , and the result shows that the emergent spectra do not sensitively depend on the amount of mass outflow .
however , these models are still one - dimensional models and they only incorporated the mass loss effect of outflow into their models , so the results still remain debatable .
we think that two - dimensional solutions considering the hydrodynamical processes and radiative transfer along both the radial and the @xmath1 direction are required to examine the topic , which we are planning to do in our future work .
we also investigated the dependence of the accretion flow structure on different input parameters respectively . @xmath3 and @xmath55 ( consequently @xmath43 ) both have some transition values as discussed in 3.2.3 and 3.2.4 , at which the outflow region reaches a maximum size .
generally when @xmath168 and @xmath169 ( for intrinsic @xmath89 ) , the accretion flow will have a large outflow region and strong outflow , as the aforementioned solutions corresponding to the slim disk model .
however , in what conditions the transition happens depends on other input parameters and it wo nt happen in case of large @xmath2 due to the existence of the critical value @xmath131 .
solutions with @xmath34 do nt contain outflow and they correspond to the solutions in ny95 , which is only applicable to adafs .
solutions with @xmath82 have the aforementioned three - region ( or two - region for @xmath132 ) structure .
solutions with @xmath83 are not likely to happen in real cases . finally , we would like to mention some caveats in our work .
the solutions obtained in this paper are steady solutions , and thus it is not guaranteed that they are stable solutions .
these solutions do not cover the whole space , so for the unresolved region near the axis , we know nothing about the detailed structure other than that it contains the material blowing out of the upper boundary in the form of wind .
our work is based on the simple @xmath170 prescription of viscosity , which is not bad according to recent mhd simulation works(hirose et al .
2009 ; ohsuga et al .
2009 ) , especially when the flow is steady . however , it may be necessary to consider other components of the viscous stress tensor for solutions with large @xmath9 .
we also used the advective factor @xmath3 to simplify the energy equation , and assumed @xmath3 to be a constant .
realistically @xmath3 should vary with both @xmath1 and @xmath31 , which could be improved by considering the details of radiative transfer .
we used newtonian gravitational potential in our work , and it needs to be corrected when applied to study the structure close to the innermost stable circular orbit . in our calculation
we did nt consider the effects of convection ( e.g. stone et al .
1999 ; igumenshchev & abramowicz 2000 ; yuan & bu 2010 ) or magnetic field ( e.g. blandford & payne 1982 ; naso & miller 2010 ) , which may be important in studying the accretion flow and the generation of outflows . finally , it is necessary to abandon the self - similar assumptions if we want to investigate the disk and outflow structure more quantitatively and precisely .
these caveats will be improved in our future work .
we thank fu - guo xie for discussions and suggestions .
we are also grateful to marek abramowicz , feng yuan , wei - min gu and li xue for helpful comments .
we also thank the referee for his / her constructive suggestions .
this work was supported by the nsfc grant ( no .
11033001 ) and the 973 program ( no .
2007cb815405 ) . abramowicz , m. a. , chen , x. , kato , s. , lasota , j .-
regev , o. 1995 , , 438 , l37 abramowicz , m. a. , czerny , b. , lasota , j. p. , & szuszkiewicz , e. 1988 , , 332 , 646 abramowicz , m. a. , lasota , j .-
p . , & igumenshchev , i. v. 2000 , , 314 , 775 blandford , r. d. , & begelman , m. c. 1999 , , 303 , l1 blandford , r. d. , & begelman , m. c. 2004 , , 349 , 68 blandford , r. d. , & payne , d. g. 1982 , , 199 , 883 chartas , g. , brandt , w. n. , & gallagher , s. c. 2003 , , 595 , 85 igumenshchev , i. v. , & abramowicz , m. a. 2000 , , 130 , 463 jiao , c .-
, & lu , j .- f . 2009 , chinese physics letters , 26 , 049701 jiao , c .-
l . , xue , l. , gu , w .- m . , & lu , j .- f .
2009 , , 693 , 670 kato , s. , fukue , j. , & mineshige , s. 2008 , black - hole accretion disks : towards a new paradigm ( kyoto : kyoto univ . press ) kawabata , r. , & mineshige , s. 2009 , , 61 , 1135 king , a. r. , pringle , j. e. , & livio , m. 2007 , , 376 , 1740 kluzniak , w. , & kita , d. 2000 , arxiv : astro - ph/0006266 naso , l. , & miller , j. c.2010 , , 521 , a31 okuda , t. , teresi , v. , toscano , e. , & molteni , d. 2005 , , 357 , 295 ohsuga , k. , mori , m. , nakamoto , t. , & mineshige , s. 2005 , , 628 , 368 ohsuga , k. , & mineshige , s. 2007 , , 670 , 1283 ohsuga , k. , mineshige , s. , mori , m. , & kato , y. 2009 , , 61 , l7 stone , j. m. , pringle , j. e. , & begelman , m. c. 1999 , , 310 , 1002 takahara , f. , rosner , r. , & kusunose , m. 1989 , , 346 , 122 takeuchi , s. , mineshige , s. , & ohsuga , k. 2009 , , 61 , 783 xie , f .-
, & yuan , f. 2008 , , 681 , 499 | in order to study the outflows from accretion disks , we solve the set of hydrodynamic equations for accretion disks in the spherical coordinates ( @xmath0 ) to obtain the explicit structure along the @xmath1 direction . using self - similar assumptions in the radial direction , we change the equations to a set of ordinary differential equations ( odes ) about the @xmath1-coordinate , which are then solved with symmetrical boundary conditions in the equatorial plane , and the velocity field is obtained .
the @xmath2 viscosity prescription is applied and an advective factor @xmath3 is used to simplify the energy equation.the results display thinner , quasi - keplerian disks for shakura - sunyaev disks ( ssds ) and thicker , sub - keplerian disks for advection dominated accretion flows ( adafs ) and slim disks , which are consistent with previous popular analytical models . however ,
an inflow region and an outflow region always exist , except when the viscosity parameter @xmath2 is too large , which supports the results of some recent numerical simulation works .
our results indicate that the outflows should be common in various accretion disks and may be stronger in slim disks , where both advection and radiation pressure are dominant .
we also present the structure dependence on the input parameters and discuss their physical meanings .
the caveats of this work and possible improvements in the future are discussed . |
quarkonia , bound states of a quark @xmath4 and its anti - quark @xmath5 , are the qcd equivalents of positronium ( bound @xmath6 ) in qed . just as we have learned much about qed from the spectra of positronium , so too we probe the strong force with @xmath7 bound states .
quarkonia form the simplest , most symmetric strongly interacting system with only two constituents ( unlike baryons ) and those being identical ( unlike mesons with `` open '' flavor ) .
the qcd potential is much richer than that of qed . in the simplest formulation , the so - called `` cornell '' potential
, we have : @xmath8 the first term mimics the single - photon exchange term in qed and corresponds to one - gluon exchange in qcd , dominating at short distances ( @xmath9fm ) with large momentum transfers .
this is the asymptotically `` free '' regime which has @xmath10 small , making calculations perturbative ( pqcd ) .
the second term , important at larger distance scales ( @xmath11 fm ) , leads to `` confinement '' and is in a regime where @xmath10 is large , making calculations non - perturbative . in this presentation
i will discuss only _ heavy _
quarkonia , @xmath12 , with @xmath13 ( charmonium ) or @xmath14 ( bottomonium ) .
the top quark is not sufficiently stable to form hadrons before decaying . ]
a big advantage of such @xmath12 systems is that they are not very relativistic : @xmath15 and @xmath16
. therefore non - relativistic qcd ( nrqcd ) can be used to good approximation , greatly simplifying calculations .
of particular note is the fact that lattice qcd ( lqcd ) is making important advances in predictive power for these systems .
further , the physical sizes of @xmath12 states span the range of `` free '' and `` confined '' , allowing both aspects of qcd to be studied . while this presentation concentrates on spectra and decays
, one can not ignore the importance of production in that much is to be learned from the formation of such @xmath12 systems from ( virtual ) photons and gluons .
the spectra for charmonia and bottomonia are shown in figure [ fig : charmbottom ] , with some typical transitions indicated .
note that the qcd wells for such states are much deeper than the corresponding case for positronium , leading to a much richer spectrum , involving many _
hadronic _ transitions in addition to those associated with photon emission .
in particular , the @xmath17 spectrum has three radial excitations below open bottom threshold and enough energy splitting to allow multi - pion or @xmath18 emission in transitions among states .
this document is a faithful replication of the pic04 presentation being a non - exhaustive _ selection _ of topics .
discoveries or updates made after pic04 will not be included ; this is therefore a snapshot of the field on 27 june 2004 .
one exception is that in the references i give the _ current _ status of the work .
of course this review builds on previous efforts , of which two good examples are the pic03 contribution of mahlke - krger @xcite and that from the 2003 lepton - photon conference by skwarnicki@xcite .
there is also a comprehensive review of quarkonia underway by the quarkonium working group@xcite ; this cern yellow report is to be completed by september 2004 .
these states have the spins of @xmath19 and @xmath20 in an anti - symmetric state , so that @xmath21 and @xmath22 .
the states with @xmath23 ( pseudoscalars ) are called `` @xmath24 '' and those with @xmath25 are called `` @xmath26 '' . for @xmath17 there is no news on these singlets ; the cleo limits@xcite from 2003 have been slightly updated@xcite but none of the five have been observed as yet . in the @xmath27 sector
these have been some new measurements of the mass , width and decays of the ground state of charmonium , the @xmath28 , but no startling developments .
the lone singlet @xmath29 state of charmonium , the @xmath30 , remains elusive , with active programs in @xmath31 production , @xmath1 decay and @xmath32 ( and maybe others ! ) all in the hunt .
the biggest news from the spin singlets is the ( re)discovery of the @xmath0 .
this had been previously reported by the crystal ball experiment @xcite in the m1 radiative transition @xmath33 .
but cleo , with similar sensitivity to the crystal ball , at @xmath34 significance . ]
does not confirm that observation.@xcite but four _ new _ observations of this radially excited singlet have now been _
published_. first , belle presented a very clean signal ( see fig .
[ fig : etacpr ] ) in @xmath35 decays @xcite ; they also observed the state in continuum production of double charmonium , @xmath36 @xcite .
this was followed by both cleo@xcite and babar@xcite publishing their measurements of the @xmath0 in two - photon fusion .
the four new measurements of the mass of the @xmath0 are shown in the right portion of fig .
[ fig : etacpr ] , presented@xcite as the hyperfine mass splitting between it and the spin - triplet @xmath1 .
the weighted average for this splitting is @xmath37 which is roughly _ half _ the magnitude for this splitting based on the older crystal ball measurement of @xmath38 , and which is to be compared with @xmath39 mev .
mev for this hyperfine splitting . ]
the splittings @xmath40 and @xmath41 are of some interest because , while the @xmath28 and @xmath42 have small radius and are rather deep in the coulomb - like qcd well , the @xmath0 and @xmath1 start to sample the confinement region of the qcd potential .
most theoretical estimates of @xmath41 are higher than the new measurements@xcite , perhaps because they tend to assume a scalar qcd potential in the confinement region .
a recent lattice result@xcite , which is quenched and thus ignores dynamical light quarks , gives a span of values of @xmath43 mev , depending on input parameters .
this situation warrants a new look at models and an unquenched lattice calculation .
the vector @xmath3 and @xmath44 resonances are directly produced by @xmath6 colliders via annihilation to a virtual photon . with large data sets recently obtained by cleo at cesr and bes at bepc ,
it is not surprising that these two experiments dominate this portion of the onia news .
good measurements of @xmath46(ns ) are particularly important .
they play a crucial role in determining the resonance widths , in that these states are significantly more narrow than the collider beam energy spread .
cleo plans to measure @xmath47 to within a few percent from its careful scans of the resonant line shapes .
learning @xmath48 to @xmath49 thus means measuring @xmath45 to similar precision .
also , @xmath45 is needed to evaluate many other important branching fractions , in that often what is really measured is the product @xmath50(ns)@xmath51(ns)@xmath52 .
cleo has both the statistical power ( over 5 million of each of the three bound - state @xmath44 resonances ) and the control of systematics to make such precise measurements of @xmath45 .
their preliminary results@xcite are @xmath45(ns ) = @xmath53 , @xmath54 , and @xmath55 for n = 1 , 2 , 3 , respectively . for the @xmath44(1s )
this agrees with the current pdg average@xcite , but the results for the two radial excitations are significantly higher than in the pdg .
these results are shown graphically in fig .
[ fig : bmumu ] .
the _ color octet _ model was developed to help explain large charmonium production rates at hadron colliders by having @xmath27 produced from a single gluon ( hence , as a color _ octet _ state ) and then perturbatively shedding a soft gluon before emerging from the interaction .
models having two gluons form the @xmath27 meson as a final state particle are called _ color singlet _ models .
a good place to test such octet@xcite and singlet@xcite models is the glue - rich environment of @xmath44 decay .
while both predict roughly equal rates for @xmath57 , the singlet momentum spectrum is much softer in that a second @xmath58 and @xmath59 must hadronize as charmed mesons .
cleo@xcite has measured @xmath60 which includes feed - down from other charmonia ( @xmath1 , ... ) ; this is consistent with the predictions of the models .
however the measured @xmath42 momentum spectrum is much too soft to support the color octet model .
further the observation of @xmath61 , with @xmath62 , and the decay rates of @xmath44 to @xmath63 and @xmath64 are above the octet model predictions . major progress has been made on a long standing puzzle in vector charmonium .
because three - gluon decay to hadrons of @xmath27 and electro - magnetic production / decay of such states via a virtual photon both depend on the wave - function overlap ( @xmath66 ) one naively expects : @xmath67 there are many complications , caveats and considerations , form factor dependences on @xmath68 , helicity conservation issues , non - relativistic effects , interference with continuum , ... ] to this `` equality '' , so one should not be too surprised at small deviations . however , two particular hadronic modes , @xmath69 and @xmath70 , both vector - pseudoscalar ( v - p ) , are blatant violators of eqn . [
eqn : rhopi ] , with limits@xcite of @xmath71 . big new data sets at bes / bepc ( 14 million @xmath1 decays and 6.4 pb@xmath72 of continuum ) and cleo - c / cesr - c ( @xmath73 million @xmath1 decays from 5.5 pb@xmath72 of luminosity and @xmath74 20 pb@xmath72 of continuum ) have come to rescue .
bes has presented evidence@xcite for the modes @xmath75 and @xmath76 in the final state @xmath77 , as depicted in their dalitz plot in fig .
[ fig : bescleopv ] .
there is a large signal in the neutral mode , giving @xmath78 and @xmath79 ( slight suppression relative to @xmath42 ) .
the charged mode has a significant signal as well ( @xmath80 ) and gives @xmath81 and @xmath82 ( heavily suppressed relative to @xmath42 @xmath83 ) .
thus , there is a large difference in these isospin - related final states .
the small continuum sample of bes does not allow for a subtraction of this possible background source .
cleo has presented@xcite results for @xmath84 for a large number of modes , including , for the first time , @xmath69 in the @xmath85 final state .
they have made a continuum _ subtraction _ , but assume no interference contribution .
after projecting the dalitz plot to obtain samples of @xmath86 and @xmath87 , they extract the yield by demanding no missing energy , as shown to the right of fig .
[ fig : bescleopv ] .
their results in the @xmath85 final state are @xmath88 and @xmath89 .
the status of this `` puzzle '' is represented pictorially in fig . [
fig : puzzle ] . some of the important features : _ ( i ) _ the @xmath69 and @xmath90 modes have now been _ measured _ ; _ ( ii ) _ the isospin violating ( and hence , electro - magnetic ) modes seem to support the `` 12% rule '' ; _ ( iii ) _ the isospin - related modes @xmath75 and @xmath76 seem quite different in their behavior ; _ ( iv ) _ the a - p states do not seem suppressed , while v - t states are suppressed by about a factor of @xmath91 ... not nearly as much as the v - p states @xmath69 and @xmath76 .
an open question is whether the suppression is due to @xmath68-@xmath92 mixing , meaning a possible enhancement in v - p decays of the @xmath3(3770 ) , by virtue of common virtual @xmath93 loops .
another interesting comparison of the @xmath42 and @xmath1 wave - functions is the relative phase for each of these to decay strongly via @xmath27 annihilation to three gluons ( `` s '' ) as opposed to electro - magnetically via a single virtual photon ( `` em '' ) .
the decay @xmath94 can proceed by either of these two routes , but @xmath95 is purely a strong decay ( su(3 ) symmetry ) and @xmath96 is purely electro - magnetic ( g - parity ) .
hence @xmath97 previously bes@xcite has measured @xmath98 . with @xmath99
@xmath1 events they have observed a large sample of mono - chromatic @xmath100 candidates from which they obtain@xcite @xmath101 = @xmath102 . combining this with various prior results on the @xmath103 and @xmath104 channels gives @xmath105 or @xmath106 .
thus , @xmath107 and @xmath108 are consistent .
bes also re - measured@xcite @xmath109 from their 55 million @xmath42 sample as @xmath110 , meaning @xmath101/ @xmath109 = @xmath111 , somewhat large for the `` 12% rule '' represented by eqn .
[ eqn : rhopi ] .
bottomonium ( see fig . [ fig : charmbottom ] ) is the only qcd system with a stable member that has two units of orbital angular momentum .
this makes these `` d '' states important to test spin - orbit and high - l effects in both models and lqcd . the cleo analysis of the four photon cascade @xmath44(3s)@xmath112 `` d '' @xmath113 is now final@xcite .
a single state is observed at @xmath114 mev at @xmath115 significance .
the decay rate and intermediate @xmath116 and @xmath117 assignments are consistent with this being the @xmath118 state .
the impact of this result on lqcd@xcite is shown in the left portion of fig .
[ fig : lqcd ] . plotted first results for nine `` golden '' quantities from analyses that were `` quenched '' , thus not incorporating the effects of dynamical light quarks .
next to this are recent _ un_quenched results with @xmath119 quarks included .
none of the nine quantities are used as `` input '' .
these show lqcd being able to attain better than 5% accuracy in these quantities , one of which is the d - s splitting in bottomonium . turning to charmonium
, there is the continuing puzzle of the nature of the @xmath3(3770 ) , which one can characterize as being @xmath120 .
clearly @xmath121 in that the @xmath3(3770 ) _ is _ directly produced in @xmath6 annihilation .
belle@xcite has recently observed a solid @xmath3(3770 ) signal in @xmath122 , from which they extract @xmath123 .
this comparison would imply that there is large s - d mixing , although color - octet models might be able to accommodate this rate for a pure `` d '' state ( @xmath124 ) . an important mode to investigate would seem to be @xmath3(3770)@xmath125 . a compilation of markii and bes results@xcite gives @xmath126 kev , which is near the _ upper bound _ for this rate as analyzed by cleo@xcite .
the kuang - yan prediction@xcite for this rate is 20 - 107 kev , depending on the level of s - d mixing ; however , the cleo limit@xcite , depicted in fig .
[ fig : lqcd ] , for @xmath44(1d)@xmath125 , is some seven times below@xcite that predicted for a such a d state in the kuang - yan model , casting doubt on whether the @xmath127 measured for the @xmath3(3770 ) is _ really _ consistent with theoretical expectations .
once the decay to @xmath128 is solidly observed , the angular distributions of the decay products should help sort out@xcite @xmath129 .
although copiously produced in e1 transitions from the vector mesons directly obtained in @xmath6 annihilation , rather little is known about these @xmath131 and @xmath117 states .
cleo has finalized@xcite its analysis of @xmath132(1s ) for @xmath133 .
there is so little @xmath19 value for this decay that the process is kinematically forbidden for j=0
! nonetheless , the quoted branching fractions are large : @xmath134 and @xmath135 , to be compared to the @xmath136 branching fraction for the e1 photon transitions @xmath137(1s ) . the rates for this @xmath18 transition are therefore roughly equal for @xmath138 and @xmath139 , as predicted@xcite .
the bes collaboration has used its @xmath131 sample to test the color - octet model in baryonic decays .
the color singlet model is unable to generate the observed rates for , _
@xmath140 . adding in color octet contributions@xcite
predicts that the ratio @xmath141 is @xmath142 and @xmath143 for @xmath63 and @xmath64 decays , respectively .
the bes collaboration @xcite last year reported on the @xmath144 channel ; now they have very clean signals for all three @xmath145 states decaying into @xmath146 .
these lead to experimental values of @xmath147 and @xmath148 for @xmath138 and @xmath139 . while the uncertainties are large , they nonetheless tend to show the @xmath144 channel enhanced , not suppressed , with respect to @xmath146 .
the hottest onia news at last year s lepton - photon meeting@xcite was the observation ( @xmath150 ) by belle@xcite of a new `` charmonium - like '' state at 3872 mev produced in @xmath151 and decaying to @xmath152 . since then
, it has been seen clearly by cdf@xcite in hadro - production ( as reported at qwgii@xcite and shown in fig .
[ fig : x3872 ] ) as well as by d0@xcite and babar@xcite . of particularly note is the nature of the @xmath153 system in the @xmath149 decay .
belle@xcite reported a dipion mass ( @xmath154 ) consistent with the decay @xmath155 , which would imply that @xmath156 .
this tendency toward large @xmath154 was actually _ used _ by cdf and d0 to enhance their signals .
[ fig : x3872 ] also shows the @xmath154 spectrum as reported by babar@xcite for the @xmath149 and for their large @xmath1(3686 ) sample .
is the high - mass region _ really _ a @xmath157 or is it just a mimic of the similar shape in @xmath1 decay ?
is there a second , low - mass peak in the @xmath154 spectrum , similar to what has been observed@xcite in @xmath44(3s)@xmath158(1s ) ?
do experiments see the more difficult channel @xmath159 , which would imply @xmath160 ?
one should keep watching this di - pion system as the sample sizes increase with more data at the tevatron and the @xmath161-factories .
d0@xcite has looked at six aspects of the production and decay of the @xmath149(3872 ) as compared to the @xmath1(3686 ) : @xmath162 to the jet , range of rapidity , isolation , decay length distribution , helicity angles of the muons and of the pions . in all cases
the @xmath149 looks like the well - established @xmath27 state .
the four - experiment average has @xmath163 mev , tantalizingly close to the mass of a @xmath164 pair at@xcite @xmath165 mev but significantly below the corresponding charged system , with a mass of a @xmath166 pair being @xmath167 mev .
this `` coincidence '' , coupled with the fact@xcite that @xmath168 leads to the conjecture that the @xmath149(3872 ) has a @xmath27 `` core '' and a large @xmath164 `` molecular '' component .
there has been a lot of activity ( with a lot more ongoing ! ) to determine the quantum numbers of this new state .
searches by belle@xcite for the decays @xmath169 , @xmath170 , and @xmath171 have shown it unlikely that @xmath149(3872 ) is @xmath118 , @xmath172 , or @xmath173 , respectively .
other studies of bes and cleo data@xcite in radiative return and two - photon fusion production tend to disfavor vector states ( @xmath174 ) and those with even @xmath145 and positive @xmath175-parity .
this is certainly a most curious state and hopefully much more will be known about it by pic2005 !
as is evident , heavy quarkonia continue to be a source of great energy and excitement , with many recent advances and many open questions : a stable `` d '' state has been firmly established and agrees nicely with the lqcd prediction ; the long - awaited @xmath176 has also been firmly established with a small hyper - fine mass splitting to the @xmath1 ; several analyses are confronting the color octet production model ; v - p states have finally been observed in @xmath1 decay with suppressions relative to @xmath42 decay of roughly a factor of 50 ; while more is being learned about the nature of the @xmath3(3770 ) , it still remains a puzzle ; even more of a mystery is the very narrow state at 3872 mev !
lots yet to do ! 99 h. mahlke - krger ( pic03 ) , _ heavy quarkonia spectroscopy and decays _ , + hep - ph/0309015 . t. skwarnicki ( lp03 ) , _ heavy quarkonium _ ,
a*19 * , 1030 ( 2004 ) .
the quarkonium working group ( qwg ) web site is www.qwg.to.infn.it . | with new data sets , recently completed analyses and renewed interest , there has been significant progress in addressing existing questions about quarkonia production and decay ... but also new questions and new confrontation with theory .
some highlights include the firm establishment of the @xmath0 and of the `` missing '' @xmath1 decays to @xmath2 final states , improved information on the nature of the @xmath3(3770 ) , and the observation of a new , puzzling charmonium - like state at 3872 mev .
= 14.5pt |
as illustrated in the title of this conference , _ from 30-doradus to lyman - break galaxies _ , the term _ lyman - break galaxy _ ( lbg ) is almost synonymous with _
high redshift starburst galaxy_. however , as has been stressed by many authors , including the lead researchers behind the lyman - break technique , current samples of lbgs consist of starbursts that are extremely luminous in the uv and do not give a complete census of all starbursts at high redshifts . the current magnitude limit in the ground based surveys for lbgs is r(ab)@xmath4 ( steidel et al .
the luminosity function of lbgs has been extended to r(ab)@xmath5 based on data from the hubble deep fields ( e.g. , adelberger & steidel 2000 ) .
the faint end of this lbg luminosity function is very steep , with a slope @xmath6 , implying that more than 70% of the light is emitted by galaxies fainter than r(ab)@xmath4 .
furthermore , as discussed at this conference , the most vigorous starbursts at high redshifts , e.g. as observed with scuba or with spitzer , are often obscured in the rest - frame uv ( e.g. , chapman et al .
2004 ) and often hence do not fulfill the selection criteria for lbgs .
how is it then possible to locate and examine starbursts at high redshifts that are missed by the lyman - break technique ?
one other method , not mentioned at this conference , is absorption selection of galaxies .
the few galaxy counterparts that so far have been identified for damped ly-@xmath0 absorbers ( dlas ) , found in qso spectra , appear to be starburst galaxies with significantly smaller luminosities than lbgs ( e.g. , mller et al .
2004 ; weatherley et al . 2005 ) .
however , the total cross section of dlas at @xmath3 is much larger than what can be accounted for by lbgs .
this implies that most of the neutral gas available for star formation at these redshifts is located in galaxies fainter than the lbg flux limit ( fynbo et al .
other papers in these proceedings discuss the dust emission selected galaxies . here
we will discuss _ i ) _ selection of ly-@xmath0 emitting starbursts through deep narrow band imaging , and _
ii ) _ localisation of gamma - ray burst ( grb ) host galaxies .
the idea to use ly-@xmath0 to search for primordial galaxies dates back to the 1960s ( partridge & peebles 1967 ) .
the first detection of redshifted ly-@xmath0 emission from galaxies not powered by active galactic nuclei ( agn ) were serendipitous discoveries resulting from searches for galaxy counterparts of qso absorbers such as dlas and lyman - limit systems ( lowenthal et al .
1991 ; mller & warren 1993 ; francis et al . 1996 ; fynbo et al . 2001 ; francis et al .
other ly-@xmath0 emitters were serendipitously discovered during searches for intra - cluster planetary nebulae ( kudritzki et al . 2000 ) .
this curiosity reflects the fact that for many years it was considered impossible to locate galaxies by their ly-@xmath0 emission as the probability for ly-@xmath0 photons to be absorbed by dust , due to resonant scattering , is much larger than for continuum photons .
the first dedicated search for ly-@xmath0 emitters with 810 m class telescopes was conducted at the keck telescope ( hu et al .
1998 ) . in 2000 , some of the authors started the program `` building the bridge between damped ly-@xmath0 absorbers and lyman - break galaxies : ly-@xmath0 selection of galaxies '' at the european southern observatory s very large telescope ( vlt ) .
this project is an attempt to use ly-@xmath0 selection to bridge the gap between absorption- and emission - line selected galaxies by characterisation of @xmath7 galaxies , possibly corresponding to the abundant population of faint ( @xmath8 ) galaxies associated to dlas ( fynbo et al .
1999 ; haehnelt et al .
the survey consists of very deep narrow band observations of three fields at @xmath9 , @xmath10 , and @xmath11 . in each of these fields ,
we have detected and spectroscopically confirmed @xmath12 20 ly-@xmath0 emitters , or legos ( acronym for _ ly-@xmath0 emitting galaxy - building object _ ) . in fig .
[ examples ] we show six examples of legos from the @xmath9 field . of the total sample ,
85% are fainter than the flux limit for lbgs .
furthermore , as only @xmath1325% of the lbgs have ly-@xmath0 emission lines with equivalent widths large enough to fulfill our selection criterion for legos ( shapley et al .
2003 ) , it is clear that lbgs and legos are almost disjunct classes of high redshift starbursts .
ly-@xmath0 emitters from the `` building the bridge survey '' ( see fynbo et al .
2003a for more information including spectra of the sources ) .
the size of each image is @xmath14 arcsec@xmath15 .
the upper row shows narrow band images of the galaxies , middle row the b - band images and the bottom row the r - band images . about @xmath16% of the legos
remain undetected in the broad bands despite @xmath17 detection limits of @xmath18 and @xmath19 .
, title="fig:",scaledwidth=85.0% ] -0.2 cm given that most legos are extremely faint , it is very difficult to establish any property beyond the ly-@xmath0 flux for individual galaxies , even with 810 m class telescopes .
for this reason , we decided to observe a field with existing , very deep broad band observations covering most of the electromagnetic spectrum , namely the goods field south ( giavalisco et al .
2004 ) . in 2002
we obtained observations of a section of the goods field south . due to bad weather
, we did not reach the same flux limit as for the `` building the bridge '' fields , but still we detected nearly 20 candidate @xmath11 legos in the field .
the analysis of the broad band properties of these candidates constitute the thesis work of two of the authors ( b. krog and k. nilsson ) . here
we report a few preliminary results .
so far the objects have been studied in x - rays ( chandra x - ray observatory , 1 ms exposure ) , near - ir ( vlt / isaac ) , and the optical broad bands ( hst / acs ) . the hst images ( fig .
[ goods ] ) confirm that the legos have extremely faint continua in the range v(ab)@xmath2029 .
furthermore , _
i ) _ these galaxies are bluer than most lbgs , with spectra that rise toward the blue ( @xmath21 ) implying younger ages and/or lower dust content than what is typical for lbgs , and _
ii ) _ they have extremely compact morphologies and sometimes consist of several knots similar to the lbgs ( e.g. , lowenthal et al . 1997 ) .
only upper limits were found in the x - ray and near - ir images so we can exclude that the galaxies harbour agn and significant older populations of stars .
we are currently working on deriving stronger constraints from stacking of the individual sources .
arcsec@xmath15 around the positions of 15 of the candidate @xmath11 legos in part of the goods field south .
the v(ab ) magnitudes range from @xmath1226 to @xmath2229 .
, title="fig:",scaledwidth=90.0% ] -0.2 cm
grbs are short , extremely energetic bursts of @xmath23-rays associated with energetic core - collapse supernovae ( hjorth et al .
if the grb rate is directly proportional to the ( massive ) star formation rate , then the properties of grb hosts should reflect the diversity of all star - forming galaxies in terms of luminosity , environment , internal extinction and star formation rate .
grb hosts therefore constitute a central clue for our understanding of galaxy formation and evolution ( e.g. , ramirez - ruiz et al .
2002 ; tanvir et al .
2004 ; see also the papers by tanvir et al . and trentham et al . in these proceedings ) .
the prompt burst of @xmath23-rays is followed by a so - called afterglow , emitting over a very wide spectral range from radio through the optical / near - ir to x - rays ( see van paradijs et al .
2000 for a review ) .
the afterglow can be extremely bright , allowing a precise localisation on the sky .
more importantly , spectroscopy of the afterglow can give information about the redshift and the physical conditions within the host , and in intervening absorption systems along the line of sight ( see , e.g. , vreeswijk et al .
2004 and jakobsson et al .
2004a for examples ) .
when the grb has faded , the host galaxy itself can be observed .
the measured redshifts for grbs cover a very broad range from @xmath24 to @xmath25 with a median around @xmath26 ( e.g. , jakobsson et al .
2004a ) .
, r(ab)@xmath27 ) , 011211 ( @xmath28 , r(ab)@xmath29 ) , and 021004 ( @xmath30 , r(ab)@xmath31 ) .
the contours show the morphology of ly-@xmath0 emission , as measured by ground - based narrow band imaging ( fynbo et al .
2002 ; fynbo et al .
2003b ; jakobsson et al . in prep ) .
grb 000926 occurred near the centre of the right - most knot .
the locations of grb 011211 and grb 021004 are marked with crosses .
the field sizes for the grb 011211 and grb 021004 images are the same .
, title="fig:",scaledwidth=35.1% ] , r(ab)@xmath27 ) , 011211 ( @xmath28 , r(ab)@xmath29 ) , and 021004 ( @xmath30 , r(ab)@xmath31 ) .
the contours show the morphology of ly-@xmath0 emission , as measured by ground - based narrow band imaging ( fynbo et al .
2002 ; fynbo et al . 2003b ; jakobsson et al . in prep ) .
grb 000926 occurred near the centre of the right - most knot .
the locations of grb 011211 and grb 021004 are marked with crosses .
the field sizes for the grb 011211 and grb 021004 images are the same .
, title="fig:",scaledwidth=26.7% ] , r(ab)@xmath27 ) , 011211 ( @xmath28 , r(ab)@xmath29 ) , and 021004 ( @xmath30 , r(ab)@xmath31 ) .
the contours show the morphology of ly-@xmath0 emission , as measured by ground - based narrow band imaging ( fynbo et al .
2002 ; fynbo et al .
2003b ; jakobsson et al . in prep ) .
grb 000926 occurred near the centre of the right - most knot .
the locations of grb 011211 and grb 021004 are marked with crosses .
the field sizes for the grb 011211 and grb 021004 images are the same .
, title="fig:",scaledwidth=26.7% ] -0.2 cm how do grb hosts compare to lbgs and legos ? so far
, the redshift has been measured for 10 @xmath2 grbs .
hst - images of the three brightest of these are shown in fig .
[ hosts ] . for the remaining seven ,
the host is either undetected to faint magnitude limits or have a magnitude fainter than r(ab)@xmath20 , so the majority of these grb hosts are fainter than lbgs .
remarkably , all current evidence is consistent with the conjecture that grb hosts are ly-@xmath0 emitters ( fynbo et al .
2003b ) . as shown in fig .
[ hosts ] , the three brightest high redshift grb hosts are ly-@xmath0 emitters , and for the remaining seven , ly-@xmath0 emission has either been detected ( 2 hosts ) or not yet searched for to sufficient depth to allow detection of even a large equivalent width emission line ( 5 hosts ) .
taken at face value , this suggest that faint , lego - like galaxies in total account for the majority of the star formation at these redshifts
. however , there could be other explanations for why grbs have , so far , mainly been localised in such galaxies ( fynbo et al .
in particular , some of the so - called dark bursts could be located in more massive and dust - obscured galaxies ( e.g. , tanvir et al 2004 ; jakobsson et al .
2004b and references therein ) . the recently launched swift satellite ( gehrels et al .
2004 ) offers a unique chance to resolve this issue .
in conclusion , surveys for legos and for grb host galaxies reveal that a major fraction of the starburst activity at @xmath2 may be located in galaxies fainter than the flux limit of the lbg survey . we have here shown that ly-@xmath0 emission and grb selection are two viable methods to probe this population of faint starbursts at high redshifts .
we acknowledge benefits from collaboration within the eu fp5 research training network `` gamma - ray bursts : an enigma and a tool '' .
this work is supported by the danish natural science research council ( snf ) .
1 adelberger , k.l . & steidel , c.c . ,
apj , 544 , 218 ( 2000 ) chapman , s. , et al . , apj , 611 , 732 ( 2004 ) francis , p. , et al . ,
apj , 457 , 490 ( 1996 ) francis , p. , et al . , apj , 614 , 75 ( 2004 ) fynbo , j.p.u . ,
et al . , mnras , 305 , 849 ( 1999 ) fynbo , j.p.u . ,
a&a , 374 , 443 ( 2001 ) fynbo , j.p.u . ,
et al . , a&a , 388 , 425 ( 2002 ) fynbo , j.p.u . , et al . ,
a&a , 407 , 147 ( 2003a ) fynbo , j.p.u . ,
et al . , a&a , 406 , l63 ( 2003b ) gehrels , n. , et al . ,
apj , 611 , 1005 ( 2004 ) giavalisco , m. , et al . , apj , 600 , l93 ( 2004 ) haehnelt , m. , et al . ,
apj , 534 , 594 ( 2000 ) hjorth , j. , et al . ,
, 423 , 847 ( 2003 ) hu , e. , et al . , apj , 502 , l99 ( 1998 ) jakobsson , p. , et al . , a&a , 427 , 785 ( 2004a ) jakobsson , p. , et al . ,
apj , 617 , l21 ( 2004b ) kudritzki , r .-
, et al . , apj , 536 , 19 ( 2000 ) lowenthal , j.d . ,
et al . , apj , 377 , 73 ( 1991 ) lowenthal , j.d . , et al . , apj , 481 , 673 ( 1997 ) mller , p. & warren , s.j . ,
a&a , 270 , 43 ( 1993 ) mller , p. , et al . ,
a&a , 422 , l33 ( 2004 ) van paradijs , j. , et al . ,
ara&a , 38 , 379 ( 2000 ) partridge , r.b .
& peebles , p.j.e . ,
apj , 147 , 868 ( 1967 ) ramirez - ruiz , e. , et al . ,
mnras , 329 , 365 ( 2002 ) shapley , a. , et al .
, apj , 588 , 65 ( 2003 ) steidel , c.c .
, et al . , apj , 592 , 728 ( 2003 ) tanvir , n. , et al . ,
mnras , 352 , 1073 ( 2004 ) vreeswijk , p. , et al . ,
a&a , 419 , 927 ( 2004 ) weatherley , s. , et al .
, to appear in mnras ( 2005 ) | selection of starbursts through either deep narrow band imaging of redshifted ly-@xmath0 emitters , or localisation of host galaxies of gamma - ray bursts both give access to starburst galaxies that are significantly fainter than what is currently available from selection techniques based on the continuum ( such as lyman - break selection ) .
we here present results from a survey for ly-@xmath0 emitters at @xmath1 , conducted at the european southern observatory s very large telescope .
furthermore , we briefly describe the properties of host galaxies of gamma - ray bursts at @xmath2 .
the majority of both ly-@xmath0 and gamma - ray burst selected starbursts are fainter than the flux limit of the lyman - break galaxy sample , suggesting that a significant fraction of the integrated star formation at @xmath3 is located in galaxies at the faint end of the luminosity function .
galaxies : high redshift |
the factorization approximation is widely used in the heavy quark physics . as a calculational scheme ,
it is very convenient and there are numerous models based on it .
however , the theoretical basis for the approximation itself is not completely understood . the usual motivation for it is a color transparency argument due to bjorken @xcite which suggests that the quark pair produced in the decay does not have enough time to evolve to the real - size hadronic entity , being a small size bound state with a small chromomagnetic moment .
that is taken to imply that the qcd interaction is suppressed .
for the heavy - to - heavy - light transition ( we shall hereafter refer to the color allowed decay @xmath0 for the sake of clarity ) the matrix elements in question are @xmath1 here the usual notations are used @xmath2 where we have explicitly written out color indices . in the usual factorization approximation one
simply separates out currents by inserting the vacuum state and disregards any qcd interactions between them on the basis of bjorken s conjecture .
that implies that the @xmath3 contribution in ( [ firfor ] ) is dropped , as it has no factorizable color singlet form .
new non - factorizable corrections appearing in the @xmath3 matrix element can arise if a hard or soft gluon is emitted or absorbed by the escaping pair .
obviously , a more quantitative estimate of these contributions is needed .
the first attempt to put a factorization approximation on the solid theoretical base was made in @xcite .
the authors considered a limit of large energy transferred to light quarks @xmath4 as @xmath5 . here
@xmath6 is the velocity of the decaying heavy quark @xmath7 and @xmath8 is the momentum transferred to the light quark pair . in this limit
the energy scales as a heavy quark mass . proposing a large energy effective theory ( leet ) as an extension of the heavy quark effective theory and making use of a convenient gauge condition @xmath9 ( with @xmath10 being a null vector , @xmath11=0 ) they proved factorization for the physical amplitude dominated by collinear quarks .
it is not clear , however , how the pair of collinear quarks hadronizes into the pion .
also , the validity of the leet itself is not yet established .
the important step towards the sought theoretical description was made in @xcite .
the authors considered a specific small velocity limit where @xmath12 which implies different scaling of ( [ energy ] ) : @xmath13 as @xmath5 .
it turns out that this limit is theoretically `` clean '' for the application of the qcd sum rule method and provides a theoretical justification of the `` rule of discarding @xmath14 corrections '' @xcite on dynamical grounds .
unfortunately , the actual @xmath15 decays are at the borderline of the formalism , bringing large uncertainties into the estimation of the matrix elements . in what follows
we develop a qcd - based model which works in the limit @xmath16 as @xmath5 .
the idea is to study the propagation of the light quarks emerging to form the final hadron while passing through the ( perturbative or non - perturbative ) gluonic fields present in the decay .
interestingly , results can be obtained which are independent of the nature of the gluonic fields .
the light quark pair , produced at @xmath17 , interacts strongly with the background gluonic field while escaping from the point they were produced and hadronizes into the pion . since the energy transferred to the quark pair is large it is not necessary that the hadronization occurs at @xmath18 . as a result , it is not sufficient to take the operators of the lowest dimension in the operator product expansion ( ope ) but rather one must sum up a whole series of operators .
the novel technical point of our method is to use introduce `` generating ( distribution ) functions '' which incorporate this infinite series of matrix elements , and then to use heavy quark symmetry to restrict the form of the matrix elements enough that useful statements can be made . in principle , these functions could be modeled or perhaps fixed by other measurements . however , at the leading order in @xmath19 the color octet matrix element vanishes due to cancellations and factorization holds .
the method should be able to be extended beyond leading order , allowing one to also see corrections to factorization .
the amplitude for the singlet or octet part is with @xmath21 being the appropriate color matrix .
the quantity @xmath22 describes the transition of the light quarks into a pion in the presence of the gluonic fields which are either internal to the b , d hadrons or produced during the transition .
the idea is to treat these gluonic fields as external fields with respect to the light quark pair produced in the decay of virtual @xmath23-meson . from the diagrams fig.1
one may write down the expression for one gluon matrix element of the produced quark pair in this general background field .
this gives a contribution to the octet matrix element only : where we set @xmath25 = the external field .
as one can see , in this approximation we neglect higher fock state contributions .
our normalization is such that @xmath26 .
thus we can rewrite the equation ( [ onegluon ] ) as @xmath27 with non - local correlators @xmath28 shifting the @xmath29-dependence of the second correlator @xmath30 we shall work in the fock - shwinger gauge @xcite which gives us the advantage of expressing the @xmath31 in terms of gauge - invariant quantities ( such as the gluon stress tensor @xmath32 ) : @xmath33 we treat the light quarks as massless , so the bare propagator of a light quark is given by @xmath34 this allows us to rewrite ( [ onegl ] ) as the correlator @xmath36 entering the expression above is non - perturbative and nonlocal but can be related to the wave function of the forming pion . following @xcite
we expand it near the light cone ( @xmath37 ) .
the general form of expansion is with @xmath39 being the full set of dirac matrices .
we would like to note , however , that since @xmath40 ( or @xmath41 ) contributes to the trace , only @xmath42 and @xmath43 terms enter the final expression for @xmath44 . in eq .
( [ expansion ] ) , the non - gauge invariant terms are exactly zero by the fs gauge condition ( [ amu ] ) which allows us to use analysis similar to @xcite .
moreover , in this gauge @xmath45 in ( [ expansion ] ) .
thus , parameterizing unknown matrix elements in the conventional form @xmath47 where @xmath8 is a momentum of the pion we arrive at the simple form for the @xmath48-correlators @xmath49 with the function @xmath50 introduced through its moments @xmath51 the zeroth moment gives the overall normalization of @xmath50 : @xmath52 with @xmath53 being the decay constant of @xmath54-meson . inserting ( [ gcorr ] ) into
the ( [ onegl ] ) we obtain @xmath55 this gives for the octet amplitude of @xmath56 transition : @xmath57 b ( 0 ) | b
\rangle = \nonumber \\ \frac{i g_s}{\pi^2 } \int dz d^4 x \frac{\phi(z)}{x^4 } e^{izxq } \bigl [ x_\mu q_\nu - ( x \cdot q ) g_{\mu \nu } - i \epsilon_{\mu \nu \alpha \beta } x_\alpha q_\beta \bigr ] \langle d | \bar c ( 0 ) \gamma_\mu a_\nu ( x ) b ( 0 ) | b \rangle - \nonumber \\ \frac{i g_s}{\pi^2 } \int dz d^4 x \frac{\phi(z)}{x^4 } e^{i(1-z)xq } \bigl [ x_\mu q_\nu - ( x \cdot q ) g_{\mu \nu } + i \epsilon_{\mu \nu \alpha \beta } x_\alpha q_\beta \bigr ] \langle d | \bar c ( 0 ) \gamma_\mu a_\nu ( x ) b ( 0 ) | b \rangle \end{aligned}\ ] ] where @xmath58 and @xmath59 can be written as @xmath60 where @xmath10 is a null vector ( as in the approximation of collinear u and d - quarks ) and @xmath61 is energy transferred to the light quark system ( it is large in the kinematical limit under consideration ) .
observe that contrary to @xcite there are terms in ( [ octet ] ) which are not proportional to @xmath62 . ] .
this amplitude is also nonlocal , so we use the same idea as in eq .
( [ expansion ] ) expanding it around @xmath37 and introducing `` generating functions '' .
it is clear that the only matrix element to be parameterized is @xmath63 in fs gauge one can expand @xmath64 about @xmath65 using `` gauge invariant '' decomposition @xcite @xmath66 this implies the matrix elements @xmath67 b ( 0 ) | b \rangle = \right . & \nonumber \\ & \left .
\frac{i}{2 g_s } \sum_{n=0}^{\infty } \frac{1}{(n+2)n ! } x^{\nu_1} ...
x^{\nu_n } x^{\rho } m_{\mu \nu_1 ...
\nu_n \rho \nu } \right . & \end{aligned}\ ] ] in order to parameterize the matrix elements appearing in the expression above we note that there are a total of @xmath68 indices and that the whole expression is anti - symmetric with respect to @xmath69 interchange .
also , there are only two vectors ( or their combinations ) available , so we choose them to be @xmath70 and @xmath8 .
this leads to @xmath71 with @xmath72 and @xmath73 being unknown coefficients . performing the contraction and dropping terms proportional to @xmath74 and @xmath75 we have for the first matrix element in ( [ octet ] ) @xmath76 note that in this limit the dependence of the matrix element on @xmath73 cancels out .
these sums can be parameterized in terms of some distribution function @xmath77 . using the binomial formula , we obtain @xmath78 where @xmath79 is a binomial coefficient .
this gives for the matrix element @xmath80 \sum_n \frac{(\nu p_b x + qx)^n}{(n+2)n ! }
f ( n , \nu)\end{aligned}\ ] ] performing resummation with respect to @xmath10 we arrive at @xmath81 this results in @xmath82 \int d^4 x \frac{x_\mu x_\nu}{x^4 } e^{i ( \alpha \nu p_b + ( \alpha + z ) q)x}\end{aligned}\ ] ] similarly , for the second matrix element one obtains @xmath83 \int d^4 x \frac{x_\mu x_\nu}{x^4 } e^{i ( \alpha \nu p_b + ( 1 + \alpha + z)q)x}\end{aligned}\ ] ] at this stage we have classified the infinite number of the unknown matrix elements in terms of a few continuous functions .
the strategy at this point in general would be to consider the structure of these functions and make reasonable smooth guesses as to their form .
for example @xmath84 is related to wavefunction used in the studies of the perturbative limit of the pion form factor and has the asymptotic form @xmath85@xcite .
however , we find that an important cancellation occurs which is independent of the form of these structure functions .
taking @xmath86 , @xmath87 and integrating over @xmath29 we obtain for the @xmath88 : ] ] at the leading order in @xmath74 for any @xmath90 and @xmath91 .
this implies that the semi - soft - one - gluon matrix elements do not bring new contributions to the factorization approximation result in the kinematical limit under consideration .
we would like to point out that the two - pseudoscalar final state is somewhat special : for instance , @xmath92 does not fall into the same category - there is a good chance that under the similar conditions there is no complete cancellation of the diagrams .
this is due to the fact that there can be extra terms in ( [ param ] ) that include polarization vectors of @xmath93 meson . in conclusion ,
we have constructed a qcd - based model for the factorization approximation in the limit @xmath94 . in the framework of the model the complete cancellation of the non - factorizable terms
is observed at the leading order in @xmath95 .
the ope is resummed by introducing two generating ( distribution ) functions that can be independently parameterized .
the model gives a consistent way to estimate non - leading ( non - factorizable ) contributions . | we propose a qcd - based model for calculation of the non - perturbative corrections to the factorization approximation in the decays of heavy mesons . in the framework of the model ,
factorization in pseudoscalar transitions holds exactly at the leading order leaving the opportunity to calculate non - leading corrections consistently . |
bhabha scattering is one of the theoretically best studied scattering processes at @xmath4 colliders and can also be measured with a high precision .
the accuracy of the monte - carlo programs developed originally for physics at lep is about @xmath5 , and with a complete two - loop calculation one may reach @xmath6 .
the latter number is indicative of efforts for the international linear collider ( ilc ) , here especially in the gigaz option running at the @xmath7 boson resonance , but also for meson factories running at much smaller energies of about 1 or 10 gev .
recent years brought considerable progress in the determination of the virtual nnlo corrections .
the virtual @xmath8 contributions to the massless differential bhabha cross section have been determined in @xcite . shortly after , this result was used for deriving the @xmath9 ( @xmath10 ) corrections to massive bhabha scattering in @xcite .
the missing photonic correction terms of order @xmath11 were derived , also from @xcite , in @xcite .
the virtual corrections from fermion loop insertions , including the corresponding double - box diagrams , could not be covered by that method . for @xmath12 , the case of only electron loops ,
the corresponding diagrams were evaluated analytically in @xcite , and the net @xmath12 cross section in @xcite . at this stage the numerically most important two - loop corrections were known . for a complete treatment one needs additionally the @xmath13 two - loop corrections with heavy fermion insertions , including the hadronic corrections which replace the loop insertions from light quarks .
the leptonic @xmath13 contributions have been derived quite recently in two papers in the limit @xmath14 ; with a direct feynman diagram calculation in @xcite , and using a factorization formula that relates massless and massive amplitudes in @xcite ( for that method see also @xcite ) .
it might be interesting to mention that the original expectations on the necessity of a complete , direct two - loop massive feynman diagram evaluation were not fulfilled .
after the analytical evaluation of a massive planar and a massive non - planar double - box diagram ( both with seven propagators ) in @xcite and in @xcite , resp . , there was hope to evaluate all the remaining diagrams soon .
there are 33 two - loop box master integrals , nine of them with seven lines @xcite .
in fact , from recent studies on the exact and mass expanded treatment of two - loop box master integrals in @xcite and @xcite , resp .
, it became clear that an evaluation of all the massive master integrals is a more complicated task than was expected .
quite recently , the case of non - planar master integrals was successfully treated in another , but related context @xcite .
proceeding similarly for bhabha scattering seems feasible now . as a matter of fact , due to these reasons
, the direct feynman diagram approach was not used for the phenomenologically needed two - loop predictions and the above - mentioned papers @xcite-@xcite remained so far a mere interesting , challenging theoretical development . in this paper , we report on the evaluation of the leptonic @xmath13 two - loop contributions with _ arbitrary mass _ of the heavier fermion , exploring the extended kinematical region @xmath15 .
we use the dispersion approach , so that our formulae may be applied without further modification also to hadronic corrections . for a review on the status of monte - carlo studies for bhabha scattering at this conference
we refer to @xcite , and for a discussion on progress related to radiative loop corrections to @xcite .
the classes of two - loop @xmath13 corrections are shown in figure [ fig - nf2 ] .
classes of two - loop diagrams for bhabha scattering containing at least one fermion loop . ] the four direct and four crossed fermionic two - loop box diagrams , obtained applying proper permutations to the sample diagram shown in figure [ fig - nf2 ] , are infrared ( ir ) divergent , and they have to be combined with other ir - divergent factorizable corrections in order to get an ir - finite contribution to the cross section . in particular , we have to add the interference of two - loop box ( class @xmath16 ) and reducible vertex diagrams ( class @xmath17 ) with the tree - level amplitude to the interference of one - loop vertex and box diagrams with one - loop vacuum polarization diagrams .
finally , we construct an ir - finite quantity taking into account also the real emission of one soft photon from one - loop vacuum polarization diagrams .
sample contributions are given in figure [ fig - irfin ] .
the fermionic two - loop boxes combine with other diagrams to an infrared - finite cross - section contribution .
, scaledwidth=50.0% ] the net contribution of pure self - energy corrections ( classes @xmath18 ) , irreducible vertex diagrams ( class @xmath19 ) , and the aforementioned ir - divergent contributions reads as @xmath20 where @xmath21 can be splitted in two components , @xmath22 we concentrate now on the renormalized two - loop box diagrams of class e , whose total contribution to the cross section may be written as @xmath23 here the auxiliary functions @xmath24 and @xmath25 can be conveniently expressed through three independent form factors @xmath26 , with @xmath27 , evaluated with different kinematical arguments , @xmath28 the particular contribution of the diagram of figure [ fig - nf2 ] coming from the interference with the tree - level s - channel is @xmath29 , and from the t - channel is @xmath30 . for the evaluation of hadronic corrections
, we observe that each term of eq .
( [ split ] ) can be written through the convolution of a kernel function @xmath31 , @xmath32 , with the hadronic cross - section ratio @xmath33 , @xmath34 for leptons and the top quark , we have to replace @xmath35 and @xmath36 , given by @xmath37 where @xmath38 is the electric charge in units of @xmath39 and @xmath40 is the color factor .
being the box diagrams ir divergent , and showing poles in @xmath41 , one should take into account higher orders in @xmath41 for @xmath42 .
however , after assembling box diagrams with factorizable corrections , ir poles cancel and @xmath42 can be evaluated at order @xmath43 .
the three box kernels are our main technical result .
they have been derived with the aid of the master integrals of figure [ fig-1loopmasters ] , in the limit @xmath44 .
the master integrals were determined with idsolver and evaluated with the mathematica packages ambre @xcite and mb @xcite , and eventually mass expanded with a mathematica package .
we also made use of form @xcite .
the master integrals for the two - loop box kernel functions .
, scaledwidth=99.0% ] we reproduce one of the kernels here , @xmath45 where @xmath46 , @xmath47 and @xmath48 is the normalization factor @xmath49 here @xmath50 is the t hooft mass unit and @xmath51 is the euler - mascheroni constant .
the sum of the box contributions with ir - divergent factorizable corrections ( see figure [ fig - irfin ] for sample cases ) is infrared - finite and can be cast in the following form , @xmath52\nonumber\\ & + & \pi\ , \textrm{im}\ , \int_{4m^2}^{\infty}\ , dz\ , \frac{r(z)}{z}\ , \frac{1}{s - z+i\,\delta } \ , f_4(z ) \bigr\}.\end{aligned}\ ] ] we may show here one of the auxiliary functions @xmath53 , @xmath54\ ,
\ln\bigl(\frac{s}{m_e^2}\bigr ) + \bigl[-z^2\bigl(\frac{1}{s}+\frac{2}{t}+2\,\frac{s}{t^2}\bigr ) \nonumber\\ & + & z\bigl ( 4 + 4\frac{s}{t}+2\frac{t}{s}\bigr ) + \frac{1}{2}\frac{t^2}{s}+6\frac{s^2}{t}+5s+4t\bigr ] \ln\bigl(-\frac{t}{s}\bigr ) + s\bigl(-\frac{z}{t}+\frac{3}{2}\bigr ) \nonumber\\&\times&\ln\bigl(1+\frac{t}{s}\bigr ) + \bigl[\frac{1}{2}\,\frac{z^2}{s}+2\,z\,\bigl(1+\frac{s}{t}\bigr)-\frac{11}{4}\,s-2\,t \bigr]\ , \ln^2\bigl(-\frac{t}{s}\bigr ) \nonumber\\&-&\bigl[\frac{1}{2}\,\frac{z^2}{t } -z\,\bigl(1+\frac{s}{t}\bigr)+\frac{t^2}{s } + 2\,\frac{s^2}{t}+\frac{9}{2}\,s+\frac{15}{4}\,t
\bigr]\,\ln^2\bigl(1+\frac{t}{s}\bigr ) \nonumber\\&+ & \bigl[\frac{z^2}{t } -2\,z\ , \bigl(1+\frac{s}{t}\bigr ) +
2\,\frac{s^2}{t}+5\,s+\frac{5}{2}\,t \bigr]\,\ln\bigl(-\frac{t}{s}\bigr)\,\ln\bigl(1+\frac{t}{s}\bigr ) \nonumber\\&-&4\,\bigl[\frac{t^2}{s}+2\,\frac{s^2}{t}+3\,\bigl(s+t\bigr ) \bigr]\ , \bigl[1 + \textrm{li}_2\bigl(-\frac{t}{s}\bigr)\bigr ]
\nonumber\\&-&\bigl [ 2\,\frac{z^2}{t}-4\,z\,\bigl(1+\frac{s}{t}\bigr ) -4\,\frac{t^2}{s } -2\,\frac{s^2}{t}+s-\frac{11}{2}\,t\bigr]\,\zeta_2\nonumber\\&- & \bigl [ \frac{t^2}{s}+2\,\frac{s^2}{t}+3\,\bigl(s+t\bigr)\bigr]\,\ln\bigl(\frac{z}{s}\bigr)\ , \ln\bigl(1+\frac{t}{s}\bigr ) + \bigl [ z^2\,\bigl ( \frac{1}{s } + 2\,\frac{s}{t^2 } + \frac{2}{t}\bigr)\nonumber\\&-&z\,\bigl ( \frac{t}{s}+2\frac{s}{t}+2 \bigr)\bigr ] \ln\bigl(\frac{z}{s}\bigr ) -\bigl [ z^2\,\bigl(\frac{1}{s}+\frac{1}{t}\bigr ) + 2\,z\,\bigl(1+\frac{s}{t}\bigr ) + s + 2\,\frac{s^2}{t}\bigr]\,\nonumber\\&\times&\ln\bigl(\frac{z}{s}\bigr)\ , \ln\bigl(1+\frac{z}{s}\bigr ) + \bigl[\frac{z^2}{s } + 4\,z\,\bigl(1+\frac{s}{t}\bigr)-\frac{t^2}{s}-4\,\bigl(s+t\bigr)\bigr]\,\nonumber\\&\times&\ln\bigl(\frac{z}{s}\bigr)\,\ln\bigl(1-\frac{z}{t}\bigr ) - \bigl [ z^2\,\bigl(\frac{1}{s}+2\frac{s}{t^2}+ \frac{2}{t } \bigr ) -2\ , z\,\bigl(\frac{t}{s}+2\,\frac{s}{t } + 2\bigr)+\frac{t^2}{s } \nonumber\\&+&2 ( s+t)\bigr]\ , \ln\bigl(1-\frac{z}{t}\bigr ) + \bigl [ \frac{z^2}{t}-2z\bigl(1+\frac{s}{t}\bigr)+2\frac{t^2}{s } + 8s + 4\frac{s^2}{t}+7 t \bigr ] \nonumber\\&\times & \ln\bigl(1-\frac{z}{t}\bigr)\,\ln\bigl(1+\frac{t}{s}\bigr ) -\bigl [ z^2\,\bigl(\frac{1}{s}+\frac{1}{t}\bigr ) + 2\,z\,\bigl(1+\frac{s}{t}\bigr ) + s + 2\,\frac{s^2}{t } \bigr]\ , \nonumber\\&\times&\textrm{li}_2\,\bigl(-\frac{z}{s}\bigr ) + \bigl[\frac{z^2}{s}+4\,z\,\bigl(1+\frac{s}{t}\bigr ) -\frac{t^2}{s}-4\,\bigl(s+t\bigr)\bigr]\ , \textrm{li}_2\,\bigl(\frac{z}{t}\bigr)\nonumber\\&- & \bigl[\,\frac{z^2}{t } -2\,z\,\bigl(1+\frac{s}{t}\bigr ) + \frac{t^2}{s } + 5\,s+2\,\frac{s^2}{t } + 4\,t\ , \bigr]\,\textrm{li}_2\,\bigl(1+\frac{z}{u}\bigr ) \bigr\}\nonumber\\&+&4\,\bigl(\frac{1}{3}\,\frac{t^2}{s}+\frac{2}{3}\,\frac{s^2}{t}+s+t\bigr)\ , \ln\bigl(\frac{2\,\omega}{\sqrt{s}}\bigr)\ , \bigl [ \ln\bigl(\frac{s}{m_e^2}\bigr ) + \ln\bigl(-\frac{t}{s}\bigr)\nonumber\\&-&\ln\bigl(1+\frac{t}{s}\bigr)-1\bigr ] . \ ] ] table [ nums1 ] and table [ nums2 ] contain numerical results for small- and large - angle scattering at a variety of energy scales .
we report the qed tree - level prediction , as well as the process - dependent contributions at nnlo of eq .
( [ totremainder ] ) ; in other words , we exclude from the tables pure self - energy corrections , which can be described introducing a running fine - structure constant and were deeply investigated in the past ( see @xcite ) , and irreducible vertex contributions ( see @xcite and @xcite ) .
a complete phenomenological analysis requires also to add the corresponding terms arising from unresolved real fermion pair production .
the standard model cross sections shown here rely on born formulae with @xmath7 boson and photon exchange a la zfitter @xcite . moreover , although the contributions from electron loops have been obtained by exact ( in @xmath0 ) evaluation of the feynman diagrams , we report here for consistency the approximated results for @xmath55 .
we further compare the analytical results of @xcite with those obtained with the dispersion approach and it is nicely seen that the former approach the latter in regions where the former are expected to become good approximations . in both tables , for each fermion flavour , we show the result obtained through the dispersion - based approach ( first line ) and the one coming from the analytical expansion ( second line ) , neglecting @xmath56 , where @xmath57 . when @xmath58 , the entry is suppressed .
we switch off in the tables the effect of the logarithm containing the energy of soft photons setting @xmath59 .
the heavier fermions have less influence on the net result , and the top quark decouples nearly completely .
between 1 and 500 gev the sum of boxes with factorizable diagrams with muon loops contributes , roughly speaking , at the order of permille to the net pure qed cross section , and the tau lepton contributes less .
the @xmath7 resonance distortes this figure , by making the influence of two - loop contributions less influential for large - angle scattering where the resonance dominates .
.numerical values for the differential cross section in nanobarns at a scattering angle @xmath60 , in units of @xmath61 ; @xmath62 gev .
bold - face entries are obtained with dispersion relations .
the energy of the soft photon is chosen to be @xmath59 . [ cols=">,>,>,>,>",options="header " , ]
we have evaluated the @xmath13 virtual two - loop corrections to bhabha scattering due to fermions with arbitrary mass @xmath1 in the limit of vanishing electron mass @xmath0 .
we have not combined these ( infrared finite ) virtual contributions with those arising from irreducible vertices ; the latter are logarithmically enhanced by terms of order up to @xmath63 , but are independent of @xmath64 , being collinear finite .
they have to be assembled with the unresolved real heavy fermion emission , which is known to cancel the @xmath63 and might lead to a suppression of the net effect . for the phenomenological analysis we will have also to take into account the effect of the running of the fine - structure constant . concerning the results shown in the tables
, the numerical contributions do not exceed the per mille level , and depend strongly on the kinematics .
the formulae presented here apply to the leptons @xmath50 and @xmath65 , but also to the top quark .
the latter decouples at small energies , but has to be taken into account at the ilc .
z. bern , l. dixon , and a. ghinculov , _ phys .
* d63 * ( 2001 ) 053007 , http://www.arxiv.org/abs/hep-ph/0010075[hep-ph/0010075 ] .
n. glover , b. tausk , and j. van der bij , _ phys .
lett . _ * b516 * ( 2001 ) 3338 , http://www.arxiv.org/abs/hep-ph/0106052[hep-ph/0106052 ] .
a. penin , _ phys .
* 95 * ( 2005 ) 010408 , http://www.arxiv.org/abs/hep-ph/0501120[hep-ph/0501120 ] .
a. penin , _ nucl .
* b734 * ( 2006 ) 185202 , http://www.arxiv.org/abs/hep-ph/0508127[hep-ph/0508127 ] .
r. bonciani , p. mastrolia , and e. remiddi , _ nucl .
phys . _ * b661 * ( 2003 ) 289343 , http://www.arxiv.org/abs/hep-ph/0301170[hep-ph/0301170 ] .
r. bonciani , p. mastrolia , and e. remiddi , _ nucl .
* b676 * ( 2004 ) 399452 , http://www.arxiv.org/abs/hep-ph/0307295[hep-ph/0307295 ] .
r. bonciani , a. ferroglia , p. mastrolia , e. remiddi , and j. van der bij , _ nucl .
* b681 * ( 2004 ) 261291 , http://www.arxiv.org/abs/hep-ph/0310333[hep-ph/0310333 ] .
m. czakon , j. gluza , and t. riemann , _ phys .
_ * d71 * ( 2005 ) 073009 , http://www.arxiv.org/abs/hep-ph/0412164[hep-ph/0412164 ] .
r. bonciani , a. ferroglia , p. mastrolia , e. remiddi , and j. van der bij , _ nucl .
* b701 * ( 2004 ) 121179 , http://www.arxiv.org/abs/hep-ph/0405275[hep-ph/0405275 ] .
r. bonciani , a. ferroglia , p. mastrolia , e. remiddi , and j. van der bij , _ nucl .
* b716 * ( 2005 ) 280302 , http://www.arxiv.org/abs/hep-ph/0411321[hep-ph/0411321 ] .
s. actis , m. czakon , j. gluza , and t. riemann , _ nucl .
phys . _ * b786 * ( 2007 ) 2651 , http://www.arxiv.org/abs/arxiv:0704.2400v.2 [ hep - ph][arxiv:0704.2400v.2 [ hep - ph ] ] . t. becher and k. melnikov , _ jhep _ * 06 * ( 2007 ) 084 , http://www.arxiv.org/abs/arxiv:0704.3582 [ hep - ph][arxiv:0704.3582 [ hep - ph ] ] .
a. mitov and s. moch , _ jhep _ * 05 * ( 2007 ) 001 , http://www.arxiv.org/abs/hep-ph/0612149[hep-ph/0612149 ] .
v. smirnov , _ phys .
* b460 * ( 1999 ) 397404 , http://www.arxiv.org/abs/hep-ph/9905323[hep-ph/9905323 ] .
b. tausk , _ phys .
lett . _ * b469 * ( 1999 ) 225234 , http://www.arxiv.org/abs/hep-ph/9909506[hep-ph/9909506 ] .
m. czakon , j. gluza , and t. riemann , _ nucl .
( proc . suppl . )
_ * b135 * ( 2004 ) 83 , http://www.arxiv.org/abs/hep-ph/0406203[hep-ph/0406203 ] .
g. heinrich and v. smirnov , _ phys .
_ * b598 * ( 2004 ) 5566 , http://www.arxiv.org/abs/hep-ph/0406053[hep-ph/0406053 ] .
m. czakon , j. gluza , and t. riemann , _ nucl .
phys . _ * b751 * ( 2006 ) 117 , http://www.arxiv.org/abs/hep-ph/0604101[hep-ph/0604101 ] .
m. czakon , a. mitov , and s. moch , _ phys .
lett . _ * b651 * ( 2007 ) 147159 , http://www.arxiv.org/abs/arxiv:0705.1975 [ hep - ph][arxiv:0705.1975 [ hep - ph ] ] . g. montagna , `` status of precision monte carlo tools for luminosity monitoring at meson factories '' , talk held at this conference .
j. gluza , k. kajda , and t. riemann , _ comput .
* 177 * ( 2007 ) 879893 , http://www.arxiv.org/abs/arxiv:0704.2423 [ hep - ph][arxiv:0704.2423 [ hep - ph ] ] . m. czakon , _ comput .
* 175 * ( 2006 ) 559571 , http://www.arxiv.org/abs/hep-ph/0511200[hep-ph/0511200 ] .
j. vermaseren , `` new features of form '' , math - ph/0010025 .
a. arbuzov , d. haidt , c. matteuzzi , m. paganoni , and l. trentadue , _ eur .
phys . j. _ * c34 * ( 2004 ) 267275 , http://www.arxiv.org/abs/hep-ph/0402211[hep-ph/0402211 ] .
* b209 * ( 1988 ) 337 .
g. burgers , _ phys .
* b164 * ( 1985 ) 167 .
d. bardin _ et al .
_ , _ comput . phys
. commun . _
* 133 * ( 2001 ) 229395 , http://www.arxiv.org/abs/hep-ph/9908433[hep-ph/9908433 ] . a.
arbuzov , m. awramik , m. czakon , a. freitas , m. grnewald , k. mnig , s. riemann , and t. riemann , _ comput .
* 174 * ( 2006 ) 728758 , http://www.arxiv.org/abs/hep-ph/0507146[hep-ph/0507146 ] .
r. bonciani , a. ferroglia and a. a. penin , arxiv:0710.4775 [ hep - ph ] | we derive the two - loop corrections to bhabha scattering from heavy fermions using dispersion relations .
the double - box contributions are expressed by three kernel functions .
convoluting the perturbative kernels with fermionic threshold functions or with hadronic data allows to determine numerical results for small electron mass @xmath0 , combined with arbitrary values of the fermion mass @xmath1 in the loop , @xmath2 , or with hadronic insertions .
we present numerical results for @xmath3 at typical small- and large - angle kinematics ranging from 1 gev to 500 gev . |
we perform _ ab initio _ calculations based on density functional theory ( dft ) to find the band structures of bi@xmath0auo@xmath43 , sn(pbo@xmath0)@xmath0 , pb@xmath1o@xmath2 , and gamo@xmath1 .
we use the perdew burke ernzerhof
type generalized gradient approximation @xcite as found in the quantum espresso package @xcite . norm conserving ,
optimized , designed nonlocal pseudopotentials are generated using the opium package @xcite .
an energy cutoff of 680 ev is used for the plane wave basis . a fully - relativistic non
collinear scheme is employed to describe spin - orbit interaction .
the atomic structures are obtained from the inorganic crystal structure database ( icsd ) database @xcite . here
we present our dft band structures of real materials that host the double dirac point .
we have searched the materials project @xcite to find stable crystals in the space groups 130 , 135 , 218 , 220 , 222 , 223 , and 230 , resulting in the following set of candidate materials .
the energy band structure of bi@xmath0auo@xmath43 in space group 130 is shown in fig .
the band structure is drawn along the high symmetry lines of the tetragonal bz , shown in the upper panel .
the fermi level is set to @xmath46 ev .
the double dirac point ( indicated by a blue circle ) appears at @xmath7 in good agreement with the results of the four - site tight - binding model .
there is an additional dirac point in @xmath44-@xmath45 , guaranteed by symmetry , which is magnified in the upper inset .
we expect that bi@xmath0auo@xmath43 , which has been synthesized in a single - crystal phase @xcite , should realize a double dirac semimetal . in fig .
[ fig2 ] , the electronic energy band structure of sn(pbo@xmath0)@xmath0 in space group 135 is presented .
the fermi level is set to @xmath46 ev and the valence bands are plotted along high symmetry lines of the tetragonal bz .
the double dirac point is enclosed by a blue - colored circle .
the number of sn , pb , and o atoms in the unit cell are 4 , 8 , and 16 , respectively , and the number of valence electrons of sn , pb , and o are 14 , 14 , and 6 , respectively , and thus the total number of valence electrons is 72 .
although sn(pbo@xmath0)@xmath0 is an insulator with filling zero ( mod 8) , a double dirac point appears in the valence band at 0.54 ev below the fermi level . the atomic structure of sn(pbo@xmath0)@xmath0 @xcite allows for the possibility of utilizing material design to shift the fermi energy closer to the double dirac point .
for example , when sn is substituted with in or sb , which have one less or one more valence electron than sn , respectively , the electron filling can change from zero to four ( mod 8) , possibly hosting the double dirac point close to the fermi energy .
figure [ fig3 ] shows the band structure of pb@xmath1o@xmath2 in space group 135 .
pb@xmath1o@xmath2 is isostructural with sn(pbo@xmath0)@xmath0 with sn being substituted by pb .
the band structure also resembles that of sn(pbo@xmath0)@xmath0 .
a double dirac point resides at @xmath47 in the valence band region near @xmath48 -0.57 ev .
the electronic energy bands of gamo@xmath1 @xcite are drawn in fig .
[ fig4 ] along the high symmetry lines of the cubic bz , depicted in the inset .
the position of the double dirac point is indicated by a blue - colored circle . | we study a class of dirac semimetals that feature an eightfold - degenerate double dirac point .
we show that 7 of the 230 space groups can host such dirac points and argue that they all generically display linear dispersion .
we introduce an explicit tight - binding model for space groups 130 and 135 , showing that 135 can host an intrinsic double dirac semimetal
one with no additional degeneracies at the fermi energy .
we consider symmetry - lowering perturbations and show that uniaxial compressive strain in different directions leads to topologically distinct insulating phases .
in addition , the double dirac semimetal can accommodate topological line defects that bind helical modes .
potential materials realizations are discussed .
a striking consequence of symmetry and topology in the electronic structure of materials is the existence of protected degeneracies that guarantee semimetallic behavior .
such degeneracies occur in graphene @xcite ( in the absence of spin - orbit interactions ) as well as at the surface of a topological insulator ( ti ) @xcite . in 2011 ,
wan et al .
@xcite pointed out that twofold degenerate weyl points could occur in bulk 3 dimensional ( 3d ) materials .
such weyl points are topologically protected , though they are symmetry prevented " in that they require broken inversion or time - reversal ( t ) symmetry to exist .
crystal symmetries can lead to a richer variety of nodal semimetals .
dirac semimetals @xcite , which feature fourfold degenerate dirac points protected by crystal symmetry , occur in two varieties .
topological dirac semimetals , such as cd@xmath0as@xmath1 and na@xmath0bi @xcite , exhibit dirac points on a rotational symmetry axis due to a band inversion .
nonsymmorphic dirac semimetals , predicted in bio@xmath0 @xcite and in biznsio@xmath2 @xcite , conversely have dirac points at high - symmetry points which are guaranteed by the underlying space group ( sg ) symmetry .
additional classes of nodal semimetals include line nodes @xcite in 3d and dirac semimetals in 2d @xcite .
in this paper we introduce and analyze a double dirac semimetal that exhibits a single eightfold degeneracy point at a brillouin zone ( bz ) corner .
we show that 7 of the 230 sgs host double dirac points ( ddps ) and argue that all of them generically have linear dispersion . for two of the sgs ( 130 and 135 ) a ddp
is guaranteed whenever the band filling is an odd multiple of four , while for the remaining five sgs the presence of ddps depends on the band ordering .
we introduce an explicit tight - binding model for sgs 130 and 135 that demonstrates the ddp , and we study its low energy structure in detail .
like the single ( nonsymmorphic ) dirac semimetal , the double dirac semimetal can be gapped into a trivial or topological insulator by applying strain . in double dirac semimetals ,
both phases can be achieved with compressive strain oriented along two different directions . moreover , in the double dirac semimetal , spatially modulating the symmetry - breaking energy gap can lead to topological line defects that bind 1d helical modes .
materials hosting ddps are discussed at the end of the paper .
the existence of symmetry - protected degeneracies at a point @xmath3 in the bz can be ascertained by determining the dimension of the appropriate double - valued projective representations of the little group of @xmath3 .
this information has been tabulated for all 230 sgs @xcite .
table [ table1 ] lists all of the sgs with symmetry points that host four - dimensional irreducible representations ( 4dir ) that are also doubled by t symmetry .
.space groups that host ddps .
sgs are indicated in international notation as well as in shnflies notation , which indicates the crystal system and point group .
the momenta @xmath3 are listed with symmetry labels for the 8dirs , as well as for some 4dirs .
the final column indicates the t - invariant vector representations of the point group contained in the tensor product @xmath4 of the 8dir at @xmath3 , indicating that in each case a linear dispersion is generic . [ cols="<,<,<,<,^,^",options="header " , ] we next consider the long - wavelength symmetry - breaking perturbations that open energy gaps and identify the resulting phases that arise .
general symmetry - breaking perturbations are classified by their symmetries under the @xmath5 point group , which can be determined from @xmath6 , as above .
the possible t - invariant perturbations at @xmath7 are listed in table [ table3 ] .
there are many terms , and their effects depend on the form of the velocity terms . in order to organize the behavior
, we first fix the velocity terms and examine the terms that can open a gap in the spectrum .
we find that there are precisely four mass terms that arise due to perturbations with specific symmetries .
the hamiltonian has the form @xmath8 the four mass terms are the unique terms from table [ table3 ] that anticommute with all three of the velocity terms and open a gap .
they fall into two groups . @xmath9 and @xmath10 anticommute among themselves but commute with each other , leading to an energy gap
@xmath11 @xmath12 and @xmath13 thus define topologically distinct phases . by evaluating the parity eigenvalues in the tight - binding model @xcite
, we conclude that @xmath12 defines a @xmath14 weak ti , while @xmath13 defines a @xmath15 strong ti @xcite . while these topological indices depend on the details of the band structure away from the ddp , the difference between them for the @xmath12- and @xmath13-dominated phases is robust . to visualize the phases , fig .
[ fig2](a ) shows a 3d phase diagram for @xmath16 . in this restricted space ,
the boundary between the sti and the wti is a cone , with the sti ( wti ) inside ( outside ) the cone . in the more general 4d phase diagram ,
the wti and sti phases appear symmetrically .
different combinations of velocity terms lead to different choices for the anticommuting mass terms .
however , for any choice of velocity terms , there is always one anticommuting term ( or linear combination of terms ) with each of the symmetries in eq .
[ masses ] .
this means that in the absence of fine tuning , a generic perturbation with a given symmetry induces the corresponding mass term and opens a gap .
the general structure of fig . [ fig2](a ) remains , except that when the inversion - symmetry - breaking term @xmath17 is present the boundary between the sti and wti phases broadens to include an intermediate weyl semimetal phase @xcite .
the dependence of the topological state on @xmath18 and @xmath19 provides a mechanism for controlling topological states using strain .
as indicated in fig.[fig2](b , c ) , uniaxial strain along the @xmath20 or @xmath21 directions induces a perturbation with a combination of @xmath22 and @xmath23 symmetry , while uniaxial strain along the diagonal @xmath24 directions has @xmath22 and @xmath25 symmetry . therefore these two kinds of compressive strain lead to topologically distinct insulators .
ddps can also be differentiated from single dirac semimetals by the existence of two distinct anticommuting mass terms that lead to the same topological phase . for a single dirac semimetal with @xmath26 dirac matrices ,
the general structure of clifford algebras predicts that there is only a single t - invariant mass term . for @xmath27 dirac matrices
there are _ two _ independent t - invariant mass terms .
this means that the space of gapped states has a nontrivial first homotopy group , indicated by the dashed circle in fig .
[ fig2](a ) , allowing topologically nontrivial _ line defects _
( fig .
[ fig2](d ) ) .
line defects in a 3d insulator in class aii have a @xmath28 topological invariant characterizing the @xmath29d @xmath30 @xcite .
when nontrivial , this guarantees that a 1d helical mode is bound to the line , similar to the helical mode bound to a lattice dislocation in a weak ti @xcite . without a lattice dislocation
, this @xmath28 invariant is inaccessible in a 4 band system because it derives via dimensional reduction @xcite from a _ third _ chern number in @xmath31d , which requires at least 8 bands . to establish that a line defect binds a 1d helical mode , we follow the analysis of ref . , and consider a simple model with @xmath32 , @xmath33 and @xmath34 .
@xmath35 in ( [ masses ] ) then has the form of a harmonic oscillator , and there is a single pair of modes with @xmath36 localized near @xmath37 .
we finally briefly consider perturbations in table [ table3 ] with the remaining symmetries , which lead to weyl or dirac semimetals .
the @xmath38 perturbations lead to either dirac points or a dirac ring , with fourfold - degenerate crossings .
the inversion breaking - perturbations generally lead to a weyl semimetal with twofold crossings , except for @xmath39 , where the remaining @xmath40 and glide mirror symmetries guarantee doubly - degenerate states for @xmath41 with the same @xmath40 eigenvalue .
this allows degenerate bands to cross along @xmath7-@xmath42 , protecting a dirac point even though inversion is violated .
auo@xmath43 in sg 130 , obtained from first - principles calculations .
the ddp appears at @xmath7 with extra dirac points along @xmath44-@xmath45 .
see the supplementary material for the methods.,width=302 ] encouragingly , the ddp appears to be feasible in known materials @xcite .
for example , a ternary bismuth aurate , bi@xmath0auo@xmath43 in sg 130 , which has been synthesized in a single crystal @xcite , hosts a ddp at @xmath7 close to the fermi level , with additional four - fold degeneracies appearing on @xmath45-@xmath44 ( fig .
[ fig3 ] ) . as for materials in sg 135 , the materials project
@xcite shows that a class of oxide materials isostructural with pb@xmath1o@xmath2 @xcite , including sn(pbo@xmath0)@xmath0 , pb@xmath1o@xmath2 , and mg(pbo@xmath0)@xmath0 , host the ddps in the valence energy regime .
although they are semiconductors with electron filling 8 , their atomic structure allows for a potential route towards a material design that shifts the dirac point near the fermi level .
the number of atoms per each species in a unit cell is 4 ( mod 8) and thus allows for filling 4 when suitably substituting atoms with an odd number of valence electrons . in sg 223 , gamo@xmath1 @xcite hosts a ddp .
there is reason for optimism that with appropriate band structure engineering , an intrinsic double dirac semimetal can be realized .
we thank eugene mele and steve young for helpful discussions .
this work was supported by nsf grant dmr 1120901 and a simons investigator grant to clk from the simons foundation .
amr acknowledges the support of the doe , under grant de - fg02 - 07er46431 .
a. h. c. neto , n. m. r. peres , k. s. novoselov , and a. k. geim , rev .
mod .
phys . * 81 * , 109 ( 2009 ) .
m. z. hasan and c. l. kane , rev .
mod .
phys . * 82 * , 3045(2010 ) ; x .-
l .
qi and s .- c .
zhang , rev .
mod .
phys .
* 83 * , 1057 ( 2011 ) .
x. wan , a. m. turner , a. vishwanath , and s. y. savrasov , phys .
rev .
b * 83 * , 205101 ( 2011 ) .
s. m. young , s. zaheer , j. c. y. teo , c. l. kane , e. j. mele , and a. m. rappe , phys .
rev .
lett . * 108 * , 140405 ( 2012 ) .
z. wang , y. sun , x .- q .
chen , c. franchini , g. xu , h. weng , x. dai , and z. fang , phys .
rev .
b * 85 * , 195320 ( 2012 ) .
z. wang , h. weng , q. wu , x. dai , and z. fang , phys .
rev .
b * 88 * , 125427 ( 2013 ) .
z. k. liu , b. zhou , y. zhang , z. j. wang , h. m. weng , d. prabhakaran , s .- k .
mo , z. x. shen , z. fang , x. dai , z. hussain , and y. l. chen , science * 343 * , 864867 ( 2014 ) .
z. k. liu , j. jiang , b. zhou , z. j.wang , y. zhang , h. m. weng , d. prabhakaran , s .- k .
mo , h. peng , p. dudin , t. kim , m. hoesch , z. fang , x. dai , z. x. shen , d. l. feng , z. hussain , and y. l. chen , nat .
mater .
* 13 * , 677681 ( 2014 ) .
s. borisenko , q. gibson , d. evtushinsky , v. zabolotnyy , b. bchner , and r. j. cava , phys .
rev .
lett .
* 113 * , 027603 ( 2014 ) .
j. a. steinberg , s. m. young , s. zaheer , c. l. kane , e. j. mele , and a. m. rappe , phys . rev .
lett . * 112 * , 036403 ( 2014 ) .
h. weng , y. liang , q. xu , r. yu , z. fang , x. dai , and y. kawazoe , phys .
rev .
b * 92 * , 045108 ( 2015 ) .
g. e. volovik , j. supercond .
nov .
magn .
* 26 * , 2887 ( 2013 ) .
h .- s .
kim , y. chen , and h .- y .
kee , phys .
rev .
b * 91 * , 235103 ( 2015 ) . y. chen , y .- m .
lu , and h .- y .
kee , nat .
commun .
* 6 * , 6593 ( 2015 ) .
h. weng , c. fang , z. fang , b. a. bernevig , and x. dai , phys .
rev .
x * 5 * , 011029 ( 2015 ) .
y. kim , b. j. wieder , c. l. kane , a. m. rappe , phys .
rev .
lett . *
115 * , 036806 ( 2015 ) .
r. yu , h. weng , z. fang , x. dai , and x. hu , phys .
rev .
lett . * 115 * , 036807 ( 2015 ) .
l. s. xie , l. m. schoop , e. m. seibel , q. d. gibson , w. xie , and r. j. cava , ( 2015 ) , apl mat .
* 3 * , 083602 ( 2015 ) .
z. wang , d. gresch , a. a. soluyanov , w. xie , s. kushwaha , x. dai , m. troyer , r. j. cava and b. a. bernevig , arxiv:1511.07440 ( 2015 ) .
l. m. schoop , m. n. ali , c. strasser , v. duppel , s.s .
p. parkin , b. v. lotsch and c. r. ast , arxiv:1509.0086 ( 2015 ) .
s. m. young and c. l. kane , phys .
rev .
lett . * 115 * , 126803 ( 2015 ) . c. j. bradley and a. p. cracknell , the mathematical theory of symmetry in solids " ( clarendon press oxford , 1972 ) .
h. watanabe , h. c. po , a. vishwanath , m. p. zaletel , proc .
natl .
acad .
sci . * 112 * , 14551 ( 2015 ) .
s. groves and w. paul , phys .
rev .
lett . * 11 * , 194 ( 1963 ) .
c. fang , y. chen , h .- y .
kee , l. fu , phys .
rev .
b * 92 * , 081201 ( 2015 ) . l. fu and c.l . kane , phys .
rev .
b * 76 * , 045302 ( 2007 ) .
l. fu , c.l .
kane and e.j .
mele , phys .
rev .
lett . * 98 * , 106803 ( 2007 ) .
s. murakami , new journ . of phys .
* 9 * , 356 ( 2007 )
. j.c.y .
teo and c.l .
kane , phys .
rev .
b * 82 * , 115120 ( 2010 ) .
y. ran , y. zhang , and a. vishwanath , nat .
phys . * 5 * , 298 ( 2009 ) .
x .- l .
qi , t. l. hughes , and s .- c .
zhang , phys .
rev .
b * 78 * , 195424 ( 2008 ) .
see supplemental material for the band structures .
j. geb and m. jansen , j. solid state chem .
* 122 * , 364 ( 1996 ) .
a. jain , s. p. ong , g. hautier , w. chen , w. k. richards , s. dacek , s. cholia , d. gunter , d. skinner , g. ceder and k.a .
persson , apl materials * 1 * , 011002 ( 2013 ) .
j. r. gavarri , j. p. vigouroux , g. calvarin and a. w. hewat , j. solid state chem .
* 36 * , 81 ( 1981 ) .
j. bornand , r. siemens , and l. oden , journal of the less common metals * 30 * , 205 ( 1973 ) .
j.p .
perdew , k. burke , and m. ernzerhof , phys .
rev .
lett .
* 77 * , 3865 ( 1996 ) .
p. giannozzi _
et al . _ , j. phys .
condens . mat .
* 21 * , 395502 ( 2009 ) .
a. m. rappe , k. m. rabe , e. kaxiras , and j. d. joannopoulos , phys .
rev .
b * 41 * , 1227 ( 1990 ) . n. j. ramer and a. m. rappe , phys . rev .
b * 41 * , 12471 ( 1999 ) .
inorganic crystal structure database ( icsd ) , ( fachinformationszentrum karlsruhe , karlsruhe , germany , 2015 ) . |
the recent claim from the pamela collaboration of an excess in the ratio of positrons to electrons plus positrons ( positron fraction ) , seems to support what heat and ams01 data suggested regarding the existence of a possible primary cosmic ray electrons source .
however some uncertainties remain , e.g the possible enhancement of the positron flux due to the relatively poor knowledge of the propagation parameters set @xcite .
a large number of candidates have been suggested that may be able to reproduce the observed positron fraction . among the exotic ones , annihilating dark matter
may be the most widely studied .
dark matter is assumed to annihilate in the galactic halo and its annihilation products should reach the earth and be measured .
as it has been pointed out in previous works @xcite , pamela data seems to accommodate preferably leptonic channels rather than dark matter decaying to quark pairs , but typically large boost factors are required to adjust the shape to the data .
this boost factors are related to an increase of the annihilation rate due to e.g. the recently invoked sommerfeld effect , annihilation of a non thermal wimp , or the usual inhomogeneities of the dark matter halo ie , clumps , dark matter mini spikes etc ... moreover , to accomodate theory to the existing data , different normalizations are required for the different experiments .
more conventional scenarios can also reproduce the spectral shape . specifically in this work we analyse the possible contribution of nearby astrophysical objects i.e. pulsars , and for leptophilic dark matter .
+ in the case of multiple sources , a natural explanation for the spectral shape in the atic data is found , namely the presence of peaks in the spectrum can be reproduced in case we have a convenient selection of pulsars , or in case we have a distribution of clumps contributing to the total flux with the required luminosity , however it is necessary to fine tune the pulsar parameters to fit the peaks present in the atic spectrum . on the contrary ,
the fermi data shows a much flatter spectrum that can be easily reproduced by pulsars with standard parametrizations .
we calculate the predicted dipole anisotropy produced in both scenarios , i.e annihilating dark matter clumps and nearby pulsars , and assess the fermi detection capability of such signature at least at the 2 sigma confidence level .
cosmic ray propagation is a diffusive process due to the random galactic magnetic fields .
if we denote the number of particles of type @xmath1 per unit volume found in a time t at @xmath2 with energy @xmath3 by @xmath4 @xmath5 $ ] , the evolution equation as initially written by @xmath6 and @xmath7 @xcite can be expressed as : @xmath8 where @xmath9 is the diffusion term with @xmath10 @xmath11 $ ] the diffusion tensor ; @xmath12 is the convection term with @xmath13 @xmath14 $ ] the velocity of the galactic wind ; @xmath15 represents the energy loss with @xmath16 @xmath17 $ ] the rate of energy change ; @xmath18 is the fraction of particles lost by radioctive decay with a characteristic life - time @xmath19 @xmath20 $ ] ; @xmath21 is the number density of particles created or destroyed by spallation processes in the propagation due to interactions with the interestellar medium of density n , with a characteristic cross section @xmath22 @xmath23 $ ] .
finally we have the source term @xmath24 @xmath25 $ ] that injects particles of type @xmath1 into the galaxy . for electron and positron propagation ,
the relevant processes are diffusion , convection , energy losses and the source term .
propagation can be approached in two complementary ways .
the first one is to solve the transport equation using numerical methods in the same way the package _ galprop _ @xcite does .
the second one is to analytically solve the transport equation with a set of realistic simplifying assumptions . in this work
we use the standard _
galprop _ code to get the positron and electron backgrounds and an analytical solution of the propagation equation for the primary positron s flux . for primary positron sources ,
we have used the green functions formalism as described in @xcite where the two main processes , diffusion and energy losses are considered . the resulting diffusion - loss equation for this process
is given by : @xmath26 [ diff - eq ] where @xmath27 codifies the energy losses due to ( a ) inverse compton and synchrotron radiation , ( b ) bremsstrahlung and ( c ) ionisation .
at the energies we are interested @xmath28 gev the energy loss is very well aproximated by the synchrotron losses in the interstellar ( ism ) magnetic fields and inverse compton off the cmb and starlight at optical and ir frequencies .
the energy loss due to these processes is calculated as @xcite : @xmath29 \end{aligned}\ ] ] where the energy density @xmath30 is the energy density of the photon background [ @xmath31 with @xmath32 .
@xmath33 stands for the ism magnetic field energy density @xmath34 for b=@xmath35 g .
+ for the diffusion coefficient @xmath36 , three setups max , med and min can be considered @xcite which are consistent with the b / c .
.diffusion setups [ cols="^,^,^ " , ] the resulting fluxes are shown ( figs .
[ pulsars ] and [ young ] ) for the considered scenarios . in order to account for the experimental data in the hundreds of gev ,
the conversion efficiency into pairs of monogem and geminga must be above the nominal value of 3% but inside the standard range ( 1%-30% ) considered in @xcite .
this values may change depending on the diffusion setup considered , the pwn @xmath37 release delay , pulsar cutoff etc ... so it should be considered as a single realisation of the multiple possibilities that can reproduce the data @xcite .
according to the standard @xmath38cdm cosmological model , a @xmath39 of the energy content of the universe is in the form of cold dark matter ( cdm ) .
probably the leading candidates to account for it are weakly interacting particles ( wimps ) , with the neutralino and the kaluza - klein boson @xmath40 the most extensively studied ones . the relic density of these particles is determined by their annihilation cross section .
observations of the cmb and large scale structure surveys estimate the relic density constraining the annihilation cross section up to a canonical value of @xmath41 .
this value sets the rate of production of standard model particles , e.g leptons , that can be measured as tracers of dark matter annihilation .
+ the signal that results from dm annihilation , depends on the squared dm density from the astrophysical side , and on the dm particle mass and cross section from the particle physics side . since the annihilation rate depends on the squared density , the presence of clumpiness or substructure imply an enhancement of the signal compared to a smooth density distribution .
the present dark matter structure is considered to have its roots in small amplitude quantum fluctuations during inflation . in the accepted `` bottom - up '' hierarchical structure formation , smaller clumps
gather together to form larger systems , completely determined by the initial power spectrum of the primordial fluctuations .
galaxies are thus embedded in large dark matter halos that in turn are made up of self - bound substructure or subhalos .
the mass distribution , abundance and internal structure of clumps is determined by means of high - resolution numerical simulations as the one conducted by diemand et al.@xcite . in this work
we make use of the mass distribution of clumps that results from diemand s simulation as in cumberbatch et al .
@xcite expressed as : @xmath42\ ] ] where the lower mass cutoff is @xmath43 and , inspired by diemand et al .
simulation , we take an upper mass cutoff at @xmath44 .
we normalise the distribution such that we have a local clump density ( r=@xmath45 ) of 500 @xmath46 between @xmath47 and @xmath48 .
we assume for the spatially dependent number density a spherically symmetrical navarro , frenk&white ( nfw ) profile so the density of clumps is given by @xmath49^{2 } } \frac{dn_{d}}{d\log(m / m_{\odot})}\ ] ] where r=20kpc and @xmath50 for a correct normalization . for the internal structure within a clump ,
we adopt a nfw density profile that gives a reasonable fit to the lightest clumps in diemand s simulation and reproduces well the outer parts of the large scale subhalos .
@xmath51^{2}}\ ] ] where the maximum radius of the clump is defined as the radius @xmath52 where the density equals 200 times the critical density .
we assume a universal concentration parameter c for all the clumps , found to lie in diemands simulation within the @xmath53 range , and a constant density core below @xmath54kpc . with all these considerations ,
the source term can be expressed as : @xmath55 where f(e ) denotes the number of positrons generated per annihilation and energy interval ( for each branching channel ) and @xmath56 is the integrated squared density for each dm clump @xmath57 once we have the source term we proceed to calculate the flux at the earth by solving the steady state ( @xmath58 ) of eq.[diff - eq ] as already done e.g in @xcite or @xcite .
the eq.[diff - eq ] can be greatly simplified translating the energy into s(e ) .
@xmath59 the equation then reads : @xmath60 where @xmath61 and @xmath62 .
the solution , is given by the convolution of the green function over the sources : @xmath63 where the green function is given by : @xmath64}{d[e(s_0)]}\frac{e^{-(\vec{x}-\vec{x_0})^2/d_{diff}^{2}}}{[\pi d_{diff}^2]^{3/2}}\ ] ] as a function of the integrating variable @xmath65 .
@xmath66 is the diffusion distance scale given by @xmath67 in terms of the variable s(e ) with e the observed energy and @xmath68 the energy at injection .
as previously stated , pamela data favours leptophilic dm , ie , candidates which annihilation products are predominantely leptons . the case of direct annihilation into electron and positron pairs can provide good fits to the atic data but it is excluded if we take into account the fermi results due to a sharp drop at the end - point ( see e.g. @xcite ) . therefore we are left with annihilations into @xmath69 and @xmath70 which provide a much softer injection spectrum and can accomodate both pamela and fermi . for such injection spectrums , large annihilation rates are required
. the astrophysical boost factor , i.e , boost based on the presence of clumpiness , has been proved to be inssufficient to account for the required normalization to adjust the data . in this respect , proposals of particle physics boost factors as velocity dependent cross sections , have been recently considered .
+ in this work our purpose is to evaluate the expected anisotropy resulting from sources that reproduces the observed abundances , so we will assume the required normalization to fit the data as a result of any of the considered mechanisms that can boost the annihilation rate . here
, we consider a dm candidate that annilates into @xmath70 with a mass of 3.6 tev ( fig .
[ fig : fig1 ] ) .
[ fig : fig1 ] we also address the posibility of a nearby dm clump being the responsible of the bulk of positrons found in the fermi , atic and pamela data . for this purpose , we treat here the clump as a point - like object at a given distance _
d _ contributing to the source term with luminosity _
l _ in the same fashion as already done by @xcite .
this possibility has been found to be very unlikely unless the sommerfeld effect is at play , but such a source would imprint a signature in the electron arrival direction that could constitute a signal of dark matter annihilation given the absence of pulsars in the neighbourhood . in this sense ,
the evaluation of such a signal makes the study meaningful .
the expected flux for a clump of dm annihilating into @xmath70 or @xmath69 with masses around 3 tev and 2 tev respectively can reproduce the fermi data with the observed drop by hess , and with slightly different normalizations the pamela data . as a posibility
, we show the flux produced by the annilation in the @xmath70 channel of a dm clump located at 0.9 kpc with a dm mass of 3.6 tev . in this case , as in a pulsar scenario , a med diffusion setup is apropiate , as the min and max cases show prominent bumps in the spectrum at high and low energies respectively , that are not observed in the fermi data .
[ fig : fig2 ]
large scale anisotropies of cr have been measured so far to be less than 1% @xcite , but it is known that if it were produced by sources with some spatial structure , small anisotropies should be present in the arrival directions and could be correlated to the potential sources , whether it be known or not . in the case of high energy electrons , very light charged particles ,
diffusion competes with large energy losses resulting in relatively short paths @xmath71o(kpc ) , so it is expected that we can use them to sample only sources within a certain distance and age .
the cosmic ray intensity can , in general , be expanded over the celestial sphere in spherical harmonics . at first order ,
when we have a clear directionality , we have a dipole anisotropy as could be the case of a single source dominating the spectrum . in this case
, the intensity can be expressed as @xmath72 where @xmath73 being this maximum and minimum intensities related to a forward - backward measurement .
the calculation of the anisotropy produced by a single source has been carried out by @xcite and is given by : @xmath74 where d(e ) is de diffusion coefficient and n is the electron number density . in order to detect a statistically significant anisotropy , at the @xmath75 level , @xmath76 where @xmath77 is the number of events above a given energy threshold , that is a function of the detector acceptance and the exposition time .
@xmath78 .
estimates of the expected anisotropy in the case of a dominant pulsar have been previously shown in e.g. @xcite,@xcite but they fail to take into account the possible effects that could arise from a realistic collection of pulsars , there could be e.g. systematic cancellations .
we have calculated the anisotropy in the case of a collection of pulsars taken from the atnf catalogue and , for the first time , in the framework of dark matter annihilation of clumps thoughout the halo , both in the case of a dominating point - like source or a distribution of clumps . in the presence of an isotropic background plus a number of contributing sources to the total flux
, the expected dipole anisotropy is given by : @xmath79 where the index i runs over all the discrete sources that contribute to the full dipole .
the product @xmath80 represents the projection of the individual dipole over the direction of maximum intensity that is energy dependent and @xmath81 is the total flux observed at earth . the projection can be easily calculated taking into account the angular separation @xmath82 of two sources on the celestial sphere given by : @xmath83 where @xmath84 , @xmath85 and the right ascension and declination are denoted by @xmath86 and @xmath87 .
+ in this work , we evaluate the anisotropy ( fig .
[ fig : indivani ] ) produced by the collection of pulsars considered in fig .
[ pulsars ] ( a ) for a med diffusion setup .
+ the dipole anisotropy will change direction depending on the energy , but the main contributor to the full dipole is shown to be monogem above a few tens of gev up to the tev and in second place geminga . a big contribution to the anisotropy ( and the positron flux ) can be expected also at higher energies by younger objects like vela or cta1 , although it is still unclear if such an object can produce a sizeable amount of electrons due to the pwn confinement . the full dipole ( fig .
[ fig : fig3 ] ) will be given by the projection of the individual anisotropies in the direction of the maximum intensity at each energy , resulting in a clear signal in the direction of monogem at energies above 20 gev .
the contribution due to geminga has an addition effect to the full anisotropy due to the relative position between it and monogem , so , changes in their individual contributions to the flux , codified for instance in the pair conversion efficiency , should not change the anisotropic pattern .
contributions from other pulsars do nt result in a systematic addition of their signal to the full anisotropy into a particular direction , due to their spatial distribution , so they constitute a kind of isotropic background . in this scenario , measurements of a possible privileged incomimg direction should point out an excess in the monogem / geminga direction , roughly oposite to the direction of the mw .
it is possible however that we are observing the contribution of some yet undiscovered pulsars that could show up in a potential study of anisotropies . in this respect ,
searches for gamma ray sources as the one conducted by fermi will help to support or disfavour the test . (
a ) for a med diffusion setup.also shown the fermi sensitivity to such an anisotropy at the 2 and 5 @xmath88 cl in 5 years . ] for a med diffusion setup.also shown the fermi sensitivity to such an anisotropy at the 2 and 5 @xmath88 cl in 5 years . ] in the case of a number of clumps contributing to the full dipole , as in the case of diemands simulation , due to the symmetry of the clump distribution we would expect a dipole anisotropy in the direction of the milky way ( mw ) center .
the expected dipole anisotropy will be given by @xmath89 where @xmath90 is the projection of the clump contribution to the full dipole in the direction of the mw center and r,@xmath91 are the cylindrical coordinates of the clump distribution .
if we introduce the clump distribution we have used in the previous section , the expected dipole anisotropy for the direct annihilation channel will be : @xmath92 it must be noted that in case we have an initial source distribution f(e ) of electrons as the one that comes from annihilation through the @xmath93 channel , an equivalent result is obtained after the convolution of the source term with the @xmath94 term of the anisotropy definition . the resulting signal ( fig .
[ fig : fig5 ] ) for the parametrization considered in fig .
[ fig : fig1 ] , would not reach the fermi sensitivity for a 5 years survey as the fermi spectrum is too flat .
on the contrary , fits to the much bumpier atic data would provide a signature at the 2@xmath88 c.l . towards the mw center providing a hint to the origin of the positron excess .
+ we can also consider a dm point source as the responsible of the bulk of pamela electrons as could be the case of a large dm clump @xcite or minispikes around an intermediate mass black hole @xcite . in this case
the expected anisotropy is reduced to @xmath95 where @xmath96 being d the distance to the source ( fig .
[ fig : fig4 ] ) . +
although the probability of finding such a bright clump in our neighbourhood is rather small , there are some scenarios , with the sommerfeld effect at play , where this probability can be boosted up to a 15% . in this case
, the anisotropy would exceed the @xmath75 level pointing toward the existence of a dominant source .
this measurement should also be complemented with searches for gamma ray emission to achieve a consistent prediction .
.also shown the fermi sensitivity to such an anisotropy at the 2 and 5 @xmath88 cl in 5 years . ]
.also shown the fermi sensitivity to such an anisotropy at the 2 and 5 @xmath88 cl in 5 years . ]
a number of posibilities have been proposed as potential candidates to account for the positron excess . from the standard astrophysical point of view , pulsars seem to be the most promising candidates .
just considering the already known gamma - ray pulsars is enough to explain the spectrum for reasonable model assumptions , but the features in the atic data require quite a large conversion factor into pairs for the considered model . on the other hand ,
fermi data shows a much flatter spectrum that can be reproduced with a reasonable set of pulsar parameters .
if we consider dark matter annihilation , the contribution from the clumpy halo can reproduce the observed patterns once we have the required normalization issue solved by means of particle physics boosts on the thermal averaged cross section .
one way or the other , the measured spectrum can be reproduced and no clear signatures can be found to distinguish between a dark matter scenario or pulsars , as the spectral shape seems to lack information of the origin of the electrons . in order to distinguish between the proposed candidates , the study of anisotropies is proposed , as a not performed yet test , that can provide useful information . in this work
we have analysed the expected anisotropy from a configuration of pulsars as in the atnf catalogue .
we have also derived the anisotropy in a general dark matter scenario , both in case we have a very bright point source and in the case of a clumpy distribution as ilustrated by n - body simulations . for the dm point - like source and the pulsar scenarios ,
the fermi observatory should be able to detect a dipole anisotropy at 2 sigma cl in five years at least .
on the other hand , we would expect an excess towards the mw center in the case of a clumpy halo as the main contributor to the primary positron flux , but fits to the fermi spectrum does nt provide enough events in a 5 years survey .
in contrast , the case of the atic spectrum would open the posibility of direct annihilation into @xmath97 implying a much harder spectrum .
even annihilations through @xmath98 channels would imprint a signature in the electron anisotropy detectable by fermi that are not expected to be observed in view of the the fermi data .
+ anisotropies have been shown to contribute in a valuable way to disentangle the positron excess problem , nonetheless there are still a lot of theoretical uncertainties ( e.g. the dark matter halo distribution or the mechanism of pair production in pulsars ) .
it must be borne in mind that we have not taken into account the proper motion of the pulsars or even that of the dark matter clumps , that could result in an enhancement or suppression of the anisotropy . moreover ,
the case of a dipole anisotropy toward a pulsar can not exclude a dark matter scenario , as it is possible to have a large dark matter cloud in the same direction masking the signal . in any case
, the precise study of electron anisotropies should be conducted together with pulsar surveys to help to discriminate between astrophysical and more exotic sources .
i would like to acknowledge the support from javier berdugo , carlos ma , jorge casaus and carlos delgado for their useful suggestions and endless patience and aurelio carnero for the proofreading .
i would also like to thank d.grasso , d.gaggero and g.dibernado for their useful remarks and the anonymous referee for his helpful comments .
this work is partially supported by the centro de investigaciones energeticas , medioambientales y tecnologicas ( ciemat ) and the micinn fpi grant number bes-2007 - 151967 .
observation of an anomalous positron abundance in the cosmic radiation boezio m. on behalf of pamela s collaboration nature , arxiv:0810.4995v1 astro - ph j. chang et al . ,
nature 456 , 362 ( 2008 ) .
s. torii et al . , submitted to astropart .
phys arxiv:0809.0760 [ astro - ph ] .
measurement of the cosmic ray e+ plus e- spectrum from 20 gev to 1 tev with the fermi large area telescope , fermi / lat collaboration , arxiv:0905.0025 .
s. w. barwick et al .
[ heat collaboration ] , astrophys .
j. 482 , l191 ( 1997).[arxiv : astro - ph/9703192 ] ; j. j. beatty et al .
93 , 241102 ( 2004)[arxiv : astro - ph/0412230 ]
. m. aguilar et al .
[ ams-01 collaboration ] , phys .
b 646 , 145 ( 2007 ) [ arxiv : astro - ph/0703154 ] . galactic secondary positron flux at the earth t. delahaye , f. donato , n. fornengo , j. lavalle , r. lineros , p. salati , and r. taillet,[arxiv:0809.5268 ] .
model - independent implications of the e+ , e- , anti - proton cosmic ray spectra on properties of dark matter marco cirelli , mario kadastik , martti raidal , alessandro strumia [ arxiv:0809.2409v2 ] .
the origin of cosmic rays.v.l .
ginzburg and s.i .
syrovatskii , pergaemon , oxford,1964 .
cosmic - ray propagation and interactions in the galaxy andrew w. strong , igor v. moskalenko , vladimir s. ptuskin , [ astro - ph/0701517 ] .
positrons from particle dark - matter annihilation in the galactic halo : propagation green s functions
. i. v. moskalenko , a. w. strong , phys.rev . d60 ( 1999 ) 063003 .
positron propagation and fluxes from neutralino annihilation in the halo .
baltz , e. a. & edsjo , j. 1999 , phys .
d , 59 , 023511 a cosmic - ray positron anisotropy due to two middle aged nearby pulsars , i.busching , o.c
. de jager , m.s .
potgieter and c. venter , the astrophysicaljournal,678:l39-l42,2008 may .
pulsars as the sources of high energy cosmic ray positrons , dan hooper , pasquale blasi and pasquale dario serpico , fermilab - pub-08 - 429-a .
tev gamma rays from geminga and the origin of the gev positron excess , [ arxiv:0810.2784v3 ] .
a. k. harding and r. ramaty , in international cosmic ray conference ( 1987 ) , vol . 2 of international cosmic ray conference , pp .
x. chi , k. s. cheng , and e. c. m. young , apj 459 , l83 + ( 1996).m possible contribution of mature gamma ray pulsars to cosmic ray positrons .
( 2008),no 1 , 87 - 95 .
dissecting pamela ( and atic ) with occam s razon : existing , well known pulsars naturally account for the `` anomalous '' cosmic - ray electron and positron data profumo stefano [ arxiv:0812.4457 ] . on posible intepretations of the high energy electron - positron spectrum measured by the fermi large area telescope .
d. grasso , s. profumo , a.w .
strong , l. baldini , r. bellazzini , e.d .
bloom , j. bregeon , g. di bernardo , d. gaggero , n. giglietto , t. kamae , l. latronico , f. longo , m.n .
mazziotta , a.a .
moiseev , a. morselli , j.f .
ormes , m. pesce - rollins , m. pohl , m. razzano , c. sgro , g. spandre , t.e .
stephens [ arxiv:0905.0636 ] l.zhang and k.s.cheng,a&a 368,1063 ( 2001 ) . the atnf pulsar catalogue , manchester , r. n. , hobbs , g. b. , teoh , a. and hobbs , m. , astron . j. , 129 , 1993 - 2006 ( 2005 ) [ astro - ph/0412641 ] .
l. zhang and k.s .
cheng , astrophys.j .
487,370 ( 1997 ) .
cosmic - ray electron excess from pulsars is spiky or smooth ?
continuous and multiple electron / positron injections.norita kawanaka , kunihito ioka and mihoko m. nojiri.[arxiv : astro - ph/0903.3782v1 ] .
electrons and positrons in the galactic cosmic rays , a. m. atoyan , f. a. aharonian and h. j. volk , phys.rev .
d 52 , 3265 , ( 1995 ) .
pulsars versus dark matter interpretation of atic / pamela , d.malyshev,i .
cholis and j. gelfand 2009.[arxiv:0903.1310 ] .
juerg diemand , ben moore and joachim stadel , 2005 , nature 433,389 .
local dark matter clumps and the positron excess , daniel cumberbatch and joseph silk , mon.not.roy.astron.soc.374:455-465,2007 [ arxiv : astro - ph/0602320v3 ] dark matter interpretations of the electron / positron excesses after fermi .
patrick meade , michele papucci , alessandro strumia , tomer volansky .
ifup - th-2009 - 9 , may 2009 .
[ arxiv:0905.0480v1 [ hep - ph ] ] the cosmic ray lepton puzzle in the light of cosmological n - body simulations , pierre brun , timur delahaye , jurg diemand , stefano profumo and pierre salati [ arxiv:0904.0812v2 ] anisotropy and corotation of galactic cosmic rays , amenomori et al .
science ( 2006 ) vol . 314 .
5798 , pp .
439 - 443 .
anisotropy and diffusion of cosmic ray electrons , mao , c.y . , and shen , c.s chinese journal of physics , vol 10 , no1 , april 1972 . the pamela and atic excesses from a nearby clump of neutralino dark matter , dan hooper , albert stebbins and kathryn m.zurek . , [ arxiv : hep - ph/0812.3202 ] .
intermediate mass black holes and nearby dark matter point sources : a myth - buster , torsten bringmann , lulien lavalle and pierre salati , [ arxiv : astro - ph/0902.3665v1 ] . measuring 10 - 1000 gev cosmic ray electrons with glast / lat , alexander a.moiseev , jonathan f.ormes , igor v.moskalenko,proc .
30th icrc ( merida ) , 2 , 449 - 452 ( 2007),[arxiv:0706.0882v1 ] | recent results from the pamela @xcite , atic @xcite , ppb bets @xcite and fermi @xcite collaborations extend the energy range in the @xmath0 measurement up to unexplored energies in the hundred gevs range confirming the bump starting at about 10gev already suggested by heat @xcite and ams01 data @xcite . this bump can be explained by annihilating dark matter in the context of exotic physics , or by nearby astrophysical sources e.g. pulsars . in order to discriminate between competing models for primary positron production ,
the study of anisotropies , complementary to the spectrum determination , shows up as new tool to look for the origin of the lepton excess . in this letter
we calculate the contribution to the electron flux given by the collection of all known gamma ray pulsars ( as listed in the atnf catalogue ) and by annihilating dark matter both in case we have a clumpy halo or in case the excess can be atributed to a nearby sizeable dark matter clump .
we address the problem of the electron anisotropy in both scenarios and estimate the prospect that a small dipole anisotropy can be found by the fermi observatory .
dark matter , anisotropy , pamela , atic , fermi / lat |
in order to understand the basic mechanisms of a growth process , it is a common practice to associate it to a certain universality class @xcite .
this is usually done with the calculation of scaling exponents of surface roughness and assuming that family - vicsek scaling is valid @xcite ( with a suitable generalization when anomalous scaling holds ) .
however , huge corrections to scaling make this calculation difficult in finite - size systems , both for simulation work and for the analysis of images of real surfaces .
these problems and the interest in finding more universal properties of growth processes motivated recent studies of other quantities , such as the distributions of surface roughness @xcite .
most efforts were devoted to calculation of those distributions in steady states ( roughness saturation ) , but the comparison of distributions obtained under conditions that parallel experimental work ( the so - called window boundary conditions ) is also a promising tool @xcite . it is expected that roughness distributions scale as @xmath3 where @xmath4 is the probability density of the roughness @xmath5 of a given configuration to lie in the range @xmath6 $ ] , @xmath7 is a scaling function , the overbars denote spatial averages and the angular brackets denote configurational averages .
( [ deff ] ) is usually obtained in analytical work , such as in the calculation of distributions for @xmath8-noise interfaces @xcite .
an alternative scaling relation is @xmath9 where @xmath10 is the root mean square ( rms ) deviation of @xmath11 .
it is reasonable to claim that comparison of full roughness distributions is better than the calculation of one or two exponents to determine the class of a growth process .
however , these distributions may also be deviated from the expected scaling relations by finite - size corrections .
for instance , finite - size data of some models in the kardar - parisi - zhang ( kpz ) class @xcite show small deviations from relation ( [ deff ] ) , while good data collapse of numerical data for different lattice sizes is obtained with relation ( [ defg ] ) @xcite .
the advantage of a scaling relation equivalent to ( [ defg ] ) was also illustrated in the study of maximal height distributions ( mhd ) of @xmath8 signals @xcite .
moreover , the leading corrections to scaling in mhd of some one - dimensional solid - on - solid models were analytically predicted and numerically confirmed in ref .
@xcite , which clearly shows that such corrections are not related to low accuracy of simulation data .
the presence of huge scaling corrections may be much worse in distributions of other quantities , such as height distributions @xcite . for the above reasons , investigation on the failure of finite - size scaling of roughness distributions
is essential to ensure their reliability for comparison with real or model systems .
moreover , understanding the reasons for the failure of an scaling relation and the advantages of other ones may also be helpful to find the form of the main finite - size corrections and to propose alternative methods of analysis .
one particularly important case is kpz growth in @xmath1 dimensions , since this class contains many models with large corrections to scaling of the average roughness , such as ballistic deposition ( bd ) and eden growth , as well as a large number of applications ( see e. g. refs .
@xcite ) .
the aim of this work is to study finite - size corrections in roughness distribution scaling by analyzing numerical simulation data of various interface growth models and to discuss the possible relations with other quantities that characterize those interfaces .
the numerical approach is essential to study the three - dimensional systems considered here , as well as some two - dimensional systems ( the usual systems where analytic work is possible are gaussian interfaces ) . first , we will study a group of ballistic - like models where the deviations from relation ( [ deff ] ) are much larger than those previously found in other models @xcite .
this means that no evidence on their universality class can be obtained by assuming that scaling relation . on the other hand , good collapse of large systems data
is obtained with eq .
( [ defg ] ) , which shows that they are actually in the kpz class . moreover
, we successfully propose an alternative scaling relation accounting for the effect of the intrinsic width , which is a constant correction term previously included in the scaling of the average roughness @xcite .
this shows how the main finite - size corrections may be extracted from roughness distribution scaling .
however , we discard the usual interpretation that the intrinsic width is a consequence of high surface steps by analyzing data of restricted solid - on - solid ( rsos ) models , where this correction is not present . instead , we observe that the deviations from finite - size scaling in the roughness distributions are usually accompanied by significant finite - size dependence of dimensionless amplitude ratios of moments of height distributions , as well as of scaling exponents and average local slopes .
this conclusion is not restricted to kpz systems , as shown in the analysis of the molecular - beam epitaxy model of das sarma and tamborenea ( dt ) @xcite in @xmath2 dimensions , which is well - known for the slow convergence of numerically estimated exponents @xcite . indeed , for this model none of the proposed scaling relations
is able to produce good collapse of small systems data .
although we are not able to determine all the reasons for the failure of scaling relations in a given universality class , we will show how deviations in different quantities are connected and will suggest alternative methods of analysis .
this may be helpful not only in future work on complex models and real systems , but also to search for models with minimal finite - size effects , e. g. following the ideas of ref .
anyway , an important conclusion is that roughness distributions are actually superior to other quantities for a reliable search of the universality class of a growing system , although it is not the final solution for all models .
the rest of this work is organized as follows . in sec .
ii we discuss the above scaling relations and propose an alternative one for cases where intrinsic width is expected . in sec .
iii we present the discrete models analyzed in the subsequent sections and the equations defining their universality classes . in secs .
iv , v and vi , we discuss the finite - size effects in the roughness distributions , the role of local slopes and the finite - size effects in the height distributions , respectively . in sec .
vii we summarize our results and present our conclusions .
if the scaling relation ( [ deff ] ) holds , the function @xmath12 is a function of the variable @xmath13 .
the moment of order @xmath14 of this function is defined as @xmath15 , so that the first moments are given by @xmath16 this means that any finite - size dependence of the ratio @xmath17 ( between the average square roughness and its rms fluctuation ) will lead to a finite - size dependence of the second moment . however , in the scaling relation ( [ defg ] ) , @xmath18 is a function of the variable @xmath19 , so that the first moments of @xmath20 are @xmath21 this means that the finite - size dependence of @xmath22 do not lead to deviations in the second moment of this scaling function , leaving possible corrections to the third moment . in works based on data collapse methods ,
the fluctuations in the first few moments of a scaling function are the ones most easily related to visual deviations of the plots of data from different system sizes . consequently , scaling relation ( [ defg ] ) is expected to be better than ( [ deff ] ) by avoiding that finite - size effects are reflected in its second moment .
this was already observed for some kpz models in @xmath1 dimensions in relatively small system sizes @xcite .
recent work on maximal height distributions of @xmath8 signals also illustrated this feature and stressed the fact that finite - size scaling of the higher cumulants was related to the deviations @xcite .
the knowledge of the particular properties of a growth model is certainly important to guide the proposal of suitable scaling relations .
for instance , this was the case of a recent work on elastic lines in random environments , in which the scaling relation ( [ deff ] ) was generalized to incorporate time and temperature effects @xcite . here
, we will consider a series of ballistic - like models whose deposits contain holes and overhangs , among other steep features , thus our approach is based on properties of related models
. simulation work on models which produce deposits with holes and overhangs , such as the eden model , suggested that the leading correction to the scaling of the average squared roughness is a constant term @xcite .
the time evolution of the average squared roughness , starting from a flat interface , obeys @xmath23 where @xmath24 is a constant called the intrinsic width , @xmath25 is the roughness exponents and @xmath26 is the dynamic exponent .
the usual interpretation of relation ( [ fv ] ) is that the first term at the right - hand side ( the original family - vicsek relation @xcite ) is associated with fluctuations of all wavelengths , while the intrinsic width is thought to be related to short wavelength features , represented by overhangs , holes and high steps at the film surface - see e. g. refs .
@xcite .
the present work is only concerned with steady state properties , where @xmath27 in eq .
( [ fv ] ) and the saturation roughness [ @xmath28 scales as @xmath29 with @xmath30 constant .
large values of @xmath24 are responsible for remarkable deviations of finite - size estimates of exponent @xmath25 from its asymptotic value .
however , numerical work on ballistic - like models show that finite - size corrections to the scaling of @xmath31 are smaller than the corrections in @xmath0 @xcite .
these results lead us to propose that , for the ballistic - like models , the roughness of a given interface configuration is a sum of a fluctuating value , which is distributed according to the universality class of the model , and a constant contribution , which is the intrinsic width . following this reasoning , we change relation ( [ deff ] ) by subtracting a constant from @xmath11 and @xmath0 : @xmath32 here @xmath33 must be viewed as a fitting constant and @xmath34 is a scaling function .
the first moments of the function @xmath35 , where @xmath36 , are @xmath37 if our assumptions are correct and @xmath33 is chosen to match the intrinsic width @xmath24 , then the finite - size dependence of the second moment @xmath38 will be cancelled , or at least it will be greatly reduced .
consequently , the quality of a data collapse plot will be similar to that based on eq .
( [ defg ] ) .
certainly this proposal has to be validated by numerical ( or eventually analitic ) work , and it is also important to notice that it is independent on the particular interpretation given to the intrinsic width .
the discrete models analyzed in this paper are the original bd model , the rsos model , a conserved rsos model , the grain deposition models and the dt model . in all cases
, deposition begins with a flat substrate . in bd
, each incident particle is released from a randomly chosen position above the deposit , follows a trajectory perpendicular to the substrate and sticks upon first contact with a nearest neighbor occupied site @xcite . in the rsos model @xcite
, the incident particle can stick at the top of a column only if the differences of heights of all pairs of neighboring columns do not exceed @xmath39 after aggregation .
otherwise , the aggregation attempt is rejected . in this work
, results for @xmath40 , @xmath41 and @xmath42 will be presented .
the original conserved rsos model ( crsos ) was proposed in ref .
@xcite , but here we will consider an extension of that model that has the same symmetries and , consequently , belong to the same universality class . in this generalized crsos model @xcite , hereafter simply called crsos , if the column of incidence does not obey the above condition for aggregation , with @xmath40 , then the incident particle executes random walks among neighboring columns until finding a position where it is satisfied and aggregation can occur . in the @xmath43 model @xcite ,
the arriving particle sticks at the top of the column of incidence if it has one or two lateral neighbors at that position .
otherwise , the neighboring columns are consulted and if one of them satisfies that condition , then the incident particle aggregates at that point . however , if no neighboring column satisfies the condition , then the particle sticks at the top of the column of incidence , and if both neighboring columns satisfy the condition , then one of them is randomly chosen .
the grain deposition models studied here were introduced in ref .
@xcite for the study of a crossover in local roughness scaling similar to experimental systems .
they are defined in a simple cubic lattice where the length unit is the lattice parameter .
the grains have cubic shapes and lateral size @xmath44 .
they sequentially incide perperdicularly to an initially flat substrate , with two of their faces parallel to the substrate .
the incident grain permanently aggregates to the deposit when its bottom touches a previously aggregated grain ( thus , there is no lateral aggregation ) .
the process is shown in fig . 1
. here we will simulate the model with grain sizes ranging from @xmath45 to @xmath46 . despite the different aggregation rules , the grain deposition models and the bd model form the group to which hereafter we will refer as ballistic - like models .
the symmetries of the ballistic - like models and of the rsos models indicate that they are in the kpz universality class ( see e. g. the discussion in ref .
@xcite , in which the continuous description of the growth process is provided by the kpz equation @xmath47 here , @xmath48 and @xmath49 are constants , @xmath50 is a gaussian noise with zero mean and variance @xmath51 , with @xmath52 constant , and @xmath53 is the dimension of the substrate . in the following sections , we will study those models in @xmath1 dimensions ( in @xmath2 dimensions , some exact results and highly accurate numerical results are already available ) .
the dt and the crsos models in @xmath2 dimensions are in the class of the nonlinear fourth order growth equation @xcite @xmath54 where @xmath55 and @xmath56 are constants .
this is also known as villain - lai - das sarma ( vlds ) equation @xcite , and the discrete models are said to belong to the vlds class .
the dt model is particularly interesting in @xmath2 dimensions due to the presence of strong finite - size corrections which leads to exponents estimates far from the vlds values ( although noise reduction methods are able to provide estimates closer to the vlds ones @xcite ) . for all models , only steady state properties will be studied in this work .
the roughness distributions are typically obtained from @xmath57 to @xmath58 different configurations , and height distributions from a number of configurations which is larger by a factor @xmath59 ( the square of the lattice size ) .
in order to compare the roughness distributions of the above models , we use two of them as representatives of their growth classes : the rsos model with @xmath40 , whose distribution hereafter will be referred as the rsos / kpz one ( @xmath1 dimensions , @xmath60 ) , and the crsos model with @xmath61 , whose distribution hereafter will be referred as the crsos / vlds one ( @xmath2 dimensions , @xmath60 ) .
the use of these models data as standards is possible because previous work have already shown that they have very small finite - size dependence @xcite .
the original bd model illustrates the need of choosing a suitable scaling relation for the analysis of roughness distributions from small system sizes . in figs .
2a , 2b and 2c we show its steady state distributions in @xmath1 dimensions obtained from @xmath62 to @xmath63 and scaled according to eqs .
( [ deff ] ) , ( [ defg ] ) and ( [ defh ] ) , respectively . for comparison ,
the rsos / kpz scaled distributions are also shown . in fig . 2c
, @xmath64 is used to match the peaks of the curves for bd and for rsos / kpz .
it is clear that the scaling with the rms fluctuation @xmath31 ( eq . [ defg ] ) is superior if compared with the scaling with eq .
( [ deff ] ) .
indeed , there is no data collapse in fig .
2a , thus that plot is unable to confirm the kpz scaling of the bd model . from the discussion of sec .
2 , the deviations may be related to finite - size corrections in the ratio @xmath22 , and our estimates of @xmath22 for different sizes @xmath65 confirm this large size - dependence .
however , the quality of the data collapse using eq .
( [ defh ] ) is similar to the one using eq .
( [ defg ] ) .
this result suggests that the assumptions leading to the proposal of eq .
( [ defh ] ) are reasonable , i. e. the roughness of each configuration is a constant plus a kpz - distributed fluctuating part ( sec .
it also gives an estimate of the intrinsic width , which is the constant @xmath64 ( fig .
the scaled distributions of the grain deposition model with @xmath46 in lattice sizes @xmath66 and @xmath67 are shown in figs .
3a , 3b and 3c , considering eqs .
( [ deff ] ) , ( [ defg ] ) and ( [ defh ] ) , respectively . the rsos / kpz distribution is also shown .
deviations from the scaling relation ( [ deff ] ) are also large here , but using eqs .
( [ defg ] ) and ( [ defh ] ) we obtain good data collapse with those large lattice sizes .
similar results were obtained with an extended model where the grains could have poisson - distributed sizes , with an average size @xmath68 ( the aggregation mechanism was the same of fig .
1 ) . in bd and in this grain deposition model
we observe deviations from data collapse with sizes smaller than those shown in figs .
the deviations for a certain lattice size @xmath65 become smaller when the grain size decreases , thus good data collapse is obtained for intermediate values of @xmath44 ( @xmath45 to @xmath69 ) with lattice sizes @xmath65 intermediate between those in figs .
2a - c and 3a - c .
typically , significant deviations appear only for @xmath70 ( @xmath71 for bd ) , probably due to further corrections to scaling , whose effects are reduced as @xmath65 increases .
our results for this group of ballistic - like models suggest that the roughness is actually distributed in a kpz - like form except for an additional constant , which may be interpreted as the intrinsic width
. additional support to the constant correction term in the scaling of @xmath0 ( eq . [ fvsat ] ) is given in fig .
4 , where we plot @xmath72 versus @xmath65 , using @xmath73 @xcite and the values of @xmath33 obtained from collapse of roughness distributions .
the saturation value of @xmath74 for each model gives an estimate of the amplitude @xmath30 in eq .
( [ fvsat ] ) .
these estimates and the correction constants @xmath33 are shown in table i , where we also present the corresponding values of both terms on the right - hand side of eq .
( [ fvsat ] ) for the largest lattice size used for each model ( @xmath75 ) .
the ratios between the leading correction ( @xmath33 ) and the dominant term @xmath76 are very large for all model , in some cases exceeding @xmath77 , which quantitatively confirm the strong finite - size effects . however , it is important to notice that the scaling with eq .
( [ defh ] ) advances over eq .
( [ defg ] ) because it not only confirms the universality class of the model ( by data collapse ) but also provides evidence on the leading finite - size correction to the roughness scaling .
these results also advance over previous work where the intrinsic width was identified only as a correction to scaling of the average roughness @xcite .
the intrinsic width is usually associated to steep surface features , such as large local slopes , holes and overhangs . in order to test this hypothesis
, it is interesting to consider other models which also have large local slopes .
this is the case , for instance , of the rsos models with large @xmath78 . in figs . 5a and
5b we show the roughness distributions for the rsos models with @xmath41 and @xmath42 scaled according to eqs .
( [ deff ] ) and ( [ defg ] ) , respectively .
both plots show excellent data collapse with very small lattice sizes , such as @xmath79 .
thus , there is no correction term similar to the intrinsic width or , equivalently , @xmath80 in eq .
( [ defh ] ) for the rsos models .
finally , we analyze the dt model in @xmath2 dimensions . in figs .
6a and 6b we show the roughness distributions for that model in lattice sizes @xmath81 and @xmath62 , scaled according to eqs .
( [ deff ] ) and ( [ defg ] ) .
no significant finite - size effect is found in the scaling with the fluctuation @xmath31 ( fig .
however , both plots show that the distribution of the dt model is very different from the crsos / vlds distribution in @xmath2 dimensions ( @xmath60 ) @xcite .
moreover , there is no evidence that the former will converge to the crsos / vlds curve as @xmath65 increases .
we observe the same large deviations when the dt distributions are scaled according to eq.([defh ] ) for several values of the constant @xmath33 .
thus , the assumption of an intrinsic width is not sufficient to represent the main finite - size effects in that model , at least in the accessible range for simulation , which is @xmath82 .
however , previous work showed that simulation of noise - reduced dt models provides estimates of scaling exponents closer to the vlds values @xcite .
thus , we also calculated the dt distributions using noise - reduction parameters @xmath83 , @xmath84 and @xmath85 . here , @xmath86 is the number of times that a certain column has to be chosen for aggregation of a new particle before the aggregation actually takes place @xcite . in fig .
6c we show the scaled distributions for these models in @xmath81 , which converge to the crsos / vlds curve as @xmath86 increases ( although reasonable data collapse is not obtained in the tails ) .
this noise - reduction scheme is not expected to change the universality class of the process because symmetries are preserved ( see e. g. the discussion in ref .
thus , the above changes in the distributions when the parameter @xmath86 is increased ( fig .
6c ) are evidence that the asymptotic behavior of the dt distribution was not attained yet .
consequently , the above discrepancies with the crsos / vlds curve do not mean that the dt model does not belong to the vlds class , but only that the finite - size effects on the roughness distributions have much more complex forms than those suggested here .
indeed , we recall that ref . @xcite derived the vlds equation from the master equation of the dt model in @xmath2 dimension . in the following sections
, we will show that finite - size corrections obtained here in the roughness distributions are accompannied by corrections in other quantities .
here we characterize the interfaces at small lengthscales by the average squared height difference between nearest neighbor columns , @xmath87 , also calculated at the steady states .
our main aim is to test the hypothesis that the intrinsic width is related to local height fluctuations of the interface . in the ballistic - like models
, @xmath88 has a significant finite - size dependence for small lattice sizes .
however , it is approximately constant in the largest lattice sizes analyzed for each @xmath44 in sec .
iii ( the same sizes used to collapse the roughness distributions ) .
however , in the rsos models @xmath88 attains a saturation value for very small system sizes . in fig .
7 we plot the correction term @xmath33 obtained in the scaling with eq . [ defh ] ( i. e. the intrinsic width ) as a function of the large @xmath65 value of @xmath88 for both groups of models .
the result for the ballistic - like models could naively suggest that the intrinsic width was only related to local height fluctuations .
however , the result for the rsos models show this interpretation is not valid in general : @xmath80 in that case and the values of @xmath88 are in the same range of those for ballistic - like models with @xmath89 .
in the dt model in @xmath2 dimensions , we observe a rapid increase of the average local slope with the lattice size : @xmath90 for @xmath79 , @xmath91 for @xmath81 , and @xmath92 for @xmath62 . a comparison with @xmath88 for the other models shown in fig .
7 is not possible because we are not able to perform a reliable extrapolation of these data to @xmath93 . on the other hand ,
these data confirm that the dt model is very far from its asymptotic limit ( the continuum vlds model ) in this range of lattice sizes .
the above results suggest that the presence of large fluctuations at small lengthscales is insufficient to explain the intrinsic width as the main correction term of roughness scaling in the ballistic - like models , since this is also a property of models which have much weaker scaling corrections ( the rsos ones ) .
on the other hand , we observe that finite - size effects on roughness distributions are accompannied by the same type of effect on @xmath87 .
finite - size effects in height distributions are typically larger than those in roughness distributions , as illustrated for kpz models in ref . @xcite . thus , in order to study finite - size effects on steady state height distributions , the best procedure is to compare their skewness @xmath94 and kurtosis @xmath95 in different lattice sizes ( other previous works which aimed at characterizing height distributions also focused on those quantities - see e.g. refs .
they are defined from the moments of the height distributions , @xmath96 , as @xmath97 and @xmath98 the sign of the skewness indicates which type of singular feature is dominant at the interface : deep valleys or grooves lead to negative @xmath94 , while sharp peaks give positive @xmath94 . in kpz models the sign of @xmath94 is the same of the coefficient of the nonlinear term @xmath99 of the corresponding kpz equation ( see e. g. the discussion in ref .
@xcite ) . thus ,
since ballistic - like models correspond to positive @xmath99 and rsos models to negative @xmath99 , we must compare values of @xmath94 from the former with @xmath100 from the latter . in figs .
8a and 8b we show @xmath101 and @xmath95 versus @xmath102 , respectively , for bd , for the grain deposition model with @xmath69 and for the rsos model with @xmath41 ( @xmath100 only for the rsos model ) .
data for bd in the largest sizes were extracted from ref .
@xcite . in figs . 8a and 8b ,
the variable @xmath102 was used in the abscissa because it is the one which provides the best linear fits among other variables of the form @xmath103 , with one digit variation in @xmath104 .
since the asymptotic values of @xmath94 and @xmath95 are not known exactly , as well as their finite - size corrections , such numerical extrapolations are necessary .
we observe very weak finite - size corrections in the rsos data , so that a simple extrapolation procedure to @xmath93 ( @xmath105 ) leads to an asymptotic estimate of @xmath94 in agreement with the best known value @xmath106 of the kpz class @xcite . however , there are large finite - size corrections in the data for the bd model .
it is remarkable that up to @xmath107 , the skewness of bd is negative , while the values for larger @xmath65 are positive .
the finite - size dependence of the data for the grain deposition model is much stronger : only for @xmath108 the skewness crosses over from negative to positive values .
the change in sign of @xmath94 of the ballistic - like models shows that the typical steady state configurations for small @xmath65 are very different from the typical ones for large @xmath65 .
in other words , the steady states for small @xmath65 are not representative of the asymptotic limit of the model ( large @xmath65 and @xmath109 ) , which is the kpz limit .
for the grain deposition model with @xmath46 , even the steady state for @xmath67 has @xmath110 and , consequently , is very far from the kpz limit .
despite these problems , the data for the bd ballistic - like models indicate that @xmath94 is positive when @xmath93 .
asymptotic estimates ( not very accurate ) of @xmath94 can be obtained from extrapolations of those data ; they are not very distant from the best known values for kpz in @xmath1 dimensions
. however , the finite - size effects on roughness distributions are certainly smaller than those found in height distributions . indeed , the good data collapse in figs .
2b , 2c , 3b and 3c is reflected in values of skewness ( of the roughness distributions ) close to the rsos / kpz curve . the analysis of the kurtosis of the above models ( fig .
8b ) leads to the same conclusions , although there the finite - size effects are much larger , as well as the error bars of the data .
the values of @xmath94 and @xmath95 for the dt and the crsos models are shown in figs . 9a and 9b as a function of @xmath102 .
they were estimated up to @xmath60 for the dt model ( this work ) and up to @xmath66 for the crsos model ( ref .
@xcite ) , with a relatively low accuracy for the largest @xmath65 data . here
the variable in the abscissa was chosen by analogy with fig .
8 , but not for extrapolation purposes .
the asymptotic skewness of the vlds class , @xmath111 , can be obtained from extrapolation of the crsos data to @xmath93 @xcite .
however , it is clear that the data for the dt model in fig .
9a do not converge to the same value , even if the sign of @xmath94 is changed ( the actual sign of @xmath94 is related to that of @xmath56 in eq .
[ vlds ] ) .
there is also no evidence that the kurtosis of the dt model will converge to the same value of the crsos model .
thus , the height distributions of the dt model up to @xmath60 are not representative of the vlds class , even if we account for simple finite - size corrections .
the overall conclusion from the above results is that small or large finite - size corrections are simultaneously present in roughness and height distributions , but the convergence of roughness distributions to the asymptotic limit is certainly much better when it occurs , as illustrated with the ballistic - like models .
the amplitude ratios of height distributions are interesting to show that , when the corrections are large , the finite - size configurations of hills and valleys are not representative of the universality class of the model .
we are not able to explain why the roughness distributions ( for large @xmath65 ) represent the correct class of the model when the height distributions do not , but certainly there is no contradiction in this finding because we are comparing distributions of local and global quantities , whose scaling properties may be very different .
for various interface growth models , height and roughness distributions and average square neighboring heights differences @xmath112 were numerically calculated in the steady states , in order to analyze their finite - size effects .
first , we considered a group of ballistic - like deposition models in which the most typical scaling relation for roughness distributions ( eq . [ deff ] ) fails to provide collapse of finite - size systems data . on the other hand ,
good data collapse was obtained with a scaling relation involving the roughness fluctuation ( eq .
[ defg ] ) and with an alternative relation which includes the effect of the intrinsic width .
this shows that this constant term is the main correction to the asymptotic kpz distribution , thus extending previous work which suggested that quantity only as a correction to the scaling of the average roughness .
the comparison with other models with large @xmath113 ( the rsos ones ) show that the average local slope is not sufficient to explain the intrinsic width , in contrast to previous interpretation .
height distributions for those models also show strong finite - size effects , although results for sufficiently large lattice sizes indicate convergence of amplitude ratios to the expected asymptotic values .
we also studied the model of das sarma and tamborenea ( dt ) in @xmath2 dimensions .
its roughness distributions up to size @xmath62 are very different from the ones representative of the asymptotic vlds class .
strong finite - size effects are also observed in skewness and kurtosis of height distributions and in @xmath88 , with no evidence of convergence to reliable asymptotic values .
our results show that the deviations from finite - size scaling in the roughness distributions are usually accompanied by significant finite - size dependence of dimensionless amplitude ratios of moments of height distributions and of average local slopes .
one important point is that this conclusion is not restricted to the kpz class .
the analysis of simple quantities , such as @xmath113 , is useful to search for finite - size effects and may reveal slow convergence to a continuum limit .
in some situations ( e. g. the dt model here ) , this convergence may be so slow that no one of the quantities analyzed here are able to show the correct class of the growth process . in the other cases , roughness distributions are certainly better than other quantities ( including scaling exponents ) to classify the process .
( circles ) , @xmath60 ( squares ) and @xmath63 ( triangles ) , scaled according to : ( a ) eq .
( [ deff ] ) , ( b ) eq .
( [ defg ] ) and ( c ) eq .
( [ defh ] ) . in each plot , the solid curve is the rsos / kpz distribution scaled accordingly.,width=377 ] and lattice sizes @xmath66 ( squares ) and @xmath67 ( triangles ) , scaled according to : ( a ) eq .
( [ deff ] ) , ( b ) eq .
( [ defg ] ) and ( c ) eq .
( [ defh ] ) . in each plot , the solid curve is the rsos / kpz distribution scaled accordingly.,width=377 ] versus lattice size @xmath65 for the ballistic like models , where @xmath33 is the constant used to fit roughness distributions to the new scaling relation .
symbols correspond to the ballistic deposition model ( squares ) and the grain deposition models with @xmath45 ( crosses ) , @xmath114 ( diamonds ) , @xmath69 ( circles ) and @xmath46 ( triangles).,width=377 ] ( squares ) and @xmath62 ( triangles ) scaled according to eq .
( [ deff ] ) and eq .
( [ defg ] ) , respectively ; ( c ) distributions for the original and the noise - reduced dt models in lattice size @xmath81 : @xmath116 ( original model - dashed line ) , @xmath83 ( crosses ) , @xmath84 ( squares ) and @xmath85 ( triangles ) . in all plots ,
the solid curve is the crsos / vlds distribution scaled accordingly.,width=377 ] ( a ) and kurtosis @xmath95 ( b ) of height distributions for the rsos model with @xmath117 ( crosses ) , the bd model ( squares ) and the grain deposition model with @xmath69 ( triangles ) .
@xmath100 is plotted only for the rsos model.,width=377 ] | we study numerically finite - size corrections in scaling relations for roughness distributions of various interface growth models . the most common relation , which considers the average roughness @xmath0 as scaling factor , is not obeyed in the steady states of a group of ballistic - like models in @xmath1 dimensions , even when very large system sizes are considered .
on the other hand , good collapse of the same data is obtained with a scaling relation that involves the root mean square fluctuation of the roughness , which can be explained by finite - size effects on second moments of the scaling functions .
we also obtain data collapse with an alternative scaling relation that accounts for the effect of the intrinsic width , which is a constant correction term previously proposed for the scaling of @xmath0 .
this illustrates how finite - size corrections can be obtained from roughness distributions scaling .
however , we discard the usual interpretation that the intrinsic width is a consequence of high surface steps by analyzing data of restricted solid - on - solid models with various maximal height differences between neighboring columns .
we also observe that large finite - size corrections in the roughness distributions are usually accompanied by huge corrections in height distributions and average local slopes , as well as in estimates of scaling exponents .
the molecular - beam epitaxy model of das sarma and tamborenea in @xmath2 dimensions is a case example in which none of the proposed scaling relations works properly , while the other measured quantities do not converge to the expected asymptotic values .
thus , although roughness distributions are clearly better than other quantities to determine the universality class of a growing system , it is not the final solution for this task . |
surface studies using grazing incidence angles offer a high degree of surface sensitivity .
this the case for x - rays@xcite , high energy electrons@xcite and ions@xcite as well as fast atoms @xcite . in these techniques
the angle of incidence is only few degrees . for most diffraction experiments ,
the required high positioning accuracy of the sample is provided with a goniometer but this is hardly compatible with sample transfer under uhv conditions and with high temperature needed for annealing .
the real challenge appears when the structural information lies in an intensity variation during an azimuthal rotation of the surface such as in ion beam triangulation @xcite and even more drastically in atom beam triangulation @xcite .
ensuring stability of the incidence angle better than 0.1 requires active correction .
we describe here a method designed for atom beam triangulation in a grazing incidence fast atom diffraction setup ( gifad ) that solves this issue . , is not exactly perpendicular to the rotation axis @xmath3 , the specular beam follows characteristic orbits during azimuthal scan.,width=264 ]
grazing incidence fast atom diffraction ( gifad or fad ) has recently emerged as a new surface science technique with exclusive surface sensitivity and high structural resolution @xcite . in this technique
a beam of @xmath4kev atoms impinges the surface at incidence angles @xmath1 of about one degree and with a angular spread ( divergence ) typically around 0.01 degree .
when the beam is aligned with a low index crystal direction , diffraction can take place provided that the surface quality is good enough in terms of flatness and coherence length .
the latter quantity is defined here as the mean separation between crystal defects .
first observed on single crystals of wide band gap ionic insulators such as lif @xcite , nacl @xcite , gifad has been shown to be successful in measuring metal surfaces @xcite , semi - conductors @xcite and even single layers of inorganic compounds @xcite and organic molecules @xcite . for surfaces of organic molecules ,
diffraction is not as sharp as for bulk crystal surfaces or inorganic layers probably due to the presence of a higher density of structural defects .
fortunately , the triangulation technique @xcite easily reveals the alignment direction of the adsorbed molecules while the shape of the azimuthal profile gives hints on the detailed molecular assembly @xcite .
even faint diffraction is enough to derive the lattice parameter of the molecular organization @xcite .
also , it has been suggested @xcite that azimuthal scans around channeling directions could be use to measure the range @xmath5 of the interaction potential above the surface .
this fundamental parameter can be very useful for quantitative analysis since it governs the length @xmath6 of the atom trajectory above the surface@xcite . at high energies ,
selective sputtering of step edges is also sensitive to the crystallographic direction @xcite .
in addition , gifad has demonstrated pronounced intensity oscillations during layer - by - layer molecular beam epitaxy growth@xcite . following these oscillations during azimuthal rotation
would allow a better uniformity of the layers .
ideally , these azimuthal scans should be performed without changing the angle of incidence .
this turns out to be more difficult than expected . on one hand
it is not straightforward to put a goniometer under uhv condition . on the other hand ,
sample transfer and rotation devices do not always allow for ultra - precise positioning of the sample .
most often the surface normal has a small residual tilt angle @xmath0 of about one degree with respect to the rotation axis .
this is sufficient to prevent straight - forward application of purely azimuthal scan . when the control of both the azimuthal and incidence angle is motorized , an on - line correction can be performed and is described below .
for each azimuthal angle , the plane of incidence is defined as the plane containing both the incident beam and the surface normal .
it also contains the specularly reflected beam , so that the location of the specular beam on the detector indicates the direction of the surface normal . during an azimuthal scan the specular beam spot follows a curve that can be described as an orbit . using the standard description of specular scattering .
@xmath7 \overrightarrow{k_{spec } } = \overrightarrow{k_{in } } - 2(\overrightarrow{k_{in}}.\overrightarrow{n } ) \overrightarrow{n}\ ] ] where @xmath8 is the incoming wavevector , @xmath9 is the specular one and @xmath10 is the surface normal . the exact calculation is not difficult but becomes even simpler by taking into account that the detector is located far away from the target surface so that the variation of the location of the impact on the crystal surface can be neglected .
we consider here that the beam impacts the surface exactly on its intersection with the rotation axis .
accordingly , the coordinates ( y , z ) on the detector depend only on the scattering angles .
we only need to focus on the vector @xmath11 describing the surface normal tilted by an angle @xmath0 with respect to the rotation axis @xmath3 during azimuthal rotation @xmath2 around this axis . defining the rotation around a vector @xmath12 by an angle @xmath13 as @xmath14
, @xmath15 can be written as a product of three rotations describing the tilt angle angle @xmath16 , the azimuthal scan @xmath17 and the angle of incidence @xmath18 , @xmath19 ; @xmath20 @xmath21 considering for simplicity , unit vectors @xmath22 and @xmath23 , i.e. @xmath24 , the scalar product @xmath25 is the first coordinate of the vector @xmath26 so that , within eq.[eq1 ] , the @xmath27 and @xmath28 components of the specular wave - vector write ; @xmath29 @xmath30 with @xmath2 the azimuthal angle . for small angle @xmath18 and @xmath0 the formula simplify and the coordinates @xmath31 on the detector form the parametric equations of the orbit .
@xmath32 = 1@xmath33 and to a tilt angle @xmath0 of 1@xmath33 , 0.66@xmath33 , 0.33@xmath33 and 0.1@xmath33 . for each panel ,
the horizontal scale is zoomed @xmath4 ten times compared with the vertical one .
the rightmost panel plots the four orbits at the same scale.,width=302 ] the eq.[eq2 ] shows that the full vertical extension @xmath34 and full width at half height @xmath35 are simple quantities while figure [ iresp ] illustrates that the orbits are not simple ellipses . .
twice per revolution , at angles ( b ) @xmath36 and ( a ) @xmath37 , this plane contains the primary beam , and coincides with the incidence plane .
the outgoing angles differ by @xmath38 and the associated spots are separated by @xmath39 . at 90@xmath33 from these values , ( c ) and ( d ) , the beam momentum @xmath40 is almost perpendicular to the reference plane with only a tiny component @xmath41 parallel to the rotation axis ( in dashed red ) . in this projected plane ,
the variation in the deflection of the specular beam is again 4@xmath0 but the amplitude on the detector will only be @xmath42 , i.e. almost two order of magnitude smaller than @xmath39.,width=226 ]
the tilt angle has to be measured in situ because each transfer can introduce slight variation of this value .
the simplest reference that can be measured with high accuracy is the location of the primary beam on the detector .
this is achieved by removing the target from the beam for a fraction of a second . from eq .
[ eq2 ] the tilt angle @xmath0 can be calculated as the ratio of @xmath39 to @xmath43 .
another option is to measure accurately several values to fit the parametric curve onto these data points .
intuitively , in the elliptic case , five values are needed corresponding to the coordinates @xmath44 of the two focal points of the ellipse and to the length of the chord .
these are linked to the experimental parameters : the value of the tilt angle @xmath0 , the mean scattering angle @xmath18 , the direct beam position and the mean inclination of the scattering plane @xmath45 . as depicted in fig.[scheme_2 ] these values can be determined as simple ratios between geometric distances measured on the detector .
in particular the mean scattering angle @xmath46 can be estimated as @xmath47 or via the center of the ellipse @xmath48 while @xmath0 can be estimated as @xmath49 or @xmath50 . ) .
taking the direct beam location as a reference @xmath51 and @xmath52 with @xmath53 and _ l _ the distance from the surface to the detector .
the small angle approximation is used , i.e. @xmath54 and @xmath55 .
the orbit in dashed red corresponds to the orbit of the line defined as the intercept of the incidence plane with the surface.,width=226 ] the main difficulty is that , quite often , a new surface has so many defects that the scattering profile of the helium atoms can be very diffuse with full width half maximum ( fwhm ) @xmath56 and @xmath57 such that the associated angular widths @xmath58 and @xmath59 exceed one degree .
a possible solution is to track the mean value of the scattering distribution either by moment analysis or via fit procedures which can usually bring the statistical error down to @xmath60 or even @xmath61 if the number of data points is large enough @xcite . when the tilt is larger than the angle of incidence , closed orbit
does not exist because the reflection is not possible in a selected azimuthal range . even in this case , the reduced variation of the scattering plane can be measured by tracking the mean value of the scattering profile along _
y_. of the scattering distribution ( ex . in insert ) is reported as a function of the azimuth @xmath2 follows a smooth sinusoidal curves having a full amplitude of 4@xmath0 .
note that the quasi elliptic shape of the scattering profile is accidental and has nothing to do with the orbits discussed above.,width=226 ] such a condition is illustrated in the insert of fig .
[ swing ] recorded at an incidence angle @xmath18 = 0.7@xmath33 .
surprisingly , a tilt angle of @xmath62 that is almost two times larger than the angle of incidence could be measured . in this case
the azimuthal scan was impossible without adjusting manually the incidence angle when it approaches the value near 0 ( almost parallel to the surface ) .
therefore , the angle of incidence is discontinuous and no orbit can be drawn . as a consequence
, the scattering pattern is not analyzed in the cartesian coordinates ( @xmath63 ) but in the polar coordinates ( @xmath64 ) illustrated in fig .
[ scheme_2 ] .
the smooth oscillation of the collision plane follows a sinusoidal curve indicating an inclination @xmath45 = 0.15@xmath33 of the position sensitive detector with respect to the mean scattering plane , a tilt angle @xmath65 .
to test the software of the correction procedure while avoiding possible damage inside the uhv chamber during test , a simple system outside vacuum has been built .
a small manipulator holds a laser pointer with its beam directed onto a polished silicon wafer placed on top of a rotating stepper motor .
a second miniature linear stepper motor is used to control the incidence angle @xmath1 .
the decisive advantage is that at the micron scale most surfaces are quasi perfect so that the specularly reflected laser beam is a single spot that can be imaged onto a paper screen at a distance _
l _ a few meters downstream .
another advantage is that the impact on the surface is visible making adjustments straightforward .
the motors are controlled by an inexpensive micro - controller arduino mega board with a 3d printer reprap 1.4board capable of hosting up to five `` allegro a4988 '' stepper motor controllers .
the arduino is connected to the host pc via an usb - link through which it receives order and transmit results .
the program is written in c inside the open source , integrated arduino development environment hosted in a pc and uploaded to the micro - controller by the usb - link .
various ratios of tilt angle to incidence angle have been explored . in these test conditions
the full trajectory can be recorded in few seconds .
we found that the curvature radius is so large compared with the horizontal width that the easiest starting strategy is to consider the orbit as a quasi ellipse .
the center @xmath66 of the orbit as well as its inclination @xmath67 are measured directly while the mean scattering angle @xmath18 and the tilt angle @xmath0 are determined from the amplitudes @xmath43 ( at half height ) and @xmath39 .
when the sample is placed in - between two view - ports located opposite to each other on the uhv chamber , shining a laser at the surface and observing the reflected beam is as easy as on the test bench .
a simple screen a few meters away will convert mm resolution to @xmath68 of a degree . a single azimuthal rotation , even partial , is sufficient to measure the tilt angle and the azimuthal reference .
if the relative positions of the laser and atom beam axis are known , the azimuthal reference can be transferred to the atom beam .
from eq . [ eq2 ] , the correction to apply to the angle of incidence during rotation is a simple sinusoidal function having a full amplitude of 4@xmath0 and a phase reference @xmath69 .
these parameters are independent of the incidence angle and have to be measured for each new sample surface before applying any correction .
the only practical issue is the user interface . to avoid possible problems during sample transfer or large amplitude movement , the default settings are that motor actions are performed without corrections . when an azimuthal scan is to be programmed , the azimuthal angle @xmath70 is , a priori , at a random position with respect to the reference plane . at this position ,
the mean scattering angle @xmath18 defined above , i.e. as the average incidence during an uncorrected azimuthal scan , has no particular meaning . at variance
, the user should simply assume that the correction will maintain the effective angle of incidence @xmath71 at the moment where the azimuthal scan is decided .
it means that the system should evolve on a quasi perfect straight line associated with the incidence angle @xmath72 and not along the correction contour associated with the mean value @xmath18 of the initial orbit .
care has to be taken to adjust the phase with the proper sign as illustrated in fig.[correction ] .
note that the corrected orbit is now almost a perfect line with less than @xmath73 degree vertical amplitude .
the compensation software have been tested successfully on the test bench but , due to a problem with the manipulator , we have no results for atoms scattering inside uhv . and a mean angle of incidence @xmath74 .
when correction is required by the user , the system is at the location indicated by a circle corresponding to an effective scattering angle of 1.2@xmath33 .
the system is now expected to evolve along the corrected circle ( looking here as a straight segment at this scale ) corresponding to the dashed blue orbit associated with a mean incidence angle @xmath75.,width=264 ]
it should be noticed that the specular angle can be ill - defined .
this occurs , for instance , in the case when surface twist and tilt mosaicity are present @xcite . in this case
there are several specular beams corresponding to the local surface orientation illuminated by the primary beam so that the correction can not be well defined . with a single crystal surface correction
is straightforward but is severely limited by the quality of the mechanical transmission .
this includes the difficult problem of mechanical backlash of the vacuum transmission because the correction applied to the angle of incidence has to be reversed at least once per turn .
there is still the possibility of a 180@xmath33 azimuthal scan without backlash . in practice ,
the presence of two view - ports allowing optical measurement is certainly a favorable condition to measure the tilt angle rapidly and accurately .
we are grateful to paola atkinson for help in reading the manuscript , to ala husseen for help during the measurements and to christophe charrire for his assistance in implementing the correction in the user interface software .
5 g. renaud , r. lazzari , f. leroy , probing surface and interface morphology with grazing incidence small angle x - ray scattering .
surface science reports * 64 * , p 255380 ( 2009 ) doi:10.1016/j.surfrep.2009.07.002 j.m .
van hove , p.r .
pukite , p.i .
cohen , reflection high energy electron diffraction measurements of algaas growth instabilities and roughening rates on misoriented substrates .
journal of vacuum science and technology b * 3 * ( 2 ) , 563 - 567 ( 1987 ) .
p. roncin , a. g. borisov , h. khemliche , a. momeni , a. mertens , and h. winter .
evidence for @xmath76 formation by simultaneous double - electron capture during scattering of @xmath77 from a lif(001 ) surface .
* 89 * , 043201 ( 2002 ) n. kalashnyk , h. khemliche and p. roncin .
atom beam triangulation of organic layers at 100 mev normal energy : self - assembled perylene on ag(1 1 0 ) at room temperature .
applied surface science .
* 364 * , p 235240 doi:10.1016/j.apsusc.2015.12.134 ( 2015 ) .
j. seifert , a. schller , h. winter , r. wodarczyk , j. sauer , and m. sierka , diffraction of fast atoms during grazing scattering from the surface of an ultrathin silica film on mo(112 ) .
b * 82 * , 035436 ( 2010 ) .
m. debiossac , a. zugarramurdi , h. khemliche , p. roncin , a. g. borisov , a. momeni , p. atkinson , m. eddrief , f. finocchi , and v. h. etgens , combined experimental and theoretical study of fast atom diffraction on the @xmath78(24 ) reconstructed gaas(001 ) surface .
b * 90 * , 155308 ( 2014 ) .
b. lalmi , h. khemliche , a. momeni , p. soulisse and p. roncin , high resolution imaging of superficial mosaicity in single crystals using grazing incidence fast atom diffraction . j. phys . : condens .
matter * 24 * 442002 ( 2012 ) .
a. schller , m. busch , s. wethekam , and h. winter , fast atom diffraction from superstructures on a fe(110 ) surface .
* 102 * , 017602 ( 2009 ) .
h. khemliche , p. rousseau , p. roncin , v. h. etgens and f. finocchi , grazing incidence fast atom diffraction : an innovative approach to surface structure analysis .
. lett . * 95 * , 151901 ( 2009 ) .
p. atkinson , m. eddrief , v. h. etgens , h. khemliche , m. debiossac , a. momeni , m. mulier , b. lalmi , and p. roncin , dynamic grazing incidence fast atom diffraction during molecular beam epitaxial growth of gaas . applied physics letters * 105 * , 021602 ( 2014 ) . | a procedure to measure the residual tilt angle @xmath0 between a flat surface and the azimuthal rotation axis of the sample holder is described .
when the incidence angle @xmath1 and readout of the azimuthal angle @xmath2 are controlled by motors , an active compensation mechanism can be implemented to reduce the effect of the tilt angle during azimuthal motion .
after this correction , the effective angle of incidence is kept fixed , and only a small residual oscillation of the scattering plane remains .
grazing incidence , surface science , fast atom diffraction , triangulation . |
degeneracies of eigenvalues of matrices dependent on parameters is a problem of great importance in physics .
the interest in this problem goes back to hamilton @xcite who pointed out a conical refraction due to coincident eigenvalues @xcite .
later , the significance of degeneracies of matrix spectra has been revealed in quantum physics @xcite , crystal optics @xcite and chemistry @xcite .
eigenvalue degeneracies are also essential for an understanding of the spectral fluctuations @xcite and the onset of quantum chaos @xcite . among those degeneracies ,
one finds the exceptional points ( eps ) @xcite where two riemann sheets of the eigenvalues are entangled by the square - root type of singularity .
the eps are generic singularities and , hence , can be found in numerous situations , e.g. the resonance problems @xcite or the quantum phase transitions @xcite .
much effort has been devoted to studies of degeneracies associated with avoided crossings in quantal spectra , focusing mainly on the topological structure of hilbert space and the geometric phase @xcite which a quantum system acquires when transported adiabatically around the singularity in parameter space @xcite .
these novel effects have attracted a considerable experimental interest as well @xcite . in many - body systems , eps have been studied in schematic models , such as the lipkin model @xcite , the interacting boson model @xcite and , recently , the 3-level pairing model @xcite which belongs to a general class of the richardson - gaudin models @xcite . in the latter case , it has been found that eigenvalue crossings in the complex - extended parameter space of the hamiltonian and the reduction in the number of eps uniquely define the quantum integrable system @xcite .
if true , this conjecture opens a possibility for studies of fingerprints of the chaotic dynamics in the quantum regime for small dimensional hilbert spaces . in the case of closed quantum system ( cqs ) described by a hermitian hamiltonian ,
mixing of two eigenvalues leads generically to the level repulsion and the avoided level crossing . in the open quantum system ( oqs )
, eigenvalues corresponding to unbound levels may exhibit not only crossing of either energies or widths but also true degeneracies due to a joint crossing of both energies and widths .
the latter case corresponds to a degenerate double pole of the scattering matrix ( s - matrix ) @xcite .
hence , oqs is a natural system to look for singularities of the ep type .
first attempt to search for eps in the realistic many - body model of atomic nucleus has been made recently using the real - energy continuum shell model @xcite , the so - called shell model embedded in the continuum ( smec ) @xcite , which is defined in projected hilbert spaces with at most two particles in the scattering continuum .
the total function space of the @xmath0-particle system in smec consists of two sets : the set of square - integrable functions @xmath1 , used in the standard nuclear shell model ( sm ) , and the set of continuum states ( the environment of the cqs ) @xmath2 embedding the sm . in one - particle continuum problems , channels
@xmath3 are determined by the motion of an unbound particle in a state @xmath4 relative to the @xmath5 nucleus in a certain sm state @xmath6 .
an index + denotes an outgoing asymptotics of a continuum state @xmath7 .
eigenstates of the oqs are found by solving the eigenvalue problem for an effective hamiltonian : @xmath8 in the function space @xmath1 of discrete states .
indices @xmath9 and @xmath10 in ( [ heff ] ) denote the projected hilbert subspaces with 0 and 1 nucleon in the scattering continuum , respectively .
@xmath11 is the green function of a single nucleon in the @xmath10 subspace . whereas the sm hamiltonian ( @xmath12 ) is hermitian
, @xmath13 is an energy dependent , non - hermitian operator above the particle emission threshold and a hermitian operator below the emission threshold . for each total angular momentum @xmath14 and parity @xmath15 separately , the @xmath16 coupling term in ( [ heff ] ) gives rise to an external mixing of different sm states @xmath17 via the coupling to common decay channels , both closed and opened .
the amount of this external mixing depends on the position of considered sm states with respect to the decay channel thresholds @xcite .
more details about the smec approach and its recent applications can be found in refs .
@xcite . in this work ,
we study the one - proton continuum of @xmath18ne to look for a possible ep with @xmath19 in a physically relevant region of excitation energies and parameters of the smec . for the sm hamiltonian ( @xmath12 in ( [ heff ] ) )
we take the zbm hamiltonian @xcite .
the residual coupling @xmath20 between @xmath9-subspace and the embedding continuum is given by : @xmath21 .
the true degeneracy of two @xmath22 resonances is searched for by varying the strength parameter @xmath23 in @xmath20 and the total energy @xmath24 of @xmath18ne .
@xmath25 is chosen at the lowest one - proton emission threshold .
in present studies , we consider the coupling of @xmath22 sm states to three lowest decay channels in @xmath18ne : @xmath26}^{{0}^{+}}$ ] ( the elastic channel ) , @xmath27}^{{0}^{+}}$ ] , and @xmath28}^{{0}^{+}}$ ] .
the energy dependence of @xmath13 is due to the coupling of discrete states of @xmath18ne to the decay channels @xmath29}^{{j}^{\pi''}}$ ] .
this coupling depends explicitly on @xmath24 . for @xmath30 ,
@xmath13 is complex - symmetric and its eigenvalues are complex . fig .
[ fig1 ] shows energies and widths of @xmath31 , @xmath32 eigenvalues as a function of the total energy of @xmath18ne for a fixed value of the strength @xmath33 mev@xmath34@xmath35 of the residual coupling @xmath20 .
@xmath31 , @xmath32 eigenstates and eigenvectors of @xmath36 coalesce at @xmath37 , forming an ep . encircling this singularity in a parameter space of @xmath13 , reveals a topological phase in @xmath31 , @xmath32 eigenvectors .
one can see a strong variation of @xmath31 and @xmath32 eigenvalues in the vicinity of thresholds for channels @xmath38}^{0^+}$ ] and @xmath39}^{0^+}$ ] , respectively .
the coulomb interaction shifts the near - threshold minima of re@xmath40 slightly above the threshold in channels @xmath38}^{0^+}$ ] ( @xmath25 ) for @xmath31 and @xmath39}^{0^+}$ ] ( @xmath41 ) for @xmath32 .
one should also notice that the energy @xmath24 of the lowest ep with @xmath19 does not necessarily correspond to the resonance energy which in smec is determined by solving a fixed - point equation for eigenvalues @xmath40 of the effective hamiltonian @xmath36 @xcite .
real part of energies in fig .
[ fig1 ] are plotted with respect to an arbitrarily chosen zero on the axis @xmath42 and only relative differences are meaningful . in the case considered in fig .
[ fig1 ] , if the ground state energy with respect to the elastic threshold ( @xmath25 ) is fixed at @xmath43 mev or @xmath44 mev , then the energy of the @xmath32 state obtained from the fixed - point equation becomes @xmath45 mev or 6.251 mev , respectively .
one should stress an essential role of non - hermiticity of @xmath13 in this problem .
the hermitian hamiltonian can be always diagonalized and eigenvalues , even if degenerate , always correspond to distinct eigenvectors .
on the contrary , in the non - hermitian hamiltonian one may find non - trivial jordan blocks of eigenfunctions and singular points in a parameter space of the hamiltonian where both eigenvalues and eigenvectors coalesce .
these eps have drastic effects on systems behavior , as we shall examine below on the example of spectroscopic factors .
[ fig2 ] shows real and imaginary parts of the one - proton spectroscopic factor @xmath46^{^+}}\rangle$ ] for the four lowest @xmath22 states in @xmath18ne .
this spectroscopic factor comes from a coupling of @xmath47 states ( @xmath48 ) of @xmath18ne to the ground state @xmath49 of @xmath50f .
the strength of the residual @xmath51-force coupling between @xmath9-subspace and the embedding continuum ( @xmath52 mev@xmath34@xmath35 ) is chosen to reproduce an experimental width of @xmath53 state in @xmath50f .
analogously , fig .
[ fig3 ] and [ fig4 ] exhibit the @xmath54^{^+}}\rangle$ ] and @xmath55^{^-}}\rangle$ ] spectroscopic factors , respectively .
these spectroscopic factors result from a coupling to the first ( @xmath56 ) and the second ( @xmath57 ) excited states in the daughter nucleus @xmath50f .
imaginary part of the spectroscopic factor , which appears for energies above the first one - proton emission threshold ( @xmath30 ) , yields the uncertainty of a real part and is related to the decaying nature of considered @xmath22 states . for deeply bound states , an external mixing of sm states due to the coupling to common decay channels can be safely neglected and , hence , the spectroscopic factor calculated in smec becomes close to the sm value .
this tendency can be seen in figs .
[ fig2]-[fig4 ] for @xmath58 . figs .
[ fig2]-[fig4 ] exhibit the external mixing of @xmath19 sm states by showing the @xmath24-dependence of spectroscopic factors for corresponding @xmath22 smec eigenstates .
this mixing is strongly dependent on the structure of individual @xmath22 sm states , the position of different one - proton emission thresholds , and the total energy @xmath24 of @xmath18ne .
one can see in figs .
[ fig2 ] , [ fig3 ] that the one - proton spectroscopic factors are large for @xmath31 and @xmath32 .
the eigenstate @xmath59 has only a sizeable @xmath60^{^-}}\rangle$ ] spectroscopic factor .
one can notice two distinct phenomena in the @xmath24-dependence of spectroscopic factors : the threshold effect and the ep effect .
the particle - emission threshold is a branch point in oqss @xcite . in figs .
[ fig2]-[fig4 ] , one may notice a strong effect of the coupling to all three decay channels @xmath61}^{0^+}$ ] ( the elastic channel ) , @xmath62}^{0^+}$ ] , and @xmath63}^{0 ^ -}$ ] for a pair of eigenstates @xmath31 , @xmath32 .
remaining @xmath22 states are effectively decoupled showing a weak @xmath24-dependence of spectroscopic factors in the vicinity of thresholds .
the threshold effect is well localized in @xmath24 and leads to mutually compensating variations in spectroscopic factors for strongly coupled @xmath31 and @xmath32 eigenstates of @xmath13 .
these compensating variations of single - particle occupancies is a signature of an entanglement of the two externally coupled states in the scattering continuum . in other words , one can not define any matrix element or expectation value of an operator which does not commute with the hamiltonian on an individual state of the entangled couple of eigenstates .
the effect of an ep at @xmath37 leads to much stronger variations of all three spectroscpic factors : @xmath46^{^+}}\rangle$ ] , @xmath54^{^+}}\rangle$ ] , and @xmath55^{^-}}\rangle$ ] for @xmath31 and @xmath32 states .
the same compensation mechanism as seen in the vicinity of the thresholds is found for spectroscopic factors in @xmath31 and @xmath32 states .
however , since the strength of the residual coupling ( @xmath52 mev@xmath34@xmath35 ) does not correspond to an exact value ( @xmath33 mev@xmath34@xmath35 ) for which the true degeneracy of eigenvalues has been found , the variations seen are caused by the proximity of the ep and _ not _ by the ep itself . as a result , the ep effect is weaker in magnitude and less concentrated in energy @xmath24 than expected at an exact position of the ep .
it is interesting to notice that the intricate entanglement of two continuum states caused by the ep is seen not only at the singularity point in parameter space of the hamiltonian ( @xmath64 mev@xmath34@xmath35 ) but also relatively far away from it . in that sense , it is a robust phenomenon relatively weakly dependent on a fine - tuning of the hamiltonian parameters and apt to be observed in nuclear physics experiments .
hamilton , dublin univ .
rev . q. mag . * 1 * , 795 ( 1833 ) .
m. berry , r. bhaduri , and s. klein , eur .
* 20 * , 1 ( 1999 ) .
j. von neumann , and e.p .
wigner , z. phys . * 30 * , 467 ( 1929 ) .
m.v . berry , _ quantum chaos _ ( 1985 ) , ed .
g. casati ( plenum ) . c. herring , phys
. rev . * 52 * , 365 ( 1937 ) .
e. teller , j. phys .
chem . * 41 * , 109 ( 1937 ) .
m.v . berry , ann .
( ny ) * 131 * , 163 ( 1981 ) ; + m.v .
berry , _ chaotic behaviour of deterministic systems _
, les houches lectures xxxvi ( 1985 ) , r.h .
helleman and g. ioss eds .
( north - holland ) .
t. kato , _ perturbation theory for linear operators _ ( springer verlag , berlin , 1995 ) .
zirnbauer , j.j.m .
verbaarschot and h.a .
weidenmller , nucl .
a * 411 * , 161 ( 1983 ) .
heiss , and w .- h .
steeb , j. math . phys . * 32 * , 3003 ( 1991 ) . a. mondragon , and e. hernandez , j. phys .
a : math . gen . * 29 * , 2567 ( 1996 ) .
e. hernandez , a. jauregui , and a. mondragon , j. phys .
a : math . gen . * 39 * , 10087 ( 2006 ) .
heiss , j. phys . a : math
. gen . * 39 * , 10081 ( 2006 ) .
p. cejnar , s. heinze , and m. macek , phys .
lett . * 99 * , 100601(2007 ) .
p. dorey , c. dunning , and r. tateo , j. phys , a : math . gen . * 34 * , l391 ( 2001 ) .
m.v . berry , proc .
london , ser .
a * 392 * , 45 ( 1984 ) .
lauber , p. weidenhammer , and d. dubbers , phys .
lett . * 72 * , 1004 ( 1994 ) ; + d.e .
manolopoulos , and m.s .
child , phys .
lett . * 82 * , 2223 ( 1999 ) ; + f. pistolesi , and n. manini .
rev . lett . * 85 * , 1585 ( 2000 ) .
r. jackiw , commun . at .
* 21 * , 71 ( 1988 ) ; + s.j .
vinitsky et al . , sov .
usp . * 33 * , 403 ( 1990 ) ; + d.j .
moore , phys .
rep . * 210 * , 1 ( 1990 ) ; + g. alden mead , rev .
phys . * 64 * , 51 ( 1992 ) . c. dembowski et al .
lett . * 86 * , 787 ( 2001 ) ; _ ibid . _ * 90 * , 034101 ( 2003 ) .
heiss , a.l.sanino , phys .
a*43 * , 4159 ( 1991 ) .
j. dukelsky , j. okoowicz , and m. poszajczak , arxiv : 0811.4105 .
j. dukelsky , j. okoowicz , and m. poszajczak , arxiv : 0811.4051 ; + j. okoowicz , m. poszajczak , and j. dukelsky , arxiv : 0811.4250
. j. dukelsky , s. pittel , and g. sierra , rev .
phys . * 76 * , 643 ( 2004 ) .
et al . _ ,
b * 707 * , 421 ( 2005 ) .
j. okoowicz , and m. poszajczak , arxiv : 0811.4060 .
lassila , and v. ruuskanen , phys .
* 17 * , 490 ( 1966 ) .
n.j . kylstra , and c.j .
joachain , phys .
a * 57 * , 412 ( 1998 ) .
i. rotter , phys .
e * 65 * , 0261217 ( 2002 ) ; _ ibid .
_ e * 67 * , 026204 ( 2003 ) .
m. znojil , j. phys .
* 39 * , 441 ( 2006 ) .
h. w. bartz , i. rotter , and j. hhn , nucl .
a * 275 * , 111 ( 1977 ) ; _ ibid . _ a * 307 * , 285 ( 1977 ) .
k. bennaceur _ et al .
a * 651 * , 289 ( 1999 ) ; nucl .
a * 671 * , 203 ( 2000 ) .
j. okoowicz , m. poszajczak , and i. rotter , phys . rep . * 374 * , 271 ( 2003 ) .
j. rotureau , j. okoowicz , and m. poszajczak , phys .
lett . * 95 * , 042503 ( 2005 ) ; nucl .
a * 767 * , 13 ( 2006 ) .
r. chatterjee , j. okoowicz , and m. poszajczak , nucl .
a * 764 * , 528 ( 2006 ) .
zuker , b. buck , and j.b .
mcgrory , phys .
* 21 * ( 1968 ) 39 .
baz et a. , _ scattering , reactions and decay in nonrelativistic quantum mechanics _ , ipst ( 1969 ) . | we demonstrate existence of exceptional points in many - body scattering continuum of atomic nucleus and discuss their salient effects on the example of one - nucleon spectroscopic factors . |
lipid bilayers are the essential parts of any living cell .
they constitute the main body of the cell membrane while being found in different organelles inside the cell .
the cell membrane hosts collections of proteins and lipid rafts and it is crowded with a variety of biomolecules . in such non - homogenous and diverse environments , the diffusion of protein molecules in lipid bilayers plays a vital role in different biological processes like cell signaling .
the diffusion of lipids and proteins is not a distinct phenomenon and depends on the environment and neighboring molecules @xcite , and even changes from cell to cell @xcite .
transmembrane proteins diffuse as dynamic complexes with lipids @xcite , and their interactions with lipid molecules mediates traffic in cell membranes .
experiments show that the hydrophobic mismatch between proteins and lipids controls the diffusion coefficient of molecules inside a bilayer @xcite .
anomalous sub- and super - diffusion processes are more efficient scenarios for finding a nearby target than normal diffusion @xcite , and they enhance the formation of protein complexes and signal propagation . according to experiments , the mean square displacement ( msd ) of membrane channel proteins of human kidney cell exhibits subdiffusion @xcite .
the addition of cholesterols to lipid membranes @xcite , and the augmented area coverage of membrane proteins @xcite also lead to subdiffusion of lipids and proteins .
superdiffusivity has been observed in several physical systems and is often associated with lvy flights .
prominent examples are the chaotic motion of particles in a rotating laminar flow @xcite with long - range flights and horizontally vibrated grains which exhibit lvy flight with small jumps compared to their diameter @xcite
. nevertheless , an active component such as molecular motor can also be the source of superdiffusivity . a recent study by @xcite shows that in a gel composed of actin filaments , fascin molecules and myosin - ii filaments , the diffusion of small actin and fascin clusters are superdiffusive because of the work done by molecular motors . in many conditions membranes
are under shear .
when a red blood cell ( rbc ) migrates through vessels smaller in diameter than itself , the rbc membrane is under shear .
the blood flow exerts tangential shear stresses on vascular endothelia , and initiates cellular processes like activating g protein - coupled receptors .
these receptors are able to sense the fluid shear stress as an increase in the lateral membrane tension , and subsequently go through conformational changes @xcite .
the temporal and spatial changes in the membrane fluidity , in response to shear flow , have been observed experimentally @xcite . in this study
we are interested in the diffusivity of lipid and protein molecules in flat membranes under shear flow , and attempt to answer three fundamental questions using molecular dynamics ( md ) simulations : ( i ) do lipid molecules have different diffusion coefficients parallel and perpendicular to flow direction ? ( ii ) how does a simple shear flow influence the random motions of transmembrane proteins ? ( iii ) is there any correlation between the population of proteins and their diffusion in the membrane ?
we simulate lipid membranes utilizing a flexible lipid model @xcite and triple - strand rigid proteins @xcite [ fig . [
fig : vprofile](a ) ] .
although different coarse grained models have been developed over years @xcite , the model adopted here has the ability to mimic the physical properties of lipid membranes .
the model is not a true coarse grained model but can qualitatively describe phenomena related to lipid membranes and associated transmembrane proteins as we will compare some of our results with true coarse grained and atomistic models .
our goal is to explore the effect of shear flow on the motion of lipid and protein molecules over nanosecond time scales .
we perform md simulations of an @xmath0 ensemble , where the number of particles @xmath1 , the volume @xmath2 and the temperature @xmath3 are held constant .
lees - edwards boundary condition is employed to generate simple shear flow with the shear rate @xmath4 @xcite .
other boundary conditions are periodic .
the temperature is set to @xmath5 so that the system is safely above the gel to liquid phase transition temperature of different phosphatidylcholine lipid bilayers .
a detailed description of the model can be found in @xcite .
we express the position and velocity vectors of particles in the cartesian @xmath6 coordinate system whose origin is located at the center of our cubic simulation box . the @xmath7 and @xmath8 axes lie in the membrane plane and the @xmath9 axis is perpendicular to that .
md scales of length , time , mass and energy are @xmath10 , @xmath11 , @xmath12 and @xmath13 , respectively .
@xmath14 is the avogadro s number . in all simulations
, the dimensions of the box along the coordinates axes , @xmath15 , @xmath16 and @xmath17 , are set to @xmath18 .
the total number of particles equals @xmath19 , which gives a fixed number density , @xmath20 , and a constant average fluid pressure of @xmath21 for all simulations . here
, we note that isotropic pressure control is not appropriate in the simulation of lipid membranes since their volume is constant in the laboratory and biological conditions .
however , a fixed number density will give a constant average pressure and physical properties of lipid membranes with different number of lipids and proteins can be compared .
physical quantities are measured using a run - time of @xmath22 .
an important mechanical property of every membrane is the surface tension @xmath23 that mainly affects the diffusion of lipid molecules .
we compute the tension of our model membranes from @xmath24l_z\ ] ] where @xmath25 ( @xmath26 ) are the components of the pressure tensor @xcite . in this study
we apply msd to determine the diffusion properties of randomly moving particles .
the diffusion coefficient is thus calculated using the einstein expression @xmath27 ^ 2\rangle , \label{eq : class - diff}\ ] ] where @xmath28 and @xmath29 is the displacement due to the random motion of the @xmath30th particle in the @xmath31-direction .
the summation in eq .
( [ eq : class - diff ] ) is taken over the particles of the same type . from here
on , we will drop the summation sign for brevity .
the operator @xmath32 denotes the canonical average .
( [ eq : class - diff ] ) describes the regular brownian motion when the msd is linearly proportional to @xmath33 .
the diffusion process is anomalous should the msd deviate from the linear form , and obey the relation @xmath34
^ 2\rangle=2d^a_{\alpha\alpha}t^a , \label{eq : class - diff-2}\ ] ] where @xmath35 is the diffusion exponent and @xmath36 is the fractional diffusion coefficient .
the regimes with @xmath37 and @xmath38 are subdiffusive and superdiffusive , respectively . to obtain smooth msd curves , we evolve systems of 80 different initial conditions and report their ensemble - averaged diffusion coefficients and msds .
the diffusion of lipids is investigated by tracing the motion of their head groups .
proteins are traced using their center of masses . in equilibrium models without external shearing ,
there is no streaming in the solvent membrane system .
therefore , the flux of particles is associated with their random motions , and any displacement is due to thermal fluctuations . in sheared membranes ,
however , there is a combination of streaming and diffusive fluxes .
we thus need to distinguish and eliminate the streaming flux when calculating the msd .
let us define the actual velocity components of the @xmath39th particle as @xmath40 where @xmath41 is the average streaming velocity , and @xmath42 is the peculiar velocity whose time integral gives the displacements in eqs .
( [ eq : class - diff ] ) and ( [ eq : class - diff-2 ] ) . in equilibrium models ,
the average velocities @xmath43 vanish and we obtain @xmath44 . with external shearing ,
the flow is always imposed in the @xmath7-direction .
therefore , @xmath45 and @xmath46 are directly integrated to find the corresponding displacements . when the simulation box is uniformly filled with one type of particles ( let us says solvent particles ) , one readily finds @xmath47 . in the presence of a lipid bilayer , the vertical velocity profile in the @xmath9-direction
is no longer linear [ see fig . [
fig : vprofile](b ) ] .
therefore , to obtain the msd of lipids , we define @xmath48 as the average velocity of the layer where the head groups of phospholipds reside , and obtain @xmath49 .
it is remarked that transmembrane proteins do not experience streaming movements , @xmath50 , because the shear forces exerted on their end points from the upper and lower solvent columns are equal and in opposite directions .
the lateral diffusion of lipids in equilibrium conditions is enhanced as the membrane tension increases . this has been observed in simulations by atomistic @xcite and coarse grained models @xcite and is because by stretching the membrane , the area per lipid increases and more space is provided for the free motion of lipids .
our simulations with this very simple model shows the same pattern . by turning on the shear flow ,
lipid molecules undergo an initial ballistic motion that transforms into an interval of subdiffusion with @xmath51 [ fig .
[ fig : msd - l ] ] .
the transient anomalous state has been observed in atomistic simulations @xcite as well .
after the transient anomalous diffusion and over longer time scales a normal diffusion with @xmath52 is observed [ fig .
[ fig : msd - l ] ] .
it is noted that we have found similar msd profiles for lipid molecules in equilibrium and sheared systems , and in both cases lipids ultimately develop normal diffusion .
although @xcite reported a permanent subdiffusive behavior by the atomistic simulation of lipid membranes in equilibrium , experiments support a final regular diffusion regime , as we do , even in the presence of obstacles @xcite .
we conclude that the diffusion regime of lipid molecules is invariant with and without external shearing .
the diffusion coefficients obtained from the normal diffusion region of msd plots , are larger for smaller shear rates .
for example , for a membrane of @xmath53 lipid molecules , we find @xmath54 , @xmath55 and @xmath56 .
for the same system under a shear flow of @xmath57 , the membrane tension drops to @xmath58 and diffusion coefficients reduce to @xmath59 and @xmath60 .
the reason is that the membrane thickness increases for higher shear rates and the tension decreases without any change in the area per lipid @xcite .
consequently , the fluidity of the membrane decreases and slows down the diffusion process . after applying the shear force
, we find that @xmath61 drops for about @xmath62 while @xmath63 remains almost constant with only @xmath64 change which is not statistically significant since it is less than the mean standard error for diffusion coefficients which is less than @xmath65 .
the difference between @xmath61 and @xmath63 is indistinguishable in fig .
[ fig : msd - l ] .
we speculate that the alignment of lipid chains with the flow breaks the isotropy and yields @xmath66 .
atomistic simulation of lipid membrane @xcite has shown that increasing the tension of membrane , induced by altering the area per lipid , result in larger lateral diffusion coefficients .
this change is not linear with tension and they have reported @xmath67 increase in lateral diffusion coefficient .
coarse grained simulations @xcite showed that for larger tensions the increase in diffusion coefficient slows down and it depends on the range of tension . in our simulation , shear flow induces @xmath68 change in tension and consequently result in a different diffusion coefficient in the flow direction . in the @xmath9-direction ,
perpendicular to the membrane plane , our msd plots always show a confined motion as is expected .
we add rod - like proteins to the membrane , and simulate models with different protein concentrations that vary significantly from cell to cell .
since proteins increase the membrane tension as they perturb the distribution of lipids @xcite , we increase the number of lipids ( proportional to proteins ) to keep the membrane almost tensionless . in equilibrium and for a membrane with a single embedded protein with @xmath69
, we can measure the diffusion coefficients @xmath70 since the msd of protein shows an ultimate regular diffusion .
we find @xmath71 and @xmath72 , which are equivalent to @xmath73 with less than @xmath65 error .
these values are larger than experimental values , by two orders of magnitude .
the obvious reason is the effect of coarse graining that has reduced the interdigitation and friction between molecules , and allows for faster movements of particles .
the reduced friction affects both membrane component and solvent motion and consequently decreases both solvent and membrane viscosity .
viscosity of a fluid is a determinant of mobility or diffusion in that medium .
another minor source of discrepancy is the smaller size of our model proteins compared to real integral proteins .
we note that the average diffusive behavior of lipids is unaffected by the presence of proteins .
we computed the average msd profile of lipids for the above system and found the same pattern of initial ballistic regime , transient subdiffusive region and final normal diffusion .
the diffusion coefficients for our model proteins is smaller than model lipid molecules by a factor of @xmath74 .
this is expected because of the larger size and mass of proteins compared to lipids .
an initial ballistic motion the same as what has been observed for lipids is recovered for proteins by using time step as small as @xmath75 for @xmath76 .
this regime is shared by all the systems regardless of the shear rate and the number of proteins . by putting the system under simple shear flow ,
proteins undergo brownian motion when only two proteins are used [ fig .
[ fig : msd - p](a ) ] .
one could anticipate this result , for single proteins can not remarkably perturb the distribution of lipids , and change the diffusion properties of the membrane .
with 4 proteins , however , we observe that they form two double - protein clusters ( due to the depletion force ) , and exhibit a strong superdiffusive motion parallel to the flow .
[ fig : msd - p](b ) shows how after @xmath77 the normal diffusion regime transforms to strong superdiffusion with @xmath78 in the @xmath7-direction .
interestingly , this exponent is the same as the superdiffusion exponent found by @xcite for active diffusion of protein clusters by molecular motors . because of crowding effect and increase in the concentration of proteins @xcite , our results show a subdiffusive behavior along the @xmath8 axis with @xmath51 .
@xcite observed @xmath79 in experiments with channel proteins of human kidney cell , and @xcite found @xmath80 by molecular simulations of aggregating nak channel proteins . for clarity , the upper parts of the msd profiles in figures [ fig : msd - p](a ) and [ fig : msd - p](b ) are plotted together in fig .
[ fig : msd - p](c ) . to rule out the effect of statistical errors in the development of anomalous behavior
, we have divided the msd profiles of the system with 4 proteins to smaller intervals , and separately calculated @xmath35 and its error over each interval using the curve fitting toolbox of matlab .
we have then assigned the mean value to the center of the time interval and plotted the calculated diffusion exponent versus time in fig .
[ fig : a ] . for instance , over the initial ballistic zone
, we have obtained @xmath81 from @xmath82 to @xmath83 , and assigned this value to @xmath84 .
[ fig : a ] shows that @xmath35 approaches @xmath85 and @xmath86 in the @xmath7- and @xmath8-directions , respectively . to understand the physical mechanism behind the observed anomaly , we conduct the following analysis .
let us define the local concentration of the head particles of lipid and protein molecules at the position @xmath87 and time @xmath33 as @xmath88 $ ] where @xmath89 , and @xmath90 is the position vector of the @xmath30th head particle .
@xmath91 denotes the total number of head particles in the monolayer , @xmath92 is the heaviside step function , and @xmath93 is the typical size of the cross section of a protein cluster ( or a protein lipid complex ) .
our numerical experiments show @xmath94 is the best choice .
we examine the trajectories of protein molecules and the spatial variation of the normalized distribution @xmath95 to explain the physics of observed superdiffusion . here
@xmath96 and @xmath97 are the minimum and maximum values of @xmath98 at a given time @xmath33 .
[ fig3 ] demonstrates contour plots of @xmath99 for the upper and lower monolayers at a randomly chosen time .
[ fig4](a ) demonstrates the trajectories of a single protein molecule and two double - protein clusters , in equilibrium and sheared systems , respectively .
the equilibrium trajectory corresponds to regular diffusion because it covers a definite area . in the sheared system
the trajectories are elongated and aligned with the flow direction , indicating a fractional random walk : local isotropic wanderings followed by small - step jumps mainly in the flow direction .
these successive jumps can be interpreted by inspecting the contour plots of @xmath100 over a long duration of time .
the hydrophobic mismatch between protein clusters ( which have asymmetric big cross sections ) and the membrane disturbs the bilayer thickness and the arrangement of nearby lipids .
moreover , proteins are able to transport their neighboring lipids with them and behave as dynamic complexes @xcite .
these two effects collaborate to create transient voids whose distribution can be described by @xmath101 ( light shades in fig .
[ fig3 ] ) . when the bilayer is sheared , protein
lipid groups are pushed into the voids created by themselves or other groups / complexes and experience accelerated , and therefore , superdiffusive movements .
it should be noted that during our simulations , the center of mass of the membrane and embedded proteins remains almost at the center of the coordinate system .
none of our samples show long - step straight motions of protein clusters .
what we have seen are small - step jumps , which are comparable with the mean distance between protein clusters and voids [ compare figs .
[ fig3 ] and [ fig4](a ) ] .
we have computed the probability distribution function ( pdf ) of protein displacements and plotted it in fig . [
fig4](b ) .
a gaussian function has been fitted to the data by setting its maximum to the maximum of pdf , and its variance is found using the full width at half minimum of pdf .
the pdf exhibits a deviation from normal distribution and it has tails .
we have also applied the kolmogorov - smirnov test @xcite to confirm that the pdf is not a normal gaussian .
this is a clear indication of anomalous diffusion . as we noted before , the ends of proteins
are pulled in opposite directions by the two sheared solvent columns .
an important question is why do protein clusters prefer to jump into the voids when the membrane is sheared ?
we have a simple explanation for this behavior : two ends of proteins attract lipids from two different layers of the membrane .
thus , the symmetry in the distribution of the upper and lower protein - bound lipids is likely to break .
moreover , the shear force is exerted by the solvent on both the lipid and protein heads . the mentioned symmetry breaking thus leads to different force components at the upper and lower ends of protein lipid complexes , and they will jump into a void in the direction specified by the broken symmetry . to quantify this process ,
we take a sample two - protein cluster and define @xmath102 and @xmath103 as the center of masses of its protein heads in the upper and lower leaflets , respectively .
we then compute @xmath104 which is proportional to the net shear force exerted on the cluster at the time step @xmath105 : the local effective area in contact with a solvent column is determined by the number of head particles , and the shear force is calculated by multiplying the effective contact area by the shear rate and viscosity .
@xmath106 will be zero if the concentrations of the head particles of lipids are identical around the two heads of the cluster . defining @xmath107 as the average of @xmath106 , the autocorrelation function @xmath108 \cdot \left [ \xi(t_i)-\bar \xi \ , \right ] } { \sum_i \left [ \xi(t_i)-\bar \xi \ , \right ] ^2},\end{aligned}\ ] ] plotted in fig .
[ fig4](c ) carries interesting information about the shear force experienced by the cluster : the correlation time @xmath109 shows a sustained accelerated motion of the cluster over @xmath110 , independent of the initial time @xmath111 .
moreover , the oscillatory decaying profile of @xmath112 shows a random symmetry breaking in the sense of deterministic chaos @xcite .
the existence of a correlation time can also be verified by studying the cross - correlation function @xmath113 over one of the leaflets .
4(d ) shows that @xmath114 of the upper leaflet steeply rises from @xmath115 and peaks at @xmath116 , which is the earliest time span that protein lipid complexes need to occupy their nearest voids .
this is quite consistent with the acceleration time scale of protein clusters predicted by the autocorrelation function @xmath112 . for @xmath117
the cross - correlation function remains almost flat because the size of a void is always bigger than the distance that a protein cluster travels during @xmath118 .
in this work , we use a toy model of proteins and lipids to simulate the dynamics of cell membranes undergoing shear flow .
we calculate the msd profile of lipids and proteins in equilibrium and sheared system and compare our result with existing works in the literature .
this simple model beautifully captures the basic regimes in the diffusive behavior of lipid molecules : short initial ballistic regime , transient subdiffusive regime and final normal diffusion .
all - atom md simulations show the same diffusive regions in the msd profile of lipid molecule @xcite .
moreover , we show that shear flow reduces the tension of membrane and consequently decreases the diffusion coefficient in the direction of flow .
the fact that reducing membrane tension slows down lipids movement has been reported as the result of all - atom md simulations @xcite and experimental measurements @xcite .
since there were small number of protein molecules in the system , we did extensive md simulations to obtain the final msd profiles for proteins .
our c++ code , which has already been calibrated @xcite for membranes under simple shear flow , is not parallel and allows only for small - scale simulations .
we have recovered our results by lammps for models in equilibrium conditions .
however , lammps has not the capability of imposing simple shear flow conditions and we could not use it to run our sheared systems over longer time scales . despite these limitations , the smooth msd profiles obtained from our numerous samples clearly show the distinction between the regular diffusion of single proteins and anomalous diffusion of protein clusters .
we explain this anomaly by deliberately examine the distribution of head particles of lipids and proteins and introducing void generation mechanism .
our simulations cover timescales of order of nanoseconds .
cell responses to stimuli are fast due to enhanced mobility of protein receptors .
consider our findings , superdiffusion of proteins under shear flow can play a dominant role in the process of signaling in endothelium cells , rbcs , liposomes used for targeted drug delivery and other sheared membranes .
we note that since the length and shape of proteins and their ability in attracting neighboring lipids controls the sizes of voids there was no other mechanism in our models for void generation superdiffusion properties reported in this study are not universal and they highly correlate with the properties of embedded proteins .
simulations using detailed structures of lipids and proteins are needed to better assess the superdiffusive behavior in realistic cell membranes .
experimental exploration of our results can be done using single particle tracking @xcite which can give the msd of proteins directly .
we thank anonymous referees for their insightful comments that helped us to improve the presentation of our results .
was partially supported by the national elites foundation of iran .
we express our sincere thanks to prof .
ali meghdari for his helpful discussions and encouragements .
42 j. a. dix , and a. s. verkman , annu .
biophys . ,
* 37 * , 247 ( 2008 ) .
s. wieser , j. weghuber , m. sams , h. stockinger , and g. j. schuetz , soft matter , * 5 * , 3287 ( 2009 ) .
s. niemel , m. s. miettinen , l. monticelli , h. hammaren , p. bjelkmar , t. murtola , e. lindahl , and i. vattulainen , j. am .
soc . , * 132 * , 7574 ( 2010 ) .
a. prasad , j. kondev , and h. a. stone , phys .
fluids , * 19 * , 113103 ( 2007 ) .
s. ramadurai , a. holt , l. v. schafer , v. v. krasnikov , d. t. s. rijkers , s. j. marrink , j. a. killian , and b. poolman , biophys .
j. , * 99 * , 1447 ( 2010 ) .
y. gambin , m. reffay , e. sierecki , f. homble , r. s. hodges , n. s. gov , n. taulier , and w. urbach , j. phys .
b , * 114 * , 3559 ( 2010 ) .
g. guigas and m. weiss , biophys .
j. , * 94 * , 90 ( 2008 ) .
f. bartumeus , j. catalan , u. l. fulco , m. l. lyra , and g. m. viswanathan , phys .
lett . , * 88 * , 097901 ( 2002 ) .
a. v. weigel , b. simon , m. m. tamkun , and d. krapf , proc .
, * 108 * , 6438 ( 2011 ) . j .- h .
jeon , h. m .- s .
monne , m. javanainen , and r. metzler , phys .
lett . , * 109 * , 188103 ( 2012 ) .
s. ramadurai , a. holt , v. krasnikov , g. van den bogaart , j. a. killian , and b. poolman , j. am .
soc . , * 131 * , 12650 ( 2009 ) .
t. h. solomon , e. r. weeks , and h. l. swinney , phys .
lett . , * 71 * , 3975 ( 1993 ) .
f. lechenault , r. candelier , o. dauchot , j. p. bouchaud , and g. biroli , soft matter , * 6 * , 3059 ( 2010 ) .
s. khler , v. schaller , and a. r. bausch , nat .
, * 10 * , 462 ( 2011 ) .
m. chachisvilis , y. zhang , and j. a. frangos , proc .
, * 103 * , 15463 ( 2006 ) .
p. j. butler , g. norwich , s. weinbaum , and s. chien , am .
j. physiol .
cell physiol .
, * 280 * , c962 ( 2001 ) . m. a. haidekker , n. lheureux , and j. a. frangos , am . j. physiol .
heart circ .
physiol . ,
* 278 * , h1401 ( 2000 ) .
r. goetz and r. lipowsky , j. chem .
phys . , * 108 * , 7397 ( 1998 ) .
a. khoshnood , h. noguchi , and g. gompper , j. chem .
, * 132 * , 025101 ( 2010 ) . s. j. marrink , h. j. risselada , s. yefimov , d. p tieleman , and a. h. de vries , j. phys
b , * 111 * , 7812 ( 2007 ) . m. p. allen , and d. j. tildesley , _ computer simulation of liquids _ , ( oxford university press , oxford , uk . ,
j. neder , b. west , p. nielaba , and f. schmid , j. chem .
phys . , * 132 * , 115101 ( 2010 ) .
h. s. muddana , r. r. gullapalli , e. manias , and p. j. butler , phys .
phys . , * 13 * , 1368 ( 2011 ) .
e. flenner , j. das , m. c. rheinstadter , and i. kosztin , phys .
e , * 79 * , 011907 ( 2009 ) .
g. r. kneller , k. baczynski , and m. pasenkiewicz - gierula , j. chem .
phys . , * 135 * , 141105 ( 2011 ) .
m. j. skaug , r. faller , and m. l. longo , j. chem .
phys . , * 134 * , 215101 ( 2011 ) .
m. javanainen , h. hammaren , l. monticelli , j .- h .
jeon , m. s. miettinen , h. martinez - seara , r. metzler , and ilpo vattulainen , faraday discuss . ,
w. h. press , s. a. teukolsky , w. t. vetterling , and b. p. flannery , _ numerical recipes : the art of scientific computing _ , 3rd edition , ( cambridge university press , 2007 ) j. h. argyris , g. faust , and m. haase , _ an exploration of chaos : an introduction for natural scientists and engineers _ , ( north - holland , 1994 ) j. m. crane , and a. s. verkman , biophys .
j. , * 94 * , 702 ( 2008 ) . | we use coarse grained molecular dynamics simulations to investigate diffusion properties of sheared lipid membranes with embedded transmembrane proteins . in membranes without proteins , we find normal in - plane diffusion of lipids in all flow conditions .
protein embedded membranes behave quite differently : by imposing a simple shear flow and sliding the monolayers of the membrane over each other , the motion of protein clusters becomes strongly superdiffusive in the shear direction .
in such a circumstance , subdiffusion regime is predominant perpendicular to the flow .
we show that superdiffusion is a result of accelerated chaotic motions of protein lipid complexes within the membrane voids , which are generated by hydrophobic mismatch or the transport of lipids by proteins . |
the question of the importance of quenched random field ( rf ) disorder in ferromagnets traces back to the primary paper by imry and ma @xcite from mid - seventies .
they argued using energy minimization for an excitation to the ground state , that the randomness in the fields assigned to spins changes the lower critical dimension from the pure case with @xmath14 to @xmath15 .
after that a number of field - theoretical calculations suggested that the randomness increases @xmath16 with two to be @xmath17
. finally came rigorous proofs first by bricmont and kupiainen @xcite in 87 that there is a ferromagnetic phase in the three - dimensional ( 3d ) random field ising model ( rfim ) and in 89 by aizenman and wehr @xcite that there is no ferromagnetic phase in 2d rfim .
thus it was established that the lower critical dimension is two .
this means that the ground state is a paramagnet , but the problem as to how to describe the structure of the ( ground state of ) 2d rfim still persists .
some recent work concerns the scaling of the correlation lengths @xcite and there is a suggestion of a ferromagnetic phase , but with a magnetization that is in the thermodynamic limit below unity @xcite .
the point is that due to the ( relevant ) disorder there are no easy arguments that would indicate , say , how the paramagnetic ground state should be characterized .
this is different from the thermal ising case , which is quite trivial in 1d . in two dimensional ising magnets , in the presence of quenched random fields , the problem of determining the ground state ( gs ) becomes more difficult .
finding the true ground state with any standard monte carlo method is problematic due to the complex energy landscape .
even with the exact ground state methods , as the one used in this paper , the thermodynamic limit is difficult to reach , since the finite size effects are strong . in the typical case of square lattices
the only way not to have a massive domain , which would scale with the euclidean dimension of the system , is to have enough interpenetrating domains of both spin orientations .
however , with decreasing strength of the randomness the ferromagnetic coupling constants between spins start to matter , the domains become `` thicker '' , and thus one enters an apparent ferromagnetic regime , and the paramagnetic ( pm ) phase is encountered only at very large length scales
. should there be large clusters with a fractal ( non - euclidian ) mass scaling they nevertheless can contribute to the physics in spite of the fact that the total fraction of spins can be negligible in the thermodynamics limit .
thus such clusters may even be measurable in experiments or be related to the dynamical behavior in non - equilibrium conditions .
therefore it is of interested to study the structure of the large(st ) clusters in the ground state , since it is not simply paramagnetic like in normal ising magnets above @xmath18 .
the true ground state structure gives also some insight into the physics at @xmath19 , since the overlap between the gs and the corresponding finite-@xmath20 state is close to unity for @xmath20 small , in contrast to the thermal chaos in spin glasses @xcite . in this paper
we want to shed some light on the character of the ground states of 2d rfim .
we have done extensive exact ground state calculations in order to characterize how the ferromagnetic ( fm ) order vanishes with increasing system size .
we have also studied the effect of the application of an external field , that is the susceptibility of the 2d rfim .
allowing for a non - zero external field makes it possible to investigate a percolation -type of critical phenomenon for the largest clusters .
we propose a phase diagram in the disorder strength and external field plane for the percolation behavior .
the presence of clusters of the size of the system , i.e. percolation type of order brings another correlation length in the systems and thus makes the decay of ferromagnetic order more complicated than at first sight .
the hamiltonian of the random field ising model is @xmath21 where @xmath22 ( in this paper we use @xmath23 for numerical calculations ) is the coupling constant between nearest - neighbor spins @xmath24 and @xmath25 .
we use here square - lattice .
@xmath2 is a constant external field , which if non - zero is assigned to all of the spins , and @xmath26 is the random field , acting on each spin @xmath24 .
we consider mainly a gaussian distribution for the random field values @xmath27 , \label{eq : gaussian}\ ] ] with the disorder strength given by @xmath6 , the standard deviation of the distribution , though in some cases the bimodal distribution , @xmath28 is used , too .
the results presented below should not be too dependent on the actual @xmath29 , in any case . to find the ground state structure of the rfim means that the hamiltonian ( [ eq : h ] ) is minimized , in which case the positive ferromagnetic coupling constants prefer to have all the spins aligned to the same direction .
on the other hand the random field contribution is to have the spins to be parallel with the local field , and thus has a paramagnetic effect .
this competition of ferromagnetic and paramagnetic effects leads to a complicated energy landscape and the finding the gs becomes a global optimization problem .
an interesting side of the rfim is that it has an experimental realization as a diluted antiferromagnet in a field ( daff ) . by gauge - transforming the hamiltonian of daff @xmath30 where the coupling constants @xmath31 , @xmath32 is the occupation probability of a spin @xmath24 , and @xmath33 is now a constant external field , one gets the hamiltonian of rfim ( [ eq : h ] ) with @xmath11 @xcite .
the ferromagnetic order in the rfim corresponds to antiferromagnetic order in the daff , naturally . as background , it is of interest to review a few basic results .
imry and ma used a domain - wall argument to show that the lower critical dimension @xmath15
in order to have a domain there is an energy cost of @xmath34 from the domain wall . on the other hand
the system gains energy by flipping the domain from the fluctuations of random fields , which interpreted as a typical fluctuation means that the gain is @xmath35 .
whenever @xmath36 , i.e. @xmath37 , it is energetically favorable for the system to break into domains .
however , in this paper we will point out , as it has been shown in 1d @xcite , that the @xmath35 scaling can be used only in relation to the sum of the random fields in the `` first excitation '' , but not to the droplet field energy when the gs consists of domains of different length scales . grinstein and ma @xcite derived from the continuum interface hamiltonian that the roughness of the domain wall ( dw ) in rfim scales as @xmath38 , which is consistent with @xmath15 .
later fisher @xcite used the functional renormalization group ( frg ) to obtain the roughness exponent @xmath39 and argued that due to the existence of many metastable states the perturbative rg calculations and dimensional reduction fail .
another , microscopic calculation by binder @xcite optimized the domain - wall energy in two - dimensions .
the net result is a total energy gain from random fields , @xmath40 due to domain wall decorations , which implies that the domain wall energy @xmath41 vanishes on a minimal length scale @xmath42 , \label{eq : l_b}\ ] ] where @xmath43 is a constant of order unity . for @xmath44
the expectation is that the system spontaneously breaks up into domains .
similarly the energy of a domain with a constant external field @xmath2 becomes @xmath45 setting @xmath46 and assuming that the critical length scale @xmath47 scales as @xmath0 , without the field , i.e. @xmath48 , the critical external field becomes @xmath49 .
\label{eq : hc_binder}\ ] ] note , that in this case the first two terms in ( [ eq : u_h ] ) assume that the domain is compact .
these results imply , together with the notion that the ground state is paramagnetic , that the magnetization should not as such display any `` universal '' features .
the results of this paper show that the magnetization is _ not _ dependent on the system size and has a smooth behavior of @xmath50 $ ] and the susceptibility vanishes with the system size as @xmath51 $ ] , where @xmath52 const , and @xmath53 .
these imply that there is a length scale , related to the rate at which clusters `` melt '' when @xmath2 is changed from zero .
the presence of such a length scale is qualitatively similar to the one discovered in the context of the percolation transition .
it turns out so that when the external field is varied , the universality class is that of the ordinary short - range correlated percolation universality class . the external field threshold for spanning @xmath54 with respect to decreasing random field strength approaches zero external field limit from the site - percolation limit of infinite random field strength value , suggesting a behavior for gaussian disorder of @xmath55 , where @xmath8 and @xmath9 .
below this value the lattice effects of site percolation are washed out , and there is yet another length scale that characterizes the percolation clusters , the size of the `` red clusters '' defined below in analogy to the usual red or cutting sites in percolation .
now a whole cluster is reversed due to the forced reversal of a `` seed '' cluster , when the sample is optimized again .
the length scale is however finite , indicating that the global optimization of the ground state creates only finite spin - spin correlations as is the case in 1d as well .
this paper is organized so that it starts by introducing in section [ method ] the exact ground state calculation technique . in section [ breakup ]
the breaking up of ferromagnetic order is discussed , based on a nucleation of droplets -picture which follows from a level crossing between a fm ground state and one with a large droplet .
the relevant @xmath0 scaling ( [ eq : l_b ] ) is derived from extreme statistics .
above the break - up length scale the domains have a complex structure that is briefly discussed .
the effect of an external field , in the case the system size is above @xmath0 , is studied in section [ suskis ] for gaussian disorder .
the percolation aspects of the 2d rfim are studied in detail in section [ perkolaatio_h ] . the phase diagram for the percolation behavior as functions of the external field and the random field strength
is sketched and the properties of the transition are discussed .
the zero - external field percolation probability is studied in section [ perkolaatio_0 ] . in the section also
the structure of spanning clusters is studied using the so called _ red clusters _ whose scaling and properties are discussed .
the paper is finished with conclusions in section [ concl ] .
for the numerical calculations the hamiltonian ( [ eq : h ] ) is transformed to a random flow graph with two extra sites : the source and the sink .
the positive field values @xmath26 correspond flow capacities @xmath56 connected to the sink ( @xmath57 ) from a spin @xmath24 , similarly the negative fields with @xmath58 are connected to the source ( @xmath59 ) , and the coupling constants @xmath60 between the spins correspond flow capacities @xmath61 from a site @xmath24 to its neighboring one @xmath25 @xcite . in the case
the external field is applied , only the local sum of fields , @xmath62 , is added to a spin towards the direction it is positive .
the graph - theoretical combinatorial optimization algorithms , namely maximum - flow minimum - cut algorithms , enable us to find the bottleneck , which restricts the amount of the flow which is possible to get from the source to the sink through the capacities , of such a random graph .
this bottleneck , path @xmath63 which divides the system in two parts : sites connected to the sink and sites connected to the source , is the global minimum cut of the graph and the sum of the capacities belonging to the cut @xmath64 equals the maximum flow , and is smaller than of any other path cutting the system .
the value of the maximum flow gives the total minimum energy of the system .
the maximum flow algorithms are proven to give the exact minimum cut of all the random graphs , in which the capacities are positive and with a single source and sink @xcite . in physical situations
this means the systems are without local frustration .
the algorithm was actually used for the first time in this context by ogielski @xcite , who showed that the 3d rfim has a ferromagnetic phase .
the best known maximum flow method is by ford and fulkerson and called the augmenting path method @xcite .
we have used a more sophisticated method called push - and - relabel by goldberg and tarjan @xcite , which we have optimized for our purposes .
it scales almost linearly , @xmath65 , with the number of spins and gives the ground state in about minute for a million of spins in a workstation . when we have added an external field in the systems our system sizes
are restricted to @xmath66 , for @xmath2 small but nonzero due to the range of integer variables ( for numerical reasons we use a discrete representation of real fields ) .
when the high precision for field values is not needed the computations extend up to system sizes @xmath67 .
we have used periodic boundary conditions in all of the cases . also the percolation is tested in the periodical or cylindrical way , i.e. , a cluster has to meet itself when crossing a boundary in order to span a system .
when the red clusters are studied , section [ red_clust ] , we have applied a technique , which allows us to take advantage of the so called _ residual _ graph @xcite .
after the original ground state is searched a perturbation is applied .
this means that _
e.g. _ , a spin is forced to be to the other direction with a large opposite field value .
then the ground state is searched again .
this time all the flow need not to be constructed from scratch , but instead one can utilize the final situation of the first ground state search ( the residual graph ) .
only the extra amount of flow , needed since the capacity of the large opposite field value is added , has to be forced through the system to the sink .
one has to subtract also the flow from the original field value ( retrace it back to the source ) .
it is thus convenient to reverse only the fields which originally were negative . for the positive field values one would have to study a mirror copy of the system ( @xmath68 ) .
thus we have analyzed the red clusters only from the spanning clusters of down spins , which does not disturb the statistics , since the spin directions are symmetrical .
the use of the residual graph reduces the time to calculate the next ground state considerably , although approximately a half of the spanning clusters have to be neglected .
notice that since the ground state energy is a linear function of the capacity of the saturated bonds ( or the field of the spins aligned along the local field ) one can compute the `` break - point '' field @xmath69 , at which a change takes place from the original ground state to the new one .
we have not paid attention to this , however , due to our main interest in the geometry of the red clusters .
one interesting additional question would be what is the smallest @xmath69 and its disorder - averaged distribution .
in this section we will derive the scaling for the break - up length scale @xmath0 , eq .
( [ eq : l_b ] ) , from extreme statistics ( as done in the paper by emig and nattermann @xcite ) . and
confirm it with exact ground state calculations .
we also discuss the ensuing domain structure qualitatively . if one picks a ( compact ) subregion of area @xmath70 of a ferromagnetic 2d rf system the energy is drawn from a gaussian distribution @xmath71 where the variance @xmath72 is due to the fluctuations of random fields , and @xmath73 . for a system of size
@xmath74 we have @xmath75 ways of making such a subregion .
the probability that a subregion has the lowest energy @xmath76 is given by @xmath77 where @xmath78 @xcite .
the distribution @xmath79 is in fact a gumbel distribution @xcite .
the average value of the lowest energies is given by @xmath80 which can not be solved analytically .
the typical value of the lowest energy follows from an _ extreme scaling _ estimate .
the factor inside the curly brackets in ( [ global ] ) is close to unity if @xmath81 becomes small enough ( for similar applications , see @xcite ) .
thus @xmath82 which yields , @xmath83 i.e. the energy gain from the fluctuations is @xmath84 a fm system would tend to take advantage of such large favorable energy fluctuations by reversing a domain , which requires breaking bonds .
this is assumed to have a cost of @xmath85 equating eqs .
( [ gain ] ) and ( [ cost ] ) yields the estimate of the parameter values at which the first `` imry - ma '' domain occurs , @xmath86 it can be easily understood that the most preferable domain is the one , which maximizes the area and minimizes the bonds to be broken , which gives @xmath87 . fig .
[ fig1 ] illustrates this , as we increase ( with a fixed random field configuration and system size ) the strength of the randomness or decrease the ferromagnetic couplings until the first domain appears .
it turns out to be so that the droplet is of the order of the system size .
this kind of nucleation with a critical size is reminiscent of a first order transition , and is related to a level - crossing , when either the random field strength or the system size is varied , similarly to random elastic manifolds , when an extra periodic potential @xcite or a constant external field @xcite is applied . by substituting @xmath88 and @xmath75 to ( [ imryjump ] ) , we get for the length scale @xmath89 , \label{eq : l_b_2}\ ] ] which is in fact @xmath0 as in ( [ eq : l_b ] ) .
this result , ( [ eq : l_b_2 ] ) , is surprising in the sense , that the extreme statistics calculation for the formation of a domain leads to the exactly same scaling as the optimization of a domain wall energy on successive scales in binder s argument . due to the extensive size of the first domain - like excitation the destruction of the ferromagnetism
resembles a first order transition .
the magnetization for a certain disorder strength and system size would be averaged over systems , in which the excitation has and has not been formed yet , with @xmath90 and @xmath91 , respectively .
hence we define a simpler measure for the break up of fm order : the probability of finding a purely ferromagnetic system , @xmath92 , i.e. , for a fixed random field strength and system size we calculate the probability over several realizations that magnetization @xmath93 @xcite .
if the transition to the pm state would be continuous , this would not make much sense , since already small fluctuations would cause @xmath94 .
however , due to the first order behavior , and to the fact that the smallest energy needed to flip a domain causes the excitation to be large , @xmath95 is a good measure and has a smooth behavior .
we have checked that @xmath96 vs. @xmath95 does not depend on @xmath13 .
we have derived the break - up length scale @xmath0 by varying the random field strength @xmath6 from the probability of finding a pure ferromagnetic system as @xmath97 .
the data is shown in fig . [ fig2 ] for gaussian and bimodal disorder ( in both cases @xmath23 ) , and the exponential scaling for @xmath0 vs. inverse random field strength squared is clearly seen .
the prefactors are @xmath98 and @xmath99 for gaussian and bimodal disorder , respectively .
to check that the probability @xmath100 is not due to so called stiff spins , i.e. , single spins for which @xmath101 , we next derive an extreme statistics formula for their existence . the probability of finding @xmath101 @xmath102 .
the extreme statistics argument , @xmath103 , with @xmath104 , gives @xmath105^{-1/2}. \label{eq : l_1}\ ] ] for gaussian @xmath106 for which @xmath107 in fig .
[ fig2 ] @xmath108 from eq .
( [ eq : l_1 ] ) .
@xmath109 also grows much faster than @xmath0 for decreasing @xmath6 which is both easy to see from eq .
( [ eq : l_1 ] ) and to check numerically . for @xmath110
for which @xmath111 @xmath109 becomes as huge as 22300 . to confirm
further that the origin for the break - up is a large domain , one can extend the argument to small domains .
the length scale @xmath112 at which one is able to find a cluster of two neighboring spins flipped , i.e. , @xmath113 , where @xmath114 , @xmath115 , becomes even greater than @xmath109 .
these small clusters are present in large system sizes , but do not play a role in the probability of first excitation , since the energy minimization prefers extensive domains .
it is amusing to note that the critical droplet size reminds of critical nucleation in ordinary first order phase transitions .
it is also worth pointing out that the reasoning for stiff spins does not work for the bimodal distribution ( since it is bounded ) , and indeed we observe as expected a similar @xmath0 scaling for both the gaussian and bimodal disorders . when a system size is well above the break - up length scale the imry - ma argument is no longer applicable to the structure . in fig .
[ fig3 ] we depict two systems with a large gaussian random field strength value @xmath116 and with a smaller one @xmath117 for a system size @xmath118 .
one can see that a system breaks to smaller and smaller domains inside each other from the case seen in fig .
the feature of having clusters in different scales is familiar from the percolation problem @xcite .
in fact in both of the examples in fig .
[ fig3 ] there is a domain , which spans the system in vertical direction , drawn in gray . for the stronger random field value one can see also smaller domains of different sizes
however , the width of the spanning cluster is greater than in a standard site or bond occupation percolation problem . later in subsection [ perc - clust ]
we discuss the scaling properties of the largest clusters in the ground state , above @xmath0 .
in fig . [ fig4 ] we show what happens in system , with a system size well above @xmath0 when an external field @xmath2 is applied . now
the clusters `` melt '' smoothly when the external field strength is increased and such first order type of phenomenon is not seen as when a first imry - ma droplet appears in the zero - field case .
the magnetization has a continuous behavior , see fig .
[ fig5](a ) , where we have the magnetization with respect to the external field for several gaussian disorder strength values .
all the magnetization values for different system sizes lie exactly on top of each other , when @xmath44 , and as long as the statistics is good . for smaller system sizes @xmath119 one could study `` avalanche''-like behavior , i.e. the sizes of the areas that get turned with the magnetic field ( see @xcite )
. however , these are due to the first order break - up , defined by ( [ eq : hc_binder ] ) and one should bear in mind that such behavior does not exist in the thermodynamic limit , @xmath120 , when the system sizes are above the break - up length scale . for @xmath3 our results indicate that the size distribution of the flipped regions as @xmath2 is swept is not that interesting . in order to find the scaling between the external field and the random field strength we have taken from fig .
[ fig5](a ) the crossing points of magnetization curves with a fixed magnetization values at external fields @xmath121 for different random field strength values @xmath6 .
the external field @xmath121 scales exponentially with respect to the random field strength , @xmath122 see fig .
[ fig5](b ) .
this is also an evidence of _ non - existence _ of a critical point in @xmath6 , in which case there should be a power - law behavior if the transition was continuous , thus no pm to fm transition is seen .
the data - collapse using the scaling ( [ eq : h_d_magn ] ) is shown in fig .
[ fig5](c ) confirming the prediction of the scaling .
the magnetization has a linear behavior with respect to the external field for small field values @xmath2 and exponential tails .
the exponential behavior of eq .
( [ eq : h_d_magn ] ) implies that there is a unique `` melting rate '' at which the cluster boundaries get eroded as @xmath2 increases and that the process is otherwise similar for all @xmath6 .
we have no analytical argument for the melting rate [ or the slope of the @xmath123-curve ] , and note that it is not seemingly , at least , related to @xmath0 .
we have also studied the susceptibility , @xmath124 , with respect to the external field . in fig .
[ fig6](a ) the susceptibility is shown for a fixed random field strength @xmath125 and varying system size .
the data has been collapsed with the area of the systems , @xmath126 . in fig .
[ fig6](b ) we have data - collapsed the susceptibility versus random field strength by scaling the external field with ( [ eq : h_d_magn ] ) as for magnetization and the susceptibility with @xmath127 .
again the exponential behavior is a sign of non - existence of any critical point , due to the lack of power - law divergence at any @xmath10 .
although the shape of the data - collapse of the susceptibility looks almost gaussian , it is actually not .
it has a constant value for small external field values @xmath2 and exponential tails for large values , as seen in fig .
[ fig6](c ) .
this results straightforwardly from the magnetization , since @xmath128 . to summarize the behavior of the susceptibility , it is @xmath129 where @xmath121 is from eq .
( [ eq : h_d_magn ] ) and @xmath130 therefore the fluctuations of the magnetization are associated with yet another scale , which is almost but not quite an inverse of that related to the magnetization . from the suscpetibility one gets the magnetization correlation length @xmath131 , which has an exponential dependence on the random field strength .
it should be noted , finally , that we have here studied only the case with gaussian disorder . with any other distribution
we would expect that the prefactors in eqs .
( [ eq : h_d_magn ] ) , ( [ eq : chi ] ) , and ( [ eq : chi_g ] ) would change .
motivated by fig . [ fig3 ] , where the domains resemble the percolation problem we next study the percolation behavior in the 2d random field ising magnets with gaussian disorder .
the usual bimodal distribution could be studied as well , but since it is susceptible to some anomalous features we concentrate on the gaussian case which does not have these problems .
the bimodal case suffers from the fact that the ground states are highly degenerate at fractional field strength values .
thus there are some ambiguities in how define percolation clusters @xcite .
when the random field strength is well above the coupling constant value , @xmath132 , the percolation can be easily understood by considering it as an ordinary site - occupation problem .
this means that , only the random field directions are important and the coupling constants may be neglected .
the site - percolation occupation threshold probability for square - lattices is @xmath133 @xcite , i.e. , well above one half . applied to the strong random field strength case it means that there must be a finite external field in order to get a domain spanning the system .
however , when the random field strength is decreased , the coupling constants start to contribute and in some cases systems span even without an external field , as in fig .
hence , we propose a phase - diagram fig .
there we can take the limit @xmath134 so that the ordinary site - percolation problem is encountered .
this is true for distributions for which one can control the fraction of `` stiff '' spins ( i.e. @xmath135 ) systematically . in the case of gaussian disorder
there will be of course , even for @xmath6 very large a small fraction of `` soft '' spins where this criterion is not fulfilled .
thus the exact point that the percolation line approaches in the @xmath136-limit will depend on the distribution , but we expect that the `` @xmath137 '' is different from one - half , that is @xmath138 .
notice that again the binary distribution presents a problem . when @xmath139 the percolation threshold lines start to approach each other and the @xmath11 line .
now there arises two questions .
the first one is , what kind of a transition is the percolation here ? is it like the ordinary short - range correlated percolation , suggested by the site - percolation analogy for the strong random field strength case , or are there extra correlations due to the global optimization relevant here ?
examples about similar cases can be found from ref .
the second question is , do the lines meet each other at finite @xmath10 , i.e. does there exist a spanning cluster also when @xmath11 and @xmath140 ?
our aim is to answer these two questions in this section , where we study the percolation problem in the vertical direction in the phase - diagram fig .
[ fig7 ] , and in the next section , where the horizontal direction , @xmath11 line , is considered . in fig .
[ fig8](a ) we have drawn the spanning probabilities of up spins @xmath141 with respect to the external field @xmath2 for several system sizes @xmath13 , which are greater than @xmath0 , for a fixed random field strength @xmath142 .
the curves look rather similar to the standard percolation .
when we take the crossing points @xmath143 of the spanning probability curves with fixed spanning probability values for each systems size @xmath13 , we get an estimate for the critical external field @xmath54 using finite size scaling , see fig .
[ fig8](b ) .
there we have attempted successfully to find the value for @xmath54 using the standard short - range correlated 2d percolation correlation length exponent @xmath144 . using the estimated @xmath145 for @xmath146 we show a data - collapse of @xmath141 versus @xmath147 in fig .
[ fig8](c ) , which confirms the estimates of @xmath54 and @xmath144 @xcite .
we get similar data - collapses for various other random field strength values @xmath6 as well . in order to test further the universality class of the percolation transition studied here , we have also calculated the order parameter of the percolation , the probability of belonging to the up - spin spanning cluster @xmath148 . using the scaling analysis for the correlation length @xmath149 and for the order parameter , when @xmath150 , @xmath151 we get the limiting behaviors , @xmath152 and thus the scaling behavior for the order parameter becomes @xmath153 \nonumber \\
\sim l^{-\beta/\nu } f\left(\frac{h - h_c}{l^{-1/\nu}}\right ) .
\label{p_infty_scale}\end{aligned}\ ] ] we have done successful data - collapses , i.e. plotted the scaling function @xmath154 , for various @xmath6 using the standard 2d short - range correlated percolation exponents @xmath155 and @xmath156 , of which the case @xmath116 with @xmath157 is shown in fig .
note , that the values @xmath158 are not shown , since cut - offs appear , due to the fact that @xmath148 is bounded in between @xmath159 $ ] and the non - scaled @xmath160 values after scaling saturate at different levels depending on the system size .
hence , we conclude that the percolation transition for a fixed @xmath6 versus the external field @xmath2 is in the standard 2d short - range correlated percolation universality class @xcite .
this is confirmed by the fractal dimension of the spanning cluster , too , as discussed below .
also other exponents could be measured , as @xmath161 for the average size @xmath162 of the clusters , and @xmath163 and @xmath164 for the cluster size distribution .
note , however , that then the control parameter should be the external field @xmath2 instead of the disorder strength @xmath6 .
see ref .
@xcite for an example of the cluster size distribution for a non - critical case ( @xmath11 , but @xmath165 ) .
the other exponents should be measurable , too , like the fractal dimension of the backbone of the spanning cluster , the fractal dimension of the chemical distance , the hull exponent etc .
in addition to the correct control parameter also the break - up length scale , has to be considered , too .
notice that there is an slight contradiction hidden in the notion that the hull exponent could be measured .
namely , both the work of ref .
@xcite and studies of domain walls enforced with appropriate boundary conditions give no evidence thereof .
it seems likely that to recover the right exponent ( 4/3 ) one has to resort to studying the spanning cluster geometry itself at the critical point with @xmath44 .
it is amusing to note that the standard 2d percolation hull exponent can be recovered in non - equilibrium simulations of 2d rfim domain walls @xcite .
we have now shown , that there exists a line of critical external field values @xmath166 for percolation .
the corresponding correlation length @xmath167 diverges as ( [ xi_perc ] ) with a correlation length exponent @xmath144 . on the other hand it was shown in the previous section ,
that there is no critical external field value for the magnetic behavior , i.e. , no pm to fm transition , and the magnetic correlation length @xmath168 has an exponential dependence on @xmath6 .
the percolation correlation length @xmath167 may cause some confusion , when studying the pm structure of the gs , since it brings in another length scale . to answer the question how the percolation critical external field @xmath166 behaves with respect to the random field , we have attempted a critical type of scaling using the calculated @xmath54 for various @xmath1692.0 , 2.2 , 2.6 , 3.0 , and 4.5 . for smaller @xmath6 @xmath0 becomes large and @xmath54 approaches the vicinity of zero , being thus numerically difficult to define .
we have been able to use the ansatz behavior of @xmath170 where @xmath9 . in fig .
[ fig10](a ) we have plotted the calculated @xmath6 values versus the scaled critical external field @xmath171^{1/2.05}$ ] and it gives the estimate for @xmath172 .
this indicates that the percolation probability lines for up and down spins meet at @xmath173 and for @xmath6 below the critical @xmath10 there is always in the systems spanning of either of the spin directions , even for @xmath11 .
actually one should note , that the only way that neither of the spin directions span is to have a so called checker - board situation , which prevent both of the spin directions to have neighbors with the same spin orientation .
however , the another scenario with an exponential behavior for @xmath166 fits also reasonably well .
this would suggest , that there is no finite @xmath10 .
[ fig10](b ) shows a behavior of @xmath174 .
this can be compared with eq .
( [ eq : hc_binder ] ) , where the break - up external field was derived .
notice that the derivation was for compact domains and the spanning clusters here are by default fractals . besides that
, the factor 13 in front of @xmath175 is much larger than @xmath176 in the scaling form for @xmath0 .
the difference implies that the @xmath177 , at which length scale the spanning probability vanishes , scales as @xmath178 .
the @xmath0 is already exponentially large length scale for small @xmath6 , so @xmath177 should be large enough that one can be below it in experiments , and thus a system can `` apparently percolate '' @xcite .
to understand how the percolation transition is seen when there is no external field and the random field strength is changed we study the phase diagram in section [ span_perk_0 ] in the direction of the horizontal arrow in fig . [ fig7 ] .
the structure of the spanning clusters are studied in sections [ perc - clust ] and in [ red_clust ] with the help of the so called _ red clusters_. in fig .
[ fig11](a ) we have plotted the probability for spanning of either up or down spins @xmath179 as a function of the gaussian random field strength @xmath6 . the probabilities are calculated up to @xmath180 , but only the interesting part of the plot is shown . there is a drop from @xmath181 at @xmath6 , which corresponds @xmath182 for each system size , to a value about @xmath183 .
we have also calculated the @xmath141 in this case and it is approximately one half of @xmath179 . for the larger random field strength values the probabilities @xmath179
decrease and the lines get steeper , when the system size increases . in order to see , if the spanning probabilities are converging towards a step function at some threshold value , we have calculated the probabilities up to the system size @xmath184 and each point with 5000 realizations . for each system size @xmath13
we have searched the crossing points @xmath185 of the spanning probability curves in fig .
[ fig11](a ) with fixed probability values @xmath1860.1 , 0.15 , 0.2 , 0.25 , 0.3 , 0.4 , and 0.5 . using finite size scaling for @xmath185 of the form @xmath187 we have estimated @xmath10 for each @xmath179 value , see fig .
[ fig11](b ) .
there we have plotted the @xmath185 versus the scaled system size @xmath188 .
one sees that the different threshold values for spanning probabilities @xmath179 approach different critical random field strength values @xmath10 .
the threshold @xmath179 s have been plotted with respect to @xmath10 in fig .
[ fig11](c ) . in the ordinary percolation
, this should be a step function , and the correlation length exponent @xmath189 independent on the criterion @xmath179 .
however , here also @xmath190 is dependent on the criterion @xmath179 and varies with respect to @xmath191 .
we believe , that this surprising phenomenon is due to that we are approaching the part in the phase diagram fig .
[ fig7 ] , where the percolation lines of up and down spins are getting close to each other . in terms of the two control parameters @xmath6 and @xmath2 one can think about the `` percolation manifold '' : it has a line of unstable fixed points @xmath166 .
usually @xmath2 is a good control parameter close to @xmath54 .
having @xmath6 as a control parameter seems to have the problem that one moves almost parallel to the actual line @xmath166 .
when considering the percolation probability of up or down spins , it actually consists of probabilities of up - spin spanning @xmath141 and down - spin spanning @xmath192 as @xmath193 , since they are correlated with each other . assuming that @xmath141 ( and @xmath192 respectively ) has a value about one - half at the critical line of percolation at the thermodynamic limit , we get @xmath194 . in standard percolation
such a value is not actually universal ( and we have not confirmed it ) but depends on the boundary conditions , etc .
however , whatever the values for @xmath141 and @xmath192 are at the thermodynamic limit , as long as they are below unity , @xmath179 is below unity , too .
this may be the reason , why there is an immediate drop in fig .
[ fig11](a ) from @xmath181 at @xmath195 for each system size , to a value about @xmath183 . if we approximate with a linear behavior the @xmath179 versus @xmath191 in fig .
[ fig11](c ) , the critical value estimated in the previous section @xmath8 , when the percolation threshold @xmath196 for up - spin spanning , has a value about @xmath197 .
another interesting point in fig .
[ fig11](c ) is that the @xmath190 is about 3/4 , when @xmath179 versus @xmath191 approaches zero .
thus the standard correlation length exponent would be reached far enough away from the area , where the percolation threshold lines for up and down spins touch each other . in order to test the break - up length scale type of scaling for the percolation behavior ( [ eq : l_b ] ) , we have taken from fig . [ fig11](a ) the estimated @xmath185 for various @xmath179 values and plotted the system sizes in double - logarithm scale versus the logarithm of the inverse of the critical @xmath185 , see fig . [
fig11](d ) .
the exponent , which is @xmath198 in @xmath0 scaling , @xmath199 , is now dependent on @xmath179 again .
at least this does not solve the problem here , and the break - up length scale type of scaling can be ruled out . in order to see if the thickness of the spanning cluster affects the scaling of the standard percolation we have measured the fractal dimension of the spanning cluster when @xmath11 .
by now unsurprisingly , the standard two - dimensional short - range correlated percolation fractal dimension @xmath200 fits very well in the data , as can be seen in fig .
[ fig12 ] .
the least - squares fit gives a value of @xmath201 .
we have measured also the sum of the random fields in the spanning cluster and found that the sum scales with the exponent @xmath200 , too .
this is in contrast to the imry - ma domain argument , where the sum is taken scale as @xmath202 .
the prefactor for the scaling of the sum of the random fields approaches slowly zero with decreasing random field strength , opposite to the mass of the spanning cluster , which increases with decreasing @xmath6 .
hence , the imry - ma argument defines only the _ first _ excitation , and is irrelevant when it comes to domains when the system has broken up to many clusters on different length scales .
then the structure is due to a more complicated optimization .
the domains are no longer compact and as noted above for large enough domains the domain - wall length should be characterized by the percolation hull exponent . so far all the evidence points out to the direction that the percolation transition is exactly of the normal universality class .
to further investigate the nature of the clusters in the presence of the correlations from the gs optimization , we next look at the so - called _ red clusters_. the structure of a standard percolation cluster can be characterized with the help of the `` colored sites '' picture in which one assesses the role of an element to the connectivity of the spanning cluster .
this picture has been also called the _ links - nodes - blobs _ model with dead - ends @xcite .
the red sites , or links and nodes , are such that removing any single one breaks up the spanning cluster . to compare with the original ground state we investigate what happens if one inverts , by fixing the local field @xmath26 to a large value opposite to the spin orientation , any spin belonging to the spanning cluster .
then the new gs is found with this change to the original problem .
the effect is illustrated in fig .
[ fig3 ] , the crucial difference to site - percolation is that now a whole sub - cluster can be reversed .
the spin drawn in yellow is the inverted , `` seed '' spin in the spanning cluster , and the spins painted red form the rest of the red cluster , which are flipped from the original ground state when the energy is minimized the second time .
we do the investigation whether the original cluster retains its spanning property for each spin or trial cluster in analogy with ordinary percolation .
those spins that lead to a destructive ( cluster ) flip , define then _ red clusters _ ( rc ) as all the spins that reversed simultaneously .
the finite size scaling of the number of the red clusters , @xmath203 , is shown in the fig .
[ fig13](a ) for different field values .
@xmath204 is in practice calculated as the number of the seed spins , which cause the breaking of the spanning cluster , since two different seed spins may both belong to the same red clusters without the red clusters being identical .
the technique to find the red clusters was introduced in section [ method ] and although it is efficient , only up to the system size @xmath205 can be studied , since each of the spins in spanning clusters has to be checked separately , because one can not know beforehand , whether it is critical or not - this is what we want to find out . for smaller field values the spanning cluster
is `` thicker '' and the red clusters get larger .
one can see from the fig .
[ fig13](a ) that @xmath206 scales with @xmath207 , where @xmath208 as in ordinary percolation , for field values @xmath209 , when @xmath44 .
the amplitude is larger the smaller the field , as is the average mass of red clusters @xmath210 .
@xmath210 is independent of the system size @xmath13 and depends only on the field @xmath6 , see fig .
[ fig13](b ) .
the other elements of the spanning cluster , dead - ends and blobs , could be generalized , too . here
blobs , which are multiple connected to the rest of the spanning cluster , are such that in order to break a spanning cluster , several seed spins are needed to flip simultaneously instead of a single one .
links , nodes , and blobs form together the backbone of the spanning cluster , and the rest of the mass of the cluster is in the dead - ends .
the red cluster size scale defines the average smallest size of any element of the spanning cluster .
in this paper we have studied the character of the ground state of the two - dimensional random field ising magnet .
we have shown that the break - up of the ferromagnetism , when the system size increases , can be understood with extreme statistics .
this length scale has been confirmed with exact ground state calculations .
the change of magnetization at the droplet excitation is naturally of `` first - order '' -kind . above the break - up length scale
we have studied the magnetization and susceptibility with respect to a constant external field .
the behavior of the magnetization and the susceptibility is continuous and smooth and does not have any indications of a transition or a critical point , in agreement with the expectations of a continuously varying magnetization around @xmath11 , and a paramagnetic ground state .
we thus conclude that the correct way of looking at the susceptibility is to study it with respect to the external field and above the break - up length scale instead of as a function of the random field strength , when the first order character of the break - up length scale may among others cause problems
. however , we are able to find another critical phenomenon in the systems , in their geometry . for square lattices sites do not have a spanning property in ordinary percolation , when the occupation probability is one half .
this corresponds to the random field case with high random field strength value without an external field . when an external field is applied and the random field strength decreased , a percolation transition can be seen .
the transition is shown to be in the standard 2d short - range correlated percolation universality class , when studied as a function of the external field .
hence , the correlations in the two - dimensional random field ising magnets are only of finite size .
we also want to point out that in these kind of systems , the random field strength is a poor control parameter and the systems should be studied with respect to the external field , and after that map to the random field strength . by doing so we have been able to find a critical random field strength value , below which the systems are always spanning even without an external field . when the percolation transition is studied without an external field and tuning the random field strength lots difficulties
are encountered .
this might cause puzzling consequences when studying the character of the ground states , not only because of the bad control parameter , but also because the percolation correlation length may be mistaken for something as the magnetization correlation length
. also note that the `` true behavior '' is seen only for system sizes large enough ( @xmath211 ) .
the percolation character of the ground state structure can be measured by the standard percolation fractal dimensional scaling for the mass of the spanning domain .
the existence of such a large cluster is not against the paramagnetic structure of the ground state , since the fractal dimension is below the euclidean dimension . in order to be consistent with the aizenman - wehr argument in the zero - external field limit
the spins in the opposite direction from the external field may form the spanning cluster at low random field strength values .
in fact we have found cases of finite systems for @xmath11 , where the magnetization of the system is opposite to the orientation of spins in the spanning cluster .
notice that this does not imply that the critical lines @xmath212 actually cross each other at @xmath10 continuing on the opposite side of the @xmath11-axis ( see fig .
[ fig7 ] ) . by considering
the red clusters it seems that in the td - limit the spanning cluster should be broken up at @xmath213 , @xmath214 since the field needed to flip such a critical droplet should go to zero with @xmath13 .
also , since the sum of the fields in the spanning cluster is shown to scale with the same fractal dimension as the mass , we conclude that the imry - ma argument does not work any more after the system has broken up on several domains .
it works only for the first domain to appear .
we have also generalized the red sites of the standard percolation to red clusters in the percolation studied here . a red cluster results from the energy minimization by flipping a whole cluster although only a single spin has been forced to be flipped , and breaks up the spanning character of a percolating cluster .
actually the finite size of the red clusters indicate also the presence of only short - range correlations in the systems .
such a lack of long - range correlations maybe explains , why we can see an `` accidental '' percolation phenomenon in a zero - temperature magnet whose physics is governed by the disorder configuration . the normal percolation universality class is tightly connected to conformal invariance , which is most often destroyed by long - range correlations or randomness @xcite .
to finish the paper , we would like to raise some open questions related to the percolation behavior of the ground states of the two - dimensional random field ising magnets . as noted , an interesting problem is the exact relation of the rfim percolation to conformal invariance .
the percolation characteristics of the ground state might be experimentally measurable since the overlap of the ground state and finite temperature magnetization should be close to unity for small enough temperatures .
the structure and relaxation of diluted antiferromagnets @xcite in low external fields are suitable candidates : there one would presume it to be of relevance that there are large - scale structures present in the equilibrium state .
in particular in coarsening , it is unclear how the eventual hull exponent of 4/3 would affect the dynamics
. it would be interesting to see what kind of phenomena can be seen in the structure on triangular lattices since here @xmath215 even in the ordinary site - percolation .
one open question or application is the 3d rfim .
the percolation transition of the minority spins is expected to take place along a line in the ( @xmath2 , @xmath6 ) -phase diagram as well , since @xmath216 for site - percolation in the case of the cubic systems most often studied numerically .
thus in low fields only one of spin orientations percolates whereas at high fields both do , see a review of 3d rfim experiments in @xcite . the role of this transition is unclear also when it comes to the ferro- to paramagnet phase boundary , and the nature of the phase transition .
this work has been supported by the academy of finland centre of excellence programme .
we acknowledge dr .
cristian moukarzel for discussions . | the ground state structure of the two - dimensional random field ising magnet is studied using exact numerical calculations .
first we show that the ferromagnetism , which exists for small system sizes , vanishes with a large excitation at a random field strength dependent length scale .
this _ break - up length scale _
@xmath0 scales exponentially with the squared random field , @xmath1 . by adding an external field @xmath2 we then study the susceptibility in the ground state .
if @xmath3 , domains melt continuously and the magnetization has a smooth behavior , independent of system size , and the susceptibility decays as @xmath4 .
we define a random field strength dependent critical external field value @xmath5 , for the up and down spins to form a percolation type of spanning cluster .
the percolation transition is in the standard short - range correlated percolation universality class .
the mass of the spanning cluster increases with decreasing @xmath6 and the critical external field approaches zero for vanishing random field strength , implying the critical field scaling ( for gaussian disorder ) @xmath7 , where @xmath8 and @xmath9 .
below @xmath10 the systems should percolate even when @xmath11 .
this implies that even for @xmath11 above @xmath0 the domains can be fractal at low random fields , such that the largest domain spans the system at low random field strength values and its mass has the fractal dimension of standard percolation @xmath12 .
the structure of the spanning clusters is studied by defining _
red clusters _ , in analogy to the `` red sites '' of ordinary site - percolation .
the size of red clusters defines an extra length scale , independent of @xmath13 .
0.4pt _ pacs # 05.50.+q , 75.60.ch , 75.50.lk , 64.60.ak_ 2 [ ] |
when studying the extremal limit of black hole solutions in gravity one immediately encounters a rather unusual geometric feature .
the geometry develops a throat which becomes infinitely deep at extremality .
this can be seen in the simple example of a @xmath6-dimensional reissner - nordstrm black hole whose metric is : @xmath7 the inner and outer horizon lie at @xmath8 and @xmath9 . in the extremal limit @xmath10
tends to @xmath11 and the proper distance between some finite distance @xmath12 and the horizon diverges .
one can isolate the geometry parametrically near the horizon within the infinitely deep throat and find a new solution to the einstein - maxwell system , namely the bertotti - robinson geometry ads@xmath13 .
the ads@xmath1 structure is not contingent upon supersymmetry but rather extremality , i.e. that the temperature of the horizon vanishes .
turning on the slightest temperature will destroy the precise ads@xmath1 structure and bring the horizon back to a finite distance . in the extremal near horizon region , the @xmath14 isometry group is enhanced to an @xmath15 . while the hawking temperature of the black hole vanishes , the entropy , given by one - quarter of the size of the horizon in planck units , is still macroscopically large .
it is interesting that a full @xmath2 symmtetry ( rather than just a dilatation symmetry ) emerges in the near horizon region given that we are only loosing a single scale , the temperature . from a holographic point of view
, the surprise stems from the fact that in quantum mechanics invariance under dilatations does not generally imply invariance under the full @xmath2 conformal transformations @xcite
. we can study such extremal , and in addition supersymmetric , black holes in string theory .
they are given by wrapping branes around compact dimensions such that they are pointlike in the non - compact directions .
the electric and magnetic charges of these branes are ( schematically ) given by the number of times they wrap around various compact cycles .
a simple and beautiful example @xcite is given by taking type iia string theory compactified on a small @xmath16 and wrapping @xmath17 d4-branes around the @xmath18 and sprinkling @xmath19 d0-particles on top .
in the large @xmath19 and @xmath17 limit , this gives an extremal d0-d4 black hole with an ads@xmath20 near horizon .
using @xmath21-duality on the @xmath22 , we can map this system to a d1-d5 system in type iib string theory .
since the circle becomes effectively non - compact in the iib frame , our former black hole now becomes a black string .
this black string has an ads@xmath23 near horizon with an @xmath24 isometry group .
we can express the ads@xmath25 as a hopf fibration of the real line over an ads@xmath1 base space .
the @xmath2 isometry of the ads@xmath1 base space is the left ( or right ) moving part of the isometries of the ads@xmath25 .
the real line becomes the @xmath21-dual direction and thus we see how the original ads@xmath1 naturally fits inside the ads@xmath25 . from the point of view of the ads / cft correspondence ,
the ads@xmath25 is dual to a certain two - dimensional cft and the large black hole degeneracy is counted by the supersymmetric ground states of the cft which only excite one of the two @xmath2 s .
the isometry of the near horizon supersymmetric ads@xmath1 reflects the remaining @xmath2 .
there is another class of extremal black holes which arise in string theory which can be obtained by considering a calabi - yau compactification , with _ generic _
@xmath26 holonomy , of eleven dimensional m - theory down to five non - compact dimensions .
typically , such a calabi - yau will not have @xmath27 isometries that we can easily @xmath21-dualize along . yet , wrapping branes around the supersymmetric cycles of the calabi - yau yields extremal five - dimensional black holes @xcite , which again have an @xmath2 isometry in the near horizon .
this ads@xmath1 does not naturally fit inside an ads@xmath25 with a two - dimensional cft dual . as a degenerate limit of the warped ads@xmath25/nhek near horizon geometry of the rotating black hole , in the limit of vanishing angular momentum .
the @xmath2 might then be a global subgroup of the full symmetries associated to the duals of such geometries @xcite .
] remarkably , attempts to count the microstates of such supersymmetric black holes have faced significant difficulties @xcite .
in fact , whenever the precise counting of microstates has been successful it has involved a cardy formula @xcite . lacking the larger virasoro structure ,
one may wonder where the ` isolated ' @xmath2 of these black holes originates and to what extent it is robust .
perhaps an additional motivation for understanding such an ` isolated ' ads@xmath1 geometry is the emergence of an @xmath2 symmetry in the worldline data of the static patch of de sitter space @xcite .
it is interesting to note that the static patch of four - dimensional de sitter space is conformally equivalent to ads@xmath28 , whose ` isolated ' @xmath2 does not seem to reside within a larger structure containing a virasoro algebra .
one particular way the @xmath2 isometeries of the black hole manifest themselves is in the worldine dynamics of d - particles propagating in the near horizon region .
we might then ask whether there are microscopic models , such as matrix quantum mechanics models with a large number of ground states , whose effective eigenvalue dynamics describe an @xmath2 invariant multiparticle theory .
hermitean matrix models @xcite in the double scaling limit has eigenvalue dynamics described by free fermions which naturally have an @xmath2 symmetry .
of course , such models contain only the eigenvalue degrees of freedom due to the @xmath29 gauge invariance that allows for a diagonalization of the matrix , and hence do not have @xmath30 degrees of freedom .
these models are dual to strings propagating in two dimensions ( for some reviews see @xcite ) . ]
some insight into these issues can be provided by studying certain quiver quantum mechanics models , which capture the low energy dynamics of strings connecting a collection of wrapped branes @xcite in a calabi - yau compactification of type iia string theory to four - dimensions . under certain conditions
these quiver theories have an exponential number of ( supersymmetric ) ground states whose logarithm goes as the charge of the branes squared , which is the same scaling as the entropy of a supersymmetric black hole in @xmath31 supergravity .
it has been argued @xcite that the near horizon ads@xmath1 of these supersymmetric black holes is related to the exponential explosion in the number of ground states in the quiver quantum mechanics .
the states in question are referred to as pure - higgs states since they reside in the higgs branch of the quantum mechanics , where all the branes sit on top of each other .
interestingly , going to the coulomb branch after integrating out the massive strings stretched between the wrapped branes , leads a non - trivial potential and velocity dependent forces governing the wrapped brane position degrees of freedom in the non - compact space .
moreover , whenever the higgs branch has an exponentially large number of ground states , the coulomb branch exhibits a family of supersymmetric scaling solutions @xcite continuously connected to its origin .
the equations determining the positions of the wrapped branes in such supersymmetric zero energy scaling solutions are reproduced in four - dimensional @xmath0 supergravity @xcite , believed to be the appropriate description of the system in the limit of a large number of wrapped branes . in this note
we would like to touch upon some of these issues .
we do so by discussing two aspects of the coulomb branch of such quiver theories , particularly those describing three wrapped branes containing scaling solutions .
first we derive the coulomb branch lagrangian of a three node quiver model ( see figure [ quiver ] ) , and establish the existence of a low energy scaling limit where the theory exhibits the full @xmath2 symmetry of conformal quantum mechanics @xcite .
these scaling theories have velocity dependent forces , a non - trivial potential as well as a metric on configuration space .
it is also worth noting that the full quiver quantum mechanics theory is itself _ not _ a conformal quantum mechanics ( and most certainly not a two - dimensional conformal field theory ) .
the emergence of a full @xmath2 symmetry rather than only a dilatation symmetry in the scaling limit is not guaranteed , and is reminiscent of the emergence of a full @xmath2 in the near horizon geometry of extremal black holes .
second , we study the behavior of the coulomb branch upon integrating out the strings in a thermal state , rather than in their ground state . at sufficiently high temperatures ,
the coulomb branch melts into the higgs branch .
this is reminiscent of the gravitational analogue where increasing the temperature of a black hole increases its gravitational pull , or a particle falling back into the finite temperature de sitter horizon .
in this section we discuss the quiver quantum mechanics theory and its coulomb branch .
these theories constitute the low energy , non - relativistic and weakly coupled sector of a collection of branes along the supersymmetric cycles of a calabi - yau three fold .
the wrapped branes look pointlike in the four - dimensional non - compact minkowski universe .
the @xmath32 supersymmetric quiver quantum mechanics comprises the following fields : chiral multiplets @xmath33 and vector multiplets @xmath34 .
the @xmath35 are the fermionic superpartners of the @xmath36 , the @xmath37 are the fermionic superpartners of the scalars @xmath38 , @xmath39 is a @xmath27 connection , and @xmath40 and @xmath41 are auxiliary scalar fields .
the @xmath42 transform in the @xmath43 of the @xmath44 .
the index @xmath45 denotes the specific arrow connecting node @xmath46 to node @xmath47 ( see figure [ quiver ] ) .
the chiral multiplets encode the low energy dynamics of strings stretched between the wrapped d - branes of mass @xmath48 sitting at three - vector positions @xmath38 in the non - compact four - dimensions .
the index @xmath49 denotes the particular wrapped d - brane .
the electric - magnetic charge vector , @xmath50 , of the wrapped branes depends on the particular cycles that they wrap , and the zwanziger - schwinger product of their charges are given by the @xmath51 .
the @xmath52 count the number of intersection points in the internal manifold between wrapped branes @xmath46 and @xmath47 . in what follows
we measure everything in units of the string length @xmath53 which we have set to one .
, each piece of which is given in ( [ quiverqm ] ) , ( [ quiverqm2 ] ) , and ( [ quiverqm3 ] ) .
this quiver admits a closed loop if @xmath54 , @xmath55 and @xmath56.,width=288 ] the lagrangian @xmath57 for the three - node quiver quantum mechanics @xcite describing the low energy non - relativistic dynamics of three wrapped branes which are pointlike in the @xmath58 non - compact dimensions is given by : @xmath59 and , @xmath60 in ( [ quiverqm]-[quiverqm2 ] ) and what follows we will mostly work with the relative degrees of freedom ( i.e. @xmath61 , @xmath62 , etc . ) since the center of mass degrees of freedom decouple and do not play a role in our discussion .
the notation we use is somewhat non - standard ( for example @xmath63 ) and is given in appendix [ notation ] along with all of our conventions .
we note that the relative lagrangian is only a function of two of the three vector multiplets , since @xmath64 , @xmath65 and @xmath66 .
the @xmath67 encode the orientation of the quiver . for the majority of our discussion we choose @xmath68 ,
corresponding to a closed loop like the one in figure [ quiver ] . the reduced masses @xmath69 ( which we denote in superscript notation in ( [ quiverqm]-[quiverqm2 ] ) ) are related to the masses @xmath70 , where @xmath71 is the string coupling constant , of the wrapped branes sitting at the @xmath38 by : @xmath72 the superpotential , which is allowed by gauge invariance only when the quiver has a closed loop , is given by : @xmath73 ( where we take coefficients @xmath74 to be arbitrary ) and contributes the following piece to the lagrangian : @xmath75 in this note we only consider cubic superpotentials and ignore the higher order terms .
this is consistent so long as the @xmath76 are small , which in turn can be assured by taking the @xmath77 sufficiently small @xcite .
the theory contains a manifest @xmath78 global r - symmetry . in the absence of a superpotential ,
the lagrangian is diagonal in the arrow ( greek ) indices and thus the theory also exhibits a @xmath79 global symmetry under which the @xmath76 transform as @xmath80 vectors .
the superpotential explicitly breaks this symmetry down to the @xmath81 gauge symmetry .
the theory can be obtained by dimensionally reducing an @xmath82 gauge theory in four - dimensions to the @xmath83-dimensional worldline theory .
it can also be viewed as the dimensional reduction of the @xmath31 two - dimensional @xmath84-models studied extensively , for example , in @xcite . in order for the theory to have supersymmetric vacua we also demand that the fayet - iliopoulos constants sum to zero : @xmath85 . in units where @xmath86 is dimensionless , and where we choose dimensions for which @xmath87 = 1 $ ] , the dimensions of energy are automatically set to @xmath88=-1 $ ] .
we also find the following dimensional assignments : @xmath89 = 1/2 $ ] , @xmath90=-2 $ ] , @xmath91=-1 $ ] , @xmath92=3 $ ] , @xmath93=-3/2 $ ] , @xmath94=0 $ ] , @xmath95 = -3/2 $ ] , @xmath96 = -1/2 $ ] and @xmath97 = 1 $ ] .
the mass squared of the @xmath98 fields upon integrating out the auxiliary @xmath99 fields is given by @xmath100 . for physics
whose energies obey @xmath101 we can integrate out the massive @xmath76 fields and study the effective action on the coulomb branch .
finally , notice also that the coupling @xmath74 has positive units of energy and is thus strong in the infrared limit since the natural dimensionless quantity is @xmath102 .
the contribution from @xmath98 loops grows as @xmath103 and thus the effective coupling constant at low energies is given by @xmath104 . in the large @xmath105 limit ,
the wrapped branes backreact and the appropriate description of the system is given by four - dimensional @xmath0 supergravity @xcite .
an immediate question about the above model regards the structure of the ground states @xmath106 $ ] of the theory , satisfying : @xmath107 = 0 $ ] .
though an explicit expression for the full @xmath108 $ ] remains unknown , the degeneracy of ground states has been extensively studied @xcite . in particular ,
the degeneracy of ground states localized near @xmath109 , i.e. the ground states of the higgs branch of the theory , were shown to grow exponentially in @xmath103 when the theory contains a superpotential , the quiver admits closed loops ( e.g. @xmath110 and @xmath56 ) and the @xmath103 obey the triangle inequality ( i.e. @xmath111 and cyclic permutations thereof ) .
this growth is related to the exponential explosion in the euler characteristic of the complete intersection manifold @xmath112 @xcite given by imposing the constraints from the @xmath113-term ( @xmath114 ) onto the @xmath99-term constraints ( @xmath115 ) .. then the complete intersection manifold is given by setting @xmath116 , imposing the @xmath117 @xmath113-term constraints : @xmath118 , inside a @xmath119 space coming from the @xmath99-term constraints : @xmath120 and @xmath121 .
the space is a product of @xmath122 s since we have to identify the overall phase of the @xmath98 due to the @xmath27 gauge connection .
when @xmath123 , which for large @xmath124 amounts to the @xmath103 satisfying the triangle inequality , the number of constraints become less or equal to the dimension of @xmath119 , allowing for more complicated topologies for @xmath112 .
] since @xmath103 goes as the charge squared of the associated @xmath27 gauge symmetry , the number of ground states scales in the same way as the bekenstein - hawking entropy of the associated black hole solutions in the large @xmath105 limit .
though a complete match between the ground states of a single abelian quiver model and the entropy of a bps black hole in @xmath31 supergravity is not known , the vast number of microstates makes these systems potentially useful candidate toy models to study features of extremal or near extremal black holes .
other pieces of @xmath125 localized near @xmath126 , i.e. the ( quantum ) coulomb branch of the theory , have also been studied @xcite . unlike the higgs branch quiver with a closed loop , superpotential and an exponential growth in its number of ground states , it was found that the number of coulomb branch ground states grows only polynomially in the @xmath103 . interpreting the @xmath127 as the relative positions of wrapped branes ,
these ground states can be viewed as describing various multiparticle configurations , as we will soon proceed to describe in further detail . to each ground state in the quantum coulomb branch there exists a corresponding ground state in the higgs branch , but the converse is not true . another way to view
this statement is that whenever a given @xmath125 has non - trivial structure in the @xmath128 directions and peaks sharply about @xmath126 , it will also have a non - trivial structure in the @xmath129 directions and peak sharply about @xmath109 but not vice versa .
for large enough @xmath130 we can integrate out the massive @xmath129 s ( in their ground state ) from the full quiver theory ( [ quiverqm ] ) . this can be done exactly given that the @xmath129 appear quadratically in ( [ quiverqm ] ) whenever the superpotential vanishes .
one finds the bosonic quantum effective coulomb branch lagrangian ( up to quadratic order in @xmath131 and @xmath132 ) : @xmath133 the terms linear in @xmath131 and @xmath132 follow from a non - renormalization theorem @xcite , whereas the quadratic piece in @xmath131 is derived in appendix [ cmetric ] .
recall that the system is only a function of @xmath134 and @xmath135 since @xmath64 .
the three - vector @xmath136 is the vector potential for a magnetic monopole : @xmath137 and @xmath138 is the two - by - two metric on configuration space : @xmath139 = \left ( \begin{array}{ccc } \mu^{1 } + \mu^{3 } + \frac{1}{4 } \frac{|\kappa^{1}|}{|\bold{q}^1|^3 } + \frac{1}{4 } \frac{|\kappa^{3}|}{|\bold{q}^1 + \bold{q}^2
|^3 } & \mu^{3 } + \frac{1}{4 } \frac{|\kappa^{3}|}{|\bold{q}^1 + \bold{q}^2 |^3 } \\ \mu^{3 } + \frac{1}{4 } \frac{|\kappa^{3}|}{|\bold{q}^1 + \bold{q}^2 |^3 } & \mu^{2 } + \mu^{3 } + \frac{1}{4}\frac{|\kappa^{2}|}{|\bold{q}^2|^3 } + \frac{1}{4 } \frac{|\kappa^{3}|}{|\bold{q}^1 + \bold{q}^2 |^3 } \end{array } \right ) ~.\ ] ] upon integrating out the auxiliary @xmath132-fields , we obtain a multi - particle quantum mechanics with ( bosonic ) lagrangian : @xmath140 that the quantum effective coulomb branch theory has a non - trivial potential @xmath141 should be contrasted with other supersymmetric cases such as interacting d0-branes or the d0-d4 system @xcite where the potential vanishes and the non - trivial structure of the coulomb branch comes from the moduli space metric .
the potential @xmath141 is also somewhat involved and is given in appendix [ potential ] . the supersymmetric configurations of the coulomb branch consist of time independent solutions which solve the equations @xmath142 .
for ( [ supergoop ] ) , this amounts to : @xmath143 in appendices [ sucos ] and [ finited ] we review that these supersymmetric configurations are robust against corrections of the coulomb branch theory from the superpotential and from integrating out higher orders in the auxiliary @xmath99 fields .
there are _ bound state solutions _
@xcite of ( [ bound ] ) which are triatomic ( or more generally @xmath144-atomic if dealing with @xmath144 wrapped branes ) molecular like configurations . of the original nine degrees of freedom , three can be removed by fixing the center of mass . then the bound state condition ( [ bound ] )
fixes another two - degrees of freedom .
thus , bound state solutions have a four - dimensional classical moduli space .
due to the velocity dependent terms in the lagrangian , the flat directions in the moduli space are dynamically inaccessible at low energies the particles resemble electrons in a magnetic field .
several dynamical features of the three particles were studied in @xcite .
there are also _ scaling solutions _ @xcite of ( [ bound ] ) which are continuously connected to the origin @xmath145 .
they occur whenever the @xmath103 form a closed loop in the quiver diagram ( e.g. @xmath110 and @xmath146 ) and obey the same triangle inequality ( @xmath111 and cyclic permutations thereof ) that the @xmath147 are subjected to .
these solutions can be expressed as a series : @xmath148 the coefficient @xmath149 , while the remaining @xmath150 can be obtained by systematically solving ( [ bound ] ) in a small @xmath151 expansion , and will hence depend on @xmath152 . the moduli space of the scaling solutions is given by the three rotations as well as the scaling direction parameterized by @xmath151 . though the angular directions in the moduli space are dynamically trapped due to velocity dependent forces
, the scaling direction is not and constitutes a flat direction even dynamically . requiring that the series expansion converges , i.e. @xmath153 , leads to the condition : @xmath154 because of this condition on the @xmath151 s
, one should be cautious when dealing with such scaling solutions .
they occur in the near coincident limit of the branes where the bifundamentals that we have integrated out become light . in order for the mass of the bifundamentals , @xmath155 , to remain large
we require : @xmath156 taking @xmath157 and @xmath158 the inequalities ( [ convergence ] ) and ( [ notach ] ) can be satisfied in the limit @xmath159 , with @xmath160 , @xmath161 , @xmath162 and @xmath152 fixed and furthermore @xmath163 .
notice that ( [ convergence ] ) implies that the distances between particles in a scaling regime @xmath164 is much less than the typical inter - particle distance of a bound state @xmath165 . to isolate the physics of the coulomb branch in the scaling regime we take an infrared limit of the lagrangian ( [ supergoop ] ) , pushing the @xmath127 near the origin and dilating the clock @xmath166 .
in particular , we would like the @xmath167 part of the metric in configuration space to dominate over the @xmath168 piece leading to : @xmath169 ) and ( [ notach ] ) . again taking @xmath157 , @xmath170 and in addition @xmath171 with fixed @xmath160 , @xmath161 , @xmath162 and @xmath152
, we can also satisfy ( [ scbounds ] ) in the limit @xmath159 , so long as we also ensure @xmath172 .
, @xmath173 . in this case
the metric on configuration space remains flat while the potential scales like @xmath174 .
] the rescaling of @xmath166 is required to maintain a finite action in the scaling limit . in type ii string compactifications
@xmath175 , where @xmath176 is the four - dimensional planck length and @xmath177 is the volume of the calabi - yau in string units @xcite , therefore the @xmath159 limit corresponds to a parametrically small string coupling .
furthermore the scaling throat deepens as we increase the mass of the wrapped branes . the rescaling above amounts simply to setting the @xmath152 and @xmath178 to zero in ( [ supergoop ] ) and replacing @xmath179 and @xmath166 with @xmath161 and @xmath180 .
we call the remaining lagrangian with vanishing @xmath152 and @xmath178 the _ scaling theory_. notice from equation ( [ scalpot ] ) that the potential @xmath141 in this limit becomes a homogeneous function of order one , i.e. @xmath181 and the linear in velocity term is retained
. one can infer from the supersymmetry variations @xcite that supersymmetric configurations have vanishing @xmath132 .
furthermore , as we have already noted , we have performed an expansion in small @xmath132 fields prior to integrating them out ( see appendix [ finited ] for more details ) in order to obtain the coulomb branch .
thus , in order for a scaling theory to exist and be consistent with a small @xmath132 expansion , it must be the case that zero energy scaling configurations exist .
had we considered a two - particle theory , where no such scaling solutions can exist , taking a small @xmath182 limit would be inconsistent with the small @xmath99 expansion .
that is because the dimensionless small quantity in our perturbation series is actually @xmath183 and the supersymmetric configuration @xmath184 occurs at @xmath185 .
expanding the non - linear @xmath99 equation ( [ twonoded ] ) ( see appendix [ finited ] ) in powers of @xmath186 , while imposing @xmath187 , one finds the following consistency condition : @xmath188 indeed , the first term on the right hand side is negligible by construction , since we took @xmath187 to keep the strings massive .
smallness of the second term in the equation would require @xmath189 , in contradiction with the condition ( [ scbounds ] ) required to isolate the scaling theory .
we now consider the three - node case for @xmath67 that admit a closed loop in the quiver . for small @xmath190 ( taking all masses the same and @xmath191 and again imposing @xmath192 ) the equation of motion of the auxiliary @xmath193-field ( in the form of ( [ threenoded ] ) )
is given by : @xmath194 where @xmath195 measures how close the configuration is to the scaling solution .
for sufficiently small @xmath196 we can consistently satisfy ( [ delta ] ) in addition to imposing the scaling inequalities ( [ scbounds ] ) . in this sense
the scaling theory is in fact a theory of the deep infrared configurations residing parametrically near the zero energy scaling solutions .
this is consistent with the dilation of time required to obtain the scaling theory .
in this section we uncover that the bosonic scaling theory action , i.e. ( [ supergoop2 ] ) with the @xmath152 and @xmath178 set to zero , has an @xmath2 symmetry .
this is the symmetry group of conformal quantum mechanics @xcite .
the group @xmath2 is generated by a hamiltonian @xmath197 , a dilatation operator @xmath99 and a special conformal transformation @xmath198 , with the following lie algebra : @xmath199 = - 2 i h~ , \quad \quad [ h , k ] = - i d~ , \quad \quad [ k , d ] = 2i k~.\ ] ] as we have already mentioned , the full @xmath2 symmetry is not guaranteed by the existence of time translations and dilatations alone @xcite .
this is suggested by the fact that the @xmath197 and @xmath99 operators form a closed subalgebra of the full @xmath2 .
the presence of a full @xmath2 is actually quite remarkable , particularly given the specific form of the scaling theory lagrangian which has velocity dependent forces and a non - trivial potential .
as discussed in the previous section , it is not true that , for any finite @xmath103 , @xmath178 and @xmath152 , the coulomb branch can be described precisely by the scaling theory action .
there will always be small corrections that break its manifest scaling symmetry : @xmath200 , and @xmath201 .
this is comforting , given that the full quiver theory has a finite number of ground states yet a conformal quantum mechanics has a diverging number of arbitrarily low - energy states , with a density of states that behaves as @xmath202 .
the corrections serve as a cutoff for the infrared divergence in the number of states , such that the coulomb branch can fit consistently inside the full quiver theory ( related discussions can be found in @xcite ) .
the conditions under which an action will be @xmath2 invariant ( up to possible surface terms ) have been studied extensively in @xcite . showing that a general theory with bosonic lagrangian describing @xmath144 degrees of freedom : @xmath203 has an @xmath2 symmetry is equivalent to finding a solution to the following equations @xcite : @xmath204 } \label{maxwell}~. \end{aligned}\ ] ] equations ( [ homo1 ] ) and ( [ homo2 ] ) ensure the existence of a dilatation symmetry . in particular , equation ( [ homo1 ] )
implies that the metric on configuration space allows for a conformal killing vector field ( also referred to as a homothetic vector field ) .
equations ( [ closed ] ) and ( [ maxwell ] ) ensure the that the action remains invariant under special conformal transformations , where @xmath205 is an arbitrary function of the @xmath206 .
indices are raised and lowered with the metric @xmath138 .
interestingly , equation ( [ closed ] ) imposes that the conformal killing form @xmath207 of the metric be exact , which is generically not the case .
hence the existence of a dilatation symmetry does not necessarily imply the symmetry of the full conformal group .
once a solution to ( [ homo1]-[maxwell ] ) is found , the three conserved quantities are then given by @xcite:@xmath208 where @xmath209 and @xmath210 .
the charge @xmath211 is the hamiltonian , whereas @xmath212 and @xmath213 are related to dilatations and special conformal transformations , respectively .
these three charges generate the @xmath2 algebra ( [ sl2ralg ] ) ( up to factors of @xmath46 ) under the poisson bracket . in what follows , we find @xmath214 and @xmath205 for the scaling theory described in section [ sec : scaltheor ] . as a warm up
, we can study a simple model consisting of a two node quiver with an equal number of arrows going to and from each node , with a total number @xmath215 arrows altogether .
this is the theory describing , for example , the low energy dynamics of a wrapped d4-d0 brane system ( see @xcite ) .
the bosonic lagrangian for the relative position @xmath216 on the coulomb branch , in the scaling limit , is given by : @xmath217 the above lagrangian is also the non - relativistic limit of one describing a bps particle in an ads@xmath28 background .
the wordline theory has three degrees of freedom and a diagonal metric on configuration space : @xmath218 .
in addition to the hamiltonian @xmath197 , the above theory has a dilatation operator @xmath99 and special conformal generator @xmath198 given by : @xmath219 such that the @xmath2 algebra is satisfied . it is thus a simple example of a conformal quantum mechanics . for our particular problem ,
it is useful to note that the index structure of the relative coordinates , i.e. @xmath220 , is trivially tensored with the spatial index implicit in bold vector symbols ( e.g. @xmath216 ) . introducing @xmath221 , with @xmath222 ,
our lagrangian ( [ supergoop2 ] ) ( with @xmath178 and @xmath152 set to zero , and @xmath67 that admit a closed loop ) takes the form : @xmath223 where @xmath224 and the six - dimensional vector potential can be extracted from ( [ supergoop2 ] ) .
an expression for the vector potential is simple to write down and is given by : @xmath225 where @xmath226 is the vector potential for a magnetic monopole and is given in ( [ mon ] ) .
it is straightforward to check , using mathematica for example , that the conditions ( [ homo1]-[maxwell ] ) are indeed satisfied .
we find @xmath227 and @xmath228 .
the explicit generators of the @xmath2 are given by : @xmath229 where we have used that @xmath230 .
the generator @xmath198 simplifies to : @xmath231 we thus conclude that the bosonic scaling theory is an interacting multi - particle @xmath2 invariant conformal quantum mechanics with velocity dependent forces .
the theory admits a further @xmath78 symmetry acting on the spatial three - vectors .
though we do not prove it here , it is natural to conjecture that the @xmath144-particle scaling theory will be an @xmath144-particle conformal quantum mechanics .
indeed , for any closed loop of @xmath232-particles we expect that in the corresponding throat there will be a conformal quantum mechanics with @xmath99 and @xmath198 as in ( [ generators ] ) except @xmath233 runs over all the @xmath234 bosonic degrees of freedom .
the one loop metric on configuration space in the coulomb branch , i.e. the coefficient of @xmath235 , will be equivalent to that of @xmath236 ( which is far easier to compute ) upon integrating out the massive strings for general quivers .
interestingly , when there are more than three nodes , there may be several distinct closed loops allowing for their own scaling throat and several distinct scaling throats may coexist , within its residing a decoupled conformal quantum mechanics theory . it is worth remarking that the coulomb branch is in fact a supersymmetric quantum mechanics with four supercharges @xmath237 and an @xmath78 global r - symmetry group
this follows from effective field theory .
we are integrating out the heavy chiral @xmath238 multiplet in its supersymmetric ground state .
thus , the low energy effective theory of the vector multiplet will be endowed with the four supercharges of the parent quiver theory .
of course , there could be anomalies that arise in the process .
for instance , the discrete time reversal symmetry of the full quiver theory is violated in the coulomb branch by the linear in velocity terms .
this is an anomaly which does not spoil the supersymmetry of the coulomb branch , and occurs due to a zero mode in the functional determinant of the @xmath239 fermions . as shown in @xcite for the two node case , one can compute the supercharges of the low energy effective theory in a systematic fashion using perturbation theory .
we have shown above that the bosonic lagrangian of this theory exhibits an @xmath2 symmetry . in order to establish that this extends to a superconformal quantum mechanics we must establish the existence of four supersymmetric special conformal generators @xmath240 .
these will be given by the commutator @xmath241 $ ] .
as discussed in @xcite , one must also ensure that the commutator @xmath242 = -i \mathcal{q}_s^i$ ] is satisfied in order to have a superconformal system .
this is guaranteed due to the manifest scale invariance of the supersymmetric action : @xmath200 , @xmath243 with @xmath201 . ,
which will be of the form @xmath244 in the scaling region , the linear piece is fixed by supersymmetry @xcite to be ( [ linear1 ] ) , which is enough to read off the scaling dimension of @xmath245 . ] hence the scaling theory is a superconformal multiparticle theory . as discussed in the introduction ,
the emergence of a full @xmath2 in the deep scaling regime is reminiscent of the emergence of a full @xmath2 in the deep ads@xmath1 throat in gravity , which could end at a horizon or cap off at the locations of entropyless d - particles .
the symmetry group manifests itself in the dynamics of wrapped branes moving in a ` geometry ' resulting from integrating out the interconnecting strings and extends upon similar observations made for other multiparticle systems in @xcite .
however , we must point out that the conformal quantum mechanics obtainde here comes directly from the low energy dynamics of strings interacting with wrapped branes , i.e. the quiver theory , rather than from a gravity calculation of the moduli space of a multi - black hole system
. it would in fact be interesting to repeat such gravitational ferrel - eardley type calculations @xcite for the low energy velocity expansion of the corresponding one - half bps scaling solutions @xcite in @xmath31 supergravity .
this might give an operational meaning to the ads@xmath1/cqm correspondence @xcite ( particularly in the large @xmath246 limit and within the ads@xmath1 scaling region ) .
recall that in the classical supergravity limit , the gravitational backreaction of a brane becomes parametrically small given that its mass goes like @xmath247 .
throat in the geometry and that the microscopic quiver model is vector like rather than matrix like . ]
furthermore , as emphasized in the introduction , our system is a quiver quantum mechanics that does not necessarily reside inside a larger two - dimensional conformal field theory , and thus the @xmath2 may be of the ` isolated ' type . given the existence of an @xmath2 invariant scaling theory , one natural question is whether a quantum state will stay localized within the scaling regime or if its wavefunction will spread away from the scaling regime , or even fall back into the higgs branch ( where @xmath248 ) .
one particular direction in which a wavefunction might easily spread is the scaling direction where the potential is identically zero and where there are ( classically ) no velocity dependent magnetic forces .
studying the quantum mechanics problem of the full three - particle scaling theory is difficult .
instead , we will look at the wavefunctions of the simpler wrapped d4-d0 model discussed in section [ d0d4 ] .
the zero angular momentum piece of the schrdinger equation in spherical coordinates ( i.e. @xmath249 , @xmath250 and @xmath251 , @xmath252 ) is : @xmath253 which can easily be solved : @xmath254 . with @xmath255
spherical harmonics and @xmath256 can be written in terms of bessel functions . here
we only look at @xmath257 modes , @xmath258 . ]
these energy eigenstates are non - normalizable at small @xmath259 with respect to the covariant inner product : @xmath260 non - normalizability of energy eigenstates is common for scale invariant theories @xcite , the non - normalizability means that energy eigenstates leak into the higgs branch . as in @xcite
, we can consider instead eigenstates @xmath261 of the @xmath262 operator with eigenvalue @xmath263 , where @xmath264 is a constant with appropriate dimensions .
if we define a variable @xmath265 , then the zero - angular momentum wavefunctions @xmath266 satisfy @xmath267 the normalizable wavefunctions are given by confluent hypergeometric functions : @xmath268 where @xmath269 is a finite normalization factor that depends on @xmath124 and @xmath264 .
whenever @xmath270 , the expression for @xmath271 simplifies significantly .
we can also obtain asymptotic expressions near @xmath272 and @xmath273 : @xmath274}-\frac{2 \ , \sqrt{\pi } } { { \gamma}\left[-n/2\right]}+ \mathcal{o}\left(x^{-1/2}\right)~ , \quad \text{for } x^{-1 } \sim 0~.\end{aligned}\ ] ] the dependence on @xmath124 and @xmath264 drops out of all expectation values of operators as a function of @xmath275 labeled @xmath276 , hence we can instead normalize the wavefunctions @xmath277 and reinstate the dependence on @xmath124 and @xmath264 in expectation values of operators by looking at the dependence of @xmath276 on the variable @xmath275 and how it transforms when we take @xmath278 .
examples of @xmath279 are displayed in figure [ wavefunctions ] .
eigenfunctions . left : plot of @xmath279 for @xmath280 in unit increments .
right : plot of @xmath279 for @xmath281 in unit increments.,title="fig:",width=288 ] eigenfunctions . left : plot of @xmath279 for @xmath280 in unit increments .
right : plot of @xmath279 for @xmath281 in unit increments.,title="fig:",width=288 ] to understand whether these states leak back into the higgs branch , we consider for example @xmath282 .
this is only finite if @xmath283 is an odd positive integer for which @xmath284 ( or equivalently @xmath285 ) , otherwise @xmath282 is infinite .
this can be seen from the fact that for positive and odd @xmath283 , @xmath286 $ ] diverges and the @xmath287 term in the large @xmath275 expansion ( [ largex ] ) dissappears .
however , for any @xmath288 there exists a number @xmath289 such that @xmath290 is finite .
similarly , when @xmath283 is an odd positive integer , there exists a number @xmath291 such that @xmath290 is infinite .
thus a large class of @xmath262 eigenstates do not leak into the higgs branch for large @xmath124 .
finding the wavefunctions in the three - node coulomb branch is clearly more complicated .
there are now six - degrees of freedom and a non - trivial potential @xmath141 .
however , at the classical level , the scaling direction remains flat and motion along it is uninhibited , neither by the potential nor by any velocity dependent magnetic forces .
so if the wavefunction has any chance of spreading , it will do so along the scaling direction .
our previous analysis of the two particle wavefunctions , which deals essentially with the scaling direction , suggests that at least some @xmath292 eigenstates do not leak into the higgs branch .
we end this section with an important question .
the @xmath2 symmetry we have been discussing so far resides in the coulomb branch of the full quiver theory .
in particular it resides in the scaling regime of the coulomb branch which is connected to the higgs branch at its tip , i.e. where all the @xmath179 become vanishingly small .
how does the @xmath2 structure act on ( if at all ) the higgs branch degrees of freedom ? perhaps a way to answer this question is via the higgs - coulomb map of @xcite ( further developed in the context of quiver theories in @xcite ) .
indeed , an @xmath2 symmetry acting on the states residing in the higgs branch would resonate closely with the appearance of an @xmath2 isometry of the ads@xmath1 in the near horizon region of the extremal black hole .
we have seen that the low energy excitations of the scaling regime in the coulomb branch are described by a multiparticle @xmath2 invariant quantum mechanics .
it was noted that this is reminiscent of the @xmath2 that appears in the near horizon geometry of an extremal black hole or the ads@xmath1 geometry outside a collection of extremal black holes that reside within a scaling throat .
if we heat up such a collection of black holes ( for example by making the centers slightly non - extremal ) they will fall onto each other due to gravity s victory over electric repulsion . naturally , then
, we might ask what happens to our scaling theories and more generally the coulomb branch configurations upon turning on a temperature ?
this amounts to integrating out the massive strings , i.e. the chiral multiplet , in a thermal state as opposed to their vacuum state . in this section
we assess the presence of bound states and scaling solutions as we vary the temperature . roughly speaking ,
this is the weak coupling version of the calculations in @xcite where the multicentered solutions were studied at finite temperature .. one node is ` pulled ' away from the remaining three , such that one can integrate out all connecting heavy strings in some thermal state .
the remaining three nodes , which we choose to contain a closed loop ( and hence have exponentially many ground states ) , are in the higgs branch which is also considered to be at some finite temperature . one could then study the coulomb branch theory for the position degree of freedom of the far away wrapped brane as a function of the temperature .
] we find that the bound ( but non - scaling ) configurations persist at finite temperature until a critical temperature after which they either classically roll toward the origin or become metastable .
this amounts to the ` melting ' of the bound state . scaling solutions on the other hand
do not persist at finite temperature and instead the potential develops a minimum at the origin , even at small temperatures .
as was alluded to , what we find is somewhat analogous to heating up a collection of extremally charged black holes , known as the majumdar - papapetrou geometries @xcite , by making the masses of the centers slightly larger than their charges .
doing so will cause the black holes to collapse into a single center configuration upon the slightest deviation from extremality .
consider the low energy dynamics of two wrapped branes , which is given by a two node quiver .
there can be no gauge - invariant superpotential in this case since the quiver admits no closed loops .
the relative lagrangian is : @xmath293 where @xmath294 .
the matter content is given by a chiral multiplet @xmath295 and a vector multiplet @xmath296 ( @xmath297 is a one - dimenional @xmath27 connection ) .
the index @xmath298 ( where @xmath215 ) is summed over .
if we keep @xmath299 constant and set to zero the fermionic superpartners @xmath151 , we can integrate out the @xmath300 and @xmath301 fields to get the effective bosonic action on the position degrees of freedom : @xmath302 if we wish to study the system at finite temperature @xmath303 , we must wick rotate to euclidean time @xmath304 and compactify @xmath305 .
notice we can not set the zero mode of @xmath297 to zero since one can have non - trivial holonomy around the thermal circle .
the @xmath300 fields have periodic boundary conditions and the @xmath301 fields have anti - periodic boundary conditions around the thermal circle .
the operator @xmath306 has as eigenvalues the matsubara frequencies @xmath307 for the bosonic case and @xmath308 for the fermionic case , where @xmath309 .
hence , it is our task to evaluate : @xmath310 where we have defined @xmath311 .
notice that @xmath312 depends on the connection @xmath297 which we eventually need to path - integrate over . upon evaluating the determinant ( see appendix [ thermaldet ] for details ) , one finds the effective thermal potential is given by : @xmath313 which has the correct low temperature limit @xmath314 , in accordance with the results of @xcite .
our task is to now perform the path integral of @xmath315 over @xmath99 and @xmath264 . in doing so
, we must be careful to ensure that @xmath316 such that integrating out the string is justifiable .
differentiating ( [ thermalpot ] ) with respect to @xmath264 we find that @xmath317 has a minimum at @xmath318 for all @xmath21 and @xmath319 .
solving the saddle point equations for @xmath99 must be done numerically .
the potential @xmath320 has two scaling symmetries associated to it .
configurations related by : @xmath321 have respective potentials related by : @xmath322 .
the qualitative features of the potential thus only depend on scale invariant quantities : @xmath323 and the scaling - invariant potential is given by @xmath324 .
note that our finite temperature analysis breaks down for @xmath325 in string units , where the effects of massive string modes start to kick in .
however , we may still consider large @xmath326 as long as we restrict ourselves to parameter regions where our approximation remains valid , e.g. @xmath327 and @xmath328 .
we use the scaling symmetries to fix @xmath329 and study the thermal phases of the theory as a function of @xmath326 and @xmath330 , noting that we can always extract physical quantities by properly rescaling @xmath331 and @xmath332 . having derived the thermal effective potential we can now describe the different thermal configurations for the two - node quiver .
the physically distinguishable phases have potentials @xmath333 with the following distinct properties : 1 .
a single stable minimum away from @xmath334 .[singstab ] 2 .
two non - degenerate minima away from @xmath334 .[twnondeg ] 3 .
two degenerate minima away from @xmath334 .[twodeg ] 4 . two minima with one at @xmath335 where :
1 . the minimum at @xmath335 is the global minimum,[q0globmin ] 2 .
the minimum at @xmath336 is the global minimum.[q0nonglobmin ] 5 . a single minimum at @xmath334 .[hottest ] examples of each of these thermal configurations
are displayed in figures [ phases ] and [ phases3 ] .
depending on the location in the @xmath337-plane , the system will be driven ( either through thermal activation or quantum tunneling ) to the most stable configuration .
if the system is in a metastable configuration , such a process can take exponentially long . at high enough temperatures ,
the system will eventually fall back to @xmath334 where the higgs and the coulomb branch meet , mimicking gravitational collapse or the melting of the molecule . and @xmath338 ) .
as the temperature is increased the system explores various thermal configurations of stable and metastable minima . from top
left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] and @xmath338 ) . as the temperature
is increased the system explores various thermal configurations of stable and metastable minima . from top left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] and @xmath338 ) . as the temperature
is increased the system explores various thermal configurations of stable and metastable minima . from top left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] and @xmath338 ) .
as the temperature is increased the system explores various thermal configurations of stable and metastable minima . from top
left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] and @xmath338 ) . as the temperature
is increased the system explores various thermal configurations of stable and metastable minima . from top left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] and @xmath338 ) . as the temperature
is increased the system explores various thermal configurations of stable and metastable minima . from top left to bottom right the system is of type [ singstab ] @xmath339 [ twnondeg ] @xmath339 [ twodeg ] @xmath339 [ twnondeg ] @xmath339 [ q0globmin ] @xmath339 [ hottest].,title="fig:",width=192 ] note that for bound states in @xmath31 supergravity , the effective potential of a small probe around a large hot black hole , as studied in @xcite , never exhibited two minima away from @xmath334 ( that is we never noticed potentials of types [ twnondeg]-[q0nonglobmin ] ) .
since supergravity is a good effective description at large @xmath340 , it would be interesting to see if this behavior matches qualitatively in the coulomb branch as @xmath330 is increased .
an example of this is shown in figure [ increasekappa ] . [ cols="^,^ " , ] a way to think about the high temperature melting transition is to integrate out the auxiliary @xmath99-fields first , which induces a quartic interaction @xmath341 for the bifundamentals . since this interaction is relevant , it will play a minor effect at sufficiently large temperatures , namely for @xmath342 .
thus , at high temperatures we are dealing with a collection of @xmath124 non - interacting complex bosonic and fermionic degrees of freedom with square masses @xmath343 and @xmath344 respectively .
in addition , due to the @xmath27 connection we must only consider the gauge invariant states : the spectrum is constrained to those states annihilated by the @xmath27 charge operator .
a particularly convenient gauge is the temporal gauge , @xmath345 . at some fixed high temperature ,
the number of gauge invariant modes that are nt boltzmann suppressed increases with decreasing mass and thus the dominant contribution to the free energy will come from for the lightest possible mass , i.e. @xmath346 .
this can be viewed as an entropic effect @xcite .
a similar analysis for three nodes gives rise to the following effective thermal potential : @xmath347 our main interest to understand the behavior of the scaling potential at finite temperature , and in particular the flat scaling direction .
thermal effects will kick in when @xmath348 . since zero temperature scaling solutions occur for arbitrarily small @xmath147 , it is sufficient to perform a small temperature expansion .
the dimensionless quantities capturing the thermal transition are @xmath349 . as in the two - node case
, one must integrate out the @xmath27 connections @xmath350 and the @xmath132 . for the case
where the @xmath351 we can identify a critical point along the scaling direction @xmath352 : @xmath353 , @xmath354 , @xmath355 and @xmath356 . on this saddle
the potential becomes : @xmath357~,\ ] ] which is negative definite .
of course there could be more dominant saddles that make the potential even lower than the above .
so more generally , this procedure must be done numerically , even in the small temperature expansion , as the equations governing the connections are intractable analytically .
it amounts to numerically minimizing a function of four variables , namely @xmath358 , @xmath359 , @xmath360 and @xmath193 ( recall that @xmath361 and @xmath65 ) . and @xmath338 ) .
as @xmath330 is increased we note that the first minimum disappears.,title="fig:",width=192 ] and @xmath338 ) . as @xmath330
is increased we note that the first minimum disappears.,title="fig:",width=192 ] and @xmath338 ) . as @xmath330 is increased we note that the first minimum disappears.,title="fig:",width=192 ] and @xmath338 ) . as @xmath330 is increased we note that the first minimum disappears.,title="fig:",width=192 ] since we are interested in the scaling branch we can set @xmath362 in our analysis . is for computational simplicity and numerical clarity .
we could have started a coulomb branch with non - zero @xmath152 s which has scaling solutions for @xmath363 in the limit @xmath364 and found similar results . ] the resulting potential ( with @xmath362 ) then exhibits the following scaling relation : @xmath365 such that @xmath366 . from the scaling symmetries we observe that the numerical value of the temperature is of little meaning , all that matters for the thermal phase structure is whether or not it vanishes .
this is to be expected since we are dealing with a scale invariant system . in figure [ voflambdatfin ]
we display the potential at @xmath367 ( left ) and @xmath368 ( right ) in the scaling direction @xmath369 for @xmath370 . at @xmath367
, we naturally find that it is vanishing for all @xmath151 . at @xmath371
we see that for radial values larger than the temperature , the scaling direction is still effectively flat , but for radii of order the temperature the thermal potential quickly falls .
the numerics are in good agreement with the analytic expression ( [ analytic ] ) .
hence , at finite temperature the system falls back into its higgs branch .
it would be interesting to explore this phenomenon in more generality for a larger number of nodes .
for @xmath372 at @xmath367 ( left ) and @xmath371 ( right).,title="fig:",width=288 ] for @xmath372 at @xmath367 ( left ) and @xmath371 ( right).,title="fig:",width=288 ]
we have observed the emergence of a full @xmath2 symmetry from a quiver quantum mechanics model which itself is not a conformal quantum mechanics .
the @xmath2 manifested itself in the effective theory of the position degrees of freedom of the d - particles , once the heavy strings were integrated out .
supersymmetry played an important role in the previous discussion given that the form of the quiver lagrangian ( [ quiverqm ] ) is heavily constrained by it . but
as was mentioned in the introduction , the @xmath2 symmetry appears for geometries that need not be supersymmetric . in this final speculative section ,
we consider the idea that randomness ( in the wignerian sense ) is behind the @xmath2 and observe that dilatations imply special conformal transformations for one dimensional systems whose dilatation symmetry is geometrized into an additional radial direction .
finally , we discuss some possible extensions
. it would be interesting to understand whether the emergence of such an @xmath2 from systems with large numbers of degrees of freedom , such as matrix quantum mechanics , could be more general .
for instance , if there are a large number of almost degenerate vacua in the putative microscopic dual of ads@xmath1 , one might imagine an approximate scale invariance emerging at large @xmath144 due to the formation of almost continuous bands of low energy eigenvalues .
more generally , if the hamiltonian is sufficiently complicated due to the multitude of internal cycles being wrapped by the branes , one might imagine drawing it from a random ensemble @xmath373 of @xmath374 hermitean matrices .
we can then formulate the following problem .
draw an @xmath374 hermitean matrix @xmath375 and assess ( under some as of yet unspecified measure , perhaps the frobenius norm is a possibility ) how well the matrix equations ( [ sl2ralg ] ) can be satisfied , given arbitrary hermitean matrices @xmath99 and @xmath198 .
in particular , are the equations better satisfied as we increase @xmath144 .
has no finite dimensional unitary representations , the equations can only be satisfied exactly for @xmath376 . ]
the emergence of an @xmath2 from a random ensemble of hamiltonians , if true , would be similar in spirit to the emergence of the wigner distribution of eigenvalue spacings from random matrices that is almost universal to quantum systems that become chaotic in the classical limit .
interestingly , an ensemble of random hamiltonians with a scale invariant distribution of eigenvalues was found in @xcite .
we hope to return to this question in the future . if one assumes that there exists a gravitational dual to the theory , and in addition that a scaling symmetry @xmath377 exists along with a radial transformation @xmath378 one obtains the two - dimensional metric ( see for example @xcite ) : @xmath379 which is nothing more than ads@xmath1 ( albeit in unusual coordinates ) .
the isometry group of ads@xmath1 is @xmath2 and thus , holographically a dilatation symmetry seems to imply the existence a full @xmath2 symmetry .
this is not true in higher dimensions unless one also assumes lorentz invariance of the boundary metric .
another feature of conformal quantum mechanics that contrasts with higher dimensional conformal field theory is that there is no normalizable @xmath2 invariant ground state .
on the other hand , @xmath6-dimensional conformal field theories have an @xmath380 invariant ground state wavefunctional .
we view this as a hint that whatever the holographic description of ads@xmath1 is , it is not necessarily a conformal quantum mechanics off the bat .
indeed , the quiver quantum mechanics whose ground state degeneracies count a large fraction of the microstates of a black hole with an ads@xmath1 near horizon are _ not _ conformal quantum mechanics in and of themselves . instead , the @xmath2 symmetry of ads@xmath1 might only become exact in some kind of large @xmath144 limit , might persist in a perturbative treatment of a supergravity solution but not at finite @xmath144 , unless of course it resides in a much larger virasoro structure . ] as opposed to the @xmath380 symmetries of ads@xmath381 which should persist at finite @xmath144 ( assuming the @xmath382-function of the dual cft vanishes at finite @xmath144 as is the case for @xmath383 sym ) . as a final note
, we would like to mention some open questions and extensions to our discussion .
we have left question of superpotential corrections to the coulomb branch of the three - node quiver untouched .
though the superpotential will not affect the ground state energy or existence of scaling solutions , it will correct the higher powers in the velocity expansion and it would be interesting to understand whether the @xmath2 symmetry is preserved by such corrections .
the quartic interaction coming from integrating out the @xmath99-terms , @xmath384 , is dominated by an expansion in cactus diagrams in the large @xmath124 limit .
there is no such cactus diagram expansion for the interaction coming from integrating out the @xmath113-term , the quartic bosonic interaction being : @xmath385 .
thus we are confronted with how ( if at all ) does one organize a large @xmath124 expansion in this case . by dimensional
analysis , and inspection of the two - loop feynman diagrams ( see figure [ dxdx2loop ] ) , we expect the velocity squared piece of the coulomb branch lagrangian in the scaling regime to become ( up to @xmath386 coefficients ) : @xmath387 where @xmath388 . in order for the superpotential contribution to
be subleading we would further require : @xmath389 in addition to the condition ( [ scbounds ] ) that forces the system into the scaling regime .
thus , we clearly see that the effect of the superpotential ( a relevant deformation ) becomes important in the deep infrared region of the scaling regime , where detailed stringy physics begins to manifest itself and potentially destroys the @xmath2 .
cccc dxdx2loop1 ( 20,20 ) & dxdx2loop2 ( 20,20 ) & dxdx2loop3 ( 20,20 ) & dxdx2loop7 ( 20,20 ) + dxdx2loop4 ( 20,20 ) & dxdx2loop5 ( 20,20 ) & dxdx2loop6 ( 20,20 ) & dxdx2loop8 ( 20,20 ) we have also left untouched the issue of larger numbers of nodes . in such a case one could consider mixed higgs - coulomb branches .
for example , we could consider a four - node quiver where three nodes are in their higgs branch and the remaining one residing far away in the coulomb branch ( see also @xcite ) , as shown in figure [ quiver2 ] .
given the exponentially large number of ground states in the higgs branch of a three node closed loop quiver , heating such a system up might provide a useful toy model for a d - particle falling into a black hole . ,
@xmath390 and @xmath391 .
, width=480 ]
it is a great pleasure to thank alejandra castro , jan de boer , sheer el showk , dan freedman , alessandra gnecchi , umut gursoy , sean hartnoll , diego hofman , c. martes , edgar shaghoulian , andy strominger , bert vercnocke and erik verlinde for useful discussions .
we are particularly grateful for uncountable insightful discussions with frederik denef throughout the course of this work .
the authors are also grateful to the _ cosmology and complexity _ conference in hydra for their hospitality while this work was in progress .
d.a . would like to acknowledge the hospitality of the aspen center for physics , the crete center for theoretical physics and the open questions in an open universe workshop in istanbul where part of this work was performed .
t.a . would also like to thank the centro de ciencias de benasque for their warm hospitality while this work was in progress .
this work has been partially funded by doe grants de - fg02 - 91er40654 and de - fg02 - 05er41360 and by a grant of the john templeton foundation .
the opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the john templeton foundation .
for convenience , we have used a compact notation whereby latin superscripts denote relative degrees of freedom or arrow directions in the quiver , for example : @xmath63 or @xmath392 .
the supermultiplets in this notation are : @xmath393 and the relative vector multiplet is denotes as @xmath394 .
the only exception to the rule is given by @xmath395 , @xmath396 and @xmath397 .
furthermore , we have : @xmath398 , @xmath65 , @xmath399 .
so with respect to the vector multiplet fields , the _ relative _ lagrangian is only a function of @xmath134 , @xmath135 , @xmath360 , @xmath193 , @xmath400 and @xmath401 .
we also have : @xmath402 .
the orientation of the quiver is encoded by the @xmath403 . for three nodes we deal with the case of a quiver with a closed loop , and without loss of generality the particular choice @xmath404 , @xmath405 and @xmath406 , corresponding to @xmath110 and @xmath146 .
an example of a quiver without closed loops is @xmath407 , @xmath405 , @xmath406 .
the spinors @xmath408 and @xmath409 with @xmath410 transform in the @xmath411 of the @xmath78 and the anti - symmetric symbol @xmath412 is such that @xmath413 .
the @xmath414 are the pauli matrices : @xmath415 for example @xmath416 .
finally the @xmath27 covariant derivative @xmath417 .
we will occasionally revert back to the original @xmath418 notation , particularly in the appendices .
it is interesting to compute the effects on the coulomb branch dynamics due to quantum corrections from the superpotential .
this amounts to a two - loop calculation .
it is important to note that due to a non - renormalization theorem in @xcite , the linear piece of the @xmath32 supersymmetric quantum mechanics lagrangian is constrained to be of the form : @xmath419 with : @xmath420 this form receives _ no _ corrections due to quantum effects originating from the superpotential .
this means , in particular , that the supersymmetric configurations of the coulomb branch which solve @xmath421 for all @xmath46 , both bound and scaling , are unmodified in the presence of a superpotential . on the other hand ,
the quadratic piece can receive quantum corrections in the presence of a superpotential .
an example of a feynman diagram contributing to the @xmath422 term is given in figure [ wfeynman ] .
c dd2loop ( 30,30 )
when going to the coulomb branch one integrates out the massive scalars , expands in small @xmath423 , solves for the @xmath99 equations of motion and feeds the solution back into the action .
we would like to show here that the supersymmetric configurations are in fact preserved for _ finite _ @xmath99 . in the case of two node quivers , upon integrating out the chiral multiplet degrees of freedom @xmath424
one finds the following non - linear equations for @xmath99 ( we take @xmath425 and @xmath215 ) : @xmath426 given that the above equation is a cubic equation in @xmath99 , one can find analytic solutions , which we are given in appendix [ twonoded ] .
it is straightforward to see that at @xmath427 , @xmath428 is a solution to the non - linear equations . plugging @xmath428 back into the lagrangian ( which for zero velocity is the potential itself ) shows that at @xmath427 the system has a zero energy . away from @xmath427
the lagrangian receives finite @xmath99 corrections .
it is not hard to convince one s self , however , that the potential will never acquire another minimum due to finite @xmath99 effects .
we argue this in appendix [ twonoded ] . for three nodes the non - linear equations of @xmath132 become ( we take @xmath429 and pick the @xmath124 s to form a closed loop in the quiver diagram ) : @xmath430 and cyclic permutations thereof .
these equations can no longer be solved analytically .
if we set the @xmath332 s to zero we can see that at @xmath431 , @xmath432 solves the linear equations and so the scaling solutions persist at finite @xmath99 for zero @xmath332 . for non - zero @xmath332 s , setting @xmath433 and expanding in small @xmath151 , we find that again @xmath432 is a consistent solution order by order in the @xmath151 perturbation theory .
the existence of new bound states or scaling solutions which are non - supersymmetric due to finite @xmath99 effects becomes far more intricate in this case .
a preliminary numerical scan seems to suggest there are none and we hope to report further on this in the future . for two - nodes ,
the @xmath99-term equation is given by ( [ 2nd ] ) .
this equation can be solved analytically for @xmath99 .
supersymmetric bound states are found when @xmath434 at @xmath435 . in what follows ,
we choose @xmath215 and allow @xmath332 to have any sign .
equation ( [ 2nd ] ) is equivalent to solving @xmath436 which can be turned into a depressed cubic of the form @xmath437 using the subsitution @xmath438 and identifying @xmath439 since @xmath440 , the three roots of this equation may be written as @xcite : @xmath441 for @xmath442 .
for all roots to be real , the arguments of the @xmath443 must be between @xmath444 $ ] , which implies @xmath445 or @xmath446 .
the @xmath447 branch is real for all values of @xmath448 . in figures [ dsol1 ] and [ dsol2 ]
we show some plots of the solutions .
$ ] ( blue ) and @xmath449 $ ] ( violet ) .
right : plot of @xmath450 $ ] ( blue ) and @xmath451 $ ] ( violet ) .
both plots are for @xmath452 .
notice that when complex , the solutions form a conjugate pair.,width=240 ] $ ] ( blue ) and @xmath449 $ ] ( violet ) .
right : plot of @xmath450 $ ] ( blue ) and @xmath451 $ ] ( violet ) . both plots are for @xmath452 .
notice that when complex , the solutions form a conjugate pair.,title="fig:",width=240 ] $ ] ( blue ) and @xmath449 $ ] ( violet ) .
right : plot of @xmath450 $ ] ( blue ) and @xmath451 $ ] ( violet ) .
both plots are for @xmath452 .
notice that when complex , the solutions form a conjugate pair.,title="fig:",width=240 ] in figure [ dsmall ] we display the effective potential for the @xmath453 solution and its comparison to the effective potential obtained by expanding @xmath99 to second order in @xmath183 and then evaluating the potential on - shell .
we notice that the position of the minimum is unaffected . evaluating the potential on the @xmath454 and @xmath455 results in an unphysical complex potential for small values of @xmath448 . scanning the parameters for all possible combinations of signs of @xmath456 results in no other bound states
. evaluated on @xmath453 for @xmath452 .
right : plot of @xmath320 evaluated on the perturbative @xmath99 solution in violet , compared with the full non - perturabative @xmath453 in blue.,title="fig:",width=288 ] evaluated on @xmath453 for @xmath452 .
right : plot of @xmath320 evaluated on the perturbative @xmath99 solution in violet , compared with the full non - perturabative @xmath453 in blue.,title="fig:",width=288 ]
in this appendix we present some details leading to the coulomb branch lagrangian of a three - node quiver .
as was shown in @xcite , the @xmath383 supersymmetric quantum mechanics of a @xmath27 vector multiplet has a lagrangian whose linear piece in the velocity and @xmath99 fields is completely fixed by supersymmetry .
explicit expressions are given in ( [ linear1 ] ) and ( [ linear2 ] ) .
in section [ cmetric ] below , we show that the coefficient of the @xmath457 term of the lagrangian is the same as that of the @xmath458 term , upon integrating out the chiral matter . in section [ potential ]
we integrate out the auxiliary @xmath99 fields in the coulomb theory and obtain an expression for the effective potential on the position degrees of freedom .
recall the lagrangian of a @xmath283-node quiver theory , after setting the fermionic @xmath151 fields to zero , is given by : , @xmath37 , @xmath459 , @xmath460 , @xmath461 .
and so on .
we also supress the greek indices on the chiral multiplet fields .
also , we do not decouple the center of mass degrees of freedom and switch off the superpotential . ] @xmath462~. \label{pnodelagrangian}\end{gathered}\ ] ] for the case @xmath463 there are three pairs of @xmath464 to be considered : @xmath465 , @xmath466 and @xmath467 .
we do not consider contributions from the superpotential in this appendix .
we take @xmath468 , @xmath469 and @xmath470.the @xmath459 propagator is given by : @xmath471 and the @xmath460 propagator is given by : @xmath472 where @xmath473 ( recall that each lower index denotes its corresponding node ) .
we evaluate our momentum integrals on the imaginary axis , which is guaranteed to give the same result had we evaluated the integral along a real contour with the usual feynman pole prescription , to lowest order in the the external momenta @xmath474 .
higher order terms in the external momenta will differ by factors of @xmath46 .
first let us consider the diagonal terms and more specifically the contribution to @xmath476 .
the two diagrams that contribute are shown in figure [ d1 ^ 2 ] .
cc d1selfenergy ( 30,30 ) & d1selfenergy2 ( 30,30 ) + @xmath477&@xmath478 the first diagram , series expanded for small external momentum @xmath474 , is given by @xmath479 while the second is given by @xmath480 therefore , ignoring the @xmath481 terms , the total contribution to the effective lagrangian is @xmath482 the @xmath483 and @xmath484 terms can be obtained by cyclic permutation .
now let us consider the off diagonal terms and more specifically @xmath485 .
the only diagram that contributes to 1-loop order is : cc d1d2selfenergy ( 30,30 ) this diagram differs only by a sign and a factor of 2 compared to @xmath477 of figure [ d1 ^ 2 ] .
therefore , the @xmath485 term of the effective lagrangian is @xmath486 other off diagonal terms can be obtained by cyclic permutation . in order to obtain the correction to the quadratic velocity terms in the lagrangian , we consider fluctuations the bosonic degrees of freedom @xmath488 .
terms of this form can be divided in terms that are diagonal in the quiver such as @xmath489 and others that are off diagonal such as @xmath490 . for each of these
there is a further subdivision to terms that are diagonal in the vector component such as @xmath491 and @xmath492 and terms that are off diagonal in the vector component such as @xmath493 and @xmath494 .
we begin by considering terms that are off diagonal in the quiver index , and off diagonal in the vector index .
for concreteness we will study @xmath495 .
the diagrams that contribute are shown in figure [ dxpdxqij ] .
cc dx11dx22phi ( 30,30 ) & dx11dx22psi ( 30,30 ) + @xmath496&@xmath497 the first diagram , series expanded for small external momentum @xmath474 , is given by : @xmath498 the second diagram , series expanded for small external momentum @xmath474 , is given by : @xmath499 summing the two , the only term that remains uncancelled is the term linear in @xmath474 .
next we consider term off diagonal in the quiver index , but diagonal in the vector index , for example @xmath500 .
the contributing diagrams are shown in figure [ dxpdxqii ] .
ccc dx11dx21phi ( 30,30 ) & dx11dx21psi ( 30,30 ) & dx11dx21loop ( 30,30 ) + @xmath496&@xmath497&@xmath501 the first diagram is given by ( [ eqdxpdxqij ] ) with @xmath502 replaced by @xmath503 .
the second diagram gives : @xmath504 the third diagram is given by : @xmath505 let us now consider terms that are diagonal in the quiver index .
the feynman diagrams contributing to @xmath506 are shown in figure [ dxpdxpij ] .
ccc dx1idx1jphi ( 30,30 ) & dx1idx1jpsi ( 30,30 ) & dx1idx1iloop ( 30,30 ) + @xmath496&@xmath497&@xmath501 + & & + dx1idx1jphi31 ( 30,30 ) & dx1idx1jpsi31 ( 30,30 ) & dx1idx1iloop31 ( 30,30 ) + @xmath507&@xmath508&@xmath509 one finds after similar calculations the following contribution to the effective lagrangian : @xmath510 having computed all relevant feynman diagrams that amount to integrating out the @xmath459 and @xmath460 fields we can now write the resulting quadratic piece of the effective lagrangian : @xmath511 notice that the coefficient of @xmath457 indeed matches that of @xmath512 .
this can be simply generalized to the @xmath144-particle case , where the lagrangian takes the same form as ( [ 3nodeeffective ] ) .
as was noted before , the linear piece was fixed by supersymmetry and given by ( [ linear1 ] ) and ( [ linear2 ] ) @xcite .
in this appendix we give the necessary formulae for the three - node coulomb branch potential . upon integrating out the chiral fields @xmath513 in ( [ quiverqm ] ) , while keeping the @xmath179 fixed and time independent , and expanding up to quadratic order in the @xmath514 fields , the @xmath99-dependent piece of the lagrangian is : @xmath515 noting that @xmath65 , we must integrate out @xmath360 and @xmath193 from @xmath516 .
the resulting on - shell lagrangian is simply the potential @xmath517 .
the equations of motion for @xmath132 , using @xmath518 , are given by : @xmath519 writing the above equations as : @xmath520 , the solution is simply given by : @xmath521 notice that the supersymmetric solution @xmath522 is indeed given by @xmath523 which is nothing more than equation ( [ bound ] ) . in the scaling regime ,
where @xmath524 , and @xmath525 , @xmath55 and @xmath56 ( or equivalently @xmath518 ) , the potential can be written as : @xmath526 where @xmath527
the thermal effective potential for the two - node quiver at finite temperature can be derived from ( [ hotl ] ) : @xmath528 wick rotating @xmath529 and periodically identifying @xmath530 , introduces the thermal ensemble .
the operator @xmath306 has eigenvalues on the matsubara frequencies : @xmath531 for bosons and @xmath532 for fermions , with @xmath533 . with @xmath534 and
denoting @xmath535 this yields : @xmath536 where we have now included a @xmath537 in the coefficient of the sums to cancel out the contribution from the integral over euclidean time in @xmath538 . taking a derivative with respect to @xmath539 : @xmath540 we will treat the two sums separately using contour integration methods
we write ( [ thermall ] ) as @xmath541 , denoting the two summands respectively .
consider : @xmath542 with @xmath543 and @xmath544 and @xmath545 a circle of radius @xmath546 around the origin with @xmath547 . given that @xmath548 decays rapidly enough , the integral evaluates to zero along @xmath545 and we obtain :
@xmath549 for the fermionic sum , the procedure is completely analgous , and ( using @xmath550 ) we find : @xmath551 we need to integrate these two sums with respect to @xmath539 to obtain @xmath552 . using the following identities : @xmath553 and @xmath554 , we find : @xmath555 as claimed .
j. michelson and a. strominger , `` the geometry of ( super)conformal quantum mechanics , '' commun .
phys . * 213 * , 1 ( 2000 ) [ hep - th/9907191 ] .
a. strominger and c. vafa , `` microscopic origin of the bekenstein - hawking entropy , '' phys .
b * 379 * , 99 ( 1996 ) [ hep - th/9601029 ] . j. c. breckenridge , r. c. myers , a. w. peet and c. vafa , `` d - branes and spinning black holes , '' phys . lett .
b * 391 * , 93 ( 1997 ) [ hep - th/9602065 ] .
s. h. katz , a. klemm and c. vafa , `` m theory , topological strings and spinning black holes , '' adv .
phys . * 3 * , 1445 ( 1999 ) [ hep - th/9910181 ] . m. guica and a. strominger , `` wrapped m2/m5 duality , '' jhep * 0910 * , 036 ( 2009 ) [ hep - th/0701011 ] .
d. anninos , w. li , m. padi , w. song and a. strominger , `` warped ads(3 ) black holes , '' jhep * 0903 * , 130 ( 2009 ) [ arxiv:0807.3040 [ hep - th ] ] .
d. anninos , `` hopfing and puffing warped anti - de sitter space , '' jhep * 0909 * , 075 ( 2009 ) [ arxiv:0809.2433 [ hep - th ] ]
. m. guica , t. hartman , w. song and a. strominger , `` the kerr / cft correspondence , '' phys .
d * 80 * , 124008 ( 2009 ) [ arxiv:0809.4266 [ hep - th ] ] .
m. guica and a. strominger , `` microscopic realization of the kerr / cft correspondence , '' jhep * 1102 * , 010 ( 2011 ) [ arxiv:1009.5039 [ hep - th ] ] .
s. detournay , t. hartman and d. m. hofman , `` warped conformal field theory , '' phys .
d * 86 * , 124018 ( 2012 ) [ arxiv:1210.0539 [ hep - th ] ] .
x . huang , a. klemm , m. marino and a. tavanfar , `` black holes and large order quantum geometry , '' phys .
d * 79 * , 066001 ( 2009 ) [ arxiv:0704.2440 [ hep - th ] ] . c. vafa , `` black holes and calabi - yau threefolds , '' adv .
* 2 * , 207 ( 1998 ) [ hep - th/9711067 ] .
j. m. maldacena , a. strominger and e. witten , `` black hole entropy in m theory , '' jhep * 9712 * , 002 ( 1997 ) [ hep - th/9711053 ] .
d. anninos , s. a. hartnoll and d. m. hofman , `` static patch solipsism : conformal symmetry of the de sitter worldline , '' class . quant .
* 29 * , 075002 ( 2012 ) [ arxiv:1109.4942 [ hep - th ] ] .
d. anninos , `` de sitter musings , '' int .
j. mod .
a * 27 * , 1230013 ( 2012 ) [ arxiv:1205.3855 [ hep - th ] ] .
e. brezin , c. itzykson , g. parisi and j. b. zuber , `` planar diagrams , '' commun .
phys . * 59 * , 35 ( 1978 ) . i. r. klebanov , `` string theory in two - dimensions , '' in * trieste 1991 , proceedings , string theory and quantum gravity 91 * 30 - 101 and princeton univ
. - pupt-1271 ( 91/07,rec.oct . )
72 p [ hep - th/9108019 ] .
p. h. ginsparg and g. w. moore , `` lectures on 2-d gravity and 2-d string theory , '' in * boulder 1992 , proceedings , recent directions in particle theory * 277 - 469 . and yale univ .
new haven - yctp - p23 - 92 ( 92,rec.apr.93 ) 197 p. and los alamos nat . lab .
- la - ur-92 - 3479 ( 92,rec.apr.93 ) 197 p [ hep - th/9304011 ]
. j. polchinski , `` what is string theory ? , '' hep - th/9411028 . m. r. douglas and g. w. moore , `` d - branes , quivers , and ale instantons , '' hep - th/9603167 . f. denef , `` quantum quivers and hall / hole halos , '' jhep * 0210 * , 023 ( 2002 ) [ hep - th/0206072 ] . f. denef and g. w. moore , `` split states , entropy enigmas , holes and halos , '' jhep * 1111 * , 129 ( 2011 ) [ hep - th/0702146 [ hep - th ] ] . i. bena , m. berkooz , j. de boer , s. el - showk and d. van den bleeken , `` scaling bps solutions and pure - higgs states , '' jhep * 1211 * , 171 ( 2012 ) [ arxiv:1205.5023 [ hep - th ] ] . s. -j .
lee , z. -l .
wang and p. yi , `` quiver invariants from intrinsic higgs states , '' jhep * 1207 * , 169 ( 2012 ) [ arxiv:1205.6511 [ hep - th ] ] .
j. manschot , b. pioline and a. sen , `` from black holes to quivers , '' jhep * 1211 * , 023 ( 2012 ) [ arxiv:1207.2230 [ hep - th ] ] .
j. manschot , b. pioline and a. sen , `` on the coulomb and higgs branch formulae for multi - centered black holes and quiver invariants , '' jhep * 1305 * , 166 ( 2013 ) [ arxiv:1302.5498 [ hep - th ] ] .
i. bena , c. -w .
wang and n. p. warner , `` plumbing the abyss : black ring microstates , '' jhep * 0807 * , 019 ( 2008 ) [ arxiv:0706.3786 [ hep - th ] ] .
r. jackiw , introducing scale symmetry , " physics today 25 , 2327 ( 1972 ) v. de alfaro , s. fubini and g. furlan , `` conformal invariance in quantum mechanics , '' nuovo cim . a * 34 * , 569 ( 1976 ) .
e. witten , `` phases of n=2 theories in two - dimensions , '' nucl .
b * 403 * , 159 ( 1993 ) [ hep - th/9301042 ] .
s. -j .
lee , z. -l .
wang and p. yi , `` abelianization of bps quivers and the refined higgs index , '' arxiv:1310.1265 [ hep - th ] .
m. r. douglas , d. n. kabat , p. pouliot and s. h. shenker , `` d - branes and short distances in string theory , '' nucl . phys .
b * 485 * , 85 ( 1997 ) [ hep - th/9608024 ] .
f. denef , `` supergravity flows and d - brane stability , '' jhep * 0008 * , 050 ( 2000 ) [ hep - th/0005049 ] .
a. nersessian and v. ohanyan , `` multi - center micz - kepler systems , '' theor .
* 155 * , 618 ( 2008 ) [ arxiv:0705.0727 [ math - ph ] ] .
d. anninos , t. anous , f. denef , g. konstantinidis and e. shaghoulian , `` supergoop dynamics , '' jhep * 1303 * , 081 ( 2013 ) [ arxiv:1205.1060 [ hep - th ] ] .
r. britto - pacumio , j. michelson , a. strominger and a. volovich , `` lectures on superconformal quantum mechanics and multiblack hole moduli spaces , '' hep - th/9911066 .
j. de boer , s. el - showk , i. messamah and d. van den bleeken , `` quantizing n=2 multicenter solutions , '' jhep * 0905 * , 002 ( 2009 ) [ arxiv:0807.4556 [ hep - th ] ] .
g. papadopoulos , `` conformal and superconformal mechanics , '' class .
* 17 * , 3715 ( 2000 ) [ hep - th/0002007 ] .
a. v. smilga , perturbative corrections to effective zero - mode hamiltonian in supersymmetric qed . "
nuclear physics b 291 ( 1987 ) : 241 - 255 . j. michelson and a. strominger ,
jhep * 9909 * , 005 ( 1999 ) [ hep - th/9908044 ] .
r. c. ferrel and d. m. eardley , slow - motion scattering and coallescence of maximally charged black holes , " phys .
lett . * 59 * , 14 ( 1987 ) a. maloney , m. spradlin and a. strominger , `` superconformal multiblack hole moduli spaces in four - dimensions , '' jhep * 0204 * , 003 ( 2002 ) [ hep - th/9911001 ] .
a. strominger , `` ads(2 ) quantum gravity and string theory , '' jhep * 9901 * , 007 ( 1999 ) [ hep - th/9809027 ] .
t. azeyanagi , t. nishioka and t. takayanagi , `` near extremal black hole entropy as entanglement entropy via ads(2)/cft(1 ) , '' phys .
d * 77 * , 064005 ( 2008 ) [ arxiv:0710.2956 [ hep - th ] ] .
a. sen , `` state operator correspondence and entanglement in @xmath556 , '' entropy * 13 * , 1305 ( 2011 ) [ arxiv:1101.4254 [ hep - th ] ] .
j. m. maldacena , j. michelson and a. strominger , `` anti - de sitter fragmentation , '' jhep * 9902 * , 011 ( 1999 ) [ hep - th/9812073 ] . c. chamon , r. jackiw , s. -y .
pi and l. santos , `` conformal quantum mechanics as the cft@xmath557 dual to ads@xmath1 , '' phys .
b * 701 * , 503 ( 2011 ) [ arxiv:1106.0726 [ hep - th ] ] .
t. banks , w. fischler , s. h. shenker and l. susskind , `` m theory as a matrix model : a conjecture , '' phys .
d * 55 * , 5112 ( 1997 ) [ hep - th/9610043 ] .
k. becker and m. becker , `` a two loop test of m(atrix ) theory , '' nucl .
b * 506 * , 48 ( 1997 ) [ hep - th/9705091 ] .
m. berkooz and h. l. verlinde , `` matrix theory , ads / cft and higgs - coulomb equivalence , '' jhep * 9911 * , 037 ( 1999 ) [ hep - th/9907100 ] .
j. fuchs , `` remarks on supersymmetric quantum mechanics at finite temperature , '' annals phys . *
165 * , 285 ( 1985 ) .
a. das , a. kharev and s. v. mathur , phys .
181b * , 299 ( 1986 ) d. n. kabat , g. lifschytz and d. a. lowe , `` black hole thermodynamics from calculations in strongly coupled gauge theory , '' int .
j. mod .
a * 16 * , 856 ( 2001 ) [ phys .
lett . *
86 * , 1426 ( 2001 ) ] [ hep - th/0007051 ] .
d. n. kabat , g. lifschytz and d. a. lowe , `` black hole entropy from nonperturbative gauge theory , '' phys .
d * 64 * , 124015 ( 2001 ) [ hep - th/0105171 ] .
n. kawahara , j. nishimura and s. takeuchi , `` phase structure of matrix quantum mechanics at finite temperature , '' jhep * 0710 * , 097 ( 2007 ) [ arxiv:0706.3517 [ hep - th ] ] .
s. catterall and t. wiseman , `` towards lattice simulation of the gauge theory duals to black holes and hot strings , '' jhep * 0712 * , 104 ( 2007 ) [ arxiv:0706.3518 [ hep - lat ] ] . k. n. anagnostopoulos , m. hanada , j. nishimura and s. takeuchi , `` monte carlo studies of supersymmetric matrix quantum mechanics with sixteen supercharges at finite temperature , '' phys .
lett . *
100 * , 021601 ( 2008 ) [ arxiv:0707.4454 [ hep - th ] ] .
y. -h .
lin , s. -h .
shao , y. wang and x. yin , `` a low temperature expansion for matrix quantum mechanics , '' arxiv:1304.1593 [ hep - th ] .
n. iizuka , d. kabat , s. roy and d. sarkar , `` black hole formation in fuzzy sphere collapse , '' phys .
d * 88 * , 044019 ( 2013 ) [ arxiv:1306.3256 [ hep - th ] ] . c. asplund , d. berenstein and d. trancanelli , `` evidence for fast thermalization in the plane - wave matrix model , '' phys .
* 107 * , 171602 ( 2011 ) [ arxiv:1104.5469 [ hep - th ] ] .
d. anninos , t. anous , j. barandes , f. denef and b. gaasbeek , `` hot halos and galactic glasses , '' jhep * 1201 * , 003 ( 2012 ) [ arxiv:1108.5821 [ hep - th ] ] .
b. d. chowdhury and b. vercnocke , `` new instability of non - extremal black holes : spitting out supertubes , '' jhep * 1202 * , 116 ( 2012 ) [ arxiv:1110.5641 [ hep - th ] ] . b. d. chowdhury , d. r. mayerson and b. vercnocke , `` phases of non - extremal multi - centered bound states , '' arxiv:1307.5846 [ hep - th ] .
d. anninos , t. anous , f. denef and l. peeters , `` holographic vitrification , '' arxiv:1309.0146 [ hep - th ]
. s. d. majumdar , `` a class of exact solutions of einstein s field equations , '' phys .
rev . * 72 * , 390 ( 1947 ) .
a. papapetrou , static solution of the equations of the gravitational field for an arbitrary charge distribution , " proc .
irish acad .
a ) a51 , 191 - 204 ( 1947 ) .
e. p. verlinde , `` on the origin of gravity and the laws of newton , '' jhep * 1104 * , 029 ( 2011 ) [ arxiv:1001.0785 [ hep - th ] ] .
a. amir , y. oreg and y. imry , localization , anomalous diffusion and slow relaxations : a random distance matrix approach , " [ arxiv:1002.2123 [ cond-mat.dis-nn ] ] y. nakayama , `` a lecture note on scale invariance vs conformal invariance , '' arxiv:1302.0884 [ hep - th ] .
n. iizuka , d. n. kabat , g. lifschytz and d. a. lowe , `` probing black holes in nonperturbative gauge theory , '' phys .
d * 65 * , 024012 ( 2002 ) [ hep - th/0108006 ] .
n. iizuka and j. polchinski , `` a matrix model for black hole thermalization , '' jhep * 0810 * , 028 ( 2008 ) [ arxiv:0801.3657 [ hep - th ] ] . | quiver quantum mechanics describes the low energy dynamics of a system of wrapped d - branes .
it captures several aspects of single and multicentered bps black hole geometries in four - dimensional @xmath0 supergravity such as the presence of bound states and an exponential growth of microstates .
the coulomb branch of an abelian three node quiver is obtained by integrating out the massive strings connecting the d - particles .
it allows for a scaling regime corresponding to a deep ads@xmath1 throat on the gravity side . in this scaling regime
, the coulomb branch is shown to be an @xmath2 invariant multi - particle superconformal quantum mechanics .
finally , we integrate out the strings at finite temperature rather than in their ground state and show how the coulomb branch ` melts ' into the higgs branch at high enough temperatures . for scaling solutions the melting occurs for arbitrarily small temperatures , whereas bound states can be metastable and thus long lived . throughout the paper ,
we discuss how far the analogy between the quiver model and the gravity picture , particularly within the ads@xmath1 throat , can be taken .
= 1 mm + dionysios anninos@xmath3 , tarek anous@xmath4 , paul de lange@xmath5 and george konstantinidis@xmath3 @xmath3 _ stanford institute for theoretical physics , stanford university
_ + @xmath4 _
center for theoretical physics , mit
_ + @xmath5 _ institute for theoretical physics , university of amsterdam _ + |
to determine @xmath7 , one needs to measure precise distances to galaxies sufficiently distant for local departures from a smooth hubble flow to be of negligible importance .
consequently we need to extend the distance scale to distances of order @xmath8 mpc or more .
although the _ hubble space telescope _ ( _ hst _ ) has greatly extended the range of direct cepheid based distances , it can not reach the region where the hubble flow dominates , and despite the successful determination of many new cepheid distances with _ hst _
, the dispute over the global value of the hubble constant is still not settled .
methods to reach the hubble flow via the virgo cluster through the use of various secondary distance indicators have tended to obtain values larger than @xmath9 km s@xmath5 mpc@xmath5 ( @xcite ) , whereas methods using a cepheid calibration of the peak brightness of type ia supernovae consistently have given values below @xmath10 km s@xmath5 mpc@xmath5 ( @xcite ) .
recently , methods based on correlations between light curve shapes and the luminosity of type ia supernovae ( @xcite ) , as well as the nebular sn ia method based on modeling of the forbidden line emission at late phases ( @xcite ) , have resulted in values in the range @xmath10@xmath9 km s@xmath5 mpc@xmath5 .
part of the problem is evidently the significant depth of the virgo cluster ( @xcite ) .
this uncertainty is eliminated if secondary indicators can be applied to extend distance measurements directly to more remote clusters .
hjorth & tanvir ( 1997 ) have combined the cepheid distance to the leo i group ( @xcite ) with a comparison of the fundamental plane ( fp ) of early type galaxies in the leo i group with the fp of the coma cluster ( @xcite ) . also , baum _
( 1995 , 1997 ) have directly compared the globular cluster luminosity functions for two giant ellipticals in the coma cluster with the combined luminosity function for the milky way and m31 .
these studies also find values in the range @xmath10@xmath9 km s@xmath5 mpc@xmath5 . in the same spirit
we shall , for the first time , derive the distance to the coma cluster based on surface brightness fluctuations .
the sbf method was introduced by tonry & schneider ( 1988 ) as a precise secondary distance indicator for early type galaxies . until now applications of sbf distances
have been limited to ground based @xmath0band ( @xcite ) and @xmath11band ( @xcite ) observations .
the superb resolution of the _
hst _ should , however , make the wfpc2 camera an ideal instrument for obtaining sbf distances to early type galaxies in the coma cluster , the main difficulty being the relatively modest aperture of the telescope .
a typical luminosity fluctuation must produce at least one detected photo electron during the total exposure time in order not to be overwhelmed by the photon noise .
the present work is based on eight @xmath12 s exposures of ngc 4881 obtained with the pc1 camera in the wide
band @xmath13 ( @xmath0 ) filter .
the basic data reductions are fully described by baum _ et al . _
our final sbf flux @xmath14 is equivalent to @xmath15 detected photons in the @xmath13band in a total exposure time of @xmath16 s. surface brightness fluctuations are due to the random distribution of discrete point sources ( stars ) across the field of the detector . @xmath14
is defined as the flux of a hypothetical ` fluctuation star ' which , when randomly spread across the detector , produces the same flux mean and variance per pixel as the observed stellar population .
as the statistical properties of the power spectrum are completely determined by the mean and variance of the observed flux per pixel , we are allowed to simulate luminosity fluctuations by a spatially random distribution of a suitable number ( in this case @xmath17 stars / pixel ) of ` fluctuation stars ' over the field of the detector .
we have thus obtained an accurate estimate of the random error , as well as the statistical bias , by reducing many such simulated fluctuation fields . a high resolution
poissonian stellar fluctuation field is first convolved with a point spread function ( psf ) created by the tiny tim ( v4.0 ) program ( @xcite ) .
subsequently , this array is rebinned to the resolution of the pc1 camera , resulting in a noiseless stellar fluctuation frame . finally , poissonian photon noise and gaussian read noise are added .
the whole sequence is repeated @xmath18 times , and all these frames are reduced in the same way as the real data .
the sample average of the derived sbf magnitudes is @xmath19 mag fainter than the true value ( i.e. , we must _ subtract _
@xmath20 mag from the derived values ) , and the standard deviation on a measurement of a single simulation ( @xmath21 s ) is @xmath22 mag .
a kolmogorov
smirnov test ( @xcite ) did not detect any deviation from normality .
our error estimate relies completely on the monte carlo simulations , which assume that the psf is well described by the tiny tim model .
fortunately , figure 1 shows that the power spectrum of the tiny tim psf is a nearly perfect match to a high signal to noise power spectrum of an annular region near the center of m32 .
the power spectrum is obtained from a 400 s @xmath13 exposure with the pc1 camera ( _ hst _ archive data , id5464 ) , and the fluctuation amplitude is derived from the same angular area and by the same procedure as we use for the ngc 4881 data .
this shows that it is fully acceptable to use the tiny tim model instead of an empirical psf .
we have chosen to derive the fluctuation magnitude @xmath1 for an annular region between @xmath23 and @xmath24 from the nucleus of ngc 4881 . in that region ,
the white noise will be dominated by photon statistics , not by the read noise ( @xmath25 ) , and a good fit to a model galaxy is obtained ( @xmath26 ) .
this letter is not the place for a detailed description of the sbf method , especially as many of the reduction procedures are quite similar to the ones described by sodemann & thomsen ( 1995 ) .
instead , we shall concentrate our description on the steps that are specific for this particular application .
the sbf amplitude , @xmath27 , is obtained from the radial power spectrum of the normalized residual image and the radial power spectrum of a high resolution psf ( a factor of 4 sub sampling ) as calculated by the tiny tim ( v4.0 ) program ( @xcite ) .
the volume of the psf is normalized to one , so the derived sbf amplitude ( flux ) corresponds to a total magnitude .
specifically , @xmath28 is calculated using equation ( 15 ) in sodemann & thomsen ( 1995 ) .
we must emphasize that the lowest wave numbers are omitted , as they are influenced by flat
fielding and galaxy fitting errors . in figure 2
we plot the scaled power spectrum of the psf on top of the radial power spectrum for the annular region in ngc 4881 .
the constant power spectrum of the white noise , 0.030 dn , has been subtracted from the plotted data points .
note that the high power values with wave numbers less than 12 are not used in our derivation of @xmath28 .
we wish to thank the anonymous referee for pointing out that the power spectrum of the ngc 4881 data appeared more noisy than the power spectrum of the simulated data , i.e. , the power spectrum of the white noise in the ngc 4881 data was significantly higher than expected from photon noise and assumed read noise ( @xmath29 ) .
several identifiable sources evidently contributed to this excess white noise : the actual read noise ( @xmath30 ) was higher than had been assumed , there was a poissonian noise associated with the subtracted dark current ( @xmath31 ) , and the effective number of frames was slightly less than 8 due to the rejection of pixels containing cosmic ray events .
however , these identifiable sources did not completely explain all of the excess white noise .
some of it must have been introduced by the pipeline processing . in order to evaluate the importance of the excess white noise
, we added the missing amount of gaussian noise to 25 of the already existing simulated fluctuation frames .
the standard deviation on a measurement of a single simulation increased from @xmath22 mag to @xmath32 mag .
figure 3 shows a plot of the scaled power spectrum of the psf on top of the radial power spectrum for one of the simulated fluctuation fields with the correct amount of gaussian noise added .
although greater noise reduces the precision of our sbf distance to the coma cluster , it does not alter the sbf distance itself .
our calculated amplitude is @xmath33 in @xmath12 s , which according to equation ( 7 ) of holtzman _ et al . _ ( 1995 ) gives a total fluctuation magnitude of @xmath34 .
@xmath35 mag to this value in order to bring it onto the instrumental aperture ( @xmath36 ) system defined by holtzman _
et al . _
in addition we must _ add _
@xmath20 mag because at the time of the ngc 4881 observations the ccd was operated at @xmath37 @xmath38c instead of @xmath39 @xmath38c , which was used after 23 april 1994 ( @xcite ) .
we can transform the instrumental magnitude @xmath40 to cousins @xmath1 by applying holtzman _
et al . _
s ( 1995 ) synthetic transformation , which is equivalent to @xmath41 ( ( * ? ? ?
9 ) ) for a typical red fluctuation color of @xmath42 .
although it is possible that the transformation from the @xmath13 band to the cousins @xmath0band may be different for the fluctuations than it is for stars , we derived @xmath43 for a similar region in m32 , which is consistent with the transformation in light of the @xmath1 values given for m32 by tonry _
et al . _
( 1990 ) and sodemann & thomsen ( 1996 ) .
the fact that the @xmath40 for m32 ( where @xmath44 ) , transformed according to holtzman _
_ , agrees with the ground based @xmath1 for this galaxy does not completely ensure that there will be similar good agreement for the redder stellar populations of giant ellipticals .
in order to settle this question we have compared @xmath45 , as derived from stellar population models ( @xcite ) having solar metallicity and ages between 2 gyr and 17 gyr , with @xmath46 for the giant stars used in the models . the fluctuation magnitude @xmath40 transforms , within a few hundreds of a magnitude , in the same way as the f814w magnitude of giant stars if one chooses the stars red enough to match the sbf colors .
actually , a 12 gyr model , which matches the average @xmath47 index of our annulus in ngc 4881 , has an offset of @xmath48 and a fluctuation color of @xmath49 . a red giant library
star with @xmath50 has an offset of @xmath51 .
the difference between this value and the offset @xmath52 given by holtzman _
et al . _
( 1995 ) is almost certainly due to differences in the @xmath0 passbands used to derive the synthetic colors .
et al . _
used landolt @xmath53 response curves , whereas worthey ( 1994 ) applied the @xmath53 curves of bessell ( 1990 ) .
the difference between landolt & bessell filter passbands ( @xmath54 mag ) is an indication of the remaining uncertainty in the transformation of @xmath13 for m stars . in the following
we shall use @xmath55 .
the variance per pixel due to unmasked globular clusters ( gcs ) can be written as @xmath56 ( @xcite ) , where @xmath57 is the number of detected gcs per pixel , @xmath58 is a flux defining the location of the apparent gc luminosity function ( gclf ) , and @xmath59 is a factor that is calculated from the apparent gclf and the completeness function ( a detection probability ) . for a gaussian lf , @xmath59
is given by equation ( 4 ) of sodemann & thomsen ( 1995 ) .
following the arguments of baum _
et al . _
( 1995 , 1997 ) we shall use a model based on transporting the mw+m31 gclf to a coma distance of @xmath60 ( @xcite ) . the combined gc data for the milky way and m31 are well fitted by a hyperbolic gclf of the form ( @xcite ) @xmath61 , where @xmath62 is the magnitude of the flux , @xmath58 , at the intersection of the asymptotes . as the detection probabilities were derived for the @xmath63band the completeness function had to be shifted by the mean color @xmath64 ( @xcite ) of the gc system in ngc 4881 . in a similar way ,
the variance per pixel due to unmasked background galaxies can be expressed as @xmath65 , where @xmath66 is the number of detected background galaxies per pixel and @xmath67 is a factor that is calculated from the apparent galaxy lf and the completeness function .
the slope of the galaxy lf is taken as the average of the values given for the @xmath0band by tyson ( 1988 ) and lilly _
et al . _
( 1991 ) , respectively . for galaxies
it is not so obvious that we can use our shifted completeness function for the @xmath63band to calculate @xmath67 for the @xmath0band , because galaxies have a much larger color spread than gcs .
fortunately , the relative contribution from galaxies , as compared to the one from gcs , is only @xmath68 so we can safely neglect the error introduced by assuming that background galaxies have the same colors as the gcs .
the contribution to the sbf flux from unmasked gcs and background galaxies is calculated as @xmath69 , where @xmath70 is the mean galaxy flux per pixel .
there are 13 detected gcs inside the annulus [ 3 , 5 ] , and we expect only about @xmath71 background galaxies brighter than detection threshold within the same annulus ( @xcite ) . using these values we find that we must decrease the measured sbf flux by @xmath72 dn .
we must accordingly _
@xmath73 mag in order to correct for the undetected point sources .
if we instead use the gaussian gclf ( @xmath74 , @xmath75 ) suggested by whitmore ( 1996 ) we obtain a correction of @xmath76 mag , which is not significantly different from the value derived for the hyperbolic gclf .
correction for the integrated spectrum of an elliptical galaxy at the coma distance is quite modest for the @xmath0band , but fluctuation colors are very red , so it is much larger for @xmath1 .
correction for @xmath40 is @xmath77 mag , where the redshift ( @xmath78 ) of ngc 4881 relative to the sun ( @xcite ) has been used .
the quoted error was estimated from a comparison of the k
corrections derived from stellar population models ( @xcite ) having solar metallicity and ages between 5 gyr and 12 gyr .
the galactic absorption in the direction of ngc 4881 is @xmath79 ( @xcite ) .
the final corrected fluctuation magnitude for ngc 4881 is then @xmath80 .
we shall calibrate the sbf distance to ngc 4881 in terms of the distance @xmath81 mpc ( random errors ) to ngc 3379 @xmath82 m 105 in the leo i group .
we have adopted the _ average _ sbf distance for the leo i group , as recently derived by tonry _
et al . _
( 1997 ) based on the cepheid distance scale .
et al . _
( 1997 ) obtained a direct determination of the distance ( @xmath83 mpc @xmath84 ( random errors ) @xmath85 ( systematic errors ) ) to ngc 3379 from the tip of the red giant branch ( trgb ) , but the systematic errors are relatively large .
the spiral galaxy m 96 interacts with a giant 200 kpc diameter hi ring ( @xcite ) orbiting the two central galaxies in the leo i group , the e1 galaxy ngc 3379 and the s0 galaxy ngc 3384 .
thus the distances to ngc 3379 and m 96 are tied together by a physical connection .
et al . _
( 1995 ) derived a cepheid distance ( @xmath86 mpc ) to m 96 which is very similar to the trgb distance obtained for ngc 3379 . a significantly shorter cepheid distance ( @xmath87 mpc )
has been derived for m 95 ( @xcite ) , but it is not yet clear whether m 95 is at the same distance as ngc 3379 . the prototype of a normal elliptical galaxy , ngc 3379 , has both a well determined radial @xmath47 gradient @xmath88 where @xmath89 ( @xcite ) , and reliable measurements of the @xmath0band fluctuation magnitude @xmath1 at four radial distances between @xmath90 and 485 ( ( * ? ? ?
* table 5 ) ) .
this is important because @xmath1 is very sensitive to radial differences in metallicity and/or age of the stellar population ( @xcite ) as indicated by differences in , e.g. , the @xmath47 index . as shown by a plot of @xmath1 against ( @xmath91 ) ( ( * ? ? ?
7 ) ) the innermost point , being the most difficult to measure , deviates significantly from the expected linear relation . using a linear regression on @xmath1 as a function of @xmath47 , excluding the innermost point
, we find that @xmath92 for the stellar population in ngc 3379 .
the assigned error on the constant term mainly reflects the photometric zero point error .
this formula should only be used to interpolate @xmath1 in the metallicity range @xmath93 .
fortunately , thomsen & baum ( 1987 ) have measured the magnesium gradient for ngc 4881 to be @xmath94 mag arcsec@xmath95 . combining this gradient with dressler s ( 1984 ) measurement @xmath96 of the @xmath47 index within a @xmath97 region centered on the nucleus of ngc 4881 we obtain the following relation : @xmath98 , where the mean of @xmath99 for a disk with the same area as dressler s slit is @xmath100 , and the mean of @xmath99 for the annulus between @xmath101 and @xmath102 is @xmath103 .
our best estimate of the @xmath47 index for the sbf annulus in ngc 4881 is thus @xmath104 . using our relation between @xmath1 and @xmath47 , a k correction of @xmath105 , and a galactic extinction of @xmath106 ( @xcite ) , we find that a stellar population in ngc 3379 with the same @xmath47 index as the one in our sbf annulus in ngc 4881 has a fluctuation magnitude of @xmath107 . based on our sbf measurements the best estimate of the relative distance modulus between ngc 3379 and ngc 4881 is @xmath108 ( random errors only ) , corresponding to a ( luminosity ) distance ratio of @xmath109 . using our adopted distance to ngc 3379 , @xmath81 mpc ,
we finally obtain a distance modulus for the coma cluster of @xmath110 ( random errors only ) , corresponding to a luminosity distance of @xmath111 mpc .
at small redshifts the distance modulus is given by @xmath112 , where @xmath113 ( @xcite ) , @xmath114 is the redshift of the coma cluster with respect to the cosmic background radiation ( @xcite ) , and @xmath7 is the hubble constant in km s@xmath5 mpc@xmath5 . if we assume that @xmath115 then @xmath116 .
we adopt @xmath117 km s@xmath5 ( @xcite ) as likely values for the recession and peculiar velocities of the coma cluster .
finally , using these values and our distance modulus for the coma cluster , @xmath110 , we estimate the hubble constant to be @xmath6 km s@xmath5 mpc@xmath5 .
ajhar , e. a. , & tonry , j. l. 1994 , , 429 , 557 baum , w. a. , _ et al .
_ 1995 , , 110 , 2537 baum , w. a. , _ et al .
_ 1997 , , 113 , in press bessell , m. s. 1990 , , 102 , 1181 branch , d. , romanishin , w. , & baron , e. 1996 , , 465 , 73 burstein , d. , & heiles , c. 1984 , , 54 , 33 davies , r. l. , sadler , e. m. , & peletier , r. f. 1993 , , 262 , 650 de vaucouleurs , g. , _ et al .
_ 1991 , third reference catalogue of bright galaxies ( springer ) dressler , a. 1984 , , 281 , 512 freedman , w. l. , _ et al .
_ 1994 , , 371 , 757 graham , j. a. , _ et al .
_ 1997 , , 477 , 535 hamuy , m. , _ et al .
_ 1996 , , 112 , 2398 han , m. , & mould , j. r. 1992 , , 396 , 453 heckmann , o. 1942 , in _ theorien der kosmologie _ ( berlin : springer verlag ) , 63 hjorth , j. , & tanvir , n. r. 1997 , , 482 , in press holtzman , j. a. , _ et al .
_ 1995 , , 107 , 1065 jacoby , g. h. , _ et al . _ 1992 , , 104 , 599 jensen , j. b. , luppino , g. a. , & tonry , j. l. 1996 , , 468 , 519 jrgensen , i. , franx , m. , & kjrgaard , p. 1996
, , 280 , 167 kennicutt jr . , r. c. , freedman , w. l. , & mould , j. r. 1995 , , 110 , 1476 krist , j. , & hasan , h. 1993 , in _ astronomical data analysis software and systems ii _ , asp conference series 52 , ed .
r. j. hanisch , r. j. v. brissenden , & j. barnes , 530 lilly , s. j. , cowie , l. l. , & gardner , j. p. 1991
, , 369 , 79 luppino , g. a. , & tonry , j. l. 1993 , , 410 , 81 mould , j. , _ et al .
_ 1995 , , 449 , 413 pahre , m. a. , & mould , j. r. 1994 , , 433 , 567 press , w. h. , _ et al . _ 1992 , in numerical recipes in c.
( cambridge university press ) , 623 riess , a. g. , press , w. h. , & kirshner , r. p. 1995
, , 438 , l17 riess , a. g. , press , w. h. , & kirshner , r. p. 1996
, , 473 , 88 ruiz lapuente , p. 1996
, , 465 , l83 saha , a. , _ et al .
_ 1995 , , 438 , 8 saha , a. , _ et al .
_ 1996 , , 466 , 55 sakai , s. , _ et al .
_ 1997 , , 478 , 49 sandage , a. , _ et al .
_ 1996 , , 460 , l15 schaefer , b. e. 1996a , , 460 , l19 schaefer , b. e. 1996b , , 111 , 1668 schneider , s. 1989 , , 343 , 94 sodemann , m. , & thomsen , b. 1995 , , 110 , 179 sodemann , m. , & thomsen , b. 1996 , , 111 , 208 tammann , g. a. , & sandage , a. 1995 , , 452 , 16 tanvir , n. r. , _ et al .
_ 1995 , , 377 , 27 thomsen , b. , & baum , w. a. 1987 , , 315 , 460 tonry , j. l. 1991 , , 373 , l1 tonry , j. l. , ajhar , e. a. , & luppino , g. a. 1989 , , 346 , l57 tonry , j. l. , ajhar , e. a. , & luppino , g. a. 1990 , , 100 , 1416 tonry , j. l. , & schneider , d. p. 1988
, , 96 , 807 tonry , j. l. , blakeslee , j. p. , ajhar , e. a. , & dressler , a. 1997 , , 475 , 399 tyson , j. a. 1988 , aj , 96 , 1 whitmore , b. c. 1996 , in _ the extragalactic distance scale _ , stsci preprint no .
1093 worthey , g. 1994 , , 95 , 107 yasuda , n. , & okamura , s. 1996 , , 469 , l73 | we report on the first determination of the distance to the coma cluster based on surface brightness fluctuation ( sbf ) measurements obtained from _ hubble space telescope _ wfpc2 observations of the bright e0 galaxy ngc 4881 in the coma cluster and ground based observations of the ` standard ' e1 galaxy ngc 3379 in the leo i group .
relative distances based on the @xmath0band fluctuation magnitude , @xmath1 , are strongly dependent on metallicity and age of the stellar population .
however , the radial changes in the stellar populations of the two giant ellipticals , ngc 3379 and ngc 4881 , are well described by published mg@xmath2 gradients , and the ground based measurements of @xmath1 at several radial points in ngc 3379 are used to calibrate @xmath1 in terms of the mg@xmath2 index . the distance to ngc 3379 , assumed to be identical to the average sbf distance of the leo i group , is combined with the new sbf measurements of ngc 4881 to obtain a coma cluster distance of @xmath3 mpc . combining this distance with the cosmic recession velocity of coma ( @xmath4 km s@xmath5 )
, we find the hubble constant to be @xmath6 km s@xmath5 mpc@xmath5 . |
regge calculus@xcite has been proposed as an approach to classical and quantum general relativity .
it consists in approximating space - time by a simplicial decomposition .
the fundamental variables of the theory are the lengths of the edges of the simplices .
this approach has been demonstrated in numerical simulations of classical general relativity and also has inspired attractive ideas for the quantization of gravity .
for instance an extension of this framework led to the successful quantization of @xmath0 dimensional euclidean gravity through the ponzano
regge model@xcite , which can also be seen as one of the key motivations for the `` spin - foam '' approaches to @xmath1 dimensional quantum gravity .
there has been quite a bit of work devoted over the years to regge calculus , for a recent review including related formulations see loll@xcite , and for a earlier pedagogical presentations see misner , thorne and wheeler@xcite .
a canonical formulation for regge calculus has nevertheless , remained elusive ( for a review see williams and tuckey@xcite ) .
we have recently introduced a methodology to treat discrete constrained theories in a canonical fashion , which has been usually called `` consistent discretizations '' .
the purpose of this paper is to show that this methodology can be successfully applied to regge calculus without any need for modifications of the regge action .
the resulting theory is a proper canonical theory that is consistent , in the sense that all its equations can be solved simultaneously .
as is usually the case in `` consistent discretizations '' the theory is constraint - free ( although as is usual in regge calculus there are triangle inequalities to be satisfied among the variables ) .
we will see that the treatment can be applied in both the euclidean and lorentzian case . in the latter case
there is an added bonus : in order to have a well defined canonical structure one naturally eliminates `` spikes '' that have been a problem in regge formulations in the past@xcite at the time of considering the continuum limit .
this is due to the fact that our simplices only have one time - like hinge .
it is therefore not possible to construct simplices with infinitesimal volume and arbitrary length .
if one lengthens the time - like hinge one necessarily has to lengthen the space - like hinges and therefore increase the volume .
therefore one will not see the quantum amplitude dominated by long simplices of vanishing volume .
to make the calculations and illustrations simpler , we will concentrate on three dimensional gravity , but the reader will readily notice that there is no obstruction to applying the same reasonings in @xmath1 dimensions . given a simplicial approximation to a three dimensional manifold , one can approximate the einstein action ( with a cosmological term ) , as a sum over the edges ( `` hinges '' ) of the decomposition plus a sum over the simplices@xcite , @xmath2 where the first sum is over all hinges and the second over all simplices , @xmath3 is the length of the hinge @xmath4 and @xmath5 is the deficit angle around the hinge , i.e. @xmath6 where @xmath7 is the angle formed by the two faces of the simplices @xmath8 that end in the hinge @xmath4 .
@xmath9 is the volume of the simplex @xmath10 ( in our three dimensional case , a tetrahedron ) .
the constants @xmath11 and @xmath12 are related to newton s constant and the cosmological constant . a more explicit expression ( see for instance david@xcite ) can be given involving the values of the volumes of the two ( in three dimensions ) faces which share the hinge @xmath4 , @xmath13 , where @xmath14 and @xmath15 are the areas of the two triangles adjacent to @xmath4 in the simplex @xmath8 .
this in turn can be used to give an expression that is purely a function of the lengths of the hinges , using the cayley
menger determinants .
we do not quote its explicit expression for brevity . in order to have a formulation that is amenable to a canonical treatment that is uniform , in the sense that one has the same treatment at all points on the lattice
, one needs to make certain assumptions about the regularity of the simplicial decomposition chosen .
this requirement can be somewhat relaxed and our method still applies , but in a first approach we will consider a regular decomposition as shown in figure 1 .
we have divided space - time in prisms ( 1 and 2 in the figure , for example ) , and each prism in turn can be decomposed into three tetrahedra ( in the case of prism 2 the tetrahedra would be given by vertices @xmath16 , @xmath17,@xmath18 ) .
are assigned to the hinges in the following way : @xmath19 , @xmath20 , @xmath21 , @xmath22 , @xmath23 , @xmath24 @xmath25.,title="fig:",height=124 ] are assigned to the hinges in the following way : @xmath19 , @xmath20 , @xmath21 , @xmath22 , @xmath23 , @xmath24 @xmath25.,title="fig:",height=124 ] are assigned to the hinges in the following way : @xmath19 , @xmath20 , @xmath21 , @xmath22 , @xmath23 , @xmath24 @xmath25.,title="fig:",height=124 ] to construct a lagrangian picture for the previous action we consider two generic `` instants of time '' @xmath26 and @xmath27 , as indicated by the direction labeled @xmath26 in figure 1 .
we wish to construct an action of the form @xmath28 where the lagrangian @xmath29 depends on variables only at instants @xmath26 and @xmath27 .
we choose one of the fundamental cubes ( union of prisms 1 and 2 in the figure ) , choose a conventional vertex in the cube labeled by @xmath30 in the lattice .
notice that the use of the cubes is just for convenience , the framework is based on prisms that have a triangular spatial basis and therefore can tile any bidimensional spatial manifold .
the variables we will consider are the lengths @xmath31 emanating from the vertex , as designated in the figure . a similar construction is repeated for each fundamental cube .
the lagrangian that reproduces the regge action is given by a function @xmath32 that includes step functions that enforce the triangle inequalities between the hinge length variables .
up to now we have kept the discussion generic , but we should now make things more precise , in dealing with either the euclidean or the lorentzian case . in the former ,
all angles and quantities involved are real . in the lorentzian case
, angles can become complex . moreover lengths can be time - like or space - like .
null intervals can also be considered , but make the formulas more complicated , so for simplicity we do not consider them here .
we will take all lengths as positive numbers , irrespective of the space - like or time - like character of the underlying hinge . in the above construction
we have chosen the decomposition in such a way that the hinge @xmath33 is time - like and all other hinges are space - like .
the formulas presented above ( for the angles , for instance ) are valid in both the euclidean and lorentzian case , but in the latter volumes , areas and length may have to be considered as imaginary numbers .
all volumes involving a time - like direction are real , and in the construction these are the only ones involved .
areas are imaginary if they involve one time - like direction and real if they do not .
lengths are real if they are time - like and purely imaginary if they are space - like . with these conventions , dihedral
angles around time - like directions are real ( for instance around @xmath33 ) , and dihedral angles around space - like directions are complex
. some can be purely imaginary ( for instance rotation around @xmath34 in tetrahedron @xmath35 ) which correspond to lorentz boosts , or complex ( for instance rotation around @xmath36 in the aforementioned tetrahedron ) which does not correspond to a lorentz transformation ( it traverses the light cone ) .
there is one further point to consider . in the expression for the deficit angle
the term @xmath37 is present for hinges that span from the base of the elementary cube to the top cover .
for hinges lying entirely within the base or the top cover the term is @xmath38 . with these conventions the lagrangian @xmath29 turns out to be real and the sum yields the correct action avoiding over counting . for a more detailed discussion of angles in the lorentzian case see sorkin@xcite .
we now proceed to treat this action with the `` consistent discretization '' approach .
we consider as configuration variables @xmath31 and define their canonical momenta , @xmath39 here one is faced with several constraints .
notice that variables @xmath40 are `` lagrange multipliers '' since the lagrangian does not depend on their value at instant @xmath27 and therefore their canonical momenta vanish .
the @xmath41 only depend on links at level @xmath26 and therefore are constraints among the variables .
the system of equations determines variables @xmath40 and their momenta in terms of the other variables so they can be eliminated .
the resulting canonical pairs are @xmath42 @xmath43 .
the remaining equations are evolution equations for these variables and there are no constraints left ( in the sense of dynamical constraints , the variables are still constrained by the usual triangle inequalities ) .
the evolution equations are a true canonical transformation from the variables at instant @xmath26 to the variables at instant @xmath27 .
this canonical transformation has as generating function @xmath44 , viewed as a type 1 canonical transformation , where in the lagrangian the variables @xmath40 have been replaced via the equations that determine them .
the reader unfamiliar with the `` consistent discretization '' approach may question the legitimacy of this procedure in the sense of yielding a true canonical structure , however it was discussed@xcite how the canonical structure arises in detail through a generalization of the dirac procedure for discrete systems .
this concludes the classical discussion . we have reduced the regge formulation to a well defined , unconstrained canonical system where the discrete time evolution is implemented as a canonical transformation .
some of the original dynamical variables are eliminated from the formulation using the equations of motion . in the usual `` consistent discretizations '' the variables that are eliminated are the lagrange multipliers . here
the links that get determined can be viewed as playing a similar role .
the equations that determine these variables are a complicated non - linear system . as in the usual `` consistent discretization '' approach ,
one may have concerns that the solutions of the non - linear system could fail to be real , or could become unstable .
we now have experience with consistent discretizations of mechanical systems and of gowdy cosmologies , which have field theoretic degrees of freedom and the evidence suggests that one can approximate the continuum theory well in spite of potential complex solutions and multi - valued branches@xcite . we expect a similar picture to occur in regge calculus .
turning our attention to quantization , as usual in the `` consistent discretization '' approach , the hard conceptual issues are sidestepped since the theory is constraint - free . the task at hand is to implement the canonical transformation that yields the discrete time evolution as a unitary quantum operator that implements the discrete classical equations of motion as quantum operatorial equations .
this will in a generic situation be computationally intensive , but conceptually clear .
it should be noted that the resulting unitary operator differs significantly from the ones that have been historically proposed in path integral approaches based on regge calculus .
the usual approach to a path integral would be to compute , @xmath45 with @xmath46 a measure that presumably should enforce the constraints of the theory .
on the other hand , in our approach one would have something like @xmath47 where @xmath48 is obtained by substituting in @xmath49 the values of the `` lagrange multipliers '' @xmath50 obtained from their equations of motion .
@xmath46 is uniquely determined when one determines the unitary transformation that implements the dynamics ( examples of this in cosmological situations can be seen in our paper@xcite ) .
so we see that we have eliminated some of the variables and the constraints of the theory and the path integral is uniquely defined by the consistent discretization approach .
some concerns might be raised about the limitations imposed on our framework by the choice of initial lattice .
we have chosen to use a lattice that is topologically cubical .
this sets a well defined framework in which to construct a lagrangian evolution between two spatial hypersurfaces .
the cubic lattice is not a strict requirement
. it would be enough to have two `` close in time '' space - like hypersurfaces with the same simplicial decomposition in both for us to be able to set up our framework and start evolving .
this can encompass quite a range of geometrical situations .
it is however , inevitable that one should give up some arbitrariness in the space - time simplicial decomposition if one wishes to have a canonical structure .
it is interesting that the structure imposed is such that it automatically eliminates the `` spikes '' ( thin simplices arbitrarily large in the time - like direction ) .
it also worthwhile emphasizing that the framework can , with relatively simple additions , incorporate topology change .
the idea is depicted in figure 2 . there
we see a point where there is topology change where the legs of the pair of pants separate . for that to happen one would need to modify at the hypersurface the explicit form of the lagrangian @xmath29 .
it is interesting that from there on one can continue without further altering the framework and that at all times the number of variables involved has not changed .
the picture also shows how one would handle an initial `` no boundary '' type singularity . here
one would have to `` by hand '' add links as the time evolution progresses forward .
these variables are free as long as one does not wish to match some final end state for the evolution .
if one has , however , specified initial and final data for the evolution , one finds that constraints appear that determine the values of the lengths of the extra links added . as an example of the framework
, one can work out explicitly the evolution of a a @xmath0 dimensional space - time consistent of a four adjacent `` unit cubes '' of the type we have considered with fixed outer boundary conditions . in this case the initial data consists of eight lengths , the other initial lengths are determined by the boundary conditions . as we discussed , having the data at level @xmath26 and @xmath27 one can determine the `` lagrange multiplier '' links that have to be substituted in the lagrangian to generate the canonical transformation between the initial and final data .
this transformation is later to be implemented unitarily upon quantization . in our formalism
it can happen that the `` lagrange multipliers '' are not entirely determined by this procedure ( for other examples where this happens , including bf theory , see our paper@xcite ) .
in such case the resulting theory has true constraints and true lagrange multipliers . in this example
this happens .
one is finally left with a canonical transformation dependent on 3 parameters ( this result is also true if one considers an @xmath51 adjacent unit cube system ) .
the presence of free parameters also requires modifications in the path - integral formulas listed above as well .
it should be emphasized that the equations determining the lagrange multipliers , even in this simplified case , are complicated coupled non - linear equations that have a complexity not unlike those in @xmath1 dimensions .
what makes them easy to solve is the knowledge that the regge equations of motion correspond in this case to flat space - time .
the canonical transformation can be implemented unitarily and the quantization completed .
we will discuss the example in detail in a separate publication .
we have applied the `` consistent discretization '' approach to regge calculus .
we see that it leads to a well defined constraint - free canonical formulation , that is well suited for quantization .
the approach can incorporate topology change .
although we have limited the equations to the three dimensional case for simplicity , we have never used any of the special properties of three dimensional gravity and it is clear that the construction can be carried out in an arbitrary number of dimensions .
it is interesting to notice that one of the original motivations for the construction of the `` consistent discretization '' approach was the observation by friedman and jack@xcite that in canonical regge calculus the lagrange multipliers failed to be free .
it can be viewed as if this point of view has now been exploited to its fullest potential , offering a well defined computational avenue to handle classical and quantum gravity .
we dedicate this paper to rafael sorkin on the occassion of his 60th birthday . we wish to thank luis lehner and jos a. zapata for comments .
this work was supported by grant nsf - phy0244335 , nasa - nag5 - 13430 and funds from the horace hearne jr .
laboratory for theoretical physics and cct - lsu .
00 t. regge , phys . rev . * 108 * 558 ( 1961 ) . g. ponzano and t. regge , `` semiclassical limit of racah coefficients '' in `` spectroscopic and group theoretical methods in physics '' , ed . f. bloch , north holland , amsterdam , ( 1968 ) .
r. loll , living rev .
* 1 * , 13 ( 1998 ) [ arxiv : gr - qc/9805049 ] . c. misner , k. thorne , j. wheeler , `` gravitation '' , w. h. freeman , new york ( 1973 ) .
p. a. tuckey and r. m. williams , class .
* 7 * , 2055 ( 1990 ) .
j. ambjorn , j. l. nielsen , j. rolf and g. k. savvidy , class . quant
. grav .
* 14 * , 3225 ( 1997 ) [ arxiv : gr - qc/9704079 ] .
f. david , `` simplicial quantum gravity and random lattices '' in `` gravitation and quantizations : proceedings : les houches summer school on gravitation and quantizations '' , j. zinn - justin and b. julia ( editors ) , north - holland , amsterdam ( 1995 ) [ arxiv : hep - th/9303127 ] .
r. sorkin , phys .
* d12 * , 385 ( 1975 ) ; see also j. barret et al .
phys * 36 * , 815 ( 1997 ) [ arxiv : gr - qc/9411008 ] . c. di bartolo , r. gambini , r. porto and j. pullin , j.math .
phys . * 46 * , 012901 ( 2005 ) [ arxiv : gr - qc/0405131 ] .
r. gambini , m. ponce and j. pullin , phys .
d * 72 * , 024031 ( 2005 ) [ arxiv : gr - qc/0505043 ] .
r. gambini and j. pullin , class .
grav . * 20 * , 3341 ( 2003 ) [ arxiv : gr - qc/0212033 ] . c. di bartolo , r. gambini and j. pullin , class .
* 19 * , 5275 ( 2002 ) [ arxiv : gr - qc/0205123 ] .
j. l. friedman and i. jack , j. math .
* 27 * , 2973 ( 1986 ) . | we apply the `` consistent discretization '' technique to the regge action for ( euclidean and lorentzian ) general relativity in arbitrary number of dimensions .
the result is a well defined canonical theory that is free of constraints and where the dynamics is implemented as a canonical transformation . in the lorentzian case
, the framework appears to be naturally free of the `` spikes '' that plague traditional formulations .
it also provides a well defined recipe for determining the integration measure for quantum regge calculus . |
when a porous medium is plunged in an infinite fluid reservoir , the solid matrix absorbs fluid until an equilibrium state is reached . many interesting features of this swelling phenomenon have been demonstrated experimentally .
the amount of swelled fluid can be controlled via different external parameters such as the fluid pressure in the reservoir @xcite , that is via its chemical potential , the fluid velocity @xcite , or a mechanical pressure exerted on the solid @xcite .
the solid fluid segregation in consolidation is seen when , depending on the external pressure acting on the porous material , phases differing in fluid content are observed .
this problem has been addressed@xcite by the authors in the framework of continuum mechanics adopting in particular a first gradient model ; the existence of two different phases depending on the external pressure has been proved . in that context the interesting question of the coexistence of the two phases could not be posed for the first gradient nature of the model . in this paper
we propose a more general two field one dimensional second gradient model to study the profiles connecting two coexisting phases and the formation of critical droplets , if any , of one phase into the other .
the adopted approach in modeling the behavior of porous continua is essentially based on a pure solid lagrangian description of motion , referring kinematics to the reference configuration of the porous skeleton ( see section [ s : setup ] ) .
the constitutive model is purely phenomenological , which means that the overall potential energy , regarded as a function of the strain of the skeleton and the fluid mass density ( per unit volume , in the solid reference configuration ) , is built up in such a way to describe the existence of two states of equilibrium : the solid rich and the fluid rich phase .
thus no refined description of solid grain connectivity , as well as connection among regions with different porosity is available in such a model .
conversely the constitutive state parameters are selected so as to describe the showing up of the fluid rich phase , which is possibly associated to the occurrence of fluid segregation .
let @xmath0 be the _ reference _ configurations for the solid and fluid components @xcite .
the _ solid placement _ is a @xmath1diffeomorphism @xmath2 such that @xmath3 is the position occupied at time @xmath4 by the solid particle @xmath5 in the reference configuration @xmath6 .
consider @xcite @xmath7 such that @xmath8 is the fluid particle which at time @xmath4 occupies the same position of the solid particle @xmath5 .
assume also @xmath9 to be a @xmath1diffeomorphism , thus the map @xmath9 associate univocally a solid particle to a fluid one and vice versa .
the _ fluid placement _ map
@xmath10 , giving the position of a fluid particle @xmath11 , is defined as @xmath12 . the _ current configuration _
@xmath13 at time @xmath4 is the set of positions of the superposed solid and fluid particles .
let @xmath14 be the jacobian of the placement map @xmath15 measuring the ratio between current and reference volumes of the solid component ; we let @xmath16 be the _ strain field_. let @xmath17 with @xmath18 , be the solid and fluid reference _ densities _ ; we define the _ fluid mass density _ field @xmath19 . assuming that the mass is conserved , it is not difficult to prove @xcite that the field @xmath20 can be interpreted as the fluid mass density measured w.r.t .
( with respect to ) the solid reference volume .
assume , now , that the _ lagrangian density _ @xmath21 of the system is in the form @xmath22 where @xmath23 is the _ kinetic energy density _ and
@xmath24 is the _ overall potential energy density _
accounting for both the internal and the external forces . in ( [ lag ] ) we have denoted with the dot the derivative taken w.r.t .
time and with the prime the derivative w.r.t . the solid reference space variable .
the equation of motion for the two fields @xmath25 and @xmath26 can be derived assuming that the possible motions of the system in an interval of time @xmath27 are those such that the fields @xmath25 and @xmath26 are extremals for the _ action functional _ @xmath28 in correspondence of the independent variations of the two fields @xmath25 and @xmath26 on @xmath29 . in other words
any possible motion of the system in the considered interval is a solution of the euler
lagrange equations associated to the variational principle @xmath30 .
if one is interested to find equilibrium profiles @xmath31 and @xmath32 of the system , namely , the solutions of the equations of motion independent of time , since the kinetic energy associated to those profiles is equal to zero , the lagrangian density reduces to minus the potential energy density . in this case
the action is given by ( [ azione ] ) where the time integral gives a not essential multiplicative constant , and the variational principle associated to such an action gives the seeked for equilibrium profiles .
assume , now , that the effect of the internal forces exchanged by the solid and fluid particles and that of the conservative external fields can be described via a potential energy density @xmath33 depending on the kinematic fields @xmath25 and @xmath26 only through the strain and the fluid mass density fields . note that the strain depends only on @xmath25 and the fluid mass density only on @xmath26 , hence the independent variations of those primitive fields reflect on independent variations of @xmath34 and @xmath20 .
thus , limiting the study to boundary value problems expressed in terms of the fields @xmath34 and @xmath20 , we can treat the fields @xmath34 and @xmath20 as primitive , consider their independent variations and look for the equilibrium profiles @xmath35 and @xmath36 starting from the variational principle @xmath37 finally , we can derive , starting from ( [ variazionale03 ] ) , the equations governing the equilibrium profiles @xmath34 and @xmath20 . by computing the variation of the action functional on @xmath38 , with @xmath39 , we get the euler lagrange equations @xmath40 with boundary conditions ensuring that @xmath41_{\ell_1}^{\ell_2 } = 0\ ] ] where @xmath42 and @xmath43 are , respectively , the variations of the strain and fluid mass density fields . for instance
dirichelet boundary conditions would do the job , since we would have @xmath44 and @xmath45 .
but for potential energy densities @xmath24 at least quadratic in the derivatives @xmath46 and @xmath47 even neumann boundary conditions would be acceptable . in the sequel we shall refer to any solution of the euler lagrange equations ( [ moto06 ] ) with suitable boundary conditions as an _ equilibrium profile _ of the system corresponding to the chosen boundary conditions .
the model ( [ variazionale03 ] ) is called a _ first gradient _ model
if the potential energy density @xmath24 does not depend on the first derivatives of the strain and of the liquid density , otherwise the model is said a _
second gradient _ model .
this way of classifying the models is related to the fact that both @xmath34 and @xmath20 depends , by definition , on the gradient of the primitive kinetic fields @xmath25 and @xmath26 .
second gradient theories are suitable to be developed for modeling stress / strain concentration due , for instance , to the presence of geometrical singularities ( crack propagation in fracture mechanics@xcite ) or phase transitions as in the case of wetting@xcite .
in particular second gradient poromechanics has been recently formulated @xcite extending the standard arguments of the biot theory @xcite .
such a model addresses the description of those deformation phenomena which occur at the same length scale as that where high gradients in deformation can be detected .
classical poromechanics@xcite is not able to describe these phenomena : the macroscopic model is regarded in that case as the average of a microscopic one where a kind of stationarity assumption@xcite ( spatial ergodicity ) on the random field which characterizes the microscopic mechanical properties of the material has been formulated .
this allows for replacing ensemble averages with volume averages insofar as the characteristic size of the heterogeneities is much smaller than the typical length scale of the reference volume element ( rve ) .
if this is no more the case , the classical assumptions of uniform strain ( stress ) or periodic boundary conditions , for every reference volume , are no more valid , but , conversely , macroscopic strain gradient plays a crucial role in specifying the state of stress / strain inside the rve itself .
the goal , here , is to formulate a second gradient poromechanical model for describing the transition from the standard biot like equilibrium , associated to a compacted solid , versus the fluid
segregated phase describing duct thinnering in the matrix and , consequently , fluid mass concentration in the pores .
thus , we consider a model with total potential energy density in the form @xmath48 where @xmath49 is a polynomial quadratic function of @xmath46 and @xmath47 , @xmath50 is a differentiable function positively diverging along any radial direction in the plane @xmath34@xmath20 , having at least a local minimum , and whose stationary points are isolated . since @xmath49 is quadratic , we have that a constant solution of the euler lagrange problem ( [ moto06 ] ) must necessarily satisfy the equations @xmath51 and @xmath52 ; in other words a constant profile must be constantly equal to an extremal point of the first gradient part @xmath50 of the total potential energy .
we then let a _
phase _ of the model to be a constant equilibrium profile equal to one of the local minima of the function @xmath50 .
note that the euler lagrange problem ( [ moto06 ] ) and ( [ moto07 ] ) for the first gradient model associated to ( [ modello ] ) , namely the one obtained for @xmath53 , is the system of algebraic equations @xmath52 and @xmath51 .
since the stationary points of the two variable function @xmath50 are isolated , we have that the equilibrium profiles for such a model are necessarily constant functions of @xmath54 equal to the values of the stationary points of @xmath50 .
hence , in the case of a first gradient model it is not possible to discuss phase coexistence , since there exist only continuous constant equilibrium profiles . on the other hand , in second gradient models ,
different ( not constant ) equilibrium profiles can exist .
this fact allows us to pose the problem of the coexistence of two existing phases .
suppose that the model exhibits the two phases @xmath55 and @xmath56 ; a _ connection _
@xcite between those phases is an equilibrium profile @xmath57 of the action functional on @xmath58 satisfying the boundary conditions @xmath59 , @xmath60 , @xmath61 , and @xmath62 .
we say that the two considered phases _ coexist _ if and only if a connection does exist .
we study , now , a particular poroelastic model and in that framework we discuss the existence of the consolidation phase transition and prove the coexistence of the two phases for a particular value of the external pressure .
more precisely we consider the poroelastic system with overall potential energy density ( [ modello ] ) with @xmath63\ ] ] with @xmath64 , @xmath65 such that @xmath66 , and @xmath67 where @xmath68 is the biot potential energy density @xcite , @xmath69 is the ratio between the fluid and the solid rigidity , @xmath70 is a coupling between the fluid and the solid component , @xmath71 is the external pressure , and @xmath72 is a material parameter responsible for the showing up of the additional equilibrium .
we remark that the condition @xmath66 ensures that the second gradient part @xmath49 of the overall potential energy density is convex . under this assumption
there exists a minimizer for the action functional @xmath73 on a bounded domain . as we will see later to ensure the existence of a connection profile , which is a dirichelet problem on an unbounded domain
, it will be necessary to assume @xmath74 ; in the limiting case @xmath75 the existence of the connection will depend on the choice of the parameter @xmath76 , @xmath77 , and @xmath78 .
we have already studied @xcite the associated first gradient model with overall potential energy density @xmath50 and we have proven the existence of a phase transition driven by the external pressure @xmath79 . more precisely it has been shown that there exists a _ critical pressure _
@xmath80 such that for @xmath81 the system admits the single _ standard _ phase @xmath82 , while a second _ fluid rich _ phase @xmath83 , appears for @xmath84 .
the standard phase is similar to the unique phase described by the model with potential energy density@xcite @xmath85 . in fig .
[ f : fig - emme - eps ] the standard and the fluid
rich phases are depicted for @xmath86 and for a particular choice of the physical parameters @xmath87 .
( 200,150 ) ( -20,0 )
the second gradient model has the same phases as the associated first gradient model .
the main result of this paper is the existence of @xmath88 , called _ coexistence pressure _ , such that the standard and the fluid
rich phases coexist at the pressure @xmath89 and do not coexist at @xmath84 and @xmath90 .
the proof will be achived in two steps : first we shall show that there exist a unique value of the pressure such that the total potential energy densities evaluated at the two phases are equal ; the second step will consist in proving the existence of the connection , that is the equilibrium profile connecting the two phases .
we first review some of the results in the previous paper @xcite ; there we have studied the equations @xmath51 and @xmath52 looking for the minima of the function @xmath50 . we have shown that the standard phase @xmath82 is the solution of the two equations @xmath91 and @xmath92 , for any @xmath71 , where @xmath93 . on the other hand the fluid rich phase @xmath83 is the solution , with the smallest value of @xmath34 , of the two equations @xmath94 and @xmath95 , where @xmath96\ ] ] and @xmath97 -\alpha b^2\varepsilon m_+^2(\varepsilon ) \vphantom{\bigg\ { } + \frac{2}{3}\alpha bm_+^3(\varepsilon)\ ] ] for @xmath98 the function @xmath99 is positive , diverging to @xmath100 for @xmath101 , and has a minimum at @xmath102 such that @xmath103 ; this explains why the fludized phase is seen only for @xmath84 .
moreover it has been proven that for any @xmath84 the point @xmath83 is a minimum of the two variable potential energy @xmath104 with @xmath79 fixed , while it is a saddle point for @xmath105 .
we now prove the first step of the above stated coexistence result . for any @xmath84 , we let @xmath106 and @xmath107 and prove that @xmath108 that is the overall potential energy density of the standard and the fluid
rich phases are equal only at the coexistence pressure .
this statement has been tested on numerical grounds , see fig .
[ f : ene1 ] where the graphs of the functions @xmath109 and @xmath110 are depicted for a given set of physical parameters .
( 200,150 ) ( -20,0 ) in order to prove ( [ coes00 ] ) we first compute the derivative of the two functions @xmath109 and @xmath110 ( with respect to @xmath79 ) ; by using ( [ secondo010 ] ) , the chain rule , and the fact that @xmath82 and @xmath83 are solutions of the equations @xmath111 and @xmath112 , we have that @xmath113 , @xmath114 , @xmath115 , and @xmath116 .
now , since @xmath117 and @xmath118 are negative functions of the pressure , we have that both @xmath109 and @xmath110 are decreasing functions of the pressure on the interval @xmath119 .
moreover , noted that both @xmath120 and @xmath121 are decreasing functions ( of the strain ) on @xmath122 $ ] , we have that @xmath117 and @xmath118 decrease when @xmath79 increases .
it then follows that @xmath123 and @xmath124 are negative and therefore @xmath109 and @xmath110 are concave on the interval @xmath119 .
since the two functions @xmath109 and @xmath110 are decreasing concave functions on the interval @xmath119 , in order to prove ( [ coes00 ] ) it is sufficient to show that @xmath125 and @xmath126 for some @xmath79 sufficiently large .
the proof of the first remark is easy : at @xmath105 the two variable function @xmath127 has just the two stationary points@xcite @xmath128 and @xmath129 . since @xmath128 is a local minimum of @xmath127 , which tends to @xmath100 along every direction on the plane @xmath20@xmath34 , the single local minimum must be the absolute minimum ; hence , @xmath125 .
the second remark follows from the asymptotic behavior of the two functions @xmath109 and @xmath110 ; as proven in the appendix [ s : asymptotic ] , for @xmath130 we have @xmath131 and @xmath132 by comparing the two asymptotic formulas ( [ coes50 ] ) and ( [ coes60 ] ) we get immediately that for @xmath79 large enough @xmath126 .
it is worth remarking that the variational problem ( [ variazionale03 ] ) for profiles with fixed values at the end points @xmath133 and @xmath134 of the interval @xmath6 , is nothing but the hamilton principle for a two degree of freedom mechanical system with lagrangian coordinates @xmath34 and @xmath20 , kinetic energy @xmath23 and potential energy @xmath135 respectively given by @xmath136\ ] ] and @xmath137 and the space variable @xmath5 interpreted as time . in other words the function @xmath24 defined by ( [ modello ] ) , ( [ secondo000 ] ) , and ( [ secondo010 ] ) is the lagrangian for such a two degree of freedom equivalent mechanical system .
it is important to remark that the mechanical interpretation is correct only when @xmath23 is a positive definite quadratic form .
it is easy to prove that this is the case provided @xmath74 . in the limiting case
@xmath75 the form @xmath23 is positive semidefinite , indeed if we substitute @xmath138 in ( [ secondo02 ] ) the function @xmath23 becomes @xmath139 where we have set @xmath140 , and is equal to zero when @xmath141 .
we study now the case @xmath74 and postpone the degenerate @xmath75 to the following section .
let us denote @xmath5 by @xmath4 and the derivative taken with respect to @xmath4 by the dot . by using ( [ moto06 ] ) with @xmath142 and recalling ( [ secondo02 ] )
, we have that the equations of motion are @xmath143 we note that the mechanical energy of the associated mechanical problem @xmath144 is a constant of the motion . first note that the two points @xmath82 and @xmath83 , with @xmath84 , are maxima of the potential energy @xmath145 of the equivalent mechanical system .
the problem of the existence of a connection between the standard and the fluid - rich phase can be rephrased as follows : look for a solution of the equations ( [ motoeq ] ) , namely , a motion @xmath146 of the equivalent mechanical system , on @xmath147 connecting the phase space point @xmath148 and @xmath149 to the phase space point @xmath150 and @xmath151 .
the connection we are seeking for is an heteroclinic solution of the equation of motion tending to two fixed points in the phase space for @xmath152 and @xmath153 .
recall that the mechanical energy @xmath154 is a constant of the motion and remark that at the equilibrium points it is equal to @xmath155 and @xmath156 . from the results in section [ s : coesistenza ]
it follows that those two energies are equal only for @xmath89 .
this remark yields that for any @xmath84 and @xmath90 the standard and the fluid
rich phases do not coexist .
we are left with the case @xmath89 . in principle
an heteroclinic solution can exist , but to prove its existence is an highly not trivial problem which has been solved , under suitable hypotheses on the potential energy , in the recent paper@xcite whose main results have been summarized in the appendix [ s : fusco ] . since in the not degenerate case the form @xmath23 is positive definite , it is possible to find an orthogonal transformation of the coordinates in the plane @xmath20@xmath34 which diagonalize the form itself .
then , performing this transformation and subctracting to the potential energy @xmath135 of the equivalent mechanical system the constant term @xmath157 , the problem of finding a connection between the standard and fluid rich phase is transformed in a problem in the form ( [ eak ] ) with @xmath158 and @xmath159 replaced by @xmath160 $ ] .
since this function satisfies the hypotheses of the theorem 3.6 by alikakos and fusco@xcite ( see the appendix [ s : fusco ] ) we can then conclude that in the case @xmath89 there exists a connection between the standard and the fluid rich phase and hence the two phases coexist .
( 200,180 ) ( 10,0 )
consider the case @xmath161 and the change of variables @xmath162 where we recall @xmath163 , which amounts to perform a rotation of the cartesian reference system in the plane @xmath20@xmath34 .
using the new variables the two functions @xmath23 and @xmath135 become respectively @xmath164 and @xmath165 where we have used ( [ secondo05 ] ) and ( [ secondo03 ] ) with @xmath166 the expression ( [ secondo08 ] ) of @xmath167 is awful , but this will not be a problem since @xmath167 is precisely the two variable functions @xmath168 , which we have already deeply studied@xcite , written via a rotation of the cartesian reference system .
we apply , now , the variational principle ( [ variazionale03 ] ) to the total poroelastic potential energy density @xmath169 and get the analogous ( indeed it is a particularization ) of the equations ( [ moto06 ] ) @xmath170 which must be solved with the boundary conditions ( [ moto07 ] ) . by using the definition of @xmath171
the above equations become @xmath172 we remark that the second of the equations above is an algebraic equation involving the two variable @xmath173 and @xmath174 ; provided it can be solved w.r.t .
@xmath174 , the first one becomes a second order ordinary differential equation in the unique unknown function @xmath173 .
more precisely , the root locus of @xmath175 is made of a certain number of maximal components such that each of them is the graph of a function @xmath176 ; for each of them the first of the two equations ( [ secondo12 ] ) becomes a standard one dimensional conservative mechanical system with potential energy @xmath177 .
( 200,180 ) ( 10,0 ) the function @xmath167 is obtained by flipping the sign of the function @xmath50 and rotating the coordinate axes .
this implies that the function @xmath167 , at @xmath89 , has the two absolute maximum points @xmath178 and @xmath179 corresponding , respectively , to the standard and to the fluid rich phases .
since @xmath82 and @xmath83 satisfy the equations @xmath180 and @xmath181 , we have that the two points @xmath182 and @xmath183 are solutions of the constraint equation @xmath175 and hence they belong to the constraint curve .
( 200,180 ) ( -20,0 ) we consider , now , the case in which at @xmath89 the two points above fall on the same maximal component of the constraint equation ( see fig .
[ f : vincolo1 ] and [ f : vincolo1xy ] ) . using the conservation of the mechanical energy of the equivalent one dimensional conservative system allows for reducing the computation of the coexistence profile ( heteroclinic ) to the evaluation of a definite integral . since the function @xmath167 has two isolated absolute maximum points which , by hypothesis , belong to the same maximal component of the constraint curve , we have that the function @xmath177 of the real function @xmath173 has two absolute isolated maxima in @xmath184 and @xmath185 ( see fig . [
f : vdix ] ) . consider the motion of the equivalent one dimensional system corresponding to the energy level @xmath186 .
the conservation of the mechanical energy implies @xmath187 hence , the heteroclinic connecting the two maxima is given by @xmath188 } } \,{{\normalfont\textrm{d}}}x'\ ] ] for any @xmath189 and for some fixed @xmath190 in the same interval . by changing @xmath190
it is found a family of heteroclinic orbits which are the same curve up to a time translation .
( 200,250 ) ( -100,0 ) ( 200,180 ) ( -20,0 ) results are depicted in the fig .
[ f : etero ] .
the @xmath34 coexistence profile shows a bump@xcite close both to the standard and the fluid rich phase .
this behavior is due to the two dimensionality of the problem : in fig .
[ f : etero2 ] we have depicted the heteroclinic on the graph of the function @xmath191 . from the picture it is clear that the optimal path climbs the two hills going around the hills themselves . in other words
the existence of the bump in the connecting @xmath34profile is due to the shape of the constraint curve in the plane @xmath20@xmath34 .
since the problem has been reduced to the computation of the heteroclinic of a one dimensional conservative mechanical system in the @xmath173 variable , it is obvious that no bump can exist in the @xmath173profile . on the other hand by looking closely at the picture in fig .
[ f : vincolo1 ] and [ f : vincolo1xy ] , it emerges that the constraint curve is monotonic w.r.t .
@xmath174 and @xmath20 ; this implies the monotonicity of the @xmath174 and @xmath20profiles .
however , it is possible to find values of the parameters such that the @xmath174profile presents a bump .
in section [ s : deghet ] we have proven that in the not degenerate case the connecting profile does exist for any proper choice of the parameters .
a similar result does not hold true in the degenerate case , indeed it is possible to find the connection if and only if the two maxima of the function @xmath135 lie on the same maximal component of the constraint curve .
we have that this is not the case for @xmath192 large enough , see the dashed curve in fig .
[ f : kpos ] which is associated to the value @xmath193 .
it is immediate to remark that the two maxima do not lie on the same connected component , hence in this case it is not possible to find a connection between the fluid rich and the standard phase .
it is worth remarking that no evidence of this patologic behavior is found in the case @xmath194 ; see fig .
[ f : kneg ] where the constraint curve and the stationary point of @xmath135 are depicted for @xmath195 , @xmath196 , @xmath197 , @xmath198 , @xmath199 , and @xmath200 .
this case is the most interesting one from the physical point of view , indeed for @xmath201 the coupling between @xmath46 and @xmath47 is negative , hence the preferred states are such that the two fields @xmath35 and @xmath36 are both increasing or decreasing .
( 200,180 ) ( 10,0 ) ( 200,180 ) ( 10,0 ) consider the degenerate model and suppose that the pressure @xmath79 is larger than @xmath202 but different from @xmath203 .
suppose that the two local minima of the potential energy @xmath50 lie on the same connected component of the constraint curve whose equation is ( [ secondo12 ] ) .
consider the function @xmath177 as in section [ s : deghet ] and note that the two local maxima are not equal .
we consider the homoclinic solution corresponding to the lowest maximum @xmath204 . in analogy with the discussion of the above section , the homoclinic equilibrium profile can be found , see fig .
[ f : omo ] , by computing the integral @xmath205 } } \,{{\normalfont\textrm{d}}}x'\ ] ] for any @xmath206 with @xmath207 the unique ( inversion ) point in the interval @xmath208 such that @xmath209 .
the homoclinic solution corresponding to the lowest maximum is often interpreted as a `` critical nucleus . '' in the sense that , if a dynamic evolution would be taken into account , one would expect that an initial condition close to the critical nucleus would evolve into the standard or the fluid rich phase ( subcritical and supercritical behavior ) .
this behavior depends on the size of the droplet in the neighborhood of @xmath210 .
indeed in @xmath210 the profile has the value @xmath207 which , for @xmath79 close to @xmath203 , is a good approximation of the phase corresponding to the largest maximum of the function @xmath167 . in other words
the critical nucleus can be seen as a droplet of the phase correpondig to the smallest value of the potential energy @xmath50 plunged into the other phase .
( 200,250 ) ( -100,0 )
in conclusion we have studied the phase transition between the fluid poor and rich phases in the context of consolidating completely fluid saturated porous media . a second gradient model to study the existence of such
a transition has been proposed .
moreover , coexistence between the two phases at the pressure @xmath203 , defined as the pressure such that the total potential energy of the two phases is the same , has been established .
we have also shown that at different values of the pressure the two phases can not coexist . for a particular choice of the parameters of the model it is possible to reduce the problem of finding the coexistence profile to the computation of a definite integral .
we have studied the coexistence profile for different values of the physical parameters of the model and shown that non monotonic interfaces exist .
the authors are grateful to g. fusco for helpful discussions and hints .
in this appendix we discuss the asymptotic behavior of the two functions @xmath109 and @xmath110 , see section [ s : coepre ] , for @xmath79 large and , in particular , prove the equations ( [ coes50 ] ) and ( [ coes60 ] ) .
we first note that by using @xmath211 , we get @xmath212 the equation @xmath213 is a cubic equation in the form @xmath214 , with @xmath215 ; by cardano s formula , since @xmath216 , there exists a single real solution given by @xmath217^{1/3 } + \big[-\frac{1}{2}\lambda p-\sqrt{d}\big]^{1/3 } \vphantom{\bigg\{_\big\ } } } \\ & \!= & \ ! { \displaystyle \big[-\frac{1}{2}\frac{3}{\alpha b^4 } p + \sqrt{\big(\frac{1}{3}\frac{3}{\alpha b^4}\big)^3 + \big(\frac{1}{2}\frac{3}{\alpha b^4 } p\big)^2}\big]^{1/3 } \vphantom{\bigg\{_\big\ } } } \\ & & \ ! { \displaystyle + \big[-\frac{1}{2}\frac{3}{\alpha b^4 } p -\sqrt{\big(\frac{1}{3}\frac{3}{\alpha b^4}\big)^3 + \big(\frac{1}{2}\frac{3}{\alpha b^4 } p\big)^2}\big]^{1/3 } } \\ \end{array}\ ] ] by using the taylor series @xmath218 , with @xmath219 being the binomial coefficient , which is convergent for @xmath220 , it is not difficult to prove that @xmath221 for @xmath79 large . by inserting this expression in the expansion for @xmath222 , we get equation ( [ coes50 ] ) . we can perform a similar computation for @xmath110 .
accounting in particular for the qualitative of @xmath118 , which tends to @xmath223 when @xmath79 is increased , we shall study the asymptotic behavior of @xmath224 for @xmath101 and that of @xmath118 which is the solution of the equation @xmath225 when @xmath130 .
the result of this analysis will provide us with the asymptotic behavior of @xmath110 .
first of all we note that @xmath226\ ] ] for @xmath101 . by using ( [ secondo010 ] )
we then have @xmath227 for @xmath130 , where we have used that @xmath228 for @xmath130 .
the function @xmath118 is implicitly defined by the equation @xmath225 which is pretty complicated . by expanding @xmath121 for @xmath101
the equation becomes @xmath229 with @xmath230 a function having limit @xmath231 for @xmath101 .
suppose @xmath79 is large enough and let @xmath118 be the solution of the equation above ; by the qualitative study we get that @xmath228 for @xmath130 .
it is then easy to show that @xmath232 $ ] tends to zero for @xmath130 , indeed , since @xmath233 is the solution of the equation above , we have that @xmath234 for @xmath130 , where we have used that @xmath235 for @xmath101 and @xmath228 for @xmath130 . by inserting the obtained expression of @xmath118 in the above expansion of @xmath110
we get equation ( [ coes60 ] ) .
in this appendix we briefly review the main results by alikakos and fusco@xcite on the existence of connections .
let @xmath236 , with @xmath237 , be a @xmath238 positive function satisfying the following hypotheses : ( 1 ) @xmath159 has two distinct local minima @xmath239 such that @xmath240 , ( 2 ) @xmath241 for any @xmath242 , ( 3 ) @xmath243 , ( 4 ) there exists @xmath244 in the open interval @xmath245 such that for any @xmath246 such that @xmath247 the two maps @xmath248 have a strictly positive derivative for every @xmath249 . conditions ( 1 ) ( 3 ) are quite natural and physically obvious ; condition ( 4 ) is a mild technical requirement allowing for potential energies with @xmath250 contact at zeroes . consider the ordinary differential equation problem @xmath251 where @xmath252 .
solutions to the problem ( [ eak ] ) are known in the literature as _ heteroclinc _ motions of the mechanical system or _ connection _
solutions in the context of phase transitions . the theorem 3.6 by alikakos and fusco@xcite states that , under the hypotheses discussed above , the problem ( [ eak ] ) admits a solution . in other words
the theorem states the existence of a connection under very general and mild requirements on the potential @xmath159 .
the proof of the theorem is based on a direct variational computation .
more precisely the authors prove the existence of a critical profile of the action functional @xmath253\,{{\normalfont\textrm{d}}}x\ ] ] on the sobolev space @xmath254 of functions @xmath252 such that @xmath255 and its weak derivative are in @xmath256 for any bounded subsets @xmath257 .
such a critical profile is the solution of the ordinary differential equation problem ( [ eak ] ) .
compared to the standard variational calculus , see for instance the paragraph 8.2 in evan s classical book@xcite , the authors have to face the lack of compactess due to the infinite domain @xmath147 on which the solution of the variational problem is defined .
this problem is overcame by using suitable costraints that are successivley removed .
it is also worth noting that in the theorem 3.7 the authors state that the connection is a minimizer of the action functional @xmath258 .
cirillo , n. ianiro , g. sciarra ,
solid fluid segregation in saturated porous media . " poromechanics iv : proceedings of the fourth biot conference on poromechanics , destech publications , inc . , june 2009 . | the appearence of the fluid rich phase in saturated porous media under the effect of an external pressure is investigated . for this purpose
we introduce a two field second gradient model allowing the complete description of the phenomenon .
we study the coexistence profile between poor and rich fluid phases and we show that for a suitable choice of the parameters non monotonic interfaces show up at coexistence . |
despite of huge advances occurred in the last 10 years , the gamma ray bursts ( grb ) phenomenon is still far to be fully understood .
open issues include : physics and geometry of the prompt emission , unexpected early afterglow phenomenology ( e.g. plateau phase , flares ) , identification and understanding of sub - classes of grbs ( short / long , x ray flashes , sub energetic grbs ) , grb / sn connection , nature of the inner engine , cosmological use of grbs , and more .
see , e.g. , @xcite for recent reviews .
the large observatory for x ray timing , loft @xcite , was selected in 2011 by the european space agency as one of the four cosmic vision m3 candidate missions to compete for a launch opportunity at the start of the 2020s .
the current schedule of the esa / m3 programme foresees the end of the assessment study by the end of 2013 and the final down selection in the first months of 2014 .
thanks to an innovative design and the development of large - area monolithic silicon drift detectors , the large area detector ( lad ) on board loft will operate in the 230 kev range ( up to 80 kev in expanded mode ) with a fov collimated to @xmath11@xmath2 , and achieve an effective area of @xmath110 m@xmath0 at 8 kev , a time resolution of @xmath110 @xmath3s and a spectral resolution of @xmath1200260 ev ( fwhm at 6 kev)@xcite
. these capabilities will allow loft to investigate with unprecedented sensitivity the rapid x ray flux and spectral variability that directly probes the motion of matter down to distances very close to black holes and neutron stars , as well as the physical state of ultra - dense matter @xcite .
such measurements are efficient diagnostics of the behavior of matter in the presence of strong gravitational fields , where the effects predicted by general relativity are more relevant , and the physics of matter at densities in excess of that in atomic nuclei , determining its equation of state and composition .
in addition to the lad , the loft payload will include a wide field monitor ( wfm)@xcite , which will monitor more than half of the lad - accessible sky ( approximately 1/3 of the whole sky ) simultaneously at any time , will operate in the same energy range as the lad with an energy resolution of @xmath1300 ev , providing information about source status ( flux variability and energy spectrum ) , as well as arc minute positioning ( tab . 1 ) .
with such wide angle sky monitoring , the wfm will also provide long term histories of the target sources , thus facilitating both the lad observations and a series of wider science goals .
based on preliminary considerations and estimates , we show that loft , possibly in combination with other grb experiments flying at the same epoch , will give us useful and unique clues to some of these still open issues in this field , through : a ) measurements of the prompt emission down to @xmath12 kev and @xmath1arcmin localization with the wide field monitor ( wfm ) ; b ) measurements of the early afterglow emission with the large area detector ( lad ) .
as detailed below , the partly unprecedented characterization of the grb x ray prompt emission , joined with source location accurate enough for optical follow - up , by the wfm will allow to investigate the properties of the circum - burst environment , thus getting further clues on the nature of the progenitors , to provide a more stringent tests for the emission mechanisms at play , to increase the detection rate of high z grbs w / r to previous missions , to shade light on the population of x ray flashes ( xrfs ) and sub - energetic events . while the grb science with the wfm will come for free , provided the optimization of trigger logic , data modes and prompt ( within a few tens of s ) transmission of trigger time and position to ground ( which is presently in the baseline of the mission @xcite ) , the contribution from the lad , consisting in the possible further characterization of the plateau phase of the early x - ray afterglow emission , will critically depend on the follow - up capabilities and policy of the mission .
the wfm will achieve scientific goals of fundamental importance and not fulfilled by grb experiments presently flying ( e.g. , swift@xcite , konus / wind@xcite , fermi / gbm@xcite ) and future approved missions ( svom@xcite , uffo@xcite ) .
these can be summarized as follows : * measurement of the grb spectral shape and its evolution down to about 2 kev in photon energy which is crucial for testing models of grb prompt emission ( still to be settled despite the considerable amount of observations ; see , e.g. , @xcite ) ; * detection and study of transient x ray absorption column / features for a few bright grbs per year ( fig . 1 ) .
these measurements are of paramount importance for the understanding of the properties of the circum burst matter ( cbm ) and hence the nature of grb progenitors ( still a fundamental open issue in the field ) . in addition , as demonstrated by bepposax @xcite , the detection of transient x ray spectral features can allow the determination of the grb redshift to be compared , when it is the case , with that determined from the optical / nir lines ; * to provide a substantial increase ( with respect to the past and current missions ) in the detection rate of x ray flashes ( xrf ) , a sub class of soft / ultra - soft events which could constitute the bulk of the grb population ( e.g. , @xcite ) and still have to be explored satisfactorily ; * to extend the grb detection up to very high redshift ( z @xmath4 8) grbs ( fig .
2 ) , which is of fundamental importance for the study of evolutionary effects , the tracing of star formation rate , ism evolution , and possible unveiling of population iii stars ( e.g. , @xcite ) ; * to provide fast ( @xmath530s ) and accurate ( @xmath11 arcmin ) location of the detected grbs to allow their prompt multi
wavelength follow - up with ground and space telescopes , thus leading to the identification of the optical counterparts and/or host galaxies and to estimate the redshift , a fundamental measurement for the scientific goals listed above , comparing it with that determined from x - ray absorption lines .
the grb science that could be done with the lad is less automatic and is strongly dependent on the time that will be required to start a too observation .
possible grb science that could be performed by the lad include : * investigating the plateau phase of the early x ray afterglow and its transition to the normal decay ( fig .
3 ) , which is of high importance e.g. for the search of signatures of energy injection by a magnetized ns and to better investigate the correlation between time and luminosity of the transition from plateau to normal decay ( e.g. , @xcite ) ; * searching for emission lines , expected by theoretical models in case of highly metal enriched circum burst environment and detected in a few cases by bepposax , chandra and xmm ( e.g. , @xcite ) but not by swift / xrt ; * complementing the observations of the prompt emission by the wfm by exploiting the transparency of the collimator at energies @xmath4 3040 kev .
no past , present or future ( e.g. , svom , uffo ) grb experiment has such a combination of low energy threshold and high energy resolution ( and wide fov ) , which will make the loft / wfm unique also for grb science .
for instance , the bepposax / wfc or the hete2/wxm had a low energy threshold around 2 kev , but with much worse energy resolution and smaller effective area ( and smaller fov ) .
svom and uffo ( @xmath42018 ) will have a low energy threshold of @xmath15 kev and with significantly worse effective area and energy resolution w / r to loft / wfm at energies @xmath510 kev .
in addition , in particular thanks to the capability of promptly transmitting the grb trigger and a position with a few arcmin uncertainty @xcite , loft will : a ) continue the fundamental service to the astrophysical community , carried out presently by swift ( and possibly in the 2018 - 2022 time frame by svom and/or uffo ) of providing detection , prompt localization and temporal / spectral characterization of grbs , thus allowing their follow up and multi wavelength studies with the best telescopes operating in the @xmath42020 time line ( e.g. , lsst , ska , cta , erosita , maybe xmm , chandra , etc . ) ; b ) complement simultaneous observations by grb experiments flying on other satellites by providing low energy extension and grb position , in a way similar to what is presently done , e.g. , by joining data from swift , fermi and konus / wind .
in addition , the study of grbs , because of their complex phenomenology and extreme energetics , the association of at least a fraction of them with core collapse sne , their redshift distribution extending up to at least z@xmath189 , is relevant to different fields of astrophysics , ranging from plasma and black hole physics to cosmology . in particular
, the loft measurements summarized above will provide not only a step forward in the understanding of grb physics , progenitors and sub classes , but , also in combination with observations by the best instrumentation at all wavelengths , will shade light on fundamental topics like the evolution of the star forming rate and of the galactic ism up to the epoch of re - ionization , the first generation ( pop iii ) stars , the understanding of the diversity and rate of core - collapse sne , the measurement of cosmological parameters . finally , the grb science already involves , and will continue to involve at the time of loft , observers and instrumentation from different communities : optical / ir robotic telescopes ( prompt detections and localization of the optical counterparts ) , major optical / ir telescopes like vlt , gemini , hubble ( identification , redshift and ism of the host galaxies , optical afterglow decay and jet signatures ) , radio telescopes like vla ( afterglow modeling and energetics ) observations in the tev range by cherenkov telescopes ( challenging emission models ) .
p. mszaros , rep .
69 ( 2006 ) 2259 .
n. gehrels , e. ramirez ruiz , d.b .
fox , ara&a 47 ( 2009 ) 567 .
b. zhang , comptes rendus physique 12 ( 2011 ) 206 ( arxiv:1104.0932 ) .
m. feroci , l. stella , m. van der klis , et al .
, experimental astronomy 34 ( 2012 ) 415 .
m. feroci , j.w .
den herder , e. bozzo , et al . ,
proceedings of the spie 8443 ( 2012 ) id.2 .
e. bozzo , j.w .
den herder , m. feroci , l. stella , proceedings of the extreme and variable high energy sky ( proceedings of science : extremesky 2011 ) ( 2011 ) id.25 .
s. zane , d. walton , t. kennedy , et al .
, proceedings of the spie 8443 ( 2012 ) id.15 .
s. brandt , m. hernanz , l. alvarez , et al .
, proceedings of the spie 8443 ( 2012 ) id.14 .
n. gehrels , g. chincarini , p. giommi , et al . , apj 611 ( 2004 ) 1005 .
aptekar , d.d .
frederiks , s.v .
golenetskii , et al . ,
space science reviews 71 ( 1995 ) 265 . c. meegan , g. lichti , p.n .
bhat , et al . ,
apj 702 ( 2009 ) 791 .
o. godet , j. paul , j.y .
wei , et al .
, proceedings of the spie 8443 ( 2012 ) id.10 .
b. grossan , i.h .
park , s. ahmad , et al . , proceedings of the spie 8443 ( 2012 ) id.13 .
b. zhang , p. mszros , apj 581 ( 2002 ) 1236 .
l. amati , f. frontera , m. vietri , et al .
, science 290 ( 2000 ) 953 .
a. pelangeon , j .- l .
atteia , y. e. nakagawa , et al .
, a&a 491 ( 2008 ) 157 .
r. salvaterra , s. campana , s. vergani , g. , et al . ,
apj 749 ( 2010 ) 68 .
woosley , j.s .
bloom , ara&a 44 ( 2006 ) 507 . b. zhang , chinese journal of astronomy and astrophysics 7 ( 2007 ) 1 . l. piro , g. garmire , m. garcia , et al . , science 290 ( 2000 ) 955 . | loft is a satellite mission currently in assessment phase for the esa m3 selection .
the payload is composed of the large area detector ( lad ) , with 250 kev energy band , a peak effective area of about 10 m@xmath0 and an energy resolution better than 260 ev , and the wide field monitor ( wfm ) , a coded mask imager with a fov of several steradians , an energy resolution of about 300 ev and a point source location accuracy of 1 arcmin in the 230 kev energy range . based on preliminary considerations and estimates ,
we show how the scientific performances of the wfm are particularly suited to investigate some of the most relevant open issues in the study of grbs : the physics of the prompt emission , the spectral absorption features by circum - burst material ( and hence the nature of the progenitors ) , the population and properties of xrfs , and the detection and rate of high z grbs .
measurements of the early afterglow emission with the large area detector ( lad ) may also be possible depending on the mission slewing capabilities and too observations policy .
x ray astronomy : instrumentation , gamma rays : bursts [ ta1 ] |
the evaluation of the accuracy of the cesium primary frequency standards requires to take into account , among others , the shift induced on the clock transition @xmath8 by the black body radiation ( bbr ) .
the definition of the second in the international system of units ( si ) is in fact based on the above transition observed in absence of any perturbation and then , in particular , at @xmath9 k @xcite , where @xmath10 is the temperature of the environment .
gallagher and cook @xcite firstly considered the effects of bbr on the rydberg states of the atoms while itano et al . @xcite focused their attention onto the atomic frequency standards .
they reported the following expression for the relative frequency shift of the @xmath0cs clock transition ( @xmath11 hz ) : @xmath12\ ] ] with @xmath13 and @xmath14 when @xmath15 k. this expression has been adopted till now to correct all the primary frequency standards for the bbr shift at an accuracy level of @xmath16 or even better @xcite .
the bbr induces both non resonant ac stark and zeeman shifts of the atomic transitions . for the ground state @xmath17 of the alkali - metal atoms the zeeman shift is of the order of @xmath18 @xcite and
is not relevant for the accuracy estimation of the present frequency standards .
the direct experimental method to evaluate the bbr shift is based on ( [ eq : itano ] ) : the frequency shift @xmath19 is measured versus the black body temperature _ t _ and the experimental points are fitted with ( [ eq : itano ] ) whose temperature dependence follows directly from the planck radiation law .
this method has been used by bauch and schroeder @xcite employing a thermal cesium beam and in our group using an atomic fountain @xcite . in order to increase the experimental resolution ,
the bbr shift may be also evaluated indirectly by measuring the dc stark shift coefficient @xmath20 defined as : @xmath21 where @xmath19 is the frequency shift induced by the static electric field @xmath22 .
measurements of @xmath20 have been performed in this way by haun and zacharias @xcite and by mowat @xcite , using a thermal beam and the ramsey interrogation technique , and by simon et al .
@xcite using an atomic fountain .
the bbr shift is then obtained from ( [ eq : kappa ] ) assuming _ e _ : @xmath23 where @xmath24 is the mean - squared electric field of the black body radiation , @xmath25 is the stefan - boltzmann constant @xcite , @xmath26 is the vacuum permittivity and @xmath27 is the speed of light in vacuum . in ( [ eq : campoe ] ) the first identity
is valid at a level of a few percent and can be corrected as reported in @xcite , while the second identity is obtained directly by the planck radiation law ( see next section ) .
the measurement of the scalar static polarizability @xmath28 is related to the above method through the expression : @xmath29 where @xmath30 is the potential energy of a neutral atom in a static electric field @xmath22 of moderate strength . due to the very low intensity of the black body field ,
the hyperpolarizability effect is negligible @xcite and will not be considered throughout the paper .
the polarizability is obtained from the force acting on the atom when submitted to a gradient of the electric field : @xmath31 this force , in turn , can be measured through the deflection of thermal beams @xcite , through e - h gradient balances @xcite or through the time of flight of cooled atoms in a fountain apparatus @xcite . in principle
, the bbr shift can be also obtained from ( [ eq : campoe ] ) and ( [ eq : ploarizzability ] ) as follows : @xmath32\langle\mathcal{e}^{2}(t)\rangle\ ] ] where @xmath33 and @xmath34 in the cs case ; @xmath35 is the planck constant .
the indirect experimental technique allows a higher resolution , being possible to submit the atoms to dc fields of several mv / m , if compared to the black body field @xmath36 v / m at _ t _ = 300 k used in the direct method .
nevertheless , it is based on the conceptual assumption @xmath37 and may not be free from spurious effects as we shall discuss further on .
> from the theoretical point of view , the evaluation of the bbr shift requires the computation of the scalar polarizabilities of the two states defining the clock transition or , in other words , of the non resonant ac stark shift .
this is typically performed through perturbation techniques @xcite .
more precisely , according to ( [ eq : shiftalfa ] ) , we have to calcolate the differential polarizability between the two ground state hyperfine levels . as pointed out by feichtner et al .
@xcite , this requires to consider in the theoretical analysis a basis of modified eigenfunctions which account for the hyperfine interaction of the ground and @xmath38 states with other states of the same quantum numbers @xmath39 and @xmath40 . a review of the theoretical works leading to the evaluation of the ground state @xmath0cs polarizability can be found in @xcite and in @xcite .
due to the basic importance of the bbr shift correction to evaluate the present primary frequency standard accuracy at the @xmath16 level and , in the next future at the @xmath41 level , it is highly desirable to re - examine the relation ( [ eq : itano ] ) considering that : * new and more precise experimental data are now available for the electric dipole moments and for the transition frequencies of @xmath0cs involved in the theoretical computation of @xmath28 ; * recent theoretical evaluations of @xmath28 @xcite lead to a significantly lower value than that commonly assumed up to now @xcite ; * the two most precise measurements @xcite of the electrical polarizability do not agree to each other .
in this paper we report an ab - initio computation of the bbr shift for the clock transition which turns out lower by @xmath42 at 300 k than the commonly adopted value of ( [ eq : itano ] ) .
a critical comparison with the theoretical and experimental values reported in the literature will be also given , showing the present difficulty to asses the accuracy at the @xmath41 level for atomic frequency standards operating at t @xmath43 k.
the non - resonant ac stark shift @xmath44 of the energy of state @xmath45 induced by the level @xmath46 is given by @xcite : @xmath47 where @xmath48 , @xmath49 the electron charge , @xmath50 the polarization vector of the electric field , @xmath51 the atomic position vector , @xmath52 and @xmath53 the energy levels of the states @xmath45 and @xmath46 in frequency units and @xmath54 the mean - squared value of the electric field at frequency @xmath55 .
when @xmath45 represents a ground state , we have @xmath56 being @xmath57 and @xmath58 the frequency and the wavelength of the atomic transition @xmath59 the electric field per frequency unit @xmath60 is related to the spectral energy density @xmath61 by the well known relation : @xmath62 in our case the energy density @xmath63 is given by the planck radiation law : @xmath64 where @xmath65 is the boltzmann constant .
the maximum of @xmath66 is reached when ( wien law ) : @xmath67 at _ t _ = 300 k , relation ( [ eq : wien ] ) gives @xmath68 thz . in the cs case
, the minimum value of @xmath57 corresponds to the d@xmath69 optical transition ( 335 thz ) so that @xmath70 for all @xmath46 levels coupled to the state @xmath45 and it is possible to omit the @xmath71 term in the denominator of ( [ eq : acss ] ) avoiding the singularity of the integrand at @xmath72 .
we shall consider further on the effect of this approximation .
the basic relation ( [ eq : acss ] ) can then be written as : @xmath73 where the mean - squared electric field @xmath74 is given by @xmath75 the relation ( [ eq : campoenergy2 ] ) , obtained through ( [ eq : campoenergy ] ) and ( [ planck ] ) , is also known as the stefan - boltzmann law ; at _ _
t__=300 k it gives @xmath76 ( v / m)@xmath77 , as already reported in the previous section .
summing over all the excited states @xmath46 coupled to the level @xmath45 , we obtain the total stark shift for the state @xmath45 : @xmath78 moreover , taking ( [ eq : campoe ] ) and ( [ eq : ploarizzability ] ) into account , it is possible to write ( [ eq : acss3 ] ) also in the following form : @xmath79 where @xmath80 is the scalar polarizability of the state @xmath45 which may be measured experimentally . more precisely , the expression ( [ eq : alfa ] ) takes into account the contribution to the polarizability from the valence excited states . in principle , the total polarizability includes also the contribution from the core excited states @xcite ; however , this is not important for the clock transition shift since it is common to both the ground state sublevels .
the bbr shift @xmath81 of the clock transition @xmath82 we are looking for is : @xmath83 introducing relations ( [ eq : campoenergy2 ] ) and ( [ eq : acss3 ] ) in ( [ deltanu ] ) we finally obtain : @xmath84 we conclude this theoretical review observing that the dc stark coefficient @xmath20 defined in ( [ eq : kappa ] ) can be expressed through the polarizabilities of the ground state levels ( see also ( [ eq : shiftalfa ] ) ) as : @xmath85 where @xmath86 is the scalar differential polarizability introduced by sandars @xcite . in the experimental measurement of @xmath20
performed with a dc electric field a tensorial component is also present , which has to be subtracted in order to take the isotropy of the black - body radiation into account @xcite .
to evaluate numerically the bbr shift given by ( [ eq : deltanu2 ] ) we need the matrix elements of the operator @xmath87 ; writing the atomic position vector @xmath51 as an irreducible rank 1 tensor operator we have : @xmath88 where @xmath89 is expressed in terms of its spherical components @xmath90 , @xmath91 and @xmath92 ; @xmath93 ( co - latitude ) and @xmath94 ( azimuth ) are the polar angles defining the electric field direction with respect to the quantization axis ( @xmath95 axis ) . the well - known wigner - eckart @xcite theorem allows us to simplify the generic matrix element : latexmath:[\[\label{eq : wignereckart } \langle\ 6s_{1/2 } ; f , m_{f}| e r_{q } | n'p_{j ' } ; f ' , m'_{f } \rangle\ = q(f , m_{f } ; f ' , m'_{f } ; j , j ' ; q ) \langle\ 6s_{1/2 } the dipole reduced matrix element and _ q _ is the index labelling the component of @xmath51 .
in ( [ eq : wignereckart ] ) we defined : @xmath97 where the coefficients in round and in curly brackets are the 3-@xmath98 and 6-@xmath98 symbols respectively .
the primed quantum numbers refer to the excited states @xmath99 coupled to the ground state by allowed electric dipole transitions .
the polarizability of each of the two ground state levels can then be written as : @xmath100 in ( [ eq : integralone ] ) we perform an integration over the solid angle @xmath101 to take the isotropic nature of the bbr into account , consequently the integral is normalized to @xmath102 ; @xmath103 is the frequency of the transition indicated in parentheses . according to selection rules , for fixed values of the quantum numbers
only one of the three @xmath104 terms is not zero . actually , as feichtner et al .
@xcite pointed out , the @xmath105 state and the @xmath38 states wave functions are perturbed by the effect of their hyperfine interaction with other states with the same @xmath39 and @xmath40 but with different principal quantum number @xmath106 .
therefore , the dipole matrix elements should not be calculated between the unperturbed wave functions but between the perturbed ones defined as : @xmath107 the coefficients @xmath108 , @xmath109 and @xmath110 , given in @xcite , are here reported for convenience .
\bigg[\frac{\nu_{0 } } { \nu(ns_{1/2}\rightarrow 6s_{1/2})}\bigg ] \\ &
b_{njf}=\bigg(\frac{\psi_{ns}(0)}{\psi_{6s}(0)}\bigg)\bigg[\frac{63/4+j(j+1 ) - f(f+1)}{32j(j+1)}\bigg]\bigg[\frac{\nu_{0}}{\nu(np_{j}\rightarrow 6p_{j})}\bigg]\\ \\ & c_{njf}=\frac{\psi_{7s}(0)\psi_{ns}(0)}{\big(\psi_{6s}(0)\big)^{2}}\bigg[\frac{63/4+j(j+1 ) -f(f+1)}{32j(j+1)}\bigg]\bigg[\frac{\nu_{0}}{\nu(np_{j}\rightarrow 7p_{j})}\bigg ] \end{split}\ ] ] and their numerical values are reported in tables 1 , 2 and 3 .
these coefficients are very small ( of the order of @xmath112 ) and their contribution to the absolute value of the scalar polarizability is completely negligible .
however , they must be taken into account in order to obtain the correct value of the differential polarizability we are interested in .
[ taba ] .coefficients @xmath113 ( in units of @xmath112 ) for perturbed cesium @xmath114 wave function . [ cols="^,^,^,^",options="header " , ] the total scalar polarizability @xmath28 must take the core polarizability @xmath115 into consideration which accounts for approximately the 4 % of the total value in the case of @xmath0cs @xcite .
our final result for @xmath28 is then : @xmath116 where @xmath117 is given by ( [ eq : alfa0 ] ) and @xmath115 is reported in @xcite . in table 6
we report the data found in literature with an uncertainty lower than 1 % .
an agreement at a level of @xmath118 can be observed , well inside the quoted error bars , both between the theoretical computations and with respect to the experimental value of amini and gould @xcite . in figure 2 the result of our computations concerning the bbr shift parameter @xmath119
is compared with the experimental results reported in literature .
our result agrees with the direct measured values while the indirect measurements , even if in agreement among them , are higher by three standard deviations .
this discrepancy has to be deeply examined not only for the correct estimation of the differential polarizability of the ground state hyperfine transition but also for its impact into the accuracy evaluation of the primary frequency standards .
first of all , the following points have to be considered : ) of the ground state @xmath0cs hyperfine transition . with the triangle we indicated the indirect measurements , while with the square the direct ones .
@xmath120 is ref .
@xcite , @xmath121 is ref .
@xcite , @xmath122 is ref .
@xcite , @xmath123 is ref .
@xcite , and @xmath124 is ref . @xcite . ] * our theoretical result leads to a polarizability @xmath28 of the @xmath17 state in agreement with the most recent and precise evaluations as shown in table 6 , while the indirect measurements , considering that the stark effect is proportional to the polarizability at the first order , lead to a considerably higher value for @xmath28 ; * our computations lead to the result of itano et al .
@xcite if we use the old values of the transition frequencies @xcite and of the electric dipole moments @xcite ; the latter , in particular , are considerably higher than the most precise recent values ; * our @xmath119 value agrees with the values obtained with the direct measurement method . taking the above remarks into account ,
the following hypothesis may be considered in order to explain this discrepancy : * the assumption of equivalence between ac and dc stark shifts is not fully correct , in spite also of the correction term @xmath125 introduced by itano et al .
@xcite ; * a systematic shift is present in the indirect measurements due to a physical effect not considered and present in the experiment .
since we have no reason at the moment to believe in the breakdown of the equivalence between ac and dc stark shifts , as far as the second hypothesis is concerned , the following point should be examined .
all the measurements performed with the indirect method are based on a ramsey interaction scheme with a stark field applied during the free flight of the atoms between the two interaction regions .
the atoms experience then two sudden changes of the electric field which could perturb the free evolution of the hyperfine coherence or induce majorana - type transitions .
the last hypotheses must be considered more a clue than a probe to explain the discrepancy under question and obviously requires a more complete theoretical treatment .
we have reported in this paper a theoretical evaluation of the @xmath17 ground state polarizability of @xmath0cs and the bbr shift of the ground state hyperfine transition between the magnetic field independent sublevels .
the former is in really good agreement with recent theoretical and experimental values ; the latter agrees with the direct measurements reported in the literature and is lower by three standard deviations than the values obtained from dc stark measurements .
the discrepancy of @xmath126 at _ _
t__=300 k between the more recent estimations and the value accepted up to now impacts the accuracy evaluation of the primary frequency standards at the level of two or three times their believed standard uncertainties .
moreover , serious problems may be expected from the bbr shift to asses the accuracy at the level of @xmath127 for the future microwave frequency standards operating at @xmath128 k.
to avoid the singularity at @xmath72 in section ii we have omitted the @xmath71 term in the denominator of ( [ eq : acss ] ) .
this simplification , usually adopted in literature , is justified by the fact that near room temperature the energy content of the bbr is peaked at frequencies much lower than the lowest transition frequency of the atom .
an estimation of the error due to this approximation is provided in @xcite with the introduction of the corrective term @xmath125 in ( [ eq : itano ] ) .
here we re - evaluate this term following also the method described in @xcite .
relations ( [ eq : acss ] ) , ( [ eq : campoenergy ] ) and ( [ planck ] ) give : @xmath129 where @xmath130 and @xmath131 .
expanding in ( [ eq : integral ] ) the factor : @xmath132 and integrating we obtain : @xmath133 being @xmath134 the bernoulli numbers .
if we consider the main ( @xmath135 ) and the first order ( @xmath136 ) terms of ( [ eq : bernoulli ] ) and sum over all the valence states , we obtain : @xmath137 where @xmath139 the main term of ( [ deltanu3 ] ) coincides with ( [ eq : deltanu2 ] ) , as expected . as regards the corrective term @xmath125 , following the same procedure to evaluate the dipole matrix elements in the basis of modified eigenfunctions as reported in section iii ,
we obtain @xmath140 , in agreement with the value reported in @xcite .
this agreement is not surprising because the recent values of the the dipole moments affect significantly the main parameter @xmath119 , but much less the corrective term @xmath125 , as can be observed in ( [ eq : epsilon ] ) where they appear both in the numerator and in the denominator .
200 j. terrien , metrologia * 4 * , 41 ( 1968 ) and ccds 13th meeting bureau international poids et mesures , recommendation s2 ( 1996 , sevres , france ) . t. f. gallagher and w. e. cook , phys .
lett . * 42 * , 835 ( 1979 ) .
w. m. itano , l. l. lewis , and d. j. wineland , phys . rev . a * 25 * , 1233 ( 1982 ) .
a. clairon et al . , proc .
of 5th symp . on freq .
stand . and metr . , woods hole , ma ( 1997);f .
levi et al . , ieee trans .
. meas . * 52 * , 2 ( 2003 ) ; s. jefferts et al .
, metrologia * 39 * , 321 ( 2002 ) ; s. weyers et al .
, _ ibid . _ * 38 * , 343 ( 2001 ) ; .
a. bauch and r. schrder , phys .
78 * , 622 ( 1997 ) .
f. levi , d. calonico , l. lorini , s. micalizio , and a. godone , submitted to phys .
r. d. haun and j. r. zacharias , phys .
rev . * 107 * , 107 ( 1957 ) .
j. r. mowat , phys .
a * 5 * , 1059 ( 1972 ) .
e. simon , p. laurent , and a. clairon , phys .
a * 57 * , 436 ( 1998 ) .
p. j. mohr and b. n. taylor , codata recommended values of the fundamental physical constant : 1998 , j. phys .
data , * 28 * 1713 ( 1999 )
. v. g. palchikov , yu . s. domnin , and a. v. novoselov , j. opt .
b : quantum semiclass . opt . * 5 * , s131 ( 2003 ) .
w. d. hall and j. c. zorn , phys .
rev . a * 10 * , 1141 ( 1974 ) .
g. e. chamberlain and j. c. zorn , phys .
rev . * 129 * , 677 ( 1963 ) .
a. salop , e. pollack , and b. bederson , phys . rev . * 124 * , 1431 ( 1961 ) .
r. w. molof , h. l. schwartz , t. m. miller , b. bederson , phys .
rev . a * 10 * , 1131 ( 1974 ) .
j. a. maddi , t. p. dinnen , and h. gould , phys .
a * 60 * , 3882 ( 1999 ) .
j. m. amini and h. gould , phys .
rev . lett . * 91 * , 153001 ( 2003 ) .
l. d. landau and e. m. lifschitz , _ quantum mechanics , nonrelativistc theory _ , pergamon press , oxford ( 1965 ) .
c. h. townes and a. l. schawlow , _ microwave spectroscopy _ , dover publications , inc .
new york , 1975 .
j. d. feichtner , m. e. hoover , and m. mizushima , phys . rev . * 137 * , a702 ( 1965 ) a. derevianko and s. g. porsev , phys
a * 65 * , 55403 ( 2002 ) .
a. derevianko , w. r. johnson , m. s. safronova , and j. f. babb , phys .
lett . * 82 * , 3589 ( 1999 ) .
g. h. sandars , proc .
. soc . * 92 * , 857 ( 1967 ) ; j. r. p. angel and p. h. g. sandars , proc .
a * 305 * , 125 ( 1968 ) .
a. r. edmonds , _ angular momentum in quantum mechanincs _ , princeton university press , princeton , new jersey , third printing , 1974 . th .
udem , j. reichert , r. holzwarth , and t. w. hnsch , phys .
lett . * 82 * , 3568 ( 1999 ) ; t. udem et al .
a * 62 * , 31801 ( 2000 ) .
r. j. rafac , c. e. tanner , a. e. livingston , k. w. kukla , h. g. berry , and c. a. kurtz , phys .
a * 50 * , r1976 ( 1994 ) .
r. j. rafac , c. e. tanner , a. e. livingston , and h. g. berry , phys .
a * 60 * , 3648 ( 1999 ) .
l. young , w. t. hill , s. j. sibener , s. d. price , c. e. tanner , c. e. wieman , and s. r. leone , phys .
a * 50 * , 2174 ( 1994 ) . c. amiot , o. dilieu , r. f. gutterres , and f. masnou - seeuws , phys .
a * 66 * , 52506 ( 2002 ) .
m. s. safronova , w. r. johnson , and a. derevianko , phys .
a * 60 * , 4476 ( 1999 ) . v. a. dzuba , v. v. flambaum , and o. p. sushkov , phys .
a * 56 * , r4357 ( 1997 ) .
s. a. blundell , w. r. johnson , and j. sapirstein , phys . rev . a * 43 * , 3407 ( 1991 ) .
p. h. stone , phys .
rev . * 127 * , 1151 ( 1962 ) . c. e. moore , _ atomic energy levels _ , natl .
bur . std .
( u. s. ) , reprint of nbs circ .
467 ( u. s. government printing office , washington , d. c. , 1971 ) .
j. w. farley and w. h. wing , phys .
a * 23 * , 2397 ( 1981 ) . | we report the theoretical evaluations of the static scalar polarizability of the @xmath0cs ground state and of the black body radiation shift induced on the transition frequency between the two hyperfine levels with @xmath1 .
this shift is of fundamental importance in the evaluation of the accuracy of the primary frequency standards based on atomic fountains and employed in the realization of the si second in the international atomic time ( tai ) scale at the level of @xmath2 .
our computed value for the polarizability is @xmath3 in agreement at the level of @xmath4 with recent theoretical and experimental values . as regards the black body radiation shift we find for the relative hyperfine transition frequency @xmath5 at @xmath6 k in agreement with frequency measurements reported by our group and by bauch and schrder [ phys .
rev .
lett . *
78 * , 622 , ( 1997 ) ] .
this value is lower by @xmath7 than that obtained with measurements based on the dc stark shift and with the value commonly accepted up to now . + pacs number(s ) : 32.60.+i , 32.10.dk , 06.30.ft |
the elastic neutrino - photon scattering process , @xmath0 , and the corresponding crossed channels have been the subject of numerous investigations related to astrophysical applications @xcite . when the center of mass energy @xmath2 is much less than the @xmath3-boson mass @xmath4 , the amplitudes for these processes have a relatively simple form .
in particular , the forward helicity non - flip amplitudes for the elastic processes @xmath0 and @xmath1 with massless neutrinos can be expanded in powers of @xmath5 by making use of unitarity@xcite .
the series are of leading order @xmath6 , and the inclusion of higher powers of @xmath7 , such as @xmath8 , enables us to use the forward elastic scattering amplitudes for @xmath0 and @xmath1 to study the optical activity of a sea of neutrinos and antineutrinos .
this property of a neutrino sea was recognized by nieves and pal @xcite and studied by mohanty , nieves , and pal @xcite in the case where the photons move in a plasma and therefore experience dispersion .
dispersive effects give rise to mass dependent of terms of the form @xmath9 ( @xmath10 denoting the photon momentum and @xmath11 the electron mass ) , which vanish in the vacuum . here , we treat the vacuum case where @xmath10 satisfies @xmath12 , and find a rather different dependence on @xmath4 and @xmath11 @xcite . in the next section , using the results of ref.@xcite , we write the forward helicity amplitudes for the elastic scattering @xmath0 and @xmath1 .
we then present a calculation for the index of refraction and the rotary power for a sea of neutrinos and antineutrinos as a function of the energy of an incident photon , the temperature of the sea , and the neutrino degeneracy parameter @xmath13 , where @xmath14 is the chemical potential for neutrinos and @xmath15 is the temperature of the sea .
using the eqs.(24)(26 ) of the ref.@xcite , the helicity non - flip amplitudes , @xmath16 , for the forward elastic scattering of a photon with helicity @xmath17 and an electron neutrino when @xmath18 , can be written as @xmath19 where @xmath20\,.\end{aligned}\ ] ] here , @xmath21 is the weak mixing angle , @xmath22 is the fine structure constant , and we have neglected higher powers of @xmath23 and @xmath5 . using the eq.(12 ) of the ref.@xcite , the helicity non - flip amplitudes , @xmath24 , for forward elastic scattering of a photon of helicity @xmath17 and an antineutrino , can be expressed as @xmath25 which is a consequence of @xmath26 invariance . to calculate the optical activity of a neutrino - antineutrino sea , we consider a photon of helicity @xmath17 and energy @xmath27 traversing a bath of neutrinos and antineutrinos that are in thermal equilibrium at the temperature @xmath15 . to give an order of magnitude estimate of the index of refraction @xmath28 of this sea
, we write @xcite @xmath29 which is a generalization of the eq.(57 ) of the ref.@xcite . the forward - scattering amplitudes
@xmath30 and @xmath31 , from the eq.(58 ) of the ref.@xcite , are @xmath32 and @xmath33 and @xmath34 are the usual fermi - dirac distributions , @xmath35 here , @xmath15 is the temperature of the neutrino - antineutrino sea , @xmath14 is the chemical potential for the neutrinos @xcite , @xmath36 and @xmath37 are the momentum and energy of a neutrino , and @xmath38 , where @xmath39 is the angle between the incoming photon and the incoming neutrino ( with similar definitions for the antineutrino ) . from eqs.([n1])([fscatbar ] ) we obtain @xmath40 due to the apparent baryon asymmetry in the universe , it is natural to consider the possibility of the lepton asymmetry . to study this possibility , and for numerical calculations of the index of refraction and the rotary power ,
it is convenient to introduce the following ratios @xmath41 where @xmath42 and @xmath43 are the number densities for the neutrino and antineutrino , respectively , and @xmath44 is the riemann zeta function . from the eqs.([dn ] ) , ( [ dnbar ] ) , ( [ ratiol1 ] ) , and ( [ ratiol2 ] ) , we find @xmath45 where @xmath46 and @xmath13 is the neutrino degeneracy parameter . as it is clear from the eqs.([ratiol1a ] ) ( [ i ] ) , the parameters @xmath47 and @xmath48 are solely functions of @xmath49 .
these neutrino asymmetry parameters @xmath47 and @xmath48 are defined such that for @xmath50 they coincide with the customary definition of the neutrino asymmetry parameter @xmath51 @xcite for a sea of relic neutrino - antineutrino that decoupled from photons in the early universe @xmath52 here , @xmath53 is the photon number density @xmath54 and the relation @xmath55 , which holds for the present day temperatures of the relic neutrinos and photons , is used .
when @xmath49 is large , @xmath47 approaches a finite value , while @xmath48 increases . in the fig.[l1l2xi ] @xmath47 and @xmath48 are plotted as functions of @xmath49 . as it is clear from this figure , for @xmath56 , @xmath47 is close to its maximum value of @xmath57 . and
@xmath48 , as functions of the neutrino degeneracy parameter @xmath13 .
the relations @xmath58 and @xmath59 hold.,width=264 ] after using eqs.([dn ] ) and ( [ dnbar ] ) for @xmath33 and @xmath34 , eqs.([nu1])([cpt1 ] ) for the amplitudes @xmath60 and @xmath61 , and performing integrations in the eq.([n2 ] ) , we obtain @xmath62 where ( for @xmath63 ) @xmath64\nonumber\\ & & \times \left(\frac{1}{4}\,\xi^4 + \frac{\pi^2}{2}\,\xi^2 + \frac{7\pi^4}{60}\right)\nonumber\\ & \simeq & 1.9 \times 10^{-4}\left(\frac{1}{4}\,\xi^4
+ \frac{\pi^2}{2}\,\xi^2 + \frac{7\pi^4}{60}\right),\;\label{c0}\\\ c_1 & = & \frac{4\alpha^2}{9\pi^2 \sin^2\theta_w } \left[\ln\left(\frac{m_w^2}{m_e^2}\right ) - \frac{8}{3}\right]\nonumber\\ & & \times \left(\frac{1}{5}\,\xi^5 + \frac{2\pi^2}{3}\,\xi^3 + \frac{7\pi^4}{15}\,\xi\right)\nonumber\\ & \simeq & 2.2 \times 10^{-4}\left(\frac{1}{5}\,\xi^5 + \frac{2\pi^2}{3}\,\xi^3 + \frac{7\pi^4}{15}\,\xi\right).\label{c1}\end{aligned}\ ] ] notice that the @xmath49dependent parts of the @xmath48 , @xmath65 , and @xmath66 are related as @xmath67 @xmath68 in the figs.[c0c1xi][c0c1l2 ] the coefficients @xmath65 and @xmath66 are plotted as functions of @xmath49 , @xmath47 , and @xmath48 , respectively . and @xmath66 as functions of the neutrino degeneracy parameter @xmath13 . the relations @xmath69 and @xmath70 hold.,width=264 ] the range of the validity of the eqs.([np1])([c1 ] ) , as far as energy is concerned , is related to that of the eqs.([nu1])([h1 ] ) , which is @xmath71 . in the eq.([n2 ] ) , if we change the upper limits of the integrations on @xmath37 and @xmath72 from the infinity to @xmath73 , the contributions of these integrals to @xmath65 and @xmath66 in the eqs.([np1 ] ) and ( [ nm1 ] ) change , at most , by 9% and 18% , respectively , if we use @xmath74 ( for @xmath75 , the corresponding changes are at most 5% and 10% ) . here , we set the following criterion : @xmath76 which for @xmath74 is @xmath77 where @xmath27 is the photon energy in gev , and @xmath15 is the neutrino temperature in kelvin . and @xmath66 as functions of the asymmetry parameter @xmath47 .
the relations @xmath78 and @xmath79 hold.,width=264 ] from eqs.([np1 ] ) and ( [ nm1 ] ) , we have @xmath80 and the following approximate relation : @xmath81 equation([npm2 ] ) implies that the leading term of the index of refraction is independent of the helicity and the energy of the incident photon , as long as eq.([criterion2 ] ) is satisfied . and @xmath66 as functions of the asymmetry parameter @xmath48 .
the relations @xmath82 and @xmath83 hold.,width=264 ] when linearly polarized light propagates through a medium that has different indices of refraction for positive and negative helicities ( @xmath84 ) , the plane of polarization of the light rotates by an angle @xmath85 , which is @xcite @xmath86 where @xmath27 and @xmath87 are the energy and wavelength of the photon and @xmath88 is the distance travelled by photons in the medium . to estimate the specific rotary power , @xmath89 , for a sea of neutrinos and antineutrinos , we use eqs.([npm1 ] ) and ( [ phi1 ] ) to obtain @xmath90 where , again , @xmath27 is in gev and @xmath91 is in kelvin . a positive angle of rotation , @xmath92 , that is @xmath93 ,
corresponds to a clockwise rotation ( dextrorotation ) of the plane of polarization of the linearly polarized incident photons , as viewed by an observer that is detecting the forward - scattered light .
thus , the optical activity of a neutrino - antineutrino sea with @xmath94 is that of a dextrorotary medium .
in addition , it is clear from eq.([phi2 ] ) that the rotary power , @xmath89 , varies as @xmath95 , which is the same as that of quartz and most transparent substances for visible light .
is shown as a function of @xmath49 for @xmath96gev .
the curve scales as @xmath97 .
[ rot],width=264 ] to get a rough estimate of rotation angle @xmath85 for linearly polarized photons propagating through the relic neutrino and antineutrino sea ( with @xmath98 ) , we use eq.([phi2 ] ) with @xmath99 , @xmath100 m / s , @xmath101 years , @xmath102k , and @xmath103 ev , and we find @xmath104 which is exceedingly small . the dependence of @xmath85 on @xmath49 is shown in fig.[rot ] .
it should be noted that recent studies which include neutrino oscillation effects in the derivation of cosmological bounds on @xmath49 conclude that @xmath105 @xcite .
in conclusion , we note that using eqs.([ratiol2 ] ) , ( [ ratiol2a ] ) , ( [ c1 ] ) , and ( [ phi2 ] ) , for the neutrino degeneracy parameter @xmath50 , the specific rotary power , @xmath89 , for a sea of neutrinos and antineutrinos is @xmath106 \frac{\omega^2 t_\nu^2}{m_w^4}(n_\nu - n_{\bar\nu } ) \,.\ ] ] this equation differs in several respects from the corresponding result in ref.@xcite , @xmath107 where , in the non - relativistic limit , the plasma frequency @xmath108 is related to the electron number density @xmath109 as , @xmath110 from eqs.([phi4 ] ) and ( [ mnprot ] ) , it is clear that the mass scales for the case of photon dispersion versus no dispersion are very different and this favors the dispersive regime .
the size of specific rotary power in the dispersive case is independent of the frequency of the propagating photons and , apart from a dependence on the lepton asymmetry parameter @xmath111 shared by both expressions , is controlled by the electron number density @xmath109 .
in addition , the non - dispersive case depends on frequency and temperature .
these differences clearly indicate that eqs.([phi2 ] ) and ( [ mnprot ] ) represent complementary limits in the treatment of the optical activity of a neutrino - antineutrino sea . we wish to thank duane dicus for numerous helpful discussions .
one of us ( a.a . ) wishes to thank the department of physics and astronomy at michigan state university for its hospitality and computer resources .
this work was supported in part by the national science foundation under grant no .
phy-0070443 .
99 h .- y .
chiu and p. morrison , phys .
lett . * 5 * , 573 ( 1960 ) .
levine , nuovo cimento a * 48 * , 67 ( 1967 ) .
landovitz and w.m .
schreiber , nuovo cimento a * 2 * , 359 ( 1971 ) .
v.k . cung and m. yoshimura , nuovo cimento a * 29 * , 557 ( 1975 ) .
d.a . dicus and w.w .
repko , phys .
d * 48 * , 5106 ( 1993 ) .
j. liu , phys .
d * 44 * , 2879 ( 1991 ) .
a. abbasabadi , a. devoto , d.a .
dicus , and w.w .
repko , phys .
d * 59 * , 013012 ( 1999 ) .
a. abbasabadi , a. devoto , and w.w .
repko , phys .
d * 63 * , 093001 ( 2001 ) .
a. abbasabadi and w.w .
repko , phys .
d * 64 * , 113007 ( 2001 ) .
nieves and p.b .
pal , phys . rev .
d * 39 * , 652 ( 1989 ) .
s. mohanty , j.f .
nieves and p.b .
pal , phys . rev .
d * 58 * , 093007 ( 1998 ) .
g. karl and v. novikov , optical activity of a neutrino sea in the standard model , " hep - ph/0009012 .
notice that eq.([n1 ] ) is based on the standard lorentz relation that connects the index of refraction to the forward - scattering amplitude , for a medium that is at @xmath112 temperature . here
, we make a nontrivial assumption that the introduction of finite temperature in the medium , will not spoil the coherent condition that was necessary to derive the lorentz relation . for a similar approach ,
see p. langacker and j. liu , phys .
d * 46 * , 4140 ( 1992 ) .
m.l . goldberger and k.m .
watson , _ collision theory _ ( wiley , new york , 1964 ) , p. 771 . p.j.e .
peebles , _ principles of physical cosmology _ ( princeton university press , princeton , nj , 1993 ) , p. 160 . j. lesgourgues and s. pastor , phys .
d * 60 * , 103521 ( 1999 ) . for
a discussion of the rotation of the polarization plane of a linearly polarized light in a gas , see the discussion leading to the eq .
( 8) of v.g .
baryshevsky , time - reversal - violating birefringence of photon in a medium exposed to electric and magnetic field , " hep - ph/0007353 . for a general discussion of rotary power ,
mathieu , _ optics _ ( pergamon press , new york , 1975 ) , p. 235 ; m.v .
klein , _ optics _ ( wiley , new york , 1970 ) , p. 502 .
a. d. dolgov , s. h. hansen , s. pastor , s. t. petcov , g. g. raffelt and d. v. semikoz , nucl .
b * 632 * , 363 ( 2002 ) hep - ph/0201287 .
k. n. abazajian , j. f. beacom and n. f. bell , phys .
d * 66 * , 013008 ( 2002 ) astro - ph/0203442 .
y. y. wong , phys .
d * 66 * , 025015 ( 2002 ) hep - ph/0203180 . | using the one - loop helicity amplitudes for low - energy @xmath0 and @xmath1 scattering in the standard model with massless neutrinos , we study the optical activity of a sea of neutrinos and antineutrinos . in particular , we estimate the values of the index of refraction and rotary power of this medium in the absence of dispersion . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.